The Future of Heritage Science and Technologies: Design, Simulation and Monitoring 9783031175947, 3031175948

This book gathers a selection of contributions dealing with the application of mechanical engineering for preserving and

100 74 110MB

English Pages 457 Year 2022

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Preface and Acknowledgement
Organization Committee
General Chairs
Organizing Committee
Honorary Chairs
Technical and Scientific Committee
Contents
3D Acquisition, Simulation, Modelling and Printing for CH
3D Data Management and Thermographic Studies as a Knowledge Base for the Conservation of a Rationalist Architecture
1 Introduction
1.1 The Case Study: The Tower of the Faculty of Engineering in Bologna
2 Methods
2.1 Geomatic Survey
2.2 HBIM
2.3 Thermography
3 Results
4 Conclusions
References
XpeCAM: The Complete Solution for Artwork Documentation and Analysis
1 Introduction
2 Materials and Methods
2.1 XpeCAM X02
2.2 Lampa
2.3 XpeCAM Platform
2.4 Calibration Methodology
3 Results and Discussion
3.1 Acquisition
3.2 XpeCAM Platform
4 Conclusions
References
Visualising Artworks: Translating the Invisible into Diagnostic Data for Identifying and Quantifying Paint Surfaces
1 Introduction
2 The XpectralTEK System
3 Stichting Restauratie Atelier Limburg (SRAL), Maastricht
4 Technical Photography (TP)
5 Multi-spectral Imaging (MSI)
6 Baseline Projects
7 Case Studies
8 Conclusion
References
An Automatic Method for Geometric and Morphological Information Extraction and Archiving of Ceramic Finds
1 Introduction
2 State of Art
3 The Proposed Methodology
3.1 Geometrical and Morphological Features Recognition
3.2 Digital Cataloguing
3.3 Conclusions
References
Modeling Marble Artworks: The Statue “Oceanus” by Giambologna
1 Introduction
2 The Case-Study: “Oceanus” by Giambologna
3 The Seismic Excitation
4 The Numerical Models
5 Results
6 Final Remarks
References
Application of the RestArt System for Stone Statue Reassembly Validated by Shaking Table Testing
1 Introduction
2 The RestArt Stone Fragments Reassembly Method
3 Experimental Validation
4 Application to the Statue of Diana Cacciatrice
5 Conclusions
References
A BIM-Based Model for Heritage Conservation and Structural Diagnostics: The City Walls of Pisa
1 Introduction
2 Related Works
3 The City Walls of Pisa
3.1 Data Acquisition
3.2 Data Processing and 3D Modelling
3.3 Model Information
3.4 Connection Between the 3D Model and the Information
4 Conclusions
References
An Innovative Method for Dimensioning the Crossbeams of an Original Painted Panel, Based on Mechanical Testing, and on Numerically Modelling Its Distortion Tendency
1 Introduction
2 Materials and Methods
2.1 The Original Panel Painting: The Nativity by Cesare da Sesto
2.2 The OPD Guidelines for the Interventions on Wooden Supports of Panel Paintings
2.3 The Foundations of the Method
2.4 The Restoration Plan, Its Main Decisions and Steps
2.5 The Mechanical Tests
2.6 Estimating the Hygroscopic Parameters
2.7 The Numerical Modelling for Dimensioning the Crossbeams
3 Results and Discussions
3.1 The Results from the Mechanical Testing
3.2 The Hygroscopic Properties
3.3 The Hygro-Mechanical Model
3.4 The Design of the Crossbeams
4 Conclusions
References
ARTE – Augmented Readability Tactile Exploration: The Tactile Bas-Relief of Piazza San Francesco Painting
1 Introduction
2 Modelling Typology
3 2.5 D Reconstruction Procedure
3.1 Preliminary Operations
3.2 Scene Reconstruction
4 Conclusion
References
From Apulian Waste to Original Design Objects: Fused Filament Fabrication (FFF) as a Sustainable Solution
1 Introduction
2 Materials and Method
2.1 Materials
2.2 PLA/Waste Composite Filaments Preparation
2.3 Methods
3 Results and Discussion
4 Conclusion
References
Microclimatic and Thermal Assessment
Microclimatic Experimental Investigation for Assuring Museum Preventive Conservation. Effective Conceptual and Testing Means
1 Introduction
2 Methods and Materials
2.1 The Case Study
2.2 Experimental Measurements and CFD Simulation
3 Results and Discussion
4 Conclusions
References
Indoor Microclimate and Conservation Issues of the Medicean Villa La Petraia. A Preliminary Assessment
1 Introduction
2 Case Study and Research Goals
3 Methodology
4 Monitoring Campaign
4.1 Results
5 Discussion and Conclusions
References
Comparative Examples of the Evolution of Thermal Cameras in Artwork Diagnostics: An Experimental Perspective
1 Introduction
2 Thermal Cameras
3 Artwork Models
4 Experimental Results
4.1 The Wooden Panel Model
4.2 The Fresco Model
4.3 A Real Fresco
5 Conclusions
References
Challenges and Opportunities for the Integration of Photovoltaic Modules in Heritage Buildings Through Dynamic Building Energy Simulations
1 Introduction
2 Aims and Methodology
3 Case Study
3.1 Building Energy Model
4 PV Selection
4.1 Modeling of PV Modules Used in the Simulation
5 Results and Discussion
6 Conclusions
References
Energy Saving for Historical Heritage: The Domotised Lighting System of the Cathedral of Nardò (Lecce)
1 Energy Saving for Historical Heritage
1.1 Introduction
1.2 History
1.3 The Design of the New Lighting System of the Cathedral
2 The Construction Site
3 Conclusions
The Impact of Conservation Conditions Versus Thermal Comfort of Visitors on the Energy Demand of a Museum Refurbished with Geothermal Systems: A Virtual Case Study
1 Introduction
1.1 Background
2 Main Aspects Concerning the Environmental Conditions in a Museum
2.1 Microclimate for the Conservation of Cultural Heritage and Comfort for People in Museums
2.2 Thermo-Hygrometric Environmental Conditions for Distinct Types of Artifacts
3 Materials and Methods
3.1 Procedure Overview
3.2 A Virtual Case Study in Croatia
3.3 Energy Modelling
3.4 Comfort and Microclimate Analysis
4 Conclusions
References
The Medusa Parade Shield by Caravaggio: Making Its Structural Replica, Laboratory Testing, and Numerically Modelling Their Hygro-Mechanical Distortion Behaviour
1 Introduction
2 Materials and Methods
2.1 The Analysis of the Medusa Shield Structure
2.2 Making the Mock-Up
2.3 Monitoring Methods and Apparatus
3 Results and Discussions
3.1 Hygroscopic Behaviour
3.2 Modelling the Hygroscopic Behaviour of the Shield’s Mock-Up
4 Conclusions
References
Innovation in Precision Low-Energy Heat Transfer Using Flexible Heating Mats for Targeted Treatments in Paintings and Paper Conservation
1 Introduction
2 IMAT and Flexible Mat Technology and Applications
3 New Developments and Perspectives of Precision Heat Transfer in Conservation
3.1 Precision Heat Transfer Mat Application for New Approach in Paintings Conservation
3.2 Precision Heat Transfer Mat Application for a New Approach Detaching Historical Photographic Prints from Secondary Support
3.3 Precision Heat Transfer Mat in Conservation of Paper and Books
4 Discussion and Conclusions
References
Monitoring of Cultural Heritage Environments
The AMOR Project: When Technology Meets Cultural Heritage
1 Introduction
2 Safeguard: Needs, Technologies and Methodologies
2.1 Remote Observation of Surface Degradation: Satellites and Drones
2.2 The GPR System
3 Fruition
3.1 New Fruition Solutions
3.2 Circular Fruition
3.3 The Importance of 5G
4 Conclusions
References
Testing Portable NMR to Monitor the Effect of Paper Exposure to UV-Light
1 Introduction
2 Materials and Methods
2.1 Portable NMR Measurements
2.2 Raman Measurements
3 Results and Discussions
3.1 Portable NMR Results
3.2 Raman Results
4 Conclusion
References
Non-invasive Analysis of the Pigment Palette Used by the Renaissance Painter Sofonisba Anguissola
1 Introduction
2 Experimental
2.1 X-radiography
2.2 Macro-XRF
2.3 Micro-XRF
2.4 Raman Spectroscopy
2.5 Optical Coherence Tomography
3 Results and Discussion
3.1 Black Pigments
3.2 White Pigments
3.3 Dark Green Pigments - Background
3.4 Brown, Yellow and Green Pigments
3.5 Blue Pigments
3.6 Red Pigments
3.7 Retouches
3.8 Comparison of the Paintings Self-portrait with Easel and the Game of Chess
4 Conclusions
References
Monitoring and Evaluation of Pietra Serena Decay Adopting NDT Techniques: Application on Building Stones in Situ
1 Introduction
2 Pietra Serena Sandstone
2.1 Stone Description
2.2 Study Sites Description
3 Analytical Methods
4 Results and Discussion
5 Conclusion
References
Palazzo Medici Riccardi: Diagnostic Experimental Site for the Pietraforte Façades
1 Introduction
2 Study Area
3 The Pietraforte Façades
4 Diagnostic Project
4.1 Scientific Survey
4.2 Experimentation
4.3 Design
4.4 Consolidation and Restoration
5 Monitoring Plan
6 Conclusion
References
Analytical Pyrolysis: A Useful Tool to Analyze and Evaluate Consolidated Archaeological Wood
1 Introduction
1.1 Conservation Treatments for Waterlogged Archaeological Wooden Objects: An Overview
2 Materials and Methods
2.1 Materials
2.2 Apparatus
3 Results and Discussion
4 Ongoing Research and Perspectives
References
A Diagnostic Method for the Pavement Conservation of the Great Synagogue of Florence (Italy)
1 Introduction
2 The Synagogue
2.1 The Synagogue Pavement
3 Sample Collection and Analytical Methods
3.1 Mineralogy, Petrography and Chemical Analyses
3.2 Geophysical and Geognostic Surveys
4 Results
5 Discussion and Concluding Remarks
References
Conservation Strategies for the Palazzo degli Affari in Florence (Italy): The Role of Protective Treatments on the Concrete Carbonation Resistance
1 Introduction
2 Case Study
3 Research Aim
4 Carbonation Prediction Model: fib Model Code for Service Life Design
5 Experimental Survey
5.1 Materials and Methods
6 Results and Discussion
7 Conclusion
References
Monitoring and Understanding VOC Induced Glass Corrosion Using Multi-modal Imaging Techniques
1 Introduction
1.1 Aims and Objectives
2 Materials
3 Experimental Methods
3.1 Digital Photography
3.2 Reflectance Transformation Imaging
3.3 Optical Microscopy
3.4 Hyperspectral Imaging
4 Results
4.1 Digital Photography
4.2 Reflectance Transformation Imaging
4.3 Optical Microscopy
4.4 Hyperspectral Imaging
5 Discussion
6 Conclusion
References
The Terrace of Saturn in Palazzo Vecchio, Florence (Italy): Material Characterisation and Monitoring for Preventive Conservation
1 Introduction
2 Material and Methods
2.1 Sampling
2.2 In Situ Investigations
2.3 Laboratory Analysis
3 Results and Discussion
4 Conclusions
Appendices
References
Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage Artworks
1 Introduction
2 Microwave Tomography System
2.1 Simulations
2.2 Laboratory Experimental Measurements
2.3 Portable System Setup
2.4 Imaging
3 2D Imaging of Sealed Pottery
4 Conclusion
References
SWIR Reflectance Imaging Spectroscopy and Raman Spectroscopy Applied to the Investigation of Amber Heritage Objects: Case Study on the Amber Altar of the Lord's Passion
1 Introduction
2 Experimental
2.1 Materials
2.2 Techniques
3 Results
3.1 Amber Identification
3.2 Amber Degradation
3.3 Amber Classification
4 Summary
References
Methods for Enhancing CH Fruition
Folk Music of the Khơ Mú in Điên Biên Province: Characteristics and Potential for Community-Based Tourism Development
1 Introduction
2 Research Methods
3 Characteristics of Folk Music and Potential for Community Tourism Development
3.1 Folk Songs
3.2 Musical Instruments
3.3 Potential for Community-Based Tourism Development
4 Conclusion
References
Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province): Unique Cultural Characteristics and Festival Protection Solution
1 Introduction
2 Research Methods
3 Overview of the Tien Cong Festival
3.1 Origin of the Festival
3.2 Festival Space and Time
3.3 Overview of the Order of the Tien Cong Festival
4 The Unique Culture of the Tien Cong Festival
4.1 Offerings
4.2 Rituals
4.3 Performing Arts of Folk Singing Dum - Different Cultural Nuances of Coastal Residents of Quang Ninh
5 Meaning and Value of the Tien Cong Festival
6 Preserving the Tien Cong Festival in the Current Context
7 Conclusion
References
Author Index
Recommend Papers

The Future of Heritage Science and Technologies: Design, Simulation and Monitoring
 9783031175947, 3031175948

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Lecture Notes in Mechanical Engineering

Rocco Furferi · Lapo Governi · Yary Volpe · Francesco Gherardini · Kate Seymour   Editors

The Future of Heritage Science and Technologies Design, Simulation and Monitoring

Lecture Notes in Mechanical Engineering Series Editors Fakher Chaari, National School of Engineers, University of Sfax, Sfax, Tunisia Francesco Gherardini , Dipartimento di Ingegneria “Enzo Ferrari”, Università di Modena e Reggio Emilia, Modena, Italy Vitalii Ivanov, Department of Manufacturing Engineering, Machines and Tools, Sumy State University, Sumy, Ukraine Editorial Board Members Francisco Cavas-Martínez , Departamento de Estructuras, Construcción y Expresión Gráfica Universidad Politécnica de Cartagena, Cartagena, Murcia, Spain Francesca di Mare, Institute of Energy Technology, Ruhr-Universität Bochum, Bochum, Nordrhein-Westfalen, Germany Mohamed Haddar, National School of Engineers of Sfax (ENIS), Sfax, Tunisia Young W. Kwon, Department of Manufacturing Engineering and Aerospace Engineering, Graduate School of Engineering and Applied Science, Monterey, CA, USA Justyna Trojanowska, Poznan University of Technology, Poznan, Poland

Lecture Notes in Mechanical Engineering (LNME) publishes the latest developments in Mechanical Engineering—quickly, informally and with high quality. Original research reported in proceedings and post-proceedings represents the core of LNME. Volumes published in LNME embrace all aspects, subfields and new challenges of mechanical engineering. Topics in the series include: • • • • • • • • • • • • • • • • •

Engineering Design Machinery and Machine Elements Mechanical Structures and Stress Analysis Automotive Engineering Engine Technology Aerospace Technology and Astronautics Nanotechnology and Microengineering Control, Robotics, Mechatronics MEMS Theoretical and Applied Mechanics Dynamical Systems, Control Fluid Mechanics Engineering Thermodynamics, Heat and Mass Transfer Manufacturing Precision Engineering, Instrumentation, Measurement Materials Engineering Tribology and Surface Technology

To submit a proposal or request further information, please contact the Springer Editor of your location: China: Ms. Ella Zhang at [email protected] India: Priya Vyas at [email protected] Rest of Asia, Australia, New Zealand: Swati Meherishi at [email protected] All other countries: Dr. Leontina Di Cecco at [email protected] To submit a proposal for a monograph, please check our Springer Tracts in Mechanical Engineering at https://link.springer.com/bookseries/11693 or contact [email protected] Indexed by SCOPUS. All books published in the series are submitted for consideration in Web of Science.

More information about this series at https://link.springer.com/bookseries/11236

Rocco Furferi Lapo Governi Yary Volpe Francesco Gherardini Kate Seymour •







Editors

The Future of Heritage Science and Technologies Design, Simulation and Monitoring

123

Editors Rocco Furferi Department of Industrial Engineering University of Florence Florence, Italy

Lapo Governi Department of Industrial Engineering University of Florence Florence, Italy

Yary Volpe Department of Industrial Engineering University of Florence Florence, Italy

Francesco Gherardini Department of Engineering “Enzo Ferrari” University of Modena and Reggio Emilia Modena, Italy

Kate Seymour Stichting Restauratie Atelier Limburg Maastricht, The Netherlands

ISSN 2195-4356 ISSN 2195-4364 (electronic) Lecture Notes in Mechanical Engineering ISBN 978-3-031-17593-0 ISBN 978-3-031-17594-7 (eBook) https://doi.org/10.1007/978-3-031-17594-7 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface and Acknowledgement

This book presents a selection of contributions, dealing with the application of Mechanical Engineering in the field of Cultural Heritage, presented at the 3rd Florence Heri-Tech International Conference, held during May 16–18, 2022, in Firenze, Italy. Organized under the patronage of the University of Florence with the support of the Department of Industrial Engineering, Florence Heri-Tech gathered researchers and experts in the field of “heritage science and related technologies” to disseminate their recent research at an international level as well as to draw new inspiration. The Conference was part of the 2022 Florence International Biennial for Art and Restoration, an international event attracting prestigious institutions and companies and creating a unique opportunity to bring together the academic world with industry. The overarching goal of Florence Heri-Tech is to promote European mobility and cooperation among researchers, students, and practitioners, to further Europe’s growth as a multi-cultural society, and to promote the idea that scientific-cultural research must be an important part of society. Furthermore, the Conference intends to foster international networks between universities, training institutions, and businesses in order to foster long-term partnership prospects. More than 80 experts, coordinated by the General Chairs and supported by a high-level Technical and Scientific Committee, were involved in the review process, which led to the selection of 32 papers (out of 101 paper submitted at the Conference and 146 abstract received by the Scientific Committee). The overall number of authors involved was 154. Contributions focus on multi-disciplinary and inter-disciplinary research concerning the use of innovative engineering-based methodologies and technologies for documenting, managing, preserving, and communicating cultural heritage. The editors would like to personally thank everyone participating in the review process for their dedication and skill at the service of the Conference. Special appreciation goes to the members of the Technical and Scientific Committees who made it possible for the Conference to take place.

v

vi

Preface and Acknowledgement

The Conference prestige was enhanced by the participation of the Honorary Chairs: Prof. Dei, Prof. Baglioni, Prof. Bonsanti, Prof. Cappellini, Prof. Cather, Prof. Ciatti, Prof. Colombini, Prof. Rodrigues, Prof. Santos, and Prof. Sgamellotti. General Chairs extend their thanks to all authors, speakers, and those persons whose labor, financial support, and encouragement made the Florence Heri-Tech event possible. Rocco Furferi Lapo Governi Yary Volpe Kate Seymour Francesco Gherardini

Organization Committee

General Chairs Rocco Furferi Lapo Governi Yary Volpe Rodorico Giorgi Anna Pelagotti Kate Seymour

University of Florence, Italy University of Florence, Italy University of Florence, Italy University of Florence, Italy Executive Agency of the European Research Council Stichting Restauratie Atelier Limburg (SRAL), Maastricht, The Netherlands

Organizing Committee Elena Amodei Lucia Maranzana Francesco Gherardini Rocco Furferi

Salone dell’Arte e del Restauro di Firenze, Italy Salone dell’Arte e del Restauro di Firenze, Italy University of Modena and Reggio Emilia, Italy University of Florence, Italy

Honorary Chairs Baglioni Piero Bonsanti Giorgio Cappellini Vito Cather Sharon Ciatti Marco Colombini Maria Perla

Dei Luigi

University of Florence, Italy University of Florence, Italy University of Florence, Italy The Courtauld Institute of Art, London, UK Opificio delle Pietre Dure, Florence, Italy CNR-ICVBC. National Research Council of Italy – Institute for the Conservation and Valorization of Cultural Heritage Rector of the University of Florence, Italy

vii

viii

Delgado Rodrigues Jose Matteini Mauro

Primicerio Mario Santos Pedro Sgamellotti Antonio

Organization Committee

Laboratorio Nacional De Engenharia Civil, Lisbon, Portugal Former Director of ICVBC-Istituto per la Conservazione e la Valorizzazione dei Beni Culturali – CNR, Italy University of Florence, Italy Fraunhofer Institute for Computer Graphics Research IGD, Darmstadt, Germany University of Perugia, Italy

Technical and Scientific Committee Al Huneidi Hani M. Argenti Fabrizio Balocco Carla Berardi Umberto Bianconi Francesco Bici Michele

Buonamici Francesco Camilli Andrea Cassar JoAnn Cavalieri Duccio Corallo Angelo Cheung Sidney Del Bianco Corinna

Del Bimbo Alberto Di Angelo Luca Bruno Fabio

International Cultural Heritage Specialist, Jordan Department of Information Engineering, University of Florence, Italy Industrial Engineering Department, University of Florence, Italy Canada Research Chair, Ryerson University Department of Engineering, University of Perugia, Italy Department of Mechanical and Aerospace Engineering of Sapienza University of Rome, Italy Department of Industrial Engineering, University of Florence, Italy Soprintendenza per i Beni Archeologici della Toscana, Department Member, Italy University of Malta Department of Biology, University of Florence, Italy Department of Innovation Engineering, Unisalento, Italy The Chinese University of Hong Kong, Hong Kong Director of the International Institute Life Beyond Tourism and Adviser of the Romualdo Del Bianco Foundation, Italy Department of Information Engineering, University of Florence, Italy Department of Industrial Engineering University of L’Aquila, Italy Department of Mechanical Energetics and Management Engineering, University of Calabria, Italy

Organization Committee

Es Sebar Leila Fernandez Federica

Ferrise Francesco Fioravanti Marco

Frischer Bernard Garmendia Arrieta Leire Gherardini Francesco

Goli Giacomo

Guidi Gabriele Hazan Susan Iadanza Ernesto Lanzoni Luca

Lazoi Mariangela Macdonald Susan Markevicius Tomas Martín Lerones Pedro Olsson Nina Patelli Alessandro Pavia Anselmo

Pellicciari Marcello Peñalver Martínez María Jesús

ix

PhD in Metrology – Politecnico di Torino, Italy IEMEST, Istituto EuroMediterraneo di Scienza e Tecnologia, Director of Department NIMA, Italy Department of Mechanics, Polytechnic of Milan, Italy Agrifood Production and Environmental Sciences Agriculture, Food and Forestry Systems, University of Florence, Italy Department of Informatics, Indiana University, US Faculty of Engineering in Bilbao University of the Basque Country (UPV/EHU), Spain Department of Engineering “Enzo Ferrari”, University of Modena and Reggio Emilia, Italy Agrifood Production and Environmental Sciences Agriculture, Food and Forestry Systems, University of Florence, Italy Department of Mechanics, Polytechnic of Milan, Italy Coordinator, Israel Consortium for Digital Heritage University of Florence, Italy Architect and Urban Planner, International Expert/Lead Expert in the Field of Cultural Heritage Enhancements—UPCT Polytechnical University of Cartagena, Spain Department of Innovation Engineering, Unisalento, Italy Getty Conservation Institute, Los Angeles University of Amsterdam, The Netherlands Heritage Area Manager, CARTIF Technology Center, Spain Art Conservation, Portland, Oregon, US Department of Physics and Astronomy “Galileo Galilei”, University of Padua, Italy Applied Computer Group – Computer Science Department – Federal University of Maranhão, Brazil University of Modena and Reggio Emilia, Italy Department of Architecture and Building Technology - Universidad Politécnica de Cartagena, Spain

x

Razionale Vincenzo Armando Ridolfi Alessandro Riminesi Cristiano

Seymour Kate Santo Alba Patrizia Tiano Piero Tucci Grazia

Uccheddu Francesca Vaccaro Vincenzo

Organization Committee

Department of Civil and Industrial Engineering, University of Pisa, Italy Department of Industrial Engineering at the University of Florence, Italy Scienze del patrimonio culturale. Studio e sviluppo di tecniche e metodologie per la conservazione dei beni culturali – CNR Italian National Research Council, Italy Stichting Restauratie Atelier Limburg, Maastricht, The Netherlands Earth Sciences Department of University of Florence, Italy Institute for the Conservation and Valorization of Cultural Heritage (ICVBC), Italy Department of Civil, Constructional and Environmental Engineering, University of Florence, Italy Department of Industrial Engineering, University of Padua, Italy Opera del Duomo Florence, Italy

Contents

3D Acquisition, Simulation, Modelling and Printing for CH 3D Data Management and Thermographic Studies as a Knowledge Base for the Conservation of a Rationalist Architecture . . . . . . . . . . . . . Ester Barbieri, Elisa Franzoni, Alessandro Lambertini, Cesare Pizzigatti, Francesca Trevisiol, and Gabriele Bitelli XpeCAM: The Complete Solution for Artwork Documentation and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vassilis M. Papadakis, Marlene Machado, and Jani dos Santos Visualising Artworks: Translating the Invisible into Diagnostic Data for Identifying and Quantifying Paint Surfaces . . . . . . . . . . . . . . . . . . . Kate Seymour, Jani Santos, Vassilis M. Papadakis, Betlem Alaponte, Alzbeta Prochazkova, Valentine Gatto, Alice Limb, Nalini Biluka, and Luuk Hoogstede An Automatic Method for Geometric and Morphological Information Extraction and Archiving of Ceramic Finds . . . . . . . . . . . . . . . . . . . . . . Luca Di Angelo, Paolo Di Stefano, Emanuele Guardiani, and Anna Eva Morabito Modeling Marble Artworks: The Statue “Oceanus” by Giambologna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Francesco Trovatelli, Francesca Barbagallo, Edoardo M. Marino, Marco Tanganelli, and Stefania Viti Application of the RestArt System for Stone Statue Reassembly Validated by Shaking Table Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martina Pavan, Giulia Pompa, Pietro Nardelli, Silvia Borghini, Vincenzo Fioriti, Angelo Tatì, Alessandro Colucci, Massimiliano Baldini, Alessandro Picca, and Ivan Roselli

3

16

28

48

60

70

xi

xii

Contents

A BIM-Based Model for Heritage Conservation and Structural Diagnostics: The City Walls of Pisa . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anna De Falco, Francesca Gaglio, Francesca Giuliani, and Massimiliano Martino An Innovative Method for Dimensioning the Crossbeams of an Original Painted Panel, Based on Mechanical Testing, and on Numerically Modelling Its Distortion Tendency . . . . . . . . . . . . . . . . . . . Lorenzo Riparbelli, Ciro Castelli, Giovanni Gualdani, Luciano Ricciardi, Andrea Santacesaria, Luca Uzielli, and Paola Mazzanti

84

97

ARTE – Augmented Readability Tactile Exploration: The Tactile Bas-Relief of Piazza San Francesco Painting . . . . . . . . . . . . . . . . . . . . . 113 Luca Puggelli, Rocco Furferi, Lapo Governi, Chiara Santarelli, and Yary Volpe From Apulian Waste to Original Design Objects: Fused Filament Fabrication (FFF) as a Sustainable Solution . . . . . . . . . . . . . . . . . . . . . . 127 Daniela Rizzo, Francesco Montagna, Elisabetta Palumbo, Daniela Fico, Valentina De Carolis, Raffaele Casciaro, and Carola Esposito Corcione Microclimatic and Thermal Assessment Microclimatic Experimental Investigation for Assuring Museum Preventive Conservation. Effective Conceptual and Testing Means . . . . 143 Carla Balocco and Margherita Vicario Indoor Microclimate and Conservation Issues of the Medicean Villa La Petraia. A Preliminary Assessment . . . . . . . . . . . . . . . . . . . . . . 155 Anna Bonora, Vincenzo Costanzo, Kristian Fabbri, Marco Pretelli, and Eva Schito Comparative Examples of the Evolution of Thermal Cameras in Artwork Diagnostics: An Experimental Perspective . . . . . . . . . . . . . . . . 169 Dario Ambrosini, Tullio de Rubeis, Giovanni Pasqualoni, and Domenica Paoletti Challenges and Opportunities for the Integration of Photovoltaic Modules in Heritage Buildings Through Dynamic Building Energy Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Elena Lucchi and Eva Schito Energy Saving for Historical Heritage: The Domotised Lighting System of the Cathedral of Nardò (Lecce) . . . . . . . . . . . . . . . . . . . . . . . 195 Cristina Caiulo and Stefano Pallara

Contents

xiii

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors on the Energy Demand of a Museum Refurbished with Geothermal Systems: A Virtual Case Study . . . . . . . . . . . . . . . . . . . . . . 205 Gianluca Cadelano, Shabnam Javanshir, Laura Carnieletto, Francesca Bampa, Alessandro Bortolin, Michele De Carli, Eloisa di Sipio, and Adriana Bernardi The Medusa Parade Shield by Caravaggio: Making Its Structural Replica, Laboratory Testing, and Numerically Modelling Their Hygro-Mechanical Distortion Behaviour . . . . . . . . . . . . . . . . . . . . . . . . 219 Paola Mazzanti, Paolo Dionisi-Vici, Marco Fioravanti, Elisa Cardinali, Justine Mialhe, Marco Togni, Luca Uzielli, and Lorenzo Riparbelli Innovation in Precision Low-Energy Heat Transfer Using Flexible Heating Mats for Targeted Treatments in Paintings and Paper Conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Tomas Markevicius, Nina Olsson, Yuhui Liu, Maddalena Magnani, and Martina Paganin Monitoring of Cultural Heritage Environments The AMOR Project: When Technology Meets Cultural Heritage . . . . . 257 Nicole Dore, Francesco Cochetti, Ilaria Catapano, Giovanni Ludeno, Gianluca Gennarelli, Maria Elena Corrado, Carlo Cacace, Paolo Osso, Michele Luglio, and Francesco Zampognaro Testing Portable NMR to Monitor the Effect of Paper Exposure to UV-Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 Valeria Stagno, Alessandro Ciccola, Elisa Villani, Roberta Curini, Paolo Postorino, and Silvia Capuani Non-invasive Analysis of the Pigment Palette Used by the Renaissance Painter Sofonisba Anguissola . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 Anna Rygula, Marta Matosz, Alicja Mogielska, Magdalena Iwanicka, Piotr Targowski, Michał Obarzanowski, and Julio M. del Hoyo-Meléndez Monitoring and Evaluation of Pietra Serena Decay Adopting NDT Techniques: Application on Building Stones in Situ . . . . . . . . . . . . . . . . 292 Sara Calandra, Irene Centauro, Teresa Salvatici, Elena Pecchioni, and Carlo Alberto Garzonio Palazzo Medici Riccardi: Diagnostic Experimental Site for the Pietraforte Façades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Lorenzo Di Bilio, Maria Grazia Fraiese, Marco Vannuccini, Matteo Galatro, Luciana Pinzani, Carlo Alberto Garzonio, Teresa Salvatici, Irene Centauro, Sara Calandra, Francesco Pilati, Hosea Scelza, Rosella Pascucci, Federica Valentini, and Pasquino Pallecchi

xiv

Contents

Analytical Pyrolysis: A Useful Tool to Analyze and Evaluate Consolidated Archaeological Wood . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 Jeannette J. Lucejko, Irene Bargagli, Marco Mattonai, Erika Ribechini, Maria Perla Colombini, Gilles Chaumat, Susan Braovac, Magdalena Zborowska, and Francesca Modugno A Diagnostic Method for the Pavement Conservation of the Great Synagogue of Florence (Italy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 Teresa Salvatici, Carlo Alberto Garzonio, Sara Calandra, Elena Pecchioni, and Alba Patrizia Santo Conservation Strategies for the Palazzo degli Affari in Florence (Italy): The Role of Protective Treatments on the Concrete Carbonation Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 Marta Castellini, Oana Adriana Cuzman, Silvia Rescic, Marco Tanganelli, Stefano Landi, and Cristiano Riminesi Monitoring and Understanding VOC Induced Glass Corrosion Using Multi-modal Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 Deepshikha Sharma, Ulrike Rothenhaeusler, Katharina Schmidt-Ott, Marvin Nurit, Yuly Castro Cartagena, Gaetan Le-Goic, Edith Joseph, Sony George, and Tiziana Lombardo The Terrace of Saturn in Palazzo Vecchio, Florence (Italy): Material Characterisation and Monitoring for Preventive Conservation . . . . . . . 376 Sveva Longo, Marta Castellini, Federico De Luca, Claudia Conti, Alessandra Botteon, Barbara Salvadori, Giorgio Franco Pocobelli, Donata Magrini, Cristiano Riminesi, Rachele Manganelli Del Fa, Giorgio Caselli, and Emma Cantisani Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage Artworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Emanuela Proietti, Giovanni Capoccia, Romolo Marcelli, Giovanni Maria Sardi, and Barbara Caponera SWIR Reflectance Imaging Spectroscopy and Raman Spectroscopy Applied to the Investigation of Amber Heritage Objects: Case Study on the Amber Altar of the Lord’s Passion . . . . . . . . . . . . . . . . . . . . . . . 401 Paulina Krupska-Wolas, Anna Ryguła, Elżbieta Kuraś, and Julio del Hoyo-Meléndez Methods for Enhancing CH Fruition Folk Music of the Khơ Mú in Điên Biên Province: Characteristics and Potential for Community-Based Tourism Development . . . . . . . . . . . . . 419 Lam Nguyen Dinh, Son Quang Van, Son Nguyen Truong, and Dinh Luong Khac

Contents

xv

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province): Unique Cultural Characteristics and Festival Protection Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433 Hue Phan Thi, Ninh Ngo Hai, Son Quang Van, and Dinh Luong Khac Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445

3D Acquisition, Simulation, Modelling and Printing for CH

3D Data Management and Thermographic Studies as a Knowledge Base for the Conservation of a Rationalist Architecture Ester Barbieri(B)

, Elisa Franzoni , Alessandro Lambertini , Cesare Pizzigatti , Francesca Trevisiol , and Gabriele Bitelli

Department of Civil, Chemical, Environmental and Materials Engineering, University of Bologna, Bologna, Italy {ester.barbieri2,elisa.franzoni,alessandro.lambertini, cesare.pizzigatti2,francesca.trevisiol2}@unibo.it, [email protected]

Abstract. Thanks to the development of technologies for digitalization, new possibilities of management and valorization of Cultural Heritage have been opened. These issues are addressed by the project IDEHA (Innovation for Data Elaboration in Heritage Areas), funded by MIUR Italian Ministry, for Italian assets of historical interest. Among the different case studies, the library tower of the XX Century complex of the Faculty of Engineering in Bologna has a great significance as an example of rationalist architecture. The present work aims at the 3D digitization of the tower, the analysis of its state of maintenance by using active thermography and the implementation of a BIM model that contains all these information. Aerial (by UAV) and terrestrial photogrammetry was used for the development of the three-dimensional model. From the acquired images, a point cloud and a textured mesh of the Tower were processed. The point cloud was also used as a reference for the elaboration of an HBIM model representing both the interior and exterior of the tower structure. Moreover, a study of the conditions of the external facades recently treated with water repellent was conducted through active thermography. The study of the state of the external brick wall face was carried out through the multi-temporal analysis of thermal images. This allowed to identify the areas where the water repellent treatment is damaged and therefore more exposed to deterioration phenomena. The main challenge of this research work is to incorporate the information acquired through active thermography into the information model in order to better manage the building. Keywords: 3D surveying · 3D modeling · HBIM · Thermography

1 Introduction Nowadays there is a growing sensitivity towards the preservation and digitization of historical heritage. Digitization of historical architecture is an important tool both for creating digital models that can be explored remotely and for creating information models © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 3–15, 2023. https://doi.org/10.1007/978-3-031-17594-7_1

4

E. Barbieri et al.

that allow the management and the monitoring of the property in every aspect. Particularly buildings of historical interest are characterized by a variety of information that must be documented for their better protection and management [1, 2]. It is therefore essential to consider historic buildings as complex systems that need to be described in all their aspects. The integration of three-dimensional surveys and the use of non-destructive techniques are the keys to monitor existing structures and facilitate the diagnostic process leading to maintenance interventions on materials or elements [3–6]. Among the most used non-destructive techniques in monitoring and diagnostics there is thermography [7, 8]. In recent years, the use of thermography has increased due to improved resolution of thermal images. The acquisition of thermal images, which are not always easy to interpret, makes it possible to investigate the behavior of materials by analyzing the surface temperature. However, it is not always simple to integrate the information gained from a thermographic survey with geometric data. Merging these two types of data is problematic both because one is metric and the other is non-metric, and also because of differences in resolution and radiometric characteristics [9, 10]. Another challenge is to integrate these data into an information model. Thanks to the development of the HBIM (Historic Building Information Modeling) methodology [11, 12], the representation of existing structures is now accompanied by a wealth of information. The creation of an as-built model makes it possible not only to represent a building geometrically and graphically, but also to obtain advantages for its management in terms of cost estimation, maintenance and document archiving. The case study chosen for this work is the library tower of the Faculty of Engineering of the University of Bologna. In order to create an HBIM that describes the geometry and integrates other data on materials’ deterioration, two surveys were conducted. With the first survey, carried out by aerial photogrammetry by drone and complemented by terrestrial photogrammetry, the geometry of the structure was detected. In the second survey, high-resolution thermal images were acquired using an active approach. The geometric survey is fundamental both to validate the information extracted from plans, elevations and sections used to elaborate the HBIM model, and to georeference the data acquired with active thermography in order to transfer the information to the HBIM model. In particular, the active thermography survey made it possible to carry out a multi-temporal analysis and study the state of the hydrophobic treatment applied to the facades, detecting the water absorption patterns in external brick walls. 1.1 The Case Study: The Tower of the Faculty of Engineering in Bologna The case study selected for this research work is the tower of the historical building of the Faculty of Engineering in Bologna (Fig. 1). The Faculty of Engineering was built in the 1930s following the designs of Giuseppe Vaccaro [13], one of the most esteemed architects in Italy at the time. The library and clock tower of the Faculty of Engineering in Bologna is a perfect example of Italian rationalist architecture, which developed between the 1920s and 1930s: inaugurated in 1935, the structure embodies the style of the period. It consists of a main body with a comb structure [14] and a tower structure (about 45 m high) that today houses the library collection and a geodesic observatory at the top.

3D Data Management and Thermographic Studies

5

The tower structure consists of reinforced concrete frames that form the 9 structural floors. The exterior cladding is made of bricks, unlike the plaster covering the central body. Only the east façade exhibits different materials, as it includes a part made of plastered concrete with alternate glass blocks, naturally illuminating the inside of the library tower. By using a different material for the Tower’s cladding with respect to the main body of the building, Vaccaro wanted to emphasize the vertical structure as a continuation of the traditional use of bricks in Bologna. The building played a key role during the Italian fascist period and the Second World War: it was occupied in 1944 by the black shirts, and until the Allied Liberation in 1945, it was a place of torture against partisans [15]. In 1945, part of the west façade of the Tower was modified: the fascist emblems in the lower part of the north façade of the Tower were removed and replaced with a new brick decoration. The faculty only began regular teaching activities in the academic year 1947–48. Presently the entire building is safeguarded for its high artistic value by putting restrictions to its modifications, according to Italian laws.

Fig. 1. On the left: The faculty of engineering just after its construction (R. Istituto Superiore d’Ingegneria - 1935 - CAF, Bologna Fund). On the right: The library tower today; at the top the geodetic observatory is present.

In general, the absence of roof overhang characterizing most of rationalist buildings left façade materials directly exposed to rain, leading to deterioration mechanisms such as freeze-thaw cycles, biological decay and others. To cope with this problem, highly durable ready-mix renders diffused in the 1920s and 1930s, and in fact they were used in other parts of the Faculty [16]. In the towers of rationalist buildings, the problem of rain absorption and runoff was further exacerbated by the height and in some case external cladding with vitrified ceramic materials (such as Italklinker in Italy [17]) was used for protective aim. In the tower under investigation, unplastered brick masonry surfaces are

6

E. Barbieri et al.

present. Moreover, the mortar joints of the brick walls of the tower are positioned backward from the masonry surface, probably to produce a more expressive texturized effect than plain traditional masonry, and this favors the persisting and capillary absorption of water during and after the rainfall events. As a consequence, the brick walls of the tower, after about 90 years of direct exposure to rain, presently suffer from different deterioration patterns, including scaling, flaking and detachment. After some detachment and falling of layers of bricks from the façades, a water repellent treatment was applied by spraying in May 2019, using a basket crane truck. The product was a mix of polysiloxane oligomers and silanes in organic solvent. Indeed, the use of protective treatments based on the use of synthetic water repellents (e.g., silicone resins and fluorinated polymers) [18] is presently the most common repair strategy for historic masonry subject to direct rain exposure.

2 Methods 2.1 Geomatic Survey The survey of the Engineering Faculty Tower was carried out by acquiring aerial images through an Unmanned Aerial Vehicle (UAV) platform and then processing these images with photogrammetric techniques. Supplementary terrestrial images were also used. The geometry of the tower, which is up to 45 m above ground level, does not allow for a complete survey without using an aerial platform. The UAV chosen for the survey is a commercial drone manufactured by DJI and subsequently lightened under 300 g to meet the requirements of local regulations in effect at the time of the survey. It is equipped with a DJI FC 1102 optical sensor with 12-megapixel resolution and 82° Field Of View (FOV). Being a lightweight drone, it does not have a particularly high flight autonomy. The actual survey time for each fully charged battery is in fact less than 15 min. It was therefore necessary to carefully plan the survey activities considering the existing limitations. Two photogrammetric capture schemes were used: nadiral and convergent. Several flyovers were carried out up to total coverage of all the facades of the tower and its top. A total of 854 images were acquired. From the entire set of images acquired, 389 were selected in order to eliminate the images that presented problems of illumination, out of focus or excessive overlap. Image processing was performed using a structure-from-motion technique. During processing, markers were identified in natural points: using a laser scanner survey (VZ400) previously carried out, it was possible to identify natural markers to have a metrically correct model. The point cloud was then processed, which after the denoising and cleaning phase consists of approximately 35 million points. The textured mesh was then created in the Agisoft Metashape Pro software (Fig. 2).

3D Data Management and Thermographic Studies

7

Fig. 2. From left to right: point cloud, mesh and textured mesh of the Engineering Faculty Tower in Bologna.

2.2 HBIM Once the geometric data were acquired from the photogrammetric survey, the information model elaboration phase began. In order to model the as-built geometry, the documents provided by the historical archive (plans, elevations and sections) were used, as well as the point cloud of the Tower acquired by UAV. The cloud was then coloured with the artificial texture PCV (Portion de ciel visible) to highlight its geometry. The use of plans and sections made it possible to attribute the exact thicknesses of the external and internal masonry, and to model the internal partitions and vertical connections. Different representations from the HBIM model are provided in Fig. 3. Through the study of archive documents and publications by G. Vaccaro, it was also possible to acquire further information on the materials: the marble used in the entrance, the type of glass blocks present in the east wall, the stratigraphy of the brick walls, the type of plaster used for the central body. Finally, the information model was accompanied by all the information concerning the system of storage of journals and books inside the library tower and their arrangement floor by floor. 2.3 Thermography Thermography is a non-contact, non-destructive technique that can be used to detect moisture or defects in materials. In this case study, thermography was used as a method to identify areas where the hydrophobic treatment is not present, damaged or no longer effective. Water repellents are expected to prevent the rain absorption by the bricks by increasing the water contact angle between water and solid surface (>90°) while allowing a certain evaporation of the moisture possibly present behind the treated layer (high water vapor diffusion). Despite the widespread use of water repellents in conservation and repair of existing buildings, the data about their in-the-field effectiveness, compatibility

8

E. Barbieri et al.

Fig. 3. (a) Overlay of the point cloud (PCV texturization) on the HBIM model; (b), (c) Views of the HBIM model; (d) Axonometric cross-section view.

and durability are quite few, especially for brick masonry directly exposed to rainfall [18]. The effectiveness of water repellents applied on façades can be investigated by simple and cheap non-destructive procedures, such as the Karsten tube [19] or the contact sponge method, but they are punctual and require the surface to be directly accessible to the operator. For this reason, the use of IR thermography was proposed by some authors for the investigation of treated surfaces [20, 21]. In the present case study, where environmental aggression is so challenging for masonry surface, the investigation of the hydrophobic behavior sometime after the application was regarded as necessary. In particular, it was noticed that some detachments continued to occur even after the application of the protective treatment, highlighting that the fragments collapsed exhibited a variable degree of water repellency (Fig. 4). The hydrophobic/hydrophilic behavior was qualitatively investigated by releasing some water drops over the external surface of the fragments and observing the shape of the drops and the velocity of capillary absorption. To better investigate the presence and distribution of the treatment on the masonry surface, IR active thermography was proposed, as a valuable tool for an overall and non-destructive evaluation of the masonry state.

3D Data Management and Thermographic Studies

9

Fig. 4. Different water repellency exhibited by some brick fragments detached from the tower about 1.5 years after the treatment’s application.

The thermal survey was carried out with the thermal camera FLIR P620, which has a spectral range between 7.5 µm and 13 µm and 640 × 800 resolution and is equipped with the standard 24° lens. The thermal survey was performed in the winter season on a sunny day. Information on temperature and relative humidity was acquired in order to correctly set the parameters of the thermal camera. At the beginning of the survey, the ambient conditions were 14.4 °C and 59.6% RH. An active thermal survey was carried out to study the different water absorption of the bricks. The south tower surface at the level of the roof of the main building was sprayed with water. 20 images were acquired over 75 min to study the absorption/drying phase. The images were taken every 3.7 min, and the camera, mounted on a tripod, was not moved during the captures. As the survey was conducted outdoors and not in a controlled environment, during the survey the camera not only framed the area involved in the experiment but also the adjacent area. This zone was used as a reference for the temperature value of the dry part. The different images, having been taken outdoors, are in fact subject to strong changes in temperature due to sunlight, clouds, etc. To compare the 20 images of the set, each thermal image was transformed by subtracting the temperature value of the surrounding dry part from the raster. The value of dryness has been calculated by averaging the temperature at 3 different points in the dry zone for each acquisition. Having subtracted the dry value, the thermal images thus processed express the T value (Fig. 5).

Fig. 5. Images transformed to dry values: (a) before wetting, (b) during wetting, (c) during the drying phase.

10

E. Barbieri et al.

Extracting information from a set of thermal images is not easy, both because of the time component and because each image was acquired under different conditions, subject to greater or smaller variations. In order to better evaluate and visualize the results, a time series was created to describe the behavior of the materials during the active thermal survey. Being a pixel-based analysis, it was necessary to co-register the images in the same reference system in order to have a perfect match of each pixel. Once the images were co-registered, the thermal images that describe specific phases during the active thermal survey were identified. The following images were used: before wetting (t0 ), end of wetting\start of drying phase (t9 ), end of drying phase (t19 ). By associating an RGB channel for each thermal image (t0 , t9 , t19 ) and overlaying the images, the three descriptive moments of the thermal investigation can be displayed simultaneously. Via the multitemporal RGB composite of the three thermal images, it is easy to identify the areas that have absorbed the most water and have not dried out completely (Fig. 6). The areas that absorbed most water are likely those where the hydrophobic treatment is damaged or absent, while the areas that dried out quickly are those where the treatment is still effective and the water drops were not absorbed by the bricks. From

Fig. 6. Thermal images (a) before wetting (t0 , blue channel), (b) at the end of wetting (t9 , green channel) and (c) in the drying phase (t19 , red channel). (d) Multitemporal RGB composite of thermal images t0 , t9 , t19 .

3D Data Management and Thermographic Studies

11

observation of the RGB composition it is possible to identify vulnerable points where problems with the surface water repellent treatment can be assumed. Once the thermal image describing the time course of the active thermal survey was created, it was decided to integrate this information into the metric survey. In order to combine the data acquired from the IR survey with the metric survey, Meshlab software was used to project the multi-temporal RGB composition onto the meshof the area of interest. From the non-resampled point cloud, the portion also affected by the active thermal survey was extracted and the mesh was then processed. The projection of the multitemporal RGB composite was then carried out by homologous points between raster and mesh. This made it possible to texturize the mesh with the colours of the multitemporal RGB composite (Fig. 7) describing the absorption of water in the masonry.

Fig. 7. Mesh coloured with multitemporal IR composite.

In order to insert the information acquired with the thermal survey into the BIM software, it was necessary to align the mesh with the model and the point cloud already available. The mesh was then discretized into a point cloud, to make its management easier within the software. The mesh was resampled in the point cloud, choosing an average point distance of 1 mm. The cloud coloured with the multitemporal RGB composition was then imported into the BIM software where it was possible to include the information extracted from the active thermal survey.

3 Results The aim of the present study was to integrate all the information acquired through the geometric survey and the active thermal survey into a single model. By using the multitemporal RGB composite of the thermal images, it was possible to recognize and

12

E. Barbieri et al.

visually identify points where quick water evaporation did not occur. By importing the cloud coloured with IR composite into the HBIM model, it was possible to identify these points on the surface of the 3D model. To indicate the potentially damaged area, some benchmarks were placed in which the possible degradation of the hydrophobic treatment was indicated in the corresponding attribute tabs within the BIM software (Fig. 8). All information concerning materials, treatment, and any provisions for hydrophobic restoration and/or maintenance could be entered.

Fig. 8. Stages of data integration: (a) HBIM model, (b) import of coloured cloud with IR composite, (c) identification of vulnerable points, (d) benchmarks into HBIM model.

The masonry in HBIM models does not consist of individual elements that can be assigned an attribute independently from the rest of the structure. In order to identify the position of the damaged area, it was therefore essential to integrate and co-register all the products processed in the same reference system. The points identified by the benchmarks refer to those areas that experienced a greater temperature drop between the first (t0 ) and the last acquisition (t19 ). As can be seen in Fig. 9, some areas dried much slower, especially in points P6 , P3 and P4 . P2 and P7 have a similar but less pronounced behavior. Points P1 and P5 , on the other hand, although

3D Data Management and Thermographic Studies

13

not yet completely dry in the acquisition at t19 , have smaller T values between the first and last acquisition. Within the HBIM (Fig. 8d), only points P2, 3, 4, 6,7 were included, as they show a more significant variation between the first and last acquisition, in terms of T (Fig. 9).

Fig. 9. Comparison of first and last thermal imaging and identification of possible degraded areas.

The areas identified by the benchmarks were then reported as potentially exposed to atmospheric agents because not properly treated with water repellents. The HBIM model was therefore equipped not only with geometric information but also with elements that indicate the areas to pay attention to in future monitoring and/or maintenance interventions. The method used is intended to be an application example of how to integrate different information for the diagnostics of existing buildings. The methodology presented here can be applied at larger scales, using a thermal camera mounted on a drone for the thermal survey. To simulate the active thermal survey, it is preferable to plan the flight after atmospheric events that wet the structure of interest. In this way, information on water absorption in different areas could be acquired. Thermography can also be used to detect the position of thermal or plumbing systems or to detect thermal bridges. All these non-destructive investigations make it possible to acquire information stored in the HBIM model, which aims at creating an identity card for the building, complete with the history of interventions and plans for future operations.

4 Conclusions The integration of different technologies allowed to merge the different data acquired and bring them together in a single digital information system. The adopted methodology made it possible to create a 3D model of the tower structure of the Engineering Faculty in Bologna, using aerial photogrammetry. The model acquired can also be used to calculate defects in the materials such as deformations, cracking and defects of the applied treatments, which can be superimposed on the model. For example, the active thermal survey carried out showed the behavior of the bricks treated with water-repellent during the wetting and drying phase. The comparison of the thermal images acquired at the beginning and at the end of the thermal survey made it possible to identify vulnerable zones, probably subject to degradation. The processing of the RGB

14

E. Barbieri et al.

composite of three thermal images allowed a dynamic view of the state of the materials during the survey. Using images before wetting, at the end of wetting and at the end of drying, the RGB composite provides a history of the material’s behaviour at these key moments. This type of visualisation therefore makes it possible to quickly identify areas showing abnormal drying behaviour. The colouring of the mesh with the colours of the IR composite made it possible to identify points in space, thanks to the co-registration of all the entities. The areas to be monitored, marked with the benchmarks, are therefore areas that will have to be investigated with further diagnostic surveys in order to intervene with maintenance operations aimed at restoring the effectiveness of the water repellent. The opportunity to include precise information about materials in the HBIM model opens up new ways of managing existing buildings, particularly those having historic and artistic value. The method described makes it possible to integrate geometric survey and diagnostic techniques thanks to the co-registration of data in order to identify degraded areas in space and thus create a BIM model that includes information on the state of the materials and the treatments applied. This approach can be used as a method for predictive analysis, using thermography as a non-contact diagnostic technique, and the HBIM as a reference model where all information on the structure, materials and components can be centralized. Acknowledgements. This work was carried out in the framework of the IDEHA project: Innovation For Data Elaboration In Heritage Areas. The project is co-financed by the European Union - ERDF and ESF, PON Research and Innovation 2014–2020 and is led by the Italian National Research Council (CNR).

References 1. Kioussi, A., Karoglou, M., Labropoulos, K., Bakolas, A., Moropoulou, A.: Integrated documentation protocols enabling decision making in cultural heritage protection. J. Cult. Herit. 14, e141–e146 (2013) 2. Bitelli, G., et al.: The GAMHer research project for metric documentation of cultural heritage: current developments. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 42, 239–246 (2019) 3. Napolitano, R., Hess, M., Glisic, B.: Integrating non-destructive testing, laser scanning, and numerical modeling for damage assessment: the room of the elements. Heritage 2, 151–168 (2019) 4. Betti, M., Bonora, V., Galano, L., Pellis, E., Tucci, G., Vignoli, A.: An integrated geometric and material survey for the conservation of heritage masonry structures. Heritage 4, 585–611 (2021) 5. Bitelli, G., et al.: The complex of Santa Croce in Ravenna as a case study: integration of 3D techniques for surveying and monitoring of a historical site. In: Proceedings of the 9th ARQUEOLÓGICA 2.0 & 3rd GEORES, Valencia, Spain, pp. 408–413 (2021) 6. Bitelli, G., Castellazzi, G., D’Altri, A.M., De Miranda, S., Lambertini, A., Selvaggi, I.: Automated voxel model from point clouds for structural analysis of cultural heritage. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. XLI-B5, 191–197 (2016)

3D Data Management and Thermographic Studies

15

7. Grinzato, E., Ludwig, N., Cadelano, G., Bertucci, M., Gargano, M., Bison, P.: Infrared thermography for moisture detection: a laboratory study and in-situ test. Mater. Eval. 69, 97–104 (2011) 8. Kylili, A., Fokaides, P.A., Christou, P., Kalogirou, S.A.: Infrared thermography (IRT) applications for building diagnostics: a review. Appl. Energy 134, 531–549 (2014) 9. Brumana, R., et al.: Combined geometric and thermal analysis from UAV platforms for archaeological heritage documentation. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. II-5/W1, 49–54 (2013) 10. Adamopoulos, E., Rinaudo, F.: Close-range sensing and data fusion for built heritage inspection and monitoring—a review. Remote Sens. 13, 3936 (2021) 11. Bruno, N., Roncella, R.: HBIM for conservation: a new proposal for information modeling. Remote Sens. 11, 1751 (2019) 12. Tang, S., Shelden, D.R., Eastman, C.M., Pishdad-Bozorgi, P., Gao, X.: A review of building information modeling (BIM) and the internet of things (IoT) devices integration: Present status and future trends. Autom. Constr. 101, 127–139 (2019) 13. Piacentini, M.: Opere di Giuseppe Vaccaro. Architettura X (1932) 14. Vaccaro, G.: L’edificio per la facoltà di Ingegneria dell’Università di Bologna. Architettura III, 97–118 (1936) 15. Sasdelli, R.: Ingegneria in guerra. La Facoltà di Ingegneria di Bologna dalla RSI alla Ricostruzione 1943–1947. CLUEB, Bologna (2007) 16. Franzoni, E., Leemann, A., Griffa, M., Lura, P.: The “Terranova” render of the engineering faculty in Bologna (1931–1935): reasons for an outstanding durability. Mater. Struct. 50, 1–14 (2017) 17. Bernardo, G., Palmero, I.L.: Architecture and materials in the first half of the 20th century in Italy. Int. J. Herit. Archit. 1(4), 593–607 (2017) 18. Marín, J.A.C., Santos, A.G.: Application method of water repellent products and its influence on the suction process on the facade of ceramic brick. Rev. Constr. 13, 3–9 (2014) 19. Duarte, R., Flores-Colen, I., de Brito, J., Hawreen, A.: Variability of in-situ testing in wall coating systems - Karsten tube and moisture meter techniques. J. Build. Eng. 27, 100998 (2020) 20. Herrmann, T.D., Mohamad, G., Lima, R.C.A.D., Santos Neto, A.B.D.S., Lübeck, A.: Avaliação do comportamento de estanqueidade à água de argamassas e hidrorrepelentes – Parte II. Matéria 24, e-12517 (2019) 21. Sansonetti, A., Casati, M., Rosina, E., Gerenzani, F., Gondola, M., Ludwig, N.: Contribution of IR thermography to the performance evaluation of water repellent treatments. Restor. Build. Monum. 18(1), 13–22 (2012)

XpeCAM: The Complete Solution for Artwork Documentation and Analysis Vassilis M. Papadakis(B)

, Marlene Machado , and Jani dos Santos

XpectralTEK LDA, Braga, Portugal [email protected]

Abstract. Spectral imaging technology has always been a great analytical technique for the study and documentation of painted works of art, especially paintings. Although the results are very promising, this technology is only available in large laboratories and is applied specifically to expensive artworks. This is because of the high technical knowhow and the significant time required to complete a study. Additionally, the high cost of such instrumentation holds the wide spread of spectral imaging application in daily practice. The XpeCAM solution was developed to overcome the limitations mentioned above with the aim of making spectral imaging technology the tool in the hands of every conservation scientist. The solution is composed by 3 major components. A broad wavelength illumination source LAMPA build by established light bulbs, able to cover the wavelength range between 360 nm and up to 1200 nm, with a controllable selection of the individual sources. The state-of-the-art multispectral imaging system XpeCAM X02, capable of acquiring 30 images covering the range between 350 nm and 1200 nm, is fully automated, portable and user-friendly. The acquired data are uploaded to XpeCAM Platform, a cloud-based application coupled with Artificial Intelligence technology to enable automated processing, analytics, visualization, and reporting. In this work, we will discuss technical advances in multispectral imaging with applications in cultural heritage. Results will show the benefits of this complete solution in the management, restoration and documentation of an artwork. Keywords: Multispectral imaging · Pigment characterization · Artificial intelligence (AI) · Visualization

1 Introduction Works of art, and in particular paintings hold a great history as they age. Studying them, one can find information regarding their origin, author, hidden information or even an artwork, any previous attempts for conservation, and many others. Extracting such information is not easy. There is a plurality of analytic techniques to use for any specific question, but each of them requires a lot of experience and knowhow, as also specialized instrumentation [2–4, 7, 9, 10, 15, 16]. One very popular technique which provides analytic imaging information is spectral imaging (SI) [1, 5, 6, 8, 11, 13, 14]. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 16–27, 2023. https://doi.org/10.1007/978-3-031-17594-7_2

XpeCAM: The Complete Solution

17

Unfortunately, although SI is very efficient, it suffers from two main facts. Data acquisition is very time-consuming, requiring significant user effort to successfully collect data, resulting in multiple images with a lot of information. To extract useful information from such complicated datasets, advanced processing methods, complete reference databases, and smart analytical methods are required. As such, scientists have to make a decision based on the importance of the artwork and the available budget they have from their clients. If allowed, they subcontract the advanced analytics to a specialized entity, but most commonly they end up using their handy DSLR cameras for the documentation of the artwork condition before and after any conservation practice. The XpeCAM solution has progressed beyond the state-of-the-art, with a focus on empowerment art scientists with a disruptive, novel user-friendly tool, enabling its application to any artwork. Under the support of leading conservation institutes that beta tested the full solution it has adapted to the actual needs of the end user [17]. It is an intelligent and complete solution for the management of artwork or any other painted surface that aims to make spectral imaging technology the tool in the hands of every conservation scientist. It is composed of 3 main components: The multispectral imaging system XpeCAM X02, the broadband illumination source Lampa, and the XpeCAM Platform, which is a cloud and AI based application with complete solutions for the end user. With the XpeCAM solution, precision conservation is transformed through an innovative spectral imaging solution with novel features such as portability, unique design, user-friendliness, automation, measurement quality, and intelligent analytics.

2 Materials and Methods The XpeCAM solution is composed of 3 different products: XpeCAM X02 which is the multispectral imaging sensor; Lampa a complete illumination system that covers the spectral sensitivity range of the multispectral imaging system; and XpeCAM Platform our artwork and data management application. In the following paragraphs, the individual components are explained. 2.1 XpeCAM X02 The state-of-the-art multispectral imaging system XpeCAM X02 (Fig. 1) is a patented [12] portable spectral camera system based on optical band-pass interference filters. This novel multispectral imaging system is capable of acquiring 30 spectral images covering the range between 350 nm and 1200 nm. It is fully automated and has a userfriendly data acquisition interface named XpecEye. The camera has a high quality sensor of 6.4 MPixel spatial resolution and a smart auto-focus mechanism that results in sharp images across all the spectral sensitivity range. The objective lens used was a 35 mm fixed focal length, but the camera has a cmount adapter, allowing coupling with any optical system with this mount, ranging from a microscope up to a telescope. It is powered and controlled by a single USB cable (USB 3.0) that when connected to a laptop it can hold full operation for hours depending on the model. The full list of XpeCAM X02 specifications is shown in Table 1.

18

V. M. Papadakis et al.

Fig. 1. XpeCAM X02 multispectral imaging camera

Table 1. XpeCAM X02 hardware specifications XpeCAM X02

Hardware specification details

Sensor

6.4 MP monochrome array sensor

Spectral sensitivity

350 nm up to 1200 nm

Spectral bands

30 band-pass interference filters

Image autofocus

Sharp spectral images across all the sensitivity range through an internal autofocus mechanism

Area indicator

Red Laser pointer module

Status indicator

Led color code indicator

Objective lens

C-mount adapter

Interface

USB 3 connection

Tripod

Standard Tripod mount

Power

DC powered through USB interface

Housing

Robust plastic (ABS) design

Dimensions

190 × 190 × 190 mm

Weight

1.8 kg weight

XpecEye The acquisition application XpecEye has a novel interface which was designed and developed to allow users to easily interact with the camera features. The user interface (UI) was designed following a lot of feedback from multiple users experience (UX) and is shown in the next Fig. 2.

XpeCAM: The Complete Solution

19

Fig. 2. The control and acquisition software application XpecEye, for both XpeCam X02 and Lampa hardware.

As it can be seen in Fig. 2a, the user starts by adding the name of the project of the acquisition that will take place. A project folder is then created in “my documents” folder under the name of the project, where all files will be saved with a structured way in subfolders. The user can then start with the calibration process to ensure that the experimental setup is ready for sample acquisition. As seen in Fig. 2b, the user can select between automated and manual image acquisition. In manual mode, controls can modify the sensor sensitivity, select the wavelength band, use zoom for better manual focus results, and save the image. In Fig. 2c, the filter selection wheel is presented, where the user can easily select the wavelength band for imaging. In auto mode, the automated acquisition range is fixed and set from 400 nm up to 1000 nm. Lastly, as shown in Fig. 2d the user has some extra control options as opening a previously saved project, choosing the number of images that will be averaged to reduce noise in each image acquisition, turning the laser pointer on/off, and watching the help tutorial. All acquired images and their metadata are stored uncompressed with automatic named filenames. The features of the XpecEye acquisition software are presented in Table 2.

20

V. M. Papadakis et al. Table 2. XpecEye features

XpecEye acquisition software features Easy operational user interface Wavelength and exposure time control Auto calibration process Auto acquisition Image histogram Image noise removal Metadata storage Temperature and humidity monitoring Real time visual magnification Laser pointer control

2.2 Lampa The illumination system Lampa presented in Fig. 3 is a broad spectral illumination system built by established light bulbs, able to cover the wavelength range between 360 and 1200 nm, with a controllable selection of the individual sources. LAMPA was designed and developed to meet the specifications and operate with XpeCAM X02, replacing the multiple commercial light sources that were previously required for the acquisition process.

Fig. 3. LAMPA light source solution. The different illumination components are present.

XpeCAM: The Complete Solution

21

The system covers with individual components the ultraviolet, visible, and infrared bands. In the following Table 3, we can see the hardware specifications of Lampa illumination system. The illumination system can be controlled in 3 different ways. As it was mentioned before it can be controlled automatically by XpeCAM X02, and through the selection of a wavelength band, power on/off the related light component. This way the total irradiation towards the artwork is limited. In manual mode Lampa can be controlled through a keypad on the back of the housing or through a special remote controller that the user can control while close to the artwork. Table 3. LAMPA hardware specifications LAMPA

Hardware specification details

Light sources

AC tungsten and florescent light bulbs

Spectral emission

360 nm up to 1200 nm

Light intensity control

Dimmer controlled sources

Spectral illumination control

Spectral wavelength band control through: 1. XpeCAM X02, 2. Remote control, 3. Manual keypad

Indicator

LCD screen display

Tripod

Standard Tripod mount

Power

240V AC - with on/off button

Housing

Robust plastic (ABS) design

Dimensions

282 × 134 mm

Weight

2.5 kg weight

The connection between XpeCAM X02 and LAMPA is achieved using the WiFi protocol. The XpecEye acquisition software is responsible to start the connection between the devices. The communication is established during the software initialization so the user only has to power on the systems. 2.3 XpeCAM Platform Following acquisition, all acquired data can be uploaded to the XpeCAM platform. The platform is developed to automatically pre-process, process, analyze and generate new image data. Each step is triggered by the data status in to the platform. Within its cloud-based capabilities it includes AI-supported features like management, data processing, data analysis, visualization, and reporting tools. The most important features of XpeCAM platform with a brief explanation are presented in the next Table 4. The platform is capable of keeping the management of the full conservation work on an artwork including documentation, intervention, analysis and reporting, through a user friendly workspace. The process is initiated by the user, starting a new project and adding

22

V. M. Papadakis et al.

the related information. The acquired data from the multispectral imaging system can be uploaded to the XpeCAM Platform, at any time decided by the end user, to trigger the automated processes and receive all results in a short time (~1 h). The user then can download the data, or keep the full project in the cloud platform for future reference or as a management repository of his work. Table 4. XpeCAM platform features Platform features

Description

Data import

XpeCAM X02 multispectral datasets and metadata

Data preprocessing

Automated light intensity and homogeneity normalization Image co-registration within the datasets

Image processing

Image generation (Color Image, UVRFC, IRRFC) Spectral correlation mapping

Machine learning

Pigment classification Pigment mapping

Image visualization

Image visualization window Image comparison curtain tool

Diagnosis management

Artwork condition information form Artwork condition images management

Intervention management

Intervention information form Before and after intervention image management

Image management

Image and picture gallery

Reporting

Artwork full report Condition report Intervention report

Through the XpeCAM platform, users have the ability to manage their daily work through project information, diagnosis and condition documentation, intervention steps and pictures, to visualize all project images through the gallery and study all the acquired spectral data, or to generate the automated report for editing and downloading (Fig. 4). The platform’s artificial intelligence (AI) algorithms are trained through our cloud reference database, which is composed of thousands of different reflectance spectra, including historical pigments, in different binding media, layer thicknesses, varnishes, and substrates. AI is then used to classify the pigments on the sample resulting in pigment maps. Automation completes the tasks by generating a report, which includes all produced results. The end user can download this report to modify and add extra information, which saves productive time that can be invested in more important tasks. 2.4 Calibration Methodology Acquisition starts by setting up the multispectral camera XpeCAM X02 and the illumination system of the two Lampas in a geometry of around 45° angle of incidence

XpeCAM: The Complete Solution

23

Fig. 4. XpeCAM Platform main menu screen

with the artwork. The first task to do when the experimental setup is set is to use a Diffuse Reflection Target based on spectralon or through our own developed target named XDRT. We are developing those targets to meet all the requirements for a correct system calibration while at the same time being cost-efficient making them easy to replace as a consumable. Calibration requires to perform image acquisitions of the surface of the DRT. During calibration the system automatically controls the sensor shutter in order to achieve maximum exploitation of the sensors dynamic range, without any saturation. After calibration of the experimental setup, the DRT is removed and acquisition of the sample’s surface is achieved. The acquired images are then processed in the platform for normalization of the sample’s images, achieving spectrally corrected data.

3 Results and Discussion The artwork used to test the system was a painting by Bento Coelho da Silveira, a famous Portuguese painter from the 17th century, one of the fewest that survived the 1755 Lisbon earthquake, in 8.5–9.0 on the Richter scale. This unique artwork was in conservation at the Signinum Laboratory (Signinum, Portugal). 3.1 Acquisition 1. The setup of the acquisition together with acquisition times and methodology. As discussed acquisition starts by setting up the multispectral camera, and the illumination system in a geometry of around 45° angle of incidence towards the surface to be studied as seen in the next Fig. 5. Then using the DRT the system starts the automated calibration process. This process takes approximately 5 min, and it is variable based on the distance of the illumination system from the target and the angle of incidence of each Lampa. Sample’s image acquisition lasts approximately 6 min and is strongly dependent on the surface absorption and spatial features because of the autofocus control.

24

V. M. Papadakis et al.

Fig. 5. XpeCAM solution in Signinum Lab during data acquisition

In the following list, a complete acquisition of the sample is presented. As is seen, in this example, the system acquired images across the range between 370 nm and up to 1100 nm. The images at 370, 380, and 390 nm clearly show the varnish layer on the left side of the artwork, which becomes transparent as the wavelength increases (Fig. 6).

Fig. 6. The array with the spectral images acquired by XpeCAM X02 and LAMPA setup. The image sequence starts at 370 nm and goes up to 1100 nm at image #1 and #24, respectively.

The spectral images in the visible spectrum are between #4 (400 nm) and #16 (700 nm) and hold all the color information of the pigments. This is where most of the pigment classification is based. Images crossing the visible range towards the nearinfrared spectrum are between #17 (750 nm) and and up to #24 (1100 nm) show the IR channels where underdrawings become more visible.

XpeCAM: The Complete Solution

25

3.2 XpeCAM Platform When images are uploaded to the platform, they are automatically processed and analyzed, and new image data is generated. The first images that are produced are the color representation of the artwork and some false color images that conservation scientists need in their work. These images are shown in Fig. 7.

Fig. 7. The pseudo color images generated by the platform: a) is the RGB reconstruction b) is the UVRFC image; c) the IRRFC1 image; and d) the IRRFC2 image

The RGB reconstruction is a close representation of the visible photograph. Ultraviolet reflectography false color image (UVRFC) enhances the protective layers enabling in a colorful way to visualize where the varnish is already removed and to evaluate the quality of the process to ensure that there was not residual subproducts left on the surface. Infrared reflectography false color (IRRFC) images 1 and 2, allow the better understanding of the color layer thickness or to understand if the cleaning process is too aggressive before it goes obvious on the experienced Conservator’s vision. This data allowed them to prevent damages to occur, allowing to perform a safer and consequently faster cleaning process. Spectral cubes are analyzed through automated processes and pigments are characterized. This is achieved through AI algorithms and our XpeCAM platform reference database. In the following Fig. 8, a resulting list of classified pigments can be seen together with the percentage of surface coverage on the area of interest. The resulting classification map that contains all pigment classified are shown in the next Fig. 8b, and are overplayed with the related corresponding color on the color image. Each classified pigment can also be visualized individually with a bright overlay color on the color image. An example is presented in Fig. 9, where the Minium red pigment is shown.

26

V. M. Papadakis et al.

Fig. 8. a) the classified pigment list, and b) the pigment classification map

Fig. 9. Mapping of Minium red pigment. The areas where the pigment is identified are shown with bright color overplayed on the RGB image.

4 Conclusions XpeCAM Solution is expected to revolutionize cultural heritage documentation. It is a state-of-the-art, fully automated, cloud-based solution with artificial intelligence capabilities, that manages to significantly simplify the documentation and analysis process. It is anticipated that the XpeCAM solution will be constantly used in the everyday work of conservation scientists, art historians, museum curators, and collectors. The solution is constantly developing and multiple new applications will be released aiming to enable the communication and discussions between professionals from different disciplines. Furthermore, classification capabilities will continue to improve through the continuous growth of the pigments/lacquers/mixtures database. This will be achieved through

XpeCAM: The Complete Solution

27

our collaborating institutions and partners whose contributions will revolutionize art analysis. Acknowledgements. The authors would like to acknowledge that this project was funded by EC under the topic EIC-SMEInst-2018–2020 - SME instrument (Grant agreement ID: 811764).

References 1. Bonifazzi, C., Carcagni, P., Fontana, R., et al.: A scanning device for VIS-NIR multispectral imaging of paintings. J. Opt. A-Pure Appl. Opt. 10, 064011 (2008) 2. Defeyt, C., Langenbacher, J., Rivenc, R.: Polyurethane coatings used in twentieth century outdoor painted sculptures. Part I: comparative study of various systems by means of ATRFTIR spectroscopy. Herit. Sci. 5, 1–11 (2017) 3. Detalle, V., Glorieux, Q., Bruder, R., et al.: Laser induced breakdown spectroscopy (LIBS): a new analytical technique for in situ study of painted artworks. Actual Chim. 98–104 (2007) 4. Gomez, M., Reggio, D., Lazzari, M.: Linseed oil as a model system for surface enhanced Raman spectroscopy detection of degradation products in artworks. J. Raman Spectrosc. 50, 242–249 (2019) 5. Kaew-On, N., Katemake, P., Tremeau, A.: Monitoring paint and primer samples using multispectral and hyperspectral imaging techniques. ScienceAsia 46, 110–116 (2020) 6. Liang, H., Saunders, D., Cupitt, J.: A new multispectral imaging system for examining paintings. J. Imaging Sci. Tech. 49, 551–562 (2005) 7. Lungulescu, E.M., Lingvay, I., Bors, A.M., et al.: Assessment of paint layers quality by FTIR and DSC techniques. Mater. Plast. 56, 87–91 (2019) 8. Marengo, E., Manfredi, M., Zerbinati, O., et al.: Development of a technique based on multispectral imaging for monitoring the conservation of cultural heritage objects. Anal. Chim. Acta 706, 229–237 (2011) 9. Martinez-Hernandez, A., Oujja, M., Sanz, M., et al.: Analysis of heritage stones and model wall paintings by pulsed laser excitation of Raman, laser-induced fluorescence and laserinduced breakdown spectroscopy signals with a hybrid system. J. Cult. Herit. 32, 1–8 (2018) 10. Mencaglia, A.A., Osticioli, I., Siano, S.: Development of an efficient and thermally controlled Raman system for fast and safe molecular characterization of paint layers. Measurement 118, 372–378 (2018) 11. MaT, M., Osorio, G., Montes, N.L., et al.: Characterization of a multispectral imaging system based on narrow bandwidth power LEDs. IEEE Trans. Instrum. Meas. 70, 1–11 (2021) 12. Papadakis, V., Campos, C.: Spectral camera having interchangeable filters. In: USPTO, USA (2017) 13. Papadakis, V., Orphanos, Y., Kogou, S., et al.: IRIS: a novel spectral imaging system for the analysis of cultural heritage objects. SPIE (2011) 14. Pelagotti, A., Del Mastio, A., De Rosa, A., et al.: Multispectral imaging of paintings. IEEE Signal Process. Mag. 25, 27–36 (2008) 15. Pereira, M.O., Felix, V.S., Oliveira, A.L., et al.: Investigating counterfeiting of an artwork by XRF, SEM-EDS, FTIR and synchrotron radiation induced MA-XRF at LNLS-BRAZIL. Spectrochim. Acta A 246, 118925 (2021) 16. Sessa, C., Vila, A., Garcia, J.F.: Determination of detection limits for SEM-EDS and m-FTIR analysis of artwork. Anal. Bioanal. Chem. 400, 2241–2251 (2011) 17. Seymour, K., Santos, J., Papadakis, V.M., et al.: Visualising artworks: translating the invisible into diagnostic data for identifying and quantifying paint surfaces. In: Furferi, R., Governi, L., Volpe, Y., Gherardini, F., Seymour, K. (eds.) Florence Heri-Tech 2022. LNME, pp. xx–yy. Springer, Fluorence (2022)

Visualising Artworks: Translating the Invisible into Diagnostic Data for Identifying and Quantifying Paint Surfaces Kate Seymour1(B) , Jani Santos2 , Vassilis M. Papadakis2 , Betlem Alaponte1 , Alzbeta Prochazkova1 , Valentine Gatto3 , Alice Limb4 , Nalini Biluka5 , and Luuk Hoogstede1 1 Stichting Restauratie Atelier Limburg (SRAL), Maastricht, The Netherlands

{k.seymour,l.hoogstede}@sral.nl, [email protected] 2 XpectralTEK LDA, Braga, Portugal {jani.santos,vassilis.papadakis}@XpectralTEK.com 3 University of Amsterdam, Amsterdam, The Netherlands 4 Hamilton Kerr-Institute, Cambridge, UK [email protected] 5 Indian National Trust for Art and Cultural Heritage (INTACH), Bangalore, India

Abstract. The abstract should summarize the contents of the paper in short terms, i.e. 150–250 words. Multi-spectral imaging (MSI) is today a commonplace tool in the diagnosis of painted artworks. Instrumentation is becoming more accessible with a wide variety of custom and commercial equipment available. Obtained spectral data can be interpreted to provide information about pigments, painting practice, conservation history and condition of painted art works. Such data is typically analysed by the expert user (a scientist) and helps the way practitioners (conservators) visualise details presented by each unique artwork. Artificial intelligence (AI) is now being applied to the field of MSI. Machine learning enables the efficient comparison of spectral responses to pre-existing databases, providing automated interpretation of results. Clear mappings of selected pigments, damage patterns or condition phenomenon provide the conservator with a new means to visualise, diagnose and monitor artworks. An innovative technological system, provided by XpectralTEK (Portugal), was trialled at SRAL (Maastricht) in the spring of 2021. Multispectral images (360–1200 nm) were captured using the unique camera and acquisition platform. AI learning created overlays and mappings empowering the conservators by providing material-technical insights. This paper will report on case studies highlighting the functionality of the XpectralTEK’s system. We will discuss how we familiarised ourselves with the new technology, adapted our working practice and how our knowledge of the paintings studied was enhanced. We will emphasise the training, interpretation and learning that our conservation team encountered, guided closely by the XpectralTEK support staff. Keywords: Multi-spectral imaging · Pigment identification · XpectralTEK · Artificial intelligence (AI) · Pilot study · Beta testing

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 28–47, 2023. https://doi.org/10.1007/978-3-031-17594-7_3

Visualising Artworks: Translating the Invisible into Diagnostic Data

29

1 Introduction Multi-spectral imaging (MSI) is a commonplace tool used by art professionals, (art) historians and (conservation) scientists in the diagnosis of painted art works. The instrumentation required is becoming more and more accessible, with a wide variety of custom and commercial equipment available.1 When executing MSI, images are acquired at incremental, stepped wavelengths resulting in spectral data cubes. Each corresponding pixel from each image can be registered in the X and Y axes, and the Z dimension provides wavelength information for that specific pixel for each registered image taken. The images and data can be interpreted to provide information about pigments, painting practice, conservation history and the condition of painted art works. Such data is typically interpreted by the expert user, a conservation scientist, and these interpretations are provided to the practitioner, an art conservator, to help visualise details presented by each unique artwork. Art conservation professionals are well-equipped, through their academic training backgrounds, to use technical photography, acquiring images of art works at different wavelengths, to investigate artworks. However, when it comes to utilising spectral data, many conservators will rely heavily on the interpretations made by the scientist. Of course, this statement sketches extremes and many conservators have a deep understanding of MSI and are fully capable of construing conclusions from acquisitions. Furthermore, conservators and scientists often forge deep collaborative partnerships, furthering the understanding of analytical results obtained using MSI. More importantly, the results obtained should be easy to convey to other art professionals, such as curators, and clients.

