Abstract
Using laboratory-produced replicas of cultural heritage objects to examine the benefits or possible detrimental effects of novel materials and methods before applying them to original objects is considered a fundamentally important procedure in the field of conservation. The use of replicas allows for the examination of methods and materials in a controlled environment and without risking any damage to the original cultural heritage objects. However, for these studies to be relevant, it is of utmost importance that the replicas match the chemical and physical characteristics of the original objects. This is of particular significance when working with porous objects such as wall paintings, where the physical properties, e.g., the hygric behaviour, can significantly influence the outcome of conservation procedures such as cleaning and desalination.
The main objective of the present PhD dissertation is to show the importance of using accurate plaster replicas when conducting wall painting conservation research - with a particular focus on limewashed Danish medieval wall painting plaster from the Gothic period. The studies in the dissertation are divided into three sections. PART I examines the lime-rich plaster constituting the support for Gothic wall paintings in Denmark, presents an approach for the replication of this plaster and examines the replicas’ microstructural resemblance to the medieval limewashed plaster (Papers I-II). Part II examines the replicas’ hygric resemblance to medieval plaster and the influence of binder to aggregate ratio as well as type of limewash on hygric properties (Paper III). Part III demonstrates how the use of replicas with a low microstructural resemblance to the original plaster can result in inaccurate predictions of the outcome of conservation procedures (Paper IV).
PART I: With the use of different analytical techniques such as polarisation microscopy, x-ray diffraction analysis, mercury intrusion porosimetry and scanning electron microscopy, the studies show that Danish medieval plaster has a compact structure consisting of a non-hydraulic lime binder and aggregates with an average ratio of 2:1, often with an abundance of lime inclusions. It has a high open porosity with a unimodal pore size distribution in the range of 0.4–5 μm and an airy lime matrix consisting of small calcite crystals. By comparing these analytical results with analyses of plaster produced from lime putty plaster and hot-mixed mortar, and supplementing information gleaned from these comparisons with historical documents, recent research studies and practical considerations, the studies of Danish medieval plaster also revealed that this plaster, with a high probability, was produced from hot-mixed mortar and not from lime putty mortar as commonly presumed.
Replicas of different compositions were assessed for their resemblance to medieval plaster in terms of microporosity and hygric properties. This comparison established that it is possible to replicate the microstructure and the hygric properties of limewashed Gothic plaster to an acceptable degree. The best match to the medieval plaster was achieved in replicas produced from a hot-mixed mortar consisting of one part quicklime to one part aggregate. The best match to the Gothic limewash, characterised by its great thickness, was achieved by a limewash produced from a more than 67-year-old lime putty.
PART II: Measurements of water absorption rate by the contact sponge method revealed that while 1:1 hot-mixed replicas were a good match to medieval plaster in terms of hygric behaviour, replicas made from a mortar mix of one part lime putty to three parts aggregate were, on the contrary, a poor match. The studies showed that replicas made from this often-used mortar composition (for wall painting and built heritage studies) have little resemblance to that of lime-rich medieval plaster, as they absorb water at a four to five times slower rate than medieval plaster. This exemplifies why the use of inaccurate replicas can be very problematic. The predictability of the outcome of a test procedure involving wet cleaning, for example, will be very skewed if using replicas with a lower lime content than the original plaster. A more lime-rich plaster will absorb substantially greater amounts of liquid in the same given time period than less lime-rich plaster, thus potentially transporting a larger volume of cleaning agents into the plaster than expected.
PART III. The importance of using accurate replicas in terms of compatible microstructure and hygric properties is further supported in Part III by a study examining five clay poultices’ suitability as sorbents applied to medieval wall painting plaster. The study compares the pore size distribution of the medieval plaster with that of the wet/drying clay poultices (measured by pressure plate) and measures migration of liquid from the clay poultices into the plaster, documented with the help of blue dye. The immense importance of the compatibility between the medieval plaster and the replicas in terms of pore size distribution and hygric behaviour was shown. This study, together with additional migration studies presented in the dissertation, once again demonstrate that using replicas with low resemblance to the original plaster involves the risk of drawing erroneous conclusions in conservation studies. Such studies will thereby lack relevance due to their imprecise predictions or, at worst, can result in the implementation of harmful conservation treatments.
