Nanomechanical Properties of Paint from Authentic Paintings

Presenter:
Cecil Krarup Andersen: The Royal Danish Academy

Abstract:
Computational models are becoming an increasingly larger part of predictive analysis and preventive conservation for cultural heritage. The Copenhagen painting model (Lee et al 2022) showed how a relatively stiff lower ground layer can help avoid cracks and delamination. However, accurate model predictions require accurate material property data and so far, predictive models have relied on data from model paint tests. While these represent an invaluable source it is unclear to what degree they represent the range of mechanical properties that can be expected from older paints in historic paintings. The analysis of paint from authentic paintings represents a challenge since samples cannot be tensile tested and are less homogeneous than paint samples produced for research. This work presents nanomechanical tests on samples from paintings up to 157 years old to test the hypothesis that the mechanical properties of these paints could be related to the paint samples cast for research purposes.

The results showed large variability in storage modulus both in the same sample indicating poor homogeneity as well as between the original paint samples. A lower ground layer with earth colour and silicates from F.T. Kloss “A frigate in a bay” 1865 was expected to have relatively low modulus based on tensile tests on oil paint samples from the Mecklenburg’s Paint Reference Collection. Surprisingly, the ground layer was shown to have a higher modulus than almost all other paints analysed, including the lead white upper ground layer on the same painting. The results suggest that, apart from pigments, fillers, and oils, the presence of other organic compounds e.g. protein-based additives (Di Gianvincenzo et al 2023) which are difficult to detect may help explain unexpected mechanical properties.

This research represents a first step into a new realm of “real” paint properties and stressed the complexity of the challenge. Since tensile tests cannot be used on the small paint samples, nanoindentation on cross sections proved to be a most valuable tool for understanding the properties of paints removed from paintings. Ultimately, this information can be used to update current and future computational models of paintings for more accurate predictions.

References
Lee, D.-H., et al (2022). Numerical modelling of mechanical degradation of canvas paintings under desiccation. Heritage Science, 10(1), 1-13.
Di Gianvincenzo, et al. (2023). Proteomic identification of beer brewing products in the ground layer of Danish Golden Age paintings. Science Advances, 9(21), eade7686

This presentation was part of the Mechanical Insights: Shaping the Future of Museum Collection Symposium. View details and the full symposium program here: https://www.getty.edu/projects/mechan...

Part of Getty Conservation Institute’s Managing Collection Environment Initiative: https://www.getty.edu/projects/managi...