2 The XpectralTEK System XpectralTEK is a company with a focus on imaging diagnostics, that creates tools to assist professionals in their daily work.2 It offers quality proven solutions, with specialisation in spectral imaging. XpectralTEK has developed a system capable of high-resolution imaging in real time. The XpectralTEK technology consists of a multispectral, mono-chromatic camera (XpeCAM) with a complementary metal-oxide semiconductor (CMOS) sensor, autofocus capabilities, and a 35 mm C-mount lens. Narrow bandwidth filters eliminate unwanted wavelengths and are included internally within the camera housing. The hardware is connected, via a computer, to an acquisition platform (XpecEYE) which guides the conservator through a series of automated captures from 380 nm to 1000 nm. Manual settings extend the wavelength range captured from 350 nm to 1200 nm. In total, 30 bandwidths can be captured for each spectral cube. The camera is linked, via Bluetooth, to two custom light modules, named LAMPA, containing broadband visible, ultraviolet and infrared light sources, that are automatically driven by the system. All images are adjusted using acquisitions at each wavelength of a Direct Reflectance Target (DRT). Post processing of the captured images using the DRT captures increases accuracy of procurement eliminating reflectance anomalies. The system is completed with an appropriate wavelength selection system, which interprets 1 URL: https://chsopensource.org/multispectral-imaging-msi/ [accessed 13 January 2022]. 2 URL: www.XpectralTEK.com [accessed 13 January 2022].

30

K. Seymour et al.

the incoming light as a function of wavelength. The images and data acquired are transferred to a cloud-based platform (XpeCAM Platform)3 , which provides integrated and collaborative visualisation of the visible and invisible.4 The innovations provided by the XpectralTEK system aim to make MSI more accessible and hands-on in daily practice to art professionals. The camera and lights while similar to other systems available function in tandem with each other and are operated through a unique, easy to function computer system. Once the MSI images are captured, the ground-breaking system uses artificial intelligence (AI) algorithms, applied to the MSI acquisitions, to provide clear analytical results empowering diagnostic capabilities by the user. Machine learning enables the comparison of spectral responses at specific wavelengths to existing databases, providing immediate interpretation of results. The visualisation of results, generated by AI algorithms, provides clear mapping of selected pigments, damage patterns or condition phenomenon allowing the conservator a new means to diagnose and monitor artworks. An innovative technological MSI system using AI, developed and provided by XpectralTEK (Portugal), was trialled at SRAL (Maastricht) in the spring of 2021. The trial put the new instrumentation, user interface and analytical AI system through its paces, equipping the SRAL conservators with valuable insights into several artworks being treated in the conservation studio. This beta testing allowed the XpectralTEK developers to modify and improve the system and user experience. Multispectral images (350–1200 nm) were captured for each artwork, included in the trial, using the unique camera and acquisition platform. AI learning created overlays and maps empowering the conservators to identify more easily the effects of past conservation treatments, while providing material-technical insights. This mutually beneficial collaboration provided both parties with constructive feedback and has enhanced the AI potential of the system by increasing the capacity of the pigment database and the interpretation of acquired spectral data cubes.

3 Stichting Restauratie Atelier Limburg (SRAL), Maastricht SRAL is a leading, not-for-profit, conservation institution situated in Maastricht (the Netherlands).5 SRAL provides expertise in four conservation departments in areas relevant to our core business, which are Conservation-Restoration of painted artworks; Education & Training of conservators; Research & Development of conservation practice; Preventive Conservation and Advice. SRAL conservators offer expertise in the conservation of easel paintings, polychromed sculpture, historic interiors, which includes wall paintings, and modern and contemporary art. These conservation departments (Fig. 1) guide the essential activities of the institution. Each department is staffed with in-house experts who are recognised in their own fields regionally, nationally, and internationally. SRAL’s expertise is transferred to the next generation through practical mentoring and training of students and emerging professionals. The institute offers several internships 3 URL: www.xpecamplatform.com [accessed 13 January 2022]. 4 Vassilis M. Papadakis and Jani dos Santos, XpeCAM: The Complete Solution for Artwork

Documentation and Analysis. To be published in Heri-Tech Conference, May 2022. 5 URL: www.SRAL.nl [accessed 13 January 2022].

Visualising Artworks: Translating the Invisible into Diagnostic Data

31

Fig. 1. One of the conservation studios at SRAL. © SRAL

and fellowships annually and collaborates with the University of Amsterdam (UvA) in training Master students enrolled in the “Conservation and Restoration of Cultural Heritage”.6 The in-house team at SRAL, augmented by interns and fellows, participated in the XpectralTEK trial and their experiences with the system are discussed in this paper. The XpeCAM project (SME Instrument, Phase 2, H2020) was trialled first at SRAL in June 2018 at a workshop introducing the XpeCAM X01 and XpecEYE to Dutch conservators and students. The workshop focused on the then current spectral imaging technology used in active and preventive art conservation. The basic principles of spectral imaging were discussed, and the available technological solutions explained. Potential applications for cultural heritage were presented using real case studies, and future perspectives were outlined. This included the development of the cloud-based AI platform (XpeCAM Platform) that would compare acquired data cubes to a pigment database and intuitively provide pigment identification and mapping. The collaboration between SRAL and XpectralTEK evolved in the interim period, which culminated in a three-month pilot study in the spring of 2021. The pilot study allowed SRAL conservators, fellows, and interns to put the system through its paces, using the results to better understand the material and technical characteristics of the observed artworks (Fig. 2). This enabled improved decision-making for conservation choices and enhanced the reporting provided by the conservator to the client. 6 URL:

https://www.uva.nl/en/programmes/masters/conservation-and-restoration-of-culturalheritage/conservation-and-restoration-of-cultural-heritage.html [accessed 13 January 2022].

32

K. Seymour et al.

Fig. 2. The SRAL team setting up the XPECAM system. © Kate Seymour

This paper will report on two of the numerous projects included in this trial, highlighting the use of the XpectralTEK system to the conservator, scientist and curator of painted artworks. The authors will discuss how they familiarised themselves with the new technology, adapted their working practice and how their knowledge of the paintings studied was enhanced by the results obtained using the XpeCAM X02 and interpreted by the XpeCAM Platform. The projects studied cover a wide typology of art works ranging from medieval tempera panel paintings to oil paintings on plaster from historic interiors. As these are too many to do justice to within the confines of this publication, the authors will focus on the interpretation of image cubes acquired from an eighteenth-century, Dutch canvas painting, the Kapittel van Thorn (unknown artist), and a nineteenth century mural altar-frontal. The paper will also discuss the training, interpretation and learning that the team encountered, guided closely by the XpectralTEK support staff.

Visualising Artworks: Translating the Invisible into Diagnostic Data

33

4 Technical Photography (TP)

Fig. 3. XPECAM and one LAMPA. © Kate Seymour

Technical photography (TP) consists of documenting and imaging a painting through photography using a variety of different illumination sources.7 The surface’s response to interaction with incident light is recorded and the images obtained, processed and analysed through observations and often remembered comparisons, making interpretations fallible. Modern technical photography is the acquisition of a collection of scientific images realised with a modified digital camera sensitive to the spectral range from ultraviolet to infrared. Each digitally acquired image records the specific reaction of the material to the influence of light falling on that surface. Light interacts with matter in a variety of ways - reflection, absorption, scattering or transmission. The ability of a surface to reflect, absorb, scatter, or transmit energy is specific to that material and depends on the wavelength. Some materials are translucent or semi-translucent to (higher) wavelengths and thus the light reflected may be from a lower substrate and the upper layers will not appear on the photographic image. The (digital) photographic image records an amount of reflected light specific to that surface. Thus, each capture at a set wavelength provides specific information on the material’s chemical (and physical) characteristics. Taken together, an image cube which collects captures, taken at incremental steps, will provide spectral information that can be used to analyse the materiality of an object. However, the images alone provide only a visual record of the interaction of light with 7 Due to the more complex instrumentation required, x-radiography is often excluded from the

acquired image cube.

34

K. Seymour et al.

the surface. Spectral data is difficult to extrapolate from these images. The response of the surface needs to be compared with standards to be able to interpret data and the images taken at different focal distances need to be correlated. Surface responses from narrower bandwidths, such as captured with the XpectralTEK system, are required to gain accurate material information. The cubes of images captured using the XpeCAM X02 (Fig. 3) are processed using AI algorithms on the XpeCAM Platform providing far more material information than photographs acquired by Technical Photography. The simplicity or complexity of the technical imaging system will guide the quality and quantity of the spectral information gathered. Typically, at SRAL, artworks are photographed with a high-quality digital single-lens reflex (DSRL) camera, in a dedicated space with specialised lighting rigs emitting visible, ultra-violet, or infrared radiation.8 SRAL staff also produces digital x-ray images (DX/XR) to provide supplementary information and examines objects using infrared reflectography (IRR). The SRAL Hamamastu Videcon IRR system is better at penetrating paints than the modified DSRL camera, although it still struggles with visualising the underdrawing present under blues and greens. Enhanced visible lighting conditions (raking or transmitted light) are also often used to record specific conditions. Initial images are taken to document the condition of the painting prior to treatment, and used to diagnose problems that will be treated in the conservation studio. Artworks are photographed to record the treatment process and the condition of the artwork before restoration is carried out. Finally, artworks are photographed as a testament to the condition that they are returned to the client and to provide evidence of the treatment. The set of images taken with broadband illumination provide a useful record of condition and treatment but have restricted potential for diagnostic or analytical means. For instance, UV radiation at around 365–385 nm can be used to detect the presence of organic surface layers such as varnishes and to tentatively identify some pigments like red lakes.9 However, information at specific wavelengths cannot be captured, making identification of most pigments impossible. The image cubes acquired with the XpectralTEK system enhanced our ability to diagnose and identify the materials of the artworks studies without using destructive sampling methods. Technical photography requires a significant practical and knowledge base of photography. The set up requires adjustment to ensure the images are faithful to the artwork being documented. Replicating the same set up at different moments of treatment is laborious. Frequently the images taken using different lighting conditions are difficult to correlate or register because the lighting set-up (distance and angle from the object) 8 SRAL currently uses a modified Sony A7RIV (61 MP) camera and a Sony FE 55 mm F/1.8

ZEISS Sonnar T* lens with interchangeable filters of Cosentino’s Robertina set (v.2) for our Technical Photography needs. Photographs are typically taken using visible light, ultra-violet and infrared irradiation sources set at a 35°-degree angle to the artwork. Camera presets are used to ensure standardised photography in visible light (VIS), ultra-violet fluorescence (UVF) and infrared (IRP). Additionally, ultra-violet reflectance (UVR) images, infrared fluorescence (IRF) images and false colour infrared (IRFC) images can be made. The DSRL camera is tethered to a laptop which controls the camera through the Adobe Lightrooom software. An X-Rite ColorChecker Passport is used to calibrate visible images taken. The quality of these images is excellent and the colour accuracy is also optimal. 9 See https://aiccm.org.au/national-news/summary-ultra-violet-fluorescent-materials-relevantconservation.

Visualising Artworks: Translating the Invisible into Diagnostic Data

35

is different. Focusing the camera in the different lighting conditions can be challenging as the focal point shifts with wavelength. For example, a UV image will typically record the induced visible fluorescence of the upper varnish layers, meaning that the focus on the paint layers will not be as clear. The quality of the images taken is determined by the equipment (and expertise) available. Shifting away from this traditional photographic set-up to a partly automated one utilising multi-spectral imaging (MSI) has the potential to enhance and expedite the acquisition of this essential documentation process and provide the conservator with more proficiency in analysing and diagnosing materials used by the artist, as well as distinguishing these from non-original additions. Degradation phenomena may also be identified and mapped and tracked over time if image capture conditions can be standardised using MSI or similar technologies. However, the system should be capable of taking images of large surfaces, at high resolution, without requiring stitching of composite acquisitions, to be considered a true replacement for traditional TP setup. XpeCAM X02 has an internal autofocus mechanism that corrects for the focusing when shifting from UV to VIS and NIR spectrum, resulting to sharp images across all the wavelength range. Furthermore, XpeCAM Platform is equipped with advanced co-registration algorithms that solve the stitching issues across the wavelength range, enabling the direct comparison between the multispectral images.

5 Multi-spectral Imaging (MSI) Multi-spectral imaging (MSI) uses a mono-chromatic camera linked to a set of filters and lights. The filters block wavelengths, below and above a narrow bandwidth, arriving at the sensor. The same lighting rigs can be used as for technical photography. Images taken at incremental stepped wavelengths are calibrated and registered on the sensor. Once collected these images can be processed. The camera, lights, and object are not moved between captures. Filters can be placed, either externally or internally, on an automated wheel to minimise displacement. The resulting set of images needs to be correlated to ensure that the pixels from each image/wavelength can be overlaid, and sequential spectral information can be extrapolated from each pixel. Other systems utilise lights that emit in a narrow bandwidth to limit wavelength. The use of a higher number of discrete images relative to narrow bandwidths allows the shift in surface response to the incremental wavelengths to be highlighted and subsequently reference spectra obtained can be analysed.10 Pigment distributions can be mapped and inconsistencies such overpaints visualised more clearly. 10 A reflectance spectrum shows for each wavelength the ratio between the intensity of the reflected

light and the incident light, measured with respect to a standard white reference. This ratio is called reflectance and is given in percentage (%). Reflectance spectra can provide information useful for the identification of pigments since the light that is not reflected is absorbed or transmitted depending on the chemical composition of the material tested. The reflectance spectral features of materials in the UV/ vis range are attributed to electronic transitions while those in the near infrared (NIR) range to fundamental vibrational overtones and combination modes. Analysis by reflectance spectroscopy is based on building up an appropriate reference spectra database, in the case of art examination, a reflectance spectra database of historical pigments. Cosentino, Antonino. (2015). Multispectral imaging and the art expert. Spectroscopy Europe. 27. p. 6.

36

K. Seymour et al.

The collaboration with XpectralTEK allowed SRAL conservators to review how we undertake our technical photography and documentation. We wanted to streamline condition checking, diagnostic methodologies, and dissemination of results to clients, students and partners. Additionally, we wanted to extrapolate more from the images we take, and use the time invested in documentation and condition checking more efficiently to gather more information about the material-technical aspects of the paintings in our care. As part of a new Memorandum of Understanding (MOU), XpectralTEK provided SRAL with the XpeCAM X02, LAMPAs and access to the novel XpeCAM Platform for trial testing (Figs. 3 and 4). SRAL conservators and a team of young emerging professionals, led by Kate Seymour, put the system through its paces to discover the potential of this AI system.

Fig. 4. The XPECAM and LAMPAS. © Kate Seymour

XpectralTEK high resolution camera XpeCAM X02 allows for imaging in real time. The instrument’s advantage over imaging systems is its potential to carry out on-site realtime analysis of materials imaged via a spectrometer function. Moreover, the quality of the lens results in sharp images across all the wavelength sensitivity range. This allows the camera to not only to image art objects between 350–1200 nm at good quality and resolution (6.5 MP), utilising a series of narrow band width filters (in increments of 10 nm in the UV, 25 nm in the VIS, and 50 nm in the NIR range), but also to carry out monitoring functions and characterisation of (surface) pigments, including mixtures. The camera can image and obtain millions of spectra from the surface at different spatial

Visualising Artworks: Translating the Invisible into Diagnostic Data

37

points, plotting the information on a graph and image. Furthermore, post processing, such as false colour imaging, and mappings of paint classification, can be applied to digital photographs to make any alterations in underdrawing or as overpaint more easily visible.

6 Baseline Projects

Fig. 5. CHSOS Pigment Checker (2018). RGB generated image (left) and pigment classification XPECTRALTEK platform (right). © SRAL

After an initial, remote instruction session, the team began by acquiring image cubes from a number of baseline projects (Fig. 5). These baseline projects consisted of known paint-outs (pigments bound in a medium) that would provide additional capacity for the database of pigments used by the AI system to determine pigment identifications and mappings. The initial trials acquainted the SRAL team with the XpeCAM camera, setting up the LAMPAs (lights), working with the DRT and navigating the capture process. The team began by examining the CHSOS Pigment Checker (2018).11 This has known pigments bound in gum Arabic applied to a watercolour paper. These are pure pigment systems, i.e. non-mixed pigments bound in the medium, which provide clear and distinctive spectral reflectance spectra. Paints consisting of mixtures of pigments are challenging for traditional MSI systems to identify. Additionally, the SRAL team were curious to see if the XpeCAM Platform could identify accurately mixtures. Thus, 11 URL: https://chsopensource.org/pigments-checker/ [accessed 13 January 2022].

38

K. Seymour et al.

the second baseline project studied was a copy of a 17th century panel paining made as part of the SRAL training programme in 2005. The pigments used and the application processes were known, which simplified the interpretation of the images acquired. As expected, the team observed the underdrawing emerging as the wavelengths increased. The response of the Lead White and Vermilion pigments were also as expected in the near IR region. The two baseline projects assured the team that the operating system was easy to use and intuitive.

7 Case Studies 1. An eighteenth-century Dutch canvas painting, the Kapittel van Thorn by an unknown artist, was brought to the SRAL conservation studios for treatment in early 2021. This large oil painting, measuring 202 × 121 cm (height × width), dates from 1709. The painting belongs to the Abbey of Thorn and was requested for loan for the 2021– 2022 exhibition “The forgotten Princesses of Thorn” at the Limburgs Museum, Venlo.12 The painting shows the family crests of the various eighteenth-century princesses who lived at the Abbey in Thorn, depicted in an architectural setting, with a central portrait of the Princess-Abbess Kunigunde of Saxony, and a landscape showing the town of Thorn with the abbey (Fig. 6). The 2021–2022 Venlo exhibition examines the opulent daily lives of the German princesses, combining objects, letters, clothing, and jewellery with portraiture, illuminating the daily lives of these noblewoman. This painting is key to the didactic aspect of the exhibition, as it sets the scene for the environment in which the princesses lived and acts as a form of introduction to the women. The painting was badly damaged in the past and has two wax-resin linings glued to the reverse, applied to provide additional support to the fragile original canvas and paint layers. Several layers of natural resin varnish had yellowed over time, reducing the clarity of the original paint colours. Overpaint was clearly visible across the painting, due to its poor colour-match and differing texture to the original in normal light. This had mostly been applied to disguise losses and wear to paint passages or to conceal tears to the canvas. The conservation and restoration treatment (undertaken by SRAL Conservation Fellow Alice Limb, with supervision from Luuk Hoogstede) focused primarily on removing these disturbing varnishes and overpaint campaigns and on improving the textual clarity of the worn inscriptions, allowing the painting to fulfil its didactic purpose of introducing these key Abbey figures and their familial allegiances once more. As the extent and non-original nature of these overpaint campaigns was already known to SRAL conservators using conventional TP and visual methods, the Kapittel van Thorn was identified as a suitable trial project for the XpeCAM X02 and XpecCAM Platform. The hope was that the distribution of non-original material could be thoroughly mapped, with the resultant image cubes providing additional information including information pertaining to the pigment composition of the various overpaints. Due to their familiarity with materials and methods of northern European painting of the 12 URL:

https://www.limburgsmuseum.nl/nl/tentoonstelling/de-vergeten-prinsessen-van-thorn/ [accessed 13 January 2022].

Visualising Artworks: Translating the Invisible into Diagnostic Data

39

Fig. 6. Kapittel van Thorn, 1709. Abbey of Thorn. 202 × 121 cm. After Treatment. Visible light. © SRAL

40

K. Seymour et al.

early eighteenth century, as well as their close examination of the original paint surface (including under microscopy), the conservators had a good idea of the likely pigment composition of the original. However, the pigment composition of the overpaints was largely unknown. Resolving these pigment questions would help to confirm the suspected dates of these restoration interventions. The team also hoped to confirm that the detailing in the quadrants of the coats of arms was authentic. Imaging with the XpeCAM X02 took place prior to and during varnish removal, meaning that some areas imaged were free of varnish, while others were still covered by a degraded layer of yellowed natural resin varnish. Examination in UV light illustrated the distribution of the multiple overpaint campaigns present above and below the top layer of varnish, due to the quenching effect of overpaint on UV-induced autofluorescence (Fig. 7). TP (IR) images showed the extent of

Fig. 7. Kapittel van Thorn. Detail. Coat of Arms. Upper Left: TP (VIS) after treatment. Upper Right: TP (IR). Lower: False-colour generated from VIS and IR. © SRAL

Visualising Artworks: Translating the Invisible into Diagnostic Data

41

overpaint and provided insight into the material beneath these non original layers. The false-colour image, created by substituting the red channel of the TP visible image with the TP infrared image, showed pigment distribution, but could not be used to elaborate on pigment identification as many blue pigments appear red in the FCIR imaging. The AI spectral processing of the captured images successfully identified the ‘whitish’ overpaint present in many of the backgrounds of the Princesses’ crests as containing Titanium white, a pigment that was not yet in use in 1709 (titanium white was first produced for artist’s purposes during the 1920s).13 This supported the existing interpretation that the overpaints were fairly recent, likely applied after the most recent structural intervention. This wax-resin lining was dated to 1938 by an inscription on the reverse of the second, outer lining canvas, further correlating to the known date range of titanium white usage. In other areas of the Kapittel van Thorn, the AI spectral interpretations were less accurate (Fig. 8). The pigment Viridian was allocated to an area corresponding to the original paint of the blue sky by the AI algorithm. Viridian cannot be present in the original paint as this pigment was not in use in 1709 (it was developed in the midnineteenth century).14 Manual comparison of the original blue’s reflectance at a variety of known wavelengths in the IR region with the reflectance of reference paint samples at the same known wavelengths was undertaken using captures from the SRAL IR camera as well as XpeCAM captures, which were uploaded to and processed by the XpectralTEK system but not interpreted by the AI. These, coupled with microscopic examination of the paint surface, pointed to the original blue in the sky being azurite. Azurite is a copperbased blue pigment frequently used prior to 1724 (when even cheaper Prussian blue became widely available) by artists seeking to imitate the far more expensive natural ultramarine. In another area of the sky, examined prior to varnish removal, original blue passages with a layer of varnish were identified by the XpectralTEK system as containing the yellow pigment orpiment. While orpiment is possible in a painting of this date (the pigment has been used by artists since antiquity), it would be unusual to mix it into a blue area. Under the magnification provided by a stereo-microscope, the paint film appeared to contain lead white, blue particles thought to be azurite and some small black particles (probably vine or ivory black). The SRAL conservators were therefore surprised by the allocations suggested by the AI algorithms and notified the XpectralTEK team. It seems that the yellowed varnish residues on the surface of this uncleaned area gave a spectral response of this area similar to a reference spectrum that contained orpiment. However, where the yellowed varnish layers were removed, the AI algorithm accurately allocated azurite. This situation could be avoided by allowing the end-user (a conservator with knowledge of historical pigment usage) to confirm or deny suggested pigment allocations by the AI system. This feedback was given to the XpectralTEK team and hopefully will be incorporated in updates to the system. While anomalies and unusual instances can occur outside of the accepted date ranges of historical pigment usage, basing the reference database used to 13 http://www.webexhibits.org/pigments/indiv/overview/titaniumwhite.html [accessed 13 Jan-

uary 2022]. 14 The process for producing Viridan was patented in Paris in 1859. See http://www.webexhibits.

org/pigments/indiv/overview/viridian.html [accessed 13 January 2022].

42

K. Seymour et al.

Fig. 8. Kapittel van Thorn. Detail. Putti and portrait of Princess-Abbess Kunigunde. Right: TP (VIS) after treatment. Middle: Detail X02 RGB. Left: Detail. AI pigment classification. © SRAL

Visualising Artworks: Translating the Invisible into Diagnostic Data

43

inform the AI spectral allocations on existing evidence in conjunction with the known or suspected date of the painting under examination would result in more reliable inferences by the AI system. The large size of this painting brought other factors under consideration when implementing MSI. The relatively small pixel counts of the CMOS sensor (6.5 MP), relative to the high resolution of the TP camera (61 MP), means that the camera does not provide enough resolution for the spectral imaging of the entire artwork in one capture. The maximum distance that the camera can be placed, when coupled with the 35 mm objective lens, from the artwork is 2 m; this gives a field of view of approximately 50 × 50 cm. It is therefore necessary to acquire a large number of captures, should one wish to image the entire surface. The details can be mosaiced together in order to achieve an image of the whole surface. These larger images are often very MB heavy and need computers with considerable processing power to view and manipulate. As yet, although the XpeCAM Platform has the cloud processing resources available to perform analysis of very heavy datasets, it does not provide mosaic possibilities, but this option is being planned and will greatly improve the ability to map pigments across large surfaces. When working with glossy or varnished paint surfaces glare from lighting was also found to be an issue, necessitating changing the angle or position of the LAMPAS. This in turn can affect the quality of the spectral data acquired as well as the ability to replicate imaging conditions, so basic geometric photographic principles must be applied. 2. Nineteenth Century Mural Painting, Altar frontal, the chapel of the Holy Appearance, St. Servaas Basilica, Maastricht. (1893–1894). Pierre Cuypers’ studio. The Chapel of the Holy Appearance (1893–1894) is an iconographic rarity. Two carved angels stand on top of the altar holding up the venerated painted veil of Veronica above the frontal depicting the prostrate, enshrouded (painted) body of Christ entombed within a stone niche (Fig. 9). The free-standing altar is placed within the chapel with walls either side set with votive stones that match the decoration scheme in shape and colour. The ensemble, a tromp d’oeil masterpiece, was created by the studio of Pierre Cuypers, in a typical Neo-Gothic style. The painted surfaces were constructed in two phases 1858–1865 and 1886–1899. The altar frontal requires conservation treatment. Successive and ongoing salt damage has caused paint layers to flake. There is evidence of previous consolidation and overpaint, as well as considerable fresh losses in the white areas and blue background. These retouches are discoloured and becoming visually evident. There is also a discoloured, greenish, coating that is disturbing. This seems to be applied solely to the blue zone. There is also evidence of additional overpainting in the stone area. It is likely that different phases of conservation treatment exist. The research carried out as part of the XpectralTEK trail aided SRAL conservators in identifying areas of original paint and distinguishing these from later applied overpaint. While the overpaints in some areas had discoloured, in other paint passages it was very difficult to determine the retouches. The blue background was especially problematic as the ‘yellowed’ coating masked the edges of the applied overpaint. The AI mapping function was applied and differences in the spectral response in the white

44

K. Seymour et al.

and blue paint passages could be interpreted. This overview indicated the extent of the previously applied overpaint. The information provided an easy manner in which to quantify the extent of the overpaint and estimate the time-scale required to treat the art work. Cleaning tests were carried out to reduce the yellowed coating. The mapping compilation guided the conservators to test in areas where the overpaint was not present and provided assurance that any pigment removed where the overpaint was present was not original. The pigment distribution maps provided by the AI algorithms clearly show the difference between original pigment paint passages and retouches (Fig. 10). Azurite pigment was used to make the paint in blue background. The pigment distribution map clearly shows the extent of loss due to flaking paint caused by salt efflorescence in this area. The ‘voids’ in the azurite map match the map of losses. The gilded halo had been retouched in the past with paints containing cadmium yellow. The sporadic distribution of this pigment in the corresponding map clearly shows that this was applied in a non-uniform manner and thus relates to a later addition. Additional maps plot the distribution of other pigments used by Cuypers’ studio and later overpaints. The distribution maps provide a quantitative value to the amount of overpaint and losses present, information that was of considerable use to the SRAL conservation team when making condition reports and treatment plans for this artwork. The examination of this altar frontal took place on-site in the chapel. This project allowed the SRAL team trialling the XpectralTEK system to experience fully the portability of the system. The lightweight, cube-formed camera fits neatly and securely into its own carry case and the two LAMPAs, and connecting cables, into a second. Two tripods, the DRT target and the laptop complete the equipment. The compact system is easy to set up on-site and meant that new operators quickly picked up the functionality of the system. The XpectralTEK system is not without is teething problems. The collaboration provided a unique opportunity to work with the design and development team at XpectralTEK, providing feedback and commentary. The team at SRAL immediately were taken with the automated system. The unique manner in which the Bluetooth system provides connection between the lights and the camera is smart and useful, eliminating excessive cabling between the parts of the system and ensures minimal irradiation towards the artwork. Minor logistical improvements would include providing a longer cable connecting the LAMPAs. The length provided was same as the optimal distance that the light positioning required, but this optimal positioning proved difficult to maintain when working on-site, or with larger paintings. The automated switching between light sources housed within the LAMPAs was useful and user friendly. The team appreciated greatly the ability of the system to change the mode of wavelength by itself, depending on the light that is selected. This enhanced the operator experience and ensured a high level of safety. Once the lights and camera was set up the operator did not have to move away from the laptop to operate the entire system during an entire cube capture. Uploading the captured images to the platform was quite challenging due to hardware, network and server security issues, but the constant help provided by the XpectralTEK team helped overcome all issues.

Visualising Artworks: Translating the Invisible into Diagnostic Data

45

Fig. 9. Altar Frontal. Chapel of the Holy Appearance, St. Servaas Basilica. Mural Paintings. TP (VIS). © SRAL

Overall, the SRAL team was impressed with the system - the hardware, the software and the usability. While, the AI processing also requires some adjustment, particularly when dealing with varnished paint surfaces, we are confident it will improve as the reference databases used to train the AI algorithms become more extensive. We believe

46

K. Seymour et al.

Fig. 10. Chapel of the Holy Appearance, St. Servaas Basilica. Detail. Head of Christ. Upper left: AI RGB. Upper Middle TP (UV). Upper Right. AI Mapping Losses. Lower Left: AI Mapping. Cadmium Yellow. Lower Right. AI Mapping Azurite. © SRAL.

that this system will indeed develop to become a powerful tool for researching and diagnosing artworks in preparation for conservation treatments and other related tasks. We hope that in the future the system can be also tested to research microscopes to enable the study of cross-sections. Cross-sections are small sections of paint that are removed from the painting, mounted and prepared for stratigraphic and pigment studies. Using the MSI pigment mapping on these cross-sections could give a clear and accurate identification of individual crystals. Spectral data obtained can be directly compared to reference spectra without the problems that we encountered with pigment mixtures of painted surfaces or miss-attributions due to influence of non-original layers such as varnish coatings. Each grain of pigment can be mapped within the cross-section and the imaging of lower layers can also take place. We hope that this system can be adapted for these purposes in the future.

8 Conclusion While this spectral analysis of pigment data still requires interpretation and confirmation by the end-user, the hardware and growing reference database associated with the software have the potential to be powerful tools for the conservation field. The camera comes with a C-mount lens which could easily be fitted to a research microscope and thus the system has the potential to image and analyse cross-sections. The XpectraTEK system further empowers the conservator with an affordable tool that will aid the quality of the conservation process and document treatment carried out. It will provide a collaborative spectral imaging platform, pooling resources and making interpretations more accurate, through the online cloud-based system the storage of raw data acquired by the XpeCAM X02 will be facilitated and can be used as a reference allowing for powerful analytics and accurate scientific correlations. This system is very user friendly and has high

Visualising Artworks: Translating the Invisible into Diagnostic Data

47

potential for the analysis of artistic materials. Taken as a whole, these data sets represent a practical and successful, non-destructive methodology to study artworks. Combined with other filters available on the XpeCAM Platform and other features provided within the platform the XpectralTEK system will enhance and expedite the conservators’ job. Acknowledgements. The authors wish to thank the SRAL conservation team for their enthusiasm and assistance during the beta testing of the XpectralTEK system. With special thanks to: Angelique Friedrichs and Bascha Stabik.

References 1. Papadakis, V.M., dos Santos, J.: XpeCAM: the complete solution for artwork documentation and analysis. In: Furferi, R., Governi, L., Volpe, Y., Gherardini, F., Seymour, K. (eds.) Florence Heri-Tech 2022. LNME, pp. xx–yy. Springer, Cham (2022) 2. Cosentino, A.: Multispectral imaging and the art expert. Spectrosc. Eur. 27, 6–9 (2015) 3. Measday, D.: A summary of ultra-violet fluorescent materials relevant to conservation. AICCM National Newsletter No 137 March 2017 (2017)

Websites: All accessed 13 January 2022 4. 5. 6. 7. 8. 9. 10. 11. 12.

https://chsopensource.org/multispectral-imaging-msi/ https://www.XpectralTEK.com https://www.xpecamplatform.com https://www.SRAL.nl https://www.uva.nl/en/programmes/masters/conservation-and-restoration-of-cultural-her itage/conservation-and-restoration-of-cultural-heritage.html https://aiccm.org.au/national-news/summary-ultra-violet-fluorescent-materials-relevant-con servation https://chsopensource.org/pigments-checker/ https://www.limburgsmuseum.nl/nl/tentoonstelling/de-vergeten-prinsessen-van-thorn/ http://www.webexhibits.org/pigments/indiv/overview/viridian.html

An Automatic Method for Geometric and Morphological Information Extraction and Archiving of Ceramic Finds Luca Di Angelo1(B) , Paolo Di Stefano1 , Emanuele Guardiani1 , and Anna Eva Morabito2 1

Heritechne Center - University of L’Aquila, Piazzale E. Pontieri, Monteluco di Roio, 67100 L’Aquila, AQ, Italy {luca.diangelo,emanuele.guardiani}@univaq.it 2 Department of Engineering Innovation, University of Salento, S.P. 6, 73100 Lecce, LE, Italy

Abstract. The study of potteries is still today almost entirely performed manually by archaeologists. The primary limits of the traditional approach are lack of repeatability in the results, the time required for the analysis and difficulty in exchanging information between researchers. Taking advantage of the previous research results obtained by the Authors in this field, a fully automated procedure for the analysis and cataloguing of potteries is presented in this paper. The procedure allows performing the geometric and semantic analysis of sherds, starting from their 3D scanned model. The method can also determine a set of meaningful measurements of the analyzed sherds and classify them according to the analysis results. Finally, the results are collected into a public and web-based application, which can be interacted with by interested people.

Keywords: Semantic segmentation Pottery analysis · 3D database

1

· Computer-aided procedure ·

Introduction

The study of potteries is one of the significant interests in archaeology. Ceramic finds are very common in archaeological excavations. Their analysis allows scientists to establish much information about the history, the economy, the art, and other characteristics of the archaeological site where they have been discovered. In the last years, this study is become a challenge not only for archaeologists but also for engineers. The introduction of Reverse Engineering technologies in the field of Cultural Heritage has started a new era in the study of ceramics that is now possible by using computer-aided technologies. So, the development of these procedures is a big challenge for scientists because archaeological finds c The Author(s), under exclusive license to Springer Nature Switzerland AG 2023  R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 48–59, 2023. https://doi.org/10.1007/978-3-031-17594-7_4

An Automatic Method for Geometric and Morphological Information

49

are often unpredictable in shape, dimensions, materials, and corrosion, making it extremely difficult to realise automated procedures for their analysis, classification, and cataloguing. Nevertheless, the necessity of automated methodologies in this area of study is deeply felt between researchers for several reasons. First of all, the study of ceramics is performed almost manually [1]. This makes the process time-consuming but, most of all, low repeatable and highly related to the operator’s subjectivity and expertise. At the same time, the material resulting from the archaeological excavation is often composed of hundreds of pieces, making it impossible to analyse in detail all of them by the traditional process, leading to an accumulation of unstudied elements into the warehouses. Although the use of computer-aided procedures for the analysis of ceramics is experiencing increasing success in the academic community, they have a significant limitation: none of these allows to automatically manage, in the Authors’ knowledge, the entire process characterizing the study of ceramic finds. To make the analysis results useful for the researchers’ community, it is necessary to extract semantic and geometric information from the shard and identify a common standard and an accessible interface for classifying and sharing the results. Thence, this paper will discuss a complete and fully-automated methodology that allows the extraction of semantic and geometric information from potteries, which is digitized by using a 3D scanner. The extracted data are then stored into a self-developed 3D database, allowing powerful queries and easily interacting thanks to a user-friendly web interface.

2

State of Art

According to a previous Authors review [1], an automated procedure for the analysis of potteries needs three main phases: axis identification, evaluation of the dimensions and semantic segmentation. The sherd axis is one of the most critical elements to determine because a wrong evaluation significantly affects the measurement and segmentation processes. In literature, some methods are published differing from the exploited property of axially symmetric surfaces. In [2] the Authors proposed the thickness versor intersection-based method, which is based on the property that the minimum wall thickness line intersects the symmetry axis of the part. The method, compared with the state-of-the-art seems to be the most robust in the axis identification on an object affected by extensive wear, encrustations, and all the typical damage found in a common archaeological sherd. The evaluation of the dimensions of a sherd is mainly based on the evaluation of the sherd profile. Defining this latter is the first step in documenting, classifying, and reconstructing archaeological potteries. Most of the available computer-aided methodologies for evaluating profiles are based on a double step procedure that defines a first attempt profile and a final representative profile. For example, Hlavackova-Schindler et al. [3] defines, as a first attempt profile, a section of the surface with a half-plane passing through the axis. Karasik et al. [4] define the pottery profile by a projection-based method, which results in

50

L. Di Angelo et al.

quite a sensitivity to noise. Kamper et al. [7], differently from the first ones, use the profile directly in correspondence of the maximum height of the pottery. This method, although very fast to be implemented, is very sensitive to surface imperfections. The Authors, in previous research activities [5,6], have introduced an innovative procedure that, firstly, defines the first-attempt profile by evaluating the radii of circles obtained from slice planes. The use of a parabolic approximation smoothes the obtained points. The last step of an automated procedure should concern semantic segmentation. This is, undoubtedly, the most critical point for an automated procedure: archaeologists’ way of studying potteries should be translated in machine instructions; this is made even more complex by the impossibility to associate analytical surfaces to sherds requiring to define a specific set of operative rules. Several existing methodologies [4,7–9] have introduced the use of shape descriptors; these allow a fast implementation of the recognition rules but, on the other side, have no correlation with the activities carried out by archaeologists. To overcome these limitations, in a previous paper, the Authors presented a procedure that, differently from the previously mentioned, bases its working principle on the evaluation of geometric differential properties of the sherd. The method can efficiently distinguish some of the most significant elements recognizable in pottery, such as walls, rim, or fractured surfaces. It has been applied to many test cases, providing results similar to the evaluation of a skilled archaeologist.

3 3.1

The Proposed Methodology Geometrical and Morphological Features Recognition

The here presented methodology allows covering the entire process related to the analysis and cataloguing of potteries by the support of a computer-aided procedure. Figure 1 reports the main steps achievable by the use of the computeraided methodology in the following described.

3D mesh of the potteries Axis identification Features segmentation and recognition Dimensional feature evaluation Data exchange with a 3D informational database Web-based client for visualization and interaction with results

Fig. 1. Main steps of the here presented methodology.

An Automatic Method for Geometric and Morphological Information

51

All the steps in the following described have been implemented in an original Matlab application. The first phase consists of evaluating the symmetry axis of the sherd. This step is fundamental to the correct segmentation and recognition of geometrical and morphological sherd features. In Literature, several methodologies have been presented for determining the symmetry axis of tessellated geometries, but most of them are inadequate for sherds; the incomplete span, corrosion and presence of non-axially symmetric features, typical of potteries, make unreliable the results obtained from the existing methodologies [10]. In this paper, the original and customized circle and line fitting approach has been developed for the evaluation of the symmetry axis α of the sherd. This performs the evaluation by using the widest convex and concave patches of the fragment. The estimated axis α is then used to distinguish the axially-symmetric features of the sherd from the non-axially-symmetric ones. In particular, this evaluation is performed by determining the SHI, defined in [11], at each node of the mesh. To make the segmentation more robust, the distance di between the axis α and the normal is used for the segmentation. SHI index and di are compared with threshold values.

Fig. 2. Description of the geometrical and morphological features recognition process.

According to Fig. 2, an axially-symmetric surface can, in turn, be distinguished by internal wall, external wall, rim and base. To identify these elements,

52

L. Di Angelo et al.

an analysis of the differential geometrical properties of the mesh is performed [5]. The following elements are evaluated to classify the axially-symmetric features of the sherd: – – – – –

Nep : number of recognized extremal parts; Nf s : number of recognized fractured surfaces; Cas : completeness of the aximuthal span; g: topological invariant identifying the number of holes through the object; g  : number of holes of the fractured surface.

The recognized features are also evaluated from the metrological point of view. The sherd profile is identified using an original procedure, whose first step consists of aligning the vertical z-axis of the coordinate system of the mesh with the α axis. The representative sherd profile ρ = f (z) is determined by considering the radius of the circles, centred in α, that approximates the axially-symmetric parts of the mesh. At each step zi , the radius of the profile results from the best-fitting circle with the point set ιi , defined as:

ιi = Pxy (zi ) ∩ M

(1)

where Pxy (zi ) is the horizontal plane passing through zi , while M is the manifold mesh. Due to the effects of noise, the obtained profile requires a smoothing; here, a parabola approximating neighbourhood of (zi , ρi ) with a defined width is used for reducing the effects of noise on the resulting profile. The evaluation of the profile allows to obtain the following dimensional characteristics for each segmented portion of the sherd: – – – – –

thickness; the extreme diameters of the portion; the minimum (maximum) diameter for a concave (convex) surface; the minimum (maximum) diameter of the rim (if applicable); the minimum (maximum) diameter of the base (if applicable);

The presented methodology has been applied to a set of archaeological finds discovered in Alba Fucens, Italy. Figures 3, 4 and 5 show an application of the implemented Matlab procedure for three different typologies of sherd. The first one, in Fig. 3, represents a typical case where neither rim nor base features can be identified. This category of sherd cannot be studied by archaeologists due to the lack of reference elements for identifying the axis. The second case (Fig. 4) represents a sherd with a rim, evidencing the capability of the methodology in recognizing a similar feature that, due to corrosion, may result hard to identify even for a human eye. The last analyzed case (Fig. 5) evidences the possibility of analysing fragments with the base, resulting in the relative associated data, such as the base diameter. It is interesting to underline the high level of automation reached by the method; the only input required by the program is the tessellated geometry in

An Automatic Method for Geometric and Morphological Information 70

53

diameter=f(height) 144

60

142 140

50 0

z

z

40

20

30

40

50

60

70

50

60

70

50

60

70

I derivative

0.2

20

30

10

0.4

60

40

0

0

50

100

-0.2 0

External diameter

20

10

20

30

40

II derivative

0.2

10 0

0 6

7

8

9

-0.2

10

0

Thickness

10

20

30

40

Fig. 3. Resulting data from the Matlab application applied to a not-rim and not-base sherd. 45

diameter=f(height) 90

40 80

35 70 0

5

10

15

30

25

25

30

35

40

45

30

35

40

45

30

35

40

45

0.5

z

z

30

20

I derivative

1

40

20 0

10 20

0

20

40

60

80

-0.5 0

External diameter

5

10

15

20

25

II derivative

15

0.2 0

10

-0.2

5 14

16

Thickness

18

20

-0.4 0

5

10

15

20

25

Fig. 4. Resulting data from the Matlab application applied to a rim sherd.

the form of a .stl file. The three previously reported test cases have been elaborated with the same algorithm configuration. Each elaboration has required about 800 s, using a Macbook Pro 16 (2019), equipped with an Intel i7 2,6 GHz and 16 Gb of RAM. An expert archaeologist has supervised all the results. This processing time can be optimised in the future, considering that the implementation used is in the prototype stage.

54

L. Di Angelo et al. 45

diameter=f(height)

80

40 60

35 40 0

30

40

5

10

15

20

25

30

35

40

45

30

35

40

45

30

35

40

45

I derivative

1

25

z

z

30 20

0.5

10

20

0

20

40

60

0 0

5

10

15

External diameter 15

10

20

25

II derivative

0.5

0

5 6

8

10

12

Thickness

-0.5 0

5

10

15

20

25

Fig. 5. Resulting data from the Matlab application applied to a base sherd.

3.2

Digital Cataloguing

The finds cataloguing has a fundamental role in the study of potteries. It allows scientists to share the results of their analysis; in this way, for example, materials coming from different excavations can be correlated, but also quantitative and qualitative evaluations about an archaeological site can be performed. The introduction of the digitized process has helped the archaeologist in this aim, but the lack of standardized procedures and public databases represents a bottleneck in this sector. This paper introduces a 3D digital database for pottery archiving to pass the previous limitations. The application can receive the data from the previously presented Matlab application, store them into a customized and high-performance database, and interact by a user-end through a web browser. The proposed system is based on web-based technologies interfaced through customized REST web API. In this way, both data upload and data visualization are fully accessible through the internet. Figure 6 shows the component diagram of the application. One of the most interesting aspects concerning the proposed application is represented by the entities that have been identified for managing the pottery information. Most of them are the result of previous authors researches [6,12– 14]. Their relationships permit the treatment of all the aspects related to the geometric and semantic information. For the specific case, two kinds of entity models have been used (cf. Fig. 7): a first one, represented by white rectangles, that allows treating from a general point of view all the aspects related to a 3D discrete geometric model; a second one, represented by red blocks, that is specific for the case of interest and includes the archaeological information.

An Automatic Method for Geometric and Morphological Information

55

Fig. 6. Component diagram of the proposed application.

Fig. 7. UML diagram of the model entities defined into the application.

The data structure described in Sect. 3.1 has been implemented in a custom Python application based on Django, a robust open-source framework for the development of web-based applications. Its powerful Object Relational Mapper

56

L. Di Angelo et al.

Fig. 8. A snapshot of the search page, showing the sherds resulting having a certain range of thickness and belonging to Alba Fucens archaeological site.

(ORM) permits model entities to be easily interfaced with the database. A powerful and open-source semantic search engine, Elasticsearch, has been integrated with the main component Django to apply a large set of queries and filters effortlessly. For example, Elasticsearch provides the tools for searching all the potteries identified in a specific archaeological site and whose thickness is between a set of values provided by the user (cf. Fig. 8). Another possibility is to search for sherds that contains a specific archaeological feature, such as a rim. All the external services of the Applications can interact with it through the REST API. For example, the Matlab application exchanges information with Django through the matlab.net.http package. The application frontend has been developed as a web application based on an open-source WebGL, Potree Viewer. This module has been integrated and redesigned explicitly for the application goals. For example, a customized code has been written for allowing users to select customized object area surfaces. This feature permits to report of new elements of interest on catalogued items; in such a way, the database knowledge can be increased by the user interaction (cf. Fig. 9).

An Automatic Method for Geometric and Morphological Information

57

Fig. 9. A snapshot of the viewer page, where the user has selected a portion of the sherd surface to report a new feature.

Each pottery resulting from a search (cf. Fig. 8) can be displayed in a dedicated viewer page. Here, the user can interact actively with the model: it can navigate, zoom, and activate a set of specific filters. For example, in Fig. 10, a filter for evidencing the internal wall feature of the sherd has been activated.

Fig. 10. A snapshot of the viewer page, where the user has activated the internal wall filter during the visualization of a sherd element.

58

3.3

L. Di Angelo et al.

Conclusions

A procedure for the geometric analysis and cataloguing of potteries has been presented in this paper. This is the result of the Authors’ previous researches, which have been properly reorganized and integrated so that a complete and integrated methodology for ceramics management is finally available. The most important goals reached by the presented procedure are: – a fully-integrated methodology which allows, starting from the 3D mesh of the sherd, the geometric and semantic segmentation of sherds, the evaluation of meaningful measurements and the cataloguing into an interactive and publicly accessible web-based application; – the possibility to analyze sherds in a standardized and objective way; – a universal data format for sharing information related to potteries. Other efforts are required to increase the automation of the process: for example, the 3D scanning process of the sherds is still performed manually by an operator. Moreover, due to the large amount of data that are associated with each 3D object (cf. Fig. 7), more efficient data structures and exchange protocols between the different components of the application should be identified.

References 1. Di Angelo, L., Di Stefano, P., Guardiani, E., Pane, C.: A review on computer-based methods for archeological pottery classification and reconstruction. In: Rizzi, C., Campana, F., Bici, M., Gherardini, F., Ingrassia, T., Cicconi, P. (eds.) ADM 2021. LNME, pp. 909–919. Springer, Cham (2022). https://doi.org/10.1007/978-3-03091234-5 92 2. Di Angelo, L., Di Stefano, P.: Axis estimation of thin-walled axially symmetric solids. Pattern Recognit. Lett. 106, 47–52 (2018). https://doi.org/10.1016/j. patrec.2018.02.022 3. Hlavackova-Schindler, K., Kampel, M., Sablatnig, R.: Fitting of a closed planar curve representing a profile of an archaeological fragment. In: Proceedings VAST 2001 Virtual Reality, Archeology, and Cultural Heritage (2001). https://doi.org/ 10.1145/584993.585034 4. Karasik, A., Smilansky, U.: 3D scanning technology as a standard archaeological tool for pottery analysis: practice and theory. J. Archaeol. Sci. 35(5), 1148–1168 (2008). https://doi.org/10.1016/j.jas.2007.08.008 5. Di Angelo, L., Di Stefano, P., Pane, C.: An automatic method for pottery fragments analysis. Meas. J. Int. Meas. Confed. 128(April), 138–148 (2018). https://doi.org/ 10.1016/j.measurement.2018.06.008 6. Di Di Angelo, L., Stefano, P., Guardiani, E., Pane, C.: Automatic shape feature recognition for ceramic finds. J. Comput. Cult. Herit. 13, 1–21 (2020). https:// doi.org/10.1145/3386730 7. Kampel, M., Sablatnig, R., Mara, H.: Robust 3D reconstruction of archaeological pottery based on concentric circular rills (2005) 8. Cohen, F., Liu, Z., Ezgi, T.: Virtual reconstruction of archeological vessels using expert priors and intrinsic differential geometry information. Comput. Graph. (Pergamon) 37(1–2), 41–53 (2013). https://doi.org/10.1016/j.cag.2012.11.001

An Automatic Method for Geometric and Morphological Information

59

9. Gilboa, A., Tal, A., Shimshoni, I., Kolomenkin, M.: Computer-based, automatic recording and illustration of complex archaeological artifacts. J. Archaeol. Sci. 40(2), 1329–1339 (2013). https://doi.org/10.1016/j.jas.2012.09.018 10. Di Angelo, L., Di Stefano, P., Morabito, A.E.: Comparison of methods for axis detection of high-density acquired axially-symmetric surfaces. Int. J. Interact. Design Manuf. 8(3), 199–208 (2014). https://doi.org/10.1007/s12008-014-0209-4 11. Di Angelo, L., Di Stefano, P.: C1 continuities detection in triangular meshes. CAD Comput. Aided Design 42(9), 828–839 (2010). https://doi.org/10.1016/j.cad.2010. 05.005 12. Di Angelo, L., Di Stefano, P., Morabito, A.E., Pane, C.: Measurement of constant radius geometric features in archaeological pottery. Meas.: J. Int. Meas. Confed. 124(April), 138–146 (2018). https://doi.org/10.1016/j.measurement.2018.04.016 13. Eslami, D., Di Angelo, L., Di Stefano, P., Pane, C.: Review of computer-based methods for archaeological ceramic sherds reconstruction. Virtual Archaeol. Rev. 11(23), 34–49 (2020). https://doi.org/10.4995/var.2020.13134 14. Yifan, L., Gardner, H., Huidong, J., Nianjun, L., Hawkins, R., Farrington, I.: Interactive reconstruction of archaeological fragments in a collaborative environment. In: Proceedings - Digital Image Computing Techniques and Applications: 9th Biennial Conference of the Australian Pattern Recognition Society, DICTA 2007 (2007). https://doi.org/10.1109/DICTA.2007.4426771

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna Francesco Trovatelli1(B) , Francesca Barbagallo2 , Edoardo M. Marino2 Marco Tanganelli1 , and Stefania Viti1(B)

,

1 Dipartimento di Architettura (DiDA), Università di Firenze, Florence, Italy {francesco.trovatelli,marco.tanganelli,stefania.viti}@unifi.it 2 Dipartimento di Ingegneria Civile e Architettura (DICA), Università di Catania, Catania, Italy {francesca.barbagallo,edoardo.marino}@unict.it

Abstract. The assessment of the structural safety of art works requires the availability of proper models to represent their behavior after the expected excitations. Marble artworks, such as sculptures made in Florence during the Renaissance, are especially sensitive to seismic actions, due to their slenderness, weight, and shape irregularity. The reliable modeling of marble artworks, therefore, is very important for their protection. In these years, many experiences have been made aimed at representing the dynamic response of sculptures after seismic excitations. In this work, a high-fidelity Finite Element model has been compared to a new simple one, consisting of concentrated mass placed in the barycenter of the statue and connected to the soil through two “equivalent” trusses. The comparison has been made on the marble statue “Oceanus”, made by Giambologna in 1570. The sculpture, currently exhibited at the courtyard of the Museo del Bargello in Florence, has a mass of about 2 tons, and a height of over three meters, and it is the only giant statue made by Giambologna. The geometrical modeling of the case-study has been made based on a detailed laser-scanner survey, which provided a comprehensive knowledge of its geometry. The paper provides the first results obtained through the proposed simplified model and a more detailed FE representation. Keywords: Seismic performance of art works · Numerical analysis of marble sculptures · Safety assessment of sculptures

1 Introduction Art works represent the core of the artistic and cultural identity of communities, and their value is unanimously recognized [1–4]. Marble sculptures result to be even more vulnerable than other works of art, due to their slenderness, their irregular shape and the fragility of the material [5]. In case of horizontal actions, such as earthquakes, marble sculptures, depending on their setting and their base fixing, may suffer serious damage because of stress increase, overturning or sliding.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 60–69, 2023. https://doi.org/10.1007/978-3-031-17594-7_5

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna

61

The safety of marble sculptures under seismic excitations has been widely investigated in these last years; such assessment requires to represent their response to horizontal actions. The representation of the response of art works, sculptures and statues to horizontal actions have been investigated through different approaches. The simplest methods adopted for these problems are based on the rigid blocks analysis, introduced by Housner in 1963 [6], and successively developed by many other researchers [7–10]. Another popular approach is based on the Finite Element (FE) analysis. This method has been extensively applied to art works [11–14] because can be applied to complex geometrical models, which can be easily recreated through laser scanner survey. The approaches based on the FE analysis can be adopted to perform time-history analyses, and can provide accurate results, taking into account the specific interface behavior between the base of the object and its support. Lately, some further simplified methods have been proposed [15]; some of them combine a simplified representation of the system with the possibility to investigate its dynamic response. In particular, the system can be represented through a truss-model with a concentrated mass, placed at the barycenter of the system. The truss can be symmetric or not, and its mechanical properties are determined based on equivalence conditions [16, 17]. In this work, two models have been adopted to represent the response to seismic excitation of a case study, i.e. the Oceanus by Giambologna, represented in Fig. 1.

Fig. 1. Views of Oceanus.

Oceanus is a marble statue, currently located in the courtyard of the Museum of Bargello in Florence. It is the only giant sculpture made by Giambologna: it has an height equal to 3.31 m, a weight of about 2 tons, and it is made of Carrara’s marble. Its geometry has been represented after a comprehensive laser-scanner survey, which provided a rich and detailed points cloud. The assumed horizontal action has been found on the basis of the seismicity of the area and the type of foundation soil of the Museum. The response of the case-study to the assumed horizontal action has been found by

62

F. Trovatelli et al.

adopting a simplified SDOF model and a more detailed FE representation. The effect of the friction between the base of the sculpture and its pedestal has been also considered.

2 The Case-Study: “Oceanus” by Giambologna The statue of Oceano was commissioned by Cosimo I de’ Medici in 1567. He got a basin granite by Elba Isle to make a giant Fountain in the Boboli gardens. The original plan of the Fountain is not certain; there is a drawn of the project at the Ashmolean Museum of Oxford. In 1572 a plaster model of Oceanus was placed in the basin, and in the same year the marble sculpture was started. Oceanus was completed in 1576, and it was placed in the hexagonal basin (whose diameter was about 7 m) together with three smaller sculptures in the bottom part. Today the Fountain looks very different; in 1618 it was moved from its initial location; it was placed close to Grotta Madama first, and after (in 1637) in the western side of the Garden, “The Garden of the Little Island”. In 1907 the statue was moved to the Bargello Museum – its current location – and it was replaced in the Fountain with a copy made by Raffaello Romanelli. The geometry of the statue has been defined after a comprehensive laser-scanner survey, providing the points cloud which was used to define the geometrical model. In Fig. 2, the main information on the case-study is shown, including the position of the barycenter, with reference to the statue only and to the system made of the statue and the pedestal together.

Z

Statue Material: marble Height: 331 cm Base print (XxY): (76.9 x 77.4) cm Center of Mass (MC) MC position: X = 40.2 cm Y = 26.9 cm Z = 162.2 cm

Z

Y

X

Pedestal Material: concrete Sides (XxYxZ): (90x90x91) cm

Fig. 2. Geometry of the statue and its pedestal.

Oceanus is made of Carrara’s marble; the mechanical properties of the marble have been found after proper experimental investigations [18, 19] made on marble samples taken from the Carrara’s quarries. The compressive and tensile strength have been assumed, respectively equal to 50 and 5 MPa, whilst the Young Modulus has been taken equal to 50000 MPa. A special attention has been paid to the assessment of the friction coefficient (CF), which plays a special role in the dynamic response of sculptures subjected to horizontal actions [7, 18, 20, 21]. The pedestal of the statue was made in the

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna

63

occasion of the last staging of the sculpture. It is made of concrete, but there is not information regarding the fixing between the statue and the pedestal. In this work, therefore, a CF equal to 0.8 has been assumed, in the hypothesis of lack of specific fixing devices, but assuming a layer of mortar between the statue and the pedestal.

3 The Seismic Excitation The seismic excitation has been defined according to the Italian Code NTC 2008 [22], on the basis of the seismic hazard of the area and the foundation soil of the Museum. The Museum area, such as the rest of the Florence basin, consists of plio-pleistocene palustrine and alluvial deposits, followed by two sedimentary cycles related to the paleo-Arno River and the holocene geomorphic evolution. According to the information currently available on the Museum area [23, 24], therefore, the soil has been classified as “class B” according to NTC 2018. The considered seismic intensity refers to a Return Period of 949 years, i.e. the Collapse Prevention limit state, with a class of use (cu ) equal to 2 (strategic buildings). Figure 3 shows the elastic spectrum provided by NTC 2018 for the case-study with those of an ensemble of seven ground motions selected by the database Itaca [25] through the software Rexel [26]. In this work, for sake of brevity, only one ground motion of such ensemble has been used for the analyses. The spectrum of the selected record, represented in blue in Fig. 3, presents a good agreement to that stipulated by the code. 1,4

Acceleration (g)

Name SRCO_HNE

NTC 2018 MEAN

1,2 1,0 0,8 0,6 0,4 0,2 0,0 0

1

2

3

Time (sec)

4

Station Magnitude PGA (g) S.Rocco 6.0 0.250

STR_HNE

Sturno

6.9

0.316

STR_HNN

Sturno

6.9

0.225

TLM1_HNE GMN_HNE

Tolmezzo 6.4 Gemona 5.8

0.315 0.298

GMN_HNN_01 Gemona 6.0 GMN_HNE_01 Gemona 6.0

0.252 0.255

Fig. 3. The selected ground motions.

4 The Numerical Models In this work two different models have been used to represent the dynamic behavior of the statue, respectively represented in Fig. 4: Finite Element model (FE) (see Fig. 4a), developed within ABAQUS [27]; it represents the real geometry of the statue. It has been defined on the basis of the geometrical model, by reducing the number of the “surface” polygons (representing the “skin” of the sculpture), and changing the model from a “surface” to a solid one, i.e. including the volume inscribed in its lateral surface. The mesh consists of 4-nodes tetrahedron

64

F. Trovatelli et al.

elements, and the model has 24230 nodes and 105890 tetrahedrons. When the statue has been considered as simply supported with the contact friction, a thin plate has been introduced, having a high stiffness and a negligible mass, so that it does not affect the dynamic response of the statue. The sliding between the contact surfaces has been limited to linear relationship assumed for the slip, with a tolerance equal to 0.005. Truss Triangular model (TT), developed within OpenSEES platform [28]; it represents the statue as a SDOF system, consisting of three trusses (see Fig. 4b) which connect as many nodes. The first node is located at the center of mass (MC) of the system, and the other two ones are placed, respectively, at the intersection between the cross section of the statue and its support. The mass is assumed as concentrated in the MC node. The trusses are modeled as “elasticBeamColumn”, and the hinges as “zeroLenght” elements, with rotational stiffness equal to zero. The rotational inertia has been neglected. The three trusses have the same cross section, which has been set in order to achieve the same Fundamental Period provided by the modal analysis (performed through the FE model). The TT model is a plane one, and it can represent the dynamic behavior of the system in one horizontal direction only. When the statue has been assumed to be simply supported over the pedestal, two further “flatSliderBearing” elements have been added between the hinges and the soil. They have no tension stiffness and strength along the vertical direction, in order to simulate the one-directional restraint of the support. A Coulomb friction behavior has been assigned to the horizontal cyclic response of the flatSliderBearing elements, which leads the sliding at the overcoming of the friction limit. FE model

T model

Center of Mass (MC)

Fig. 4. The structural models.

In both the models, the damping has been assumed through the Rayleigh model [29, 30], where the needed α and β coefficients have been found by assuming a damping equal to 3%, whilst the first vibrational periods refer to the first two modes of the system along the selected direction.

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna

65

In order to set the α and β coefficients, and to set the TT-model, a preliminary modal analysis has been performed with the FE-model, which provided the vibrational Periods of the case-study. The First mode occurs along the Y-direction; therefore, the Y-direction has been assumed to perform the time-history analysis, and the second Period considered for determining the α and β coefficients has been assumed along such direction (it refers to the 4th vibrational mode of the statue, which is the 2nd mode along the Y-direction). Table 1. First vibrational Periods along the Y-direction and coefficients for the Rayleigh damping matrix Quantities

2nd period*

1st period

α

β

Unit

sec

sec

Hz

sec

Modal response of the statue

0.0741

0.0110

4.42999

0.0000915

2D modal response of the statue

0.0545

0.0083

6.00018

0.000069

Since the TT-model represents the dynamic response of the system along onedirection only, even the one-dimensional dynamic response of the FE model has been investigated, by introducing lateral restraints to the nodes placed along the front side of the statue. In Table 1 the main information regarding the modal analysis is provided, whilst further results provided by the modal analysis can be found in [31].

5 Results The dynamic response of the statue has been checked in terms of displacement of the center of mass. Figure 5 shows the displacement time history of the statue assumed to be fixed at its base, respectively restrained in the Y-Z plane (2D-behavior) or not. The timehistories are plotted in diagrams having different scale, since the dynamic response of the system with lateral restraint is much lower than the one without. It can be noted that the time-histories provided by the two models have a very similar trend; however, the one provided by the TT-model looks shifted compared to the other one; this difference is due to the lack of equivalence between the axial stiffness of the two models; since the system is not symmetrical, indeed, the TT-model presents an initial horizontal displacement due to the gravitational loads. 0,3

0,8

2D FIXED SYSTEM

displacement (mm)

displacement (mm)

FIXED SYSTEM

0,6

0,2 0,1 0,0 -0,1 -0,2 -0,3

TT-model

-0,4 0

1

2

3

4

time (sec)

5

6

7

FE-model 8

9

10

0,4 0,2 0,0 -0,2 -0,4 -0,6

TT-model

-0,8 0

1

2

3

4

5

6

7

8

FE-model 9

10

time (sec)

Fig. 5. Displacement time-histories provided by the considered models for the fixed system.

66

F. Trovatelli et al.

In order to compare the two time-histories without this initial gap, the response provided by the TT-model has been shifted; Fig. 6 shows the comparison between the timehistories after such correction; the comparison is limited to the range of time between 1 and 4 s, where the dynamic response is higher. Figure 7 shows the comparison between the dynamic response of the system found through the two models when the base restraint is assumed to be a simple support with friction between the contact surfaces. It should be noted that removing the fixing restraint at the base, the system experiences higher displacements, since there is some sliding at the base of the statue. In this case, the difference between the two models is higher with respect to the fixed system. 0,4

FIXED SYSTEM

0,6

0,2

displacement (mm)

displacement (mm)

0,8

2D FIXED SYSTEM

0,3

0,1 0,0 -0,1 -0,2 -0,3

FE-model

-0,4 1

2

TT-model

3

0,4 0,2 0,0 -0,2 -0,4 -0,6

FE-model

-0,8 1

4

2

time (sec)

TT-model

3

4

time (sec)

Fig. 6. Details of the displacement time-histories provided by the considered models for the fixed system (after the shift of the TT results).

4

1 0 -1 -2 -3

TT-model

-4 0

1

2

FRICTION SYSTEM

3

2

displacement (mm)

displacement (mm)

4

2D FRICTION SYSTEM

3

3

4

5

time (sec)

6

7

FE-model 8

9

2 1 0 -1 -2 -3

TT-model

-4

10

0

1

2

3

4

5

6

7

FE-model 8

9

10

time (sec)

Fig. 7. Displacement time-histories provided by the considered models for the simply supported system with friction.

The maximum displacement provided by the simplified TT-model does not exceed 1 mm, while the one provided by the FE-model exceeds 3 mm both in the 2D and in the 3D systems. The response of the FE-model, when the 3D system is considered, evidences some peaks between 2 and 3 s, settling around a horizontal displacement equal to 2 mm for a time over 5 s. In order to better understand the dynamic response of the simply supported system, Fig. 8 shows both the horizontal responses provided by the FE model. The plot is focused on the range of time between 1 s and 6 s, when the most significant response occurs. It can be noted that the response of the system along the X-direction plays an important role in its global response. Along the X-direction, indeed, the system experiences a displacement up to 1 mm, changing its fundamental configuration after the first 3 s of vibration.

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna 4

FRICTION SYSTEM (FE-model)

3

displacement (mm)

67

2 1 0 -1 -2 -3

FE, Y displacement

FE, X displacement

-4 1

2

3

4

5

6

time (sec)

Fig. 8. Details of the displacement time-histories along the X and Y directions provided by the FE model for the simply supported system with friction.

6 Final Remarks The paper presents the first results of a numerical investigation made on the marble sculpture Oceanus, currently located in the courtyard of the International Museum of Bargello in Florence. The dynamic response of the system has been checked with reference to a ground motion compatible to the seismic hazard of the area. Two different numerical models, involving different computational effort, have been adopted in the analysis: a FE model, and a simplified Truss Triangular model, consisting of three truss members, where the mass of the system is concentrated in the mass center. The TT model represents the dynamic response of the system in one plane only, and its stiffness is set on the dynamic properties of the system. The dynamic analysis has been performed along one direction only, i.e. the direction of the first vibrational mode of the statue. Two different base restraints have been considered; the statue has been assumed to be fixed and simply supported with the introduction of a proper friction coefficient. Furthermore, the statue has been assumed both as having a spatial and a 2D response, by introducing proper lateral constraints to limit its response in the X-Z plane. When the statue has been assumed as fixed at the base, the two models have provided a quite similar dynamic response, both for the spatial and for the 2D cases. The simplified model provided the same dynamic trend, with displacements slightly higher than the FE model. When the statue has assumed to be simply supported, and the friction has been introduced, the simplified TT-model provided horizontal displacements much smaller than the FE-model, especially for the free (3D) system. In this case, indeed, the dynamic response of the system along the X-direction (orthogonal to the direction of the analysis) has played a crucial role in the spatial response of the system. The study is currently limited to one case-study only, and only one response quantity has been processed; furthermore, the performed analyses refer to one ground motion only. Anyway, the study has evidenced that the simplified TT-model provides a quite satisfactory representation of the dynamic response of the system when a fixed restraint is assumed at the base. When the system is assumed to be simply standing over the

68

F. Trovatelli et al.

pedestal, and the friction is introduced at the base, the response provided by the TTmodel differs considerably from the one provided by the FE model. In order to better understand the properties of each model, the numerical responses should be compared to the results provided by an experimental test. Moreover, further analyses should be done, considering a larger number of response quantities, case studies and ground motions.

References 1. Lowry, M.K., Farrar, B.J., Armendariz, D., Podany, J.: Protecting collections in the J. Paul Getty Museum from earthquake damage. WAAC Newsl. 29, 16–23 (2007) 2. Liberatore, D.: Vulnerabilità dei beni archeologici e degli oggetti esibiti nei musei. CNRGruppo Nazionale per la Difesa dai Terremoti – Roma (2000) 3. Spyrakos, C.C., Maniatakis, C.A., Taflampas, I.M.: Application of predictive models to assess failure of museum artifacts under seismic loads. J. Cult. Herit. 23, 11–21 (2017) 4. Reinhorn, A.M., Viti, S.: Monumental buildings used as museums: protection or danger for the artifacts? Proc. Struct. Integr. 29, 40–47 (2020) 5. Wittich, C.E., Hutchinson, T.C., Wood, R.L., Seracini, M., Kuester, F.: Characterization of full-scale, human-form, culturally important statues: case study. J. Comput. Civ. Eng. 30(3), 05015001 (2016) 6. Housner, G.W.: The behavior of inverted pendulum structures during earthquakes. Bull. Seismol. Soci. Am. 53(2), 403–417 (1963) 7. Monaco, M., Guadagnuolo, M., Gesualdo, A.: The role of friction in the seismic risk mitigation of freestanding art objects. Nat. Hazards 73(2), 389–402 (2014). https://doi.org/10.1007/s11 069-014-1076-9 8. Kounadis, A.N.: New findings in the rocking instability of one and two rigid block systems under ground motion. Meccanica 50(9), 2219–2238 (2015). https://doi.org/10.1007/s11012015-0167-3 9. Bakhtiary, E., Gardoni, P.: Probabilistic seismic demand model and fragility estimates for rocking symmetric blocks. Eng. Struct. 114, 25–34 (2016) 10. Ther, T., Kollár, L.P.: Model for multiblock columns subjected to base excitation. J. Int. Assoc. Earthq. Eng. 47(2), 418–437 (2018). Edited By: Masayoshi Nakashima, Michael Fardis, and Michael C. Constantinou 11. Sorace, S., Terenzi, G.: Seismic performance assessment and base-isolated floor protection of statues exhibited in museum halls. Bull. Earthq. Eng. 13(6), 1873–1892 (2014). https:// doi.org/10.1007/s10518-014-9680-3 12. Bagnéris, M., et al.: A complete methodology for the mechanical diagnosis of statue provided by innovative uses of 3D model. Application to the imperial marble statue of Alba-la-Romaine (France). J. Cult. Herit. 28, 109–116 (2017) 13. Forcellini, D., Giardi, F., Tanganelli, M.: Seismic assessment of the historical third tower in San Marino based on a 3D laser scanner survey (3D-LSS). Innov. Infra-Struct. Solut. 4(1), art. no. 20 (2019) 14. Viti, S., Pintucchi, B., Rotunno, T., Tanganelli, M.: The seismic analysis of Cerere at the Museum of Bargello. Bull. Earthq. Eng. 18(6), 2635–2656 (2020). https://doi.org/10.1007/ s10518-020-00802-6 15. Galassi, S., Ruggieri, N., Tempesta, G.: A novel numerical tool for seismic vulnerability analysis of ruins in archaeological sites. Int. J. Archit. Herit. 14(1), 1–22 (2018). https://doi. org/10.1080/15583058.2018.1492647 16. Cocuzza Avellino, G., et al.: Numerical and experimental response of free-standing art objects subjected to ground motion. Int. J. Archit. Herit. (2021). https://doi.org/10.1080/15583058. 2021.1902019

Modeling Marble Artworks: The Statue “Oceanus” by Giambologna

69

17. Conti, S.: Protezione sismica di opere d’arte: l’uso di basi lubrificate con polvere di grafite e il caso studio della statua di Cerere. Tesi di Laurea in Ingegneria Edile-Architettura, Università di Catania (2020) 18. Domaneschi, M., Tanganelli, M., Viti, S., Cimellaro, G.P.: Developing a laboratory facility to assess friction coefficients of standing samples. Proc. Struct. Integr. 29, 142–148 (2020). ISSN 2452–3216. https://doi.org/10.1016/j.prostr.2020.11.150 19. Tanganelli, M., et al.: Dynamic analysis of artifacts: experimental tests for the validation of numerical models. In: Papadrakakis, M., Fragiadakis, M. (eds.) National Technical University of Athens (NTUA), Proceedings of the 7th International Conference on Computational Methods on Structural Dynamics and Ear-thquake Engineering. vol. 2, pp. 2865–2877 (2019). ISBN 978-618-82844-7-0, Crete, Greece, 24–26 June 2019 20. Viti, S., Tanganelli, M.: Resimus: a research project on the seismic vulnerability of museums’ collections. In: Papadrakakis, M., Fragiadakis, M. (eds.) COMPDYN Proceedings, vol. 2, pp. 2819–2829. National Technical University of Athens (2019). ISBN 978-618-82844-7-0, grc, 2019 21. Domaneschi, M., Tanganelli, M., Viti, S., Cimellaro, G.P.: Vulnerability of art works to blast hazard: the Fountain of Neptune in Florence. In: Papadrakakis, M., Fragiadakis, M. (eds.) COMPDYN 2018 (2021) 22. NTC 2018: Decreto del Ministro delle Infrastrutture 17 gennaio 2018. Aggior-namento delle «Norme tecniche per le costruzioni». Gazzetta Ufficiale della Repubbli-ca Italiana, n. 42 del 20 febbraio 2018, Supplemento Ordinario n. 8 (2018) 23. Sapia, V., Materni, V., Giannattasio, F., Marchetti, M.: Esplorazione geofisica del sottosuolo: primi risultati nel centro storico di Firenze (in Italian). In: RESIMUS: Un progetto rivolto alla vulnerabilità sismica delle opere museale, DIDAPRESS (2018) 24. Coli, M., Rubellini, P.: Geological anamnesis of the Florence area, Italy. Z. Dt. Ges. Geowiss. (German J. Geosci.), 164(4), 581–589 (2013) 25. Itaca: Database of the Italian strong motions data (2008). http://itaca.mi.ingv.it 26. Iervolino, I., Galasso, C., Cosenza, E.: REXEL: computer aided record selection for codebased seismic structural analysis. Bull. Earthq. Eng. 8, 339–362 (2009). https://doi.org/10. 1007/s10518-009-9146-1 27. Hibbit, H.D., Karlsson, B.I., Sorensen, E.P.: ABAQUS user manual, version 6.12 Dassault Systemes Simulia Corp. Rhode Island, USA (2012) 28. Mazzoni, S., McKenna, F., Scott, M.H., Fenves, G.L.: OpenSees command language manual. Pacific Earthquake Engineering Research Center, University of California, Berkeley (2007) 29. Rayleigh, L.: Theory of Sound (Two Volumes), 1954th edn., p. 1877. Dover Publications, New York (1954) 30. Chopra, A.K.: Dynamics of Structures, Theory and Applications to Earthquake Engineering. Prentice Hall, New York (1995) 31. Tanganelli, M., Galassi, S., Viti, S.: Simplified analyses for the model setting of sculptures: the “Oceano” by Giambologna. In: Proceedings of 8th ECCOMAS Thematic Conference on Computational Methods in Structural Dynamics and Earthquake Engineering. Athens, 27–30 June 2021 (2021)

Application of the RestArt System for Stone Statue Reassembly Validated by Shaking Table Testing Martina Pavan1 , Giulia Pompa1 , Pietro Nardelli2 , Silvia Borghini3 , Vincenzo Fioriti4 , Angelo Tatì4 , Alessandro Colucci4 , Massimiliano Baldini4 , Alessandro Picca4 , and Ivan Roselli4(B) 1 Ma.Co.Re’. s.r.l., Rome, Italy 2 Architect, Rome, Italy 3 Museo Nazionale Romano, Rome, Italy 4 ENEA, Rome, Italy

[email protected]

Abstract. An innovative mechatronic-based procedure for high-precision reassembly of stone fragments was applied to restore the ancient roman statue of Diana Cacciatrice (Diana the Huntress), whose fragments were stored at the repository of the Pio Capponi Museum in Terracina, Italy. The RestArt system comprises a high-accuracy 3D laser scanning of two fragments positioned on a special machine specifically designed for handling and accurately move large fragments. Then a software-simulated best-fitting of the two homologous fractured faces of each fragment provides the needed roto-translation matrix, which drives the machine control system to move one fragment to match the other one. Also, the RestArt machine integrates a numeric-controlled moving drilling device for high-precision boring of the fractured surfaces at the designated points for optimal coaxial rods insertion. This permits a very effective fixing of the fragments and allows multi-point fixing, which is practically impossible with conventional methods. The efficacy of the RestArt reassembly method was experimented through shaking table tests on 80-cm-height stone columns specimens. Some specimens were restored by the traditional method, taken as benchmark. All specimens were subjected to strong vibration tests reproducing extreme earthquakes and truck transport on pot-holed road. The RestArt system resulted less time-consuming and capable of providing a reassembly much more resistant to vibration excitation than the traditional method. After such good experimental results, the RestArt system was applied to reunite the head, the legs and the right arm to the main torso fragment of the Diana statue. Keywords: Statue restoration · Large stone fragments reassembly · Mechatronic-based system

1 Introduction The reassembly of stone statues is a complex task for restorers and comprises many aspects and problems not easy to be solved. From one side, problems typical of the © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 70–83, 2023. https://doi.org/10.1007/978-3-031-17594-7_6

Application of the RestArt System for Stone Statue

71

restoration of artworks and historic monuments must be faced and related guidelines must be followed, such as the ones about the historic study of the monument and the respect of the artistic value of the object through material compatibility and reversibility of the intervention [1]. From the other side, big statues pose the seemingly more trivial problem of moving heavy objects avoiding any risk of damage. Any historian who has ever dealt with the issues of moving and handling of big sculptures knows very well how this kind of practical problems have affected the history of the sculpture itself, from the ancient times [2] to the renaissance times [3] and until today [4]. These problems become even more relevant when massive stone fragments have to be moved with extreme accuracy to achieve the delicate task of statue reassembly [5]. In order to provide support to restorers in reassembling fragmented objects, recent computer-aided technologies and advanced 3D virtual reconstruction techniques became more and more relevant. They allow unprecedented visualization opportunities through advanced computing tools for high-precision virtual reassembly of fractured parts [6– 9]. By now, virtual reconstruction methods are essential to support the restorers [10]. Then, their experienced skills usually leads to conclude to job in excellent final results by accurate manual procedures on the real objects. But manual handling of fragments is not always possible, especially when fragments are very large and heavy. In such circumstances, it is necessary to utilize powerful machines to lift and move the fragments with safety and accuracy. To this purpose, modern mechatronics can integrate hightechnological advances of 3D virtual reconstruction tools with powerful high-precision controlled machines [11] able to safely handle the fragments. In the present work, the stone reassembly procedure pointed out in the RestArt project was applied. It integrates a mechatronic-based system, whose machine and accessories were invented and designed by P. Nardelli [12], with advanced 3D scanning and virtual fragments fitting. The RestArt method was experimented on a series of fractured stone specimens representing fragments of marble statues or architectonic elements [13]. Some specimens reassembled by the traditional method were used as a reference. All specimens were tested on shaking table at ENEA Casaccia research center [14] in order to verify the new method efficacy by assessing the restored specimens mechanical resistance to extreme vibration excitations (strong earthquakes and high-speed truck transport). After the above laboratory tests, which proved that the RestArt method achieved remarkable high-quality reassembly of stone fragments that resulted to be much more resistant than the ones by the traditional method, the innovative method was also applied to the restoration of some real statues. One of them was an ancient roman fragmented and incomplete statue representing Diana the Huntress (Diana Cacciatrice in Italian) made up of thassos marble, that was stored at the repository of the Pio Capponi Museum in Terracina, near Rome [15]. In the present paper a detailed description of all steps of the restoration intervention statue of Diana through the RestArt method is provided. The restoration intervention was carried out in collaboration with the Soprintendenza Archeologia Belle Arti e Paesaggio per le Provincie di Latina Frosinone e Rieti and with the Pio Capponi Museum. The RestArt system allowed the safely and accurate reassembly of the fragments, though this statue represented a complicated and difficult

72

M. Pavan et al.

application due to stability problems. The successful restoration of the statue made it possible to exhibit it in its regular vertical position.