Although the studies showed that it is possible to replicate the microstructure and the hygric properties of Gothic plaster to an acceptable degree, and that the best-suited replicas, confirmed by the experiments on hygric behaviour, could be considered reliable stand-ins for medieval plaster, differences were also detected in the microstructure of the best-suited replicas and the medieval plaster. Therefore, the dissertation also provides suggestions for future replica productions, aiming to improve the match to medieval plaster. These suggestions embody the accumulated knowledge obtained in the course of this PhD and are based on practical experience in replica production, analyses of samples and numerous literature studies. Among the most significant suggestions for improvements are the following: using a larger volume of lime, using lump quicklime instead of powdered quicklime, changing the hot-mixing technique from wet-slake to sand-slake and prioritising a careful selection of the substrate for the replicas (to ensure correct hygric behaviour).
While the studies in the PhD dissertation set out to examine Danish Gothic wall painting plaster, its replication and the significance of using reliable replicas for wall painting conservation studies, they also inspired new ideas for future studies. For example, the studies on the hygric properties of plaster revealed a need for a study on the repair mortar used for lime-rich plaster in wall painting and built heritage conservation. Especially in light of the fact that the studies in this dissertation clearly showed that the use of 1:3 mortar, which is the most common mortar composition used for repair mortar in wall painting conservation in Denmark, does not comply with the principal tenets of modern conservation regarding compatibility and the use of like-for-like materials for repair and reconstruction work when dealing with lime-rich medieval plaster. Hence, the use of this mortar for repair work on Danish Gothic plaster should be re-evaluated.
The main objective of the present PhD dissertation is to show the importance of using accurate plaster replicas when conducting wall painting conservation research - with a particular focus on limewashed Danish medieval wall painting plaster from the Gothic period. The studies in the dissertation are divided into three sections. PART I examines the lime-rich plaster constituting the support for Gothic wall paintings in Denmark, presents an approach for the replication of this plaster and examines the replicas’ microstructural resemblance to the medieval limewashed plaster (Papers I-II). Part II examines the replicas’ hygric resemblance to medieval plaster and the influence of binder to aggregate ratio as well as type of limewash on hygric properties (Paper III). Part III demonstrates how the use of replicas with a low microstructural resemblance to the original plaster can result in inaccurate predictions of the outcome of conservation procedures (Paper IV).
PART I: With the use of different analytical techniques such as polarisation microscopy, x-ray diffraction analysis, mercury intrusion porosimetry and scanning electron microscopy, the studies show that Danish medieval plaster has a compact structure consisting of a non-hydraulic lime binder and aggregates with an average ratio of 2:1, often with an abundance of lime inclusions. It has a high open porosity with a unimodal pore size distribution in the range of 0.4–5 μm and an airy lime matrix consisting of small calcite crystals. By comparing these analytical results with analyses of plaster produced from lime putty plaster and hot-mixed mortar, and supplementing information gleaned from these comparisons with historical documents, recent research studies and practical considerations, the studies of Danish medieval plaster also revealed that this plaster, with a high probability, was produced from hot-mixed mortar and not from lime putty mortar as commonly presumed.
Replicas of different compositions were assessed for their resemblance to medieval plaster in terms of microporosity and hygric properties. This comparison established that it is possible to replicate the microstructure and the hygric properties of limewashed Gothic plaster to an acceptable degree. The best match to the medieval plaster was achieved in replicas produced from a hot-mixed mortar consisting of one part quicklime to one part aggregate. The best match to the Gothic limewash, characterised by its great thickness, was achieved by a limewash produced from a more than 67-year-old lime putty.