2 The RestArt Stone Fragments Reassembly Method The RestArt method was conceived with the purpose to provide a substantial support to restorers in order to facilitate and to improve the quality of the their work in the difficult procedure of reassembly of large stone fragments. It is based on the use of a specifically invented and designed high-precision robotic numerical control machine integrated with advanced 3D scanning instrumentation and dedicated software. The overall procedure allows to largely overcome the usual difficulties related to the fragments handling, fitting and matching. Moreover, it permits the making of accurate coaxial holes with the highest level of safety and with the least sacrifice of original matter. The advanced 3D scanning software is able of fracture interface recognition that allows to simulate the reassembly of fragments in a virtual environment so that potential actual difficulties can be analyzed and foreseen. Besides, the fragments are always moved after fracture interface recognition in the virtual environment whose positions are memorized so that the handling is always machine controlled with the highest repeatability and safety level. The prototype machine was designed to handle large (up to about 2 m long) and heavy (up to about 1 t) stone fragments with high accuracy (in the order of about 0.01 mm). The machine is equipped with two beds, or “cribs”, designed to host one fragment each, so as to treat two fragments at a time. Each crib is mounted on properly designed and motorized slides and rotators in order to control fragments motions in all necessary translation directions and rotation angles, which are measured through accurate displacement and rotation sensors. Moreover, innovative solutions of some of the machine components guarantee high effectiveness, such as the air cooling system of the drill bits, the original dust extraction system, the fracture interface recognition and virtual handling software, the robotic arm of the drilling device and a special drill bits set. Firstly, a preliminary feasibility study is carried out by expert and specifically trained restorers who analyze the fragments in terms of state of conservation and conditions. As the procedure considers two adjacent fragments at a time, the restores must decide the sequence of reassembly couples, which is not always a trivial task and must consider geometrical and logistic aspects. Then, each fragment of the considered adjacent couple is placed on and secured to each machine crib in a proper position. When the fragments are stable, the 3D laser scanning can be carried out to acquire the two fracture interfaces (Fig. 1a), which are processed by the dedicated software to perform a virtual best fit of the fragments. Therefore, the robotic cribs are moved in accordance to the roto-translation matrix provide by the best fitting algorithm in order to physically match the two fragments (Fig. 1b). The two fragments are placed near each other with a safety margin of a few mm in order to avoid contact. Then, one of the fragments is moved very slowly toward the other until safe contact between them in order to verify the correctness of the physical matching and the final matching position is memorized.

Application of the RestArt System for Stone Statue

73

Fig. 1. The RestArt machine: fragments 3D scanning (a) and matching (b).

At this point, one of the machine cribs is moved away as far as is necessary to make room for the drilling device, which is raised from the machine bottom and positioned between the two cribs. The drilling device is controlled through the dedicated software by a high-precision robotic arm. The drill bit is accurately aligned to the cribs so that the holes can be bored perfectly coaxial on both fragments by just turning the drilling device by 180° around the vertical axis. The accurate coaxial drilling allows the fragments reassembly by three-point fixing. Consequently, the fragments interaction at the fracture interfaces is given by the area within the three points, whose indented surface opposing the torsional and slipping forces makes this solution of much greater mechanical resistance than the one with only one-point fixing. The innovative air cooling system for the special drill bits is specifically designed to avoid the overheating of the mechanical parts without using a water-based system, which would be more time consuming, complex and inappropriate for water-sensitive and polychromatic objects that could be damaged. After the drilling of all holes is accomplished, the rods can be inserted inside the holes with minimal loss of original material, which can be saved for micro-puttying as necessary, while the interfaces are glued with minimal quantities of resins. Finally, the two cribs are slowly moved again to the memorized position to put back the two fragments together for the concluding reassembly.

3 Experimental Validation An experimental validation of the RestArt method was carried out by testing six stone specimens. A shaking table was used to test the resistance to vibration of the specimens [16] with particular emphasis on their fragments reassembly. The specimens materials were intended to be representative of common Italian cultural heritage ancient stone works. More specifically, Carrara marble was used to simulate the behavior of a typical stone statue material, while the Roman Travertine was considered as representative of a typical material for ancient architectural elements in ancient Roman context. All specimens had the same dimensions, which were H × L × W 800 × 180 × 180 mm and were fractured in the middle, simulating a random fracture caused.

74

M. Pavan et al.

The specimens were fixed at the base to the shaking table and an eccentric (about 25 cm out of the specimens vertical central axis) load of 300 kg made up of steel plates was added to simulate the effect of jutting upper parts of statues (Fig. 2). The six tested specimens were as follows: • T2R: Roman Travertine specimen reassembled by RestArt method with 1-point fixing; • T3T: Roman Travertine specimen reassembled by traditional method with 1-point fixing; • C1R and C3R: Carrara marble specimens reassembled by RestArt method with 1-point fixing; • C4T: Carrara marble specimen reassembled by traditional method with 1-point fixing; • C2R: Carrara marble specimen reassembled by RestArt method with 3-point fixing. The instrumentation used to measure the motion of the specimens comprised a 3D capture motion system with passive optic markers [17] and accelerometers. The positions and number of used sensors can be seen in the shaking table setup shown in Fig. 2. The markers were located in order to monitor the relative displacements of the fragments during the vibration tests. The acceleration data were processed by Experimental Modal Analysis (EMA) methods in order to extract the specimen fundamental frequency after each test [18]. Subsequently, a global Damage Index D based on the reduction of the fundamental frequency was calculated [19]. Also, sonic testing of the specimens was carried out before and after the shaking table tests to verify the state of damage of the internal material [20]. A further monitoring analysis of the dynamic behaviour of the specimens was obtained through the application of video processing algorithms derived by the motion magnification (MM) technique, able to enhance differences in vibration amplification in different parts of the specimens [21].

Fig. 2. Shaking table setup for Session 1.

The experimental program was subdivided in three sessions characterized by different types of extreme vibration inputs that can cause damage to historical assets. In particular they are as follows:

Application of the RestArt System for Stone Statue

75

• Session 1: specimens T2R and T3T were subject to inputs based on the Amatrice 2016 earthquake record with acceleration scaled from 0.2 to 2.0; • Session 2: specimens C4T and C3R were subject to inputs based on vibration recorded on a 100-km/h-speed wheeled truck running on pot-holed road with acceleration scaled from 0.1 to 1.0; • Session 3: specimens C1R and C2R were subject to inputs based on synthetic earthquakes with 7–20-Hz 30-s-duration and peak ground acceleration (PGA) from 0.2 g to 4.5 g.

Fig. 3. Damage index (D %) of specimens in shaking table tests.

Fig. 4. Inspection of specimen C2R (reassembled with 3-point fixing) after final failure at 4.5 g of PGA in Session 3.

The results of the above tests are illustrated in Fig. 3. The RestArt specimen T2R remained substantially undamaged throughout all the testing session up to FS equal to

76

M. Pavan et al.

2.0. On the contrary, in the traditional specimen T3T a significant damage started arise at FS equal to 1.4 and increased rapidly in the following tests reaching 80% at the end of the session. Similarly, in Session 2 the RestArt specimen C3R showed only negligible damage up to the last test, in which D reached 37%, while C4T appeared more damaged since FS of 0.35 and ended up with almost 70% of D. Finally, session 3 quantified the remarkably increased resistance of the RestArt specimen reassembled with 3-point fixing with respect to the analogous specimen with only 1-point fixing (Fig. 4).

4 Application to the Statue of Diana Cacciatrice In the present section the application of the RestArt system to the restoration of the statue of Diana Cacciatrice at the Pio Capponi Museum in Terracina is described as case study. This restoration intervention permitted the exhibition of the statue in its natural vertical position. The whole intervention was carried out by the Ma.Co.Rè. Staff with the supervision of Pietro Nardelli. The High Surveillance of the restoration was accomplished by the Soprintendenza Archeologia Belle Arti e Paesaggio delle Province di Frosinone, Latina e Rieti. The statue had been discovered in several fragments of various size near Terracina. Then, the fragments were collected and transported from the founding site to the municipal depository (Fig. 5).

Fig. 5. Fragments of the statue of Diana stored at the Pio Capponi Museum before restoration.

Application of the RestArt System for Stone Statue

77

Fig. 6. 3D Laser scanning of the fragments (a). Virtual reconstruction of the torso fragment (b).

Fig. 7. Virtual simulation (a) and manual matching of the head with the torso’s neck (b).

Here the fragments were stored until the restoration intervention was decided. To such purpose the fragments and the RestArt robotic equipment was brought to a dedicated space within the Pio Capponi Museum. The statue presented a very complicated condition because of the high angles of the fracture surfaces and of the compositional anatomy. The stored fragments were four: the main fragment is represented by the torso with both legs to the knee; the right arm (not complete); the head; the base including the right leg from the knee down and an acephalous hunting dog beside. After careful surface cleaning and photographical survey, the 3D laser scanning of each fragment was carried out (Fig. 6). Initially, the head had not been considered belonging to the statue of Diana, but thanks to the virtual recognition by 3D laser scanning of the fragment (Fig. 7a) it was then verified that it perfectly matched the corresponding fracture interface of the torso’s neck (Fig. 7b). Similarly, all other fragments were confirmed as belonging to the statue and a virtual reassembly was simulated (Fig. 8a). The virtual simulation gave the restorers the possibility of a preliminary view of the final result in order to assess its overall consistency.

78

M. Pavan et al.

Moreover, the restores could evaluate the reassembly problematics and formulate proper practical solutions, such as the choice of optimized fixing rods’ positioning and angles in order to avoid possible risks for material integrity. Through the observation of the 3D virtual reconstruction of the fragments it could be possible to analyze the fracture interface surfaces, which were all under remarkable angles with the fragments longitudinal axis. Generally, this is considered a very critical condition for the application of traditional methods. A very critical condition that could be pointed out through the analysis of the 3D virtual reassembly simulation was represented by the fact that the statue appeared to have a quiet eccentric mass distribution loading more on the missing left leg. Such critical posture did not make possible to obtain the vertical stable positioning of the statue by only putting the available original four fragments back together in their correct place. Consequently, an additional support system made up of steel framework was designed to link the left knee to the left heel to compensate the missing leg and allow a balanced load transfer to the base. Another criticality was given by the right-arm fragment reassembly to the torso because of the combination of several factors, such as the relatively small available interface area, the small overall fragment volume, the specific position in the statue and the almost vertical orientation of the interface main plane. The above situations usually pose very difficult problems if using the traditional reassembly methods. Instead, the RestArt method allowed to approach these problems with an easier and safer solution. Accordingly, a choice of the most convenient sequence of fragments couples to be treated was made. Then, the first two fragments are mounted and fastened to the RestArt machine for physical matching. The same was done with the other fragments two at a time. As an example, in Fig. 8b the main torso and the right-leg fragments coupling is shown.

Fig. 8. Virtual reassembly of the fragments (a) and physical matching by RestArt machine of the leg fragment with the main torso fragment (b).

Application of the RestArt System for Stone Statue

79

The virtual reassembly and the physical matching of the right-arm with the torso can be seen in Fig. 9a and in Fig. 9b respectively. Both the right-leg and the right-arm fragments were attached to the main torso fragment by 3-point fixing, as is shown in Fig. 10. In particular, fiberglass and stainless steel rods were considered for joining the fragments (Fig. 10b). Rods number, diameter, length and position at the interface were specifically designed in order to support the fragments mass and thickness. The rods had a maximum diameter of 10 mm. After the restorers selected on the 3D virtual reconstruction of the fragments interface the most convenient points to be bored, the software calculated and sent the corresponding 3D coordinates of the homologous points on both fragments interfaces (Fig. 10a) to the robotic system so as to accurately control the drilling device, always under the supervision of the restorers.

Fig. 9. Virtual reassembly (a) and physical matching (b) of the right-arm fragment with the main fragment.

Fig. 10. Accurate virtual positioning of the drilling device for three-point fixing reassembly of the right-arm fragment (a) and physical insertion of the rods in the numerical-controlled drilled holes of the main fragment (b).

80

M. Pavan et al.

Drill device robotic arm

Dust extraction

a)

b)

Fig. 11. Positioning of the high-precision drill device (a). Drilling of a hole (b).

Fig. 12. The RestArt machine with the final reassembly of the statue of Diana (on the right).

Subsequently, the drilling of the holes was carried out using a proper drill bit size in relation to the chosen rod thickness and length (Fig. 11a). The marble dust produced by the drilling was carefully extracted (Fig. 11b) and saved to be reused for the micro-puttying needed to arrange the inevitable imperfections at the outer edges of the fragments joints. Epoxy resin was used for final gluing of the interfaces, after protection with acrylic resin in solution. Particular care was necessary for the fixing of the right-leg to the torso, the two fragments with the highest volume and mass. Moreover, from a structural point of view this fixing was the most important for the global stability of the statue. Here a 3-pointfixing was performed utilizing three stainless steel threaded rods with a diameter of 10 mm.

Application of the RestArt System for Stone Statue

81

The geometric accuracy of the coaxial drills guaranteed the minimal sacrifice of original matter, which anyway was recycled for the micro-puttying finishes. Furthermore, the extreme geometric accuracy facilitates also the functionality and the reversibility of the intervention, as it makes much more easy to insert and remove the rods. In substance, the criterion of minimum intervention, in the face of greatest effectiveness, was fully respected and accomplished. The restoration intervention was concluded by the finishing operations and the final cleaning of all parts, after surface consolidation and application of protection products. The final result is shown in Fig. 12.

5 Conclusions In the case study here presented the restoration through the innovative RestArt system of the statue of Diana Cacciatrice at the Pio Capponi Museum in Terracina is illustrated. The innovative mechatronic-based RestArt method was specifically developed for high-precision reassembly of stone fragments in the restoration of ancient statues and architectural elements. Before the actual application to a real ancient statue the method was experimented on stone specimens made up of Carrara marble and Roman Travertine. The experimentation demonstrated the remarkable effectiveness of the method in terms of reassembly precision and mechanical properties. It permitted a very resistant fixing of the parts, especially in the case of 3-point-fixing, even in case of extremely strong vibrations, much stronger than traditional reassembled specimens could withstand. Moreover, the whole procedure resulted much less time consuming and also easier for restorers than the traditional method. The proposed method allowed the virtual recognition of the fragments, with particular reference to the fracture interfaces, and provided the restorers with a virtual simulation of the reassembly intervention based on the integration of accurate 3D laser scanning and best fitting algorithms. Most importantly, this provided an objective tool to verify the belonging of the head fragment to the statue, contrarily to what was previously thought. This phase was only the starting point of an complete restoration procedure for supporting the optimal choice and the reliable planning of each operation of the following phases. Among the most innovative aspects of the proposed method are the possibility to easily identify the homologous points of the drilled holes at the fracture interface of the other fragment and the possibility to perform coaxial holes with extreme accuracy. This is made possible by the integration of high-precision 3D scanning and high-level virtual simulation software with a mechatronic numeric-controlled machine specifically invented and designed by P. Nardelli for large stone fragments reassembly. Other relevant innovations of the equipment used in the RestArt method comprise the air cooling system of the special drill bits and the original dust extraction system, which guarantee high effectiveness in the drilling operation and significantly contribute to solve practical and logistic problems posed by complex restoration interventions to traditional methods. However, it is important to underline that in the proposed method the technological innovation must be seen as a substantial support to the restorers, but in no case it is

82

M. Pavan et al.

meant to replace the human control, which is assured through the supervision of expert professionals with proper expertise and sensibility, especially needed in the case of delicate and complicated restoration interventions. The whole restoration was concluded with a very satisfactory final result and gave the museum the possibility to exhibit the statue in its natural vertical posture. Most remarkably, the method allowed the restorers to accomplish the reassembly intervention of the statue with an easier, safer and more accurate procedure and in a quite shorter timing than the traditional methods would have permitted. Also the sacrifice of original material and the reversibility of the intervention were noticeably optimized in comparison to more conventional techniques. Acknowledgements. The validation method through shaking table testing was developed within the Research Project “RIPARA Integrated systems for the seismic retrofitting of architectural heritage” (2021–2023) funded by Lazio Region within the Technological District for New Technologies applied to Cultural Heritage (DTC) Programme (Grant no. 305-2020-35586, 25/05/2020). A special thanks for collaboration to Soprintendenza Archeologia Belle Arti Paesaggio delle Province di Frosinone e Latina and the municipality of Terracina. A distinct note of merit is due to the Laborarorio di Restauro Grandi Marmi of the Museo Nazionale Romano for having believed and supported the development of the RestArt technology from the beginning and to the Istituto Superiore per la Conservazione ed il Restauro (ISCR), since both institutions collaborated to the present experimentation. The statue reassembly system was developed within the RestArt project, co-financed by EU and Regione Lazio (call POR FESR Lazio 2014–2020, Kets Tecnologie Abilitanti). The RestArt system was invented, designed and patented by architect Pietro Nardelli. Ma.Co.Re’ s.r.l. retains ownership of the RestArt machine and all its accessories, including 3D scanning system.

References 1. Petzet, M.: Principles of preservation: an introduction to the international charters for conservation and restoration 40 years after the Venice charter. In: International Charters for Conservation and Restoration. Monuments & Sites, ICOMOS, Munich, pp. 7–29 (2004) 2. National Geographic website. https://www.nationalgeographic.com/culture/article/120622easter-island-statues-moved-hunt-lipo-science-rocked. Accessed 07 Jan 2022 3. Levine, S.: The location of Michelangelo’s David: the meeting of January 25, 1504. Art Bull. 56(1), 31–49 (1974) 4. Elnady, T.: Vibration measurements during the relocation of Ramses II statue. In: Turkish Acoustical Society - 36th International Congress and Exhibition on Noise Control Engineering, INTER-NOISE 2007, Istanbul, vol. 4, pp. 2392–2400 (2007) 5. TKSST website. https://thekidshouldseethis.com/post/installing-massive-statues-engine ering-at-the-metropolitan-museum-of-art. Accessed 07 Jan 2022 6. Papaioannou, G., et al.: From reassembly to object completion - a complete systems pipeline. ACM J. Comput. Cult. Herit. 10(2), Article 8 (2017) 7. Jo, Y.H., Hong, S., Jo, S.Y., Kwon, Y.M.: Noncontact restoration of missing parts of stone Buddha statue based on three-dimensional virtual modeling and assembly simulation. Herit. Sci. 8(1), 1–12 (2020). https://doi.org/10.1186/s40494-020-00450-8 8. Rasheed, N.A., Nordin, M.J.: Reconstruction algorithm for archaeological fragments using slope features. ETRI J. 42(3), 420–432 (2020)

Application of the RestArt System for Stone Statue

83

9. Gherardini, F., Santachiara, M., Leali, F.: 3D virtual reconstruction and augmented reality visualization of damaged stone sculptures. In: IOP Conference on Series: Materials Science and Engineering, vol. 364, p. 012004 (2018) 10. Arbacea, L., et al.: Innovative uses of 3D digital technologies to assist the restoration of a fragmented terracotta statue. J. Cult. Herit. 14(4), 332–345 (2013) 11. Jäger, G., Manske, E., Hausotte, T., Mastylo, R., Dorozhovets, N., Hofmann, N.: Precision mechatronics based on high-precision measuring and positioning systems and machines. In: Proceedings of SPIE 6616 Optical Measurement Systems for Industrial Inspection V, 661–621 (2007) 12. Nardelli, P.: Apparato per la rilevazione, misurazione e movimentazione per ricomporre sezioni di oggetti fratturati e relativa tecnica di utilizzo dello stesso. Italian patent application n. 102021000001973, 1 February 2021 13. Nardelli, P., et al.: RestArt method: innovative high-precision mechatronic-based procedure for fragmented stone statues restoration. In: 2021 International Conference on Metrology for Archaeology and Cultural Heritage (MetroArchaeo 2021), Journal of Physics: Conference Series (JPCS). Milan (2022) 14. Mongelli, M., et al.: Experimental tests of reinforced concrete buildings and ENEA DySCo virtual laboratory. In: 5th International Conference on Structural Health Monitoring of Intelligent Infrastructure (SHMII-5), Turin, 9 p. (2011) 15. Pio Capponi Museum website (in Italian). https://www.beniculturali.it/luogo/museo-pio-cap poni. Accessed 07 Jan 2022 16. Rossi, M., Calderini, C., Roselli, I., Mongelli, M., De Canio, G., Lagomarsino, S.: Seismic analysis of a masonry cross vault through shaking table tests: the case study of the Dey Mosque in Algiers. Earthq. Struct. 18(1), 57–72 (2020) 17. De Canio, G., et al.: Passive 3D motion optical data in shaking table tests of a SRG-reinforced masonry wall. Earthq. Struct. 40(1), 53–71 (2016) 18. Mongelli, M., Roselli, I., De Canio, G., Ambrosino, F.: Quasi real-time FEM calibration by 3D displacement measurements of large shaking table tests using HPC resources. Adv. Eng. Softw. 120, 14–25 (2018) 19. DiPasquale, E., Ju, J.W., Askar, A., Çakmak, A.S.: Relation between global damage indices and local stiffness degradation. J. Struct. Eng. 116(5), 1440–1456 (1990) 20. Polimeno, M.R., Roselli, I., Luprano, V., Mongelli, M., Tatì, A., De Canio, G.: A nondestructive testing methodology for damage assessment of reinforced concrete buildings after seismic events. Eng. Struct. 163, 122–136 (2018) 21. Fioriti, V., Roselli, I., Tatì, A., Romano, R., De Canio, G.: Motion magnification analysis for structural monitoring of ancient constructions. Measurement 129, 375–380 (2018)

A BIM-Based Model for Heritage Conservation and Structural Diagnostics: The City Walls of Pisa Anna De Falco1 , Francesca Gaglio2 , Francesca Giuliani1(B) , and Massimiliano Martino1,2 1

Department of Civil and Industrial Engineering (DICI), University of Pisa, Largo Lucio Lazzarino, 56126 Pisa, Italy [email protected], {francesca.giuliani,m.martino}@ing.unipi.it 2 Department of Energy, Systems, Territory and Construction Engineering (DESTeC), University of Pisa, Largo Lucio Lazzarino, 56126 Pisa, Italy https://www.destec.unipi.it/, https://www.dici.unipi.it/

Abstract. The use of Historical Building Information Modeling (HBIM) is gaining much interest in the architectural heritage domain due to its ability to support the design and management of conservation activities. Any maintenance, preservation or revitalization strategy requires obtaining a complete knowledge of the site and conducting critical investigations on geometrical, physical, and documental data. The benefits of implementing HBIM comprise the possibility to integrate the data coming from multiple sources, inspections and diagnosis techniques, as well as to employ standardized and robust tools for orienting cultural heritage asset management. This study explores the challenges of developing HBIM on ancient city walls, whose great dimension and extension require adapting the conventional workflow in order to obtain results in a reasonable time. The combination of traditional geometric surveys and more innovative techniques allows for a complete and extensive photogrammetric documentation of the city walls. The acquisition process has been speeded up without compromising the accuracy of the resulting model, thus offering a reliable representation of multiple issues of the historical assets, ranging from its features and state of conservation to its structural deficiencies. By applying the novel workflow to the city walls of Pisa, the paper discusses the interoperability among different tools and the broad versatility of the proposal for large architectural heritage.

Keywords: HBIM conservation

1

· City walls · Cultural heritage · Preventive

Introduction

Planned preventive conservation entails the protection of cultural heritage and is founded on the attentive identification of situations of risk and the systematic c The Author(s), under exclusive license to Springer Nature Switzerland AG 2023  R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 84–96, 2023. https://doi.org/10.1007/978-3-031-17594-7_7

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

85

planning of minimally invasive interventions [1]. This approach ensures greater efficiency in terms of costs and results with respect to unplanned systems based only on corrective actions [2], which may not be able to remove the causal factors of disasters. Conversely, planned preventive conservation is a proactive management process to avoid unnecessary deterioration, damage and even failure by means of periodical monitoring, scheduled maintenance and integral condition assessment. A specific challenge in this field is to develop easy methods to establish a virtuous management process for large architectural heritage assets, such as ancient city walls, which require great effort to accurately document and represent the current state of conservation as well as to handle interventions avoiding economic burden. City walls played a key role in defending people and places from foreign invaders and have been significant for shaping local identities since the remotest period. Today, they hold great potential as cultural resources, as they are large preserved iconic attributes of civic pride and identity. However, the effective implementation of the planned preventive conservation of such a large number of monumental constructions spread across a vast territory is a challenging task and is still under-researched. Many failure events have recently occurred in diverse sites and countries [3,4] stressing the need to ensure systematic, condition and significance-based conservation, repair and maintenance of historical structures. The conception and implementation of novel tools for managing the conservation process of city walls may increase the operative capacity by fostering the transfer and sharing of data between different professionals, from archaeologists to engineers, and governmental actors, from regional to local ones. Significant contribution may derive from the application of Heritage Asset Management (HAM) principles, which promote a multi-disciplinary, knowledge-based decision making driven by comprehensive and up-to-date data, with explicit leadership and responsibilities [5]. However, there is still no standard procedure or framework for HAM, even though some advancement are being experimented by applying Building Information Modeling (BIM) [6]. In the case of historical buildings, the literature refers to Heritage BIM (HBIM) [7] as an effective approach to pursue the modelling of architectural elements and to handle a great amount of data coming from multiple sources. The benefits of HBIM are manifold, ranging from the documentation of the current state to the assessment and monitoring of intervention. The case study presented in this paper concerns a large sector of the city walls of Pisa, a well-preserved fortified system located in Tuscany (Italy). The walls are a remarkable example of the Medieval construction techniques and present peculiar characteristics, with slender curtain walls, high-rise towers, wide gates, and bastions. The creation of a decision support system for the city walls is based on the multidisciplinary and cross-scalar knowledge of the assets that are investigated by applying a HBIM methodology. As such, this contribution gives special emphasis to the survey and modelling phases by experimenting a novel scan-to-BIM approach. The proposed workflow is capable of ensuring a rapid,

86

A. De Falco et al.

yet accurate and reliable, result by speeding up the data acquisition and the 3D model creation and information.

2

Related Works

In the last decades, the research progress in the field of 2D and 3D digital data allowed scholars and heritage experts to rely on accurate models to document the consistency and conservation state of cultural assets. With respect to 2D media, 3D models have the great advantage of providing a comprehensive and metrically correct representation of the building geometry. They can be obtained by using solid modelling techniques, which can be directly or parametrically handled by the user according to the time and precision of the expected results, or by direct triangulation of point-cloud data, which instead are influenced by the quality of the point-cloud itself. By and large, multiple techniques have been developed for range-based (e.g., laser scanner) and image-based (e.g., photogrammetry) reconstructions of architectural objects, and the selection of the most suitable approach is usually conducted on a case by case criterion. Indeed, diverse modelling techniques are usually employed in parallel because none of them is individually able to ensure high geometric accuracy, portability, full automation, photo-realism and low cost at once, together with flexibility and efficiency in the surveys [8]. Once 3D models are created, several pieces and types of information can be registered on it, thus creating enriched or ontological models [9]. This task entails a preliminary documentation of the heritage assets and a retrospective analysis of several sources. Data rich 3D models can be included in interfaces and platforms that also collect reports, analog/digital pictures, drawings, and archival documents [10], thus producing digital libraries to organize, access and manage content. Several studies have also added a temporal dimension (diachronic and synchronic) to the 3D architectural model [11,12] in order to describe the evolution of the building during its life cycle. Few studies explored how the use of 3D models and web-based digital platforms can support restoration activities [13], also documenting the project operations and advancements. Meanwhile, BIM has become popular in the civil engineering sector thanks to its ability to produce and manage structured digital information and multidisciplinary design skills. The process is supported by several software tools - continuously improved for reaching higher quality, lower errors and cost reduction that ensure a coordinated, consistent and always up-to-date working model. In the architectural heritage domain, the advent of advanced surveying techniques is leading to the development of reality-based HBIM methodologies, that rely on as-built/as-is models. The models depict the actual state of the building and can be semantically enriched with a wide range of data. Despite these advances, critical aspects include the elaboration of data from different sources, such as material properties, historical stratifications, and damage patterns, which may inform conservation practices more effectively if coupled with 3D representations. Currently, several BIM platforms have are being used

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

87

by experts to perform the modelling, visualization, compilation, and management of the knowledge of architectural heritage. However, the interoperability among the available tools within a scan-to-BIM process and the lack of historical parametric object libraries are still debated topics.

3

The City Walls of Pisa

The city walls of Pisa were built in Medieval times, starting from the XII century, to defend the central settlement of the historical Maritime Republic. The process of extension and consolidation of the domain over the territory, as well as the continuous conflicts with nearby cities, led to the design and construction of advanced defensive systems that are considered as monuments of inventive genius and skill. Although they have progressively lost the protective function, the walls have been preserved and only limited sectors of the perimeter have been lost over the centuries. Today, the overall length of the circuit is approximately 7 km, the average height of the curtain walls is 11 m, and the thickness is about 2 m. The structure is made of multi-leaf masonry, with two stone outer-leaves with localized brick portions and the inner core of irregular masonry. It is built using local techniques, workmanship, and materials, particularly the sedimentary stone “Breccia” from Asciano, the Limestone from San Giuliano, or a yellowish calcarenite from Livorno (classified as “Panchina” Fm. Calcarenite) [16,17]. The circuit of the city walls has been classified according to the morphological type of element, thus distinguishing the curtain walls, gates, towers, bastions, and fortresses. Each element has been identified with an alphanumeric code that refers to the typology associated with a sequential numbering. A summary of the adopted classification is in Table 1. As shown in Fig. 1, the sectors investigated in this study correspond to: the curtains labelled as CUR.01, CUR.02, CUR.03, CUR.04, and CUR.05; the gates coded as GATE.01, GATE.02 and GATE.03 (i.e., Porta Nuova, Porta del Leone, and Porta Santa Maria, respectively), and the towers labelled as TOW.01, TOW.02 and TOW.03 (i.e., Torre del Catallo, Torre del Leone, and Torre Santa Maria, respectively). Table 1. Codification system for the city walls of Pisa Element type Curtain wall Gate

Tower Bastion Fortress

Code

CUR

GATE TOW BAS

FOR

Numbering

1–49

1–15

1

1–11

1–3

88

A. De Falco et al.

Fig. 1. City walls of Pisa. At the top, general view and classification of the walls; at the bottom, two emblematic sectors of the walled system.

3.1

Data Acquisition

The geometrical survey is conducted through photogrammetric techniques combined with topographic measurements to derive the 3D coordinates of several targets and control points onto the architectural heritage. The application of Structure from Motion (SfM) photogrammetry allows for reconstructing the 3D geometry of the scene starting from a rapid and less demanding data acquisition phase. This method is able to automatically and iteratively solve the positioning and orientation of cameras from a set of many overlapping images, without the need to provide a priori a network of targets. Redundancy is a key requirement in the acquisition phase [14], therefore photographs shall be been taken from a wide array of positions and in uniform lighting conditions. In the case of city walls, some issues may arise in practice owing to the limited accessibility and the great dimension of the object, which

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

89

force the surveyor to rely on long-distance shootings. The great distance between the camera and the feature of interest decreases the spatial resolution of the photographs, which in turn produces low-density point clouds. Positional data (x, y, z) are used as control points to scale the 3D reconstruction of the scene to an absolute coordinate system, as well as to manually adjust any misaligned camera. Coordinates can be collected with several range-based methods, such as a laser scanners or more traditional surveying instruments. In this study, the survey employed a Total Station that provides high-quality and precise coordinates of a series of control points. Besides, the survey exploits remote technologies like Global Positioning Systems (GPS) in the shooting of photographs in order to improve the alignment and orientation of cameras. The result is a scaled point cloud of the investigated sector of the city walls (Fig. 2), which is later used to create the 3D model. It is interesting to note that these control points are also used in the processing phase (see Sect. 3.2) to set the projection planes of orthomosaics and annotation maps. Photographs were acquired using an iPhone 11 camera having a resolution of 12MP and a 1/2.55-in. sensor, with GPS on. The distance from the wall ranged between 7 m and 12 m depending on the accessibility, which was often limited due to presence of trees, roadways, and fences. Photographs were taken in longitudinal strips ensuring a high degree of overlap (about the 70% on average), with two bottom-to-top shots wherever the shooting distance was too small to acquire the whole wall height. Additional photos of the walls have been taken from the wall walk to capture the upper parts with a sufficiently good resolution. The acquisition was done over several days at different times, to have the most uniform illumination possible, avoiding too-strong direct light and hard shadows. Photographs and positional data have been processed using the SfM software Agisoft Metashape [15]. The dense cloud of the investigated sector of the city walls is obtained from an overall number of photographs equal to 348 over a total length of approximately 400 m.

Fig. 2. Perspective view and details of the city walls of Pisa processed with SfM photogrammetry.

90

3.2

A. De Falco et al.

Data Processing and 3D Modelling

The 3D architectural model has been generated starting from the dense point cloud, after a process of refinement that entails filtering out areas with poor confidence and reducing noise generated by any reprojection error. The model derives from the automatic triangulation of the dense point cloud in Agisoft Metashape, which produces a 3D textured mesh of the investigated sectors (Fig. 1). Figure 3 shows the 3D reconstruction process of a portion of the walls, comparing the rectified real photographs (a) with the dense point cloud (b), the model in solid visualization (c), and the tiled textured model (d). The mesh and the texture preserve a good level of detail, compatible with the large dimension of the object under investigation, as they are able to reproduce the geometric irregularities of the city walls. The model has been used to generate 7 orthomosaics, namely high resolution imagery obtained from the projection of the reconstructed object over a reference surface, which can be used as a basis for the diagnosis and the annotation of wall properties. Given the approximately planar geometry of city walls, the projection surface is a plane that has been defined by assigning three Marker-Points in Agisoft Metashape. The model can be exported in three-dimensional format (e.g., obj wavefront, fbx, 3ds, ...) that guarantees the direct association of the processed texture even in a different software. In this study, the final tiled model textured has been imported in the McNeel Rhinoceros software [18] that is able to easily manage complex geometries and hosts Grasshopper, an algorithmic modeling environment supporting a wide range of useful plug-ins. The scan-to-BIM procedure herein proposed employs the Grasshopper plug-in Archicad Live Connection developed by Graphisoft, which enables the connection with the Graphisoft Archicad software [19]. Specifically, the Grasshopper-Archicad Live Connection makes it simple to generate native Archicad BIM elements in Grasshopper using familiar nodes, automating and simplifying common and complex processes. Once the model of the city wall sectors has been imported in Rhinoceros as a continuous textured mesh, the geometry has been cut to clearly distinguish the elements according to the proposed classification system (Table 1). Further operations have been performed to aid the information of the model with data that are typically annotated on 2D media and orthoimages, such as deterioration patterns (e.g., decay, degradation, damage) and materials [3,16]. All these pieces of information have been included into an “annotation plane” (Fig. 4), which can be correctly positioned in the 3D space using the same Marker-Points previously defined in Agisoft Metashape for generating the orthomosaic. Each annotation plane contains closed 2D entities that lie on that plane and that identify and delimit the different features of the architectural heritage under investigation. At the same time it is important to prepare a file on Graphisoft Archicad that will host the three-dimensional entities and its customized set of BIM features, through the creation of specific properties regarding historical, material, structural and morphological data. This set of features will constitute the semantic structure for the final HBIM model.

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

91

Fig. 3. 3D reconstruction process and level of detail on a portion of the CUR.04: (a) rectified real photograph; (b) dense point cloud; (c) model in solid visualization; (d) tiled model textured.

3.3

Model Information

The main advantage of HBIM lies in the possibility to integrate into a single model multiple type of data from diverse documentary sources, inspections and diagnosis techniques. In fact, the survey is not limited to the geometry of the city walls, but also concerns the building technology and its deterioration patterns. Site inspections can supplement historical and archive research, thus providing direct critical knowledge to experts. The diverse materials and deterioration patterns are singled out and annotated on 2D representations of the asset, by drawing the polygons or regions over the representation itself. Typically, this operation is time-consuming since it consists in the manual drawing of the areas, however recent research efforts explored the use of human-centred AI-powered tools for the semantic segmentation of 2D orthographic images thanks to the software TagLab developed by

Fig. 4. Annotation planes for the curtain wall CUR.04

92 A. De Falco et al.

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

93

the Visual Computing Lab [20]. As for the city walls of Pisa, the orthomosaics have been generated on Agisoft Metashape and later annotated by using this novel software that can significantly reduce the processing time and optimize the semantic segmentation [16]. As previously anticipated, the semantic structure of the HBIM has been set up in Archicad and finalized through the use of the Options/Property Manager, which allows for creating new specific properties that will be punctually associated to the 3D model. In this case study, the set of New Properties and New Group Properties has been defined as in Table 2. Table 2. Customized groups and properties set for the HBIM of the city walls of Pisa.

New group property

New properties

HBIM identity data Image; Comments; Tag; Element code; Periodization of historical transformations; Historical documentation; Documentation of the current state; Stratigraphic analysis; Masonry typology; Wall type; Intervention; Period of intervention; Maintenance plan; Criticality of the context or asset HBIM design parameters related to masonry types

3.4

Representation; Block - composition; Block - surface processing; Block - representative size (hxbxt); Block shape; Block - color; Joints - mortar composition; Joints morphology and surface finish; Joints - color; Joints horizontal thickness representative value (t); Joints vertical thickness representative value (t); Texture recourses; Texture - joints offset; Plaster; Section - core composition; Section - core type; Section - connection between core and block; Section - Total thickness

Connection Between the 3D Model and the Information

The mesh model imported in McNeel Rhinoceros is “splitted” with the Grasshopper Mesh Split component using the closed 2D entities previously defined in the annotation plane, referring to materials or decays (Fig. 4). This operation will produce new sub-meshes associated with different properties. These meshes are connected to the Grasshopper-Archicad Live Connection component Morph-solid (in this case study the morph entity was chosen), in which it is possible to connect relative Morph Settings. Such settings are also referred to Archicad software settings and can therefore be defined in the Grasshopper environment. Among these settings, it is possible to load all the new customized properties of the “HBIM identity data” and of the “HBIM-Design parameters related to masonry types” (Table 2), which were defined in the Archicad file, and thus proceed with the assignment and compilation of the information related to the elements of the architectural heritage, directly in the Grasshopper environment (Fig. 5).

94

A. De Falco et al.

At the end of this procedure, the Synchronize tool in Grasshopper will import the mesh and all the related HBIM information in the Archicad file. In Archicad environment it is possible to manage the information related to the historical artifact through the setting of Graphic Override Combinations to which Graphic Override Rules are associated. The latter links typed graphic views according to the assigned information (Fig. 6).

Fig. 5. HBIM in Archicad. The selection in green is the object representing the gate Porta del Leone, coded as GATE.02.

Fig. 6. Typed graphic view of the masonry classes set into the HBIM for the curtain wall CUR.04.

A BIM-Based Model for Heritage Conservation and Structural Diagnostics

4

95

Conclusions

The procedural pipeline herein presented is designed for efficiently obtaining asbuild 3D models that retain information on external dimensions, deformations and surface materials of city walls and, more generally, large architectural heritage. It allows for documenting and recording the characteristic of the building over a complete and sufficiently accurate model, which can be used to track and share data among heritage professionals. Textured 3D models can effectively support the diagnostic process by conveying data over time, enabling critical temporal evaluations and ensuring the revision of any piece of information by any actor involved in the conservation. As such, the proposed framework fully complies with the principles of HAM as it promotes knowledge-based decision making driven by comprehensive and updated data. The proposal applies a scan-to-BIM approach that performs a fully automated 3D reconstruction of the scene from oriented images, thus limiting manual operations that are time-consuming and prone to errors. The case study herein presented demonstrates the proactive and efficient interaction between the 3D mesh file produced by Agisoft Metashape and other files post-processed by other software, such as McNeel Rhinoceros, Grasshopper and Graphisoft Archicad. The workflow stands out for the speed of importing complex mesh elements in the Graphisoft Archicad BIM environment, as well as for the possibility of compiling HBIM information in the Grasshopper environment using familiar nodes system. To conclude, the procedural pipeline has the advantage of indicating a scan-to-BIM procedure that is fast, lean and easily modifiable and customizable. Besides, it ensures a rapid exchange of up-to-date information and the interoperability among different tools that are already employed by professionals in the fields.

References 1. Della Torre, S.: Italian perspective on the planned preventive conservation of architectural heritage. Front. Archit. Res. 10, 108–116 (2020). https://doi.org/10.1016/ j.foar.2020.07.008 2. Vandesande, A., Van Balen, K.: Preventive conservation applied to built heritage: a working definition and influencing factors. In: Innovative Built Heritage Models, vol. 3, pp. 63–72. CRC Press/Balkema (Taylor & Francis Group) (2018) 3. De Falco, A., Giuliani, F., Ladiana, D., Rjolli, L., Bordo, D., Di Sivo, M.: Typological characterization of ancient town walls for disaster prevention and mitigation. The Mo. MU project. In: Roca, P., Pel` a, L., Molins, C. (eds.) Proceedings of the 12th International Conference on Structural Analysis of Historical Constructions, SAHC 2020, 29 September–1 October 2021 (2021) 4. Chen, G., Li, L., Li, G.M., Pei, X.J.: Failure modes classification and failure mechanism research of ancient city wall. Environ. Earth Sci. 76(23), 1–15 (2017). https:// doi.org/10.1007/s12665-017-7150-3 5. Hull, J., Ewart, I.J.: Conservation data parameters for BIM-enabled heritage asset management. Autom. Constr. 119, 103333 (2020)

96

A. De Falco et al.

6. Piaia, E., et al.: BIM-based cultural heritage asset management tool. Innovative solution to orient the preservation and valorization of historic buildings. Int. J. Archit. Herit. 15(6), 897–920 (2021). https://doi.org/10.1080/15583058.2020. 1734686 7. Dore, C., Murphy, M.: Integration of historic building information modeling (HBIM) and 3D GIS for recording and managing cultural heritage sites. In: 2012 18th International Conference on Virtual Systems and Multimedia, pp. 369–376. IEEE (2012) 8. Remondino, F., El-Hakim, S.: Image-based 3D modelling: a review. Photogram. Rec. 21(115), 269–291 (2006) 9. Messaoudi, T., V`eron, P., De Luca, L.: An ontological model for the reality-based 3D annotation of heritage building conservation state. J. Cult. Herit. 29, 100–112 (2018). https://doi.org/10.1016/j.culher.2017.05.017 10. Stefani, C., Busayarat, C., Lombardo, J., De Luca, L., Veron, P.: A web platform for the consultation of spatialized and semantically enriched iconographic sources on cultural heritage buildings. J. Comput. Cult. Herit. (JOCCH) 6(3), 1–17 (2013). https://doi.org/10.1145/2499931.2499934 11. Apollonio, F.I.: The production of 3D digital archives and the methodologies for digitally supporting research in architectural and urban cultural heritage. In: M¨ unster, S., Friedrichs, K., Niebling, F., Seidel-Grzesinska, A. (eds.) UHDL/DECH -2017. CCIS, vol. 817, pp. 139–158. Springer, Cham (2018). https://doi.org/10. 1007/978-3-319-76992-9 9 12. Bruno, S., Musicco, A., Fatiguso, F., Dell’Osso, G.R.: The role of 4D historic building information modelling and management in the analysis of constructive evolution and decay condition within the refurbishment process. Int. J. Archit. Herit. 15(9), 1250–1266 (2021). https://doi.org/10.1080/15583058.2019.1668494 13. Apollonio, F.I., et al.: A 3D-centered information system for the documentation of a complex restoration intervention. J. Cult. Herit. 29, 89–99 (2018) 14. Westoby, M.J., Brasington, J., Glasser, N.F., Hambrey, M.J., Reynolds, J.M.: ‘Structure-from-motion’ photogrammetry: a low-cost, effective tool for geoscience applications. Geomorphology 179, 300–314 (2012) 15. Agisoft Metashape (2021). http://www.agisoft.com/ 16. Pavoni, G., et al.: Another brick in the wall: improving the assisted semantic segmentation of masonry walls. In: Spagnuolo, M., Melero, F.J. (eds.) Eurographics Workshop on Graphics and Cultural Heritage. The Eurographics Association (2020) 17. Raneri, S., Pancani, D., De Falco, A., Montevecchi, N., Gioncada, A: material characterisation for preserving cultural heritage: evidence of the 1595 fire at Pisa cathedral. Stud. Conserv. 1–13 (2021). https://doi.org/10.1080/00393630.2021.1898886 18. McNeel Rhinoceros (2020). https://www.rhino3d.com/it/ 19. Graphisoft Archicad (2021). https://graphisoft.com/it/solutions/archicad 20. TagLab (2021). https://github.com/cnr-isti-vclab/TagLab

An Innovative Method for Dimensioning the Crossbeams of an Original Painted Panel, Based on Mechanical Testing, and on Numerically Modelling Its Distortion Tendency Lorenzo Riparbelli1 , Ciro Castelli2 , Giovanni Gualdani1,2 , Luciano Ricciardi2 , Andrea Santacesaria2 , Luca Uzielli1 , and Paola Mazzanti1(B) 1 DAGRI, University of Florence, Florence, Italy

[email protected] 2 Opificio delle Pietre Dure, Florence, Italy

Abstract. Most panel paintings feature mechanical systems, typically made of crossbeams, designed to control the distortion of the panel. When the crossbeams are too stiff or too yielding in relation to the distortion tendency of the panel, significant damage to the artwork may occur. The evaluation of the factors causing the damage necessarily relies on the restorer’s experience and wisdom; however, in the framework of a long-standing cooperation between Opificio delle Pietre Dure (Florence) and the Research Group on Wood Technology of DAGRI Department (University of Florence), an innovative method has been developed to provide additional objective criteria for dimensioning the crossbeams system. A specific protocol was developed, based on an engineering approach and consisting of the following five “modules”: a) a Restoration Plan, developed by restorers and conservators, b) appropriate mechanical tests to evaluate the stiffness of the painted panel alone and of the crossbeams alone, c) hygroscopic tests to evaluate the hygro-mechanical behaviour of the painted panel alone, d) appropriate numerical modelling for dimensioning the panel-crossbeams system, and e) numerical and/or experimental verification of the dimensioning. This paper describes such protocol and its first implementation on an original panel painting, the Adorazione del Bambino e Committente (in brief called Nativity) attributed to Cesare da Sesto (1514–1515). Keywords: Panel paintings · Crossbeams · Crossbeam-induced damage · Engineering for artworks restoration · Restoration protocol · Numerical modeling of artworks · Mechanical testing of wooden artworks

1 Introduction The panel paintings are complex artworks, because of the variety of constituting materials and construction features [1–3]. They were widespread between the 12th and 16th © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 97–112, 2023. https://doi.org/10.1007/978-3-031-17594-7_8

98

L. Riparbelli et al.

century, when wood was the most common support for painting, until it was progressively replaced by canvas. Typically, the wooden support of a panel painting is made of one or more boards glued together along the edges. On the front face (and occasionally also on the back face) several layers are applied (typically formed by hide glue, gesso, sometime canvas, paint layers, protective varnish), which for the sake of brevity, from now on are collectively named paint layers; all these materials having peculiar mechanical and hygroscopic behaviour [4]. Also, a restraining system is usually present, made by crossbeams or frames, or sometime by a combination of both. Many kinds of paint layers and of restraining systems can be found, having very different composition, structure, properties, and conservation requirements. A large proportion of conservation problems of panel paintings depends mainly on the following factors, which cannot be ignored if a comprehensive analysis is to be carried out [5]: (1) biological attacks, (2) ageing of the materials making up the paint layers, and (3) the mechanical relationships between crossbeams and panel (also depending on the environmental conditions to which the artwork is exposed). The focus of this paper is on point (3). It is well-known that the panel paintings are sensitive to the relative humidity (RH) changes, since they are made of wood, a hygroscopic anisotropic material. Every time the RH changes, the wooden support reacts in the short and longer period by modifying its dimensions and its shape, so these changes affect the paint layers as well. The short period distortions are typically caused by the buildup of moisture gradients across the panel thickness and are recovered once the gradients disappear [6, 7]. Other deformative mechanisms are involved in the longer period, including the distortions induced by growth rings orientation, the compression set shrinkage, the ageing of the materials, or the stiffening action of the paint layers [6, 8]. These mechanisms and their combinations are at the basis of the permanent cupping of the panel paintings, and of many of the damages that can affect both the paint layers and the panel itself. The restraining system had (and still has) the function of partly restraining such distortions, so to prevent permanent damages to the wood and to other materials, and deterioration of the aesthetics of the artwork; in many cases it also has the function of facilitating the safe handling of the painted panel. Sometimes the restraining system (be it the original one, and/or successive replacements or modifications) is too stiff in relation to the environment in which the artwork is located, and hence might produce damages on the wooden panel, such as open joints or fractures, or on the paint layers. Of course, the risk of damage also depends greatly on the original quality of the paint layers; in some cases, paint layers are so strong that they can overcome the problems caused by excessive wood movements. These are among the reasons why changes on the restraining system may be needed; such interventions are entrusted to the restorers of the wooden supports, who must identify the causes of the problems and decide if and how the existing crossbeams should be modified. Their work necessarily combines operational skills, reasoning, and experience; however, this paper introduces a scientific method that might provide additional objective criteria to support their choices.

An Innovative Method for Dimensioning the Crossbeams

99

A long-standing, close cooperation between the Opificio delle Pietre Dure (OPD) restorers of wooden supports and the Wood Technology research group from DAGRI (University of Florence) led among other to develop an innovative engineering-based method for dimensioning the crossbeams, which was for the first time applied on an original panel painting, as described in this paper. This method consists of a protocol based on scientific analyses, whose information are processed by a specifically developed numerical model to obtain the rational dimensioning of the crossbeams. On the one hand, this method should be as general as possible, so that it can be applied to many works of art, even if they differ in terms of construction techniques, conservation history, and species and characteristics of the wood from which they are made; on the other hand, it must be able to provide specific results, useful for each individual artwork to which it is applied. To reconcile these opposing needs, the method is made up of a set of logical blocks, each of which makes it possible to consider the hygro-mechanical behaviour of the individual artwork, and the conservation requirements of its paint layers.

2 Materials and Methods 2.1 The Original Panel Painting: The Nativity by Cesare da Sesto The Nativity (Adorazione del Bambino e Committente, 1513–1520, attributed to Cesare da Sesto) has been under restoration at the OPD [9] in the years 2016–2017. It is a large painted panel (see Fig. 1) having dimensions 213 × 137 × 3.5 cm, made of three Poplar (Populus alba L.) boards, two of them plainsawn and the third one centre sawn. Generally, the boards are of mediocre quality, affected by knots, ring-shake, and pith on the centre sawn board; in addition areas of insect attack are present. The restraining system is made of three Chestnut (Castanea sativa Mill.) crossbeams nailed on the support. The top and bottom ones are 11 cm wide, and the central one is 9 cm wide; all three were originally 5 cm thick, with significant amount of wane.

Fig. 1. The Nativity front and back faces: (a) and (b) after, (c) before the restoration.

When the painting arrived in the OPD Laboratory, raking light observation revealed the presence of numerous blisters, several colour losses, and diffused surface wrinkling,

100

L. Riparbelli et al.

all indicating an enduring damaging action produced on the paint layers by the distortion of the wooden support (see Fig. 1c). During restoration work carried out some 10 years earlier, the boards had been connected with glued wedges [5], most of which were now unglued or had caused cracks in the vicinity, due to large stresses clearly produced by the impeded cupping; this was a definite sign of ongoing damage, whose cause was identified by the restorers as the excessive stiffness of the crossbeams (that clearly had already been noticed during previous restorations, since the crossbeams were found slightly thinned by rough axe-work). In addition, surface unevenness was present at some disconnected joints. The mediocre quality of the original wood and the insect attacks on the panel had also contributed to its deteriorating situation. Thus, the restorers decided that simply reconnecting the boards would have been insufficient, and that a new dimensioning of the crossbeams was necessary. 2.2 The OPD Guidelines for the Interventions on Wooden Supports of Panel Paintings The restoration methods of the OPD follow general guidelines, and in practice no strict rules are prescribed; in very short summary, the interventions on the wooden supports are based on the following general principles [5]: 1. 2. 3. 4. 5. 6. 7.

Appropriate elastic behaviour of the crossbeams. Reversible connection between crossbeams and panel. Freedom for the panel to deform in its own plane. “Controlled freedom” for the panel to distort. Forces between crossbeams and panel can be adjusted at any time. Panel and crossbeam profiles matching at time of intervention. Adhesion restored between the boards forming the panel, so that it can behave monolithically.

The study presented in this paper is based on all these points, and on the assumption that they cannot be dissociated from each other. 2.3 The Foundations of the Method The Five Modules of the Protocol. The innovative method, applied for the first time on the occasion of the restoration of the Nativity, combines the skills of the restorers with an engineering approach, including a FEM numerical simulation of the panel to explore its distortion behaviour. It consists of the following five modules: a) a Restoration Plan, developed by the artwork’s restorers and conservators, b) appropriate measurements to reliably estimate the stiffness of the painted panel and of the crossbeams, c) hygroscopic tests to identify the hygro-mechanical behaviour of the painted panel alone, and hence derive its hygroscopic parameters, d) FEM numerical hygro-mechanical modeling of the panel-crossbeams system, to dimension the new crossbeams, and e) numerical and/or experimental verification of the new panel-crossbeams system. Making the Panel Monolithic. Firstly, all open joints or cracks in the wooden panel need to be re-connected, to obtain a monolithic mechanical behaviour of the whole panel.

An Innovative Method for Dimensioning the Crossbeams

101

The Experimental Tests. Ideally, once the structural integrity of the panel is achieved, mechanical and hygroscopic tests should be carried out to provide the following main data: a) the bending stiffness of the panel and of the crossbeams, and b) the shrinkage/swelling coefficients and the diffusion coefficient of the wood forming the panel. These data are necessary to provide reliable inputs for the numerical model. If for any technical, organizational, or safety reasons, not all the experimental tests can be performed, some literature data can replace the mechanical and/or hygroscopic coefficients obtained through testing; however, since the literature data are inherently generic and non-specific, the results obtained will be less and less accurate and less representative of the particular artwork. The Restraining Index RE. To quantify objectively and unambiguously the amount of warp distortion prevented by any crossbeams system a new parameter is proposed, named Restraining Index (RE), defined as follows: RE =

f − r 100 f

(1)

where: f is the maximum deflection of the panel free from any restraining system, when subjected to a given variation of Relative Humidity (RH); r is the maximum deflection of the same panel restrained by a specific restraining system, when subjected to the same RH. In other words, RE expresses quantitatively the panel’s deflection reduction produced by the restraining system, as compared to the panel’s deflection without any restraining system, under the same RH variation. Since RE is expressed as a percentage value, for a panel without any restraining system RE = 0%, whereas for a panel of which the hygroscopic distortions are totally blocked by the restraining system RE = 100%. Please note that RE is not entirely independent of the conditions and the hygroscopic variation considered; however, for the purposes of the project described here, it has provided useful results.

2.4 The Restoration Plan, Its Main Decisions and Steps Before any restoration intervention, a specific Restoration Plan [9] for the individual artwork is needed. For the Nativity, the conflict between the panel and the too stiff restraining system was identified by the OPD restorers as the main cause of its deterioration; thus, the following technical and technological choices were initially defined for the remedial intervention on the crossbeams system: – maintaining as much of the constituent wood as possible was considered essential to preserve the historical value of the original crossbeams; – the new crossbeams therefore had to be obtained through sawing lengthwise the original ones in thin slices, then packing and gluing together groups of such slices to form laminated battens, finally stacking such battens to obtain a global stiffness determined by their number and thicknesses (see Sect. 3.4);

102

L. Riparbelli et al.

– metal springs were to be introduced in the original nailed connections between panel and crossbeams, to also make the nails elastically yielding [5]; – deciding the optimum RE for the new crossbeams was of fundamental importance; an RE value ranging between 50% and 60% was selected by the restorers, based on their experience and expertise; – the operations were to be carried out in the following sequence: restore the adhesion of the detaching paint and remove the old over-paintings – separate crossbeams from panel (following which the panel’s distortion increased by ~2 cm, showing the magnitude of the still existing internal stresses) – restore the panel’s monolithic structure – carry out the mechanical tests on panel and crossbeams – design and obtain the new crossbeams from the original ones – modify the nails to allow for springs – refinish the wooden surfaces and reassemble the whole system. 2.5 The Mechanical Tests The mechanical tests (obviously to be carried out with all the precautions for the artwork’s safety) are required by this protocol to provide specific information about the bending stiffness of the original crossbeams and of the panel. Bending Tests on the Crossbeams. The crossbeams were tested according to the threepoint bending method, following as much as possible European standards such as UNI EN 408 [10], although some modifications to the geometry of the load application points were needed. Each crossbeam was loaded, in succession, with dead loads applied at the center of the distance between the supports and at two additional points located approximately halfway on its left and on its right. The load was applied in successive steps of about 50 N, up to about 400 N. Three deflections were measured by means of three digital gauges (Mitutoyo Absolute ID-U1025, resolution 0,01 mm, accuracy 0,02 mm) placed under the loading points (see Fig. 2).

Support

1

2

3

a

b

c

Support

Deflection measurement points

Fig. 2. The geometrical scheme of the bending tests on the crossbeams. The dashed arrows on the top represent the load applied in succession (progressively from zero to 400 N) in each of the three locations (1, 2, 3); the two lateral arrows under the beam represent the constraint reaction forces exerted by the supports. The three symbols under the beam (a, b, c) represent the digital gauges used to measure its vertical deflections in the three corresponding locations.

An Innovative Method for Dimensioning the Crossbeams

103

Bending Test on the Panel. For the bending test, the panel was placed horizontally with the painted face upwards and was supported near the four corners by pinned supports, so it was free to distort. The supports were sliding, and adjustable in height to allow an approximately equal distribution of its weight when the panel was placed on them. The load, consisting of metal plates carefully stacked on top of each other, placed on Melinex® sheet and foam blocks to protect the paint layers, was progressively applied on the geometrical center of the panel. The maximum load of 110 N, applied in eleven steps, was decided together with the restorers, after a preliminary simulation confirmed that the maximum stress levels would be quite small and safe for this specific artwork. The deflections were measured by means of four digital gauges (Mitutoyo Absolute IDU1025, resolution 0,01 mm, accuracy 0,02 mm) mounted on fixed supports (see Fig. 3a). Three gauges (a, b, c) were aligned at mid height of the panel, each at the center of one of the boards, while the fourth one (d) was placed near the top of the panel, at the center of the central board; the readings of the four gauges provided data for computing both the transversal and the longitudinal deflections.

120

c

3b

d

b

100

Load [N]

80

a 60

Load application point

40

3a

c d

20

b a

Deflection measurement points

Supports 0 0

0.2

0.4

0.6

0.8

1

1.2

Deflection [mm]

Fig. 3. Geometrical scheme (a) and resulting load-deflection diagram (b) of the bending tests carried out on the wooden panel, free from crossbeams. The arrow on the top represents the applied load (in successive steps of about 10 N, up to about 110 N), the four ones near the corners represent the constraint reaction forces exerted by the pinned sliding supports. The symbols under the panel (a, b, c, d) represent the digital gauges used to measure its vertical deflections in the corresponding locations (see the main text). The results are commented in Sect. 3.1.

104

L. Riparbelli et al.

2.6 Estimating the Hygroscopic Parameters Hygroscopic tests would have been needed to identify the actual distortion behaviour of the panel when subjected to well-known RH variations, and hence derive realistic hygroscopic parameters of the specific panel [11–15]. Unfortunately, the long time required to perform hygroscopic tests was not available in the case of the Nativity. Thus, literature data for the hygroscopic characterization of the panel had to be used, as described in Sect. 3.2. 2.7 The Numerical Modelling for Dimensioning the Crossbeams The numerical modelling was carried out to achieve two distinct objectives: i) derive the elastic parameters from the mechanical test data, and ii) evaluate through hygromechanical simulation the distortion behaviour of the panel when equipped with crossbeams having various stiffness values. From ii) the stiffness of the crossbeams providing the desired RE could be identified, and hence the new crossbeams could be dimensioned. Deriving the Material’s Elastic Parameters from the Mechanical Test Data. The inverse identification process was carried out through a Nelder-Mead [16, 17] optimization algorithm on the elastic orthotropic compliance matrix expressed as in (2), where the minimization process of the functional was applied to the multipliers a and b (a, b ∈ R+ ). This characterization was applied both on the panel and on each of the three crossbeams. ⎞ ⎛ √ √ 0 0 0 aS abS abS 0 0 0 11 12 13 ⎜ √abS 0 bS 0 √bbS 0 0 0 0 ⎟ ⎟ ⎜ √ 12 √ 22 23 ⎟ ⎜ 0 0 0 0 0 ⎟ ⎜ abS13 bbS23 bS33 √ 0 ∗ (2) Sij = ⎜ ⎟ 0 ⎜ 0 bbS44 √ 0 0 ⎟ 0 0 ⎟ ⎜ 0 ⎠ ⎝ 0 abS55 0 0 0 √ 0 0 abS66 0 0 0 0 0 where Sij∗ is the optimized compliance tensor and Sij0 the components of the compliance matrix estimated based on the actual mean density of ~450 kg/m3 , following Guitard [18]; the 1-direction represents the longitudinal direction, the 2-direction and the 3direction represent two generic transversal directions, without making any difference 0 = S 0 ). between the radial and tangential stiffnesses (S22 33 Simulating the Hygro-Mechanical Distortion Behaviour of the Panel. For the simulation of moisture diffusion in wood, a simplified approach consisting of an isotropic version of Fick’s theory was used [19]. Following the approach described in [20] we define the moisture flow as: qm = −ρ0 · D · ∇mc

(3)

where ρ 0 is the wood density in dry conditions, mc the moisture content and D the tensor _

of diffusion coefficients that in our case has the following form: ⎡ ⎤ D0 D = ⎣ D0 ⎦ D0

(4)

An Innovative Method for Dimensioning the Crossbeams

105

Finally, the time-dependent form of Fick’s law is:

∂mc = ∇ · D · ∇mc . ∂t

(5)

The mechanical chaining is handled automatically by the open-source finite element solver code_aster [21], by projecting the nodal values of moisture content from the hygroscopic first-order mesh to the mechanical second-order mesh. The mechanical simulation is carried out in a linear orthotropic elastic field in small strains; the wood contacts between the panel and each individual crossbeam were treated using a Lagrange formulation stabilized with Coulomb’s friction (0.4 coefficient was used). Simplifications. Please note that in the above simulations some simplifications have been made, including the following: a) the ground layers and the paint layers are assumed to be totally impervious to moisture, so their stiffness is negligible, and b) the emissivity effect of the bare wood surface on the back of the panel has been neglected. These hypotheses might match the actual characteristics of the panel and are discussed in literature [22–24]; however, they appear to overestimate the maximum distortions the panel undergoes during the moisture equilibration transient [14] and are therefore acceptable for the purpose of evaluating its structural safety.

3 Results and Discussions 3.1 The Results from the Mechanical Testing Panel. The data resulting from the bending test on the panel are shown in Fig. 3b. Through an optimization algorithm, the experimental results have been made to coincide with the corresponding simulated results to reliably evaluate the longitudinal and transversal MOE of the panel (see Table 1). On the load-deflection diagram, the circles represent the experimental data; the straight fitted lines (numbered like the corresponding points), confirm the linearity of the results. Please note that all the measured deflections derive from the combination of both longitudinal and transversal distortions of the panel (i.e., bowing and cupping). In fact, the deflection at point (2) is approximately equal to the sum of those at points (1) (3) (whose deflection is almost exclusively caused by bowing) and at point (4) (whose deflection is almost exclusively caused by cupping).

106

L. Riparbelli et al. 500

c b a

Load [N]

400

a c

a

b

c b

300 200 100

a

0 0

1

b

2 Inflection [mm]

c 3

a 4 0

1

b

2 Inflection [mm]

a

c 3

4

0

1

b

2 Inflection [mm]

c 3

4

Fig. 4. Bending behaviour of the central crossbeam loaded in succession at three points. When the load is applied at midspan (point b) the deflections are larger, due to the larger bending moment. When the load is applied at lateral points (a and c), the bending behaviour is not perfectly symmetrical, possibly because of the crossbeam’s taper.

Crossbeams. The bending tests on the crossbeams produced nine load-deflection curves (see Fig. 4), one for each combination of load application point and deflection measurement points. The same procedure used to process the experimental data of the panel was applied to these data to estimate the MOE of the crossbeams (see Table 1).

Table 1. The MOE (Modulus of Elasticity) values of the panel and of the individual crossbeams, estimated by means of the numerical procedure explained in Sect. 2.7. The appropriateness of using them in the models presented in this paper is ensured by the fact that they have been derived from tests carried out on the specific panel and on its crossbeams. MOE [MPa] Panel, longitudinal

9376

Panel, transversal

428

Upper crossbeam

11229

Central crossbeam

5821

Lower crossbeam

9227

3.2 The Hygroscopic Properties Due to limited time available the hygroscopic properties could not be derived from experimental tests; therefore, the following data from literature [14, 25–27] were used: a) two extreme values for the shrinkage/swelling coefficient: α1 = 0.09%/% and α2 = 0.3%/% (maximum and minimum values for a generic transversal direction) b) the shrinkage/swelling coefficient in the longitudinal direction was derived from the above values, reducing them by one order of magnitude c) the value of the diffusion coefficient was assumed as D0 = 10–4 mm2 /s. These parameters were used to simulate through finite element modelling the diffusion of moisture along the panel’s thickness and the resulting distortions, as the consequences of a humidity variation from RH = 55% to RH = 35%, typically corresponding to a variation of MC (Moisture Content of wood) of about 2.6% according to [28].

An Innovative Method for Dimensioning the Crossbeams

107

3.3 The Hygro-Mechanical Model The hygro-mechanical model developed – fed with the mechanical parameters derived from the experimental tests (Sect. 3.1), and the hygroscopic parameters derived from literature (Sect. 3.2) – made it possible to reliably estimate the evolution over time of the maximum cupping deflection of the free panel which, according to the protocol described in this paper, is the most important control parameter. Considering the above mentioned great natural variability of the wood’s hygroscopic parameters and the impossibility of carrying out hygroscopic tests on the studied panel to accurately estimate its specific ones, the decision was made to explore the variation interval of the panel’s deflections depending on the value of the α parameter. For this purpose, the two extreme values derived from the literature, α1 = 0.09%/% and α2 = 0.3%/%, have been entered in the hygro-mechanical model, and hence the maximum and minimum estimated values of the deflection have been computed. The two curves shown in Fig. 5 represent the maximum and minimum deflection values over time. 0

5

10

Time [days]

15

20

25

0

-10

Deflection [mm]

-20

α2=0.09%/%

-30

-40

-50

α1=0.3%/%

-60

Fig. 5. Two simulations of the variation over time of the cupping deflection of the studied wooden panel, obtained with the two extreme values for the shrinkage/swelling coefficient (α1 = 0.09%/% and α2 = 0.3%/%) derived from the literature (see Sect. 3.2). The two peak values of deflection are quite different (–16.7 mm and –56 mm), and hence confirm the fundamental role of such coefficient in determining the entity of the deflection peak during the transient phase before reaching hygroscopic equilibrium.

3.4 The Design of the Crossbeams As mentioned in Sect. 2.4, the Restoration Plan specified that the original crossbeams were to be sawn lengthwise into thin slices, which were then to be packed, maintaining their original order, into groups glued together to form laminated battens of specified thickness. The new crossbeams were to be formed by stacking on top of each other two or more of these laminated battens without gluing, so that their overall thickness would

108

L. Riparbelli et al.

equal the thickness of the original crossbars minus that of all the saw cuts required to reduce them to slices; their global stiffness would therefore be the sum of the individual stiffnesses of such battens. To design the composite cross-sections of the new crossbeams, simplified dimensioning methods based on linear models were used, as a pre-dimensioning and general assessment of the design. Various crossbeam configurations were taken into consideration, by virtually subdividing and recombining each one of them, and then applying the simplified model to simulate the corresponding deflection of the restored and reassembled wooden support exposed to a same RH variation RH = 20%. Among the many ones virtually tested, configurations A and B are shown in Fig. 6. Configuration A consisted of two laminated battens (one with rectangular cross section, 2 cm thick, located at the base of the crossbeam, the other having the residual thickness and maintaining the original wane), and resulted in RE = 65.80% and 10.6 mm maximum deflection. Configuration B consisted of three laminated battens (two with rectangular cross section, respectively 2 cm and 1.5 cm thick, located at the base of the crossbeam, the third having the residual thickness and maintaining the original wane), and resulted in RE = 53.54% and 14.4 mm maximum deflection. Since the RE target value was between 50% and 60%, configuration B was chosen as the tentative geometry, to be verified by a non-linear FEM simulation with full consideration of the contact phenomena; such verification indicated that the behaviour of the panel with the new crossbeams (made with Configuration B) resulted in RE = 58.1%, a value that may be considered satisfactory. Cross-section diagram of the original crossbeams

50

Configuration A Configuration A

Max deflection 10.6 mm RE= 65.80 %

20

Configuration B

Max deflection 14.4 mm RE= 53.54 %

15 20

Configuration B

Fig. 6. Two of the many configurations of the new crossbeams analysed through the FEM numerical model with their dimensions, as explained in the text, and the simulated corresponding deflections. Width remains unchanged: 11 cm for top and bottom crossbeams, 9 cm for the central one.

An Innovative Method for Dimensioning the Crossbeams

109

Based on these important and innovative results, the original crossbeams were modified according to Configuration B and to the above-mentioned Restoration Plan’s specifications. Acceptability of the Results. As mentioned in Sect. 3.3, the lack of hygroscopic tests is actually a limiting factor of this study; however, a sensitivity study carried out on the models (not reported here for brevity) showed that the dynamics of deformation of the panel varies greatly as the hygroscopic parameters change, but the resulting change in the RE index is quite small; furthermore, it should be mentioned that during 4 years of follow-up of the restored painting, no further conservation problems could be observed by the restorers. Influence of Friction Phenomena. It should be noted that the FEM simulation results are significantly influenced by the wood-wood contacts and the resulting friction, by the areas of contact that evolve over time during transient hygroscopic variations, and by the geometry of the individual boards; in this case these factors influenced about 5% of the magnitude of the RE index. Practical Details. The thin sawn elementary laminates were 0.4 cm thick and hence quite flexible; each group (glued with Bindan® glue) was held in position with clamps against a counterform matching the cupped transverse profile of the restored panel (central sagitta of about 2.2 cm) until the glue was completely set. Finally, small wooden blocks were inserted between the crossbeams and the panel, to assure their reciprocal contact despite the irregularity of the panel’s back face. The Role of the Springs. As specified by the Restoration Plan, compression springs were inserted in the anchorage points between panel and crossbeams, giving the nailed connections additional elasticity following a long-standing practice at OPD [29–31]; however, in the FEM numerical analysis these springs were neglected, considering them infinitely stiff. In fact, these springs are intended for a fine and local adjustment of the restraining system, so they are outside the global behaviour that is simulated to define RE. Therefore RE, computed according to the method described in this study, is the maximum possible for the system, whereas the springs can reduce it locally and/or globally according to the choices of the restorers.

4 Conclusions The deep and long lasting collaboration between the researchers from DAGRI Department and the Sector of the wood restorers from OPD made it possible to carry out the project described in this paper, which for the first time presents an innovative method, together with the related operational protocol, for dimensioning elastic crossbeams systems according to objective scientific procedures, when for any reason those present are judged unsatisfactory, or even detrimental to the conservation of the artwork. This method has engineering bases and adopts Finite Element Models (FEM) specifically developed for such task; however, the wisdom and experience of the restorers are an

110

L. Riparbelli et al.

integral and essential part of it, also regarding the choice of the value to be attributed to the index RE. The method starts in fact with acquiring in-depth knowledge of the physical structure of the painted panel, namely its geometrical, mechanical, and hygroscopic properties; and to provide accurate, reliable, and useful results this information needs to be obtained from measurements carried out directly on the individual artwork. If it is not possible to acquire some of this information directly from the artwork, data available from the scientific literature may be used; however, it is evident that the more accurate and specific the knowledge is about these properties, the more precisely the behaviour of the individual panel may be simulated, and the more accurate and reliable will be the design of the new crossbeams. The protocol is composed of five modules; this type of organization makes it possible to adapt it to the needs of any individual artwork, and to the various other circumstances affecting the needed intervention; also, it can take advantage of more advanced methods which can be gradually developed, for example a numerical modelling considering additional characteristics. In the specific case of the Nativity, time constraints allowed tests to be carried out only on the mechanical properties of the wooden panel and of the crossbeams, while the information on the hygroscopic behaviour of the panel was derived from literature data. Therefore, the dimensioning of the crossbeams system could not provide unequivocal results for this specific panel painting; and in this paper it was suggested that the results (in this case, the simulated distortion at a precise hygroscopic variation) should be expressed by a range of values rather than a single value (which would appear to be deterministic, whereas in reality a certain amount of uncertainty would remain). In fact, the protocol described in this paper could be used, as a scientific exercise or as an intervention guideline, even without having carried out experimental tests and technological observations on the panel; however, the results obtained in this way, i.e. without specific experimental data from the specific artwork, risk being not only inaccurate, but even harmful to the artwork itself. The dimensioning of a crossbeams system to appropriately control the cupping of painted panels can be seen as the search for a balance between various requirements, including: i) to limit recoverable deformations in the short period (i.e. the transient ones), as well as those in the long term (i.e. those of the compression set type), and ii) at the same time to prevent the development of forces that could produce stresses endangering the paint layers, the wooden support, and their adhesion.

References 1. Uzielli, L.: Historical overview of panel-making techniques in central Italy. In: Dardes, K., Rothe, A. (eds.) The Structural Conservation of Panel Paintings: Proceedings of a Symposium at the J. Paul Getty Museum 1995, pp. 110–135. The J. Paul Getty Trust, USA (1998) 2. Véliz, Z.: Wooden panels and their preparation for painting from the middle ages to the seventeenth century in Spain. In: Dardes, K., Rothe, A. (eds.) The Structural Conservation of Panel Paintings: Proceedings of a Symposium at the J. Paul Getty Museum 1995, pp. 136–148. The J. Paul Getty Trust, USA (1998)

An Innovative Method for Dimensioning the Crossbeams

111

3. Wadum, J.: Historical overview of panel-making techniques in the Northern Coutry. In: Dardes, K., Rothe, A. (eds.) The Structural Conservation of Panel Paintings: Proceedings of a Symposium at the J. Paul Getty Museum 1995, pp. 149–177. The J. Paul Getty Trust, USA (1998) 4. Mecklenburg, M., Tumosa, C.S., Erhardt, D.: Structural response of wood panel paintings to changes in ambient relative humidity. In: Dorge, V., Carey Howlett, F. (eds.) Painted Wood: History and Conservation: Proceedings of a Symposium organized by the Wooden Artifacts Group of the American Institute for Conservation of Historic and Artistic Works, and the Foundation of the AIC (1994), pp. 464–483. Getty Conservation Institute, USA (1998) 5. Ciatti, M., Castelli, C., Santacesaria, A.: Panel Painting: Technique and Conservation of Wood Supports. Edifir, Florence, Italy (2006) 6. Buck, R.D.: Some applications of mechanics to the treatment of panel paintings. In: Thomson, G. (ed.) Recent Advances in Conservation, International Institute for the Conservation of Historic and Artistic Works (IIC), London, UK, pp. 156–162 (1963) 7. Hoadley, R.B.: Chemical and physical properties of wood. In: Dardes, K., Rothe, A. (eds.) The Structural Conservation of Panel Paintings: Proceedings of a Symposium at the J. Paul Getty Museum 1995, pp. 2–21. The J. Paul Getty Trust, USA (1998) 8. Hunt, D., Uzielli, L., Mazzanti, P.: Strains in gesso on painted wood panels during humidity changes and cupping. J. Cult. Herit. 25, 163–169 (2017) 9. Gualdani, G.: Il caso studio del dipinto su tavola Natività di Cesare da Sesto, un lombardo nel Rinascimento meridionale all’inizio del XVI secolo. Particolarità tecniche e costruttive. Analisi sul degrado della pellicola pittorica in relazione al comportamento del supporto ligneo. L’intervento di restauro della tavola. [The case study of the panel painting Nativity by Cesare da Sesto, a Lombard in the southern Renaissance at the beginning of the sixteenth century. Technical and constructive details. Analysis of the degrade of the pictorial film in relation to the behaviour of the wooden panel. The restoration of the painted panel] Graduation Thesis, OPD, Florence, Italy (2017) 10. Italian standard UNI EN 408:2012 Strutture di legno - Legno massiccio e legno lamellare incollato – Determinazione di alcune proprietà fisiche e meccaniche [Timber structures - Structural timber and glued laminated timber - Determination of some physical and mechanical properties], UNI, Milan, Italy (2012) 11. Dupré, J.C., et al.: Experimental study of the hygromechanical behaviour of a historic painting on wooden panel: devices and measurement techniques. J. Cult. Herit. 46, 165–175 (2020) 12. Cocchi, L., et al.: Verifying the operation of an elastic crossbeam system applied to a panel painting: the deposition from the cross by an anonymous artist from Abruzzo, sixteenth century. Stud. Conserv. 62, 150–161 (2017) 13. Dionisi-Vici, P., Mazzanti, P., Uzielli, L.: Mechanical response of wooden boards subjected to humidity step variations: climatic chambers measurements and fitted mathematical models. J. Cult. Herit. 7, 37–48 (2006) 14. Marcon, B., Mazzanti, P., Uzielli, L., Cocchi, L., Dureisseix, D., Gril, J.: Mechanical study of a support system for cupping control of panel paintings combining crossbeams and springs. J. Cult. Herit. 13S, S109–S117 (2012) 15. Uzielli, L., Cocchi, L., Mazzanti, P., Jullien, D., Dionisi-Vici, P.: The Deformometric Kit: a method and an apparatus for monitoring the deformation of wooden panels. J. Cult. Herit. 13S, S94–S101 (2012) 16. Nelder, J.A., Mead, R.: A simplex method for function minimization. Comput. J. 7, 308–313 (1965). https://doi.org/10.1093/comjnl/7.4.308 17. Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P.: Numerical Recipes: The Art of Scientific Computing, 3rd edn. Cambridge University Press, Cambridge (2007). ISBN 978-0-521-88068-8

112

L. Riparbelli et al.

18. Guitard, D.: Mécanique du Matériau Bois et Composites, Cepadues Editions, Toulouse, France (1987) 19. Saft, S., Kaliske, M.: Numerical simulation of the ductile failure of mechanically and moisture loaded wooden structures. Comput. Struct. 89, 2460–2470 (2011). https://doi.org/10.1016/j. compstruc.2011.06.004. ISSN 0045-7949 20. Fortino, S., Mirianon, F., Toratti, T.: A 3D moisture-stress FEM analysis for time dependent problems in timber structures. Mech. Time-Depend Mater. 13, 333 (2009). https://doi.org/10. 1007/s11043-009-9103-zFortino 21. Électricité De France - EDF. Finite element code_aster, Analyse des Structures et Thermomécanique pour des Etudes et des Recherches 1989–2022. Open source www.code-ast er.org 22. Allegretti, O., Bontadi, J., Dionisi-Vici, P.: Climate induced deformation of Panel Paintings: experimental. Observations on interaction between paint layers and thin wooden supports. In: IOP Conference on Series: Materials Science and Engineering. vol. 949. IOP Publishing (2020). https://doi.org/10.1088/1757-899X/949/1/012018 23. Konopka, D., Kaliske, M.: Transient multi-FICKian hygro-mechanical analysis of wood. Comput. Struct. 197, 12–27 (2018). https://doi.org/10.1016/j.compstruc.2017.11.012 24. Gebhardt, C., Konopka, D., Börner, A., Mäder, M., Kaliske, M.: Hygro-mechanical numerical investigations of a wooden panel painting from “Katharinenaltar” by Lucas Cranach the Elder. J. Cult. Heritage 29, 1–9 (2018). https://doi.org/10.1016/j.culher.2017.08.003. ISSN 1296-2074 25. Dureisseix, D., Marcon, B.: A partitioning strategy for the coupled hygromechanical analysis with application to wood structures of cultural heritage. Int. J. Num. Meth Eng. 88, 228–256 (2011) 26. Olek, W., Perré, P., Weres, J.: Inverse analysis of the transient bound water diffusion in wood. Holzforschung 59, 38–45 (2005) 27. Olek, W., Weres, J.: Effects of the method of identification of the diffusion coefficient on accuracy of modeling bound water transfer in wood. Transp. Porous Media 66, 135–144 (2007) 28. Bratasz, L., Kozlowski, R., Rachwal, B.: Sorption of moisture and dimensional change of wood species used in historic objects. In: Gril, J. (ed.) Wood Science for Conservation Braga 2008 COST Action IE0601. LNCS, pp. 15–20. Firenze University Press, Florence (2010) 29. Castelli, C., Ciatti, M.: Proposta di intervento su particolari supporti lignei, [Proposal for interventions on particular wooden supports], OPD Restauro 1:108–11, Florence, Italy (1989) 30. Mazzanti, P., Togni, M., Goli, G., Uzielli, L.: Preliminary tests for mechanical properties of wooden ‘buttons’ used for attaching auxiliary supports behind panel paintings. Stud. Conserv. 60(5), 333–339 (2015). https://doi.org/10.1179/2047058414Y.0000000139 31. Mazzanti, P., et al.: Experimental tests for mechanical properties of wooden ‘buttons’ used for attaching auxiliary supports behind panel paintings. In: IOP Conference on Series: Materials Science and Engineering, vol. 949 (2020). https://doi.org/10.1088/1757-899X/949/1/012057

ARTE – Augmented Readability Tactile Exploration: The Tactile Bas-Relief of Piazza San Francesco Painting Luca Puggelli , Rocco Furferi(B) , Lapo Governi , Chiara Santarelli , and Yary Volpe Department of Industrial Engineering, University of Florence, Florence, Italy [email protected]

Abstract. Blind and visually impaired people are mostly excluded in enjoying visual artwork yet. Even if the effectiveness of tactile supports has been proven in previous studies, these are difficult to realize, since they are commonly handmade. In this paper, a set of computer-aided interactive tools for a semi-automatic reconstruction of tactile bas-relieves is proposed. Starting from the digital picture of a painting, this set make it possible to retrieve a 2.5D reconstruction of a scene in the form of flat-layered bas relief, which means that the scene is reconstructed solely by means of geometric primitives such planes, cylindrical surfaces, conical surfaces and generic (curve) surfaces. Tools have been specifically thought to obtain tactile bas-relieves of architectural scenes. Unlike typical handmade crafting, the proposed tools do not require specific user skills or training. In fact, user is only asked to select points (i.e., to detect a vanishing point) or segments of the picture to obtain a specific surface. Tools have been designed, optimized, and adopted to realize the tactile bas-relief of the painting Piazza San Francesco (unknown artist, Museo Civico di Arte Antica - Pistoia), within the research activities related to ARTE project (Augmented Readability Tactile Exploration), co-founded by Cassa di Risparmio di Pistoia e Pescia. Keywords: Tactile bas-relief · Blind · Shape from single image

1 Introduction The accessibility of Blind and Visually Impaired People (BVIP) to culture is a fundamental target recognized by “Universal Declaration of Human Rights”. Nevertheless, BVIP are mostly excluded from equal access to 2D visual artworks, which are realized to be enjoyed solely by sight. This gap can only be filled by implementing tactile supports which are specifically designed to help the blinds in understanding, exploring and enjoying the art piece by the sense of touch. To this end, several support typologies have been developed and proposed in scientific literature, but none is officially acknowledged as the univocal code to translate pictorial language into tactile. During a previous study, which have been carried out by the authors with the help of the Italian Union of the Blinds (UIC – Unione Italiana Ciechi) [1–5], four principal typologies have been detected and © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 113–126, 2023. https://doi.org/10.1007/978-3-031-17594-7_9

114

L. Puggelli et al.

tested with the help of a panel of blind volunteers. In particular, the support typologies that have been analyzed are tactile outline, texturized pattern, flat layered bas-relief (in the following referred as FLB) and bas-relief (also called 3D printing or 3DP). The first one consists in a flat surface in which the contours of the main figures represented in the painting are in relief. In texturized pattern, each figure represented in the scene is characterized by a different 3D pattern. Flat-layered bas-relief instead each figure is represented by flat surfaces at different depths. Finally, bas-relief is an actual bas-relief representation of the scene. From the tests, it emerged that tactile outline and texturized pattern models do not offer a relevant help in discriminating contour of objects and in perceiving object position in the 3D space (e.g., which figures are in background from which figures are in foreground). FLB and 3DP resulted to be the more meaningful tactile representations. In particular, even if the first one resulted to be the more “readable” by people who have experienced sight, 3DP generally have been judged as optimal, being at the same time meaningful and pleasant to explore [1–6] (Fig. 1).

Fig. 1. Tactile bas-relief of a painting

Due to the geometrical complexity of the scenes reproduced in paintings, their realization is often committed to specifically trained artists, who craft such manufacts by hand. Despite the undoubted high quality of such manufacts, the timings required to complete them and the related costs strictly reduce their diffusion in museums [3] (Fig. 2). For this reason, several computer-based methods have been proposed in the last decade with the aim of speeding up the relief reconstruction process from 2D images [7–9], aimed at overcoming the well-known issues related to single-view reconstruction techniques. A plausible solution has been proposed by authors in [9], in which the “volume” of complex figures is obtained by means of a specifically developed shapefrom-shading technique and the overall scene is modeled as a flat-layer bas relief. In this paper, it is reported the process of the bas-relief reconstruction of the painting Piazza San Francesco (Unknown artist, XVIII century), hosted in Museo Civico di Arte Antica in Pistoia (Italy), for which specific tools have been devised and optimized.

ARTE – Augmented Readability Tactile Exploration

115

Fig. 2. Hand crafted tactile bas-relief

2 Modelling Typology (See Fig. 3).

Fig. 3. Piazza San Francesco, unknown artist – Museo Civico di Arte Antica – Pistoia (Italy)

The choice of the painting has been done with the help of experts in art-history of the Musei Civici of Pistoia. This the work of art was chosen mainly because of its significance for the city of Pistoia, as it represents a bird’s eye view of one of the most important landmarks of city life both in the 18th century and today. The subject is significantly different from those for which the procedure was adopted and used in previous works, to obtain the tactile bas-relief. In those cases, the case study has always been a Renaissance painting, in which the principal figures are one or more human figures, displaced in the foreground. Therefore, most of the procedure is focused on the reconstruction of 2.5D surfaces by analyzing and decoding information related to shading (the principal information available to understand the shape of the figure). In

116

L. Puggelli et al.

painting, the protagonist is an urban square and its architecture, with no human figures in the foreground. People is still present, but the importance of the single within the scene is marginal: dimensions are small, details are limited or even absent. For both these two reasons, the reconstruction based on shading analysis should not be used. Shape from shading, in fact, requires the presence of details and brightness accuracy to provide satisfactory reconstruction results. On the other hand, supposing that each human figure was finely detailed and accurately shaded, their dimension in the bas-relief would have been too small to be perceived and appreciated by touch. With the idea of realizing a bas-relief with approximately 800 mm × 400 mm × 100 mm dimensions, the main human figure would be approximately 70 mm × 28 mm in size, and its head maximum dimension would be only 7 mm (see Fig. 4).

Fig. 4. Human figure dimension (mm)

It is therefore evident that a detailed modeling would lead to a manufact that may is pleasant to see, but useless and confusing to understand by touch. In this case, a FLB representation is to prefer. On the modelling side, even if this operation is made easier by excluding shading analysis, specifically devised tools are required to achieve a complete reconstruction using only perspective-based information. This is due to multiple factors, other than the overall scene complexity under an architectural point of view. To make the reconstruction more difficult, a first issue is given by the irregularity of some surfaces, like the buildings’ facades in the right side of the scene (see Fig. 5). A second issue is about the presence of cylindrical and conical surfaces (e.g., the church rose window), as visible in the following picture (Fig. 6). Finally, some of the planar surfaces (or at least that can be approximated by planes) are underdetermined (e.g., roofs): in this case no information is directly available (they do not lay on the principal plane, their orientation is not clearly defined). In the following section, these issues will be further analyzed, and specifically dedicated computer-based tools will be proposed.

ARTE – Augmented Readability Tactile Exploration

117

Fig. 5. Irregular surfaces – buildings’ facades

Fig. 6. Cylindrical and conical surfaces: church rose window (left) and church portal (right)

3 2.5 D Reconstruction Procedure On the light of what discussed in the previous section, the workflow of the 2.5 D reconstruction procedure is as follow:

118

L. Puggelli et al.

After some preliminary operations, two images are obtained: the first is a segmented image, that is used in the scene reconstruction step; the second is a properly filtered gray-scale version of the painting image, which is used in the brightness details step. In scene reconstruction step, the FLB geometry is retrieved. This phase, which is the core of the proposed procedure, improvements with respect to [9] are proposed, to overcome the issues described in the previous section. The fine detail step aims at retrieving image details by considering the brightness channel of the painting digital picture. Finally, the two 2.5 D geometries are combined to obtain the virtual model of the tactile bas-relief, which is converted into .stl file to be directly manufactured by means of 3D printings or other CMM machines. In this paper, the attention will be focused in the scene reconstruction step, for which 4 specific tools are presented. 3.1 Preliminary Operations The 2.5D retrieval procedure is thought to work directly on the digital image of the painting. Before starting with any operation, three main operations must be performed on the image: correction, segmentation, filtering. Supposing that the digital picture has been shoot by means of a calibrated camera, lenticular distortions must be corrected. In this case, the acquisition device consists of a Fujifilm XT-1 camera (provided with a 16.3 megapixel APS-C sensor, 23.5 × 15.7 mm) equipped with Fujifilm XF 18-55 f/2.84R LM OIS lens. The image has been undistorted using widely recognized registration algorithm [10, 11]. At this point the image is correct, but the painting may appear foreshortened, due to a not frontal point of view. Consequently, perspective distortion must be corrected. Finally, the image of the painting is scaled to match the actual painting’s dimensions. At this point an orthophoto of the painting is obtained and it’s possible to proceed with the second operation, during which the different objects represented in the scene, such as human figures, garments, architectural elements, are properly identified. This task is widely recognized as “segmentation” and can be accomplished by adopting any of the methods proposed in literature (e.g., [12, 13]). The result of segmentation is an image that matches the original one in dimensions, where different regions (or clusters) are discriminated.

ARTE – Augmented Readability Tactile Exploration

119

Fig. 7. Segmented image in false colors

Since reconstruction operations are based on gray-scale information, the color of such an image has been discarded by performing a color conversion of the original image from sRGB to CIELAB color space and by extracting the L channel (brightness) [14]. Even if in this work shape-from-shading is not used, albedo normalization has been performed, to equalize brightness level of each segment. This can ideally be obtained by dividing the gray channel of each segment by its actual albedo value. This operation has been done with the aim of retrieve image details in the bas-relief exploiting the brightness of the image, as in [15]. 3.2 Scene Reconstruction Once the starting image has been segmented it is necessary to define the properties of its regions, to arrange them in a consistent 2.5D scene. As in [15], the FLB is created in the form of a depth-map, i.e. a grayscale image in which the height of a pixel is given by its own gray value (from 0 in the background to 1 in the foreground). The procedure begins by determining the position of the horizon line in the image. This is carried out by calculating the coordinates of the central (or principal) vanishing point (VP), expressed in the UV reference system (specified in the picture). The central (or main) vanishing point is discovered as the intersection of two vanishing lines, which are determined in the image by manually selecting two points on each. If there is any ambiguity, more vanishing lines can be utilized, and the vanishing point is calculated as the average point of all intersections between the couples of vanishing lines. A reference coordinate system (RCS) is created around the central (or main) vanishing point, using VP as its origin. The x and y directions reciprocally match U and V to maintain the UV reference system’s correspondence. The z axis is now normal and pointing inwards to the image plane.

120

L. Puggelli et al.

At this point is possible to proceed with planes discrimination. For a sake of clarity, the normal of a plane will be expressed as follows: − → n = [nx , ny , nz ]

(1)

Four main types of planar segments can be identified when looking at a general painted scene with perspective: • frontal segments (FS), the segment is parallel to the image plane (or frontal plane, FP), whose geometry is unaffected by the perspective distortion (nz = 1); • horizontal segments (HS), the segment is parallel to the ground of the scene (or horizontal plane, HP) and its normal coincides with y-axis (ny = 1): among them, the “main plane” corresponding to the virtual 2.5D scene’s ground or floor can be identified; • vertical segments (VS), perpendicular to FP/FS and to HP/HS (nx = 1); • oblique segments (OS), not belonging to the previous three categories. Unfortunately, some of the architectural structures represented in the painting are not planar. As previously mentioned, also curve surfaces are represented in the picture, such the portal and the rose window of the church (similar to conical surfaces). In addition, part of the buildings’ facades on the right side of the painting are not planar. For what concerns FS, HS and VS, the reconstruction procedure follows the method proposed in [15], which proved to be effective in other similar works. In the following sections, the attention will be focused on the reconstruction of the remaining undetermined segments, which are: • • • •

Generic oblique segments (GOS); Generic vertical segments (GVS); Curved surfaces with vertical generators (CSVG); Conical surfaces (CS).

Oblique Planes Looking at the picture [Fig. 7], it is possible to detect many segment that actually are not parallel to the principal planes of the scene (HP,VP,FP). These can be further categorized into two distinct classes: 1. Generic oblique segments (GOS); 2. Oblique segments orthogonal to HS (generic vertical segments – GVS). To retrieve their geometry, two different solving procedures have been defined and implemented. A few observations must be made before proceeding with the description. Since GVS is a generalization of vertical segments, the dedicated procedure can also be used to retrieve VS geometry. As GOS generic segments, the dedicated procedure can ideally be used to retrieve the geometry of any planar segment.

ARTE – Augmented Readability Tactile Exploration

121

Generic Oblique Segments – GOS The interactive procedure requires the user to select the segment to model (S*) and two adjacent segments (AS 1 and AS 2 ), which physically touch S* in the 3D scene (their geometry must be known at this point of the overall modeling procedure). The successive steps are almost completely automatic. Starting from the segmented image SI, the masks relative to S*, AS 1 and AS 2 are obtained (mS*, mAS 1 , mAS 2 ).  if P(x, y) ∈ S ∗ → mS ∗ (x, y) = 1 (2) mS ∗ = if P(x, y) ∈ / S ∗ → mS ∗ (x, y) = 0 Each mask corresponds to a logical matrix large as SI, i.e. a matrix in which values can be only 1 (true) and 0 (false). Masks mAS 1 and mAS 2 are merged and mAS is obtained. mAS = mAS 1 ∪ mAS 2

(3)

Successively, mS* is inverted and mNS* is obtained () mNS ∗ = not(mS ∗ )

(4)

The inner boundary BS* of mNS* is then detected. BS* represents the outside boundary of S*. At this point, the mask I* is calculated as intersection between BS* and mAS. I ∗ = BS ∗ ∩ mAS

(5)

I* represents all the pixels of BS* that also belong to mAS, and consequently it represents the pixels of the selected adjacent segments that are adjacent to S* in the picture. If necessary, user is asked to select which part(s) of I* is relative to a physical contact between S* and the adjacent segments in the three-dimensional scene (IC*, for a sake of clarity, see picture). Being known the 3D coordinates of IC* points, it is possible to proceed with the geometry reconstruction of S*. The reconstruction process takes as an assumption that the plane πS∗ (which approximates S*) passes through IC* points. This assumption is not formally true, but the occurring error is acceptable for the application thanks to the high resolution of a digital picture (hence, the small relative size of a single pixel). From a mathematical point of view, πS∗ is formulated as a cartesian plane: πS∗ → ax + by + cz + d = 0

(6)

In this formulation, the plane is defined as dot product between plane normal − → → n (a, b, c) and the position vector − r (x − x0 , y − y0 , z − z0 ) of a point P0 (x0 , y0 , z0 ), where d = −(ax0 + by0 + cz0 ). The plane πS∗ can be retrieved by optimizing the linear system obtained by applying Eq. 6 to every point of IC*. Once πS∗ is calculated, the height of each pixel of S ∗ can be evaluated.

122

L. Puggelli et al.

The main limitation of this formulation is that the geometry of (at least) two adjacent segments to S* must be known. Generic Vertical Segments – GVS With respect to GOV, GVS reconstruction can exploit more geometric information about the segment under exam. The procedure is designed to realize the geometry of a FS (1 touching point), an GVS (2 touching points) or a series of contiguous GVS, if more than 2 touching points are selected. In this last case, the obtained surface can be seen as a prismatic-like, whose directrix is a broken line and its generators are vertical lines. The interactive procedure requires the user to select the segment to model (S*) and one adjacent horizontal segment HS, which physically touches S* in the 3D scene (its geometry must be known at this point of the overall modeling procedure). In addition, the user is asked to select at least one touching point, depending on the desired surface. The successive steps are completely automatic. Starting from the segmented image SI, the masks relative to S* and HS are obtained (mS*, mHS). For the formulation, please refer to Eq. 2. If just one touching point TP is specified, the height of S* points is fixed to be the same as the height of TP. Aside from that, the procedure generates an auxiliary depthmap (ADM) and two auxiliary segmented images (ASI1 and ASI2). ADM is an all-zero depth-map, except for the columns relative to TPs, which have the same height as the relative TP. ASI1 is an all-zero labelled image, with exception for the columns relative to TPs, which are sorted from left to right. ASI2 is zero where ASI1 > 0, and the pixels before, between, and after zero columns are progressively ordered from left to right (Figs. 9, 10, 11).

Fig. 8. S* (blue), HS (orange) and two TP (green)

Fig. 9. ADM auxiliary depth map, relative to Fig. 8.

ARTE – Augmented Readability Tactile Exploration

Fig. 10. ASI1 auxiliary segmented image, relative to Fig. 8.

123

Fig. 11. AS2 auxiliary segmented image, relative to Fig. 8.

Referring to the obtained images, the geometry can be retrieved by adopting GOS procedure. In fact, at this point two connected segment (in AS1 image) have a defined geometry (in ADM). The result is shown in figure. Curved Surfaces with Vertical Generators (CSVG) With this term, authors refer to not planar surfaces whose directrix is a curved line and its generators are vertical lines (e.g., building façade in the right side of the painting). These can be addressed as an extension of GVS, in which (ideally) every column of the segment corresponds to a generator. Starting from this assumption, if the height of the pixels relative to the directrix is known, the geometry can be retrieved similarly to GVS. As a result, the height of a generator that terminates in a touching point is set at the same height as the touching point. The remaining parts of the segment is instead retrieved by means of best-fit planes (Fig. 12 and 13).

Fig. 12. GVS obtained with the devised procedure.

Fig. 13. Irregular curved segment, obtained with the devised procedure.

124

L. Puggelli et al.

Conical Surfaces (CS) This tool has been designed to retrieve the 2.5D geometry of not-linear segments with not-vertical generators, which can be approximated to conical surfaces. In projective geometry (i.e., perspective), this includes cylindrical surfaces. The procedure follows the same line as GOS until IC* definition. The surface is then formulated as follows: CS∗ → z = ax2 + by2 + cxy + dx + ey + f = 0

(7)

Analogously to GOS, CS∗ is retrieved by optimizing the linear system obtained by applying Eq. 7 to every point of IC* (Fig. 14).

Fig. 14. Generic curved surface.

This procedure requires that the geometry of (at least) one adjacent segment is known (with at least 6 contiguous pixels).

4 Conclusion In this paper, 4 surface reconstruction tools for the flat-layer bas-relief reconstruction of a painting are proposed. Starting from a segmented digital image of the painting, these tools are specifically designed to recover: 1) generic oblique segments; 2) generic vertical segments; 3) curved surfaces with vertical generators; 4) conical/cylindrical surfaces. All these specific surfaces typologies can not be retrieved by only exploiting perspective related information. If in one hand such tools have been appositely designed to reconstruct the 2.5D scene of the painting “Piazza San Francesco”, on the other there are no restriction to their applicability to the FLB reconstruction of other paintings. In the near future the physical prototype of the bas-relief will be manufactured, by means of AM techniques [16]. With the help of a panel of blind volunteers, it would be helpful and fascinating to assess the efficacy of the enrichments obtained by the proposed tools in FLB and tactile bas-relief in general in future study.

ARTE – Augmented Readability Tactile Exploration

125

References 1. Lederman, S.J., Klatzky, R.L.: Hand movements: a window into haptic object recognition. Cogn. Psychol. 19(3), 342–368 (1987). https://doi.org/10.1016/0010-0285(87)90008-9 2. Cantoni, V., Lombardi, L., Setti, A., Gyoshev, S., Karastoyanov, D., Stoimenov, N.: Art masterpieces accessibility for blind and visually impaired people. In: Miesenberger, K., Kouroupetroglou, G. (eds.) ICCHP 2018. LNCS, vol. 10897, pp. 267–274. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-94274-2_37 3. Volpe, Y., Furferi, R., Governi, L., Tennirelli, G.: Computer-based methodologies for semiautomatic 3D model generation from paintings. Int. J. Comput. Aided Eng. Technol. 6(1), 88–112 (2014). https://doi.org/10.1504/IJCAET.2014.058012 4. Romeo, K., Chottin, M., Ancet, P., Pissaloux, E.: Access to artworks and its mediation by and for visually impaired persons. In: Miesenberger, K., Kouroupetroglou, G. (eds.) ICCHP 2018. LNCS, vol. 10897, pp. 233–236. Springer, Cham (2018). https://doi.org/10.1007/9783-319-94274-2_32 5. Stanko-Kaczmarek, M., Kaczmarek, L.D.: Effects of tactile sensations during finger painting on mindfulness, emotions, and scope of attention. Creat. Res. J. 28(3), 283–288 (2016). https:// doi.org/10.1080/10400419.2016.1189769 6. Governi, L., Furferi, R., Puggelli, L., Volpe, Y.: Improving surface reconstruction in shape from shading using easy-to-set boundary conditions. Int. J. Comput. Vis. Robot. 3(3), 225–247 (2013). https://doi.org/10.1504/IJCVR.2013.056041 7. Reichinger, A., Neumüller, M., Rist, F., Maierhofer, S., Purgathofer, W.: Computer-aided design of tactile models. In: Miesenberger, K., Karshmer, A., Penaz, P., Zagler, W. (eds.) ICCHP 2012. LNCS, vol. 7383, pp. 497–504. Springer, Heidelberg (2012). https://doi.org/ 10.1007/978-3-642-31534-3_73 8. Hoiem, D., Efros, A.A., Hebert, M.: Automatic photo pop-up. ACM Trans. Graph. 24, 577– 584 (2005). https://doi.org/10.1145/1073204.1073232 9. Governi, L., Furferi, R., Volpe, Y., Puggelli, L., Vanni, N.: Tactile exploration of paintings: an interactive procedure for the reconstruction of 2.5D models. In: 2014 22nd Mediterranean Conference on Control and Automation, MED 2014, pp. 14–19 (2014). https://doi.org/10. 1109/MED.2014.6961319 10. Vassilopoulou, S., et al.: Orthophoto generation using IKONOS imagery and high-resolution DEM: a case study on volcanic hazard monitoring of Nisyros Island (Greece). ISPRS J. Photogramm. Remote Sens. 57(1–2), 24–38 (2002). https://doi.org/10.1016/S0924-2716(02)001 26-0 11. Puggelli, L., Furferi, R., Governi, L.: Low cost device to perform 3D acquisitions based on ChAruCo markers. In: Rizzi, C., Andrisano, A.O., Leali, F., Gherardini, F., Pini, F., Vergnano, A. (eds.) ADM 2019. LNME, pp. 189–200. Springer, Cham (2020). https://doi.org/10.1007/ 978-3-030-31154-4_17 12. Arbeláez, P., Maire, M., Fowlkes, C., Malik, J.: Contour detection and hierarchical image segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 33(5), 898–916 (2011). https://doi. org/10.1109/TPAMI.2010.161 13. Heath, M., Sarkar, S., Sanocki, T., Bowyer, K.: Comparison of edge detectors: a methodology and initial study. Comput. Vis. Image Underst. 69(1), 38–54 (1998). https://doi.org/10.1006/ cviu.1997.0587

126

L. Puggelli et al.

14. Schwarz, M.W., Cowan, W.B., Beatty, J.C.: An experimental comparison of RGB, YIQ, LAB, HSV, and opponent color models. ACM Trans. Graph. (TOG) 6(2), 123–158 (1987). https:// doi.org/10.1145/31336.31338 15. Governi, L., Carfagni, M., Furferi, R., Puggelli, L., Volpe, Y.: Digital bas-relief design: a novel shape from shading-based method. Comput.-Aided Des. Appl. 11(2), 153–164 (2014). https://doi.org/10.1080/16864360.2014.846073 16. Carfagni, M., Puggelli, L.: Different strategies for rapid prototyping of digital bas-reliefs. Appl. Mech. Mater. 510 (2014). https://doi.org/10.4028/www.scientific.net/AMM.510.163

From Apulian Waste to Original Design Objects: Fused Filament Fabrication (FFF) as a Sustainable Solution Daniela Rizzo1 , Francesco Montagna2 , Elisabetta Palumbo3 , Daniela Fico2 , Valentina De Carolis2 , Raffaele Casciaro1 , and Carola Esposito Corcione2(B) 1 Dipartimento di Beni Culturali, Università del Salento, via D. Birago 64, 73100 Lecce, Italy

{daniela.rizzo,raffaele.casciaro}@unisalento.it 2 Dipartimento di Ingegneria dell’Innovazione, Università del Salento, Campus Ecotekne, s.p. 6

Lecce-Monteroni, 73100 Lecce, Italy {francesco.montagna,daniela.fico,valentina.decarolis, carola.corcione}@unisalento.it 3 Dipartimento di Ingegneria e Scienze Applicate, Università di Bergamo, Viale Marconi, 5, 24044 Dalmine, Bergamo, Italy [email protected]

Abstract. Lecce stone is a natural material extracted from quarries in the Salento area (Lecce, Apulia, Italy), and used for building and ornamental elements. Its nature as a non-renewable resource justifies its considerable exploitation; but extraction and processing activities produce waste that creates a strong environmental impact. The Puglia region is also rich in olive trees. Usually, the pruning waste is used as firewood or for furniture. Moreover, the amount of olive wood waste available has increased with the spread of Xylella Fastidiosa. Starting from the problem of waste management and the depletion of non-renewable resources, the goal of the research was to find sustainable alternatives to give a new life to waste produced from local raw materials and creating exclusive handicraft products. Composite filaments with a polymer matrix (based on polylactic acid-PLA) were produced using different waste materials as additives and then used to create “made in Apulia” products, through the use of Fused Filament Fabrication (FFF) technology. Finally, an eco-friendly coating was tested to improve the durability of PLA 3D products. This work represents a window into the possibility of creating new systems for reusing local waste materials while promoting sustainable economic growth and a circular economy. Keywords: Waste · 3D printing · LCA · Reuse · Sustainability

1 Introduction Lecce stone (LS) is a Myocenic stone from the Salento area (Italy) used as a building material and as a decorative element thanks to its workability and chromatic properties [1, 2]. Historically, LS is among the most exploited materials in Apulia. In the past, it was used to build dolmens, menhirs, Roman constructions, as well as decorations, © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 127–139, 2023. https://doi.org/10.1007/978-3-031-17594-7_10

128

D. Rizzo et al.

statues, private buildings and churches during the Baroque period [3, 4]. Even today, there are many artisans specialized in the processing of this stone who can transform it into elements of great artistic value, as well as furniture and souvenirs. In fact, in this geographical area, the stone sector is among the most flourishing and includes all the phases of the processing of this stone from extraction to the marketing of the finished product [5]. However, in recent years this sector has suffered from foreign competition, which with lower labor costs is able to offer similar products at less expense. This situation has led companies to deal almost exclusively with the extraction for the marketing of blocks, penalizing the successive processing stages. This is why it is necessary to make renewals in the production chains and bring the sector back to the levels of excellence that have distinguished it over time. However, it is necessary to intervene with actions that provide for a different approach in terms of environmental sustainability, a circular economy and sustainable economic development. In fact, stone processing has a strong environmental impact, because it is a non-renewable material and its continued extraction is causing quarries to be exhausted. Moreover, it is a nonbiodegradable material and its processing produces solid, dust and sludge waste, which represents a real environmental problem, already known in the scientific literature [6, 7]. In fact, LS waste is disposed of in the dump and this contributes to aggravating various environmental problems, such as climate warming, pollution of water, as well as the increase of landfill areas [8]. Another typical resource in the Salento area is the olive tree. A large part of the Apulian economy is based on the production and export of oil produced by these species. Furthermore, the olive are long-lived trees, so much so that Puglia boasts the presence of many centuries-old specimens, that have an established historical-anthropological value [9]. To ensure the productivity and health of the olive the main action to be carried out is pruning, which serves to renew the productive branches and to make the tree more resistant to attacks by parasites. Pruning waste is generally used for combustion or composting [10]. In other cases, the larger pieces are used for craftsmanship (i.e. for the production of kitchen utensils such as spoons, cutting boards, trays) or for interior design (i.e. tables, stools and lamps, or sculptures). The processing of these products in turn produces other waste in the form of wood shavings and powders. In the perspective of the circular economy and sustainable development the quantity and type of waste does not represent worrying numbers, it becomes interesting in terms of design and craftsmanship. This work aimed at developing innovative polymer composites starting from recycling of Lecce stone (LS) or olive wood (OW) waste, for production of 3D printed “made in Apulia” objects, through Fused Fabrication Filament (FFF) technology. The objectives are encouraging the recovery of production waste from processing and promoting the use of new technologies within the production chains through relevant solutions to the issues of environmental sustainability, the circular economy and sustainable economic development. Nowadays, an amount of LS and OW waste is recycled and used as raw material in other fields in Apulia, but a part of these, it is destined for landfill [5]. Our project was born from the collaboration with the association “Fatto in Bottega” (Fasano,

From Apulian Waste to Original Design Objects

129

Apulia, Italy), a group of artisans working with stone, metal, wood and ceramics. The main aim was to propose new and sustainable programs to recycle waste generated by the craftsmen (through the production of new composite filaments for FFF printing), and to promote the introduction of new technologies in the production chains of these small companies, as well as the post-SARS-CoV-2 economic recovery of local crafts. In Fig. 1 a schematic representation of the FFF printing process phases is reported.

Fig. 1. Indicative scheme of the phases of the FFF printing process starting from the PLA/LS waste filament and of the waterproofing phases of the 3D object.

2 Materials and Method 2.1 Materials For creating FFF filaments, polylactic acid (PLA) was used as the polymer matrix (Ingeo 4043D, NatureWorks LLC, Blair, Nebraska), and waste of Lecce Stone (LS) and Olive Wood (OW) were used as fillers. LS is a Miocene limestone, classifiable as biomicrite with pack-stone or packstone-wackstone texture. It is composed exclusively of fossil fragments, whose interstitial spaces are occupied bymicrite and partly from glauconite and phosphatic nodules [11]. The LS scraps used in the work were supplied by the company Decor s.r.l. (Monteroni, Lecce, Italy) and it was a compact block of stone with a water content of 88% by weight. To obtain the LS filler powder, the stone block was first dried in oven (60 °C, 4 days), then crushed with a mortar, and the powder was sieved until of a diameter of approximately 0.25 mm. OW is a precious wood, with high mechanical properties, in particular hardness. The wood has yellow sapwood and light brown heartwood with both light streaks, and is characterized by a very fine texture, while the grains are irregular. For this paper, the OW powder waste was kindly obtained from “I lumi del Faso by Sante

130

D. Rizzo et al.

Musa”, a craftsman of the association “Fatto in Bottega”, who makes artistic lamps with olive wood. To obtain the OW filler powder, wood waste was dried in oven until costant weight (60 °C), then milled with Retsch Ultra Centrifugal Mill ZM 100 (Retsch GmbH, Germany), until of a diameter of a 0.25 mm. Finally, Hybrid (Copernico, Italy) were tested as coatings on the 3D printed specimens. 2.2 PLA/Waste Composite Filaments Preparation PLA and LS waste powder (50% wt) were mixed and fused with a Haake Rheomix 600/610 model extruder, to produce the composite filament. The extruder has six subzones, each with a length of 60 mm. The screw has a diameter of 16 mm, with a constant L/D for the extruder equal to 25. Moreover, a commercial PLA filament for FFF (Tips 3d prn) was also used for reference materials [5]. To produce a filament, whose diameter corresponds to 1.75 mm, two different temperature profiles and screw speeds were used for pure PLA and PLA/LS composites respectively. The extruder parameters are reported in Table 1. After the exit from the circular die (3 mm in diameter), the material was water cooled and coiled on a spool [12]. PLA and OW waste powder (10% and 20% wt) was mixed and extruded with a twinscrew extruder (3Devo Composer 450 Filament Maker, Netherlands) using the extrusion parameters entered in Table 1. Table 1. PLA, PLA/LS, PLA/OW filaments: extrusion process criteria.

EXTRUSION PROCESS PARAMETERS Screew speed (rpm) Feed zone temperature (°C) Compression zone temperature (°C) Metering zone temperature (°C) Die temperature (°C)

PLA 14 175 180 200 200

PLA/LS 30 220 210 190 170

PLA/OW 4 185 190 185 187

2.3 Methods All the analyses described in this paragraph were carried out on PLA and PLA/LS filaments [5]. On the PLA/OW composite filament the optimization of the filament production is still in progress. A cradle-to-gate Life Cycle Assessment (LCA) was carried out on both filament types (PLA versus PLA/LS) to investigate the environmental impact related to the printing process [13]. The assessment is performed following UNI EN ISO 14040-14044:2021 and UNI EN 15804:2021 standards. According to UNI EN 15804:2014, this study considered a cradle-to-gate LCA for facade siding slabs made with PLA and also PLA/LS waste powder. In particular, this

From Apulian Waste to Original Design Objects

131

analysis considered two slab solutions: the former made of 100% PLA and the other containing 50% PLA and 50% LS scraps. The declared unit selected is 1 m2 of finished slabs sized 0.35m x 0.49m with a thickness of 0.02 m. The system boundaries explored extraction and processing of raw materials (A1), transportation to the manufacturing site (A2) and production of materials (A3) [14]. Figure 2 shows the system boundary considered for LS production. Concerning the A3 stage, we have assumed that the transportation distance from corn production to the manufacturing of PLA granules is 200 km and the transportation distance of PLA granules to the slab printer is 500 km. Finally, the transportation distance of LS powder waste to the slab printer is 30 km [5]. The embodied phase was modeled using SimaPro software and Ecoinvent inventory data, version 2.2. The above-mentioned methodology was applied to an Italian company operating in the extraction and processing of Lecce Stone. The Life Cycle Inventory (LCI) data regarding extraction and manufacturing of the raw materials is implemented using first-hand data gathered from June to July 2015, through a limitation on the use of data secondary to previous indicators that relate to consumption, such as fuel needed for transportation and electricity. The life cycle impact assessment (LCIA) used the impact categories from the basic CML-IA method (Table 2) and normalized for the territorial unit EU25+3 (Fig. 3). Table 2. Indicators describing environmental impacts adopted in this study. INDICATOR Global Warming Ozone layer depletion Photochemical oxidation Acidification Eutrophication potential Abiotic Depletion (ADP elements) Abiotic Depletion (fossil fuels)

ABBREVIATION GWP100a ODP POCP AP EP ADPelements ADPfossil fuels

UNIT kg CO2 eq. kg CFC 11 eq. kg C2H4 eq. kg SO2- eq. kg PO43- eq. kg Sb eq. MJ, net calorific value

The calculation of the indicators was done both with and without the landfill emissions; this served to demonstrate the success of the waste recycling process relative to these indicators. By means of DSC analysis (Mettler Toledo DSC1 StareSystem) on PLA and PLA/LS extruded filaments, it was possible to evaluate the impact of LS waste on glass transition temperature (Tg ). The analyses were carried out using a temperature range of 25 to 200 °C (10 °C/min). Rheological analysis were carried out on a Rheometrics Ares rheometer (TA Instruments, New Castle, DE, USA). Steady rate tests were performed at 200 °C, with a cone and plate configuration, varying the shear rate between 0.01 and 10 s1 . A TGA/DSC1 Star and System (METTLER Toledo, Zürich, Switzerland) was used in order to evaluate the LS content and the thermal degradation, heating the samples in the range of 20–800 °C, at a heating rate of 10 C/min. Three replicas were performed on each analysis. Finally, FFF composite filaments were tested with Fused Fabrication Filament

132

D. Rizzo et al.

(FFF) additive technology, using the Creality CP-01 printer (Creality, UK). The CAD model was created using the software Rhinoceros (Robert McNeel & Associates, USA). This was converted to a.STL file using the software Cura (Ultimaker B.V., Utrecht, Netherlands) to slicing the model, define layer height, speed, quantity of materials and time to print the models. The printed specimens were brush-coated using Hybrid coating (Copernico, Italy). The performance of this green coating was tested by measuring the dynamic contact angle using First Ten Angstroms FTA1000. Analysis was performed by the sessile drop technique using double distilled water. The average of ten measurements on each sample provided the results.

3 Results and Discussion The LCA results are shown in Fig. 2 and 3. The analysis reveals that scraps are produced in all stages of the LS production cycle, generating a high environmental impact. In particular, the evaluation indicates that impact indicators for LS are significantly better than those for PLA/LS composite slabs and even more so compared to PLA alone slabs; this results from the PLA production process with the consequent energy consumption. Important is the function of the produced and recycled stone powder which is an innovative product in building facade cladding. This means a reduction in land use for quarries

Fig. 2. Steps in the production process of PLA/LS waste FFF slabs and the boundary of the analysed system in the cradle-to-gate LCA.

From Apulian Waste to Original Design Objects

133

and landfills, which is not currently considered an EN 15804 indicator, but is relevant to the impact assessment of the stone industry.

Fig. 3. Comparison of normalised impacts values (CML baseline normalisation factors for EU25+3 for the two slab solutions: PLA and PLA/LS waste slabs).

DSC thermograms of PLA and PLA/LS filaments are reported in Fig. 4. The glass temperatures (Tg ) of both materials were measured by the StarSystem software, using the inflection point as the measuring point. The results indicate an increase of Tg from 56 °C to 66 °C, thanks to the presence of LS. From the TGA curves in Fig. 5 it can be seen that the processability of the PLA filament is not altered by the presence of the inorganic filler LS. Specifically, the analyses confirm a solid LS residue in the composite filament of about 50% by weight, as expected, and the degradation temperature remains unchanged at 170 °C for both filaments.

134

D. Rizzo et al.

Fig. 4. DSC thermograms of PLA and PLA/LS waste composite.

Fig. 5. TGA thermograms of PLA and PLA/LS waste composite.

From Apulian Waste to Original Design Objects

135

Viscosity values are reported in Fig. 6, as a function of the imposed shear rate [12]. The rheological data show that the pseudoplastic behavior, characteristic of thermoplastic polymers, which is found in slurry viscosity, increases it due to the presence of LS waste. This information is very important, because the viscosity of the inorganic polymer must be comparable to that of the polymers conventionally used with FFF equipment for successful printing. Figure 6 shows in fact the 3D printed object. Furthermore, research has shown that the deformation rate achieved by PLA/LS composites is also comparable to that of conventional plastics used in FFF (approx. 1 mm/min).

Fig. 6. Rheological curves of PLA and PLA/LS composite.

In Fig. 7 an example of an object printed with the PLA/LS filament before (left) and after (right) removal of the supports is reported.

136

D. Rizzo et al.

Fig. 7. Example of an object printed with the PLA/LS filament before (left) and after (right) removal of the supports.

PLA is a material that is highly susceptible to degradation. For this reason, this research also included the application of eco-friendly coatings on the surfaces of 3D printed objects. On the surfaces of the objects printed with FFF technology from waste PLA/LS filament, a coating called Hybrid (patented by the University of Salento for Lecce stone and purchased by the company Copernico, Italy), was applied, with the aim of waterproofing objects, increasing their longevity and durability against atmospheric agents. The performance of the green coating was evaluated in terms of the dynamic contact angle (α) measurement (°). The results of the α measurement are shown in the Fig. 8 and show a slight increase in the average values after the application of the coating.

From Apulian Waste to Original Design Objects

137

Contact angle (°) 120.0 100.0 80.0 60.0 40.0 20.0 0.0 PLA/LS untreated

PLA/LS treated

Fig. 8. Mean values and standard deviation of the dynamic contact angle measured (°) for test specimens printed using the FFF technique and PLA/LS waste filament, before and after waterproofing with Hybrid.

4 Conclusion The waste management and the exhaustion of non-renewable resources are one of the biggest problems of this age and of the future. This study demonstrates the possibility to use scraps from Lecce stone (LS) and olive wood (OW), together with polylactic acid (PLA), in order to design new composite materials to create “made in Apulia” 3D printed objects, with Fused Filament Fabrication (FFF) technology. These are, in fact, very important resources for the economy of the area of Salento (Puglia, Italy), as in most cases they entail a considerable impact on the environment. The work involved the “Fatto in Bottega” (Fasano, Italy) association, which supplied the waste materials derived from their artisanal production. Dust and solid waste are in fact the waste products that derive from the craftsmanship of Lecce stone and must be taken to landfills, with considerable disposal costs and a high environmental impact. On the other hand, wood shavings, wood dust and pieces of bark are the waste products derived from the artisanal processing of olive wood, which cannot always be reused as fuel. The main aim of the research was to find innovative solutions for the reuse of materials that would not have any possible second use. The FFF printing technique was found to be adequate for the purpose, thanks to its advantages: reuse of the material, the reduction of the carbon footprint, the extension of the life cycle of a material and the reduction of production and disposal costs. In addition, the easy use of the FFF printer and the possibility to customise products facilitate the introduction of the technology into the production chain of the craftsmen involved. In this way, the FFF printer not only allows the craftsman to avoid the disposal of material and reduce costs, but also to reuse this material for their own needs, creating complementary accessories to the handmade products, and unique objects. In fact, we demonstrated that the PLA/LS composite filament involves a great reduction in

138

D. Rizzo et al.

environmental impact, as it is partially biodegradable (since it has 50% biode-gradable polylactic acid). In addition, the environmental effects are greatly improved thanks to the recycling cycle of the bio-composite material itself, after the 3D printing process starting from the PLA/LS filament. Finally, the possibility of using green, solvent-free, eco-friendly coatings increases the durability of printed objects against atmospheric agents, allowing a reduction in material waste and the possibility of building objects for the outdoor environment. Ongoing studies are aimed at further investigating the mechanical properties (bending tests, tensile tests, hardness) of 3D printed objects in PLA/LS and of the PLA/LS/Hybrid system that, according to the first data, gives greater hardness to the object in addition to a better water repellency. The applications can then affect different areas, such as construction, architecture, decoration: the 3D objects in PLA/LS can be structural elements, such as floor slabs or cladding, or ornamental objects, decorative, sculptures or non-structural rehabilitation elements. Overall, the work aims to be a reflection on the possibility of creating new solutions for the re-use of local materials, while promoting sustainable economic growth and the circular economy. Future developments will concern the possibility of creating innovative solutions for reuse of the other materials (e.g., ceramics, metal, etc.) and the optimisation of existing composite filaments.

References 1. Cotecchia, V., Calò, G., Spilotro, G.: Caratterizzazione geolitologica e tecnica delle calcareniti pugliesi. In: 3th Convegno Nazionale sull’Attività Estrattiva dei Minerali di II Categoria, pp. 209–216 (1985) 2. Stella, M., Battista, V.: Le pietre da costruzione di Puglia: il tufo calcareo e la pietra leccese: censimento delle cave attive, tecniche di estrazione, caratterizzazione geolitologica e petrografica, caratteristiche termofisiche e meccaniche, tecnologie d’impiego, processi di degradazione e diagnosi, normative. Consiglio nazionale delle ricerche, Istituto per la residenza e le infrastrutture sociali 10, Bari (1991) 3. De Pascalis, D.G.: L’arte di fabbricare e i fabbricatori: tecniche costruttive tradizionali e magistri muratori in Terra d’Otranto dal Medioevo all’età moderna. BESA editrice, Nardò (2001) 4. Mainardi, M.: L’industria del cavar pietra. Le cave nel Salento. Conte Editore (1998) 5. Esposito Corcione, C., Palumbo, E., Masciullo, A., Montagna, F., Torricelli, M.C.: Fused Deposition Modeling (FDM): an innovative technique aimed at reusing Lecce stone waste for industrial design and building applications. Constr. Build. Mater. 158, 276–284 (2018) 6. Napolano, L., et al.: Environmental life cycle assessment of lightweight concrete to support recycled materials selection for sustainable design. Constr. Build. Mater. 119, 370–384 (2016) 7. Napolano, L., Menna, C., Asprone, D., Prota, A., Manfredi, G.: LCA-based study on structural retrofit options for masonry buildings. Int. J. Life Cycle Assess. 20(1), 23–35 (2014). https:// doi.org/10.1007/s11367-014-0807-1 8. Manfredi, S., Tonini, D., Christensen, T.H., Scharff, H.: Landfilling of waste: accounting of greenhouse gases and global warming contributions. Waste Manag. Res. 27, 825–836 (2009) 9. Scortichini, M.: Predisposing factors for “olive quick decline syndrome” in Salento (Apulia, Italy). Agronomy 10 (2020) 10. Donner, M., Radic, I.: Innovative circular business models in the olive oil sector for sustainable mediterranean agrifood systems. Sustainability 13(5), 2588 (2021)

From Apulian Waste to Original Design Objects

139

11. Fico, D., De Benedetto, G.E.: Characterisation of surface finishes on ancient historical buildings in Salento (Southern Italy): an integrated analytical approach to establish the secrets of artisans. J. Cult. Herit. 35, 99–107 (2019) 12. Esposito Corcione, C., Gervaso, F., Scalera, F., Montagna, F., Sannino, A., Maffezzoli, A.: The feasibility of printing polylactic acid-nanohydroxyapatite composites using a low-cost fused deposition modeling 3D printer. J. Appl. Polymer Sci. 134 (2017) 13. Marin, M., Öchsner, A.: Handbook of Ecomaterials. Springer, Heidelberg (2019) 14. Bahramian, M., Yetilmezsoy, K.: Life cycle assessment of the building industry: an overview of two decades of research (1995–2018). Energy Build. 219, 109917 (2020). https://doi.org/ 10.1016/j.enbuild.2020.109917

Microclimatic and Thermal Assessment

Microclimatic Experimental Investigation for Assuring Museum Preventive Conservation. Effective Conceptual and Testing Means Carla Balocco(B)

and Margherita Vicario

Department of Architecture, DiDA, University of Florence, Florence, Italy [email protected]

Abstract. Microclimate knowledge of historic buildings is fundamental for preventive conservation. Considering cultural heritage in the light of the UN-Habitat’s new report 2021, allows a radical rethinking of the relationship between the built, historical space/place and its boundary, nature, and environment according to their protection, preventive conservation, health and energy-environmental sustainability. In this research a study on the performance and efficacy of practical applications of an integrated methodological approach, based on experimental investigation and CFD transient simulation, is presented and discussed. It was a fundamental support for a new conservation/protection concept based on the knowledge of dynamic interactions between macro-environment (urban, historical architectural context) and micro-environment (building, museum and its different areas). The San Marco Museum in Florence (Italy) was the case study. The study in transient conditions of the connection between building thermo-physics, plant system working, integrated with air temperature variations and moisture transfer, allowed identification of possible interventions on critical factors, for people training and orientation towards a conscious use/fruition, better management (museum and plant system), protection/conservation of the indoor environment quality. Main findings showed that it is not always possible to satisfy the indoor climate within the constraints and recommended limits, without a new control/regulation system strategy. Keywords: Microclimate · Measurements · Transient simulation · Preventive conservation · Cultural heritage

1 Introduction The knowledge and control of indoor climate is a fundamental and complex task for artworks preservation in museums, especially if they are housed in historic buildings that belong to the cultural heritage [1]. Moreover, if their original intended use has been transformed and completely converted into another one, they cannot offer the energy performances required for newly designed buildings. In this field the EU supports the complex principle of the “Adaptive re-use of the built heritage”, i.e. the wide possibility © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 143–154, 2023. https://doi.org/10.1007/978-3-031-17594-7_11

144

C. Balocco and M. Vicario

to modify the functional and distributive use of the indoor environment of any historical building [1–5]. While on the one hand they have a good mass and thermal inertia, they often have inadequate windows with very poor light transmission characteristics, heating and/or cooling plants and artificial lighting systems without controls [3, 6, 7]. Correct indoor environmental conditions for museums, galleries, archives and libraries have been provided by the revised edition of the ASHRAE Chapter 24-TC9 [8]. The optimal museum environment concept was based on two assumptions ‘the more stable, the better’ and the ‘best available technology’ to maintain constant climate conditions to guarantee preventive conservation without any degradation phenomena risk. In particular, the new ASHRAE guidelines suggest strategic solutions for cultural heritage and collection conservation by means of economic, social and environmentally sustainable ways [8]. The recent shift from ‘best available technology’ towards more sustainable museum/collection management also include semi-passive climate control especially for unoccupied spaces (e.g. storage systems and archives). The crucial issue is to control climate conditions and air quality in a museum with energy use reduction for environmental and energy sustainability. Most of the literature, mainly on energy efficiency, shows that indoor climate control results in excessive energy consumption, especially if the museum is housed in a historical building whose intended original use and spatial/functional organization has been radically changed over time [10, 11]. Furthermore, it has also been demonstrated that maintaining a constant room air temperature and humidity (i.e. fixed design conditions) throughout the year, for museums located in climatic zones with significant seasonal fluctuations, can produce thermal discomfort. Referring to sustainability and decreasing energy consumption and environmental impact, the question about indoor climate conditioning and control of museums, has changed from strict conditions to more reasonable/adaptable ones [9–11], i.e. using exhibition/room and artworks needs as a reference instead of designing and/or retrofitting the efficiency and effectiveness of HVAC plants. Recent research has demonstrated that many materials can be conserved at indoor climatic conditions with a larger indoor air temperature and relative humidity fluctuations than those suggested by standards [6–13]. Some authors showed that the permissible temperature range for conservation can be wider than the one for people (visitors, operators, staff etc.) thermal comfort [10, 11]. Research literature has demonstrated that microclimatic parameter fluctuations and high variability can be accepted by means of an adaptive temperature control strategy because it assures thermal comfort and energy saving without risks for preventive conservation [10–13]. Some crucial research has provided dynamic set-point control strategies combining comfort and conservation requirements for indoor cooling/heating of museum environments [10]. This study, based on transient simulations, considered about twenty different climatic zones in Europe, different typologies of historic buildings (i.e. construction materials, structural, architectural and physics features, etc.) and five different thermo-hygrometric set-point strategies (i.e. ultra-stable reference case and all the ASHRAE classes) combined with different Air Handling Unit (AHU) configurations. In particular, the provided set-point algorithm, based on the integration of collection/conservation requirements with thermal comfort

Microclimatic Experimental Investigation

145

requirements, allows smooth control of seasonal adjustments connected to acceptable short fluctuations of indoor air temperature and relative humidity [10]. A recent study has shown that if the current used thermo-hygrometric set-point values, concerning technical standards and guidelines, are modified with those provided by the application of a multi-objective optimization method, it is possible to obtain a balanced result between energy consumption reduction, human comfort and conservation risks reduction [7]. The authors, evaluating the effectiveness of the proposed method by dynamic simulations, validated with experimental measurements, have proved that this fact is particularly evident for historic buildings subjected to constraints and/or cultural heritage used as exhibition spaces and museums [7]. In this research context, a recent study has been oriented to a proper indoor climate identification for historic building/museum preventive conservation, taking into account the interaction between building thermo-physics, (i.e. thermal inertia, heat capacity of building materials, thermal behaviour in response to external climatic stresses, etc.) and existing plants [12]. The authors have suggested some climatic parameter indices, calculated with measured climatic parameters (i.e. the magnitude and the speed of weekly fluctuations and the magnitude of daily fluctuations). Some indices about climatic excursions have also been proposed, because they are useful for comparing the indoor climate of different buildings, taking into account outdoor climate building mitigation in terms of magnitude and delayed time [7, 12, 13]. Starting from this literature background, in our present research a case study on the performance and efficacy of the practical applications of an integrated methodological approach was presented and discussed. The proposed method based on experimental investigation and CFD transient simulation, can be used as a fundamental support for a new concept of conservation and protection which implies the knowledge of dynamic interactions between macro-environment (urban, historical architectural context) and micro-environment (building, museum and its different areas). It allows the assessment of the adaptability, acclimatization, resilience and plasticity of the building-plant system. In particular, a thorough study at transient conditions of the connection between building thermo-physics, plant system working conditions, indoor air temperature variations and moisture transfer, was developed by means of post-processing integration of experimental and simulation results, to identify new control/regulation system strategies for an optimal building-plant system management, guaranteeing preventive conservation conditions and indoor environmental quality.

2 Methods and Materials 2.1 The Case Study The museum of San Marco in Florence (Italy) was the pilot project (Fig. 1a). It is located in the monumental part of the Dominican priory complex, which was built in the 15th century on the pre-existing medieval Sylvestrian monastery of the 13th century. The experimental and numerical investigations were carried out in the zone which was the friars’ dormitories, that are located in the oldest part of the priory, on its first floor, i.e. in the zone that surrounds the Cloister of Saint Antonino. The large space, that preserves its original structure, consists of three corridors consisting of 43 cells covered with a

146

C. Balocco and M. Vicario

barrel vault, whereas the entire zone is covered with a hipped roof supported by wooden trusses (Fig. 1b). The extensive cycle of Renaissance frescoes, painted by Fra Angelico (1438–1445), in all the cells and corridors, are of great historical and artistic value. It is important to note that the cooling/heating plant is a variable refrigerant flow system (VRF) located on the extrados of the vaults. It consists of three VRF outdoor units connected to eighteen indoor ducted fan coils, six in each corridor, equipped with a centralized temperature control system but without humidity control (Fig. 1c).

Fig. 1. Photos of the San Marco Museum: a) the building; b) the cells and corridor; c) the trussed roof and the VRF plant.

2.2 Experimental Measurements and CFD Simulation Due to room dimensions and the need for reduction of cost, time and operation invasiveness of the experimental setup, six probes were placed at the most significant points. The central zone of the second corridor was identified by the information cross-referencing process obtained by a preliminary investigation. The second corridor is 45, 5 m in length, 2,50 m wide with a pitched roof 6,97 m high at the ridge. The cells placed on both sides of the corridor are 3,70 m in length, 2,75 m wide and the barrel vault is 3,35 m at the apex. Air suggested values and literature evidence, considering the type of mixed mhumidity values were acquired every minute and processed every 15 min with Tinytag Plus 2 TGP-4500 (accuracy: temperature ±0.5 °C (0 to 40 °C); relative humidity ±3%) and Extech RHT10 (accuracy: temperature ±1 °C (−10 to 40 °C); relative humidity ±3%). In Fig. 2 their location inside cell 26 and the corridor is shown. The experimental monitoring campaign took place over one year (form December 2019 to November 2020). Referring to an extensive recent study [14] it was possible to define a 3D computational model of the environments studied and develop a methodological approach that foresees the integration between the measurement and transient simulation phases. Physics and thermodynamics of the building-plant system, investigated by means of dynamic experimental measurements, were directly used for the numerical simulation models. The integration between experimental data and their representation and development in space and time by means of computational thermo-fluid dynamics modelling, allowed the knowledge of the indoor microclimate for the study of those processes that lead to degradation phenomena and risk assessment. All the thermo-physical properties of all the wall materials and building components, number and presence of visitors/operators, lighting and window management

Microclimatic Experimental Investigation

147

hourly schedules, the VRF system, the on-off regulation system and its schedule in three different periods, deduced from a previous study [14] were used as boundary conditions.

Fig. 2. Instruments location inside cell 26 and corridor: the investigated ambient; instruments location on a section view b) and on plan view c).

The validated models provided by [14] were expanded and further developed by means of both direct/dynamic experimental data and relevant results of their postprocessing. The implemented numerical simulations at transient conditions of the connection between building thermo-physics, plant system working, integrated air temperature variations and moisture transfer, allowed discovery of possible interventions on critical factors. The external climatic data, corresponding to the experimental monitoring period, used for investigation and results post-processing, were provided by the “Fondazione Osservatorio Ximeniano” in Florence. In particular, 3 periods were investigated because they are connected to the VRF operating conditions: heating period, during which the system is switched on from 04:00 to 14:00 h with a set point temperature of 21 °C, cooling period, during which the system is switched on for 24 h a day with a set point temperature of 23 °C, and the switch-off period.

3 Results and Discussion In this section crucial results obtained by means of the matching between experimental data processing and CFD transient simulations are explained and discussed. The standard EN 15757:2010 was implemented to assess the differences arising from the plant working condition on short-term fluctuations in temperature and relative humidity. The seasonal cycle was calculated with the moving average and the short-term fluctuations, by means of the difference between the measured value and the moving average: with this procedure all the data are “smoothed” and the damping of the short-term fluctuations is obtained by highlighting long term trends. The limits of the interval, considered suitable for conservation based on the historical climate, are calculated as the 7th and 93rd percentiles of short-term fluctuations. In this way, 14% of the largest and most risky fluctuations are ruled out. The seasonal cycle was calculated with the moving average of 30 days for the period relating to the entire reference year and with a moving average of 10 days for the three periods relating to the operation of the plant. Table 1 shows the limit values and the calculated intervals for the

148

C. Balocco and M. Vicario

short-term fluctuations of the air relative humidity and temperature with respect to the measurement points inside the cell and corridor. The trends of these same parameters are shown in Fig. 3. The analysis of the measured data shows that the values of the annual interval are strongly influenced by the fluctuations calculated for the heating period. The intermittent operation of the VRF plant, which only controls the air temperature, acts on both temperature and humidity, increasing the short-term fluctuations of both parameters. While for the cooling and shutdown period of the system, the internal microclimate is much more stable with reduced short-term fluctuations. Table 1. Limit values and calculated intervals for the short-term fluctuations of the air relative humidity and temperature with respect to the measurement points, inside the cell and corridor

Heating Cooling Off Annual

RH (%) -5,70/5,60 -1,75/2,27 -2,90/2,93 -6,17/4,96

Cell 26 DRH (%) T (°C) 11,30 -0,72/0,68 4,02 -0,57/0,58 5,83 -0,47/0,47 11,13 -1,29/1,22

DT (°C) 1,40 1,15 0,94 2,51

RH (%) -4,21/4,10 -2,57/2,53 -2,57/2,72 -5,08/4,78

Corridor DRH (%) T (°C) 8,31 -2,12/1,85 5,10 -0,62/0,62 5,29 -0,65/0,63 9,86 -1,97/2,03

DT (°C) 3,97 1,24 1,28 4,00

Fig. 3. Air temperature and relative humidity trends in cell 26, respectively a) and b), and corridor, respectively c) and d) compared to the corresponding external climate values

Microclimatic Experimental Investigation

149

The EN 15757:2010 standard does not consider the effect of daily temperature and RH cycles which are harmful for storage, especially for hygroscopic materials and in the presence of air conditioning systems. Therefore, the incidence of the plant operation on the daily variations of the thermo-hygrometric parameters was evaluated according to the above. The daily fluctuation (DF) was calculated as the difference between the max and min values over 24 h. The limits suggested by ASHRAE 90-1-2007, that were used as acceptable, are those corresponding to air temperature and relative humidity requirements for collections in museum environments of class A (i.e. with a low risk and DFT < 2 °C and DFRH < 5%). Using the data representation method proposed by [7] it was possible to make the quality and stability of the environmental climate explicit (Fig. 4). In particular, in Fig. 4: n is the total number of readings; nTRH the number of readings in which DT and DRH are within the fixed range; nRH the number of readings in which only DRH < 5%; nT the number of readings in which only DT < 2 °C; n0 the number of readings in which both DT and DRH are outside the fixed range. Figure 4 shows the climatic conditions connected to the daily fluctuations verified for more than 79% of the entire year in the corridor and inside the cell. This does not happen in the winter period in which the system is switched on and during which the values in the corridor remain in the established ranges for only 5% of the time, while for 45% of the time both values are outside the range and 47% of the time the daily RH fluctuation is 5◦ C Qbuil,H

1 ηe ηc ηd COP

Text ≤5 In these equations, Qbuil,H

5◦ C,

Text >5◦ C Qbuil,H

(3)

i=1



◦C

 + Qbuil,C

1 ηe ηc ηd  ηgas

1 ηe ηc ηd EER

 (4) 

+ Elight − EPV (if > 0, else0)

(5)

represents the heating requirement of the envelope

represents the envelope heating requirement for external when Text ≤ temperature higher than 5 °C, Qbuil,C the envelope cooling requirements, Elight the

186

E. Lucchi and E. Schito

electrical energy due to lighting and EPV the electrical energy produced through the various type of PV modules, discussed in Sect. 4.1. The two values of conversion factor for natural gas and electrical energy in primary energy (fgas and fel , respectively) are usually defined in national legislation (e.g., [21]): their values are shown in Table 2. Figure 4 shows the simulated model in TRNSYS 17. Table 2. Characteristics of the heating and cooling system and energy conversion factors. Heating/cooling system Terminal unit efficiency (ηe )

0.98

Control system efficiency (ηc )

0.90

Distribution system efficiency (ηd )

0.98

Natural gas boiler efficiency (ηgas )

0.96

Second-law efficiency of heat pump in heating (ηII ,H )

0.35

Second-law efficiency of heat pump in cooling (ηII ,C )

0.25

Non-renewable primary energy conversion factors Natural gas energy conversion factor (fgas )

1.05

Bought electrical energy conversion factor (fel )

1.95

Fig. 4. The simulated model in TRNSYS 17.

Challenges and Opportunities for the Integration of Photovoltaic Modules

187

4 PV Selection PV applications in the buildings are divided in Building Attached Photovoltaic (BAPV) and Building Integrated Photovoltaics (BIPV) [5, 22–24]. The first one is based on the use of conventional PV modules [10] such as crystalline silicon or thin-film cells. These systems are mounted directly on the building envelope and function as electricity generators [5]. BIPV is defined as a multifunctional building element that substitutes the existing component for providing energy and constructive functions, such as structural integrity, mechanical resistance, thermal, noisy, fire and weather protection, daylight control, security [22–24]. Both types of PV integration can be applied to heritage buildings, considering the preservation of historic and distinctive features, and the aesthetic appearance of PV systems. Several Countries published national guidelines on PV integration in heritage buildings according to national legislation, local authorization processes, and specific heritage features [8]. The evaluation criteria can be summarized in aesthetic, technical, and energy integration [8–10]. Aesthetic integration refers to the visual, spatial, and material compatibility of PV and heritage. Visibility concerns the minimization of the visual impact of PV technologies, locating the systems on hidden roofs, interior facades, behind parapets, outbuilding, or new additions. Also, PV panels must guarantee the original appearance of the building element, matching its design, color, and texture. Thus, terracotta modules are suggested for clay roof tiles, anthracite or grey cells for slate or stone, white, yellow, or light grey modules for plaster, and highresolution printed images for marble or wood [10]. Spatial compatibility indicates the respect of original shapes, and proportions. Geometrical uniformity, particularly, means the total coverage of the building element, respect of original lines, and grouping of panels [9]. Finally, material compatibility is devoted to the protection of historic materials and values, avoiding any permanent losses and changes of the original character [8]. Technical integration refers to the hygrothermal, structural, and electrical compatibility between old and new materials, avoiding any risks of moisture, crack, falling, thermal bridge, and efficiency reduction [11]. Reversibility is a very important aspect and concerns the possibility of removing the PV system, without generating losses on the historic material [9–11]. Technical aspects integration can be reached only through a detailed design project. Finally, energy integration implies energy efficiency, cost optimization, and life cycle evaluation of the PV system [11]. A synthesis of these criteria is reported in Table 3. These guidelines report general principles for PV integration without discussing specific technologies. On the contrary, some Italian workshops of the EU project “BIPV meets history” [9] with local heritage authorities showed compatible interventions with different PV technologies in traditional buildings of Lombardy Region [12, 25]. Hidden colored, thin films, semi-transparent, and textured PV modules seem very promising for heritage integration [12]. A synthesis of this work is reported in Table 4. 4.1 Modeling of PV Modules Used in the Simulation The general equation to evaluate the electrical energy produced by PV modules is: EPV = ηPV ηINV SPV Isol

(6)

188

E. Lucchi and E. Schito Table 3. Criteria for PV integration in heritage buildings.

Type of integration

General criteria

Specific criteria

Aesthetic integration

Visual compatibility

Reduced visibility Visual compatibility Chromatic compatibility Low reflectivity

Spatial compatibility

Dimensional compatibility Respect of original proportions

Material compatibility

Conservation of heritage values

Technical compatibility

Structural compatibility

Preservation of historic appearance Technical integration

Electrical compatibility Hygrothermal compatibility Reversibility Detailed design

Detailed design project Maintenance design

Energy integration

Energy performances

Energy performances

Life cycle assessment

Environmental impact of PV systems

Economic costs Environmental impact of building

where Isol is the solar radiation on the PV module, based on slope and azimuth, and ηINV is the inverter efficiency (equal to 0.85). In Table 5, the characteristics of the simulated PV modules are reported, necessary to evaluate the ηPV . The scenario n. 1 refers to the “base case scenario”, without PV modules. Scenarios n. 2 and n. 3 refer to the typical BAPV monocrystalline blue-colored modules, with high efficiency but low level of aesthetic integration. The efficiency of these modules are evaluated through the following equations, based on [26], to be used in Eq. 6:  

(7) ηPV = ηPV ,ref 1 − βT ,PV TPV − Tref ,PV TPV = Text + (219 + 819Kt )

NOCT − 20 800

(8)

Differently from these two cases, a fixed reference efficiency was used for the other types of BIPVs and BAPVs, according to technical datasheets of producers [7].

Challenges and Opportunities for the Integration of Photovoltaic Modules

189

Table 4. PV selection in the case study according to national guidelines and local workshops. Building element

PV

PV

type

technology

No PV

-

%

Heritage

In the

Scena

recomm.

building?

rio

-

1

50%

2

Convention-

100%

3

al (blue)

50%

BAPV -

Building roof

Yes 100% BIPV

Roof

Cantilevered roof Internal canopy

PV tile

50%



5

100%



6

BIPV

Thin film

100%



Yes

7

BIPV

Thin film

100%



No

-

100%

Yes

-

100%

Yes

-

Colored (grey)

façade Façade

BAPV

Colored

100%

BIPV

(grey)

100%

BAPV

Printed

25%



Transparent

100%



Balcony Signage

Stairwell o lift

BIPV

Colored (grey)

Window

BIPV

Pavement = Recommended



8 Yes No

9

Yes 100%



10

Solar concentrator

100%

BIPV

Transparent

100%

BIPV

Colored

-

Window Greenhouse

4

(terracotta)

BAPV

Internal



Colored

External façade

100%

(green) ▲ = To be specifically evaluated

No

-



Yes

11



No

-

= Not recommended

■ = Not inserted

190

E. Lucchi and E. Schito Table 5. Main characteristics of the simulated PV modules. Azimuth

Area SPV [m2 ]

Color

Reference ηPV [-]

Scenario number

PV type

Slope

1

No PV

-

-

-

-

-

2

BAPV

27°



48.8

Blue

See Eqs. 7–8

3

BAPV

27°



97.5

Blue

See Eqs. 7–8

4

BIPV

27°



16.3

Blue

0.153

5

BIPV

27°



97.5

Terracotta

0.136

6

BIPV

27°

0° and 270°

97.5 (at 0°) 10 (at 270°)

Terracotta

0.136

7

BIPV



-

15.6

Transparent

0.146

8

BAPV

90°



9.0

Silver grey

0.120

9

BIPV

90°



6.0

Transparent

0.146

10

BIPV

90°



6.0

Silver grey

0.120

11

BIPV

90°



20.0

Transparent

0.146

5 Results and Discussion Figure 5 presents the results of the simulation in terms of electrical energy produced by the different types of PV modules. PVs in Scenarios 2-3-5-6 produce a high amount of energy not self-consumed by the building itself. Instead, most of the electrical energy produced by PV in Scenarios 4-7-8-9-10-11 is self-consumed. In terms of electrical energy balance, the use of wide areas of the roof (as in Scenarios 2-3-5-6) causes a surplus of energy that can be sold to the grid. The amount of energy immediately used by the building is quite the same for Scenarios 2-3-4-5-6-7-11. The amount of energy produced instead in the three cases 8-9-10 is very low (about 2,000 kWh), as the involved PVs area drops below 10 m2 . The evaluation of the non-renewable input primary energy (as evaluated in Eq. 3) leads to the results in Fig. 6, where this parameter is compared to Scenario 1 (No PV). Note that very different solutions, in terms of PVs features and areas, leads to similar results of saved non-renewable primary energy, due to the absence of influence of sold energy in the parameter in Eq. 3. Thus, using all the available roof or just the half to pose a BAPV (Scenarios 2 and 3) leads only to a 5% difference in the saved non-renewable primary energy, compared to the absence of PV (scenario 1). Scenarios 2-3-4-5-6-7 and 11 lead to percentage of saved primary energy higher than 20%. Only Scenarios 8-9-10 reduce non-renewable primary energy less than 10%. As a result of these analyses, Table 6 a summary of the PV integration conditions in the building, through a simplified methodology. Each scenario is evaluated from the aesthetic, energy, and technical point of view distinguishing among suggested, not suggested, and acceptable solutions. The assessment of the aesthetic integration of PV

Challenges and Opportunities for the Integration of Photovoltaic Modules

191

Fig. 5. Electrical energy produced by PV modules in the different scenarios.