PART II: Measurements of water absorption rate by the contact sponge method revealed that while 1:1 hot-mixed replicas were a good match to medieval plaster in terms of hygric behaviour, replicas made from a mortar mix of one part lime putty to three parts aggregate were, on the contrary, a poor match. The studies showed that replicas made from this often-used mortar composition (for wall painting and built heritage studies) have little resemblance to that of lime-rich medieval plaster, as they absorb water at a four to five times slower rate than medieval plaster. This exemplifies why the use of inaccurate replicas can be very problematic. The predictability of the outcome of a test procedure involving wet cleaning, for example, will be very skewed if using replicas with a lower lime content than the original plaster. A more lime-rich plaster will absorb substantially greater amounts of liquid in the same given time period than less lime-rich plaster, thus potentially transporting a larger volume of cleaning agents into the plaster than expected.
PART III. The importance of using accurate replicas in terms of compatible microstructure and hygric properties is further supported in Part III by a study examining five clay poultices’ suitability as sorbents applied to medieval wall painting plaster. The study compares the pore size distribution of the medieval plaster with that of the wet/drying clay poultices (measured by pressure plate) and measures migration of liquid from the clay poultices into the plaster, documented with the help of blue dye. The immense importance of the compatibility between the medieval plaster and the replicas in terms of pore size distribution and hygric behaviour was shown. This study, together with additional migration studies presented in the dissertation, once again demonstrate that using replicas with low resemblance to the original plaster involves the risk of drawing erroneous conclusions in conservation studies. Such studies will thereby lack relevance due to their imprecise predictions or, at worst, can result in the implementation of harmful conservation treatments.
Although the studies showed that it is possible to replicate the microstructure and the hygric properties of Gothic plaster to an acceptable degree, and that the best-suited replicas, confirmed by the experiments on hygric behaviour, could be considered reliable stand-ins for medieval plaster, differences were also detected in the microstructure of the best-suited replicas and the medieval plaster. Therefore, the dissertation also provides suggestions for future replica productions, aiming to improve the match to medieval plaster. These suggestions embody the accumulated knowledge obtained in the course of this PhD and are based on practical experience in replica production, analyses of samples and numerous literature studies. Among the most significant suggestions for improvements are the following: using a larger volume of lime, using lump quicklime instead of powdered quicklime, changing the hot-mixing technique from wet-slake to sand-slake and prioritising a careful selection of the substrate for the replicas (to ensure correct hygric behaviour).
While the studies in the PhD dissertation set out to examine Danish Gothic wall painting plaster, its replication and the significance of using reliable replicas for wall painting conservation studies, they also inspired new ideas for future studies. For example, the studies on the hygric properties of plaster revealed a need for a study on the repair mortar used for lime-rich plaster in wall painting and built heritage conservation. Especially in light of the fact that the studies in this dissertation clearly showed that the use of 1:3 mortar, which is the most common mortar composition used for repair mortar in wall painting conservation in Denmark, does not comply with the principal tenets of modern conservation regarding compatibility and the use of like-for-like materials for repair and reconstruction work when dealing with lime-rich medieval plaster. Hence, the use of this mortar for repair work on Danish Gothic plaster should be re-evaluated.
Original language | English |
---|
Place of Publication | Copenhagen |
---|---|
Publisher | Royal Danish Academy - PhD School, Royal Danish Academy - Architecture, Design, Conservation Institute of Conservation |
Number of pages | 265 |
Publication status | Published - 31 May 2023 |
Note re. dissertation
PhD defends 16.10.2023.Assessment committee:
Kristin Balksten. Docent, Tech. Dr., Kulturvård – Campus Gotland, Uppsala University, Campus Gotland, Visby, SwedenBarbara Lubelli, Dr. Associate Professor, University of Technology, Delft, The NetherlandsMikkel Scharff, Associate Professor, Royal Danish Academy - Institute of Conservation, Copenhagen, Denmark (Chair of the Assessment Committee).
Principal Supervisor:
Jane Richter, Associate Professor, Royal Danish Academy - Institute of Conservation, Copenhagen, Denmark (Chairperson of the defence).
External supervisor:
Isabelle Brajer, Senior researcher National Museum, Denmark.
Isabelle Brajer, Senior researcher National Museum, Denmark.
Funding
The publication has been made possible by financial support from the Augustinus Foundation (Grant no. 16-4313)