Fig. 6. Percentage of saved primary energy due to PV modules.

systems refers to international guidelines and local focus groups (Table 4), and acceptable means intervention that must be evaluated case-by-case, whereas not suggested and suggested respectively means not compatible and compatible on the basis of Heritage Authorities recommendations. Energy evaluation consider the percentage of saved nonrenewable primary energy, dividing in not suggested (0%), acceptable (0–20%), and suggested (>20%) interventions. Finally, evaluation of technical integration differentiates between suggested (integrated or BIPV solutions) and not suggested (not integrated or BAPV). Scenario 1 is acceptable from the conservation of the original appearance of the roof, but not considering energy transition requirements especially considering damage and

192

E. Lucchi and E. Schito Table 6. PV integration in Rustico Macchi as result of the analysis.

decay of the building element. On the contrary, scenarios 2 and 3 are not acceptable both from heritage and technical point of view. Also, the energy-surplus produced by the systems cannot be used directly by the building, resulting in an unnecessary overproduction. Scenarios 4–5-6–7 and 11 present a high level of integration on both an energy and technological point of view, whereas for the aesthetic integration a specific analysis is needed. Scenarios 5 and 6 have significant energy production and are high conservation-compatible with traditional clay roofs. Finally, scenarios 8 (balcony), 9 and 10 (stairwell) are not compatible with the specific building from the conservation point of view, as the traditional wood materials have an aesthetical appearance of PV modules.

6 Conclusions Even if traditional PV modules are not used in historic buildings due to preservation and aesthetical reasons, their possibility of installation on heritage buildings has been evaluated in this paper, demonstrating the match between aesthetic, energy and technological integration. In particular, a set of solutions involving BIPVs on roof and windows is identified as capable of reducing non-renewable primary energy more than 20% (compared to a no-PV solution), and maintaining at the same time a high aesthetic and technological integration with the building. Future prospects of this work include an optimized integration of the heating/cooling system with PV modules and an economic analysis. Funding. Operation co-financed by the European Union, European Regional Development Fund, the Italian Government, the Swiss Confederation and Cantons, as part of the Interreg V-A ItalySwitzerland Cooperation Program for the Project “BIPV meets history. Value-chain creation for the building integrated photovoltaics in the energy retrofit of transnational historic buildings” (ID n. 603882).

References 1. European Parliament, Directive 2018/844 of the European Parliament and of the Council of 30 May 2018 adding Directive 2010/31/EU on the energy performance of buildings (EPDB) and Directive 2012/27/EU on energy efficiency. Official Journal of the European Union (2018) 2. European Parliament, Directive 2018/2002 of the European Parliament and of the Council on energy efficiency improving Directive 2012/27/EU on energy efficiency (EED) (2018)

Challenges and Opportunities for the Integration of Photovoltaic Modules

193

3. European Parliament, Directive 2018/2001 of the European Parliament and of the Council on the promotion of the use of energy from renewable sources (RED II) (2018) 4. Kabir, E., et al.: Solar energy: potential and future prospects. Renew. Sustain. Energy Rev. 82, 894–900 (2018) 5. IEA-PVPS T15: Enabling framework for the acceleration of BIPV. http://www.iea-pvps.org. Accessed 07 Dec 2021 6. Shakeel, S.R., Rajala, A.: Factors influencing households’ intention to adopt solar PV: a systematic review. In: Kantola, J.I., Nazir, S., Salminen, V. (eds.) AHFE 2020. AISC, vol. 1209, pp. 282–289. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-50791-6_36 7. BIPV meets history: value-chain creation for the building integrated photovoltaics in the energy retrofit of transnational historic buildings. http://www.bipvmeetshistory.eu. Accessed 07 Dec 2021 8. Lucchi, E., Polo López, C.S., Franco, G.: A conceptual framework on the integration of solar systems in heritage sites and buildings. IOP Conf. Ser. Mater. Sci. Eng. 949, 012113 (2020) 9. Durante, A., Lucchi, E., Maturi, L.: Building integrated photovoltaic in heritage contexts award: an overview of best practices in Italy and Switzerland. IOP Conf. Ser. Earth Environ. Sci. 863(1), 012018 (2021) 10. Pelle, M., Lucchi, E., Maturi, L., Astigarraga, A., Causone, F.: Coloured BIPV technologies: methodological and experimental assessment for architecturally sensitive areas. Energies 13, 4506 (2020) 11. Polo López, C.S., et al.: Risk-benefit assessment scheme for renewable solar solutions in traditional and historic buildings. Sustainability 13, 5246 (2021) 12. PV accept. http://www.pvaccept.de. Accessed 07 Dec 2021 13. IEA-SHC T41. Solar energy and architecture. http://task41.iea-shc.org. Accessed 07 Dec 2021 14. IEA-SHC T51. Solar energy in urban planning. http://task51.iea-shc.org. Accessed 07 Dec 2021 15. EFFESUS: Energy Efficiency for EU Historic Districts’ Sustainability. http://www.effesus.eu. Accessed 07 Dec 2021 16. IEA-SHC T59. Deep renovation of historic buildings towards lowest possible energy demand and CO2 emission (nZEB). http://task59.iea-shc.org. Accessed 07 Dec 2021 17. Klein, S.A.: TRNSYS 17: A Transient System Simulation Program. Solar Energy Laboratory, University of Wisconsin, Madison, USA (2016). http://sel.me.wisc.edu/trnsys 18. Ente Italiano di Normazione, UNI/TS 11300-1: Energy performance of buildings. Part 1: Evaluation of energy need for space heating and cooling. UNI, Milan, Italy (2014) 19. CTI (Italian Thermotecnical Committee). Italian Typical Meteorological Years (2012). https:// www.cti2000.it/index.php?controller=news&action=show&newsid=34985. Accessed 01 Dec 2021 20. MATLAB, version R2021a. The MathWorks Inc, Natick, Massachusetts (2021) 21. Decreto Interministeriale 26 giugno 2015 – Applicazione delle metodologie di calcolo delle prestazioni energetiche e definizione delle prescrizioni e dei requisiti minimi degli edifici. Gazzetta Ufficiale della Repubblica Italiana (2015) 22. BSI EN 50583-1:2016 Photovoltaics in buildings. BIPV modules (2016) 23. European Parliament and European Council Regulation (EU) No 305/2011 (2011) 24. Ghosh, A.: Potential of building integrated and attached/applied photovoltaic (BIPV/BAPV) for adaptive less energy-hungry building’s skin: a comprehensive review. J. Clean. Prod. 276, 123343 (2020)

194

E. Lucchi and E. Schito

25. Peluchetti, A., et al.: Criteria for building types selection in preserved areas to pre-assess the building integrated photovoltaics solar potential - the case study of Como land area. IOP Conf. Ser. Earth Environ. Sci. 863, 012003 (2021) 26. Evans, D.L.: Simplified method for predicting photovoltaic array output. Solar Energy 27, 555–560 (1981). https://doi.org/10.1016/0038-092X(81)90051-7

Energy Saving for Historical Heritage: The Domotised Lighting System of the Cathedral of Nardò (Lecce) Cristina Caiulo and Stefano Pallara(B) “Studio AERREKAPPA Ltd” Engineering Company, Lecce, Italy [email protected]

Abstract. Our goal was to use domotics to transform the Cathedral of Maria SS.ma Assunta of Nardò into a “smart building”, that’s why the design of the new lighting system mainly focuses on energy saving systems. The Cathedral dates back to the VII-XI centuries and the last significant interventions were carried out by the famous architect Ferdinando Sanfelice, brother of the bishop Antonio, in the XVIII century and by the bishop Giuseppe Riccardi at the end of the XIX century. The chosen illuminating devices are characterised by extreme flexibility, allowing to modify the luminous fluxes through a continuous regulation of the load. Particularly, all illuminating devices are ad-hoc devices specifically conceived for the Cathedral and equipped with specific DALI (Digital Addressable Lighting Interface) feeders and therefore all adjustable and adaptable to each “functional scenario”. Having chosen the BUS system allows to save material, optimise the path of the conduit pipes and carry out quick and simple modifications at any time, simply excluding or adding illuminating sources in a “scenario” or creating brand new “scenarios”. The project provides for an electronic system of interconnection and management made of serial communication on conductors, with a DALI open protocol to manage the illumination of the Cathedral and an electronic system of management and interconnection, with a KNX open protocol to supervise and light the accessory areas and the managing of over 50% of all the sockets. The two domotic systems are connected by means of specific KNX/DALI gateways and can be programmed by PC. The “functional scenarios” designed according to the needs of the Cathedral (liturgic or not), each of them operated by means of one single control, will be carried out through lighting devices controlled by a program control unit, which will attribute a specific configuration to each single “scenario”, though always modifiable, with no changes in the system infrastructure. All the systems can be controlled via software and each device can be eliminated at any time from one “scenario” and added to another one, or can belong to two or more “scenarios” at the same time with different illumination levels. Keywords: Lighting · Domotics · Smart building · Energy saving systems

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 195–204, 2023. https://doi.org/10.1007/978-3-031-17594-7_15

196

C. Caiulo and S. Pallara

1 Energy Saving for Historical Heritage 1.1 Introduction Before restoring and/or general upgrading buildings of particular historical, architectural, artistic and/or symbolic importance, countless critical issues arise. All the more so in the case of sacred buildings, which in addition to the specific demands listed before, also express important spiritual values.

Fig. 1. Interior lighting/rendering with staggered colours

Therefore, it is not a question of simply conserving and enhancing their value as a material or immaterial symbol, but a question of honouring the mystery they contain.

Energy Saving for Historical Heritage

197

The light is the immaterial matter that is most of all capable of doing so, as required by the specific rules of the Italian Episcopal Conference (Fig. 1). We have tried to reach an important goal: to transform Nardò Cathedral, through Domotics, into a “smart” building, an “intelligent” building: in this way, we have put forward the creation of a “smart city”. Given the historical intrusiveness of technological infrastructure in a historical listed building such as the Cathedral, the insertion of engineering equipment was a significant problem. Domotics currently represents the most innovative planning philosophy, but above all, the “lighter” planning philosophy both quantitatively and qualitatively. The considerable reduction in the quantity of cables compared to a traditional system, together with the possibility to manage and to monitor a large number of lighting fixtures, for then to create “scenarios” of light, according to the most varied necessities, has made Domotics the best choice to achieve the goal we preset. 1.2 History Here are some snippets of history, taken from the present bibliography, very useful to understand the stylistic complexity of this sacred building.

Fig. 2. Side-chapels (ph Studio Aerrekappa Ltd)

The original structure of Nardò Cathedral, dedicated to the Assumption of the Virgin Mary, dates back to the VII–XI centuries.

198

C. Caiulo and S. Pallara

One of the most renovations were accomplished by the Benedictine monks after the earthquake of 1245; more or less in the same period, the “Black Christ” crucifix was carved from walnut and mounted on an even more ancient cross made from oak, which is one of the most ancient wooden crucifixes in southern Italy (Fig. 2). Further renovations were accomplished after the earthquakes of 1350 and 1456; successively, the Cathedral underwent more renovations made directly by various Bishops, including the Neapolitan Bishop Antonio Sanfelice (1708–1736), who made use of the work of his brother, the eminent architect Ferdinando. In 1892, Bishop Giuseppe Ricciardi (1888–1908) decided to demolish the Cathedral: as soon as the first walls were removed, the original medieval structure appeared (Fig. 3). Once the baroque stuccos were detached, the ancient Nardò Basilica came into view, which until now had been hidden, showing the architecture of that era. Bishop Ricciardi then commissioned the distinguished Sienese painter Cesare Maccari (1840–1919) to fresco the choir, the apse and the ogival vault in the presbitery. 1.3 The Design of the New Lighting System of the Cathedral It is obvious how intrinsically complex the lighting project of the Cathedral was. To which prevailing style did we have to refer to, to which artistic or architectural element did we have to favour without neglecting the others?

Fig. 3. Interior lighting/usable surface area/lighting graphics [lx]

In a sacred building so rich in stratifications, the danger is that by trying to give value to everything, we risk giving no value to anything.

Energy Saving for Historical Heritage

199

The solution to this problem, which affects the lighting project in all buildings of significant historical, artistic and architectural interest, derives from fundamental parameters on the basis of which we make the choices of “what”, “how” and “how much” to light up. The criterion we used, beyond the ever prevailing liturgical needs, was both to elaborate a hierarchy of importance of the various elements composing and characterizing the building, as historically recognized by eminent specialists of this field, but also as emotionally perceived by the community within which the building was built and modified over time, and to illuminate each element according to the relevance attributed to it. More generally, the first planning rule to be followed presupposes a study of daily light in the various hours of the day with an ad hoc photographic documentary research in order to understand which lighting effects due to sunlight have characterized the Cathedral over its historical and architectural events. In light of this, a good lighting project must never omit a decisive element for its success: the “lighting control”, which consists of the pre-established map of the contemporaneity of the lighting, namely, the planning of the different “scenarios” for every need, in the belief that a good light can lend more quality (Fig. 4).

Fig. 4. Design of the lighting system: plan (by Studio Aerrekappa Ltd)

200

C. Caiulo and S. Pallara

The main basic criteria, which we kept in mind in planning the lighting system of Nardò Cathedral (Fig. 5), concern the reduction of energy consumption and the conformity of the lighting characteristics of the appliances to the specific needs, through an adequate choice of lighting fixtures with suitable luminous efficiency, color rendering index (CRI), duration, luminous flux and color temperature, according to the specific necessities of each “scenario”, to be positioned and oriented in order to obtain the optimal light level for each necessity. The chosen lighting fixtures have intrinsic characteristics of flexibility, and are therefore equipped with both the possibility of varying the luminous flux by means of continuous load regulation (dimmers), halogen bulbs and LED (Light Emitting Diode), with double KNX supervision and DALI (Digital Addressable Lighting Interface) in order to permit the adoption of the most suitable lighting for any future requirements too. The choice of the BUS system allows a considerable economy in the quantity of materials and an optimization of the canalization paths, as well as quick and very simple changes at any time with the exclusion or insertion of lighting fixtures in a “scenario”, or the creation from scratch of other “scenarios” with elementary programming from the control panel, without the help of technicians, who are responsible for managing the control unit, and any configuration updates requested by the Customer. In the case of Nardò Cathedral the redevelopment passed through the study of a structure which, as well as providing the Cathedral with a new and innovative lighting system, was also able to optimize the consumption producing significant energy savings. Because of these reasons, we chose Domotics, which we already used in 1998 in the Sanctuary of San Giuseppe da Copertino (in the small town of Copertino in the province of Lecce) appearing on the list of the heritage listed buildings. In short, Domotics is a control and management system for electrical equipments or devices or otherwise electrifiable apparatus which allows - among other things - to optimize consumption.

Fig. 5. Design of the lighting system: longitudinal section (by Studio Aerrekappa Ltd)

By means of an integrated set of sensors and digital actuators, this system can be controlled, even remotely, through software, according to certain data transmission protocols (Fig. 6).

Energy Saving for Historical Heritage

201

Lighting, heating, irrigation, roller shutters, electrified windows and anti-theft systems, everything can be managed with Domotics, an “intelligent” technology, specifically programmed through software, which is able to respond automatically to an event or a change of state of controlled devices. A common example: if the irrigation system is equipped with a digital humidity sensor installed in the ground, it will start only when the sensor detects an insufficient humidity level, and it will turn off in case of rain or, if the lighting system is equipped with a twilight sensor, the lights, which we want to automatically turn on at nightfall, will do it autonomously with a balanced level of illumination, according to the external light or according to our needs. This is why Domotics can today be considered the best system for controlling energy efficiency. If this is combined with very low consumption appliances such as LED lights and continuous load regulators (dimmers) to calibrate the levels of light intensity, it is possible to obtain remarkable and easily measurable results from the point of view of energy savings (F).

Fig. 6. Design of the lighting system: cross section (drawing by Studio Aerrekappa Ltd)

For Nardò Cathedral, we planned a specific electronic system for BUS interconnection and management, that is, a serial communication system on conductors of proper sections of suitable insulation to be housed in the same (open) DALI ducts of the power plant in order to manage the lighting of the building, and a BUS electronic interconnection and management system on unshielded twisted pair (UTP) of suitable insulation to be housed in the same ducts of the power plant, with (open) protocol KNX for the supervision, for the lighting of the accessory compartments, and for the management of more than 50% of the sockets placed in all settings, both sacred and additional, which are controlled or regulated by a dimmer for optimal management of energy consumption.

202

C. Caiulo and S. Pallara

The two home automation systems also communicate with each other through appropriate KNX/DALI gateways and are programmable via PC, connectable via dedicated KNX/USB interfaces, also in order to provide the necessary save of the configurations for easy reproducibility of the “scenarios” created in the event of a fault. In addition, through supplementary dedicated interfaces (KNX/Internet Gateway and KNX/GSM), these systems can be switched on, off and monitored also remotely. All of them are supported by a mixed wired and wireless intrusion/anti-theft alarm system in RF (Radio Frequency) consisting of proper control units, sensors, sirens, communicators, and are equipped with a dedicated interface for integration into the KNX system, besides a CCTV system (Closed Circuit Television, also known as video surveillance) consisting of an integrated web server allowing to easily connect 4 composite video inputs and an audio source, and making them available on the Internet via an accessible IP address in order to also shoot and file both on-site and remote. In this way, all the switching on and off of the lighting bodies and of any other electric device will potentially be automated by suitable sensors, dimmers and actuators. All the systems can be included in the home automation management of the building in order to allow all the necessary supervision and integration. Moreover, in the case of a sacred building, the home automation system responds perfectly to the liturgical needs as expressed in a series of ad hoc pastoral annotations issued by the CEI (Italian Episcopal Conference): for example, in that of February 18, 1993 on “The planning of the new churches”, in which the most frequent “scenarios” are listed, such as daily, weekday, holiday celebrations, artistic or prayer use, sacred liturgies (Christmas or Easter masses) or extraordinary events such as prayer vigils or sacred representations. The same concept is expressed in the Pastoral Note of 31 May 1996 on “The adaptation of churches according to the liturgical reform”, where the importance of lighting planning in the enhancement of sacred buildings is reaffirmed, which presupposes an accurate study of natural light and of the light suitable for that particular building, according to the prevailing historical period (of construction or transformation). The “functional scenarios”, planned according to liturgical and non-liturgical needs of the Cathedral, each one of them activated through the actuation of a single command, will be achieved by means of lighting devices, controlled by actuators through a programming unit, which will give each “scenario” a precise, but always modifiable configuration without changing anything of the plant infrastructure. The devices can be managed via software, and can be removed from one “scenario” at any time, and inserted in another one. In any case, they can always belong to different “scenarios” at the same time, with different levels of lighting: Domotics allows it. So in the future, if the demands of the Cathedral change, it will be sufficient to update these “scenarios”, according to the new needs, without realising masonry or installation works, but simply reprogramming the system, with an indisputable benefit in terms of reduction of material inconvenience and costs.

Energy Saving for Historical Heritage

203

2 The Construction Site More importantly is the physical impact of the plant infrastructures on the precious wall texture and on the decorative structure of the Cathedral. Beyond the due prescriptions by the competent Superintendency, the responsibility of looking for much lighter installation solutions for ducts, devices, electrical panels and cables, when it is impossible to use the existing undercurrent pipes, has led us to a result which has found the appreciation of many and our satisfaction as professionals and experts in Restoration. We paid the utmost attention to the paths of the external pipes privileging, when possible, positions enhancing internal corners; in any case, the same pipes and plastic channels were always painted by our Restorer in the colour support and texture of the material, detecting in person the corresponding RAL (Reichs-Ausschuss für Lieferbedingungen)1 . In the same way, we proceeded with the colour of all the lighting fixtures, which the supplier made as special pieces, not only because equipped with DALI ballasts, but also because provided with the colour chosen by us with the aforementioned modality. We paid particular attention to the lighting solution reserved for the high altar: a light structure of box-shaped beams, ad hoc designed in thin laminated wood panels for the recessed installation of LED lighting fixtures was built and assembled together the original one, then painted by our Restorer with a marbled effect in the style of the original beams (Fig. 7).

Fig. 7. Overall view (ph. Aristide Mazzarella)

The result is double: on the one hand, the original canopy above the altar has not been affected at all, and the final effect is that of the light that mysteriously comes from nowhere; on the other hand, the new structure is camouflaged, but perfectly recognizable and removable respecting the fundamental principles of a good restoration.

3 Conclusions Home automation has once again been confirmed as the best technology both for the management of reduced plant infrastructures and for the enormous potential of the 1 RAL can be translated in English as “National Commission for Delivery Terms and Quality

Assurance”. It is a European colour matching system which defines colours for paints, coatings and plastics.

204

C. Caiulo and S. Pallara

countless “light scenarios” that can be imagined and created, and for the possibility of remote control. Certainly, it is the best technology in the case of historic buildings, and in historical listed buildings in particular, firstly because of its minimal invasiveness, but also for the great versatility which allows to refunctionalize in the best way various architectural typologies and/or with different purposes. Once again, we are using it in our project of restoration and refunctionalization, currently underway, of the Church and half of the seventeenth-century Convent of San Domenico, also in Nardò, where the Curia intends to move part of its offices, where we face many other problems, not only plant engineering, but trying to combine the past with the near future. Acknowledgments. The authors wish to thank Monsignor Giuliano Santantonio, parish Priest of the Cathedral and Vicar of his Excellency the Bishop of the Diocese of Nardò-Gallipoli.

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors on the Energy Demand of a Museum Refurbished with Geothermal Systems: A Virtual Case Study Gianluca Cadelano1(B) , Shabnam Javanshir1 , Laura Carnieletto2 , Francesca Bampa3 , Alessandro Bortolin1 , Michele De Carli2 , Eloisa di Sipio4 , and Adriana Bernardi1 1 Institute of Atmospheric Sciences and Climate, National Research Council of Italy, Rome,

Italy [email protected] 2 Department of Industrial Engineering, University of Padua, Padua, Italy 3 UNESCO Regional Bureau for Science and Culture in Europe, Venice, Italy 4 Department of Geoscience, University of Padua, Padua, Italy

Abstract. The energy demands of a museum are related mainly to the need for space heating and cooling, to provide adequate comfort to visitors. However, it is mandatory to ensure the correct microclimate for the conservation of perishable artifacts. The temperature and humidity conditions in the exhibition rooms must be compatible with each item. In this regard, each of them requires specific conditions, which sometimes could be vastly different from those of thermal comfort for people. The cultural heritage material is the parameter that most determines both the optimal average conditions and the allowed range of variability. Such requirements are widely described in the literature. Conversely, the energy demands, energy cost and comfort conditions are parameters that depend on the specific complexity of each case study, as they are linked to the local climate, the characteristics of the buildings and the energy policy of each country. Therefore, for the purpose of this research, a specific case study in the town of Split (Croatia) was selected. A simplified energy model of a museum was used to assess how much the microclimatic needs of the various items affect the energy demands. The boundary conditions were determined by the intersection between the need to preserve the cultural heritage and guarantee thermal comfort of visitors. Materials for which this compromise cannot be achieved have been identified. They represent the items that must be kept in air-conditioned showcases. Finally, the costs were estimated for different types of energy systems. In particular, the traditional generation systems were compared with the new geothermal systems explicitly developed for refurbishments during the EU H2020 project GEO4CIVHIC. Keywords: Microclimate · Building energy · Historical building

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 205–218, 2023. https://doi.org/10.1007/978-3-031-17594-7_16

206

G. Cadelano et al.

1 Introduction 1.1 Background Museums are complex environments, which must fulfill many functions and are subject to many constraints. The main goals of a museum are to guarantee the conservation of the preserved artefacts, while allowing the best comfortable experience by the visitors. At the same time, the management of the museum must deal with the architectural and structural characteristics of the exhibition spaces, and ensure the economic sustainability, as far as possible, of the museum as a business. An essential part of the budget of a museum is due to the energy used for air conditioning. In fact, it is an important aspect, because the conservation of the artefacts, as well as the comfort of the visitors, are directly dependent on the environmental conditions. Furthermore, the typical architectural features of a museum usually include large halls and open spaces, which leads to high energy consumption to guarantee suitable temperature and humidity conditions. Moreover, even stronger constraints related to the building itself may arise if the museum is located in a historic or listed building. For example, many technological solutions are not always directly applicable in these contexts (i.e. photovoltaic panels, envelope thermal insulation may not be installed). Shallow geothermal systems can be proposed in some cases [1], and some European projects [2, 3] are trying to enlarge the basin of historic buildings that can apply such technology.

2 Main Aspects Concerning the Environmental Conditions in a Museum 2.1 Microclimate for the Conservation of Cultural Heritage and Comfort for People in Museums In general, exhibition spaces and museums have the specific purpose of making the cultural heritage available and visible to people, thus, the need for conservation is strongly connected to the need to provide adequate comfort to the visitors. The first aspect, which is conservation, should be always prioritized, but on some occasions an additional committed management is adopted. Hopefully, just in a few cases the needs of the people are prioritized, most probably compromising the conservation of artifacts. Although previous literature and international standards exist to define quantitatively suitable environmental conditions for both conservation and thermal comfort, it is still difficult for curators to find a good compromise to guarantee these two aspects simultaneously. 2.2 Thermo-Hygrometric Environmental Conditions for Distinct Types of Artifacts The control of the microclimate is fundamental for the conservation of works of art and archaeological heritage [4]. This is even more important in case of composite materials or extremely sensitive to thermo-hygrometric variations [5]. Organic materials (e.g., wood, paper, leather, etc.) need stable relative humidity and temperature conditions. People could experience thermal discomfort by the deviation of setpoint regulated by air

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

207

conditioning; however, the same variation could lead to irrecoverable damage and the destruction of artefacts. For this reason, during museum opening hours, the thermal comfort of visitors, as well as the thermal condition required to conserve the artefacts, should be considered. Moreover, during the closing hours, the thermo-hygrometric parameters should be regulated to satisfy the condition of material conservation as complete priority. Thermo-Hygrometric Comfort of People in Indoor Environment The quantitative evaluation of thermo-hygrometric comfort for people is defined by two widely recognized standards which are based on the work carried out by Fanger [6]: ANSI/ASHRAE Standard 55 [7] and ISO 7730:2005 [8]. A human being’s thermal sensation is mainly related to the thermal balance of his or her body as a whole. This balance is influenced by physical activity and clothing, as well as the environmental parameters: air temperature, mean radiant temperature, air velocity and air humidity. When these factors have been estimated or measured, the thermal sensation for the body as a whole can be predicted by calculating the Predicted Mean Vote (PMV). The PMV is an index that predicts the mean value of the votes of a statistically relevant group of persons on a 7-point thermal sensation scale ranging from −3 (cold) to +3 (hot). Zero is neutral and represents the most comfortable condition of the human body. The Predicted Percentage Dissatisfied (PPD) index provides information on thermal discomfort or thermal dissatisfaction by predicting the percentage of people likely to feel too warm or too cool in a given environment. As for people’s thermal comfort, it is obvious that buildings do not need to be air-conditioned for comfort in closing times, i.e. when visitors are not present. During the opening hours, suitable thermo-hygrometric parameters (operating temperature, relative humidity, etc.) should consider the results in terms of PPD and PMV. Specific Conservation Needs of Materials Constituting the Cultural Heritage Under certain conditions, it is just impossible to guarantee the correct conservation requirements of any type of cultural heritage in an exhibition room. For example, very extreme conservation conditions (e.g. required for very fragile materials) may be quite different from those of human thermal comfort. In other cases, these are reasons related to the difficulty of air conditioning in exceptionally large rooms for a continuous time without interruptions, i.e., even when the museum is closed to visits. In these cases, it is preferable to use air-conditioned showcases, which create a more easily controllable micro-environment. This solution requires some precautions nonetheless. For example, the risks associated with the greenhouse effect in case of nearby lighting sources, that emit in the infrared spectral range, must be avoided. Moreover, air ventilation inside the showcases should be finely controlled in order to avoid the formation of degradation triggered by stagnation of humidity. Therefore, once the compatibility between conservation needs and comfort conditions has been ascertained, it is interesting to evaluate whether avoiding showcases can be sustainable from an energy and economic point of view. The management of a museum must also consider the economic aspects, that are by large related to the cost of heating and cooling of the building. In Fig. 1 the range of temperature and humidity to conserve different types of materials that could be found in a museum has been illustrated. Materials such as leather, wood, and stone must be conserved in a range of temperature and humidity similar to the one required for human

208

G. Cadelano et al.

thermal comfort. The relative humidity and temperature to conserve the materials such as gold, porcelain, ceramics, and stoneware are not very strict as they can be varied in a wide range. However, materials such as paper and photos need to be conserved within the temperature range of 13 °C to 18 °C and 0 °C to 15 °C [9, 10], respectively, which might create strong human thermal discomfort. The items should be displayed in a showcase to meet the conservation requirements in such cases. Temperature and relative humidity ranges comparison

Fig. 1. Comparison between ranges of temperature and relative humidity proposed by literature as proper for human comfort (yellow box) and materials commonly found in cultural heritage items (other colored boxes). Stone, leather and wood are almost overlapping, and therefore barely distinguishable in figure.

On this basis to examine different climate control strategies for the rooms of a museum, the study excluded the needs of the materials that should be kept in display cases. Materials such as wood and leather are very sensitive to relative humidity, and in some cases, also to temperature. Therefore, they are often in cases, as in the proposed case study, presented in Sect. 3.2. Then, the following scenarios have been evaluated: • Scenario 1 – conservation minimum effort. The required energy demand to conserve items made from stone, wood, or leather is considered. Thermal conditions are maintained in a certain range of 19–24 °C within twenty-four-hour to prevent damage to the artifacts. Although indoor thermal comfort during visiting hours has not been considered as driving requirement, it should be guaranteed by the same condition as shown in Fig. 1. In this scenario the indoor temperature variation is followed by the outdoor temperature trend to avoid excessive energy consumption. However, the slight variation of temperature could lead to deterioration of the material over time. The relative humidity is constantly equal to 55% over 24 h. • Scenario 2 – comfort only. The energy consumption is limited to the visiting hours to satisfy the thermal comfort of the visitors. The indoor air temperature has been set to achieve a PPD of 30% or less during opening hours, leaving the museum with no heating or cooling during the remaining hours. The relative humidity is constantly equal to 55% over 24 h. The daily variation in the indoor conditions is not suitable for the conservation outside a showcase because of the significant thermal condition

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

209

variation between opening and closing hours. This scenario is applicable only to materials insensitive to abrupt temperature and humidity changes. • Scenario 3 – comfort, maximum stability. The indoor temperature and relative humidity level off at 22 °C and 55% respectively within twenty-four-hour. This condition is stabilized to facilitate the optimal conservation of cultural heritage and prevent damage to those items. As a consequence of such selected setpoints, thermal comfort should also be assessed. The study emphasizes that scenarios 1 and 2 intend to cover their objectives while also considering the issues related to energy saving, which should also be considered mainly when the museum is housed in a historic building [11]. The temperature can be adjusted within a certain range in such cases. This freedom made it possible to choose the minimum possible temperature variation compared to that which would occur without conditioning, as long as it was sufficient to achieve the conservation or comfort requirements respectively. As a result, the lowest energy possible was used. The chosen temperature of set point for scenario 2 has been obtained by mean of an adaptive optimization algorithm [12] having a specific PPD value (i.e. 30%) as constraint condition.

3 Materials and Methods 3.1 Procedure Overview The three scenarios have been simulated with a dynamic software that allows the calculation of the energy demand of the building, further elaborated to evaluate the energy uses. The first section involved data mining and analysis of a museum to define and create the energy model of the building. The second part of the work consists of the development of the energy model of the buildings in the simulation tool. In this phase, simulations for three different scenarios have been carried out to evaluate the dehumidification, heating and cooling peak load, and energy demand of the buildings. In the last step of the study, the integration of the heat pump and the electrical energy demand of the heat pump has been evaluated. 3.2 A Virtual Case Study in Croatia For the present study, a real museum has been considered. The Museum of Croatian Archaeological Monuments Muzej Hrvatskih Arheoloških Spomenika (Fig. 2) is located in Split, Croatia. In 1979, the historical center of Split was recognized as a UNESCO World Heritage Site and, contextually, the Museum has been listed on the Register of Protected Cultural Heritage of the Republic of Croatia. Today the museum stores a huge collection of artifacts (about 20000 items), and 25% of which is displayed in permanent exhibitions (Fig. 3). The Museum’s inventory consists predominantly of exemplars of jewelry, weapons, and daily use items and include many stone artifacts that once belonged to the old Croatian church interiors. The collection of early medieval wicker, clay figures, and ancient Croatian Latin epigraphic monuments

210

G. Cadelano et al.

Fig. 2. Overhead view of the building

Fig. 3. Examples of cultural heritage items from the Museum collection

(from the IX to the XII century where names of Croatian kings were engraved) is the largest collection of its kind in Europe. An architectural schematic of the building has been used to obtain a simplified model of the exhibition halls of the museum as needed for the software simulation. 3.3 Energy Modelling A model of the museum was created with TRNSYS simulation tool [13] to calculate the energy demand in three scenarios which have been briefly discussed in Sect. 2.2. Different setpoints of air temperature and relative humidity have been considered according to the principles of energy optimization and fulfilling conservation and comfort requirements. The building consists of three floors, with the primary areas of an office building, the exhibition hall, the museum workshops, and the event space. The gross footprint off the building is about 5232 m2 and total net volume of 27473.6 m3 , while the heated useful area is about 4093.1 m2 . The materials and properties which are used to construct the museum model are assumed to carry out the energy building model. The external wall is supposed to be made of concrete without any added layer of thermal insulation (U = 1.63 W m−2 K−1 ) and the internal wall consists of hollow bricks (U = 1.73 W m−2 K−1 ). The first and second floor were composed of concrete, cement, and pavement with a transmittance of 1.98 and 2.3 W m−2 K−1 , respectively. The roof was constructed with reinforced concrete (U = 1.54 W m−2 K−1 ). Double glazed windows with an aluminum frame (U = 2.5 W m−2 K−1 ) are considered in this model. Thermal Load Analyses of the Building The three scenarios described in Sect. 2.2 have been implemented in the virtual simulation as control strategies for the heating and cooling mode as follows: • In scenario 1, the air conditioning is switched on continually to satisfy the thermal comfort within the temperature range of 19 °C to 24 °C and 55% of relative humidity.

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

211

In this case, during the closing hours, the thermal setpoint is regulated dynamically, followed by the outside temperature to decrease energy consumption. • In scenario 2, thermal comfort during the visiting hours of the museum is considered from Monday to Friday, from 9 AM to 4 PM. Therefore, the air conditioning is switched on two hours before the visiting hours to guarantee the thermal comfort of the visitors immediately at 9 AM and switched off at 4 PM. The rest of the day and during the weekend the system is completely switched off. • In scenario 3, the air-conditioning system is switched on h24 maintaining the temperature at 22 °C and the relative humidity at 55%. For each building element, the geometry data (surface areas, volumes, etc.), the thermal properties of glazing elements, the internal loads due to lights, electronic devices, and visitors in the different rooms have been collected. The required data of energy gained by electrical equipment, lighting, and activity of people, as well as the metabolic rate and clothing factor have been estimated using the value suggested in ASHRAE 55 [7]. The thermal load profiles for different control strategies have been calculated by means of TRNSYS simulations. The thermal peak load of the buildings and the annual energy load profile are obtained from the simulations of one year of operation period of the museum building. 3.4 Comfort and Microclimate Analysis The energy simulations discussed in Sect. 3.3 required the creation of dynamic annual thermo-hygrometric trends based on a time resolution of one hour. Figure 4 shows the indoor (in red) and outdoor (in blue) air temperature trends over a one-year period in the museum, which will be used as input data for the energy model to compare scenario 1 (top), scenario 2 (center) and scenario 3 (bottom). The suggested range for the conservation of stone (but quite valid also to leather and wood) is given as reference and it is represented as a green band. While scenario 1 and scenario 3 required keeping the indoor temperature within certain limits or strictly at a selected set point, scenario 2 also required to implement the occupancy schedule (based on museum opening hours) and optimization of the conditions to reach a maximum PPD threshold (imposed equal to 30%) calculated using custom algorithms coded in the Matlab® software environment. It is quite evident that, except for scenario 3, there are significant daily variations in temperature, which are particularly large in scenario 2 due to the temperature drop when the heating is turned off at night and during the weekends. The environmental parameters produced by the simulation software for each scenario have been implemented according to standard methods for the calculation of PMV and PPD as presented in Sect. 2.2. Amongst the parameters needed to calculate the comfort parameters (Metabolic rate and Clothing index), we considered the former as constant and equal to 1.2 met = 69.84 W/m2 , which corresponds to the energy produced per unit surface area of an average person while standing and slowing moving though the exhibition rooms. The latter act as thermal insulation when clothes are worn by a person and that has a substantial impact on thermal comfort. Considering the typical clothing variability during the year, this value was adjusted on a daily basis and ranged from 1 Clo = 0.155 m2 ·K/W down to 0.5 Clo, linear, dependent on the average outdoor temperature.

212

G. Cadelano et al. Indoor vs outdoor temperature- Scenario 1: Conservation, minimum effort

Indoor vs outdoor temperature- Scenario 2: Comfort only

Indoor vs outdoor temperature- Scenario 3: Conservation, maximum stability

Fig. 4. Indoor and outdoor air temperature yearly trends for scenario 1 (top), 2 (center) and 3 (bottom).

These limit values correspond to trousers, long-sleeved shirt, long-sleeved sweater, tshirt (Clo = 1), and full-length trousers and long-sleeved shirt, light jacket (Clo = 0.74) to replicate winter and summer indoor clothing. Since there is not significant air leakage through the building envelope of the actual building, we considered a value of air velocity equal to 0.1 m/s. The figures below show the PMV (blue) and PPD (red) values during opening hours, for each scenario (Figs. 5, 6 and 7). For each scenario, the PPD values remain below 30%. This happens not only in scenario 2 where that was the imposed restriction, but also in scenario 1 and in scenario 3, the latter rarely going beyond 10–15%. The PMV trends indicate a similar situation for scenario 1 and scenario 2, where the sensation of comfort varies from −1 (slightly cold) in winter to +1 (slightly warm) in summer. However, in scenario 1 the spring and autumn seasons fall within the limits of neutral sensation, while in scenario 2 the feeling of cold covers all year except summer, when people sensation turns abruptly

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

213

PMV and PPD - Scenario 1: Conservation, minimum effort

Fig. 5. PMV (in blue) and PPD (in red) yearly trends for scenario 1. PMV and PPD - Scenario 2: Comfort only

Fig. 6. PMV (in blue) and PPD (in red) yearly trends for scenario 2.

warm. Scenario 3 is the most favorable from the point of view of comfort, in fact the sensation is equal to 0 ±0.5 all year round. Energy Demand and Primary Energy Use Analysis Total energy demand required for heating and cooling in three scenarios are summarized in the Table 1.

214

G. Cadelano et al. PMV and PPD - Scenario 3: Conservation, maximum stability

Fig. 7. PMV (in blue) and PPD (in red) yearly trends for scenario 3.

Table 1. Energy demand based on the prioritization of either comfort or conservation. Scenario

Heating (MWh)

Cooling (MWh)

1 – Conservation, minimum effort

265

269

2 – Comfort only 3 – Conservation, maximum stability

77

25

432

312

In the cases where material conservation is prioritized, cooling and heating are required to keep the handcrafts within a certain temperature and humidity in all seasons. Therefore, the energy required for heating and cooling is not limited to the cold and hot seasons. As it can be seen from Table 1, the required energy of heating and cooling in the third scenario reaches the highest consumption corresponding to 432 MWh and 312 MWh for space heating and cooling respectively. However, in the first scenario, those values are remarkably lower and decrease almost by half at 265 MWh for heating and at 269 MWh for cooling, due to the lower number of operating hours required. In the third case, in fact, the air conditioning system is switched on within 24 h to guarantee the constant temperature at 22 °C, whilst in the first scenario the indoor temperature trend is conservatively adjusted considering the thermal control of material conservation. This allows a significant energy saving, nevertheless temperature variation could have detrimental effects on material conservation. In the second scenario, where air conditioning is switched off during closing hours the energy consumption is minimized. The scenario that prioritizes the visitors’ thermal comfort requires the lowest energy demand, corresponding to the 29% and the 9% of the heating and cooling energy required when applying the first scenario. This result indicates the different impact on

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

215

the energy demand of the conservation of materials compared to the thermal comfort of visitors, thus highlighting the importance of using adequate control strategies and generation systems to satisfy the energy requirements. In fact, if heating is provided by a natural gas boiler, commonly coupled to high-temperature terminal units, the overall expected efficiency for generation, distribution and emission system could be around 85%. On the contrary, considering the mild climate, heat could be provided by electric heaters, whose system efficiency is around 95%. Assuming the use of a heat pump coupled with the existing high-temperature emission system to minimize the retrofit actions inside the building while introducing renewable energy sources, a constant coefficient of performance (COP) of 2.5 can be used, as suggested by previous studies held during Horizon 2020 Cheap-GSHPs project [14]. However, electrical heaters and Natural Gas Boiler systems can provide only heating. Thus, the comparison considered only space heating energy demand (Table 2). Table 2. Non-renewable primary energy use for space heating based on the prioritization of either comfort or conservation. Different energy sources are compared: electrical, natural gas and geothermal. Scenario

Electrical heaters (MWh)

Natural gas boiler (MWh)

Geothermal heat pump (MWh)

1 – Conservation, minimum effort

404

342

153

2 – Comfort only

117

99

45

3 – Conservation, maximum stability

659

559

250

The non-renewable primary energy conversion factor of natural gas is fixed equal to 1.1, while the conversion coefficient for electricity production is equal to 1.45 [15]. For this reason, the primary energy demand required by electrical heaters is the highest among the possible solutions. Although heat pumps have the same primary energy conversion factor due to the electrical energy required to operate, the efficiency rate is 2.5; thus, this system allows a significant primary energy reduction. If there is no constraint conceiving urban or historical importance, coupling ground source heat pumps with photovoltaic systems (when possible) further increases renewable energy share. At the same time, primary energy use is reduced because the energy needed by the system for the daily operation corresponds to the period of maximum solar energy production. Furthermore, geothermal heat pumps can supply cooling energy demand with the same system, reducing operating costs and maintenance. Given the cost of energy in Croatia, where the case study is located, it is possible to assess the annual cost of each scenario. In particular, Eurostat suggests 0.132 e/kWh as electric energy cost, while 0.038 e/kWh can be considered for the price of natural gas. Table 3 presents the estimated costs for the three scenarios, showing the different impact of the operating schedules of the systems. It is worth to remind that for the sake of this study, only the current operating cost has been considered. Installation, maintenance, and ownership expenses are not included.

216

G. Cadelano et al.

Table 3. Cost per year demand based on the prioritization of either comfort or conservation. Different energy sources are compared: electrical, natural gas and geothermal. Scenario

Electrical heaters (ke/year)

Natural Gas Boiler (ke/year)

Geothermal Heat Pump (ke/year)

1 – Conservation, minimum effort

36.8

11.8

14

2 – Comfort only

10.7

3.4

4.1

3 – Conservation, maximum stability

60

19.3

22.8

In addition, the choice to prioritize comfort or conservation, and the source of energy used, have an impact on the environment. Table 4 represents the carbon footprint of one year of operation, based on the calculation of the emission associated with the production of each type of energy in Croatia, which is represented in grams of CO2 . Similarly to the results presented concerning the cost of energy, the environmental impact is highly dependent on the country, due to the different national share of renewable energy sources. ISPRA gives a conversion factor of 210 gCO2 per kWh of thermal energy used, whereas 199.1 gCO2 /kWh is the value used for electrical energy [16]. Table 4. CO2 emissions per year demand based on the prioritization of either comfort or conservation. Different energy sources are compared: electrical, natural gas and geothermal Scenario 1 – Conservation, minimum effort 2 – Comfort only 3 – Conservation, maximum stability

Electrical heaters (Mg CO2 /year) 80

Natural Gas Boiler (Mg CO2 /year) 72

Geothermal Heat Pump (Mg CO2 /year) 31

23

21

9

131

117

50

4 Conclusions The temperature and humidity of the indoor air of a museum must be rigorously controlled in order to ensure adequate microclimatic conditions for the conservation of the housed collections and acceptable thermal comfort for visitors and staff. Regardless of the specific characteristics of the building, and the types of materials that constitute the preserved cultural heritage, after all, every possible strategy for controlling the indoor microclimate is a compromise between comfort, conservation and energy costs. The case study presented is located in Split, Croatia, and allows the evaluation of three different scenarios. When considering a temperature range of 19–24 °C within twenty-four-hour

The Impact of Conservation Conditions Versus Thermal Comfort of Visitors

217

to prevent damage to the museum’s items that guarantees overall thermal comfort for visitors, energy demands are generally acceptable. However, if the conservation of materials is prioritized, the energy needs for space heating and cooling almost doubles (scenario 3). Lastly, if thermal comfort is considered as a driving requirement, the energy need is reduced by 70% and 90% in heating and cooling, respectively. The related results, converted in energy use considering the efficiency of different generation systems, highlight the benefits achievable with the installation of ground source heat pumps (GSHP). In fact, it can reduce primary energy use and CO2 emissions of more than 50% when compared to gas boilers and more that 70% if compared to electrical heaters. Although the investment cost related to the installation of ground heat exchangers has not been considered in this analysis and it is actually the main drawback associated with geothermal systems, the difference shown in cost savings among the three generation systems could guarantee interesting payback times for GSHP. Moreover, the use of GSHPs allows the supply of cooling energy demand that is needed during the warmest period of the year in July and August. Further studies will be needed to size the borehole heat exchangers field and the possibility to couple the system with photovoltaic panels to exploit the contemporaneity between solar availability and energy demand of the building. Funding. GEO4CIVHIC project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 792355.

References 1. Perry, T., Jay, C.A.: Sustainable solutions for historical buildings: geothermal heat pumps in heritage preservation. APT Bull. 40, 21–28 (2009) 2. Cheap-GSHPs EU Project. www.cheap-gshp.eu. Accessed 10 Oct 2021 3. GEO4CIVHIC EU Project. www.geo4civhic.eu. Accessed 10 Oct 2021 4. Thomson, G.: The Museum Environment. Butterworths-Heinemann, London (1978) 5. Bernardi, A.: Microclimate inside cultural heritage buildings. Il Prato, Padova (2008) 6. Fanger, P.O.: Assessment of man’s thermal comfort in practice. Br. J. Ind. Med. 30, 313–324 (1973) 7. The American Society of Heating, Refrigerating and Air-Conditioning Engineers. ANSI/ASHRAE Standard 55—Thermal Environmental Conditions for Human Occupancy. ASHRAE, Atlanta, GA, USA (2017) 8. The International Organization for Standardization. ISO 7730:2005—Ergonomics of the Thermal Environment—Analytical Determination and Interpretation of Thermal Comfort Using Calculation of the PMV and PPD Indices and Local Thermal Comfort Criteria. ISO, Geneva, Switzerland (2005) 9. UNI 10829:1999: Beni di interesse storico e artistico - Condizioni ambientali di conservazione - Misurazione ed analisi 10. The American Society of Heating, Refrigerating and Air-Conditioning Engineers. ASHRAE Handbook—HVAC Applications Chapter 21: Museums, Galleries, Archives, and Libraries. ASHRAE, Atlanta, GA, USA (2011) 11. Comite Europeen de Normalisation. EN 16883: Conservation of Cultural Heritage—Guidelines for Improving the Energy Performance of Historic Buildings. Comite Europeen de Normalisation, Brussels, Belgium (2017)

218

G. Cadelano et al.

12. Cadelano, G., et al.: Improving the energy efficiency, limiting costs and reducing CO2 emissions of a museum using geothermal energy and energy management policies. Energies 12, 3192 (2019). https://doi.org/10.3390/en12163192 13. Klein, S.A., et al.: TRNSYS, a transient simulation program. Solar Energy Laboratory, University of Wisconsin, Madison (1994). http://www.trnsys.com/. Accessed 10 Oct 2021 14. Emmi, G., et al.: A novel ground-source heat pump with R744 and R1234ze as refrigerants. Energies 13, 5654 (2020). https://doi.org/10.3390/en13215654 15. Eurostat. Calculation methodologies for the share of renewables in energy consumption (2017). https://ec.europa.eu/eurostat/statisticsexplained/index.php?title=Calculation_ methodologies_for_the_share_of_renewables_in_energy_consumption. Accessed December 2021 16. Istituto Superiore per la Protezione e la Ricerca Ambientale, ISPRA, Fattori di emissione atmosferica di gas a effetto serra nel settore elettrico nazionale e nei principali Paesi Europei, Rapporti 303/2019. https://www.isprambiente.gov.it/files2019/pubblicazioni/ rapporti/R_303_19_gas_serra_settore_elettrico.pdf. Accessed 15 Dec 2021

The Medusa Parade Shield by Caravaggio: Making Its Structural Replica, Laboratory Testing, and Numerically Modelling Their Hygro-Mechanical Distortion Behaviour Paola Mazzanti(B) , Paolo Dionisi-Vici , Marco Fioravanti , Elisa Cardinali, Justine Mialhe, Marco Togni , Luca Uzielli , and Lorenzo Riparbelli DAGRI, University of Florence, Florence, Italy [email protected]

Abstract. The parade shields are artefacts present in the 15th century, up to the 17th . Even if their exact function is not clear, it is known they were not used in battle; according to most scholars, they might have been used as display objects, carried by soldiers and retainers of princes and noblemen in military and religious parades, or as tournament prizes. Both the shape and the materials are distinctive aspects of these parade shields and depend on the historic period they were manufactured. Among them, is the Medusa shield painted in Italy by Caravaggio around 1598, and presently exhibited in the Uffizi Gallery (Florence). During the restoration campaign it underwent in 1998–2002, several information were collected about its wooden structure, geometry and materials used. Computed Tomography showed that the wooden structure consists of wooden lamellae forming two adjacent almost spherical shell layers glued together and having the grain directions approximately perpendicular to each other. Anatomic examination confirmed that in Medusa’s and in other cases the lamellae were made of Poplar wood (Populus alba L.). Based on such information a hypothesis was developed about the construction technique originally used, and by applying this technique (which included soaking the Poplar wood lamellae in water and adapting them on a spherical counter form) an exact structural replica was manufactured. Several experimental tests were then carried out on such replica to interpret its mechanical behaviour, including some load tests to study its elastic behaviour, and some hygroscopic ones (i.e. variations in ambient humidity) to observe the induced distortions. The experimental results thus obtained were key to understand the mechanical and hygro-mechanical behaviour of this type of structure, and to explain its tortoise-shell shape, which is typical for the shields consisting of two (or even more, provided it is an even number) wooden layers; a numerical model was also developed, to explain the observed hygro-mechanical behaviour. Keywords: Panel paintings conservation · Caravaggio’s Medusa shield · Testing and numerical modelling

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 219–234, 2023. https://doi.org/10.1007/978-3-031-17594-7_17

220

P. Mazzanti et al.

1 Introduction The parade shields are artefacts already present in the 15th century, whose use lasted up to the 17th . Even if their exactly function is not clear, it is known they were not used in battle and possibly they were used as display objects, carried by soldiers and retainers of princes and noblemen in military and religious parades [1], or as tournament prizes. Such artworks are niche objects; however, there are several examples exhibited in various museums worldwide such as the Uffizi, Stibbert and Bardini Museums in Florence (Italy); Victoria and Albert Museum and British Museum in London (United Kingdom); Metropolitan Museum in New York (USA) just to mention a few. The shape and the materials are distinctive aspects of these parade shields. The geometric shape of these shields changes with time passing; for the 15th century they have various shapes, such as short and square, or long and pointed [2], or even round-shaped, while during the 16th and 17th they mainly were round-shaped [3]. In addition, a large variety is encountered among the materials used to manufacture the shields, such as papier-mâché, leather, wood, iron [3]; however, this paper focuses on the parade shields made of wooden lamellae. The round wooden parade shields (known in Italian as rotelle) are characteristic objects for both their 3D shape and their structure; the original shape, round and convex, is peculiar and represents a challenge for the artist, who needs to accurately control the proportions of the images. The restorers and conservators, who need to repair or control possible damages on such convex objects, face the same challenge today. The structure is also peculiar, since such shields are usually made of wooden lamellae organised in various layers, up to four, and each layer is approximately perpendicular to the adjacent ones. The main direction of the lamellae approximately corresponds to the longitudinal anatomical direction of the wood and the wooden structure contributes to the present shape of these shields; some of them are still round-shaped as when they were manufactured, while some others have gained a tortoise-shell shape. It is plausible that the shields with an odd number of layers are still round, while the ones with an even number of layers have a tortoise-shell shape. This paper focuses on one specific shield made of two crossed layers of wooden lamellae and characterised by the tortoise-shell shape, the well-known Medusa shield, painted by Caravaggio (Michelangelo Merisi, 1571– 1610) and exhibited at the Uffizi Gallery in Florence (Italy). Together with the Medusa shield, some others shields exist that have the same shape, such as the two shields (n. 1412 and 4899) at the Stibbert Museum (Florence, Italy), four (including n. 458) at the Bardini Museum (Florence, Italy) or the “other Medusa” shield, known as Medusa Murtola (private collection, Milan, Italy) [4]. For conservative purposes, several aspects need to be considered including a) the structure, b) the shape, and c) the materials of these parade shields. As for the Medusa shield, the painting materials were deeply investigated especially during the restoration campaign that took place in 1998–2002 [3, 5]; on the other hand, few information are present in literature describing the construction techniques and the main distortions the shield has undergone since its manufacturing, which could be just as useful for conservators and restorers. This paper aims to help reduce such information gap; for this purpose, at the wood technology laboratories of DAGRI (University of Florence) a wooden shield was manufactured as a faithful replica of the Medusa one, regarding both the materials and the structure [6]. Such

The Medusa Parade Shield by Caravaggio

221

replica underwent several hygroscopic and mechanical tests to analyse and explain its behaviour, and an ad-hoc numerical model was also recently developed to explain it.

2 Materials and Methods 2.1 The Analysis of the Medusa Shield Structure During the restoration campaign the Medusa shield underwent in 1998–2002, lots of information were collected on the paint layers and the wooden structure. The survey of the paint layers provided information on the materials used for the ground and the painted layers both on the recto and the verso of the shield; also, the conservation conditions of such layers were assessed. For this specific subject [3, 5] should be consulted, since this paper is focused on the wooden structure specifically. As for the wooden structure, at present the shield has a typical tortoise-shell shape, and its perimeter does not lay on a plane. The two main perpendicular diameters are 570 mm, the longer one, and 550 mm, the shorter one; the average curvature radius of the shell is 403 mm, and the height of its top (with respect to a plane surface on which it rests when placed horizontally) is around 115 mm. In addition to this macroscopic analysis, the Medusa shield underwent a CT (Computed Tomography) scan within the same restoration campaign [7] and the over 80 images (see e.g. Fig. 1) were processed by means of a 3D CAD software to reconstruct the wooden structure (see Fig. 2). The CT scans showed that the shield has the approximate shape of a spherical shell and is formed by two layers of lamellae glued on each other. The outer layer consists of 7 lamellae, the inner one of 8. The lamellae have the shape of fusiform spherical segments (similar to the time zones of the Earth), variously truncated at the ends because the shield is a spherical cap, not a complete hemisphere; their maximum width is between 60 and 92 mm. In crosssection (see Fig. 1) the layers have the shape of arcs of circumference; each layer is approximately 6 mm thick throughout all its surface. The lamellae have straight grain, oriented along their longitudinal axis; their growth rings are randomly oriented; at the top of the shell the two layers are oriented approximately one perpendicular to the other, deviating from the perpendicular by an angle of about 6°. The total thickness of the shield is 20 mm approximately, when the linen and gesso layers are included, and thins down to 5 mm at the edges. The glue utilised to join the layers was hide glue. As for the wood species, Heikamp [8] reported in 1966 that it had been identified by Prof. Romano Gellini (Institute for Agricultural and Forestry Botany, University of Florence) as Poplar; however, he did not specify the method used for such identification. One of the Authors of this paper (M. Fioravanti) confirmed such identification as Poplar (Populus alba L.) by non-invasive thorough observation of a wood surface that was temporarily uncovered during the restoration [9]. 2.2 Making the Mock-Up The mock-up was made to understand both the manufacturing process and the hygroscopic and mechanical behaviour of this kind of shields. Before starting this work (in 2002) a thorough bibliographic research was carried out to gather any information on

222

P. Mazzanti et al.

Fig. 1. A few scans from a Computed Tomography of the Medusa shield: (a) front view of the shield, the white spots are the original nails which were used for fixing the handling straps on the rear; (b) a cross-section of almost the whole shield, the two superimposed wooden layers are visible; the concave shape below is the sliding bed on which the patient to be scanned - in this case the Shield – is placed; (c) detail of a cross-section, clearly showing the two layers of wooden lamellae: the patterns of the annual rings are visible in the external layer because the cross-section is perpendicular to the wood’s grain, whereas no such pattern is visible in the internal one because the cross-section is parallel to the wood’s grain.

Fig. 2. An exploded view of the wooden structure of the Medusa shield (the growth rings shown are purely fictitious), reconstructed by a 3D CAD software.

the construction of these wooden shields; however, no information were retrieved at that time. Therefore, some hypotheses on the manufacturing technique have been thoughtfully made and applied [6]. Later, in 2011, a few information were mentioned in [4], which reported similarities with the method already applied by the Authors. Of great importance was the structural analysis carried out during the restoration campaign (see Sect. 2.1), which made it possible to build the mock-up model through a sort of a reverse engineering process. Firstly, several roughly rectangular oversized slabs were obtained by sawing along the grain some seasoned boards of Poplar (Populus alba L.) wood supplied by local trade; then individual lamellae were planed to 6 mm thickness and sawn slightly oversize to the shapes identified through the CT scans. After some preliminary tests (including boiling and directly shaping the lamellae on a spherical mould – effective but too time-consuming), it was decided to initially curve each lamella with a single longitudinal curvature having a radius of approximately 403 mm, the same as the shield; to this end a mould and a counter-mould were obtained at the same time by sawing from an 80 mm thick wooden plank a cylindrical segment, having a slightly smaller radius to consider the elastic spring-back of the lamellae.

The Medusa Parade Shield by Caravaggio

223

The lamellae were soaked during 12 h in water at room temperature (15–20 °C), then were individually clamped between mould and counter-mould tightened against each other by means of a carpenter’s clamp and left there to partially dry during 30 min under the action of a jet of hot air from a hairdryer (similar results could likely be obtained by leaving the clamped assembly near a wood fire). After having removed the partially dried and curved lamella from the mould, the elastic spring-back produced the desired longitudinal curvature. The still warm and slightly moist lamella was quickly and provisionally fixed on the assembling wooden mould having the shape of a spherical cap with the desired radius of 403 mm, using nails driven in only halfway and folded in such a way as to keep it tightly fitting against the mould. Thus, the lamella was obliged to fully adapt to the sphere’s double curvature, which was quite easy because of its limited width. Starting from the central ones, thus all the eight lamellae of the internal layer were applied on the spherical mould; before being provisionally nailed each lamella was given its final shape and angle on the edges with a hand-held smoothing plane, to fit closely to the adjacent ones. When the internal layer was completed, minor surface irregularities were removed with a well sharpened hand chisel. Then the external layer was applied perpendicular to the first one, starting again from the central lamella and adopting the same procedure, except that it was glued to the first one by using traditional hide hot glue. Before gluing each lamella, only the corresponding provisional nails of the first layer were removed and new nails were partially driven in and folded with the same technique, so to keep firmly in place both layers. When the replica was completed, it was moved in a climatic chamber at standard conditions (65% RH and 20 °C) still nailed on the mould, and left there for approximately 30 days to reach its equilibrium moisture content. Two perpendicular “meridians” (roughly corresponding to the longitudinal directions of the central lamellae of the two layers) and an arbitrarily chosen “parallel” (near to the shield periphery) were then drawn, and several stainless-steel threaded screws with hexagon socket (Allen screws) were applied to be used as reference points for successive measurements (see Fig. 3). Finally, the shield’s periphery was trimmed by sawing the shell’s base along a parallel circle and minor surface irregularities were removed with the hand chisel. After completion of the first set of measurements, all the nails were extracted and the mock-up (which until then had been constrained to maintain a spherical shape) was removed from the mould and started deforming: instantly due to elastic spring-back, subsequently due to viscoelastic recovery. 2.3 Monitoring Methods and Apparatus Manual Measurements The manual measurements on the shield mock-up were performed using as reference points the already mentioned Allen screws applied along the two perpendicular “meridians” (in fact corresponding to the longer and the shorter diameter of the deformed shell) and on the arbitrarily chosen “parallel” near to the shield’s periphery. To perform properly the measurement regardless of the shield’s distortion, a digital calliper (Mitutoyo Absolute IP66, 0.01 mm resolution, 0.02 mm accuracy) was equipped with purpose-made spherical feelers, fitting freely without any slack into the hexagon socket

224

P. Mazzanti et al.

of the Allen screws, so the reference points were materialized by the centres of the steel spheres. The measurements were manually repeated in different climatic and constraint conditions.

Fig. 3. The mock-up still constrained by nails on the spherical mould, with Allen screws placed along the two perpendicular “meridians” and the “parallel”, to be used as reference points for the manual measurements.

Method and Apparatus for Monitoring Distortions The method applied to automatically monitor the mock-up’s distortions is based on a) allowing it to deform freely under mechanical or hygroscopic stresses and b) measuring the distance variations between selected reference points. On the horizontal plane, the length variations of the two main perpendicular diameters were directly measured; whereas the vertical displacement of the top central point of the mock-up (its “pole”, so to say) was measured in relation to the supporting diameter (the longer or the shorter one, depending on the case) on which the mock-up rested while tested. To automatically perform such measurements a specific apparatus was designed at DAGRI and made in a local workshop. The basic idea was to have the ends of the supporting diameter resting on spherical hinges, freely running along a guide rail. To prevent the mock-up from overturning a third spherical hinge (free to slide vertically but prevented from moving horizontally) inhibited any horizontal displacement of the top of the mock-up, which was hence externally isostatic and free to distort. This apparatus, named Shield Measurement Apparatus (SMA, see Fig. 4) was mounted on a thick fibreboard panel, and was made of some commercial components and some special parts made on purpose, designed at DAGRI.

The Medusa Parade Shield by Caravaggio

225

Details of the SMA Apparatus Each support at the end of the supporting diameter (alternatively the shorter or the longer one) was conceived and built to behave as a sliding spherical hinge, and was made of a spherical plain bearing holding a special clamp, shaped to be fixed on the periphery of the mock-up. The bearing’s base was fixed on a runner block freely sliding on a guide rail. A displacement potentiometric transducer (Penny&Giles, SLS190, resolution virtually infinite, independent linearity ±0.25%) placed between the two blocks monitored the distance between the supporting diameter’s ends. The whole system was designed so that the bearing’s centre of rotation fell exactly on the periphery of the mock-up. Two different couples of clamps were made available, featuring appropriate angles (22° and 66°, respectively) to fit the slope of the mock-up being tested, alternatively, when resting on the longer or on the shorter diameter (see detail A in Fig. 4).

A

C

D

B

E

Fig. 4. The Shield Measurement Apparatus (SMA) is here represented with a 3D CAD drawing. Detail A shows the supporting point’s spherical hinge, made of a roller (a runner block bearing mounted on a guide rail, detail D) and a pinned support (a spherical bearing mounted on the roller) to allow the mock-up’s free displacement and distortion. The aluminium clamps supporting the mock-up were made with two different slopes to fit the shape of the mock-up at the two main diameters. Detail B shows the spherical elements mounted on the ends of the free diameter, made to contact flat elements connected to the vertical and the horizontal displacement transducers. Detail C shows the upper hinge, preventing the mock-up from overturning and connected to a third vertical displacement transducer.

The ends of the free (non-supporting) diameter were equipped with spherical contact elements, against which two opposite vertical plates were gently kept in contact by a tension spring. These, like the supports described in the previous detail A, were fixed on blocks sliding along a guide (see detail D in Fig. 4) perpendicular to the previous one, so that the distance variations between the free diameter’s ends could be monitored by a displacement transducer, identical to the previous one. Above the extremes of the same

226

P. Mazzanti et al.

free diameter two other displacement transducers (Penny&Giles, SLS095, resolution virtually infinite, independent linearity ±0.25%) were vertically fixed, the feelers of which consisted of flat surfaces resting on the corresponding spherical elements, so that also their vertical displacements were monitored (see detail B in Fig. 4). The upper spherical hinge consisted of a steel ball housed between a support plate fixed on the top of the mock-up, and a metal rod free to slide vertically, but allowing no horizontal displacement; a 0,5 kg mass weighing on the bar eliminated the slacks of the restraint system and prevented the ball from coming out of its seat. A third displacement transducer (Penny&Giles, SLS095, resolution virtually infinite, independent linearity ±0.25%) monitored the vertical displacement of the rod, and hence of the mock-up’s top (see detail C in Fig. 4). To prevent the wood from getting indented, large aluminium contact areas were provided at all the supporting/clamped points. Through the above mentioned five displacement transducers both the horizontal and vertical changes of shape were monitored; a PACE XR-8X-SE (accuracy ±0.25%) data logger was used for both supplying the transducers and recording the data. Environmental Conditions The mock-up underwent various environmental climatic variations: a) manufacturing conditions: no environmental control, MC (Moisture Content of wood) ≥ FSP (Fibre Saturation Point); b) equilibration period at 65% RH and 20 °C (typical EMC – Equilibrium Moisture Content of wood) ≈ 12%; c) successive equilibration period at 40% RH and 30 °C (typical EMC ≈ 7%); d) successive re-equilibration period at 65% RH and 20 °C (typical EMC ≈ 12%). The tests were carried out in a climatic box where RH was controlled by a humidity controller apparatus, in turn placed in a climatic chamber where T was controlled. The climatic parameters were recorded every 15 min by means of an Onset HOBO U12-013 (resolution: 0.03 °C at 25 °C and 0.05% RH, accuracy: ±0.35 °C from 0° to 50 °C and ±2.5% from 10% to 90% RH typical).

3 Results and Discussions 3.1 Hygroscopic Behaviour When the mock-up was fabricated by nailing the lamellae on the spherical mould and gluing them together its MC was above and not far from the FSP. When it was removed from the mould it was in equilibrium with air at 65% RH and 20 °C. While nailed on the mould it was forced to remain spherical in shape; however, once nails were removed the internal stresses were no more balanced by their mechanical constraint, and finding a new equilibrium caused a change in shape. Thus, the mock-up reached a new equilibrium shape similar to a tortoise shell, characterised by the shorter diameter being oriented parallel to the grain at the top of the external layer, and the longer diameter perpendicular

The Medusa Parade Shield by Caravaggio

227

to that same grain direction. Moreover, its perimeter, which initially was circular and laid on a plane, was roughly elliptical and raised in correspondence with the shorter diameter. Such shape changes were amplified or reduced by the dehumidification/humidification cycles that followed (see Fig. 5); however, since then the mock-up maintained a tortoiseshell shape. Other original shields very likely featuring the same wooden structure show the same shape.

nailed on the mould immediately after being removed from the mould stabilized at 65 % RH and 20 °C stabilized at 40 % RH and 30 °C

4≡13 5

10 11

3 14

2

9 1

Heights [mm]

6 15

16

17

0

Measured points

Fig. 5. Schematic profiles of the mock-up shield recorded at selected significant points in time, along the longer (left) and the shorter (right) diameters. For simplicity, only the ordinates (i.e. the heights measured with respect to the respective diameter) are shown to scale, while in the abscissa the horizontal distances measured along the diameters are all shown at fixed values. When the RH decreases, the distortion increases: the sagitta, i.e. the height of the vertex (here named point 13) increases with respect to the diameter that decreases, while the sagitta at the same point (here named point 4) decreases with respect to the diameter that increases

The tests concerning hygroscopic distortions following the imposed environmental humidity changes were carried out on the mock-up placed horizontally on the SMA, resting on two supports placed at the two ends of its longer diameter, and continuously monitored through several displacement transducers (see Sect. 2.3), which allowed an accurate study of its hygroscopic behaviour. Processed data showed that when the RH decreases the tortoise-shell shape becomes more marked, whereas when the RH increases the mock-up tends towards a shape approaching a spherical cap. In Fig. 6, an example of an adsorption test is presented, with the mock-up supported on its longer diameter. Before the test began the mock-up was equilibrated in air at 40% RH and 30 °C, and a pronounced tortoise-shell shape was observed: the length difference between the two diameters was large, and the shorter diameter was in a lower position than the longer one. For the test the mock-up was exposed to more humid air, having 65% RH and 20 °C, and was supported at the ends of the longer diameter; the measured displacements confirmed the asymmetric hygroscopic behaviour of such a structure, with the longer diameter shortening and the shorter one lengthening. At the equilibrium, the short diameter’s lengthening was four times larger than the long one’s shortening.

228

P. Mazzanti et al.

In addition, in the vertical direction the shorter diameter had approached the longer one, uplifting almost twice as much as the mock-up’s top. Thus a more spherical cap’s shape was reached. When an opposite desorption cycle was imposed, the reverse behaviour was observed in terms of both modality and quantities, and the previous tortoise-shell shape was again accentuated (graph not shown here). A possible intuitive interpretation of the above-described hygroscopic behaviour is discussed below, based on both the shape and the internal structure of this mock-up. The Shape-Based Geometric Effect. For a thin and low-height spherical cap, regardless of the internal structure of its constituting material (provided it is at least a little flexible), it is intuitive (and easily verifiable e.g. on a spherical cap cut from a rubber ball) that when it rests on two opposite points of its circular perimeter (i.e. the ends of a diameter) a vertical displacement applied on its top will result in the two main perpendicular diameters having opposite variations in length, and moving away vertically from each other. The diameter defined by the two supporting points will lengthen, whereas the perpendicular one will shorten; such distortion taking place with negligible length variations on the cap surface. Since our mock-up can be considered a slightly distorted spherical cap, the same behaviour can be expected from it, and remain valid for distortions of both mechanical and hygroscopic origin. Elongaon of the short diameter Not supported

15 13

Upliing of the ends of the diameter

11

Displacement [mm]

9 7 5

Upliing of the top

3 1

0.00001

0.0001

0.001

0.01

0.1

-1 1

Time [log gg] -3

10

100

1000

Shortening of the long diameter Supported

-5

Fig. 6. The graph shows the hygroscopic behavior of the mock-up undergoing an RH variation from 40% (T = 20 °C) to 65% (T = 30 °C); the test was carried out with the longer diameter supported. The blue and the yellow lines represent the length variation of the longer and of the shorter diameter, respectively. The shorter diameter lengthens while the longer one shortens; since the longer diameter shortens 5 times more than the shorter one lengthens, the shape of the mockup becomes more spherical. At the same time in the vertical plane, the shorter diameter uplifts approximately 2 times more than the top, thus the mock-up’s perimeter tends to flatten.

Mechanical tests have also been carried out on our mock-up to explore and confirm such behaviour; however, they are not reported in this paper due to space constraints.

The Medusa Parade Shield by Caravaggio

229

The Internal Structure-Based Hygroscopic Effect. Our mock-up can be considered a thin and low-height spherical cap, made of two layers of wood with mutually perpendicular grain. Due to the well-known wood anisotropy, when RH (and hence their Moisture Content MC) increases all the lamellae tend to swell in their own transverse anatomical direction by a much greater amount than in the longitudinal direction, the latter being practically negligible. Let us virtually separate two narrow arches having the full thickness of the mock-up, each centered on one of the two mutually perpendicular meridians oriented according to the longitudinal directions of the two layers of wooden lamellae; and let us call LongExt the arch parallel to the external lamellae, and LongInt the one parallel to the internal lamellae. Figure 7 shows the behaviour of such a mock-up brought to reach hygroscopic equilibrium first in drier air, then in more humid air, in which the MC of wood increases, and hence the wood tends to swell: – the length of the arch LongExt tends to increase by a negligible amount in the external layer (having longitudinal anatomical direction), and to increase significantly in the internal layer (having transversal anatomical direction); since the two layers are glued together, this differential elongation results in a tendency to increase the radius of curvature of the arch (in other words: to straighten, thus reducing its sagitta, i.e. its height), and therefore to increase the corresponding diameter; – vice versa in the arch LongInt, perpendicular to LongExt, the differential elongation is inverted, and in short, the corresponding diameter tends to decrease; – in the areas of the mock-up progressively further away from the two main meridians, the longitudinal directions of the two layers form angles smaller than a right angle, so that their distortion behavior progressively moves away from one of those described above, and progressively approaches the other. The opposite behaviour can be observed when the RH decreases. Thus, the wooden structure, when made of only two crossed layers, strongly affects the shape variation of the mock-up and of the original parade shields.

65 % RH

LONG DIAMETER

SHORT DIAMETER 40 % RH

Fig. 7. The hygro-mechanical distortion behaviour of the mock-up when RH increases. The mockup cross-section is represented by a dashed line at lower RH and by a solid line at higher RH. The cross-section parallel to the top lamellae of the external layer (shorter diameter, right) shows diameter decreasing and sagitta increasing, whereas the cross-section parallel to the lamellae of the internal layer (longer diameter, left) shows diameter increasing and sagitta decreasing.

230

P. Mazzanti et al.

3.2 Modelling the Hygroscopic Behaviour of the Shield’s Mock-Up The shield is a mechanically complex body and it is difficult to construct a predictive model of it. We can also think that for reasons related to the variability of the characteristics of the wood, the specific construction techniques adopted, the phenomena of damage and aging over the centuries, our mock up is only partially representative of the real Medusa’s shield; similar considerations lead us to believe that any model can only give qualitative information. The objectives of the numerical modelling of the shield were therefore: 1. to verify that the anisotropy of the wood brings a shield initially having the shape of a spherical cap when its MC is above the FSP, to a shape similar to that of a turtle shell, found in the original shield and obtained in our mock-up when MC becomes lower, 2. to evaluate how much the distortion resulting from a linear elastic modelling differs from the distortion measured on our mock-up, to understand if the distortion effects we observe are due to important non-linear not reversible phenomena, and 3. the numerical model presented here deals only with distortions originating from hygroscopic variations. In a future development it might be used to explain the distortions originating from mechanical external stresses, mentioned in Sect. 3.1. The modelling of the hygroscopic behaviour of the shield’s mock-up was performed by the Ansys numerical finite elements code and based on a hygro-mechanical chaining based on an isotropic formulation of Fick’s law of diffusion, whith the following premises: 1. the geometry of our mock-up is defined as a spherical cap with the sphere’s radius 403 mm, and the radius of its base 288 mm; the mock-up is considered symmetrical with respect to the two main perpendicular diameters, so that four equal portions/area/section were obtained and the modelling was developed taking into account this double symmetry; 2. the cylindrical coordinates were defined, one for each layer of the mock-up; unfortunately, no prevailing anatomical direction can be identified, so an average direction is considered; moreover, the hygroscopic parameters used were obtained from literature data, because no specific tests were carried out on this wood; 3. the numerical model was developed on a hexahedral mesh, with discretization of 2 mm, of linear elements for the diffusion computation, second order for the chained mechanical computation; 4. the boundary conditions were defined according to the double symmetry discussed above, to prevent rigid body motions; 5. for the modelling, a 30% MC value was imposed as initial condition, representing the state of FSP induced by the gluing process; then, the EMC, calculated from climatic conditions of the test, was imposed on the internal and external surfaces of the numerical model immediately after the RH variation, to make a comparison possible between the numerical model and the experimental tests.

The Medusa Parade Shield by Caravaggio

231

Modelling from Fibre Saturation Point to EMC with 45% RH The first phase of the modelling was to interpret the dehumidifying process from FSP to the experimental condition of EMC with 40% RH and 30 °C (see environmental conditions at the end of Sect. 2.3). The saturation state occurred at least once, when the mock-up was assembled. Indeed the application of hide glue for gluing together the two perpendicular wooden layers required a large amount of water, which easily saturated the thin wooden lamellae. The simulation included a period of 30 days and a MC variation of −12.95% based on Merakeb’s formulation of the isotherms [10]. At the initial moment the mock-up was considered a perfect spherical cup with a base diameter of 563. 8 mm, as when it was nailed on its mould. As the RH varied, the numerical model began to deform, and after the 30 days of simulation, it reached its new shape with a major diameter of 578.5 mm and a minor diameter of 534.4 mm. However, the actual dimensions of the mock-up were 557 mm for the major diameter and 485 mm for the minor one, when the same RH variation and time interval were considered. The large difference in dimensions between the experimental test and the calculation results implies that the model, which is a linear model, did not correctly interpret the mechanical behaviour of the mock-up. Apparently, the non-linear phenomena are prevalent and could have caused some plastic deformations and possibly structural damages; in fact in the external and in the internal layer of the mock-up multiple traction micro fractures are observable. The Unsteady State/Short Term Modelling The following step was modelling the transient distortion of the two main diameters for a humidity change from 50% to 65% and a moisture variation of +1.8%. Once again this linear model has not proved to be adequate to interpret the hygroscopic distortion of the mock-up. Indeed, the calculated displacements of the two main diameters at equilibrium were not consistent with the experimental data (Table 1). Table 1. The experimental data of an adsorption test are compared to the data from the numerical model. The model shows a deviation from the experimental data possibly due to the presence of open-joints and cracks on the lamellae, which modify the diffusion behaviour of the shield. Longer diameter [mm]

Shorter diameter

Experimental data

−1.7

9.4

Modelling

−2.3

5.7

The same behaviour shows up when the transient phase is observed. When the first 80 h are analysed, the deviation of the model from the experimental data is particularly large for the shorter diameter, whose displacement is double compared to the numerical model. Such difference can be attributed to the presence of many cracks on the wooden lamellae or to the open-joints (disconnections between two adjacent lamellae), which strongly contribute to make the real diffusion far from an unidirectional model and faster as deep layers of wood are directly exposed to external humidity.

232

P. Mazzanti et al.

The Long Term Modelling Thus, in order to analyse a linear behaviour and to avoid the diffusion problems, a model was performed on an experimental desorption test (from 65% to 44% RH, 20 °C, see Fig. 10).

Fig. 10. Numerical results, described in Table 2, for the displacement in mm of the two main diameters. In the simulation with the purpose to reduce the degrees of freedom of the system it has been exploited the double geometric and hygro-mechanic symmetry of the system, arranging opportune boundary conditions both for hygroscopic and mechanical computations.

For this test, the MC variation was −2.6% and the results are listed in Table 2. Table 2. The experimental results of an adsorption test, compared with those of the numerical model, indicate that the model is consistent with the experimental data. Longer diameter [mm]

Shorter diameter [mm]

Experimental data

1.4

−6.3

Modelling

3.0

−5.9

This simulation is consistent with the experimental data in terms of absolute error and physical behaviour, even if the longer diameter shows a significant difference from the data in terms of relative error. The conservative consequence of finite element modelling can therefore be summarised in the following points: 1. For small moisture variations the distortion behaviour of the shield is reversible and well represented by a linear model. 2. Significant variations in humidity are associated with important internal stresses that have led to irreversible deformation and damage phenomena. 3. In general, diffusion analysis methods based on the non-cracked condition are not capable of representing the transitional phases.

4 Conclusions The conservation of the parade shields, such as the well-known Medusa painted by Caravaggio, can benefit from the knowledge of their structure and construction features,

The Medusa Parade Shield by Caravaggio

233

and of their distortion behaviour. To improve such knowledge an accurate structural replica of the Medusa shield was built, based on a sort of reverse engineering process. The direct observation and several TC scans of the original Medusa shield provided essential information on its structure and geometry, which has approximately the shape of a spherical cap peculiarly distorted, somehow like a turtle shell. TC scans clearly showed its internal structure, made of thin spindle-shaped wooden lamellae organised in two layers, glued together according to crossed directions. No information about the technique originally implemented to manufacture such wooden artworks could be found in the literature, so within the framework of the work described here a technique was developed to manufacture the structural replica; such technique has proven to be effective, and compatible with the technologies widely available in the 15th century. The mock-up then underwent several laboratory tests and measurements, by means of methods and equipment developed on purpose for this research, to ensure it could freely deform under the action of climatic or mechanical stresses. For the climatic tests the shape measurements took place – after hygroscopic equilibrium was reached – in a climatic chamber with wetter and drier climatic conditions. The test results highlighted the distortion behaviour of the mock-up, and hence presumably of the original shields featuring the same internal structure, under the action of environmental variations. Such behaviour, substantially consisting of a greater or lesser accentuation of the turtle-shape distortion, has been explained with reference to the shield’s shape and structure, and in the light of basic Wood Technology knowledge. Finally, a numerical modelling of the hygroscopic distortion behaviour of the shield’s mock-up was performed by the Ansys numerical finite elements code, based on a hygro-mechanical chaining and on an isotropic formulation of Fick’s law of diffusion. Such modelling showed how the spherical shape and the two crossed wood layers enhance the anisotropic deformation components, generating large internal stresses which can produce plasticization and damage in the material, namely for large RH variations; the fact that the results from linear modelling fit well the experimental data only for small humidity variations (in the range of RH 65%–44%) indicates that in this range distortion phenomena are generally reversible and no damage should occur. However, the model shows that in general the shape of this kind of artwork makes them very vulnerable to environmental thermo-hygrometric variations and very fragile. Acknowledgements. The Authors gratefully acknowledge Liye Deng, Enrico Innocenti, Elisa Millacci and Vincenzo Sferra, who carried out some of the tests and processing reported in this paper as part of the requirements for obtaining their degrees.

References 1. Victoria and Albert Museum. http://collections.vam.ac.uk/item/O39946/parade-shield-unk nown/. Accessed 30 Nov 2021 2. Sanders, K., Harrison, L., Higgitt, C., Young, C.: Purely decorative? Technical analysis of a fifteenth-century northern European parade shield. Stud. Conserv. 57(Suppl. 1), S268–S278 (2012). https://doi.org/10.1179/2047058412Y.0000000019 3. Caneva, C. (ed.): La Medusa del Caravaggio restaurata. Retablo Cultura Arte Immagine, Rome (2002)

234

P. Mazzanti et al.

4. Zoffili, E. (ed.): The First Medusa. 5 Continents Edition, Milan (2011) 5. Favaro, M., Vigato, P.A., Andreotti, A., Colombini, M.P.: La Medusa by Caravaggio: characterisation of the painting technique and evaluation of the state of conservation. J. Cult. Herit. 6, 295–305 (2005) 6. Uzielli, L., Cardinali, E., Dionisi Vici, P., Fioravanti, M., Salvioli, N.: Structure, mockup model and environment-induced deformations of Italian laminated wood parade shields from the 16th century. In: Gril, J. (ed.) Wood Science for Conservation Braga 2008 COST Action IE0601, LNCS, pp. 242–248. Firenze University Press, Florence (2010) 7. Dionisi Vici, P., Fioravanti, M., Uzielli, L.: La struttura lignea dello Scudo In: Caneva, C. (ed.) La Medusa del Caravaggio restaurata, pp. 161–168. Retablo, Rome (2002) 8. Heikamp, D.: La Medusa del Caravaggio e l’armatura degli Scià Abbâs di Persia, “Paragone”, XVII, 199, pp. 62–76, September 1966 9. Signorini, G., Di Giulio, G., Fioravanti, M.: Il legno nei Beni Culturali, guida alla determinazione della specie legnosa. Aguaplano, Perugia (2014). ISBN 9788897738404 10. Merakeb, S., Dubois, F., Petit, C.: Modeling of the sorption hysteresis for wood. Wood Sci. Technol. 43(7–8), 575–589 (2009)

Innovation in Precision Low-Energy Heat Transfer Using Flexible Heating Mats for Targeted Treatments in Paintings and Paper Conservation Tomas Markevicius1,2(B)

, Nina Olsson2 , Yuhui Liu3 , Maddalena Magnani4 and Martina Paganin5,6

1 University of Amsterdam, Amsterdam, The Netherlands

[email protected] 2 Precision Mat, LLC, Portland, OR, USA 3 Institute of Conservation IBR, Bavarian State Library, Munich, Germany [email protected] 4 Università degli studi di Torino, Turin, Italy 5 Accademia di Belle Arti di Bologna, Bologna, Italy 6 London, UK

Abstract. The paper discusses innovative approaches and targeted structural treatments of works on paper and paintings made possible by temperature management technologies based on low-energy flexible mats for precision heat transfer. Flexible silicone MAT (Multipurpose Accurate Thermal Management, developed by the Precision Mat, LLC) laminates, and transparent carbon nanotube-based IMAT (Intelligent Mobile Multipurpose Accurate Thermoelectrical Device for Art Conservation, developed by IMAT project, Horizon Europe) prototypes with the associated mobile MAT and IMAT thermal management consoles were ideated and designed specifically for the field to offer accuracy, mobility and smart nanotech for new CH remediation approaches over the unreliable traditional tools of the past, such as hand irons and heating tables, with the benefit of precision, steadiness, uniformity, and control in heat transfer from ambient to the customary temperatures used in heat activation (25 °C−65 °C). The varied MAT and IMAT mat dimensions, their flexible and thin profile combined with accuracy in the low temperature range allow conservators in diverse fields to formulate novel targeted treatments that exploit the effects of sustained precision mild heat transfer over time without the risks and unnecessary stress related to uncontrolled heat-transfer of the past. Case studies examine the application of flexible warming mats in diverse treatments of works on paper, and photographs conducted by the authors. The operational parameters and practical advantages offered by the low-energy heat transfer nanotechnology and targeted approaches taken in each particular treatment show the broad versatility of the new heat transfer method and how easily it could be tailored for the specific needs of each particular case, opening new opportunities for art conservators to refine their treatments within the margins of minimal intervention and risk. Keywords: Low energy heat transfer · Thermal management · Paintings conservation · Book & paper conservation · Photographic materials · Modern © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 235–253, 2023. https://doi.org/10.1007/978-3-031-17594-7_18

,

236

T. Markevicius et al. materials · Structural treatments · Nanotubes · Green technology · Heritage science and technology · Sustainability

1 Introduction Conservation treatment methodologies in both paintings and paper conservation have traditionally exploited heat as an essential factor for the effective remediation of structural damage, planar deformations, consolidation of the paint and the substrate in paintings, works on paper and other cultural heritage objects. Historically, and up to the present day, diverse methods of heat transfer have been used, such as hand-held irons, heated spatulas, sandbags, water bags, hair dryers, heat guns, infrared lamps, custom-fabricated hot tables, and multipurpose heating tables from the 1980s on. It must be emphasized, that diverse past (and present) interventions involving heat were not without considerable risk, since heat was applied with rudimentary tools that provided quite limited control over the set temperature, the steadiness of delivery, and uniformity of distribution over treatment surface area, that led to highly undesirable results ranging from incomplete treatments to irreversible changes in surface morphology of the paint and ground layers. Most electrical hand irons are not precision tools, as they lack steadiness in heat transfer, due to the regulation by thermostatic on-off mechanisms that deliver only an average set temperature, which in actuality is a fluctuating series of over and undershoots which trigger the on-off function. Variance as high as ±15 °C were observed on hand irons (Fig. 1).

Fig. 1. Heating pattern of widely used handheld irons, regulated by on-off switch which delivers fluctuating heat with 30 °C swings. The heating pattern was measured using a type-T thermocouple and IMAT prototype.

Innovation in Precision Low-Energy Heat Transfer

237

This lack of steadiness is even more problematic when the conservator attempts to heat a larger area with sweeping movement of the iron, which necessitates the set temperature to be considerably higher than the desired surface temperature. Furthermore, most hand irons are designed for household textiles with operational parameters between 120 °C−180 °C, and therefore do not have settings in the safe and low ambient temperature range for paint films. Other temperature controls such as rheostats are based on energy output and are not corrected by local temperature readings, and the actual surface temperature remains unknown. The risks and beneficial effects of heat transfer in conservation treatments must be assessed within a comprehensive understanding of the nature of treated objects as complex physical systems, which is crucial to evaluate their long-term stability and to formulate treatment protocols. In 1991, Mecklenburg published experimental data indicating that artist paints are viscoelastic systems (Mecklenburg and Tumosa 1991). The viscoelastic behavior of paintings has been investigated in the context of their cleaning and mechanical properties (Hedley and Odlyha 1989), Michalski (Michalski 1991), Hagan (Hagan 2006, 2017), Phenix (Phenix 2011); the effects of temperature and humidity have been investigated in contexts of art transportation and preventive conservation. However, the effects of temperature and viscoelastic behavior were less investigated in the context of structural treatments of paintings on canvas. These aspects were addressed by Berger and Russel (Russell and Berger 1982; Berger and Russell 1984, 1986, 1988) and Goddard (Goddard 1989), Olsson and Markeviˇcius (Olsson and Markeviˇcius 2010, 2013, 2017, 2018), Magnani (2021), Magnani et al. (2022), Paganin et al. (2021), Percival-Prescott (2003). Once fully dry, paint films behave as elastic, viscoelastic, or viscoplastic materials depending on the chemical nature of their components. With aging and in variation with the effects of pigments, oil films acquire a higher degree of crosslinking and stiffness with a corresponding rise in glass transition temperature from ambient temperature (0– 50 °C) to as high as 75–100 °C for paints containing lead white (Phenix 2011). Glass transition temperatures of acrylic, alkyd and oil primers were found to be in the range 21–44˚ C (Hagan et al. 2007). In structural conservation treatments of paintings canvas, the constituent viscoelastic materials such as paint films may suffer structural failure from the rapid application of force to paint films while in the glassy state. However, it is possible to shift from a strain causing failure to one avoiding failure by slowing the rate of deformation (1), adding moisture or other agent acting as a plasticizer (2), or increasing the amount of energy in the system, i.e. temperature (3). The relationship of stress and strain, described as Young’s modulus is influenced by plasticizers, such as water (i.e. humidity treatments) and temperature. Stiffness may be reduced by precisely controlling temperature and the rate of deformation. In the former, temperature control becomes paramount, as beneficial heat effects cannot be safely exploited without the precision instrumentation, discussed in this paper.

238

T. Markevicius et al.

2 IMAT and Flexible Mat Technology and Applications Historically used and currently available heat transfer instrumentation was designed for non-conservation applications and lacks accuracy in the low temperature range, steadiness, and uniformity, which limits options and treatment approaches. The surface areas of heat transfer instruments are limited to the size of spatula heads, tacking irons, hand irons, and heating tables, with no alternatives in between. In response to the gaps in currently available instrumentation, a precision heat transfer technology in the form of flexible silicone-clad heating mats and associated temperature control consoles was developed specifically for art conservation applications. Since 2003, Olsson and Markevicius have experimented and worked to develop a new approach using precision low-energy heat transfer mats and associated temperature control consoles (Olsson and Markevicius 2010, 2013, 2017, 2018, 2022). In 2010, the authors first proposed using carbon nanotubes (CNT) for precision heat transfer (Markevicius et al. 2011).

Fig. 2. IMAT system with flexible heating mat, wireless Bluetooth thermocouple, powered by the temperature control console with touchscreen. It is composed of a custom electrical textile with nanocarbon yarns, laminated with clear silicone. The mat is connected to a Proportional Integral Derivative or PID controller, considered the most accurate for conservation application. The PID controller takes readings up to 40 times per second and self-corrects, providing the ultimate in even and steady heating pattern.

Innovation in Precision Low-Energy Heat Transfer

239

From 2011–2014 European Commission’s IMAT Project (Intelligent Mobile Multipurpose Accurate Thermo-Electrical Device, call ID: 283110), coordinated by the University of Florence advanced and refined the design of flexible low-energy heating laminates for cultural heritage conservation (Fig. 2). In order to enable the adoption of precision heat transfer in conservation practice, available wound wire technology has been again employed in the design of a series of silicone heating mats and associated consoles for conservation used in 2019, to launch the production of MAT (Mobile Accurate Thermal Management for Art Conservation) system production in 2022 (Fig. 3). The operational parameters and design features have been further adapted to prioritize performance in the low temperature range, portability, and use of standard domestic power input. The thermocouples are mounted externally. The mat dimensions range from 5 × 13 cm to 64 × 76 cm which function at 120 V for North American use or 240 V for EU use, while the largest mat (76 × 103 cm) is designed only for EU 240 V power input. Overall, the mats provide the conservator with novel advantageous temperature control in the low temperature range (ambient to 40 °C) making prolonged heat transfer possible to permit a “low and slow” approach, which allows conservators to innovate safe and nuanced treatments of both lined and unlined paintings. The mats are designed to deliver heat to the temperatures slightly above those used customarily to activate thermoplastic adhesives (to 70 °C).

Fig. 3. MAT precision heat transfer system.

For the novel treatment methodology, heating mat systems were designed to provide an accurate, versatile, and mobile heat transfer technology, and are all based on three key components: the heating mat, the thermocouple (TC), and the temperature control console. The heating mat is the thermal output surface. The mats are laminates composed of resistive elements, embedded within an exterior cladding of vulcanized silicone. The resulting heating mats are flexible, with a non-tack surface and thin profile. While flexible mats were proposed originally for structural treatment of paintings (Olsson and Markeviˇcius 2010, 2013, 2017, 2022), more recently, they have begun to find broader application in other fields, and were adopted for conservation of plastics (Krumrine 2019), paper and rare books (Liu 2019), photographic materials (Paganin

240

T. Markevicius et al.

2021), and modern materials, discussed further. Since 2003, flexible mats have been used in numerous lining treatments in paintings conservation: in combination with low pressure vacuum envelopes for lining of a 17th c. painting by Orazio Gentileschi (Olsson and Markeviˇcius 2017), and a series of six paintings by Kenneth Hudson (Markeviˇcius et al. 2017), and in the 2010 loomed treatment of Veronese’s Petrobelli Altarpiece at the National Gallery of Canada (Olsson and Markeviˇcius 2010).

3 New Developments and Perspectives of Precision Heat Transfer in Conservation New cases studies include conservation treatments on paintings by N. Olsson at Nina Olsson Art Conservation in the United States, and research from three Master’s theses in Italy and Germany by M. Magnani at Università degli Studi di Torino in collaboration with Centro Conservazione e Restauro (CCR) “La Venaria Reale” on conservation of modern media, such as PVA and sand paint matrixes by Giulio Turcato; on temperature optimized treatments on bound manuscripts and large-scale paper objects by Y. Liu at Technical University of Munich in collaboration with Bavarian State Library, and innovative methods for the detachment of silver gelatin prints from the secondary support by M. Paganin at Accademia di Belle Arti di Bologna. 3.1 Precision Heat Transfer Mat Application for New Approach in Paintings Conservation Precision Heat Transfer: New Opportunities to Exploit Time, Temperature and Moisture Factors for Safer and more Effective Treatments. In practice, remedial treatment to reduce diverse planar distortions such as cupped paint, and surface deformations of the paint, ground and canvas layers may be conducted by exploiting their viscoelastic properties with controlled increase in temperature to cause the paint film to transition from a glassy state to a pliable leathery state when pressure may be applied safely and effectively. The transition temperature may be reduced by introduction of a plasticizer such as humidity. Accuracy is critical for safe and effective heat treatment, to maintain the set temperature within the Tg range to achieve the transition to a compliant state, while avoiding overheating the paint film where unwanted deformation may occur. Thick or aged films may require a longer period of heat activation to achieve the even warming throughout the entire painting stratigraphy. With aging and in variation from the effects of pigments, oil films acquire a higher degree of crosslinking and stiffness with a corresponding rise in glass transition temperature from ambient temperature 0–50 °C to as high as 75–100 °C for paints containing lead white (Phenix 2011). Therefore, in many instances the ideal operational temperature will be in the low or ambient temperature range, and in some cases even a small viscoelastic response can be significant.

Innovation in Precision Low-Energy Heat Transfer

241

Remedial Treatment of Beva Lined Modern Paintings. Robert Motherwell’s, Open No.16 in Ultramarine with Charcoal Line (1968) (acrylic on canvas, 252.7 × 473.7 cm, Dedalus Foundation), was damaged during transport, leaving a 13 cm concave and dent in the center of the composition, and two areas of surface distortions in the lower corners.The work had been previously cold-lined with BEVA-Gel onto cotton canvas. A thin profile mat was introduced between the wood stretcher and the canvas verso and heated to 40 °C for 10 min to soften the lining adhesive and relax the canvas and paint layers in a safe range for acrylic films. Once softened and smoothed, the deformation sites were cooled between two heat sink plates held in place for 30 s from either side. The same procedure was used to smooth the large concave dent in the center (Markevicius et al. 2017) (Fig. 4). Fig. 4. Robert Motherwell’s Open No.16 in Ultramarine with Charcoal Line (1968): mild heat is applied with a mat from the verso to treat planar deformations.

Treating Severe Cupping and Planar Distortions. John E. Stuart, Mt. St. Helens from a hill back of Portland, 1885, oil on canvas, 18 × 30 in. (45.7 × 76.2 cm), Oregon Historical Society Museum. Planar distortions from cracks and severe cupping had formed in the extremely brittle paint and ground layers, and the tacking edges were degraded at the return edge. The scope of structural treatment was to stabilize cracks, reduce or eliminate planar distortions, and reinforce the tacking edges with a strip lining. Lascaux P550 (20% in Naptha) was introduced into the cracks from the recto in an area with particularly unstable paint and ground cracks. Following full evaporation of the solvent, localized humidification with a small chamber preceded application of localized mild heat transfer at 30 °C for 40 min, which produced a pliable state where mild pressure was used for consolidation (Fig. 5). Once unstable areas were treated, the painting was removed from the stretcher for a general structural treatment. The same P550 resin was applied to the verso and allowed to dry. The verso was lightly humidified with a moistened blotter with an interleaf of Polartec® microporous membrane for 30 min. A custom cut platform of museum board and ½” foam core was created to support the fragile tacking edges. The painting was then placed in a

242

T. Markevicius et al.

Fig. 5. Localized low and slow heat transfer treatment at 30 °C for consolidation and remediation of cupping and planar distortions (left). Subsequent general treatment of painting using a lowpressure vacuum envelope, and mild heat transfer while supported with a custom platform to protect fragile tacking edges (right). Published with permission of Oregon Historical Society Museum.

low-pressure envelope and warmed to 40 °C for 40 min (Fig. 5), with replacement of museum board midway to capture introduced moisture. Finally, strip lining supports of crepeline were prepared with BEVA-film prior to bonding to the tacking edges, using an angled support to prevent flattening of the return edge.

Fig. 6. Schematic representation of the layers of the collage: [1] canvas (sized); [2] alkyd-based ground; [3] bitumen; [4] PVAc and sand; [5] carbon paper (carbon side); [6] sketch paper; [7] crêpe paper; [8] carbon paper (red waxed side).

New Approaches Treating Unconventional Modern Materials: Treatment of Giulio Turcato Collage on Canvas. 1970’S collage by Giulio Turcato (Intesa Sanpaolo Art Collection, Gallerie D’Italia, Milan). The collage is composed of carbon, crêpe, and sketch paper, glued to a layer of sand and PVAc (likely Vinavil®), over bitumen, and an alkyd-based ground1 . (Fig. 6). It is laid on a cotton canvas stretched on a wooden stretcher. The most challenging part of the treatment consisted in solving cracks and delamination in the bitumen layer. In the central area, critical delamination areas were cupped up to 4 mm high from the ground and were up to 15–20 cm wide. The PVAc and sand layer was very stiff, rigid, the bitumen was brittle, and the alkyd ground was stiff, while the papers were fragile and extremely thin (carbon paper −25 µm). Removal of the canvas from the stretcher was not possible, because of the rigidity of the complex canvas-paint matrix and risk to disrupt the fragile equilibrium of diverse constituent 1 M. Magnani Master’s thesis (2021) University of Turin, in collaboration with Centro

Conservazione e Restauro “La Venaria Reale”.

Innovation in Precision Low-Energy Heat Transfer

243

materials. Targeted heat transfer using flexible IMAT mat was deemed optimal to flatten the cupping of delaminating areas with 3D geometries. Evenly applied low temperature was crucial to avoid damages to heat sensitive materials, such as wax in carbon paper, and thermoplastic PVAc resin. The goal was to apply mild heat transfer, slowly reaching the temperature where the paint will become sufficiently elastic, and applying pressure to reduce the deformations, thus avoiding the risk of cracking the papers or the paint layers. Even, steady warming provided by the mat was essential to diffuse heat in the entire matrix. To verify the treatment concept, two warped Vinavil 59® (PVAc based) and sand layer mockups were prepared with different thicknesses (−0.7 mm and 1 mm), to test the efficacy and the correct time-temperature set to reduce their deformations, together with the application of weights (Figs. 7, 8 and 9). Table 1(A–B) shows the time-temperature settings used to flatten the deformations in the mockups. The time-temperature setting used in the actual treatment was similar: Table 1(C). Table 1. IMAT time-temperature settings used to flatten the deformations in the mockups Table A – IMAT settings for −0.7 mm mockup (using orange IMAT mat) Set

Time (min)

Temperature (°C)

Tr

5

22.4 (ambient T)

Ton

50

35

Tf

5

22.4 – (ambient T)

Table B – IMAT settings for −1 mm mockup (using orange IMAT mat) 1° CYCLE Set

Time (min)

Temperature (°C)

Tr

5

22.4 (ambient T)

Ton

60

35

Tf

5

22.4 – (ambient T)

Set

Time (min)

Temperature (°C)

Tr

5

22.4 (ambient T)

Ton

30

35

Tf

5

22.4 – (ambient T)

Tr

5

22.4 (ambient T)

Ton

30

35

2° CYCLE

3° CYCLE

(continued)

244

T. Markevicius et al. Table 1. (continued)

Table B – IMAT settings for −1 mm mockup (using orange IMAT mat) Tf

5

22.4 – (ambient T)

Table C – IMAT set on artwork (transparent mat) 1° CYCLE Tr

15

22.4 (ambient T)

Ton

20

35

Tf

15

22.4 – (ambient T)

Tr

20

22.4 (ambient T)

Ton

40

38

Tf

20

22.4 – (ambient T)

Tr

20

22.4 (ambient T)

Ton

40

40

Tf

20

22.4 – (ambient T)

2° CYCLE

3° CYCLE

Fig. 7. “Thin” mockup (left) and “thick” mockup (right) before treatment.

Fig. 8. “Thin” mockup (left) and “thick” mockup (right) during treatment.

The second cycle at a higher temperature was necessary because the first one was not fully effective2 , while the third one was made in an attempt to reduce cupping of delaminating paint. The cupping was measured after the treatment and was reduced by 2 PVAc constituting the artwork most likely has a higher glass transition temperature than the one

of the mockups, since it is a fifty years old material still undergoing a natural ageing process

Innovation in Precision Low-Energy Heat Transfer

245

Fig. 9. “Thin” mockup (left) and “thick” mockup (right) after treatment.

50% (Fig. 10). Treated areas were kept under the weight during paint consolidation, which was carried out using Eva Art, injected under the PVAc and sand layer through the cracks, using syringes with ultra-thin needles.

Fig. 10. One of the delaminated and cupped areas, before (left) and after (right) the treatment.

Treating Flour-Paste Lined 18th Century Painting. The painting by André Bouys A Woman Knitting, 1700 ca., oil on canvas, 90.8 × 71.4 cm, Portland Art Museum, Portland, Oregon) had been flour-paste lined in a prior treatment dating to the late 19th century, which was stable and well adhered. However, the recto had several areas of lifted, cupped and curled paint, caused by animal glue residues from the facing probably dating to the relining treatment. The recto was humidified through Polartec® microporous membrane with a moist blotter. Subsequently, mild heat of 36˚ C was applied to the recto for 40 min. Once the paint layers had transitioned to a compliant state, the cupped and curled paint was brought into plane with a silicone tipped heated spatula. Having regained planarity, the animal glue residues were removed from the paint layer. 3.2 Precision Heat Transfer Mat Application for a New Approach Detaching Historical Photographic Prints from Secondary Support The detachment of historical photographic prints from the mounting cardboard, have to be considered an extreme treatment and should be limited as much as possible. However, certain emergency situations, mainly related to the state of conservation of the typical of polymers. See: Scalarone, D.: Materials for conservation and restoration, University of Turin, AY 2018/2019 (2019).

246

T. Markevicius et al.

photograph and the need to ensure its maximum durability over time, sometimes lead the conservator to take such a decision. This was the case of a silver gelatin print of the famous actress Lyda Borelli, belonging to the Vittorio Martinelli Collection conserved at the Cineteca di Bologna (Fig. 11). Therefore, some of the best known and most consolidated photograph detachment techniques were tested to examine the specific advantages and criticalities of each of them. The research was developed both by consulting specific contributions offered by contemporary and not, specialized literature, and by carrying out detachment tests on original historical mountings - the selected samples were in a mediocre state of conservation and presented comparable characteristics to the Lyda Borelli’s mounting. The traditional detachment systems tested have been dry mechanical detachment, detachment by humidification with ultrasound humidifier, detachment after water immersion, detachment by cardboard delamination, detachment by verso imbibition, detachment by recto imbibition, detachment after humidification in humidity chamber. All of them showed different issues and the incompatibility with Lyda Borelli’s mounting.

Fig. 11. The front and the reverse of L. Borelli photograph before treatment.

For this reason, the research has been directed through new possibilities and alternative methods, exploiting the advantages offered by the most advanced technologies. The experimentation has been focused on a different detachment system, based on the combination of the well-known humidification sandwich, consisting of breathable membranes and wet blotters, with the innovative IMAT device. The idea behind the sandwich - IMAT system, developed as a result of the considerations that emerged during the previous tests, concerns the possibility of detaching a chemically degraded silver gelatin print from the mounting cardboard exploiting the action of humidity and thermal energy

Innovation in Precision Low-Energy Heat Transfer

247

at the same time. This guarantees maximum safety of the silver image and of the cardboard - which is often sacrificed during detachment treatments in favor of safeguarding only the photograph. The sandwich - IMAT has been realized according to the following arrangement of the different layers: polyester Melinex sheets on top and on bottom, to insulating the package from the surrounding environment; silicone sheet at the base, useful to limit the thermal dispersion; heated silicone mat, as a source of thermal energy; polyester Melinex sheet, as a protective cover of the mat; wet blotter, source of humidification; breathable Polartec NeoShell membrane, to permit the capillary diffusion of the humidity; the photograph, facing with the emulsion side upwards; dry blotter, to reduce the humidification of the recto of the object (Fig. 12).

Fig. 12. Humidification using Polartec membrane and IMAT in progress (left), detachment of the photograph (right).

The experimentation of this system was carried out through several modalities, testing different temperatures on different types of samples. The tests included facsimiles of photographic mountings and historical silver gelatin prints mounted on original cardboards. The results obtained were satisfactory. The advantages offered by the use of the sandwich - IMAT system concerned the possibility of carrying out a homogeneous humidification of the object, which is necessary to uniformly reduce the mechanical resistance of the adhesion point. The creation of a sandwich consisting of a single wet blotter placed in indirect contact with the verso of the object makes it possible to reduce the impact of humidity to the water-sensitive gelatin photographic emulsion. The decision to use a microporous breathable membrane ensured better control and more gradual humidification. The presence of a thermal energy source inside the sandwich allowed the vapor to propagate more quickly, reducing the time needed to carry out the treatment, without creating interference. The humidification system allowed the prints to be detached from the cardboard without compromising the structure of the cardboard itself, which is preserved in good condition. The sandwich - IMAT system, set at 40 °C and used for 90 min time, was therefore used to detach the Lyda Borelli’s print from the mounting board, allowing the photograph to be safely removed. The preservation in good condition of the mounting cardboard was also guaranteed.

248

T. Markevicius et al.

3.3 Precision Heat Transfer Mat in Conservation of Paper and Books Old mending and backing in bound manuscripts and rare prints, which are glued with proteinaceous adhesives, can result in heavy damage to papers due to embrittlement and hardening. Sometimes it may also interfere with readability when old repairs were carried out imprecisely so that the original text and illustrations are misaligned or concealed. A combination of controllable moisture and heat application is often required for the detachment of old repairs. Treatment methods combine hydrogels/Gore-Tex sandwich with the innovative carbon nanotube-based IMAT prototype has been successfully tested and introduced to the conservation of originals (Liu, 2018). Detaching Old Mends Glued in the Spine Fold of a Bound Manuscript. The first application was taken on a Latin manuscript (Munich, Bavarian State Library, Clm No. 18199), which was an anthology and written in the 15th century on paper with handcolored illustrations. Several old mends with protein-based glue were found in the spine fold. The shrinkage, stiffness and inflexibility of the aged adhesive have caused heavy distortions in paper (Fig. 13 a, b). In addition, some mending papers also covered the illustrations. Therefore, a detachment of the old mends was considered as necessary. A treatment-sandwich combined an agarose gel with a heating mat was designed (Fig. 14). Agarose was prepared in 4 wt% and casted in 2 mm, then cut into a rectangle shape with the same width as the mending papers and a length of 2 cm. A sheet of Rayon paper was applied to generate a more uniform humidification and to minimize the possible existence of residue. The temperature of the mat was set at 40 °C (Fig. 13e). After treated for 3– 5 min, the adhesive layer could swell, and the mending papers could be removed by a micro-spatula piece by piece (Fig. 13f). The residues of the proteinaceous glue were then carefully reduced with Tylose® MH 300 gel (7 wt%). After the treatment, the tension of the area and the distortions on the pages were considerably reduced (Fig. 13 c, d). Detaching the Old Backing from a Large-Scale Woodcut Map Print. Another application of this precision heat transfer was carried out on a large fold-out woodcut map of Venice in the Peregrinatio in Terram Sanctam (Bavarian State Library, Munich, No. 2 Inc.c.a. 1726 a), which is an incunable account of a pilgrimage to Jerusalem undertaken by Bernhard von Breydenbach and Erhard Reuwich together with other companions in the year of 1483 to 1484. The book featured several large fold-out woodcut panoramas of cities in Europe and maps of Jerusalem, Palestine, and Egypt. The Venice woodcut map in this copy (Bavarian State Library, Munich, No. 2 Inc.c.a. 1726 a) was damaged with several missing areas, splits, and tears (Fig. 15a). It was reinforced at an early point of time with two backing-papers: The first backing was a wastepaper, whose text side was glued with protein-based adhesive to the back of the map. The glue had already penetrated the map surface in many places. This backing was later reinforced once more by a grey-colored paper with a starch-based paste (Fig. 15b). The aging of the adhesives leaded to stiffness of the whole map as well as new tears on the edges and the fold areas. Based on the result of a preliminary test, it was determined that the ability of water absorption of the starch-based paste was not satisfying. The two backings were therefore decided to be detached simultaneously by combining moisture introduction with input of thermal energy. In this case, a treatment-sandwich combined a Gore-Tex sandwich

Innovation in Precision Low-Energy Heat Transfer

249

Fig. 13. Treatment on manuscript BSB, Clm 18199: Heavy distortions caused by the shrinkage of the mending glue (a, b); Pages after the conservation treatment (c, d); Images during the conservation treatment (e, f). ©BSB/IBR/Liu.

Fig. 14. Schematic diagram of the humidification sandwich with heat transfer for the detachment of the old mending papers in BSB, Clm 18199.

with heat transfer was selected (Fig. 16), as it generated a more uniform humidification due to the large size and high stiffness of the backings. A 2 mm thick blotting board was cut into a rectangle shape with the length same as the backing’s width and a width of 10 cm. The target temperature was set at 40 °C. During the treatment, two sandbags were placed onto the treated area to ensure an ideal contact between the mat and the Gore-Tex sandwich. A sheet of polyester non-woven wadding was placed in between to prevent

250

T. Markevicius et al.

Fig.15. Treatment on manuscript BSB, 2 Inc.c.a. 1726 a: Woodcut map of Venice folded out of the incunable (a, b); Images during the conservation treatment (c, d). ©BSB/IBR/Liu.

Fig. 16. Schematic diagram of the humidification sandwich with heat transfer for the detachment of the backings from the woodcut map of Venice (BSB, 2 Inc.c.a. 1726 a).

heat loss. Several extra sandbags were placed directly onto the area, where the mat had contact only to the polyester foam book support and air, to avoid it from being partially overheated (Fig. 15c). The adhesive layer could swell circa 20 min, and the backing could be readily removed piece by piece (Fig. 15d).

Innovation in Precision Low-Energy Heat Transfer

251

4 Discussion and Conclusions The paper discusses new approaches and targeted treatments of diverse tangible cultural heritage objects – paintings, photographs, books and works on paper made possible by innovative low-energy heat transfer, using flexible precision mats. Ongoing studies provide a new understanding of the viscoelastic properties of constituent art materials and identify behavioral changes caused by external factors such as heat and humidity, and this research investigates how these factors may be employed by controlling the temperature and the treatment time to shift aged and brittle artwork materials into more elastic physical states that allow safe remedial treatment. This new knowledge may be applied by conservators to improve safety and effectiveness of structural treatment outcomes with the proposed methodology using MAT and IMAT low-energy heat transfer technology for low temperature range (below 40 °C) up to those customary used for thermoplastic adhesive activation (65 °C). The new methodology represents a radical shift from low-tech, poorly controlled heating methods to an approach where the low-energy heat transfer is targeted, safe and commensurate with the desired result. While the traditional tools, such as hand irons overheat the top surface, and operate at unsafe high temperatures, the MAT, thanks to steady and even heat diffusion over large areas, allows for a novel “low and slow” approach, using safe low temperature over an extended period of time, tailored to specific treatment. Beyond the temperature control, novel access to sophisticated microporous membranes to acquire control in the humidification of materials under treatment as a plasticizing agent provides conservators with alternatives to formulate new treatment methodologies of paintings on canvas, modern materials, works on paper, books, photographs, and more, within the margins of minimal intervention and risk, while achieving the maximum result. The variety of case studies shows the broad spectrum of application for the mild heat transfer technology in structural treatment and beyond, from the use of mild heat over extended period (low and slow) to treat planar and surface distortions, to safe treatment of works previously treated with natural and synthetic crystalline waxes, to the utility of the thin, flexible profile to reach between canvas and stretcher bars for treatments that conserve the original mounting and structural integrity of the piece. The compact dimensions and portability of the device allow the conservator to easily and simply work in the laboratory or conduct state of the art treatments in the field, advancing best practices in art conservation and treatment of artworks. It is the authors’ hope that further study will deliver better understanding how to exploit the beneficial effects of targeted low-energy heat transfer applied to structural treatment, while the introduction of precision heating mats, and low and slow approaches, developed and advocated by the authors, numerous collaborators who contributed to this research, will make the future treatments safer, more efficient, greener and sustainable. Acknowledgements. The authors are grateful to the three Master thesis supervisors for the contribution for: Y. Liu Master thesis: prof. K. Eckstein, prof. T. Allscher, Munich Technical University and Bavarian State Library; M. Magnani Master thesis: Prof. A. Bassi, Prof. D. Scalarone, Prof. F. Belloni; Dr. T. Poli, Dr. L. Avataneo; University of Turin and Centro Conservazione e Restauro “La Venaria Reale”, Intesa Sanpaolo, Archivio Turcato and paintings conservator Laura Amorosi,

252

T. Markevicius et al.

who lent the IMAT prototype; M. Paganin Master thesis: Prof. C. R. Monaco, Prof. A. Panzetta, Accademia di Belle Arti di Bologna and Cineteca di Bologna.

References Berger, G.A., Russell, W.H.: The new stress tests on canvas paintings and some of their implications on the preservation of paintings. In: ICOM Committee for Conservation 7th Triennial Meeting, Copenhagen, 84/2/7–9 (1984) Berger, G.A., Russell, W.H.: Investigations into the reactions of plastic materials to environmental changes. Part I: the mechanics of decay of paint films. Stud. Conserv. 31, 49–64 (1986) Hagan, E., Charalambides, M., Learner, T., Murray, A., Young, C.: Factors affecting the mechanical properties of modern paints. In: Modern Paints Uncovered, Proceedings from the Modern Paints Uncovered Symposium, Tate Modern, London, 16–19 May 2006, pp. 227–228 (2006) Hagan, E.W.: Thermo-mechanical properties of white oil and acrylic artist paints. Prog. Org. Coat. 104, 28–33 (2017) Hedley, G., Odlyha, M.: The moisture softening of paint films and its implications for treatment of fabric supported paintings. In: Traitement des support. Travaux interdisciplinaires, Paris, 2, 3 et 4 novembre 1989. ARAAFU, pp. 157–162 (1989) Goddard, P.: Humidity chambers and their application to the treatment of deformations in fabricsupported paintings. Conservator 13, 20–24 (1989) Krumrine, K.: An Innovative Technique for Reforming Cellulose Acetate in an Architectural Model of Rockefeller Plaza: Challenges of Preserving Modern, Unstable Restorations AIC’s 47th Annual Meeting, 13–17 May 2019 (2019) Liu, Y.: An innovative heat transfer method for solving old mending glued with proteinaceous adhesives in manuscripts and rare books, Thesis for: Master of Art; Advisor: Prof. Erwin Emmerling, Dr. Irmhild Ceynowa, Dr. Thorsten Allscher, Karin Eckstein, M.A. (2019). https:// doi.org/10.13140/RG.2.2.20332.31363 Magnani, M.: Study and conservation treatment of a multi-material collage by Giulio Turcato: consolidation and reintegration issues of an artwork with complex stratigraphy, Master’s thesis, University of Turin – La Venaria Reale, Turin, Italy. Supervisors: Prof. Bassi, A., Scalarone, D., Belloni, F., Olsson, N., Markevicius, T. (2021) Magnani, M., Bassi, A., Scalarone, D., Poli, T., Markevicius, T.: Study and restoration treatment of a collage by Giulio Turcato: from precision mild heat transfer using IMAT nanotechnology to novel sustainable methods and strategies for consolidation and reintegration, American Institute for Conservation, 2022 Annual Meeting (2022) Markevicius, T., Meyer, H., Olsson, N., Furferi, R.: Conductive transparent film heater as alternative to heating table: towards new intelligent mobile accurate thermo-electrical (IMAT) device for structural conservation of paintings, in Art’11 by AIPnD, Florence, Italy, 13 April 2011 (2011) Markeviˇcius, T., Meyer, H., Saborowski, K., Olsson, N., Furferi, R.: Carbon nanotubes in art conservation. Int. J. Conserv. Sci. 4(Special Issue), 633–646 (2013). ISSN:2067–533X Markevicius, T., et al.: New approaches to an old problem: a precision mild heat transfer method for the Nuanced treatment of contemporary and modern art works. In: Conference: The 18th Triennial Conference of International Council of Museums - Conservation Committee ICOMCC, Copenhagen (2017) Markeviˇcius, T., Syversen, T., Chan, E., Olsson, N., Skov Hilby, C., Šimait˙e, R.: Cold, warm, warmer: use of precision heat transfer in the optimization of hydrolytic enzyme and hydrogel cleaning systems, Gels in conservation conference. In: Angelova, L., Ormsby, B., Townsend, J.H., Wolbers, R. (eds.) International Academic Projects (2018). ISBN 9781909492509

Innovation in Precision Low-Energy Heat Transfer

253

Olsson, N., Markeviˇcius, T.: Low and slow: the role of targeted precision heat transfer and innovative flexible mat technology for new methodology in conservation of paintings on canvas. In: Coddington, J., McClure, I., Schwarz, C. (eds.) Conserving Canvas Symposium. Yale University. Getty Publications (2022) Mecklenburg, M. F., Tumosa, C.S.: Mechanical behavior of paintings subjected to changes in temperature and relative humidity. In: Mecklenburg, M. (ed.) Art in Transit: Studies in the Transport of Paintings, Washington DC, National Gallery of Art, pp. 173–216 (1991) Michalski, S.: Paintings: their response to temperature, relative humidity, shock and vibration. In: Art in Transit: Studies in the Transport of Paintings, Washington DC, National Gallery of Art, pp. 223–248 (1991) Olsson, N., Markeviˇcius, T.: Flexible thermal blanket and low-pressure envelope system in the structural treatment of paintings on Canvas. In: Buckley, B. (ed.) Postprints of the Paintings Specialty Group, American Institute for Conservation meeting, Milwaukee, Washington DC. American Institute for Conservation, pp. 63–71 (2010) Olsson, N., Markeviˇcius, T.: Studio tip: innovative applications of heating mat and low-pressure systems for lining and minimal structural treatment of paintings. In: Partridge, W. (ed.) Postprints of the Painting Specialty Group, American Institute for Conservation meeting, Chicago, Washington DC: American Institute for Conservation, pp. 127–135 (2017) Paganin, M., Gianferrari, M., Gioia, R., Markevicius, T.: Sandwich – IMAT: un nuovo sistema di distacco delle stampe fotografiche storiche dal supporto secondario, XIX Congresso Nazionale IGIIC – Lo Stato dell’Arte 19 – Udine 2021 (2021) Paganin, M.: Il restauro di fotografe d’autore montate su supporto secondario. I ritratti di Lyda Borelli e Maria Melato a confronto, Master’s thesis in paper conservation. University of Bologna, Bologna, Italy (2021) Percival-Prescott, W.: Foreword: origins and outcome of the lining conference. In: Villers, C. (ed.) Lining Paintings. Papers from the Greenwich Conference on Comparative Lining Techniques, pp. v-ix, London. Archetype Publications Ltd in association with the National Maritime Museum, Greenwich (2003) Phenix, A.: Thermal mechanical transitions in artist’ oil paints and selected conservation materials: a study by Dynamic Mechanical Analysis (DMA). In: Buckley, B. (ed.) 2009 Postprints of the Paintings Specialty Group, American Institute for Conservation Meeting, Los Angeles, Washington DC. American Institute for Conservation, pp. 72–89 (2011) Russell, W.H., Berger, G.A.: The behavior of canvas as a structural support for painting. Science and Technology in the Service of Conservation, IIC-Washington Congress, London, IIC, pp. 139–45 (1982)

Monitoring of Cultural Heritage Environments

The AMOR Project: When Technology Meets Cultural Heritage Nicole Dore1(B) , Francesco Cochetti2 , Ilaria Catapano3 , Giovanni Ludeno3 Gianluca Gennarelli3 , Maria Elena Corrado4 , Carlo Cacace4 , Paolo Osso5 , Michele Luglio6 , and Francesco Zampognaro6

,

1 NAIS – Nextant Applications and Innovative Solutions, Lazio Rome, Italy

[email protected] 2 CoopCulture, Rome, Lazio, Italy [email protected] 3 CNR-IREA - National Research Council - Institute for Electromagnetic Sensing of the Environment, Campania Naples, Italy {catapano.i,ludeno.g,gennarelli.g}@irea.cnr.it 4 ICR – Central Institute for Restoration, Lazio Rome, Italy {mariaelena.corrado,carlo.cacace}@beniculturali.it 5 ESRI Italy, Rome, Lazio, Italy [email protected] 6 NITEL - National Interuniversity Consortium for Transport and Logistics, Lazio Rome, Italy [email protected], [email protected]

Abstract. AMOR project focuses on the development of pre-commercial services addressed to the Institutions responsible for the conservation and the valorisation of the Cultural Heritage. Services, developed both for Safeguard and Fruition, will be tested at the Ancient Baths of Caracalla and on Aurelian walls, that will be surveyed with several technologies for the detection of superficial (satellites, UAV) or sub-superficial (GPR) damages/anomalies affecting them. Fruition services will offer to tourists a mobile app equipped with Institutional approved contents and with Mixed Reality experience (3D model) directly available on personal mobile devices. Moreover, the anonymous (previously authorised) tracking of tourists will allow to obtain further information useful for safeguard aspects. Keywords: Cultural heritage · GPR · Satellites · Mobile APP · Mixed reality · 5G · Data analytics

Short Intro The AMOR project [1] was born with the aim of offering services in the areas of Safeguard and Fruition of Cultural Heritage. These services are about circular safeguard (Fig. 1) and fruition, fully responding to the strategies of MiC.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 257–265, 2023. https://doi.org/10.1007/978-3-031-17594-7_19

258

N. Dore et al.

Fig. 1. Circular safeguard

1 Introduction The AMOR - Advanced Multimedia and Observation services for the Rome cultural heritage ecosystem - project, co-financed by the Italian Space Agency (ASI) and coordinated by the European Space Agency (ESA) and the Ministry of Culture (MiC), falls within the scope of the ESA 5G for L’ART (Business Applications program) call. Project activities, lasting 24 months, are coordinated by NAIS Srl, leader of a group composed of public and private entities (ICR, CNR-IREA, CoopCulture, NITEL, ESRI Italy), each of them with extensive experience and expertise in the various aspects addressed in AMOR. The project mission was conceived because of the more and more evident need to take technologies and digital tools available today, apply them systematically, and put them at the service of Public Bodies responsible for the conservation and enhancement of the Italian cultural heritage, as well as for stakeholders and private users interested in such sectors. The monuments of Rome, application scenario of AMOR services, are the suggestive archaeological complex of the Baths of Caracalla and a southern stretch of the mighty Aurelian Walls (from Porta San Sebastiano to Porta Latina).

The AMOR Project: When Technology Meets Cultural Heritage

259

2 Safeguard: Needs, Technologies and Methodologies Among the main needs of end users, which AMOR services aim to satisfy, there is undoubtedly that of a greater awareness of the conservation status of monuments, and the nature and severity of the main criticalities that impact them. To meet these needs, thanks to satellites, AMOR makes use of technologies for the observation of the territory over a large area, and also offers a high detail of the surfaces (few cm) through the use of UAV (Unmanned Aerial Vehicle) systems. Then comes a sub-surface investigation of both soil and vertical structure, such as walls, through the use of GPR (Ground Penetrating Radar). 2.1 Remote Observation of Surface Degradation: Satellites and Drones For the remote observation of damage and dangers of the territory, impacting on cultural heritage and the areas adjacent to them, as well as for the observation of their evolution over time, AMOR makes use of instruments such as satellites, equipped with multispectral sensors and SAR (Synthetic Aperture Radar) sensors, along with UAV systems, also equipped with appropriate sensors selected on the basis of the degradation phenomenon to be detected. SAR sensors, through the use of the PS-InSAR (Permanent Scatterer - Interferometric Synthetic Aperture Radar) technique [2], are now widely recognized and used; they allow the extraction of information on the slow millimeter displacements of the ground, and that of the buildings placed above it, which otherwise are not detectable if not through the installation of dedicated sensors that, in any case, provides punctual information unlike the vast area coverage guaranteed by satellites. Therefore, periodic monitoring using this technique allows for an improvement in knowledge of the critical issues related to the instability of the ground, and the reflections that these may have on the monuments (see Fig. 2).

Fig. 2. PS-InSAR analysis on the historic center of Rome - GIS St’ART platform

260

N. Dore et al.

The multispectral sensors mounted on board satellites are, on the other hand, used in AMOR with two main purposes: a) analysis of the infesting vegetation on roofs or wall ridges of the monuments (1st level analysis); b) change detection, over a vast period of time, of a specific urban sector. In the latter case, the use of commercial satellites with high spatial resolution (VHR) is essential for the identification and mapping of changes in the urban fabric, which cannot be observed otherwise. These have the function of extracting valuable information for the refinement of the calculation of the vulnerability of urban units and the associated risk (Risk Card - ICR/MiC). In fact, Risk Map [4, 5] is now evolving testing an extension of the analysis of the conservation status from the single monument to the urban unit. AMOR project has a specific case study in the area nearby via della Lungara (Rome historical centre, in Trastevere zone). However, for the purpose of further information or more precise observations that the satellite is unable to provide, the use of UAV systems refines and completes the range of services. Through these, in fact, it is possible to detect and map the damage that occurs on a specific monument, with a view to improving knowledge of the conservation status of a monument and also contributing to the calculation of the vulnerabilities of individual monuments. 2.2 The GPR System The georadar - or GPR - is another technological tool used for safeguard purposes in AMOR. It is a radar system designed for carrying out sub-surface surveys, increasingly used in the Cultural Heritage sector [3]. By exploiting the ability of microwaves to penetrate non-metallic materials and sophisticated data processing procedures, often optimized for the specific application of interest, the georadar provides high resolution images of the investigated region in a non-invasive way. These images make it possible to locate buried objects, whose shape and dimensions can be reconstructed, and to characterize degradation phenomena that can damage the structure under examination. In AMOR, GPR is used both to carry out surveys of the subsoil - aimed at increasing knowledge of the site, thanks to the identification of underground objects of which memory has been lost, for example walkways and/or underground cisterns, as in the case of the archaeological complex of the Baths of Caracalla - and to carry out structural investigations aimed at characterizing the cracks, as in the case of the Aurelian Walls. The AMOR GPR system is a highly flexible system, as it can be equipped with antennas operating at different frequencies (from a few tens of MHz to a few GHz), thus allowing to reach different depths of penetration and to obtain images at different spatial resolutions, in order to satisfy the requirements dictated by the purpose of the survey requested by the user. The acquired data are processed using procedures developed by CNR - IREA researchers. These procedures involve both filtering algorithms that aim to emphasize the useful signal, meaning the signal due to objects of interest, and tomographic reconstruction approaches which, by solving an inverse problem of electromagnetic diffusion, produce focused images of the identified objects, making

The AMOR Project: When Technology Meets Cultural Heritage

261

their accurate geometric characterization possible. These images therefore provide useful information for the safeguarding and optimized maintenance of the site, and can be used to make (not otherwise visible) artifacts available to visitors. At the present stage of the project, two GPR measurement campaigns have been carried out, one at the archaeological complex of the Baths of Caracalla and the other at a portion not accessible to visitors of the southern part of the Aurelian Walls, affected by a non-negligible phenomenon of fracturing. Figure 3 shows the measurement campaign carried out at the Baths of Caracalla, where attention was focused on the area of gardens and avenues. A preliminary result, obtained by applying only the filtering procedures to data acquired along one of the avenues, is shown. Several sub-superficial anomalies can be noted: the first, on the left, located at a depth of about 1.2 m – and the others, at about 2 m. The anomalies could be associated with cisterns and/or tunnels. Further investigations are underway in order to arrive at their correct interpretation.

Fig. 3. Prospecting with GPR at the baths of Caracalla

3 Fruition As mentioned in the introduction, AMOR also aims to develop services dedicated to fruition. In fact, in a historical period characterized by a pandemic still in progress, the possibility of giving users multimedia contents certified and approved by the authorities directly responsible for the monuments, undoubtedly gives added value to the experience of a cultural tourist.

262

N. Dore et al.

The service proposed by AMOR therefore provides for the development of a mobile app that offers certified information content and emotionally engaging experiences (e.g. Mixed Reality (MR)), for a more in-depth and immediate knowledge of cultural heritage. 3.1 New Fruition Solutions Today, technological innovation in the digital field allows the creation and development of new tools for the enhancement of cultural heritage, capable of offering the public contents with previously unimaginable communication potential and, at the same time, of collecting useful data for the protection of the interested sites. The Mobile App designed and developed for the AMOR project is able to offer an innovative form of fruition, which combines advanced multimedia content and simplification of use through itineraries and georeferencing. The tool is designed to be used in total autonomy by visitors from all over the world and enjoyed in different languages. The App provides a catalog of contents created to enjoy urban tourist itineraries in different cultural places, such as museums, archaeological areas, monuments. A map of the city indicates the user his position (in case of active GNSS and in full compliance with current privacy regulations) allowing him to orient himself and decide which itinerary to follow, or which site to visit and - if interested - the user can also download and enjoy the various contents available directly on the device. Today, with a smartphone, anyone can reach an unimaginable amount of information in a short time. But what contents are actually reliable? The app of the AMOR project, developed with cultural institutions, allows the user to access scientifically approved and quality content, together with the proposal of experiences not otherwise accessible online. The multimedia contents, created by storytelling professionals, guide the user throughout the tour, focusing his attention on the most significant details for a correct understanding of the monument. Satellite and graphic maps, audio commentary, image galleries, texts, virtual reconstructions, elements of Mixed Reality, are from time to time combined together to create an emotionally engaging visiting experience. In particular, in AMOR the use of Mixed Reality (MR) will allow users to view 3D models of the Baths of Caracalla complex, facilitated by 5G and 4G connections. The informative and educational purposes of the application are closely linked to those of the protection and enhancement of the sites visited, by collecting useful data about visitor flows (see next paragraph) (Fig. 4). 3.2 Circular Fruition According to what has just been described, what contributes to making AMOR innovative is precisely the circularity of the data collected through different sources, a circularity that leads to the generation of further information material that flows into the safeguard sector, feeding the knowledge of “indirect vulnerability” of the property, also impacted by anthropogenic causes. The anonymous collection of visitor positioning data will allow for an analysis of flows (data analytics) relating to the places they visit.

The AMOR Project: When Technology Meets Cultural Heritage

263

Fig. 4. Mobile APP for viewing 3D models in MR at Baths of Caracalla

This analysis process is integrated with a geofencing activity, which consists in identifying crossing areas of which the positioning data are recorded, giving additional information. In addition to data collected by the mobile app, the analysis of flows will also be fed by data available on the market relating to the gravitation on telephone cells of the various mobile devices, the so-called gravitational databases, or data made available by some telephone operators or companies which enrich the databases after some analysis processes. This latter product was taken into consideration in AMOR to enrich the project (and the service in the future) with a component of historical data analysis, useful for building behavior patterns over longer times and which also takes into consideration visitors not downloading or installing the app. In particular, the GeoMobile DB product was used for the project, as it offers data on a quarterly basis, referring to the census sections from which it is possible to obtain information on time bands of influx, age, nationality (if Italian or foreign), type of user (whether business or private). For the analysis and visualization of this data, the project will use the ArcGIS Enterprise platform, containing the modules defined as Geotools that have been developed. Within the platform, used for the definition and management of the geofencing areas and for the analysis and visualization of data, applications will be developed for the visualization on the map, and also diagrams with the analysis results, through dashboards that will allow the user manager of the area to have a synoptic picture about the trend of the visitors’ flow.

264

N. Dore et al.

The assessments that may be made by the area manager will be useful for making decisions in order, for example, to enhance less visited places, affix information/promotion signs, place information points, reorganize visitor routes and so on. 3.3 The Importance of 5G 5G supports the AMOR project with a view to a more fluid mobile fruition. The fifth generation of mobile network - namely 5G - brings with it important advances and revolutions in the telecommunications sector. 5G, in addition to offering a higher transmission speed and lower latency compared to the previous generation 4G/LTE (and other technological improvements), offers significant innovations from the infrastructural, functional and services point of view. In fact, 5G includes innovative design paradigms such as Software Defined Networking (SDN) and virtualization of network functions (NFV). These technologies allow the creation of a virtualized network infrastructure that is completely computerized as regards the functionality and management of the network itself [6]. All this makes the 5G ecosystem flexible and adaptable to the characteristics of the service, allowing the design of independent portions of the network, called “slices” [7]. The result is being able to provide advanced services that go beyond voice, message and data services offered right now by mobile networks, which meet the needs of new virtual operators (called “verticals”) in the automotive sectors (e.g. autonomous driving), smart cities (e.g. sensor networks), eHealth (e.g. telemedicine), media (e.g. augmented reality), Industry 4.0, etc. [8]. In this context, the use of the 3D multimedia contents offered in itinere during city tours, foreseen by the AMOR project, can enjoy the great potential offered by 5G networks, allowing to define a new “Vertical” in the culture/infotainment sector. Specifically, the full availability of 5G will allow the application to interact directly with a “slice” of the network, making use of the processing of large amounts of data in the network itself, at low latency, offering a user experience that is even superior to what is possible today.

4 Conclusions The complexity of the AMOR system sees its own simplification in the way it is made available to end users. Access to information relating to traditional and “collateral” safeguarding (from the mobile app) takes place through a single data consultation point: the St’ART® platform [9], which provides all the tools for reading the acquired data (satellite, UAV and GPR), facilitating the user in the subsequent actions to be taken, where necessary, and supporting him through a specially designed and calibrated workflow. A simplification in the context of fruition is also implied for the end user, in this case the cultural visitor. The possibility of consulting official multimedia data, as well as the availability of innovative solutions (Mixed Reality) “within reach of a smartphone” and in full compliance with the anti-Covid-19 regulations, offers quality fruition that is

The AMOR Project: When Technology Meets Cultural Heritage

265

aimed at everyone; just think of the current reticence in wearing a 3D viewer, or handling a device used by a third party. The products derived from these services, once put into the system, will flow into what can be characterized as the Site Knowledge Hub, which over time, through participation in other projects that will see these same sites as a scenario for further activities, will be enriched and become a point of collection, correlation and consultation of information (for use and safeguarding). Therefore, the preventive approach proposed by AMOR, which aims to trigger a reduction in costs for the aspects related to safeguarding (costs no longer oriented to the onerous emergency, but to ordinary maintenance) and an “enlarged” fruition, are designed to trigger a virtuous circle for what is called circular safeguarding.

References 1. https://business.esa.int/projects/amor 2. Ferretti, A., Prati, C., Rocca, F.: Permanent scatterers in SAR interferometry. IEEE Trans. Geosci. Remote Sens. 39(1), 8–20 (2001) 3. Catapano, I., Gennarelli, G., Ludeno, G., Soldovieri, F.: Applying ground-penetrating radar and microwave tomography data processing in cultural heritage: state of the art and future trends. IEEE Sig. Process. Mag. 36(4), 53–61 (2019). https://doi.org/10.1109/MSP.2019.289 5121 4. Cacace, C.: La Carta del Rischio del Patrimonio Culturale in Il Futuro dei Centri Storici Digitalizzazione e strategia conservative, a cura di Donatella Fiorani, Roma 2019, Edizioni Quasar di Severino Tongon S.r.l, via Ajaccio 43, Roma, pp. 65–74. ISBN 978-88-7140-925-2 5. Accardo, G., Cacace, C., Rinaldi, R.: Il Sistema Informativo Territoriale della carta del Rischio’ in ARKOS – Scienza e Restauro dell’Architettura Nardini Editore Anno VI – Nuova Serieaprile/giugno 2005 6. Carugi, M.: ITU, 2019 Distinguishing features and high level requirements of 5G/IMT 2020 networks. https://www.itu.int/en/ITU-D/Regional-Presence/ArabStates/Documents/events/ 2019/ET/S1-%20ITU%20Reg%20Forum-Tunis-5G%20IMT2020-presentation-Marco-Car ugi-v1.pdf 7. Ericcson, Applied network slicing scenarios in 5G. https://www.ericsson.com/en/reports-andpapers/ericsson-technology-review/articles/applied-network-slicing-scenarios-in-5g 8. https://5g-ppp.eu/verticals/ 9. http://www.start-solutions.it/

Testing Portable NMR to Monitor the Effect of Paper Exposure to UV-Light Valeria Stagno1,2(B) , Alessandro Ciccola3 , Elisa Villani4 , Roberta Curini3 , Paolo Postorino5 , and Silvia Capuani2 1 Department of Earth Sciences, Sapienza University, Rome, Italy

[email protected] 2 National Research Council-Institute for Complex Systems (ISC), Rome, Italy 3 Department of Chemistry, Sapienza University, Rome, Italy 4 Department of Environmental Biology, Sapienza University, Rome, Italy 5 Department of Physics, Sapienza University, Rome, Italy

Abstract. Paper-based works of art can be considered the most important carrier of information about culture, science, business, politic and history. Therefore, it is highly important to preserve the integrity of the paper these objects are made of. Paper is an organic material mainly made of cellulose fibers, whose durability depends on pH, heat, humidity, oxygen, pollution, metal ions, lignin, and UVvisible light. Cellulose absorbs more in the near UV region, therefore radiation with wavelength of 300–550 nm produces most of the paper damage. The aim of this work was to test the potential of single-sided portable NMR to highlight the effect of UV light on paper. To this end, the longitudinal relaxation time (T 1 ) and the transversal relaxation time (T 2 ) of structural-known filter paper before and after the exposure to UV light were measured and supported by Raman spectroscopy. The decrease of both T 1 and T 2 parameters with the increase of the UV exposure time indicates a modification of the cellulose chains, which was confirmed by Raman spectra. Moreover, this study presents a preliminary non-invasive protocol for assessing the effect of artificial UV irradiation on paper by using a portable NMR sensor. Keywords: Portable NMR · UV-light · Paper · Raman spectroscopy

1 Introduction Paper represents one of the many organic materials which have been used in the past for artwork creation. Paper-based works of art can be considered the most important carrier of information about culture, science, business, politic and history [1]. For this reason, it is important to preserve paper artworks and understand the main factors that can endanger paper durability. The production process of paper changed during the centuries in response to market needs [1], therefore the raw materials also changed. Cellulose, hemicellulose and lignin are the three polymers considered the main components of paper [2, 3]. Depending on its final use, paper can contain different percentage of these © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 266–276, 2023. https://doi.org/10.1007/978-3-031-17594-7_20

Testing Portable NMR to Monitor the Effect of Paper Exposure

267

polymers [2]. However, in many cases cellulose, a polysaccharide made up of β-glucose units linked together by β(1 → 4) glycosidic bonds [3], is the major structural component of paper and its degradation produces an alteration of the mechanical properties of paper [1]. Indeed, cellulose polymer chains exist in organized crystalline regions, which are impermeable to water, but may coexist with non-crystalline regions. The non-crystalline regions can take up water into molecular scale spaces between individual polymer chains [4] and therefore they are responsible for paper degradation induced by water. In addition to the production process, a sheet of paper can undergo chemical treatments, such as coating, smoothing, and bleaching [1, 2]. All these treatments together with the type and quality of raw materials used affect the stability of paper. Nevertheless, the durability of cellulose fibers remains highly dependent on their acidity/alkalinity [1]. For this reason, a deacidification process can be performed as a basic conservation technique of paper [1, 5, 6]. Beyond the paper pH, there are many other factors that influence the paper durability, such as heat, humidity, oxygen, UV-visible light, pollution, metal ions, lignin and degradation products [7, 8]. Among the aforementioned factors, the effect of UV-visible light can be considered one of the major source of decay for cellulosic materials, such as paper [9]. In particular, it is known that pure cellulose absorbs more in the near UV spectral region [9]. Therefore, light in the range 300–550 nm induces most of the damage in cellulose. There exist two types of light-induced degradation reactions of cellulose: direct photolysis and photosensitized degradation. Photolysis occurs when the energy of a photon is high enough to induce dissociation of a bond. For wavelengths higher than 340 nm direct photolysis of cellulose does not take place [1, 10]. On the other hand, in the photosensitized degradation a chemical species, called photosensitizer, absorbs a photon and the energy from electronically excited state is transferred to initiators, or oxygen, resulting in the formation of reactive species (free radicals) and in the photooxidative decay of cellulose [1, 9–14]. Philips et al. [10] suggested that most carbon-centered radicals are formed at wavelength of 360 nm. In the presence of oxygen, these radicals convert into peroxyl radicals that become hydroperoxydes when hydrogen is abstracted (mainly from O-H and C-H bonds) [9, 13]. At λ = 360 nm, hydroperoxydes can decompose yielding to highly reactive hydroxyl radicals [9, 14]. Photooxidative reactions lead to gradual depolymerisation of cellulose [8, 9]. Degradation is not random [8, 9] and results in an increase of carbonyl content, carboxyl and hydroperoxide groups and a decrease of degree of polymerization and alpha-cellulose content [9, 13]. Moreover, evaluating the effect of UV-visible light on paper artworks is of great importance to act stabilization processes and long-term preservation strategies [9]. This work aims at testing the potential of single-sided portable NMR to highlight the effect of UV-light on paper. To the best of our knowledge, no work about the application of portable NMR to monitor the UV artificial aging effect on paper can be found in the literature, whereas other techniques were widely used to investigate this effect [2, 15, 16]. On the other hand, several NMR studies to characterize the cellulose-water system of paper have been carried out [17–24]. Considering that in the cultural heritage field, non-invasive and non-destructive in-situ analyses are preferred [25–27], in this work the longitudinal relaxation time (T 1 ) and the transversal relaxation time (T 2 ) of structuralknown filter paper before and after the exposure to UV light were measured by using

268

V. Stagno et al.

portable NMR. This study suggests a preliminary non-invasive protocol for assessing the effect of UV irradiation on paper by using a portable NMR sensor. The results were corroborated by Raman spectroscopy.

2 Materials and Methods A sample of filter paper (Filter-Lab 1244), corresponding to Whatman 44 paper, which is characterized by an alpha-cellulose content around 100% and pore size of 2–4 μm, was studied before and after exposure for 48 h, 52 h and 109 h to UV light produced by a lamp emitting at λ = 365 nm. The lamp power was 36 W and the voltage 230 V– 50 Hz. All measurements were performed at room temperature T = 23 ± 1 °C and relative humidity RH = 30.0 ± 3.5%, which were monitored by using the TROTEC BC06 thermo-hygrometer. 2.1 Portable NMR Measurements The Whatman paper was analyzed before and after the UV exposure for 48 h, 52 h, and 109 h by using NMR relaxometry. A Bruker minispec mq-ProFiler with a single-sided magnet that generates a static magnetic field of 0.35 T was used. It was equipped with a RF-probe for performing experiments with a 2 mm depth from the sample surface, characterized by a 1 H-resonance frequency of 15.78 MHz and dead time of 2 μs. Before and after the paper exposure to UV radiation, the longitudinal relaxation time (T 1 ) and the transversal relaxation time (T 2 ) were measured. T 1 was acquired by a SaturationRecovery (SR) sequence with 64 points from 0.05 to 5000 ms, repetition time (TR) = 0.02 s and number of scans (NS) = 1024. T 2 was acquired by using a Carr-PurcellMeiboom-Gill (CPMG) sequence with TR = 1 s, echo time TE = 30 μs, 200 echoes and NS = 2048. Each T 2 experiment was repeated three times to test the reproducibility, whereas only one T 1 experiment was performed for each sample due to its longer acquisition time. All data were elaborated by using the Inverse Laplace Transform (ILT) algorithm [28, 29] in MATLAB 2021a to obtain the T 1 and T 2 distribution. 2.2 Raman Measurements The Whatman paper was analyzed before and after the UV exposure for 52 h and 109 h by using Raman spectroscopy. The instrumental setup is represented by a Horiba JobinYvon HR Evolution micro-Raman spectrometer. This is equipped with a He-Ne laser (λ = 632 nm), coupled with a microscope with a set of interchangeable objectives. A 20x objective was used. Intensity of radiation has been set at 15 mW by neutral filters; the acquisition time was set at 20 s for each scan, while the number of acquired scans at 30 scans. Moreover, spectra were acquired in four different points of the unaged (not exposed to UV light) and aged (exposed to UV light for 52 h and 109h) samples, then averaged, and a polynomial function was fitted to the baseline and subtracted.

Testing Portable NMR to Monitor the Effect of Paper Exposure

269

3 Results and Discussions This work tested the potential of low-field single-sided NMR to assess paper conservation state after an artificial UV-light aging process. To this end, we measured the T 1 and T 2 parameters before and after the UV exposure of Whatman paper and we compared the NMR results with those obtained from Raman spectroscopy. 3.1 Portable NMR Results In Fig. 1 the longitudinal relaxation time (T 1 ) distribution of the filter paper before the UV exposure (black dashed line) and after the UV exposure for 48 h (green line), for 52 h (red line) and for 109 h (blue line) is shown. The unaged sample and the UV-aged for 48 h sample are characterized by two T 1 components, whereas the samples exposed for 52 h and 109 h to UV light present only one T 1 component. The unaged sample has one longer T 1 of about 3 s and one shorter T 1 of 0.10 ± 0.02 s. After UV-exposure for 48 h, the longer T 1 component shortened around 0.3 s and the shorter T 1 component (=0.089 ± 0.01 s) did not significantly change. This T 1 component, which is around 0.1 s, was detected for all the samples and remained the same after all the different UV-aging times. According to the literature [20, 22], the NMR relaxation times components of paper change with its moisture content. Under environmental condition at T = 20 °C and RH = 50% the water content of paper varies from 5% to 8%. At this hydration level, three different hydrogens populations called free water, bound water and cellulose protons, associated with three different values of T 1 , can be identified. Free water is more mobile and confined in the amorphous cellulose [17] and has one long T 1 component of about 3–4 s, which decreases by increasing the paper hydration. Bound water shows one intermediate T 1 of less than 1 s, which does not change with paper hydration. Cellulose protons reservoir is characterized by one short T 1 of few ms, which increases with paper hydration and tends to zero when the water content of the paper is very small. In accordance with the T 1 dependence on the magnetic field strength, in this study at low-field (15 MHz), we detected shorter T 1 values than those measured by Mallamace et al. [22] at high-field (700 MHz). Moreover, we only measured two T 1 components of about 3 s and 0.1 s, but we did not detect the short T 1 component of cellulose protons. This can be due to the shortening of T 1 with the decrease of the magnetic field strength and/or to the fact that our samples were maintained at T = 23 °C and RH = 30%, which corresponds to a hydration level below 5% at which the T 1 component of cellulose protons is approximately zero or too short to be detected. Mallamace et al. [22] explained that the T 1 of free water exhibits a first increment followed by a linear reduction as function of the thermal-degradation days. This is because the first step of thermal-degradation consists in the rupture of intramolecular hydrogen bonds of paper. Moreover, the T 1 of bound water and cellulose protons show a slight decrease with degradation. In our case, we only observed the shortening of T 1 peak associated with free water after 48 h of UV-exposure and its disappearance in the UV-aged for 52 h and 109 h samples. This behaviour can be ascribable to the effect of the UV-aging, which produces the loss or reduction of the free water amount. It may be possible that after 48 h of UV-aging, free water was completely lost from

270

V. Stagno et al.

paper or that the T 1 peak of free water was overlapped to the peak at T 1 = 0.1 s [17]. This T 1 component of 0.1 s associated with dynamically restricted water did not exhibit a significant shortening after the UV-exposure, indicating that bound water was not affected by significant variations induced by the UV-aging.

Fig.1. T 1 distribution of not aged paper sample and of UV-aged paper sample for 48 h, 52 h and 109 h.

In Fig. 2 the transversal relaxation time (T 2 ) distribution of the filter paper before the UV exposure (black dashed line) and after the UV exposure for 48 h (green line), for 52 h (red line) and for 109 h (blue line) is displayed. It is possible to notice that both the unaged and aged samples of paper are characterized by three T 2 components: around 0.2 ms, 1 ms and tens of ms. The component of tens of milliseconds is characterized by the lowest intensity, whereas the component of about 0.2 ms has the highest intensity. Moreover, Fig. 2 shows a clear decrease of T 2 with the increase of UV exposure only after 52 h. The component mainly affected by the UV-aging is that one around tens of milliseconds which exhibits a significant shortening passing from about 40 ms for the unaged sample to about 1.7 ms for the UV-aged paper for 109 h, in accordance with the T 1 trend (Fig. 1). According to the literature [22], this component can be associated with the hydration water (free water) present in the pores of paper (with diameter around 2–4 μm), which at RH = 30% occupies around the 3% of weight of cellulose. After the UV-exposure of 109 h, this T 2 component has a higher probability which can be due to a greater amount of free water but characterized by low mobility compared to that of the paper samples UV-aged for shorter times. Since free water molecules only occupy the amorphous domain of cellulose, the increase of intensity of the free water component can be ascribed to the increase of the water accessible sites of the amorphous

Testing Portable NMR to Monitor the Effect of Paper Exposure

271

cellulose as also detected by Tortora et al. [17]. However, due to the low intensity of this T 2 component, our result must be taken with care.

Fig. 2. T 2 distribution of not aged paper sample and of UV-aged paper sample for 48 h, 52 h and 109 h.

The two most intense T 2 components may be interpreted as two different hydrogen populations characterized by a different mobility. In accordance with the two-phase water model in the cellulose network described in the literature [4, 17, 24, 30, 31], the two T 2 components in Fig. 2 correspond to one more mobile hydrogen compartment (T 2 around 1 ms), almost exclusively confined in the amorphous cellulose, and one less mobile hydrogen compartment (T 2 around 0.2 ms) fast exchanging with hydroxyl groups of the cellulose chains [17]. The shorter component of T 2 associated with less mobile hydrogens does not change passing from the unaged sample to the UV-aged for 48 h sample. On the contrary, it shows a significant decrease both after the UV exposure for 52 h and for 109 h. In particular, the unaged and aged for 48 h samples are characterized by T 2 of 0.25 ± 0.01 ms, whereas the UV-aged for 52 h sample has T 2 of 0.21 ± 0.01 ms and the UV-aged for 109 h sample has T 2 of 0.19 ± 0.01 ms. This indicates that some modifications in the paper structure and/or in the paper-water interaction took place after 48 h of UV exposure, which are reflected in the transversal relaxation time. Furthermore, the longer T 2 component seems to be more affected by the UV-aging for 109 h since its shortening is stronger. Indeed, this T 2 component, associated with more mobile hydrogens, is 1.00 ± 0.02 ms, 0.96 ± 0.05 ms, 0.91 ± 0.08 ms and 0.64 ± 0.05 ms, for unaged paper, UV-aged for 48, 52 and 109 h paper, respectively. This trend suggests that the bound water mobility significantly decreases at 109 h of UV-exposure.

272

V. Stagno et al.

The NMR results discussed above suggest that the UV exposure of paper for at least 52 h affects the water dynamics and content with a general decrease of both T 1 and T 2 in agreement with Paci et al. [21]. Specifically, the decrease of the T 2 component associated with the cellulose protons domain and the increase of the probability associated with the T 2 peak of free water are index of the UV-induced structural modification of cellulose chains. Indeed, according to Tortora et al. [17], the T 2 component of the cellulose protons is almost independent from relative humidity, therefore its decrease indicates that a modification of the cellulose hydroxyl groups took place and was observed only at 52 h of UV-aging. As consequence the exchange between hydroxyls and water molecules slows down causing a faster spinspin relaxation. The increase of the peak intensity of the free water suggests an increment of the accessible sites of the amorphous cellulose. 3.2 Raman Results In Fig. 3 the Raman spectra of the unaged (black line), UV-aged for 52 h (red line) and UV-aged for 109 h (blue line) paper is shown. The main differences before and after the UV exposure are observable for the signals between 1200 and 1600 cm−1 , indicated by black arrows. The comparison of Raman spectra obtained for Whatman paper before and after UV-aging in Fig. 3 shows some clear variations. The Raman peaks modes at 1094 and 1119 cm−1 are considered indicative of β-(1,4)-glycosidic linkage stretching and, consequently, of interest in several aging studies [32, 33], so the peak at 1094 cm-1 was used for normalization of the spectra in the range 720–1800 cm−1 . With reference to this normalization, the main differences before and after the UV exposure are observable for the signals between 1200 and 1600 cm−1 . In particular, the signals at 1290, 1339 and 1370 cm−1 , attributable to CH2 bending modes, are subjected to a decrease of intensity. The weak band at 1479 cm−1 , which is considered representative of high crystallinity arrangement, decreases in intensity, while its analogue at 1467 cm−1 increases in intensity with a clear broadening, especially at higher wavenumbers. In the C-H stretching range, between 2500 and 3200 cm−1 , no evident variations of signal is observable except for a higher intensity signal at 2968 cm−1 . With reference to the fundamentals of Chiriu model proposed for the study of cellulose aging [34], the intensity ratio between the peak at 1094 cm−1 and the CH2 bending at 1370 cm−1 was calculated before and after the UV exposure. Usually, a decrease in this ratio is considered indicative of standard aging because it is related to the decrease of polymerization degree in cellulosic polymers. However, as observable from the spectra, this ratio increases from 1.62 to 1.87 in the 52 h aged paper, while after the 109 h aging it is stable at around 1.86. This suggests a diminution of β-(1,4)-glycosidic linkage amount in comparison to the CH2 content of single glucose units. This is also confirmed by the increase from 1.30 to 1.43 of the intensity ratio between the signal at 1094 cm−1 and the C-H stretching peak at 2894 cm−1 , also considered, according to the literature [32], a marker of breaking of glycosidic bonds. The intensity decrease of the glycosidic signals is reported by Pandey et al. [35] in the case of UV-irradiated cellulose and it could be considered indicative of photodegradation processes induced by this radiation. It is also

Testing Portable NMR to Monitor the Effect of Paper Exposure

273

Fig. 3. Raman spectra of not aged (black line), UV-aged for 52 h (red line) and UV-aged for 109 h paper (blue line). Black arrows indicate the major variations of Raman shift of the UV-aged paper compared to the unaged paper.

interesting to notice that, in other studies where natural aged paper were investigated [34], the decrease of this intensity ratio was considered a marker of progressive aging, but this should not be considered contradictory: in fact, the natural aging of cellulose could not depend from UV exposure, because this typology of material is usually stored in indoor environments, where the UV radiation could be totally (UVC and UVB) or partially (UVA) filtered. This was also confirmed, in literature, from the comparison of UV aging with Xenon-lamp induced one [35]. Moreover, in the presented spectra, the cleavage of glycosidic bond could be confirmed also from the slight broadening in the wavenumber range between 1000 and 1070 cm−1 , which could suggest the formation of new distinct chemical environments. The increase in intensity of the signal at 1467 cm−1 ,

274

V. Stagno et al.

related to amorphous arrangement in cellulosic materials [32], could be indicative of a lower crystallinity in the paper during the aging, with the disappearance of the crystalline cellulose fraction in favour to the non-crystalline one; this would be confirmed also by the broadening of the band over 1500 cm−1 . It is also worth to notice that no typical signal of oxidation around 1630 cm−1 is observable in the spectra after the aging, even if this was reported in previous study [32]. However, this should not be considered indicative, because in the presented work the UV exposure time was not prolonged, so it is possible that this phenomenon is not clearly detectable.

4 Conclusion In this work the potential of low-field single-sided NMR in assessing and monitoring the conservation state of paper during the UV-aging process was tested and the NMR results were compared with Raman spectroscopy. The two techniques provided complementary information [36] about the paper state after UV exposure: changements in the watercellulose interaction were detected by portable NMR, whereas modifications of the cellulose bonds were assessed by Raman spectroscopy. The preliminary non-invasive protocol based on portable NMR T 1 and T 2 relaxation showed that Whatman paper at T = 23 °C and RH = 30% is characterized by three T 2 components: free water, bound water and water in fast exchange with the hydroxyls of cellulose. The T 1 and T 2 associated with these three compartments undergo a decrease with the increase of UV-exposure time. The minimum exposure time needed to observe changes in the NMR relaxation times is 52 h. Most of the NMR signal comes from water interacting with the OH groups of cellulose and, despite its T 1 was not detected, the decrease of T 2 associated with this domain indicates a slowdown of the exchange rate among the hydroxyls of cellulose and water, which is a consequence of the UV-induced modification of the cellulose chains. Raman spectra suggested that both after the UV exposure of paper for 52 h and 109 h, the β-(1,4)-glycosidic linkage amount decreases and the increase in intensity of the signal at 1467 cm−1 , related to amorphous arrangement in cellulosic materials, could be indicative of a lower crystallinity in the paper. This is in agreement with the increased intensity at 109 h of UV-aging of the free water T 2 peak. In conclusion, this work suggests that portable NMR T 1 and T 2 relaxation times are sensitive to the effect of UV irradiation on paper. For future work, in order to study the overall dynamics of paper decay caused by UV-light it is recommended to increase the exposure time. However, it is worth to mention that portable NMR was able to detect small changes in the sample even if the exposure time was not sufficiently long to induce a strong degradation of the Whatman paper.

References 1. Kolar, J., Strliˇc, M., Novak, G., Pihlar, B.: Aging and stabilization of alkaline paper. J. Pulp Pap. Sci. 24(3), 89–94 (1998)

Testing Portable NMR to Monitor the Effect of Paper Exposure

275

2. Robotti, E., Bobba, M., Panepinto, A., Marengo, E.: Monitoring of the surface of paper samples exposed to UV light by ATR-FT-IR spectroscopy and use of multivariate control charts. Anal. Bioanal. Chem. 388(5–6), 1249–1263 (2007). https://doi.org/10.1007/s00216007-1370-4 3. Micheli, L., Mazzuca, C., Palleschi, A., Palleschi, G.: Combining a hydrogel and an electrochemical biosensor to determine the extent of degradation of paper artworks. Anal. Bioanal. Chem. 403(6), 1485–1489 (2012). https://doi.org/10.1007/s00216-012-5885-y 4. Ibbett, R., Wortmann, F., Varga, K., Schuster, K.C.: A morphological interpretation of water chemical exchange and mobility in cellulose materials derived from proton NMR T2 relaxation. Cellulose 21(1), 139–152 (2013). https://doi.org/10.1007/s10570-013-0106-1 5. Blüher, A., Vogelsanger, B.: Mass deacidification of paper. Chimia (Aarau) (2001) 6. Strlic, M., Kolar, J.: Evaluating and enhancing paper stability - needs and recent trends. Cult. Herit. Res. a pan-European Chall. (2003) 7. Bara´nski, A., Konieczna-Molenda, A., Łagan, J.M., Proniewicz, L.M.: Catastrophic room temperature degradation of cotton cellulose. Restaurator (2003).https://doi.org/10.1515/ REST.2003.36 8. Kolar, J.: Mechanism of autoxidative degradation of cellulose. Restaurator (1997).https://doi. org/10.1515/rest.1997.18.4.163 9. Malešiˇc, J., et al.: Photo-induced degradation of cellulose. Polym. Degrad. Stab. 89(1), 64–69 (2005). https://doi.org/10.1016/j.polymdegradstab.2005.01.003 10. Phillips, G.O., Hinojosa, O., Arthur, J.C., Mares, T.: Photochemical initiation of free radicals in cotton cellulose. Text. Res. J. 36(9), 822–827 (1966). https://doi.org/10.1177/004051756 603600909 11. Phillips, G.O., Arthur, J.C.: Chemical effects of light on cotton cellulose and related compounds: part I: primary processes in model systems. Text. Res. J. 34(6), 497–505 (1964). https://doi.org/10.1177/004051756403400604 12. Phillips, G.O., Arthur, J.C.: Chemical effects of light on cotton cellulose and related compounds: part II: photodegradation of cotton cellulose. Text. Res. J. 34(7), 572–580 (1964). https://doi.org/10.1177/004051756403400702 13. Hon, N.S.: Formation of free radicals in photoirradiated cellulose VIII. Mechanisms. J. Polym. Sci. Polym. Chem. Ed. 14(10), 2497–2512 (1976). https://doi.org/10.1002/pol.1976. 170141015 14. Hon, D.N.S.: Photooxidative degradation of cellulose: reactions of the cellulosic free radicals with oxygen. J. Polym. Sci. Polym. Chem. Ed. 17(2), 441–454 (1979). https://doi.org/10. 1002/pol.1979.170170214 15. Yatagai, M., Zeronian, S.H.: Effect of ultraviolet light and heat on the properties of cotton cellulose. Cellulose 1, 205–214 (1994). https://doi.org/10.1007/BF00813508 16. Lee, S.B., Bogaard, J., Feller, R.L.: Damaging Effects of Visible and Near-Ultraviolet Radiation on Paper, pp. 54–62 (1989). https://doi.org/10.1021/bk-1989-0410.ch004 17. Tortora, M., Sfarra, S., Casieri, C.: NMR relaxometry and IR thermography to study ancient cotton paper bookbinding. Appl. Sci. 9(16), 3406 (2019). https://doi.org/10.3390/app9163406 18. Capitani, D., Segre, A.L., Attanasio, D., Blicharska, B., Focher, B., Capretti, G.: 1H NMR relaxation study of paper as a system of cellulose and water (1996) 19. Bana, B.K.: Molecular dynamics of cellulose-water systems investigated by NMR relaxation method, vol. 54, pp. 501–504 (2000) 20. Froix, M.F., Nelson, R.: The interaction of water with cellulose from nuclear magnetic resonance relaxation times. Macromolecules 8(6), 726–730 (1975) 21. Paci, M., Federici, C., Capitani, D., Perenze, N., Segre, A.L.: NMR study of paper. Carbohydr. Polym. 8617, 289–297 (1995)

276

V. Stagno et al.

22. Mallamace, D., Vasi, S., Missori, M., Mallamace, F., Corsaro, C.: NMR investigation of degradation processes of ancient and modern paper at different hydration levels. Front. Phys. 13(1), 1–10 (2018). https://doi.org/10.1007/s11467-017-0686-6 23. Grunin, L.Y., Grunin, Y.B., Nikolskaya, E.A.: An NMR relaxation and spin diffusion study of cellulose structure during water adsorption. Biophysics 62, 198–206 (2017). https://doi. org/10.1134/S0006350917020087 24. Conti, A., Palombo, M., Parmentier, A., Poggi, G., Baglioni, P., De Luca, F.: Two-phase water model in the cellulose network of paper. Cellulose 24(8), 3479–3487 (2017). https://doi.org/ 10.1007/s10570-017-1338-2 25. Stagno, V., Mailhiot, S., Capuani, S., Galotta, G., Telkki, V.V.: Testing 1D and 2D single-sided NMR on Roman age waterlogged woods. J. Cult. Herit. 50, 95–105 (2021). https://doi.org/ 10.1016/j.culher.2021.06.001 26. Stagno, V., Genova, C., Zoratto, N., Favero, G., Capuani, S.: Single-sided portable NMR investigation to assess and monitor cleaning action of PVA-borax hydrogel in travertine and Lecce stone. Molecules 26(12), 3697 (2021). https://doi.org/10.3390/molecules26123697 27. Stagno, V., Ciccola, A., Curini, R., Postorino, P., Favero, G., Capuani, S.: Non-Invasive Assessment of PVA-Borax Hydrogel Effectiveness Portable NMR (2021) 28. Venkataramanan, L., Yi-Qiao Song, M.D., Hurlimann: Solving Fredholm integrals of the first kind with tensor product structure in 2 and 2.5 dimensions. IEEE Trans. Signal Process. 50(5), 1017–1026 (2002) 29. Berman, P., Levi, O., Parmet, Y., Saunders, M., Wiesman, Z.: Laplace inversion of lowresolution NMR relaxometry data using sparse representation methods. Concepts Magn. Reson. Part A 42(3), 72–88 (2013). https://doi.org/10.1002/cmr.a.21263 30. Topgaard, D., Söderman, O.: Diffusion of water absorbed in cellulose fibers studied with 1H-NMR. Langmuir 17(9), 2694–2702 (2001). https://doi.org/10.1021/la000982l 31. Lindh, E.L., Terenzi, C., Salmén, L., Furó, I.: Water in cellulose: Evidence and identification of immobile and mobile adsorbed phases by 2H MAS NMR. Phys. Chem. Chem. Phys. 19(6), 4360–4369 (2017). https://doi.org/10.1039/c6cp08219j 32. Ana, P.P., Alves, L.P.Z., de Oliveira, A.A.N., Castro, R.N., de Oliveira, L.F.C., Edwards, H.G.M., Antonio, C., Sant’Ana: The structure of different cellulosic fibres characterized by Raman spectroscopy. Vib. Spectrosc. 86, 324–330 (2016). https://doi.org/10.1016/j.vibspec. 2016.08.007 33. Wiley, J.H., Atalla, R.H.: Band assignments in the raman spectra of celluloses. Carbohydr. Res. 160, 113–129 (1987). https://doi.org/10.1016/0008-6215(87)80306-3 34. Chiriu, D., Ricci, P.C., Cappellini, G., Salis, M., Loddo, G., Carbonaro, C.M.: Aging of ancient paper: a kinetic model of cellulose degradation from Raman spectra. J. Raman Spectrosc. 49(1), 1802–1811 (2018). https://doi.org/10.1002/jrs.5462 35. Pandey, K.K., Vuorinen, T.: Comparative study of photodegradation of wood by a UV laser and a xenon light source. Polym. Degrad. Stab. 93(12), 2138–2146 (2008). https://doi.org/10. 1016/j.polymdegradstab.2008.08.013 36. Longo, S., Capuani, S., Corsaro, C., Fazio, E.: Silver fir characterized by micro-imaging NMR and FTIR spectroscopy. In: IOP Conference Series: Materials Science and Engineering (2020). https://doi.org/10.1088/1757-899X/777/1/012004

Non-invasive Analysis of the Pigment Palette Used by the Renaissance Painter Sofonisba Anguissola Anna Rygula1(B) , Marta Matosz1 , Alicja Mogielska2 , Magdalena Iwanicka3 , Piotr Targowski4 , Michał Obarzanowski1 , and Julio M. del Hoyo-Meléndez1 1 Laboratory of Analysis and Non-destructive Investigation of Heritage Objects, National

Museum in Kraków, Kraków, 3 Maja 1, 30-062 Kraków, Poland [email protected] 2 Museum – Castle in Ła´ncut, Art Conservation Department, Zamkowa 1, 37-100 Ła´ncut, Poland 3 Faculty of Fine Arts, Nicolaus Copernicus University in Toru´n, ul. Sienkiewicza 30/32, 87-100 Toru´n, Poland 4 Institute of Physics, Faculty of Physics, Astronomy and Informatics, Nicolaus Copernicus University in Toru´n, ul. Grudzi˛adzka 5, 87-100 Torun, Poland

Abstract. A multi-disciplinary project focusing on the non-invasive analysis of the early painting of Sofonisba Anguissola (1532–1625), “Self-portrait at an Easel” (Museum-Castle in Ła´ncut, Poland) was implemented out applying macroand micro-analytical techniques. Technical photography using UV radiation and X-radiography were used along with macro-XRF to better understand the materials and techniques of this painting. Then, microscopic non-invasive techniques like single-point micro-XRF, micro-Raman and optical coherence tomography (OCT) were used as complementary methods to enhance the level of detail of results obtained with macro-techniques. The pigments used by Sofonisba Anguissola include: vermilion, lead white, smalt with indigo, goethite with impurities like jarosite and anatase, vegetable and bone carbon black, a copper green pigment, and massicot. Macro-XRF technique detected the presence of Ca and Pb in the ground of the painting suggesting the application of an oil-gypsum ground layer with addition of lead white. The palette used for “Self-portrait at an Easel” was compared to another painting from the same period, “The Game of Chess” (National Museum in Poznan, Poland) showing similarities in the techniques. Keywords: Sofonisba Anguissola · XRF · Raman spectroscopy · OCT · Non-invasive analysis

1 Introduction Sofonisba Anguissola was born in 1532 in Cremona as the oldest daughter of Bianca Ponzoni and Amilcare Anguissola. Her relatively poor but noble family provided her a © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 277–291, 2023. https://doi.org/10.1007/978-3-031-17594-7_21

278

A. Rygula et al.

good education. From a young age, Sofonisba was interested in the humanities and fine arts, thus her talent was recognized. Her first teacher was Bernardino Campi. Initially, she mostly painted small religious paintings, but quickly she became noticed as a portrait painter. Sofonisba was thirty years old when she became famous as official court painter to the Spanish king Filip II. In her long life she also lived and worked in Palermo (1573– 1579), had a short stay in Piza (1579–1580), then moved to Genoa (1580–1615), and later returned to Palermo (1615–1625). Before she died at age 93, she had taught young Anthony Van Dyck for a short time and had given him some advice on portrait paintings [1]. In the painting Self-portrait at an Easel, (Fig. 1A) Sofonisba Anguissola shows herself in three-quarter view, while working. The artist in her twenties calmly looks at the viewer, dressed in dark clothes with a white lace collar and cuffs. The background is dark green. At the left, there is an oil painting on an easel. She is painting Madonna and Child and fortunately, there is a chance to have a look at her palette: white, red, brown, brown-green, and black paints can be recognised. The artist painted Self-portrait at an Easel between 1554–1556 [2], in her formative period. There are stylistic similarities to other Sofonisba’s paintings, such as Self-portrait at the Spinet (1556–57, Naples, Museo e Real Bosco di Capodimonte) or Self-portrait with an inscription written in an open book held by the depicted figure: SOPHONISBA/ ANGUSSOLA/ VIRGO/ SIPSAM/ FECIT/ 1554 (1554, Vienna, Kunsthistoriesches Museum Wien, Gemadegalerie, inv. no 285) [3]. Unfortunately, there is no information about the provenance of the examined painting or how it arrived at the Ła´ncut Castle [4]. It is likely that this work was purchased by Princess Izabella Lubomirska (El˙zbieta Izabela née Czartoryska), probably in the second half of the 18th century as part of one of her numerous journeys around Europe. Izabella was the greatest art patron in eighteenth century Poland. The Ła´ncut Castle became her main residence where she gathered a valuable collection of works of art [5]. In 1944 the painting was taken over from the Potocki family when the Soviets entered Ła´ncut. Subsequently, the museum began to be formed with all left behind Potocki family property and artefacts [6]. Now the painting Self-portrait at an Easel by Sofonisba Anguissola is exhibited in the Red Corridor on the first floor in the eastern wing of the Ła´ncut Castle. This work has been shown at numerous European exhibitions, first in 1994 in Cremona, then in Vienna, Madrid, Warsaw, and Washington DC [7]. The first documented conservation of the Self-portrait at an Easel comes from 1964. At that time, the painting was wax-lined and the varnish layer was cleaned [8]. In the next documentation from 1994, it is mentioned that before the 1964 intervention, the painting was overcleaned (especially the Madonna’s clothes) [9]. Therefore, Sofonisba’s painting was restored at least three times. Because there is no additional information, the Castle Museum in Ła´ncut decided to conduct research under the Polish Research Infrastructure for Heritage Science (E-RIHS.pl) project. This multi-institutional national consortium provides research services to cultural institutions in Poland. A multi-disciplinary project was carried out using non-invasive micro- and macroanalytical techniques. Besides UV-excited luminescence photography (not shown) and X-radiography, macro-XRF spectrometry measurements provided important information about the overall structure of this cultural heritage object. Furthermore, microscopic

Non-invasive Analysis of the Pigment Palette

279

non-invasive techniques including single point micro-XRF, micro-Raman spectroscopy and optical coherence tomography (OCT) were used to support and expand the results obtained using macro-analytical techniques. Additionally, to obtain a more comprehensive knowledge about the painter’s workshop the research results obtained for the paintings Self-portrait at an Easel and The Game of Chess were compared. The latter was previously analysed in the LANBOZ Laboratory in 2015 (Fig. 1B).

Fig. 1. Sofonisba Anguissola paintings: A: Self-portrait at an Easel (1554–1556, 65.7 cm × 59 cm, oil on canvas), B: The Game of Chess (1555, 72 cm × 97 cm, oil and tempera on canvas) – National Museum in Pozna´n, Poland

The Game of Chess painting was created in 1555 when the artist was 23 years old. Sofonisba introduced her three sisters and a nurse who are in the garden or on the terrace, against the background of an oak tree in the middle of the landscape. Careful examination and research of the object allowed us to state that the painting was subjected to conservation treatments several times over the centuries. Micro-XRF measurements were conducted together with technical photography of UV-excited fluorescence, and in the VIS and IR ranges. A summary of the analytical results is available in an earlier publication [10]. According to our best knowledge, there is no published research about the technique and technology of Sofonisba Anguissola’s paintings.

2 Experimental 2.1 X-radiography X-radiographs were obtained using a DIX-Ray system (EXAMION GmbH, Niemcy), which is based on a wireless panel with dimensions 46×38 cm. The distance between the X-ray tube’s forehead and the detector surface is 200 cm. The exposures were carried out using voltage and current of 40 kV and 40 mA, respectively.

280

A. Rygula et al.

2.2 Macro-XRF Two-dimensional macro-XRF (MAXRF) scans of the painting were carried out using an M6 Jetstream spectrometer (Bruker, DE). The instrument consists of a measuring head containing a Rh-target X-ray tube and a 30-mm2 X-Flash® silicon drift detector (SDD) mounted on an XY-motorized stage. The X-rays are focused through polycapillary optics, which allow to set the beam size by changing the working distance. The pixel size of the elemental maps, determined by the step size, was set to 2 mm, with the beam spot size of 650 μm. The voltage and current of the X-ray tube were set to 30 kV and 600 μA, respectively. The acquisition time was 20 ms/pixel. The data was analysed with use of Bruker M6 1.6.615.0 software. 2.3 Micro-XRF Micro-XRF analysis were conducted using an ARTAX 800 spectrometer (Bruker, DE) equipped with a Rh tube and policapillary optics, which provide 410 nm) at high magnification (1x, 4x, 10x and 20x) was used. OM has been performed also for cross-sections observations (C2, C4-5, C7-8, C10, C12-14, C16-21). X-Ray Diffraction (XRD) XRD analysis was adopted for the detection of degradation products present on the analysed samples. An X-ray diffractometer PANalytical X’PertPRO (Cu anticathode, λ = 1.54 Å), equipped with an X’Celerator multidetector with a current intensity of 30 mA, voltage 40 kV, explored 2θ range between 3 and 70°, step size 0.04°, time to step 50 s and acquisition time of 17 min was employed. High Score software was utilized for acquisition and data processing and the ICDD diffraction database was taken as reference. XRD was performed both on powder and a zero-background plate. Fourier-Transform Infrared Spectroscopy (FTIR) FTIR spectroscopy was chosen to provide information about alteration products, binders, varnishes and consolidants used during restoration treatments. A portable ALPHA FTIR (Bruker Optics GmbH) device with 4000–400 cm−1 operating range, 4 cm−1 resolution, 24 scans, SiC Globar source and a DTGS detector equipped with the ATR and Transmission modules were employed. Bruker OPUS 7.2 software and Thermo Electron’s OMNIC spectroscopy software were employed for spectra analysis. Raman Spectroscopy Micro-Raman spectroscopy was involved for achieve final information on pigments, degradation, and restoration materials. A micro-Raman setup using a Senterra dispersive Raman microscope (Bruker Optik GmbH) equipped with a Peltier cooled CCD detector (1,024 × 256 pixels) and a 785 nm excitation laser. Spectra were acquired with a laser power ranging from 10 to 100 mW, 20x, 50x and 100x magnification, 50 μm pinhole, acquisition times between 50–200 s and spectral range between 90–3500 cm−1 .

3 Results and Discussion By visual inspection, restoration interventions, such as facings, are evident (Fig. 2a). Chromatic inhomogeneities, yellowing of the varnish layers, cracking and detachment of the pictorial layers are visible on the entire surface of the ceiling (Fig. 2b). Salt efflorescence and darkening of pigments were also present (Fig. 2c). Traces of burnt areas of the 1690 fire are still observable (Fig. 2d). This complex number of various materials is also confirmed by XRF collected spectra. Copper presence in C4; gilded samples (C7, C8) investigations showed the presence of Au, Fe and Ca suggesting a bolo gilding; red colour samples (C10 and C12) spectra

380

S. Longo et al.

Fig. 2. Visual observation of alteration phenomena. a) facings and yellowing; b) cracking and detachment; c) chromatic inhomogeneities and salt efflorescences; d) cracking, detachment, chromatic inhomogeneities and burnt material.

highlight the presence of Hg suggesting the existence of cinnabar or vermilion red pigment while orange colour samples (C13 and C20) were probably made by iron-based

Fig. 3. Comparison of monitored parameters by dataloggers versus the registered data by the centre Florence weather station of CRF (Functional Regional Centre of Tuscany Region). (a) Plot of temperature (T) values with relation to the rainfall; (b) Plot of relative humidity (RH) values with relation to the rainfall.

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

381

earth pigments. Furthermore, XRF spectra acquired on blue colour samples (C10, C19, C21) showed the presence of cobalt probably due to blue smalt. The presence of Pb and Ca can be referred to as the preparation layers. Samples description, XRF major, minor and trace elements are reported in Table 1 in appendix. As concerns, the thermo-hygrometric monitoring, the acquired data, started in March 2021, in the attic (Tint and RHint ) and in the loggia close to the ceiling (Text and RHext ), have been compared with the data - temperature (T), relative humidity (RH) and rainfall - registered by the CRF weather station located in the centre of Florence. Comparing the results of monitoring performed by datalogger in outdoor and the data registered from the weather station no-significant difference can be reported. The green lines, both for temperature and relative humidity, represent the parameters registered by the datalogger inside the attic, whilst the blue lines represent the environmental parameters in outdoor. A different behaviour in temperature and humidity is evident between the attic and outdoor in particular during the winter when the heating is turned on. However, the variations on the relative humidity inside the attic are less than in the loggia (Fig. 3a–b). During the summer, the temperature inside the attic can reach values higher than in the the loggia environment (Fig. 4a), but similarly at the relative humidity the night/day spread of the registered values are less than to the outdoor environment parameters. The trend of the relative humidity follows with a delay of about four days the rainfall event (Fig. 4b). The different exposure to environment factors of the wooden ceiling between attic side and outdoor side can activate strengths and strains favoring the cracks and craquelure formation on the painted surface. Preliminary optical microscopy observation on samples showed that they are composed of various layers. Multi-layers stratigraphies were confirmed also by cross-section observation. About the colour’s palette, traces of green pigment particles together with darker ones could be observed in C4 (Fig. 5a). Moreover, traces of blue particles in C10 (Fig. 5b) were found. OM analysis allows the identification of wood anatomical details on sample C21 (Fig. 5c). In Fig. 5d cross-section of C21 shows a semi-diffuse porosity and monolayered rays. These morphological characteristics, compared with references [3] make possible the identification of a deciduous wood from the Salicaceae family probably related to Populus L. genus. Green particles observed at OM in sample C4 were identified as malachite, together with tenorite, its degradation matrix, by XRD patterns. Hydrocerussite and cerussite lead matrices were observed on samples (C1, C2 and C17). Palmierite and plattnerite, degradation products of lead were also detected. The presence of baryte (C14, C17) could be probably ascribed to the 1860s restoration. [4] Moreover, epsomite (C14), a soluble salt connected to the formation of efflorescence on painted surfaces, was detected. [5] Crystalline compounds such as gypsum, anhydrite, calcite and quartz are assigned to the ground layer (Table 2 of the appendix). FTIR spectroscopy results provided relevant information on degradation products, the binders, varnishes and consolidants products used during restoration treatments. In particular, a widespread presence of calcium oxalates (1622, 1315 cm−1 ) was attested

382

S. Longo et al.

Fig. 4. Magnification of the summer period at the end of August. (a) Plot of temperature (T) values cons to the rainfall; (b) Plot of relative humidity (RH) values cons to the rainfall.

Fig. 5. Optical Microscopy images: a) sample C4 image of green pigment particles (RL, 10x); b) sample C10 image of multi-layer cross-section (RL, 10x); c) and d) sample C21 semi-diffuse porosity and monolayered rays are visible (RL 4x, 10x)

on the majority of the samples. Its origin could be caused by binding media degradation (especially if lipidic) or to wrong conservation treatments and/or by atmospheric pollution [6]. Zinc oxalates, another dangerous alteration product was detected on samples (C6, C20). The presence of zinc oxalates is also confirmed by the disappearance or decrease of characteristic oil binder frequencies (e.g., C-H stretches (3000–2800 cm−1 ), C = O stretches (1750–1730 cm−1 ), C-O stretches (1200–1100 cm−1 ) and the alkyd

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

383

peak at 1270 cm−1 ). A recent paper suggests a possible mineralisation process describing the formation of metal oxalates as a further conversion of metal soaps [7]. The presence of linseed oil as a binder (assignments 1740, 2856 and 2930 cm–1 ) probably could be referred both to Vasari’s painting technique [8, 9] and during recent interventions as well. A metal soap of calcium (calcium oleate) was found on sample C4 (2917, 2847 cm−1 , 1585, 1407 cm−1 (COO- stretches). An oleate is an unsaturated fatty acid formed after hydrolysis of fatty esters in the oil medium, causing the so-called blooming effect. Analyses revealed the possible presence of calcium phosphate; literature suggests that this alteration product could be caused by the interaction of phosphates from carbon black with other materials [30]. Acrylic resins (methyl methacrylates and PMMA) were found on samples interested in restoration treatments and some restoration reports documented the use of Primal AC33. Moreover, analysis performed on sample C23 showed the presence of gypsum with possible lead white as imprimitura ground layer (Fig. 6) [10].

Fig. 6. FTIR-T spectrum of sample C23 in KBr pellet. Gypsum and white lead bands are visible.

Micro-Raman spectroscopy analysis provided some final information on materials. Due to a strong fluorescence, several samples (C2, C3 and C13) could be not observed, while analyses on blue colour samples (C14 and C19) revealed the presence of phthalocyanine blue, a synthetic organic pigment that was first synthesised and marketed in 1933 [11]. This information permits to date back Area 3 (Fig. 1) to the last restoration campaign conducted from 1948 to the 1960s. In sample C14, the blue pigment was present together with baryte and gypsum, while in sample C19, pyrolusite (658 cm−1 , 753 cm−1 ), gypsum, calcite, goethite (from the yellow grains) and carbon black were detected as well. No reliable bands could be observed for a cobalt-based pigment, therefore results obtained by XRF spectroscopy could not be confirmed. The band for pyrolusite was weak but in accordance with the literature [12]. Analyses revealed the presence of Prussian blue on samples C4 and C17 (527 cm−1 ). In sample C4, the pigment was found in a thin white layer on top of the malachite green layer (429 cm−1 , 268 cm−1 , 175 cm−1 ,

384

S. Longo et al.

148 cm−1 ), together with gypsum, lead white and grains of amorphous goethite [13] (545 cm−1 , 393 cm−1 ) and hematite (411 cm−1 , 295 cm−1 ). Raman analysis of sample C17 confirmed the presence of plattnerite (515 cm−1 ) in the inner blue paint layer, also identified as azurite mixed with particles of cinnabar. Bands were compared with the literature and found compatible [14]. The use of these two pigments and the degradation product formed in this layer could suggest the presence of original materials employed when the Terrace ceiling was created by Vasari and his helps. The blue grains in the upper pink and beige layers were identified as ultramarine blue, most likely artificial, due to the presence of other typical restoration materials in the same layers. Finally, the black pigment on sample C12 was characterised as carbon black, present on the paint layer together with cinnabar, hematite and anglesite (975 cm−1 ), a lead sulphate usually formed from alteration (whitening) of lead-based pigments, mostly red lead [15]. The causes of the formation of this product are unclear, especially since red lead was not detected on the paint layer and should be further investigated in the future. Moreover, some questions about the presence of complex stratigraphyes are still not answered. SEM-EDS analysis is commendable in future.

4 Conclusions In this research study the multi-disciplinary approach allowed to identify several materials, such as azurite, cinnabar, white lead, malachite, and red and yellow earth pigments how possible original. Linseed oil was likely used as binding medium in the painting layer, while the composition of the organic component in the ground layer could not be assessed, probably due to complete degradation of the materials used, and to the limitations of the techniques employed. An imprimitura layer constituted of lead white mixed with gypsum was found. The original wood from one of the round angular panels was identified as probable poplar (populus sp.), compatible with historical records [2]. Restoration materials could be identified as phthalocyanine blue mixed with black manganese oxide, Prussian blue and ultramarine blue (most likely of synthetic origin) for blue/dark blue areas, red and yellow ochre for red and orange areas and yellow decorations, lead white for white and lighter paint layers, baryte and calcite present as fillers/extenders in the pigments. Identification of varnishes and consolidant products was achieved too: natural (beeswax) and synthetic waxes not better identified were found on some areas, while acrylic resins such as Primal AC-33 and Paraloid B-72 were applied in other areas, with and without facings. Probably these two resins were used in more recent conservation interventions, while waxes were employed during older treatments, also considering their evident degradation state (cracking and yellowing phenomena were observed). The artworks seem to be significantly affected by salt attacks (zinc, calcium and other oxalates were observed, as well as soluble salts like gypsum and epsomite), caused by the direct outdoor exposition. The thermo-hygrometric variations affect the painted surface of the ceiling, in particular the strong and rapid humidity variation in summer can be a dangerous source of decay, indeed the presence of hygroscopic salts and compounds can activate decohesion phenomena due to their volume changes. About the wood structure supporting the ceiling the differences in RH and T between attic environment and the outdoor during the winter can determine strains on the wood panels

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

385

which can cause μm-movements on the structure that can induce cracks on the painted surface. No problems of stability of the wood ceiling have been recognized thanks to the steel reinforcement realized in the attic side, but the aforementioned movements could bring to the detachment of small portions of the preparatory layer and subsequently the fall of the painting layer. These preliminary scientific results, testify how a restoration campaign is necessary for the Terrace of Saturn in Palazzo Vecchio, Florence. Acknowledgements. This work has been carried out under partial financial support of project Smart4CHˆ2 (POR FSE 2014–2020, Tuscany Region). Authors would like to thank the Municipality of Firenze for provide the logistic support during the diagnostic campaign; Prof. Verdiani (University of Florence) for providing point cloud and Dr. Adriana Iuliano (University of Bologna) for her contribution in analysis, in particular on Raman spectroscopy.

Appendices

Table 1. Samples description and XRF results Sample Description

Major elements Minor elements

Trace elements

C1

Dark colour

Pb

Fe, Ni

Ca, Cr, Cu, Zn, Sr

C2*

Pinkish colour

Pb

Ca, Fe, Cu

Ni

C3

Green colour

Fe, Zn, Ca

Cr, S

K, Ti, Ni, Pb, Sr

C4*

Black colour

Cu

Pb, Fe, Ca, Ba

S, K, Mn, Ni, As, Sr

C5*

Green colour

Fe, Ni

Ca, Cr

Ti, Cu, Zn, Au, Sr

C6

Green colour

Zn

Fe, Ca

S, Ba, Cr, Pb, Sr

C7*

Gilding

Fe

Au, Ca

S, K, Zn, Sr

C8*

Gilding

Fe

Ni, Au, Ca

Cu, Zn, Sr

C9

Wax/glue

Pb

K, Ca, Fe, Co, Ni, As S, Cu, Zn, Bi, Sr

C10*

Red colour and blue Pb colour layers

C11

Wood

C12*

Red colour

C13*

Ni, Fe

Ca, Co, Cr, Cu, Hg

Ca, Hg, Pb, Fe

Zn, Sr

K, Mn, Ni, Cu

Orange colour

Fe, Pb

Ca

K, Ni, Cu, Sr

C14*

Blue colour

Ni, Fe, Mn, Ca

Ba

Cu, Zn, Pb, Sr

C15

Blue colour with efflorescence

Ni, Fe, Ca

Cu, Zn

Cr, Mn, Sr

C16*

Green colour, canvas

Fe, Ca

Cr, Zn, Ni

Mn, Cu, Pb, Sr (continued)

386

S. Longo et al. Table 1. (continued)

Sample Description

Major elements Minor elements

Trace elements

C17*

Dark colour

Pb

Ca, Ba, Cu, Sr

C18*

Dark red colour

Zn

Fe, Ni, Ca

Ti, Mn, Cu, Pb, Sr

C19*

Dark blue colour

Mn, Ca, Zn

Fe, Ni, Co, Sr

Ti, Cu, Pb

C20*

Orange colour

Ca, Fe, Zn, S

Sr

Pb

Ca, Fe, Zn

S, Pb, Sr

Ba

C21*

Burnt part

Pb

Fe, Ni, Co, Ca

K, Bi, As, Sr

C22

Wood

C23

Ground layer

C24

Glue/varnish with facing

Fe, Ni, Zn

* Cross-section samples

Table 2. XRD, FTIR and Raman results of samples Sample

Description

Location

C1

Dark colour

Pictorial layer

Ground layer

C2

XRD

Raman

Linseed oil, sulphate (probably potassium sulphate) CaSO4, CaSO42H2O, PbCO3, K2Pb (SO4)2

Pinkish colour Pictorial layer Ground layer

FTIR

Anhydrite, natural wax, proteic substance Gypsum, synthetic wax, calcium oxalate

CaSO4, CaSO4.2H2O, CaCO3, SiO2, PbCO3

Gypsum, traces of oxalate, traces of calcium carbonate, cellulose (wood) (continued)

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

387

Table 2. (continued) Sample

Description

Location

XRD

FTIR

C3

Green colour

Pictorial layer

nylon-11

Synthetic polymer compatible with PMMA

Ground layer

CaSO4.2H2O, SiO2

Gypsum, lipidic substance compatible with linseed oil and/or egg, traces of calcium oxalate and calcium carbonate

Pictorial layer

CuO, Cu2CO3(OH)2

Calcium oleate, Prussian blue, calcium Malachite oxalate, silicates, possible calcium phosphate,

Ground layer

CaSO4.2H2O, CaSO4, CaCO3, SiO2

Gypsum, calcium carbonate

CaSO4.2H2O, SiO2

Gypsum

C4

C5

Black colour

Green colour

Green colour

Gypsum, calcite, white lead, hematite

Pictorial layer Ground layer

C6

Raman

Pictorial layer

Ground layer

Organic substance compatible with oil, zinc oxalate CaSO4.2H2O, CaCO3, SiO2

Gypsum, calcium carbonate (continued)

388

S. Longo et al. Table 2. (continued)

Sample

Description

Location

XRD

FTIR

C7

Gilding

Pictorial layer

Au

Red bole with silicates, organic substance compatible with synthetic wax

Ground layer

CaSO4.2H2O, SiO2

Gypsum, calcium oxalate

C8

C9

Gilding

Wax/glue

Pictorial layer

Linseed oil

Ground layer

Gypsum

Pictorial layer

n-paraffin

Raman

Beeswax

Ground layer C10

C11

Red colour and blue colour layers

Pictorial layer

Lead white, linseed oil

Ground layer

Gypsum, calcium oxalate

Wood

Pictorial layer Ground layer

C12

Red colour

Cellulose, gypsum, synthetic polymer compatible with PMMA

Pictorial layer

HgS

Ground layer

CaSO4.2H2O, CaCO3, CaSO4

Cinnabar, carbon black, hematite, anglesite Gypsum, calcite (continued)

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

389

Table 2. (continued) Sample

Description

Location

XRD

C14

Blue colour

Pictorial layer

BaSO4, MgSO4

Ground layer

CaSO4.2H2O

gypsum

CaSO4.2H2O

Gypsum

Ground layer

CaSO4.2H2O, SiO2 (trace)

Cellulosic fibre compatible with cotton

Pictorial layer

Pb3(CO3)2(OH)2, PbO2, BaSO4

Ground layer

CaSO4.2H2O

C15

C16

C17

C18

C19

Blue colour with efflorescence

Pictorial layer

Green colour, canvas

Pictorial layer

Dark colour

Dark red colour

Dark blue colour

Ground layer

Pictorial layer

Ground layer

Gypsum CaSO4.2H2O, CaCO3, SiO2

Pictorial layer

Pyrolusite MnO2

FTIR

Raman Phthalocyanine blue, baryte Gypsum, quartz

a Multi-layered

sample [1]: Gypsum, anhydrite, calcium carbonate, organic substance compatible with natural wax

Gypsum

Acrylic resin compatible with methyl methacrylate

Red ochre

Gypsum, Anatase linseed oil, calcium oxalate Phtalocyanine blue, baryte, pyrolusite, carbon black (continued)

390

S. Longo et al. Table 2. (continued)

Sample

C20

Description

Location

XRD

FTIR

Raman

Ground layer

CaSO4.2H2O, CaCO3, SiO2

Gypsum, calcium oxalate, calcium carbonate

Gypsum, calcite, amorphous goethite

Orange colour Pictorial layer Ground layer

C21

Burnt part

Goethite and hematite CaSO4.2H2O

Pictorial layer

Gypsum, zinc oxalate

Gypsum

Natural wax mixed with organic substance compatible with oil

Ground layer C22

Wood

Pictorial layer Ground layer

CaSO4.2H2O, SiO2

Cellulose

CaSO4 .2H2 O

Gypsum, beeswax, white lead

C23

Ground layer

Ground layer

24

Glue/varnish with facing

Pictorial layer

Cellulose (paper), beeswax

Ground layer a 1) inner blue: plattnerite, azurite, cinnabar; 2) pink layer: artificial ultramarine blue, hematite,

lead white; 3) green layer Prussian blue, amorphous goethite, ultramarine blue.

References 1. Allegri, F.: Palazzo Vecchio e i Medici: guida storica di Ettore Allegri, Alessandro Cecchi. S.P.E.S, Firenze (1980) 2. Lensi, A.: Palazzo Vecchio. Casa editrice d’arte Bestetti e Tumminelli, Milano (1929) 3. IAWA List of Microscopic Features for Hardwood Identification, IAWA Bull N. S., vol. 10, pp. 219–232 (1989)

The Terrace of Saturn in Palazzo Vecchio, Florence (Italy)

391

4. Feller, R.L.: Artist’s Pigments: A Handbook of Their History and Characteristics, vol. 1. Cambridge University Press, Cambridge (1986) 5. Zehnder, K., Schoch, O.: Efflorescence of mirabilite, epsomite and gypsum traced by automated monitoring on-site. J. Cult. Herit. 10(3), 319–330 (2009) 6. Simonsen, K.P., et al.: Formation of zinc oxalate from zinc white in various oil binding media: the influence of atmospheric carbon dioxide by reaction with 13 CO2. Herit. Sci. 8(1), 1–11 (2020) 7. Platania, E., et al.: Investigation of mineralization products of lead soaps in a late medieval panel painting. Spectrochim. Acta. A. Mol. Biomol. Spectrosc. 228, 117844 (2020) 8. Vasari, G., Maclehose, L.S., Brown, G.B.: Vasari on technique: being the introduction to the three arts of design, architecture, sculpture and painting, prefixed to the lives of the most excellent painters, sculptors and architects. Unabridged and Unaltered Republication, New York (1960) 9. Spring, M., Higgitt, C., Saunders, D.: Investigation of pigment-medium interaction processes in oil paint containing degraded smalt. Natl. Gallery Tech. Bull. 26, 56–70 (2005) 10. Cennini, C.: Il libro dell’arte. Neri Pozza Ed, Vicenza (1971) 11. Phthalocyanine blue - CAMEO. http://cameo.mfa.org/wiki/Phthalocyanine_blue. Accessed 06 Dec 2021 12. Bernardini, F., et al.: Raman spectra of natural manganese oxides. J. Raman Spectrosc. 50(6), 873–888 (2019) 13. Rinaudo, C., Croce, A.: Micro-raman spectroscopy, a powerful technique allowing sure identification and complete characterization of asbestiform minerals. Appl. Sci. 9(15), 3092 (2019) 14. Burgio, L., Clark, R.J., Firth, S.: Raman spectroscopy as a means for the identification of plattnerite (PbO2), of lead pigments and of their degradation products. Analyst 126(2), 222– 227 (2001) 15. Costantini, I., et al.: Darkening of lead-and iron-based pigments on late Gothic Italian wall paintings: energy dispersive X-ray fluorescence, μ-Raman, and powder X-ray diffraction analyses for diagnosis: Presence of β-PbO2 (plattnerite) and α-PbO2 (scrutinyite). J. Raman Spectrosc. 51(4), 680–692 (2020)

Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage Artworks Emanuela Proietti1(B) , Giovanni Capoccia1 , Romolo Marcelli1 Giovanni Maria Sardi1 , and Barbara Caponera1,2

,

1 Consiglio Nazionale delle Ricerche - Istituto per la Microelettronica E Microsistemi, Via

Fosso del Cavaliere 100, 00133 Rome, Italy [email protected] 2 Istituto VA-VE, Piazza Trento 5, 00019 Tivoli, Rome, Italy

Abstract. A prototype portable tomographic system for microwave imaging is presented in this paper as non-destructive testing technique for the full-volume inspection of a work of art, as it can give morphological and physical information on the inner structure of the investigated sample. A custom-made 3D-printed structure has been built to support the antenna elements of the systems and enclose the target being tested. Transmission measurements between pairs of antennas have been acquired through a Vector Network Analyzer connected to a modular switching matrix. Finally, data have been inverted by an open-source reconstruction algorithm to generate the image. Keywords: Microwave tomography · Antenna · Cultural heritage imaging

1 Introduction Many disciplines from ever investigate the possibility to access the internal properties of materials. Biology, material science, and sometimes medicine and palaeontology can slice specimens and study materials’ internal structure, but this is impossible for ancient cultural heritage artifacts. Investigation of cultural heritage is a multidisciplinary work that needs the contribution of different research fields. The archaeological considerations are essential to classify the aesthetic style and identify the historical and geographical context in which an art object can be framed. Restorers, on the other side need some information on materials before starting the conservation activities. They often need to know what their eyes can’t see, what is inside an object that the engineers call inclusions. X-ray imaging is routinely used to analyze museum artifacts providing contrasted images depending on the radiodensity of the materials. Similarly, neutron imaging techniques are available as efficient non-destructive testing tools in cultural heritage science. In this work, microwave radiation has been applied as a possible powerful candidate in art conservation to investigate various art-related materials. Microwave tomography is a noninvasive technique that can reveal the internal structure of the target by analyzing the microwave signal absorption in the material. In the study of an ancient artifact, microwave tomography can provide information about the © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 392–400, 2023. https://doi.org/10.1007/978-3-031-17594-7_29

Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage

393

work of art composition and density without damage or sampling. Furthermore, it can reveal invisible defects, cracks, or damages, showing if there are any inclusions or understanding if the object is a single piece or is made of different parts joined together. Microwave imaging is based on the ability of electromagnetic waves to penetrate nonmetallic media and retain information about the internal dielectric and conductivity properties of the object under test. It represents a potentially powerful noninvasive diagnostic tool that finds application in countless scenarios [1, 2]. In addition, it offers several advantages as compared to other diagnostic methodologies since it relies on non-ionizing radiation and cheap and portable hardware systems. Therefore, during the last decades, microwave imaging has triggered many research activities, which focused on the hardware and signal processing requirements. In cultural heritage artworks, it is crucial to detect the presence of pieces of metal or different stones inside the handwork to address restoration operation. Nails and metals in general and insulators and stone slivers are examples of typical inclusions, whose dimensions are generally almost a few cubic centimeters. Ligneous artifacts often have voids inside or rotten areas and this information is very important in the restoration process [3]. For this specific application, targets are made up by inhomogeneous material and, as electromagnetic media, they are lossy and dispersive. The target attenuates the propagating signal, and complex data on the physical properties of the material are stored in the scattered field. The unknown physical properties of the material composing the target are often related to high permittivity or conductivity. For this reason, to retrieve meaningful data from the reconstruction algorithm, system setup needs to be carefully designed. Multiple reflections can generate artifacts on the sample/air/interface of the medium where the electromagnetic wave is travelling, leading attenuation into the sample (mismatch at the sample interface). In literature are present studies using both matching layer or antennas in contact with the samples to overcome these effects [4]. The prototype presented in this paper consists of a custom-made 3D printed versatile imaging system holding an array of lightweight sensing antenna elements. A compact transceiver has been used to transmit and receive wideband signals through the antennas, and a switching network controls the data acquisition from the antenna array. Data storing is performed by a processing unit. The work presented has been developed in collaboration with the Autonomous Institute of Villa Adriana and Villa d’Este (VA-VE), UNESCO World Heritage Site since 1999 in Tivoli, Italy, shown in Fig. 1. Villa Adriana was the imperial residence of the Roman Emperor Adriano. It was built between 127 and 138 AD and extended over 120 hectares. The variety and quantity of material in this site allowed us to set the prototype, and we present in this paper some preliminary investigation on archeological finds.

394

E. Proietti et al.

Fig. 1. Villa adriana “maritime theater”, Tivoli, Rome, ITALY

2 Microwave Tomography System The core of the microwave tomography system is a multi-antenna architecture. The antennas have been designed based on a self-complementary wideband monopole with a diameter of 30 mm by CST Microwave Studio and their layout by Altium Designer 20. Details of design and fabrication are available on [5]. 2.1 Simulations A measurement setup in microwave tomography typically includes a series of receiving (Rx) and transmitting (Tx) antennas to measure the scattered fields needed for the reconstruction algorithms. The antennas can be arranged on planar, cylindrical, or hemispherical surfaces. We have chosen a cylindrical setup with the antennas placed on linear slots along the circumference. In this way, antennas can translate to be into direct contact with the sample under test and accurately sample the scattered fields. The target is a hollow cylinder (10 cm diameter, 5 mm thick) filled by the library material (marble, dry soil, wood, etc..). We have placed random metal inclusions inside this homogenous host material as the experiment’s goals are to detect metals inside the artifacts. The antennas are placed equidistant around the cylinder at the same quote. In the simulation process with N antennas, the path loss (Sxy ) and reflections (Sxx ) are collected with a single transmitting antenna x and the remaining receiving antennas y (with y = 1 to N−1), collecting N reflection signals, with x = 1…N, and (N−1) + (N−2) +…. + 1 transmitting signals. The more the number of antennas, the more the simulation time and the more the data collected. In Fig. 2, N = 8 antennas have been used for the simulation. In Fig. 3 the electromagnetic response of the system (S21 -transmission coefficient) shows how the inclusions are detected in a bandwidth of almost 1 GHz [5].

Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage

395

Fig. 2. 3D model of the simulated configuration: the host material has been hidden to highlight the inclusions’ locations.

Fig. 3. Simulated transmission response for the system using 8 antennas (S21 , S42 , S73 ): continuous line is the response of a homogenous sample of dry soil, while the dotted line corresponds to three metallic inclusions.

2.2 Laboratory Experimental Measurements The first test has been done on a 8 antennas simplified setup placed on a specific ad hoc support structure. To validate the simulations, the considered target is a becker filled with dry soil with three metallic inclusions inside The antennas are placed equidistant around the cylinder at the same quote as in Fig. 4a. Transmission parameters S21 have been recorded by a Keysight P9573 Vector Network Analyzer (Fig. 4b) and they are shown in Fig. 5. 2.3 Portable System Setup In the microwave imaging system, each antenna acts as a transmitter and a receiver, in order to have multi-static multiview measurements of the sample. Hence, all the antennas are connected to the VNA through a switching matrix. All the switches are connected to a control board to get the appropriate connections between the antennas and the VNA.

396

E. Proietti et al.

Fig. 4. Experimental antenna arrangement with N = 8 antennas (a), Keysight P9573 vector network analyzer (b).

Fig. 5. Experimental measurements of the transmission response for the system using 8 antennas (S21, S42, S73): continuous line is the response of a homogenous sample of dry soil while the dotted line corresponds to three metallic inclusions.

This approach has been followed to develop a portable prototype to enable a software integration of the measurement data and the imaging algorithm. In this way, complete automated scans of the target can be run, providing the reconstructed images almost real time. A description of the system is shown in Fig. 6. The 16 antenna array is connected to the Vector Network Analyzer, by two levels Radio-Frequencies switch matrix. A notebook PC hosts the switching logic program, the acquisition software, and the imaging algorithm.The switch matrix comprises two layers, a Switch Control Board level and 4 RF Switch Bank level as shown in Fig. 7a. Each RF switch bank 4:2 is connected to 4 antennas and sends the control board’s TX/RX signal of the pair antennas selected

Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage

397

Fig. 6. Block diagram of the portable setup

to the PC. The Switch Control Board is directly connected to the transmit and receive ports of the VNA and sets the RF Switch Bank by 2 RF switch 4:1 (Fig. 7a). RF switches Peregrine Semiconductor’s PE42441 have been used in the switchboards. The working frequency band of such components ranges from 10 MHz up to 8 GHz.

Fig. 7. RF switch 4:2 bank and control board (a) and antennas hosted in a 3D printed holder (b)

The antennas pair are sequentially connected to the transmitting and receiving port of Keysight P9375A compact VNA for measuring the field scattered by the target, as in Fig. 8. Each inactive antenna is switched to a 50  load ground-connected to minimize the interferences in the measurement. The switch control board is equipped with an ATmega328 microcontroller programmed to transparently pass the USB data to an I2C bus connected to four 16-bit MCP23017 input/output (I/O) expanders used to drive the RF switches. We have used SPXT (single-pole X-throw) switches. The nomenclature means Single Pole (pole is the the independent signal), X throw (number of different signal paths for each pole), one signal from or towards X = 2 or X = 4 different ports (throws) [6, 7]. The insertion losses of SP2T-SP4T are in Table 1 [6]. The total insertion loss measured (in dB) of the critical signal path L for a given switching architecture can be computed as (1), where S Ti is the number of the switch in

398

E. Proietti et al.

Fig. 8. Setup composed by a vector network analyzer (VNA), a radio frequency (RF) switch matrix connected to a set of 16 antennas. A notebook PC manages the system by selecting the active antenna pair connected to the transmitting and receiving port of keysight P9375 compact VNA to measure the field scattered by the target.

Table 1. Basic SPXT switches Switch type

Model [6]

Insertion losses

SP2T

PE42422

0.25 dB

SP3T

PE42430

0.45 dB

SP4T

PE42441

0.45 dB

the Ti throw, N is the number of different basic switches and L Ti denotes the insertion loss in dB introduced by a basic switch with Ti throws. Therefore, an accurate characterization L=

N 

STi · LTi

(1)

i=1

of the maximum number of throws per switching stage is crucial since they determine the overall insertion loss of the critical signal path, i.e. the signal path with the most significant power losses. The electronic boards have been designed to keep as low as possible the acquisition time using solid-state electronic components. The software can acquire 192240 measures in approximately 2 min (S = 16; F = 801), i.e., it requires about 0.6 ms for each batch of 801 frequency samples. The total insertion loss measured in dB of the critical signal path L according to (1) is L = 2 * 0.45 + 1 * 0.25 = 1.15 dB.

Microwave Imaging Applied to Noninvasive Diagnostic of Cultural Heritage

399

2.4 Imaging To reconstruct the sample image and the inclusions inside it, the data collected by the system have been processed through an inversion algorithm. The open-source MATLAB framework MERIT Microwave Radar-based Imaging Toolbox has been used [8]. MERIT contains a complete package for microwave radar imaging, including the signal processing and parameter search algorithms for dielectric properties. For imaging, two sets of measured data or scans are required, one taken at a fixed rotational offset from the other to remove the same artifacts and noise from the image. This offset is the distance between two consecutive antennas in our case. The image’s resolution can be improved by maximum elimination of the noise from the signals and by keeping the minimum distance between the focal points in the imaging domain. Each set of measured data is represented by F × C array, where F is the length of frequency point in each scan, and C is the number of Channels, i.e. the Transmission Coefficient obtained by the combination of N Antennas (N−1) + (N−2) +…. + 1. It also requires data consisting of all the frequency points used for scan files, the location of the antennas, and the combination of channels. The instruments management and acquisition software have been developed using Microsoft Visual Studio 2019 with Vb.NET language.

3 2D Imaging of Sealed Pottery The portable setup has been transferred to Villa Adriana laboratory (Fig. 9a) to perform in situ two-dimensional (2D) imaging of actual artworks, such as pieces of pottery, jars, amulets, etc. An ancient roman sealed pottery there has been measured to demonstrate the potential of microwave imaging for detecting hidden materials in historical museum pieces. The operative setup of the imaging system is shown in Fig. 9a, on ancient sealed pottery of Istituto VA-VE.

Fig. 9. Operative setup of the imaging system (a). Image of the inclusion detected inside an ancient work of art obtained by MERIT software [8]. Qualitative indicator map of the inclusions (b).

S21 parameters of each active antenna pair have been acquired in frequency range between 1 and 8 GHz with 801 frequency steps. The measured data have been processed

400

E. Proietti et al.

using the MERIT inversion algorithm to obtain a 2D reconstruction of the sample and its inclusions as in Fig. 9b. The prototype has been successfully tested, clearly revealing the presence of two types of inclusions scattering different signal levels. After removing the soil inside, the following restoration shows the sealed pottery has a metal inclusion and a stone inside. This result is particularly encouraging because of the detection of hidden in the field of art conservation for the investigation of various art-related materials.

4 Conclusion A microwave system prototype for tomographic imaging, with the final goal to investigate cultural heritage artifacts and possible inclusions, has been presented in this paper. The imaging setup has been tested by measuring a soil sealed pottery. Future developments will aim to integrate the system with a z-axis movement and perform more comprehensive experimental measurements, including a detailed validation of the local measurement setup and volumetric characterization with metallic and dielectric inclusions. Acknowledgement. This research was supported by COOLTURE Project funded by italian Regione Lazio POR FESR LAZIO 2014–2020 AVVISO PUBBLICO “GRUPPI DI RICERCA 2020”.

References 1. Pastorino, F.: Microwave Imaging: Methods and Applications, pp. 1–310. MA:Artech House, Boston (2018). ISBN: 9781630813482 2. Bisio, I.: Variable-exponent Lebesgue-space inversion for brain stroke microwave imaging. IEEE Trans. Microw. Theory Tech. 68, 1882–1895 (2019). https://doi.org/10.1109/TMTT. 2019.2963870. ISSN: 0018–9480 3. Boero, F.: Microwave tomography for the inspection of wood materials: imaging system and experimental results. IEEE Trans. Microw. Theory Tech. https://doi.org/10.1109/TMTT.2018. 2804905 4. Lanini, M.: Design of a slim wideband-antenna to overcome the strong reflection of the air-tosample interface in microwave imaging. In: Proceedings of the 2015 International Conference on Electromagnetics in Advanced Applications, Turin, Italy, pp. 1020–1023 (2015) 5. Proietti, E.: A tomograph prototype for microwave imaging applied to cultural heritage: Preliminary experimental results. In: AIP Conference Proceedings, vol. 2416 (2021). https://doi. org/10.1063/5.0068509 6. Peregrine semiconductor, Portfolio of RF switches. http://www.psemi.com/products/rf-swi tches 7. Skyworks Solutions, Portfolio of RF switches. http://www.skyworksinc.com/products switches.aspx 8. Radar-based Imaging Toolbox available on GitHub. https://github.com/EMFMed/MERIT

SWIR Reflectance Imaging Spectroscopy and Raman Spectroscopy Applied to the Investigation of Amber Heritage Objects: Case Study on the Amber Altar of the Lord’s Passion Paulina Krupska-Wolas1,2(B) , Anna Rygula1 , El˙zbieta Kura´s3 , and Julio del Hoyo-Mel´endez1 1

Laboratory of Analysis and Nondestructive Investigation of Heritage Objects, National Museum in Krak´ ow, Krak´ ow, Poland [email protected],[email protected] 2 Faculty of Physics and Applied Computer Science, AGH University of Science and Technology, Krak´ ow, Poland 3 Decorative Arts Conservation Studio, National Museum in Krak´ ow, Krak´ ow, Poland

Abstract. The chosen object from the National Museum in Krak´ ow collection (the Amber Altar of the Lord’s Passion) was examined by Raman spectroscopy (RS) and reflectance imaging spectroscopy (RIS) in short wavelength infrared (SWIR) spectral range. These studies allowed to identify the type of amber used and to assess the existence of a degraded layer. Based on the analysis of the reflection spectra and using the NMF (Nonnegative Matrix Factorization) algorithm, the amber of the Altar were assigned to two different classes. It may be associated with the use of different technologies for the preparation of amber plates. Moreover, as a result of the PCA (Principal Component Analysis) dimension reduction algorithm, an initial classification of the object’s amber into succinite (both native and German) and/or gedan-succinite and/or glessite was performed. Analysis of Raman spectra obtained by deep profiling method allowed to state the existence of a degraded layer on the surface of the amber figures of Saints Peter and Paul. The change in the spectrum profile depending on the depth of the signal collection indicates, among others, changes in the structure related to the exomethylene group and changes in the CH2 and CH3 groups. Thanks to the analysis of reflectance imaging spectroscopy and Raman spectroscopy, it was possible to identify the places of occurrence of amber imitations. The bands distinguished in the Raman spectra indicate the use of an epoxy resin. Keywords: Amber · Heritage objects · Reflectance imaging spectroscopy · SWIR · Raman spectroscopy · Degradation · Amber imitation c The Author(s), under exclusive license to Springer Nature Switzerland AG 2023  R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 401–416, 2023. https://doi.org/10.1007/978-3-031-17594-7_30

402

1

P. Krupska-Wolas et al.

Introduction

Amber [21,27] is a fossil resin of conifers, less often deciduous trees. The most popular type of amber is Baltic amber, also known as succinite. It can be said that amber is, in a way, a chronicle of processes taking place during its solidification. This is reflected in the differences in the chemical structure of amber, which is successfully studied mainly by Fourier transform infrared spectroscopy (FTIR) and also Raman spectroscopy (RS) [4,22]. X-Ray fluorescence and scanning electron microscopy with EDS [5] are also commonly used to describe and identify amber type. These methods, as well as analytical photography techniques, e.g. UV light-induced fluorescence imaging, are employed to identify amber imitations [6,7,12,14,24,26]. It is also worth to mention that amber is called a living mineral because it undergoes various chemical processes constantly, changing its properties and appearance. Some of these processes involve degradation [3,16,28], which can occur as a result of oxidation with oxygen from the air and as a result of the light impact. The issue of identifying the type of amber used to decorate a historic object may determine the provenance of the work itself. What is also important - determining the degree of degradation helps in planning the necessary conservation works and can help in determining the most optimal strategy for the protection of the work. These are the main reasons for researching amber objects. The use of noninvasive and nondestructive methods, such as Raman spectroscopy or reflectance imaging spectroscopy (RIS) has additional advantages, as it does not violate the integrity of the valuable work. To the best of the authors’ knowledge, apart from analytical photography, no other imaging methods have been commonly used to study amber, and in particular to study historic amber objects. The aim of this work is to explore the use of hyperspectral imaging techniques and the evaluation of the data in a broader context, when used in conjunction with analytical photography. It is believed that the reflectance imaging spectroscopy technique could prove to be a valuable tool in the scientific development of amber craftsmanship.

2 2.1

Experimental Materials

The Amber Altar of the Lord’s Passion from the National Museum in Krak´ ow collection of European Decorative Arts was selected as a case study. The Altar is attributed to the authorship of Michael Schodelock. It is dated to the middle of the 17th century. The structure and the base of the Altar are wooden, the front and sides are lined with amber tiles of various shades and transparency, as well as plaques with figural scenes and ornaments carved in ivory (Fig. 1). A set of reference amber samples from the collection of the Museum of the Earth of the Polish Academy of Sciences in Warsaw was also examined with reflectance imaging spectroscopy technique. The collection counts in total 63

SWIR Reflectance Imaging Spectroscopy

403

samples. Table 1 presents a list of selected reference amber (and copals). All 63 reference spectra were taken for RIS analysis, but for clarification purposes only some of them (listed in Table 1) are shown further in this study.

Table 1. List of selected amber and copals references Amber type

Sample description

Succinite

Sambia (today Kaliningrad) Bitterfeld from Goitshe mine auccinite, white amber Saxon white Ukraine Adam´ ow Poland Kazimierz Poland Jantary Poland

Gedanite

Sample id. 20710

Gedan-succinite Sample id. 2222 Ukraine Glessite

sample id. 21268 Striped

Crancite

K¨ onigssee Profen

Copals

Philippines Dominican Republic Borneo

Others

Valchovite Goitshite Lebanese

Fig. 1. The visible light photo of the examined object. The numbers represent the places of point measurements by RS.

2.2

Techniques

Reflectance Imaging Spectroscopy (RIS). The measurement was carried out using an imaging system with short wavelength infrared reflectance spectroscopy (SWIR 1000–2500 nm, 256 spectral channels, spectral resolution 6.32 nm; Specim, Oulu FIN). The system is equipped with six point halogen light sources arranged symmetrically on both sides of the camera at an angle of 45◦ [11]. Figures of Saints Peter and Paul were measured independently. The cross was not measured. The data obtained as a result of more than one measurement are merged using the homographic function [15]. The scanning system was positioned in vertical geometry. The typical scan parameters for the measurement cycle were used: exposure time 7.8 ms, repetition frequency 10 Hz, camera

404

P. Krupska-Wolas et al.

movement speed 5.5 mm/s, which allowed to obtain the appropriate measurement geometry in relation to the set distance of the camera from the object and a satisfactory image quality. The obtained hyperspectral images were normalized to the white (Spectralon white plate, Labsphere, New Hampshire, US) and reduced with dark current. Reference and dark current measurements were taken under the same parameters as for the actual object’s measurements. Data analysis was perform with the Python scripting language (Python Software Foundation. Python Language Reference, version 3.7.1) and with the ENVI software (Harris Corporation, Melbourne, Florida, US).

Fig. 2. The process of RIS data analysis.

SWIR Reflectance Imaging Spectroscopy

405

The input data were processed to reduce scattering effects and highlight important spectral structures. Moreover, the analysis was limited to the spectral range 1350–1800 nm, which turned out to be specific enough to classify the amber against the reference base. The data presented below were subjected only to the algorithms which gave the best results in terms of classification quality. Firstly, the authenticity of the amber was checked. The details can be found in the Amber identification section. In the next step, the data were subjected to the parallel operation of two algorithms: Nonnegative Matrix Factorization (NMF) and Pixel Purity Index (PPI, Fig. 2). The first one allows to visualize the differences in the spectral structure of the material constituting the object. The PPI algorithm indicates areas of spectrally pure pixels of the analyzed hyperspectral image. The pixels can be grouped into clusters based on the shape of their spectrum. The authors used the PPI algorithm to obtain set of representative object spectra needed to perform further analysis. This is a very desirable operation due to the huge amount of pixels in the original image (1200 × 772), which gives almost a million spectra to analyse. Further, the calculated representative spectra were grouped into classes obtained by NMF analysis. The classification of Altar’s amber in relation to reference amber samples was performed by Principal Component Analysis (PCA). A flowchart presenting steps of the analysis of hyperspectral data can be found in Fig. 2. Raman Spectroscopy (RS). Raman spectra were collected with a portable Raman spectrometer (DeltaNu, Wyoming, USA) equipped with a thermoelectrically cooled charge-coupled device detector working in the range 200–2000 cm−1 . Spectra were acquired using a 785 nm diode laser with a maximum output power between 24–50 mW. Spectra were recorded using an 8 cm−1 resolution. The applied acquisition time was 1 s with 10 accumulated spectra for every point. The automatically background correction was applied. The measurements were carried out in a distance about 1 cm which is the value for the collection of the most intensive Raman signal. An attempt was made to register the Raman depth profiling for figures of Saints Peter and Paul to try to collect spectral information about amber surface degradation. The scans were collected from a distance of the most intensive Raman signal (called middle distance). Then, the spectra were registered from the distance about 0.5 cm (near ) and about 1.5 cm (far ).

3 3.1

Results Amber Identification

A. Golloch in his work [10] shows a set of typical amber reflectance spectra and presents a method of identifying amber imitation by determining two coefficients I1,2 for each of the analyzed spectra:

406

P. Krupska-Wolas et al.

I1,2 =

x1,2 − a · 100, b−a

(1)

where: a–reflectance intensity for 1200 nm, b–reflectance intensity for 1315 nm, x1,2 –reflectance intensity for 1415 and 1555 nm, respectively. Based on the above (1), the values of the coefficients for the reference amber collection were determined. All of the reference amber reflectance spectra are presented on Fig. 3. The factors are respectively: −29 ± 33 for the 1415 nm band and 1 ± 38 for the 1555 nm band. The values of the coefficients for the Altar were determined analogously. gedanites and gedan-succinites

1000

reflectance [a.u.]

reflectance [a.u.]

succinites

1200

1400

1600

1800

2000

2200

2400

1000

1200

1000

1400

1600

1800

2000

2200

2400

2200

2400

reflectance [a.u.]

wavelength [nm] black amber

reflectance [a.u.]

wavelength [nm] glessites and crancites

1200

1400

1600

1800

2000

2200

2400

1000

1200

wavelength [nm]

1400

1600

1800

2000

wavelength [nm]

reflectance [a.u.]

other kinds amber

1000

1200

1400

1600

1800

2000

2200

2400

wavelength [nm]

Fig. 3. The reflectance spectra of all amber references. Based on them, the coefficient factors I1,2 were calculated.

Figure 4 shows the graphic interpretation of coefficients I1,2 for the Amber Altar. The areas for which the coefficient value is within the calculated reference range are marked with color. Black indicates values outside this range. Wood

SWIR Reflectance Imaging Spectroscopy

407

areas and ivory plaques (yellow and white markings in Fig. 4) remain black, as expected. During RIS measurements the statue of the Virgin Mary (purple arrow, Fig. 4) were wrapped with an additional layer of material to eliminate any possible damage. This is the reason for excluding the material of the figure as amber. Therefore, the statute of the Virgin Mary was not further analyzed with RIS.

1800

1600

1400

1200

1000

800

600

Raman shift [cm−1 ]

400

Fig. 5. Raman spectrum collected from point 5. Highlighted bands indicate the presence of epoxy resin.

200

1800

1600

1400

1200

1 2 3 4

976

1450 1304 1205 1143

1645

Raman intensity [a.u.]

662 582 442

826

1439

1289 1245 1106

Raman intensity [a.u.]

1596

Fig. 4. The intensity ratio map for the Amber Altar. The arrow markings can be found in the text.

1000

800

600

Raman shift [cm−1 ]

400

200

Fig. 6. Raman spectra collected from points 1–4. Highlighted bands indicate the presence of amber.

Also some shining areas with high reflection degree are black (blue arrows, Fig. 4).

408

P. Krupska-Wolas et al.

Table 2. Assignments of Raman bands Table 3. Assignments of Raman bands for amber [20]. for epoxy resin [25]. Bands position [cm−1 ]

Approximate assignment

Bands position [cm−1 ]

Approximate assignment

1596

ν(C=C) aromatic

1737

ν(C–O) stretching ester

1439

deformation CH2

1645

ν(C=C) non-conjugated

1289

ν(CC) ring breathing

1630

ν(C=C) conjugated with C=O

1245

ν(CC) ring breathing

1615

ν(C=C) aromatic

1106

the side string vibrations

1450, 1359, 1327, 1315

δ(CH2 ), δ(CH3 )

826

ρ(CH2 )

582



1442

δ(CH2 ), δ(CH3 ) scissors

442



1304, 1298

δ(CH2 ), δ(CH3 ) twisting

1205

δ(CCH), δ(C–O)

1143

ν(CC) ring breathing

1074

ν(COH)

1063

ν(C–C), ν(COH)

1034

ν(C–O)

976

ρ(CH2 ), ρ(CH3 )

928



878

ρ(CH2 )

820



Selected areas (green arrows, Fig. 4) of the Amber Altar also remain black. It is suspected that these are places where the losses in plates are filled with imitations of amber. Raman spectroscopy showed the presence of epoxy resin at the selected measurement point (no. 5, Fig. 1). The spectrum is shown in the Fig. 5 and the band assignments can be found in Table 2. Raman spectra (Fig. 6, Tab. 3) collected from selected places on the Altar confirm the presence of amber in the remaining areas marked with color in Fig. 4. 3.2

Amber Degradation

Contrary to visible light photography (Fig. 1), in the false-color image (Fig. 7a) it was observed that the alternately positioned amber plates differ. This is also confirmed by further analysis. Hyperspectral images were processed with the NMF algorithm [13,18], which allowed to initially assess the material diversity of the examined object (Fig. 7b). Based on the obtained images, it can be assumed that there are two classes of amber (green and red areas in Fig. 7b). The way in which amber’s plates were arranged in the object may indicate a lack of randomness - they are arranged alternately, as noted previously. The set of reflectance representative spectra was determined using the PPI algorithm [2,8,9]. As a result of PPI a cluster map was obtained (Fig. 7c). Based on the PPI cluster map, the average spectra for each specified cluster were calculated. Further, the PPI cluster map image and NMF result image were compared.

SWIR Reflectance Imaging Spectroscopy

409

Fig. 7. (A) False colors image, R = 1551.01 nm, G = 1217.55 nm, B = 1299.43 nm; (B) NMF result based on three chosen loadings obtained; (C) PPI cluster map.

410

P. Krupska-Wolas et al.

The representative PPI spectra have been grouped into the red and green class according to NMF results (as shown in Table 4). Table 4. Assignments of PPI clusters to distinguished by NMF red and green classes. Cluster number

Cluster color Class color according to PPI according to NMF

Description

cluster 1

amber plates on the edges, the first tires

cluster 2

details on figures of Saints Peter and Paul amber throughout the Altar, even plates in the first and second tires

cluster 5 cluster 8

amber on odd plates, the first and second tires

cluster 9

shiny fragments throughout the Altar

cluster 11

fragments of upper tires abrasions on two plates of the first and second tires

cluster 14 cluster 16

left foot of the Altar a plate different from other amber plates under the central ivory plaque

cluster 17 cluster 18

the middle amber plate, the third tires

cluster 20

figures of Saints Peter and Paul

Spectra show the greatest difference in the intensity of the 1415 and 1639 nm bands (Fig. 8). The presence of the 1415 nm band is related to the stretching vibrations of C–H conjugated with O–H water. It is suspected that the 1639 nm band indicates the presence of organic acids, which are characterized by the presence of combination bands and overtones of O–H stretching vibrations, among others in the range of 1400–1650 nm [23]. Both, the 1415 nm and 1639 nm bands are shallower for the NMF green class than for the red one. Moreover, double the maximum absorption can be observed for 1405 and 1440 nm for the red class. It may be related to decrease of water and organic acids (including succinit acid also) content in green class amber.

1350

1400

1639 1440

1404

1415

standardized reflectance [a.u.]

object spectra green class NMF object spectra red class NMF

1450

1500

1550

1600

1650

1700

1750

1800

wavelength [nm]

Fig. 8. Reflectance spectra of clusters for both, red and green NMF amber classes.

It is also known that white amber contains more air bubble inclusions than amber of a different color [1]. The greater amount of air bubbles inside an amber

SWIR Reflectance Imaging Spectroscopy

411

878

1034

1400

1205

1450

1600

St.Paul St.Peter

Raman intensity [a.u.]

1630

stone may cause less amount of structures characteristic for amber itself, i.e. structures which can be described by reflectance spectrum bands detailed above. According to that, it is suggested that originally NMF green class amber was milky or butter color while the NMF red class was rather cogniac or cherry color (as it is now). The difference between green and red class amber was quite visible then. It should be mentioned that the spectrum c14 is the average calculated from area of the plates that have been rubbed across the surface (Fig. 1, Fig. 7). These plates were classified to the red (cogniac) NMF class, abrasions to the green class. It is suggested that the current shape of the c14 spectrum is due to mechanical degradation, not related to the original milky color of amber. It is believed that initially c14 amber was dark (cogniac or cherry). It is also noteworthy that the c17 and c18 represent the mean spectra of the plates clearly different from the others (Fig. 1, Fig. 7). The c5 spectrum is an average spectrum collected from plates throughout the Altar that were originally likely to be milky colored. The assignment of the c9 spectrum to both, the red and green NMF classes, is probably due to the fact that c9 describes the shiny details of an object. The c9 spectrum is not specified enough to be classified in only one NMF class. The remaining spectra (c1, c2, c8, c11, c16, c20) were assigned to the red NMF class, originally cogniac or cherry colored.

1800

1200

1000

800

600

400

200

Raman shift [cm−1 ]

Fig. 9. Raman spectra collected from points 6–11, Saints Paul and Peter. Highlighted bands indicate the presence of amber.

The interesting dependence can be also observed on the obtained Raman spectra (Fig. 9). For the statues of Saints Peter and Paul (NMF class red), good quality spectra were collected. This was not possible for the statues of the Virgin Mary (examined without additional material around) and of Saint John (data not shown) assigned to NMF green class with originally milky or butter color. The differences in water content and an amount of organic acids may cause the different stage of amber degradation.

412

P. Krupska-Wolas et al.

Raman intensity [a.u.]

St.Paul St.Peter

878

1205

1450

1359 1290

1630 1595

In addition, depth profiling by Raman spectroscopy was performed for the figures of Saints Peter and Paul. The change in the band profile in the range of about 1630 and 1450 cm−1 (Fig. 10) proves structural changes in the amber itself, i.e. changes in the structure related to the exomethylene group and changes in the CH2 and CH3 groups [20]. It indicates amber degradation, which is expected to be larger on the surface than in the interior amber.

far middle near

far middle near 1800

1600

1400

1200

1000

800

Raman shift [cm

1

600

400

200

]

Fig. 10. Raman spectroscopy depth profiling for the figures of Saints Peter and Paul.

3.3

Amber Classification

The PCA was performed on 10 representative reflectance spectra of the Amber Altar (obtained by PPI algorithm) and 63 reference amber spectra. As a result of the PCA [17,19] it was possible to successfully classify most of the object amber spectra into the group of succinite reference spectra (Fig. 11). Only a fragment of the PCA plot is presented, which includes all the spectra of the object. Not all of the reference amber spectra are shown, because they are out of range of the presented plot. Some of the object spectra (c1, c9, c11, c16, c20) are located among the reference spectra of Polish Baltic amber (Jantary, Kazimierz, Adam´ ow), and also close to those from Ukraine and Germany (Bitterfeld). The spectra of c5, c14 and c17 are located close to the spectra of white succinite sample and Saxon white. In addition, the c5 spectrum is also located relatively close to the reference gedan-succinite and glessite reference spectra. In contrast, the c2, c8 and c18 spectra remain on the edge of the plot, close to the reference spectra of succinites (Sambia, Saxon and white amber) as well as gedan-succinites and glessites. The division into NMF red and green class was also be taken into account while presenting PCA results. Assignment to the specific NMF class was marked in the PCA plot (Fig. 11). The NMF green amber class is indeed located close to the spectra of the reference white succinites, however, c5 also remains in close

Scores on PC2 (explained variance: 22.99%)

−1.5

−1

−0.5

0

0.5

1

1.5

−1

c2

−0.5

0.5

1

1.5

Scores on PC1 (explained variance: 55.59%)

0

white succinite c17 Saxon white c14

c5

Sambia

2

2.5

Fig. 11. PCA plot for two first components: object amber versus reference amber.

−1.5

c8

c18

object spectra red class NMF object spectra green class NMF object spectra both class NMF copals gedanite gedan-succinite glessite crancite succinite c1 others Bitterfeld Poland Adamów c20 Ukraine c11 c16 c9 Poland Jantary Poland Kazimierz

3

SWIR Reflectance Imaging Spectroscopy 413

414

P. Krupska-Wolas et al.

proximity to glesites and gedano-succinites. Most of the spectra of the NMF red amber class (c1, c11, c16, c20) are surrounded by native succinites. Some of the red class spectra (c8 and c18) are close to succinites (Saxon white, Sambia) and gedan-succinites and glessites. The c2 spectrum describes the shiny parts of the figures of Saints Peter and Paul, therefore its position on the PCA plot may not be reliable enough. Based on the PCA analysis, it can be concluded that amber plates can be assigned to succinites and/or gedan-succinites and/or glessites. It is suspected that part of the original cogniac color amber plates (c1, c11, c16, c20) are native succinites, while others (c8 and c18) may be glessites or gedan-succinites. The spectrum of c14 can probably be assigned to white succinite, while the c5 classification remains unclear–most of the original milky amber plaques of the Altar may be white succinite or glessite or gedan-succinite.

4

Summary

The Amber Altar of the Lord’s Passion was subjected to SWIR RIS and Raman spectroscopy in order to assess the material composition of the object. It was possible to distinguish areas of interference with the original structure of amber plates and to determine that the losses were filled with epoxy resin. The amber plates of the Altar have been divided into two classes, which differ probably in water content. It may indicate that originally some of the amber plates were white and others - dark (cogniac or cherry). White amber contains many inclusions of air bubbles, which cause the amount of structures characteristic for amber is lower. Due to RIS analysis, it is suspected that two colors of amber were used white (milky or butter) and dark (cogniac or cherry). As a result of the degradation process, the differences between the amber colors are no longer visible. The arrangement of amber of both NMF classes in the object shows that the differences within amber plates are not the result of randomness, but rather the intention of the creator. The existence of a degraded amber layer for the figures of Saints Peter and Paul was also found. The representative reflectance spectra of the Amber Altar were classified against the reference spectra. Based on PCA analysis, it is suspected that the amber used may be native (or other from Middle Europe) succinite and/or gedansuccinite and/or glessite. Acknowledgements. The authors would like to thank Barbara L  yd˙zba-Kopczy´ nska and the Museum of the Earth of the Polish Academy of Sciences in Warsaw for providing amber references for research, and Agata Mendys-Frodyma for her contribution to the research and analysis of the RIS data.

SWIR Reflectance Imaging Spectroscopy

415

References 1. Features and properties of Baltic amber. https://ambermuseum.eu/en/cechy-iwlasciwosci-bursztynu-baltyckiego/. Accessed 15 Dec 2021 2. Ahmad, F.: Pixel purity index algorithm and n-dimensional visualization for ETM and image analysis?: a case of district vehari. Glob. J. Hum. Soc. Sci. Art Humanit. 12, 23–32 (2013) 3. Badea, G., et al.: Fourier transform Raman and statistical analysis of thermally altered samples of amber. Appl. Spectrosc. 69, 1457–1463 (2015). https://doi.org/ 10.1366/15-07866 4. Truic˘ a, G.I., et al.: A multi-analytical approach to amber characterisation. Chem. Pap. 68(1), 15–21 (2013). https://doi.org/10.2478/s11696-013-0415-8 5. Beck, C.W.: Spectroscopic investigations of amber. Appl. Spectrosc. Rev. 22(1), 57–110 (1986). https://doi.org/10.1080/05704928608060438 6. Beck, C., Wilbur, E., Meret, S., Kossove, D., Kermani, K.: Infrared spectra of amber and the identification of Baltic amber. Archaeometry 8, 96–109 (2007). https://doi.org/10.1111/j.1475-4754.1965.tb00896.x 7. Botfeldt, K., Kalsbeek, N.: Identification of amber and amber imitations by infrared spectroscopy. Meddelelser Konservering 1, 3–11 (2007) 8. Chang, C., Plaza, A.: A fast iterative algorithm for implementation of pixel purity index. IEE Geosci. Remote Sens. Lett. 3, 63–67 (2006) 9. Cucci, C., Delaney, J., Marcello, P.: Reflectance hyperspectral imaging for investigation of works of art: old master paintings and illuminated manuscripts. Acc. Chem. Res. 49, 2070–2079 (2016). https://doi.org/10.1021/acs.accounts.6b00048 10. Golloch, A., Heidbreder, S., L¨ uhr, C.: Identification of amber and imitations by near infrared reflection spectroscopy. Fresenius J. Anal. Chem. 361, 545–546 (1998). https://doi.org/10.1007/s002160050943 11. Klein, M., Aalderink, B., Padoan, R., de Bruin, G., Steemers, T.: Quantitative hyperspectral reflectance imaging. Sensors 8, 5576–5618 (2008). https://doi.org/ 10.3390/s8095576 12. Liu, Y., Shi, G., Wang, S.: Color phenomena of blue amber. Gems Gemol. 50, 134–140 (2014). https://doi.org/10.5741/GEMS.50.2.134 13. Muller, A., Guido, S.: Introduction to Machine Learning with Python. O’Reilly, Sebastopol (2017) 14. Musa, M., Kaye, T., Bieri, W., Peretti, A.: Burmese amber compared using microattenuated total reflection infrared spectroscopy and ultraviolet imaging. Appl. Spectrosc. 75, 2 (2021). https://doi.org/10.1177/0003702820986880 15. Nie, L., Lin, C., Liao, K., Liu, M., Zhao, Y.: A view-free image stitching network based on global homography. J. Vis. Commun. Image Representation 73, 102950 (2020). https://doi.org/10.1016/j.jvcir.2020.102950 16. Pastorelli, G., Richter, J., Shashoua, Y.: Photoageing of Baltic amber-influence of daylight radiation behind window glass on surface colour and chemistry. Polym. Degrad. Stabil. 96, 1996–2001 (2011). https://doi.org/10.1016/j.polymdegradstab. 2011.08.013 17. Sinduja, R., Chidambaram, S., Sumathi, A.: Analysis of dimensionality reduction techniques for hyperspectral image classification. Int. J. Eng. Trends Technol. 21, 111–115 (2015). https://doi.org/10.14445/22315381/IJETT-V21P219 18. Rajabi, R., Ghassemian, H.: Spectral unmixing of hyperspectral imagery using multilayer NMF. IEEE Geosci. Remote Sens. Lett. 12, 1 (2014). https://doi.org/ 10.1109/LGRS.2014.2325874

416

P. Krupska-Wolas et al.

19. Rodarmel, C., Shan, J.: Principal component analysis for hyperspectral image classification. Surv. Land Inf. Sci. J. 62, 115–125 (2002) 20. Rygula, A., Klisinska-Kopacz, A., Krupska, P., Kuras, E., del Hoyo-Melendez, J.: The surface degradation of Baltic amber: the depth-profiling analysis and its application to historical object. J. Raman Spectrosc. 52, 1 (2020). https://doi.org/10. 1002/jrs.5942 21. Sadowski, E.M., et al.: Conservation, preparation and imaging of diverse ambers and their inclusions. Earth-Sci. Rev. 220, 103653 (2021). https://doi.org/10. 1016/j.earscirev.2021.103653, https://www.sciencedirect.com/science/article/pii/ S0012825221001549 22. Shashoua, Y., Berthelsen, M.B., Nielsen, O.: Raman and ATR-FTIR spectroscopies applied to the conservation of archaeological Baltic amber. J. Raman Spectrosc. 37, 1221–1227 (2006). https://doi.org/10.1002/jrs.1586 23. Sikorksa, E.: Spektroskopia optyczna w ocenie jako´sci z˙ ywno´sci. Uniwersytet Ekonomiczny w Poznaniu, Pozna´ n (2019) 24. Teodor, E.D., Litescu, S.C., Neacsu, A., Badea, G., Camelia, A.: Analytical methods to differentiate Romanian amber and Baltic amber for archaeological applications. Cent. Eur. J. Chem. 7, 560–568 (2009). https://doi.org/10.2478/s11532009-0053-8 25. Vaskova, H., Køes´ alek, V.: Quasi real-time monitoring of epoxy resin crosslinking via Raman microscopy. Int. J. Math. Models Methods Appl. Sci. 5, 1197–1204 (2011) 26. Vogler, A.: Photoluminescence of Baltic amber. Zeitschrift f¨ ur Naturforschung B 73, 673–675 (2018). https://doi.org/10.1515/znb-2018-0106 27. Wagner-Wysiecka, E.: Mid-infrared spectroscopy for characterization of Baltic amber (succinite). Spectrochim. Acta Part A-Mol. Biomol. Spectrosc. 196, 418– 431 (2018) 28. Winkler, W., Kirchner, E., Asenbaum, A., Musso, M.: A Raman spectroscopic approach to the maturation process of fossil resins. J. Raman Spectrosc. 32, 59– 63 (01 2001). https://doi.org/10.1002/1097-4555(200101)32:159::AID-JRS6703. 0.CO;2-D

Methods for Enhancing CH Fruition

Folk Music of the Kho, Mú in Ðiê.n Biên Province: Characteristics and Potential for Community-Based Tourism Development Lam Nguyen Dinh1

, Son Quang Van2(B) , Son Nguyen Truong1 and Dinh Luong Khac3

,

1 VNU University of Social Sciences and Humanities, Hanoi, Vietnam 2 Institute of Cultural Heritage and Development Studies, Van Lang

University, Ho Chi Minh City, Vietnam [email protected] 3 Faculty of Information Technology, Ha Long University, Quang Ninh, Vietnam [email protected]

Abstract. The Kho, Mú people in Vietnam’s Ðiê.n Biên province still preserve and promote their unique forms and genres of folk music, which has an important place and role in the cultural life of the people and acts as a means of identifying their traditional culture. Amid the current strong development of domestic and international tourism, research aiming to identify the unique values of the traditional music heritage of the Kho, Mú should be further promoted to introduce their folk music to a wider audience and develop community-based tourism. This article aims to identify the entire folk music of the Kho, Mú and discover the unique values of this ethnic group. In addition, the study also aims to analyze the tourism potential and connect the introduction of the Kho, Mú’s folk music with the promotion of community tourism, further contributing to the preservation and development of folk music in the area. The study will use ethnomusicological and cultural anthropology methods, combined with research from ecotourism, to analyze folk music and ways to bring it into the tourism development of the local community. Direct field surveys and other fieldwork methods have served as the key research methods of this study. The results show that the folk music of the Kho, Mú people of Ðiê.n Biên province has unique features compared to other Kho, Mú groups and other ethnic groups in Vietnam, as the folk music of the Kho, Mú people still preserves its traditional and unique values. The results also show that bringing folk music into community tourism development can help to preserve and develop these musical genres in contemporary society. Based on the research objectives [7], this report does not delve into the decoding of folk music in terms of musicology, which is a limitation of this research. However, through studying the folk music of the Kho, Mú and proposing connections to develop community-based tourism, this report offers important findings in the context of today’s Vietnam. Keywords: Kho, Mú people · Community tourism · Ethnomusicological · Cultural anthropology

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 419–432, 2023. https://doi.org/10.1007/978-3-031-17594-7_31

420

L. N. Dinh et al.

1 Introduction The Khmu (Vietnamese Kho, mú), also known as the Xá, are an ethnic group belonging to the Mon-Khmer language group, the Austroasiatic language family, who live in several Asian countries, such as China, Vietnam, Laos, and Thailand. In Vietnam, this ethnic group lives mainly in the northern mountainous provinces and Nghe An. The population of the Khmu people in Vietnam amounts to over 70,000 people, of which the largest concentration is in Nghê. An province (35,600 people) and Ðiê.n Biên province (16,200 people). In Ðiê.n Biên province, the Khmu people live in 36 communes in six districts (but there are no Khmu in Mu,`o,ng Lay). This article’s survey was conducted in Kéo, Công, and Ten villages in Mu,`o,ng Ph˘ang commune, Ðiê.n Biên district, with a population of approximately 1,500 people [1, 2, 4, 5]. Local surveys show that the Khmu people in Ðiê.n Biên [2] live mainly on agriculture, in which their primary crop is wet rice. Currently, the Khmu people are transforming their economic structure from pure rice farming to the VAC farming model, which includes fruit trees, fishponds, and livestock. Today, the Khmu still maintain their traditional handicraft such as weaving brocade bags and knitting household items like mats and baskets [4]. The houses of the Khmu are important tangible cultural heritage symbols of the Khmu people. Their houses are four-room stilt houses, surrounded by balconies. The outermost space, right at the entrance around the front half of the house is reserved for the reception area. The entire back half of the house is used as a sleeping area. To the right of the entrance, people designate an area for an ancestral altar and a picture of President Hô` Chí Minh. On major holidays of the year, such as Lunar New Year and the National Day, they often hold festivals and pray to commemorate their ancestors and the president of Vietnam (Figs. 1, 2 and 3).

Fig. 1. Weaving frames and traditional products of the Khmu people (Source: Dinh Lam, 2018)

In addition to housing, the clothes of the Khmu people are also an important and characteristic physical cultural symbol. In the past, their skirts and shirts were made of brocade fabric, woven in many colors. There are pieces of silver and many coins attached

Folk Music of the Kho, Mú in Ðiê.n Biên Province

421

Fig. 2. House of the Khmu (Source: Dinh Lam, 2018) Fig. 3. Traditional costumes of the Khmu women (Source: Dinh Lam, 2018)

to the shirt, creating a unique look. The skirts of the Khmu are black, almost ankle-length, with some light red or brown patterns with stripes interspersed with diagonal or triangular frills. The Khmu scarf is also adorned with silver jewelry, in the form of copper coins and silver chains. Khmu women also wear a hairpin of about 20 cm in length, with a pointed tip, bigger than a thumb. This jewelry is a prominent feature that helps distinguish the makeup and costumes of the Khmu women. Nowadays, many Khmu people also wear black Thái costumes. Men’s costumes are rarely used today. Instead, they wear popular costumes like from the Kinh. In-depth interviews with locals over 70 years old showed that men’s traditional clothes were previously black or purple, woven with brocade. Besides the tangible cultural heritage exists a system of intangible cultural heritage, including traditional folk music. Ceremonies associated with the human life cycle and religious rites associated with rice-growing culture are considered very important in the cultural life of the Khmu. In the life cycle ritual, the naming ceremony (called Po,n mo,) is performed right after the child is born. The Khmu hold a ceremony to name their babies after 2–3 days of age. At that time, the shaman (who can be a family member) is invited to perform the naming ceremony. The offerings included a chicken, a bowl of rice, and a thread. When making offerings, the shaman will take (and pick up) rice and then pray while counting the grains of rice. According to the concept of the Khmu people, if the number of rice grains counted is even, it is good, otherwise, it is bad. Therefore, the teacher often has to do the worshiping ceremony again until he asks for an even number of rice grains, then the ceremony will be completed. They will then notify the family so that they can name the child. After choosing the child’s name, a string is tied around the wrist, with the implication that the child grows up quickly and does not forget his or her name.

422

L. N. Dinh et al.

One of the other important ceremonies of the Khmu people in Ðiê.n Biên is the wedding ceremony. The Khmu organize weddings for their children when they reach adulthood. In-depth interviews have shown that the welding process takes place in three steps [5]: The first step is called the entrance. In this step, the groom’s family sends two men as representatives to the girl’s house to talk about marriage. Offerings include two bottles of wine, two chickens, 1 kg of meat, and 1 kg of salt. In this first meeting, if the bride’s family accepts the invitation, the groom can ask his siblings to come and work as housekeepers and laborers a day thereafter to help the girl’s family and, at the same time, create a relationship between the two sides. After the ceremony, the groom’s family also has to bring an offering (including two chickens, wine, tea, and cigarettes) to the village head’s house to ask for permission for the couple to officially move on to the two families and ancestors to ask about the marriage ceremony. The second step is an interrogation. Similar to step one, the groom’s family prepares chicken and wine and then sends two men to the girl’s house. In this step, the two parties will agree on the date and month of the wedding. Usually, before conducting the betrothal ceremony, both families often go to the house of the shaman in the village to find out which day is a good day to conduct a wedding for their children. The third step is the actual marriage ceremony. The groom’s family has to bring the wedding ceremony to the bride’s house from the previous afternoon. The offerings include betel nut, a quintal of pigs, five coins and six silver coins, a fish cage (with eight fish), a dress, a scarf, and a belt. In which, the five coins belong to the groom. Sacrifices are then made to the ancestors of the girl’s family. After the wedding, the groom must stay at the bride’s house for a year. After that, a pig has to be slaughtered for the official ceremony to bring the bride home. However, this custom is often not performed any longer. The wedding of the Khmu people is unique. Activities such as toasting wine and singing are very exciting for all participants. When holding a reciprocal singing ceremony, usually the boy’s family will sing before the girl’s family. And finally, funerals are also religious rites associated with the life cycle and are particularly important rituals in the cultural life of the Khmu. The deceased person is placed in their house for a day. After that day, people choose an auspicious time for the burial. According to the old customs, when shrouding, people also catch a chicken, beat it to death, and place it in the coffin. The purpose is for the soul of the chicken to lead the way for the soul of the dead to return to their ancestors. When burying children and grandchildren in the family, they must wear a white mourning towel. However, once married, girls are not allowed to wear white mourning scarves next to their parent’s house, because they have to abstain from their husband’s family. The Khmu do not build graves for the dead but organize burials in the form of deep digging and burying. The Khmu hold anniversaries on New Year’s but do not celebrate the death anniversary. On New Year’s, wine and fish are the two main offerings that are placed on the altar to invite ancestors as well as the souls of the dead. Nowadays, the Khmu also follow the general regulations of the state on funeral organizations.

Folk Music of the Kho, Mú in Ðiê.n Biên Province

423

In addition to funerals and rituals associated with the human life cycle, the Khmu people in Ðiê.n Biên also preserve and practice several other important traditional festivals, notably the Rain Prayer Ceremony and Ban (Village) Offering Ceremony. The Khmu hold a ceremony to pray for rain and worship the village once a year around the middle of July when the rice in the fields is coming in. The purpose is to pray for favorable rain and wind, as well as healthy people, pets, and livestock. Praying for rain is also an important ritual of the Khmu. People prepare for the ceremony by erecting a tent. In the tent, the offering includes dog meat and cooked beef, displayed on banana leaves (in two parts corresponding to two plates) and two jars of canned wine. The shaman weaves two bamboo (or cork) slats of different shapes. One is across; the other is a circle with a diameter of about 40 cm, and then uses a bamboo pole about 80 cm high to prop it up. After the ceremony is over, the shamans bring these objects, along with two bamboo screens, to the fields to continue the follow-up ceremony. At that time, people smeared the blood of the killed animals on the two screens. The cross-shaped slat is tied with a dog’s tail and then boiled. This action shows how people offer sacrifices to the gods and ask for their help. After the Rain Prayer Ceremony, the shaman performs the Ban Offering Ceremony. This ceremony is usually held at night in the village [5]. The offering includes two chickens. After the main worshiping ceremony, people hold another ghost worshiping ceremony. The purpose of praying is for the people who are sick to wish for their disease to be cured. On this holiday, if there is a sick person in the family, the shaman and villagers will come to the house to pray and chase the evil ghosts out of the house so the sick person to get well. After the funeral ceremony is completed, the family will organize a feast with wine and rice. In all rituals associated with the life of the Khmu, folk music is an important part. Music is associated with the Khmu people from birth to death. Folk music is a means of connecting people with people, for instance by singing reciprocal during dates. Folk music is also associated with gods in rituals, mostly by singing particular tunes and performing with local musical instruments. As music is closely associated with the cultural life of the people, folk music can be considered an art form that is cherished by domestic and foreign tourists alike. Thus, this paper will further dive into the value of Khmu folk music for domestic and international tourism development in the region. ij

ij

2 Research Methods During the study, the author conducted fieldwork directly in the villages of Kéo, Công, and Ten in Mu,`o,ng Ph˘ang commune, Ðiê.n Biên district – all of which are populated by Khmu people. The research also included in-depth interviews with 25 practicing artisans, with a special focus on artisans over 70 years old who remember their traditions well enough; and three village elders/chiefs (già làng). For the research, all interviewees needed to be talented artists who have participated in singing for many years and have thus a rich and diverse understanding of local culture. With regards to musical instrument performers, the research focuses on selecting talented artisans who are not only capable of performing traditional musical instruments but also crafting these. Several village elders

424

L. N. Dinh et al.

and chiefs, along with some other artists, were also selected for in-depth interviews in this study to learn more about the migration process, living circumstances, and general cultural characteristics of the Khmu community in this locality. Thus, the artists selected in this study not only can perform local art but also have a high professional level, and often maintain and practice forms of folk music. The in-depth interviews are important to clarify the function of folk music within the Khmu community, as well as to gain a deeper understanding of the processes of creating, practicing, and maintaining folk music forms in Vietnam. The in-depth interviews also assessed the nature of the music and the different genres in the life of the people. Interviews focused on the history, names, and characteristics of musical instruments and their development; The interviewer also watched demonstrations of the instruments and observed these in each locality. The interviews asked the artisans about the presence of domestic and international tourists visiting the locality, and some other related issues. In addition to in-depth interviews, the study also included interviews with two experts on folk music. Along with that, the author conducted interviews with festival participants and could thus gain a closer look at assessing the relationship of music with the cultural and religious life of the Khmu. Furthermore, documents, books, magazines, and daily newspapers of Ðiê.n Biên province were also important sources to supplement the study. As such, the ethnomusicological and cultural anthropology methods are central to this study. The study places insiders, i.e., indigenous artisans who create and practice folk music locally, at its center. Ethnomusicological field research in the locality helped clarify the general picture of the music in the cultural life of the Khmu.

3 Characteristics of Folk Music and Potential for Community Tourism Development The Khmu people preserve and promote many genres of folk music in their cultural life. Music is also a particularly important resource to develop local communitybased tourism, helping the community socio-economically and maintain their precious heritage. The potential, as well as the treasure of folk music of the Khmu, is largely supported by their traditional musical instruments. In the process of assessing each different category and genre, the study also includes information related to tourism potential. 3.1 Folk Songs In terms of genres, the Khmu have three main folk song categories: To,m, Chom oi, and Colin. To,m is a popular folk song in the cultural life of the Khmu and can be used for solo singing and reciprocal singing in different spaces such as during weddings, celebrations for a new house, and even for lullabies. The author found that, while singing To,m, sometimes the Khmu people also use the Pí tót (horizontal flute) and the Pi s˘am roi. To,m singing reflects different aspects of life from family relationships, friends, and weddings to productive labor activities, entertainment, and festivals. Lyrics of To,m at a wedding can include the following aspects:

Folk Music of the Kho, Mú in Ðiê.n Biên Province

425

– A singer representing the groom’s side sings when the groom’s family asks for the bride. Lyrics include verses like “I took his son to my husband’s house. After the ceremony is over, the bride must return to her groom’s house.” – A singer representing the bride could thank the groom’s family on behalf of the bride’s family with: “The girl’s family is poor, so it’s not worthy of the boy’s family”, and similar phrases. Through in-depth interviews and performances, the author discovered that To,m singing follows a free rhythm. Vocals with many undulating tones create a very unique sound. The melody proceeds consecutively in the third, sixth, and seventh intervals. In lullabies, an interviewed artist practiced with the scale a - h - e - g; with an additional c sound only appearing once. One practicing artisan said, “To,m singing in the locality has been preserved for many generations. This song is often part of art programs in the district and Ðiê.n Biên province. So far, To,m singing is still maintained in many cultural activities and local beliefs.” Another practicing artisan said, “In recent years, people traveling to Ðiê.n Biên often go to Muong Phang to enjoy local food and To,m songs, as well as Pi flute songs. Some also like to listen to the Ðao (a musical instrument associated with the work of the Khmu). Followed by the To,m singing comes the Chom oi singing. This is a song with a bright and strong rhythm, expressing the fun nature of the song. This song can be sung in many different spaces, especially on happy ceremonial occasions such as weddings and new house celebrations in the village. The author learned that during the wedding of the Khmu, the young and the old still sing this song nowadays. Unlike the To,m song, when singing the Chom oi tune, people often do not accompany it with musical instruments. Chom oi is a song with a clear melody and rhythm. Some of the songs we dealt with were 2/4 tempo. The melody begins with pitch intervals: e1 - a1 - e1 - f1# - e1 - H - A - e1. The progression can easily be seen in the melody from the e1 jump to a1, then it jumps back and goes up a major second, then back to (e) and continues to jump down an octave below the first octave. Such clear melodic organization makes the chom oi song free and easy to listen to. Along with the To,m and Chom oi singing other traditional folk songs are practiced and passed on to future generations by the people. Tourists who come to this land likely will want to enjoy these songs, along with the unique local dishes. The last song that we discovered so far is still practiced in the cultural life of the Khmu: the Co, lin. This song has a slow and free rhythm, with a very melodic nature. When listening to the Co, lin, one can see similarities with the To,m tune. In terms of content, Co, lin reflects the working life and daily activities of the Khmu. People can sing this song while on the field, at a wedding, or while a child is sleeping, but not in religious and funeral spaces. It is intimately attached to the people here, along with the To,m and Chom oi, the Co, lin attracts visitors to the village. It can be said that these folk songs are an important potential for the Khmu to develop community-based tourism while preserving their cultural heritage and promoting their culture to the world.

426

L. N. Dinh et al.

3.2 Musical Instruments Along with the folk songs, the traditional musical instruments of the Khmu people are also a unique heritage, indispensable in their cultural life. The musical instruments also promote their local cultural and economic development, especially in the current context. Traditional musical instruments include, first of all, the vertical flute Pi S˘am roi. This is a native musical instrument, unique and characteristic of the Khmu. The Pi tam flute belongs to the family of blow instruments, with a free and vibrating reed shape. This instrument is intended for use by men only. It is used for solo performances or as an accompaniment to the To,m and Co, lin songs. An artist explained that when guests come, they love to hear this flute. Many tourists want to buy the instrument as a souvenir, but the people here have not developed this instrument into a commercial product for sale, although many good artisans in this locality can create them. The Pi S˘am roi is made up of a small cork piece (the people call it N´u,a Sâu chít; it is also possible to use other types of bamboo with similar small stems), then cut off the top for a length of about 60–70 cm. People cut off the top and leave only the body only about 50 cm long. Then, at the top part (the part that creates the blown reed), with a diameter of about 0.3 cm, people measure down to 1.5 cm and make this part thinner; then use a sharp knife to cut a part of about 2.5 cm to form a reed, the sound driver. From the top of the flute, measuring down to about 35 cm, people punch three holes to adjust the pitch when performing, in which the second hole is 4 cm from the first hole; The third hole is 3 cm from the second hole (Fig. 4).

Fig. 4. Solo performance of the Pi s˘am roi (Source: Dinh Lam, 2018)

When performing, people close the reed. Combined with the use of the three holes drilled in the body of the instrument, the Pi s˘am roi produces a very unique, husky, and bass sound. Regarding the performance, the Pi s˘am roi often plays melodies from Khmu folk songs, most of which follow the melodies of To,m singing.

Folk Music of the Kho, Mú in Ðiê.n Biên Province

427

Along with the Pi s˘am roi goes the Pí tót flute, also known as the nose flute. To produce such a flute, the artisan will take a cork tree (dry or fresh) with a length of about 60 cm, about 3 cm wide. The artisan will cut one stick to leave the eyes at both ends. At one end, he will make a square hole about 10–12 cm from the eye as a blowhole. Another hole is cut into the side about 7 cm from the eye of the tube as a punch hole. The nasal flute produces two tones, one when played and the other when opened. When playing, people play the flute with their noses while singing. The sound of the flute creates unique reciprocity between human voices and musical instruments. The Pí tót is used in daily life, not in funerals and religious ceremonies. In the past, the Pí tót was a male instrument, but now women also play it. Practicing artisans and locals said that when foreign tourists come to the village, they are very interested in the Pí tót performances. Together with other forms of folk music, this musical instrument offers an important potential for the development of community tourism. Next comes the H’rông instrument, the Ðàn môi. The mouth harp of the Khmu used to be made of bamboo, but now there are many bronze ones. The average size of this instrument is about 10 cm long and 1.3 cm wide at its widest part. The H’rông is made by taking a bamboo stick (or inlaid copper) and cutting out a reed in the middle of the body and then beveling the head of the instrument. When playing, people put the lip onto the oral cavity and use the right hand to hold the handle, while the left hand bounces slightly on the pointed end of the instrument, producing a unique sound. Lip lutes are mainly used by men, and sometimes they play in accompaniment to solo singing or love songs, but it is not common. Another particularly important instrument in the cultural life of the Khmu in Ðiê.n Biên is the Ðao instrument. People make this instrument by taking an old cork tree (dry or fresh) with a diameter of about 5 cm and cutting a piece with three sticks (about 80 cm long). People use a small iron bar to chisel through three segments, then use a knife to bevel on both sides of the pipe (symmetrical) to create two lutes attached to the tube (the Khmu people call it chamfered rods). About 30 cm from the end of the tube (from the bottom up) people punch two holes 10 cm apart from each other into the tube. When playing the Ðao will emit two tones spaced exactly four intervals from each other. When playing, the left hand holds the Ðao, and the thumb presses the upper hole, while the ring finger presses the bottom hole. During Lunar New Year, women play the Ðao during singing. Usually, there are only about eight people in Ðao dances, divided into two groups. In the inner circle, three people are playing the Ðao according to certain rhythms. In the outer circle, boys are holding wine bottles in their hands and singing to the rhythm of the Ðao, occasionally taking a sip of wine and creating a fun atmosphere (Figs. 5, 6 and 7). Finally, the Tu,n hét is a musical instrument that belongs to the string instruments. The neck and resonator are made entirely of bamboo. The Tu,n hét is mainly used by men, and the instrument is not used in religious ceremonies such as funerals or to pray to village deities. In addition to these six instruments, in the past, the Khmu in the Mu,`o,ng Ph˘ang commune also had some other instruments such as drums, gongs, and flutes (Pí). But now, these instruments have been lost, with only some gongs remaining.

428

L. N. Dinh et al.

- môi) (Source: Dinh Fig. 5. Performance of the Pi tót (Source: Quoc Fig. 6. H’rông (dàn Dinh, 2019) Lam, 2018)

Fig. 7. Ðao instrument (Source: Dinh Lam, 2018)

Surveys and in-depth interviews showed that the Khmu preserve their indigenous folk music heritage. Many interviewees also said that tourists from different countries come to this place and often want to enjoy the local folk music, besides the local cuisine. Thus, this is an important opportunity for the Khmu to develop a community-based tourism economy associated with the preservation of their traditional culture. 3.3 Potential for Community-Based Tourism Development As ecotourism and experiential tourism are developing on a global scale, the government of Vietnam has tried to perfect policies for tourism development, especially promoting community-based tourism [10]. The government has set out the Strategy for Vietnam’s Tourism Development by 2030, in which the focus is on turning tourism into a spearhead economic sector, creating a driving force to promote the development of other industries and sectors, and making an important contribution to the formation of the modern economic structure. As importance is attached to cultural tourism development,

Folk Music of the Kho, Mú in Ðiê.n Biên Province

429

the conservation and promotion of heritage values and national cultural identity plays a crucial role. Based on the fact that the number of tourists has been increasing year by year before the pandemic, the government has set the goal that by 2025 the country would at least receive 35 million international arrivals and 120 million domestic visitors, maintaining a high average growth of international and domestic visitors from 12–14% and 6–7% per year, respectively [11]. By 203, the government strives to receive at least 50 million international arrivals and 160 million domestic arrivals, maintaining an average growth rate of 8–10% and 5–6% per year [10]. Thus, based on this approach, exploiting ethnic cultural identities in the communitybased tourism development strategy offers many opportunities. Ethnic cultural heritage, such as the folk music of the Khmu in Ðiê.n Biên province, has many opportunities for development and integration to promote their traditions as a national treasure to domestic and international tourists, contributing to promoting local economic and social growth. Also in the tourism development strategy, the government emphasized the development of cultural tourism products in association with the conservation of the significant advantages of the nation’s historical-cultural heritage, focusing on exploiting the diverse and distinctive strengths of each region to form unique tourism products with competitive advantages, contributing to building a prominent brand identity of Vietnam’s tourism. The strategy is moreover also meant to promote the development of resort tourism, weekend tours associated with healthcare and education, as well as cultural and historical exploration and enhanced cultural exchange between regions [10]. Based on the tourism potential of the locality, since 2013, the People’s Committee of Ðiê.n Biên province [6] has assessed the local tourism resources to promote its development. Accordingly, the People’s Committee of Ðiê.n Biên province issued Plan No. 906/KH in 2013 on the tourism development strategy of the province by 2020, with a vision for 2030. The main objectives are to build Ðiê.n Biên into one of the three key tourism development areas of the northern midland and mountainous tourism region. Accordingly, the province will focus on developing sightseeing tours for visitors to learn about the traditional cultural identity of the ethnic minorities of this region [8]. The plan of the Party Committee of Ðiê.n Biên province [6] also stated that the province wants to develop tourism products, focusing on historical and spiritual tourism; cultural and ecological tourism; and entertainment and healthcare tourism [1]. It is estimated that the total number of visitors to Ðiê.n Biên in 2020 was about 351,000, reaching 38.57% of the plan for the year (910,000 arrivals); in which international visitors were estimated at 16,800 arrivals, reaching only 8% of the planned 210,000 arrivals [4] – these low margins are likely due to the pandemic-induced travel restrictions. At present, there are 215 tourist accommodation businesses in the province with 2,954 rooms, 5,139 beds, and 11 cultural villages in Ðiê.n Biên district and Ðiê.n Biên Phu city. These firms have also been implementing services for tourists, such as food and beverages (F&B), cultural and artistic exchanges, and the production and trade of traditional brocade products. Several homestays in the area also meet the general requirements for F&B, accommodation, and activities. Thus, these are ready to participate in local community-based tourism development.

430

L. N. Dinh et al.

It should be added that, at present, the province has planned out Ðiê.n Biên Phu airport as a sub-regional international airport. The road system between So,n La, Hòa Bình, and Hanoi – as well as the provinces of the northern delta to Lai Châu, Lào Cai, Yên Bái, and Phú Tho. – forms a network between provinces and regions and serves as the foundation for increased tourism development. With its geographical position, the province is also a favorable bridge between the northern region and the provinces of northern Laos, southwest China, and southern Myanmar. These are favorable conditions for Ðiê.n Biên to exploit the potential of tourism, services and trade, and the bordergate economy with neighboring countries in the region. Thanks to its infrastructure, tourists coming from nearby large cities, such as Hanoi, were able to travel to Ðiê.n Biên smoothly [2]. This highlights the particularly important conditions for promoting Ðiê.n Biên’s tourism, including the folk music of the Khmu people in the area. In addition to the Khmu ethnic group, Ðiê.n Biên province is also home to 18 other ethnic groups, each with its unique cultural identity and features, including both tangible and intangible cultural heritage. The traditional socio-cultural institutions of each ethnic group – such as the To,m singing in Mu,`o,ng village and the production, belief activities, and festivals of each ethnic group – are attractive tourism resources for upcoming visitors, especially international tourists. Like the Khmu, most festivals and cultural and religious activities of the ethnic minorities in Ðiê.n Biên province also include folk music. Thus, traditional music is one of the important means to promote cultural tourism in the province. Based on assessing the characteristics of the Khmu folk music genre and the related potential for community-based tourism development in Ðiê.n Biên, coupled with the government’s policy, the author suggests the homestay tourism model as one of the most potential models that can attract international tourists. The model not only helps preserve and promote indigenous traditional music heritage in the community but also helps artisans and people live by their craft, promote the passion for their art, and strengthen the sense of responsibility to protect their heritage for future generations. However, specific orientations and plans are necessary to avoid an uncontrolled development, leading to improper preservation of the inherent properties of each type of folk music. In other words, we should avoid “professionalized” folk music, which is incorrect for many traditional folk music repertoires, and instead preserve the existing repertoire. To do this well, we need a strategy and scientific management. For instance, traditional musical and cultural heritage can be digitalized on websites and other internet-based applications, so that visitors can look up and learn about the different types of music and the places they want to visit. If done right, tourism development can preserve and promote traditional music and, at the same time, contribute to the socio-economic development of the region and the nation.

4 Conclusion The Khmu people in Ðiê.n Biên continue to preserve and maintain many unique forms of indigenous cultural heritage, particularly their folk music. Their traditional music is one of the major factors which identify their culture and differentiate it from other ethnic groups. Closely associated with the cultural life of each person, the traditional

Folk Music of the Kho, Mú in Ðiê.n Biên Province

431

folk music of the Khmu people has been present in most of the cultural life of the Khmu, right from the birth of a child to the time when people return to the other side of the world. Thus, music performs its functions in religious ceremonies and traditional beliefs of the Khmu. Their folk music is also a means of moral education for a younger generation, for example as a means for teenage boys and girls to come together and send love to each other. Music acts as a link between humans, and humans and gods. Folk musical instruments play a particularly important role in the cultural life of the Khmu, in which there are religious rituals and beliefs. Traveling to experience and enjoy the traditional music of the Khmu people, as well as of other ethnic groups in Ðiê.n Biên province, can serve as a means to learn and enjoy the indigenous traditional culture of the region. Domestic and international tourists alike are increasingly interested in the development of ecotourism and real-life experiences, especially attracting tourists from advanced and developed countries. This has been the main reason for the government and Ðiê.n Biên province to produce many important policies to develop community-based tourism in recent years. However, to develop community-based tourism associated with cultural products in the right professional direction – while also promoting the set goals of developing the service economy and preserving and promoting traditional culture – the tourism industry needs to improve the quality of its services, from the central to local levels. It is necessary to focus on management tasks and the promulgation of a suitable legal framework through the correct marketing activities, education, and mobilization of people to raise awareness. With that, the tourism industry of Ðiê.n Biên and the development of community-based tourism [3] in the villages of the Khmu is likely to develop in a positive direction. Acknowledgement. The authors wish to express their gratitude to Van Lang University, Vietnam for funding this research.

References 1. Ðiê.n Biên Provincial Party Committee: Resolution No. 03-NQ/TU on tourism development in Ðiê.n Biên province to 2025, with a vision to 2030, 7 May 2021 2. Duong, H.: Ðiê.n Biên Tourism: Potential and Development Orientation. Ðiê.n Biên Electronic Newspaper (2014). Accessed 20 Nov 2021 3. Pujiastuti, E.E., Sriyono, S., Soeprapto, A., Nugroho, S.P.: Visiting peaceful Maghilewa: culture and nature as the potentials of community-based tourism. In: LPPM UPN “Veteran” Yogyakarta Conference Series. Proceeding on Political and social science series, vol. 1, no. 1, pp. 107–115 (2020) 4. Ministry of Culture, Sports and Tourism: Ðiê.n Biên tourism towards long-term goals. Website of the Ministry of Culture, Sports and Tourism (2021). Accessed 15 Nov 2021 5. Nguyen, H.D., Nguyen, D.L.: Additional survey by Nguyen Dinh Lam in 2018: Field Report on Folk Music and Dance of Thái, Lao, H’Mong, and Kho Dancing in Ðiê.n Biên province. Library of the Vietnamese Institute for Musicology, Hanoi, pp. 2–14 (2009) 6. Nguyen, T.T.H.: Management of Vietnamese civil servants research: a descriptive approach. J. Public Adm. Gov. 11(1), 285–315 (2021) 7. Phu, P.X., Anh, V.T., Tran, N.T.B., Van, P.H.T., Duyen, P.N.: Contribution of community tourism to poverty alleviation in Vietnam: a case study. Asia-Pac. J. Rural Dev. 21(1), 65–83 (2011). https://doi.org/10.1177/1018529120110104

432

L. N. Dinh et al.

8. People’s Committee of Ðiê.n Biên Province: Plan No. 906/KH on the Strategy for tourism development of Ðiê.n Biên until 2020, with a vision to 2030, 16 April 2013. Ðiê.n Biên (2013) 9. Population and Housing Census: Full Results, pp. 134–225. Hanoi 2009 (2010) 10. Prime Minister: Decision No. 147/QD-TTg on Vietnam’s tourism development strategy to 2030, 22 January 2020. Hanoi (2020) 11. Tran, T.H., Hoa, H.C., Nguyen, T.P.L., Do, T.N., Le, H.A., Nguyen, Q.S.: Human resources for tourism and travel business in Vietnam, business and economic research. J. Public Adm. Gov. 10(4), 63–82 (2020)

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province): Unique Cultural Characteristics and Festival Protection Solution Hue Phan Thi1 , Ninh Ngo Hai1

, Son Quang Van2(B)

, and Dinh Luong Khac3

1 Faculty of Culture, Ha Long University, Uông Bí, Quang Ninh, Vietnam

{phanthihue,ngohaininh}@daihochalong.edu.vn 2 Institute of Cultural Heritage and Development Studies, Van Lang University, Ho Chi Minh

City, Vietnam [email protected] 3 Faculty of Information Technology, Ha Long University, Uông Bí, Quang Ninh, Vietnam [email protected]

Abstract. Vietnam is a country with many traditional festivals. Festivals are the most compact accumulation of culture of a nation, a region, and are crowded community activities with religious, sacred, and solemn beliefs that take place in different spaces and times fixed, historically, and ethnically distinct. In the flow of Vietnamese traditional festivals, the Tien Cong festival has many similarities with traditional festivals in general but still has many unique features of the coastal area of Quang Ninh. The article introduces an overview of the Tien Cong festival, deeply analyzes the unique cultural features, meanings, and values of the festival through several typical offerings, rituals, and folk games, and proposed some solutions to preserve and develop the Tien Cong festival (Ha Nam island, Quang Yen town, Quang Ninh province). Keywords: Tien Cong festival · Procession ceremony · Ha Nam Island · Quang Yen town

1 Introduction Ha Nam Island, the land located south of Quang Yen town, Quang Ninh province, Vietnam, is separated from the mainland by the Chanh river - a tributary of the Bach Dang river with more than 34 km of sea dike length. This is a land famous for many relics and festivals bearing the characteristics of Northern folk. One of the unique festivals with distinctive features of the coastal region of Quang Ninh in both the ceremony and the festival addition has been honored, as a national intangible cultural heritage is the Tien Cong festival (the procession of people). Tien Cong festival is one of the biggest traditional cultural festivals of the year in Quang Yen town, showing the gratitude of the local people to the 17 Tien Cong who had the merit of “encroaching on the sea” to open the hamlet, building villages, established “Four communes” on Ha Nam island (Quang Yen town, Quang Ninh province). At the © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 433–444, 2023. https://doi.org/10.1007/978-3-031-17594-7_32

434

H. P. Thi et al.

same time, meeting the spiritual and religious needs of the people of Ha Nam Island, reminding the young generation to remember their roots, and arousing the principle of “Drink water, remember the source”, “Respect the old man and get a life” and uphold the solidarity of clans and villages. This article does not fully introduce the rituals and folk games taking place in the Tien Cong festival, but only introduces the festival in general, analyzes and clarifies its unique cultural features, meanings, and values value of the festival. Thereby, readers understand and deeply feel the intangible cultural values of the festival, and at the same time offer some solutions to preserve and promote the Tien Cong festival during the integration period.

2 Research Methods Methods of document analysis and synthesis: Understanding, accessing, and inheriting documents and research works related to the Tien Cong festival. Ethnographic field method: Attend the Tien Cong festival to observe record, record, and photograph activities taking place during the ceremony and festival part of the festival. In addition, the author also interviewed and discussed with some people in the Festival Organizing Committee, with representatives of families who held the Thuong Tho festival, local folk singers, and some local people. Cultural management to collect accurate documentary information, gain a deeper understanding of the festival’s origin, uniqueness, meaning, and value, and preserve and promote the Tien Cong festival in recent years.

3 Overview of the Tien Cong Festival 3.1 Origin of the Festival Legend has it that, in the reign of King Le Thai Tong, era Thieu Binh (1434), there were 17 families in Kim Hoa ward, Hoai Duc palace, south of Thang Long citadel (now Hanoi), living mainly by farming. Farming and fishing on the Kim Nguu River. Now, to expand the city, the king took their land, in return, allowing them to find land, establish a village anywhere, and be exempt from taxes at first. 17 families went down the Red River to the Bach Dang estuary to find a way to make a living. At first, they lived on boats by fishing and fishing. One night, they took refuge in a floating mound of the tidal flat, heard the sound of frogs, knew there was freshwater here, 17 families decided to stop at this tidal flat to reclaim land, establish a hamlet and establish a village [2]. To commemorate the merits of the Tien Cong who openly established the village, the people of Phong Luu commune, Ha Nam island established a temple in Cam La village and held the Tien Cong festival (also known as the Tien Cong temple festival, the La temple festival), the procession festival), takes place from January 4 to 7 every year. According to legend, the Tien Cong festival was born around the end of the 17th century, worshiping 17 Tien Cong, including Song Vu, Hong Tiem Vu, Huy Ngoan Bui, Ngo Bach Doan Ngo, Phuc Coc Nguyen, Phuc Thang Nguyen, Phuc Vinh Nguyen, Khep Le, Mo Le, Tam Tinh Vu, Giai Vu, Nghe Nguyen, Thuc Nguyen, Bach Nien Bui, Viet

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province)

435

Pham, Quang Tin Duong, Quang Tan Duong was the first to build three villages: Phong Coc, Cam La, Yen Dong [1]. Not only remembering the fairies who openly established villages, but the festival is also an occasion to honor the elders. 3.2 Festival Space and Time The festival takes place in the area of Tien Cong Temple and at families with an old woman (aged 80, 90, and 100) in Cam La, Phong Coc, Phong Hai, Yen Dong, Nam Hoa, and Yen Hai communes. However, the festival atmosphere covers the whole area of Ha Nam Island, Quang Yen town. Tien Cong Temple: This is the place to worship 17 Tien Cong, which is the center of the Tien Cong festival with ritual activities taking place. Tien Cong Temple is located in the center of Ha Nam island commune, built on high ground with an area of nearly 3,000 m2 , facing east, with Nhi (=) architectural style, 3-room structure 2 wings, roof tiled with comedy nose. Tien Cong Temple keeps a lot of valuable artifacts and worshiping objects. In the middle, there is a wooden altar painted with vermilion and gold, beautifully carved, inside there is a tablet “Enlightenment in the fields of ten seven fairies and gods” (the gods of the 17 fairies who openly opened the land). Above the nave, there is a large letter with four words “Vulnerable and righteous people” expressing the virtues of the people who are both affluent and chivalrous. There is also a horizontal painting, a couplet praising and remembering the great merits of immortals. There is also a stone stele in the temple about Mach Lake, the origin of the Tien Cong [4]. The family church of the Tien Cong (prayer place of the family worshiping Tien Cong) where the Co Ra ceremony takes place is also known as the to sacrifice ceremony. The 17 Tien Cong are not only worshiped by the villagers at the Tien Cong Temple, but also by their descendants, who set up their churches to worship their ancestors (one of the 17 Tien Cong and the next generations of their family). From 1434 to now, the descendants of the Tien Cong family living on Ha Nam Island have reached the 22nd generation [3]. The family church is a place of religious activities according to the custom of ancestor worship. On January 4, every year, descendants gather here to perform the ceremony to announce their ancestors and inform the clan council. Thuong Tho (Mr. Thuong). Therefore, the family churches here have contributed to the great Tien Cong festival with typical features of rural Vietnam. The family with old Thuong is the place where the celebration of the great old man’s life takes place. Families with 80-year-old, 90-year-old, and 100-year-old grandparents prepare to decorate the family premises according to traditional rituals to celebrate their longevity. On 6 January, the family organizes a celebration of the long life of Thuong at home as a day of reunion. Tien Cong festival is held for four days from the fourth to 7 January every year. This is the busiest time of the year on Ha Nam Island.

436

H. P. Thi et al.

3.3 Overview of the Order of the Tien Cong Festival To prepare for the Tien Cong festival, before that, from December 29 to January 3, the villagers had a meeting to organize the festival, prepare offerings, select sacrifices, and review the round old men and women 80, 90 years old, 100 years old, decorated Tien Cong temple; the family has the old Thuong decorating the place to celebrate the longevity ceremony and the items and sacrifices to bring the old man to the Tien Cong temple… and the main festival takes place from January 4–7 with the following rituals: On the 3rd or 4th day of January is the ceremony of “The feast of the family” or “The sacrifice of the ancestors” held by the descendants at the Tien Cong family church. January 5th: The villagers hold the festival opening ceremony at Tien Cong Temple. January 6th: Families hold a birthday celebration for the elderly at home. January 7th: In the morning, families hold the procession of Thuong to Tien Cong Temple and Thanksgiving. At noon at Tien Temple Village citizens, hold the Tien Cong sacrifice ceremony, Groundbreaking ceremony, and Wrestling ritual. In addition to the ritual part, the festival part of the Tien Cong festival is also extremely rich and unique with many folk games such as cock fighting, swinging, human chess, whore’s nest, whore, singing, and many other activities. Culture, art, and sports take place in the area around Tien Cong Temple. The festival ends on the afternoon of January 7. After the Tien Cong festival, the villagers of Ha Nam Island entered all the activities of the New Year, people here believe that after sacrificing Tien Cong, they have been supported and given the strength to continue the career that the Tien Cong left behind. The organization of the Tien Cong festival has met the needs of community cultural activities, spiritual needs, and entertainment of local people at the Bach Dang estuary.

4 The Unique Culture of the Tien Cong Festival Besides the similarities with traditional festivals in general, the Tien Cong festival also has many unique features, imbued with elements of marine culture, only found on Ha Nam Island, which are expressed through several offerings, offerings, rituals, and traditional folk performing arts. 4.1 Offerings Firstly, the smiling Long Ma - the symbol of the worshiping object in the festival, bears elements of marine culture. Offerings in the Tien Cong festival are carefully prepared by the villagers, including sticky rice, Banh giac gac, and dishes from cattle, poultry, and sea products. Among the offerings, there is an offering that is considered the symbol of the object of worship in this festival, which is the Dragon Horse. Dragon Horse is a bogus animal, the dragon head, and the horse body is a sacred animal that exists in the imagination of folk1 . To make a dragon-horse is an art of 1 According to the Buddha’s history, the dragon horse is an evil animal in the sea; the Buddha

took refuge as a disciple, assigned to control water in the East Sea, called Ananda.

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province)

437

displaying fruits of the Ha Nam - Quang Yen people. Dragon horse is made with fruits and local products (at least 5 kinds of fruits). Dragon’s head - dragon’s head is cleverly connected with green papaya. This fruit is divided into 2 parts, the beautiful, round, and large part is used as the head, and the small part is cut in half to make the ears. Eyes made of fresh, round areca nut (or dotted with longan seeds or with sparkling marbles). The nose is attached with a double-eyed custard apple. The fangs are made of red peppers. Beards are made of crochet flowers. The neck is made of banana flower folds stacked on top of each other. Dragon horse body - horse body, is folded with bunches of green pepper bananas, placed upside down, the body is dotted with yellow tangerines, red peppers, yellow, white, and red chrysanthemums. The tail and mane are tied with a bunch of bird’s nest flowers (woody plants like coconut and areca). The legs are folded with banana flower folds. The dragon-horse is placed on a rectangular platter painted in gold. The offering of Dragon Horse is highly valued by residents of Ha Nam Island. Each family has a long-life person, always has a five-fruit tray displayed in the shape of a dragon-horse on the incense burner, and is carried on a palanquin during the procession of Thuong to Tien Cong temple. The image of a dragon-horse in the procession represents the majesty and majesty of each family on Ha Nam Island. The dragon-horse is a symbol of the strength to rise above nature and master life. The dragon-horse has both the arrogant appearance of a dragon and brings water to regulate the earth, trees, and everything, to quell storms, and floods, and at the same time brings the endurance and endurance of horses to cross miles and miles deep. In the view of the ancients, the nature of the sea here hides many mysteries and disasters. It is believed that worshiping the Dragon Horse helps people to gain synergy, withstand big waves and overcome all obstacles. Second, Giay Gac cake: The color of Ha Nam Island’s culinary culture. Gac cake is an indispensable offering at the Tien Cong festival. In the Ha Nam island commune, white shoe cake is used for filial piety, Gac cake is used in festivals and for joy. The main ingredients for making banh giay are glutinous rice with yellow flowers and gac fruit. People take gac intestines soaked with glutinous rice and white wine. After the sticky rice is cooked and has a bright red color, it is pounded very finely and then molded into a cake. The cake must be large, round, bright red, both beautiful and fragrant, 2–4 cm thick, 20 cm to 50 cm in diameter. The big and red cake is the unique feature of Ha Nam banh chung, because, in other provinces and cities in Vietnam, there are banh chung, but usually, it is white and small, usually 3–5 cm in diameter. Gac cake is both an offering to the Fairy, symbolizing luck, fortune, and good things for a new year’s start, as well as a food and a gift for relatives and visitors to your home. People on Ha Nam Island with the implication that people’s hearts are full of each other. 4.2 Rituals First, is the festival of the family - the festival of the descendants of the family of the fairies. The opening ceremony of the Tien Cong festival is the festival of the ancestors, which is held on the 4th day of the first lunar month by the descendants of the Tien Cong clans at the church of the ancestors to see off the ancestors who have returned to

438

H. P. Thi et al.

Fig. 1. The smiling dragon is made from fruits (Source: Giap Luong Thuy, 2017)

Fig. 2. Bringing old Thuong to Tien Cong temple with a hammock (Source: Giap Luong Thuy, 2017)

celebrate Tet. Lunar New Year with children and grandchildren. Families with elderly people (80 years old, 90 years old, 100 years old) bring offerings to pay respects to their ancestors and ancestors who have blessed the elderly to have a long life, and at the same time, respectfully notify the association. The family members know which family will have a long life this year and invite their relatives to attend the birthday celebration on January 6 at the family. The festival of the family name means praying to the ancestors to bless their descendants with good health, prosperous business, a happy and prosperous new year, and an opportunity to show filial piety to ancestors, grandparents, parents, and family ties [2]. This ritual is the festival of descendants of the ancestors. Second, is the custom of living in the Thuong Tho festival - a unique ritual in Vietnam. The people of Hainan Island believe that a family with a long life is that family that lives in good virtue, does many good deeds and works well in production, so it is supported by the Tien Cong for longevity. Regardless of who it is, whether, in the upper or lower limbs, the person who reaches longevity is respected by the family as “the Thuong” and becomes the second name for everyone in the family to use in communicating with the people. Reach that age. Thuong Tho festival, also known as the slag ceremony, is held at the old man’s family on January 6. Around 7 a.m., Thuong Tho went to the altar to burn incense to worship his ancestors and then went out to sit on a life-long chair. Descendants near and far, clans and neighbors came in large numbers to participate in the ceremony. The eldest son reads the gratitude letter to his parents, and then his descendants and everyone have offerings to wish him long life and a life ceremony “one life is equal to a bunch of dead offerings”, wishing him long life and good health. Received fortune from Thuong. With the upper-class people coming to celebrate their longevity, after receiving the ceremony, Grandpa Thuong had to get up from his chair to invite the upper-ranked people who were his father to come to the table to drink water. Longevity Ceremony is a festival for families and clans. Children and grandchildren celebrate the life of their grandparents, and parents, and have more happiness and pride. In Ha Nam, the Thuong Tho festival is more important than both the wedding ceremony and the funeral. Therefore, people who have passed their longevity when they die, their children and grandchildren rarely cry, funerals are solemnly held by parents who return

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province)

439

to their hearts with satisfaction, funerals do not blow trumpets but have music teams. The bowl of sound is to the tune of Luu Thuy Hanh Van. When bringing the party, there are also bowls and flags of the five directions like a festival. Third, the procession ceremony - the soul of the festival of ancestors. It can be said that without the procession of longevity, there is no Tien Cong festival. The procession ceremony (process Thuong) to Tien Cong temple takes place on the morning of January 7. From 5 a.m, the procession departs from the old man’s family. The procession with drums and gongs opened the way, the wards were clad in bowls, golden umbrellas, and purple leaves, and the descendants followed Thuong in the king’s procession. Thuong holds a cane to walk next to the hammock or next to the palanquin. If the old man is weak, he will sit on the palanquin or lie on the hammock for his descendants to carry. The eldest son walked next to the old man’s hammock, followed by his brothers and sisters in the family. If the procession passes through the house of Thuong’s son-in-law, nephew-in-law, grandson of Thuong, his descendants set up a small shop on the side of the road to welcome Thuong into the restaurant to rest so that his children and grandchildren can bow down to give thanks and wish him a long life. When coming to the temple, the procession must follow the regulations: the procession goes in and goes on the right, and the procession leaves on the left. Offerings were brought into the temple, old man Thuong and his descendants solemnly entered to worship, solemnly read a poem to give thanks for the merits of the Tien Cong, and pray for good health, prosperity, and peace (Fig. 2). The procession of Thuong in the Tien Cong festival upholds filial piety, respects the elderly, and shows equality between the classes of villagers in the old countryside. Any old man or woman who used to be dignitaries, translators, or commoners, even those who live in the village until they reach their old age are called Thuong and are carried to the Tien Cong temple to celebrate the ancestors. It is the mark of “Heavenly Duke”, the “Order of Heaven” bestowed, the honor of a lifetime, family, and clan that not everyone can have. It can be seen that the unique feature of the procession in the Tien Cong festival compared to traditional festivals, in general, is also reflected in: Objects of the procession: Traditional festivals in general often have a procession of gods (the feast of the gods) with the object of the procession being gods (objects that are not real but only real in the imagination of the people) or the Citadel, the fairies (died). However, at the Tien Cong festival, the villagers do not hold a procession of worshiping objects being the fairies, but the families themselves organize the procession of the old man, who his grandfather, grandmother, father, and mother (living people) from the family. The family went to Tien Cong Temple to give thanks to the fairies. Several processions: Each traditional festival has only one procession, but the Tien Cong festival has more than one procession, even many processions (in some years, there are more than 200 processions of Thuong’s grandfather to Tien Temple). Public – source). Therefore, the procession ceremony represents a unique culture, imbued with the cultural identity of the inhabitants of the Bach Dang estuary, the soul of the Tien Cong festival. Fourth, the dyke-building ritual symbolizes eternity.

440

H. P. Thi et al.

When the Thuong elders finished the thanksgiving ceremony for Tien Cong, the village invited two healthy Thuong elders to the front door of the temple to do the dyke construction ceremony. The villagers built a small symbolic dyke in front of the gate of the Tien Cong temple and prepared squarely cut stones for the two Thuong to build on the dyke. This ceremony is also known as the groundbreaking ceremony. After this ceremony, new families and villages are allowed to dig ditches, build dikes, and cultivated fields and gardens. For families who have died during the New Year holidays, they have to wait until this groundbreaking ceremony to dare to hold funerals and dig graves. The dyke-building ritual not only reminds descendants of their roots but also symbolizes the longevity of coastal land. Because Ha Nam Island has a terrain 2 m lower than sea level, it is surrounded by 34 km of sea dyke. From an unspoiled tidal flat, the village was encroached by the fairies through the dike to form a village to become the prosperous countryside as it is today. Therefore, the descendants of the fairies and the islanders must be responsible for preserving the sea dyke system to protect the fields, increase production, and maintain the life of Ha Nam Island forever. Fifth, the wrestling ritual symbolizes the “longevity” and “the health of the people crossing the dike and encroaching on the sea”. After the groundbreaking ceremony, the Thuong elders continued to perform the wrestling ritual (symbolic wrestling). This is a unique “wrestling” in the wrestling festivals in the North. The Thuong elders followed the rules with the usual martial arts pieces, made the gesture of wrapping the thread a few times and then hugging each other and turning one or two times, whoever was strong enough to lift the other Thuong off the ground was the winner. At the senior age, being able to enter the wrestling ring, even if it was symbolic, was already extraordinary, yet the Upper Elders still won more than lost in the arena. If the old man wins, his family and family will be glorious and have the belief that he will live a long life (longevity), give health to his children and grandchildren, and at the same time remind his children and grandchildren to exercise and keep the sea dike system, fighting storms, rising tides to protect villages, from the way of family and ancestors. Therefore, the wrestling ritual not only symbolizes “Longevity”, but also symbolizes “the health of those who encroached on the sea, established villages, and established the Ha Nam island region”. 4.3 Performing Arts of Folk Singing Dum - Different Cultural Nuances of Coastal Residents of Quang Ninh During the festival, the organization of Dum singing is an indispensable activity in the Tien Cong festival. Unmarried men and women gather in groups to sing opposite each other. The boys of each village gathered in groups to sing to the girls of another village. The content is often asked about hometown, family, profession, love… There are singing festivals, and good reciprocal groups that can sing all day, all night, and many couples have become husband and wife through this festival. Dum singing is a folk performance art form in the intangible cultural heritage system of Vietnam, quite popular in the Northern Delta. In each region and region, Dum singing has its characteristics. Dum singing in the Ha Nam island area has its nuances of coastal residents of Quang Ninh with some differences as follows:

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province)

441

About the venue for singing: If singing in other rural areas (Hai Phong, Thai Binh, etc.) is only performed on land (inland), then singing in the Tien Cong festival is held to sing both on land and boat. Steps to organize the singing: Singing Dom follows the following steps: singing hello (hello and getting acquainted), singing visit - invite (ask family background, parental love and invite to play at home), sing quiz - preaching, singing pictures (often quizzing, painting with natural phenomena, landscapes, professions…), singing love songs (expressing affection for friendship or marriage), singing out (break up). Later, depending on the situation, you can sing to invite each other to visit your house, sing to advise each other to go to school, sing to go to the army, sing to send letters, sing about shopping… In addition to following the basic steps, the Dum singing of the inhabitants of Ha Island Nam also has a unique and unique singing step, which is “hat chua”. Lyrics and voice: The song system of Dum singing is very rich, with several dozen songs for a single step (singing type). The lyrics are mainly placed in the form of hexagonal, hexagonal, or hexagonal variations. Dum singing in Thuy Nguyen (Hai Phong) often uses new lyrics, but Dum singing in Ha Nam and Quang Yen often uses old lyrics and is rarely mixed, because in the verses there are many names of lands and villages of the countryside. Ha, Nam was brought in skillfully. This shows the originality and distinct locality of Ha Nam Dum’s singing. In particular, Dum singing on Ha Nam island is also different from Dum singing on other seas by the way of singing with a real voice, sweet, warm, full of love, rustic of agricultural residents, narrow scale (approx. 3 syllables), however, the melody is somewhat tangled, the accent requires the singer’s sophistication (the main vocal range of Dum singing is the correct 5th). Therefore, Dum singing has become an intangible cultural heritage, a spiritual product with historical-cultural value that needs to be preserved, and promoted the values in the treasure of folk songs in the coastal area of Quang Ninh.

5 Meaning and Value of the Tien Cong Festival The Tien Cong festival shows the coexistence of two cultural layers: the belief that worships people - worshiping the fairies and the tradition of respecting old age (respecting the elderly) - honoring the elderly. This coexistence is a distinct custom of the people of Ha Nam island commune that is nowhere to be found. The worshiping custom of “Laughing Dragon-Horse” is an expression of belief in worshiping the sea god, with the meaning of educating the islanders’ traditions to rule the water, imbued with elements of marine culture, expressing their belief in worshiping the sea god. In addition, the products of flowers, leaves, tubers, fruits… That make up the Dragon Horse reflect the agricultural economy that has existed and developed on Ha Nam Island for nearly 600 years. The ritual of wishing long life, bowing to life, and the procession of the old man of Thuong show the beauty in community cultural activities - bloodline, to honor the principle of “drinking water, remember the source” honoring the elderly, reminding everyone to respect Respecting the elderly is the unique and delicate culture of the Vietnamese people at the sea mouth. Tien Cong festival preserves many cultural activities and unique folk games (playing swings, cock fighting, human chess, shrimp nests, singing groups…) to meet the needs

442

H. P. Thi et al.

of entertainment, and community exchanges. The community of the people, contributing to the construction of cultural life at the grassroots level, has a pervasive effect on people and tourists. The Tien Cong festival carries a historical value to the long-standing tradition of the Vietnamese people in general and the people of Ha Nam Island commune in particular. The customs and rituals in the festival have long been regulated, built in the village’s convention, which everyone must follow. It is a beautiful custom with profound educational value on family ethics, encouraging all classes to live a beautiful and useful life for the village and the whole community. These are typical cultural activities of Ha Nam Island in terms of family, clan, and village relationships that need to be preserved and promoted.

6 Preserving the Tien Cong Festival in the Current Context Over the past 300 years, the vitality of the Tien Cong festival has remained sustainable with the unique cultural features of the estuary of Bach Dang. However, in recent years, with socio-economic fluctuations and the influence of the market mechanism, some families do not follow traditional rituals, for example, a procession of life with chains. Lots, of life paintings with a portrait of Thuong, offerings with roasted pigs, foreign wine… The festival scale is getting bigger and bigger with the participation of more people: each festival season in recent years has about 300 Thuong old people., with more than 5,000 attendees. If all 300 families organize the procession of Thuong, it will be costly. The organization of a long life procession, whether or not there is a family or a family; for those Thuong who cannot organize the procession, they will feel “resentful”. Along with that, some families organize a long life procession, but because of difficult economic conditions, it creates pressure and burden for their children and grandchildren. In the Tien Cong festival in recent years, Dum singing is still held, but the participants are old men and women, no boys or girls participate, and young people are not interested in this art form. Singing Dump is rarely used and is gradually disappearing. People on the island are in danger of losing a spiritual food left by their fathers. Along with that, during the festival, many shops and restaurants appear, and the road to the Tien Cong shrine area is narrow, leading to the situation that people and tourists coming to the festival often suffer from traffic congestion. Tien Cong festival is a cultural and spiritual product, a tourist product that is attractive to locals and visitors. To do this, the Quang Yen town government has implemented communication activities in many forms such as making television films, broadcasting propaganda, distributing leaflets, and leaflets and establishing clubs. to educate pride in the cultural traditions and festivals of the homeland; raising awareness of conservation so that people can organize festivals according to traditional rituals, ensure solemnity, and safety, and preserve the unique beauty of the Tien Cong festival. In addition, to overcome traffic congestion, the locality has been opening herringbone roads, connecting residential areas with highway 5B extending from Hai Phong through Ha Nam island to the province’s National Highway 18A; The expansion of the campus and the restoration and embellishment of the Tien Cong temple relic for the festival have also been taken into account. Security and environmental sanitation are focused on making the

Tien Cong Festival (Ha Nam Island, Quang Yen Town, Quang Ninh Province)

443

festival healthy, actively contributing to the construction of new rural areas, economic development, and increasing income for local people. The measures that the Quang Yen town government is taking are very encouraging. However, the change in the Tien Cong festival in recent years has put an urgent need to pay more attention to the conservation and promotion of this unique cultural heritage value. This is not only the responsibility of those in charge of heritage management, of governments at all levels but also of the entire community in the region. This article offers the following suggestions: Firstly, preserving rituals by traditional rites: The local government propagates and mobilizes people to eliminate rituals that are not by tradition, such as not organizing the process of Thuong by cyclo or offerings. Offerings must be made from local products; Encourage the clans to organize a collective procession of the Thuong elders, each group will receive a budget of 10–20 million VND (expenses for renting costumes, decorations, facilities, etc.) directly serving the procession). Thanks to this mechanism, every old man can be brought to Tien Cong Temple to celebrate his ancestors. Second, preserve and promote the value of the Dom song: The government and local people need to rekindle the source of the Dom singing and develop it into a movement through many activities such as: establishing a club of Dom singing, creating an environment for Dum singers to exchange and practice; encouraging artisans to research, collect and re-compile old Dum songs, to impart experience to the younger generation; develop a theme on Dum singing and put it into extra-curricular teaching in some high schools (mainly applied to 2 levels of education, namely junior high school and high school) in Ha Nam island village and town. Quang Yen; select gifted children for artisans to transmit their craft, join the club of Dum singing to organize performances for school students and local people to enjoy; performing Dom singing in local cultural programs (on the occasion of festivals, cultural and social events of the district and province); sing to visitors during village tours in Ha Nam island village.

7 Conclusion As the most unique spring festival in the country about filial piety, the tradition of drinking water, remembering the source, and opening the land of ancestors, the Tien Cong festival with its unique cultural features have created a widespread attraction. It has been recognized by the Ministry of Culture, Sports and Tourism as a National Intangible Cultural Heritage in Decision No. 1852/QD-BVHTTDL dated May 8, 2017 [5]. This is both an honor, and pride, but also a reminder to each islander as well as the local government to appreciate what their ancestors left behind. At the same time, there is a project to preserve the festival, exploit the values and unique features of the Tien Cong festival, and build it into a unique cultural tourism product with a special attraction to domestic and foreign tourists. Acknowledgements. The authors wish to express their gratitude to Van Lang University, Vietnam for funding this research. We gratefully acknowledge the financial support from Van Lang University, Ho Chi Minh City, Vietnam; and also, we wish to express our appreciation to the topics coded DTDL.XH.01/19; KC.09.13/16–20 for financing this research.

444

H. P. Thi et al.

References 1. Ngo, D.D.: Tien Cong festival - a unique “people procession” festival in Ha Nam (2020). https://baoquangninh.com.vn/le-hoi-tien-cong-le-hoi-ruoc-nguoi-doc-dao-o-ha-nam2468429.html. Accessed 30 Mar 2022 2. Phan, T.H.: Textbook Typical Traditional Festivals of Quang Ninh. Publishing House Thai Nguyen University, Thai Nguyen (2018) 3. Phan, T.T.V.: Churches of the Tien Cong family in the Ha Nam area, Quang Yen (2021). http://www.baotangquangninh.vn/Chuyen-de/cac-nha-tho-dong-ho-tien-cong-o-khuha-nam-quang-yen-1003.htm. Accessed 3 Apr 2022 4. Phan, T.T.V.: Tien Cong Temple (2020). http://www.baotangquangninh.vn/Chuyen-de/mieutien-cong-950.htm. Accessed 3 Apr 2022 5. Thai, B.: The Tien Cong festival has become a national intangible cultural heritage (2017). https://baoquangninh.com.vn/le-hoi-tien-cong-duoc-tro-thanh-di-san-van-hoaphi-vat-the-quoc-gia-2341349.html. Accessed 4 Apr 2022

Author Index

A Alaponte, Betlem, 28 Ambrosini, Dario, 169 B Baldini, Massimiliano, 70 Balocco, Carla, 143 Bampa, Francesca, 205 Barbagallo, Francesca, 60 Barbieri, Ester, 3 Bargagli, Irene, 316 Bernardi, Adriana, 205 Biluka, Nalini, 28 Bitelli, Gabriele, 3 Bonora, Anna, 155 Borghini, Silvia, 70 Bortolin, Alessandro, 205 Botteon, Alessandra, 376 Braovac, Susan, 316 C Cacace, Carlo, 257 Cadelano, Gianluca, 205 Caiulo, Cristina, 195 Calandra, Sara, 292, 305, 333 Cantisani, Emma, 376 Capoccia, Giovanni, 392 Caponera, Barbara, 392 Capuani, Silvia, 266 Cardinali, Elisa, 219 Carnieletto, Laura, 205 Cartagena, Yuly Castro, 359 Casciaro, Raffaele, 127 Caselli, Giorgio, 376

Castelli, Ciro, 97 Castellini, Marta, 344, 376 Catapano, Ilaria, 257 Centauro, Irene, 292, 305 Chaumat, Gilles, 316 Ciccola, Alessandro, 266 Cochetti, Francesco, 257 Colombini, Maria Perla, 316 Colucci, Alessandro, 70 Conti, Claudia, 376 Corrado, Maria Elena, 257 Costanzo, Vincenzo, 155 Curini, Roberta, 266 Cuzman, Oana Adriana, 344 D De Carli, Michele, 205 De Carolis, Valentina, 127 De Falco, Anna, 84 De Luca, Federico, 376 de Rubeis, Tullio, 169 Del Fa, Rachele Manganelli, 376 del Hoyo-Meléndez, Julio M., 277, 401 Di Angelo, Luca, 48 Di Bilio, Lorenzo, 305 di Sipio, Eloisa, 205 Di Stefano, Paolo, 48 Dinh, Lam Nguyen, 419 Dionisi-Vici, Paolo, 219 Dore, Nicole, 257 dos Santos, Jani, 16 E Esposito Corcione, Carola, 127

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. Furferi et al. (Eds.): Florence Heri-Tech 2022, LNME, pp. 445–447, 2023. https://doi.org/10.1007/978-3-031-17594-7

446 F Fabbri, Kristian, 155 Fico, Daniela, 127 Fioravanti, Marco, 219 Fioriti, Vincenzo, 70 Fraiese, Maria Grazia, 305 Franzoni, Elisa, 3 Furferi, Rocco, 113 G Gaglio, Francesca, 84 Galatro, Matteo, 305 Garzonio, Carlo Alberto, 292, 305, 333 Gatto, Valentine, 28 Gennarelli, Gianluca, 257 George, Sony, 359 Giuliani, Francesca, 84 Governi, Lapo, 113 Gualdani, Giovanni, 97 Guardiani, Emanuele, 48 H Hai, Ninh Ngo, 433 Hoogstede, Luuk, 28 I Iwanicka, Magdalena, 277 J Javanshir, Shabnam, 205 Joseph, Edith, 359 K Khac, Dinh Luong, 419, 433 Krupska-Wolas, Paulina, 401 Kura´s, El˙zbieta, 401 L Lambertini, Alessandro, 3 Landi, Stefano, 344 Le-Goic, Gaetan, 359 Limb, Alice, 28 Liu, Yuhui, 235 Lombardo, Tiziana, 359 Longo, Sveva, 376 Lucchi, Elena, 180 Lucejko, Jeannette J., 316 Ludeno, Giovanni, 257 Luglio, Michele, 257

Author Index M Machado, Marlene, 16 Magnani, Maddalena, 235 Magrini, Donata, 376 Marcelli, Romolo, 392 Marino, Edoardo M., 60 Markevicius, Tomas, 235 Martino, Massimiliano, 84 Matosz, Marta, 277 Mattonai, Marco, 316 Mazzanti, Paola, 97, 219 Mialhe, Justine, 219 Modugno, Francesca, 316 Mogielska, Alicja, 277 Montagna, Francesco, 127 Morabito, Anna Eva, 48 N Nardelli, Pietro, 70 Nurit, Marvin, 359 O Obarzanowski, Michał, 277 Olsson, Nina, 235 Osso, Paolo, 257 P Paganin, Martina, 235 Pallara, Stefano, 195 Pallecchi, Pasquino, 305 Palumbo, Elisabetta, 127 Paoletti, Domenica, 169 Papadakis, Vassilis M., 16, 28 Pascucci, Rosella, 305 Pasqualoni, Giovanni, 169 Pavan, Martina, 70 Pecchioni, Elena, 292, 333 Picca, Alessandro, 70 Pilati, Francesco, 305 Pinzani, Luciana, 305 Pizzigatti, Cesare, 3 Pocobelli, Giorgio Franco, 376 Pompa, Giulia, 70 Postorino, Paolo, 266 Pretelli, Marco, 155 Prochazkova, Alzbeta, 28 Proietti, Emanuela, 392 Puggelli, Luca, 113 R Rescic, Silvia, 344 Ribechini, Erika, 316 Ricciardi, Luciano, 97 Riminesi, Cristiano, 344, 376 Riparbelli, Lorenzo, 97, 219

Author Index Rizzo, Daniela, 127 Roselli, Ivan, 70 Rothenhaeusler, Ulrike, 359 Ryguła, Anna, 277, 401

447 Tatì, Angelo, 70 Thi, Hue Phan, 433 Togni, Marco, 219 Trevisiol, Francesca, 3 Trovatelli, Francesco, 60 Truong, Son Nguyen, 419

S Salvadori, Barbara, 376 Salvatici, Teresa, 292, 305, 333 Santacesaria, Andrea, 97 Santarelli, Chiara, 113 Santo, Alba Patrizia, 333 Santos, Jani, 28 Sardi, Giovanni Maria, 392 Scelza, Hosea, 305 Schito, Eva, 155, 180 Schmidt-Ott, Katharina, 359 Seymour, Kate, 28 Sharma, Deepshikha, 359 Stagno, Valeria, 266

V Valentini, Federica, 305 Van, Son Quang, 419, 433 Vannuccini, Marco, 305 Vicario, Margherita, 143 Villani, Elisa, 266 Viti, Stefania, 60 Volpe, Yary, 113

T Tanganelli, Marco, 60, 344 Targowski, Piotr, 277

Z Zampognaro, Francesco, 257 Zborowska, Magdalena, 316

U Uzielli, Luca, 97, 219