A Methodological Approach to Interpreting Deformations in Panel Paintings Using Reduced-Order Hygro-Mechanical Modeling

Abstract

Panel paintings respond to environmental change in ways that depend on the individual object. Environmental monitoring is routine, but climatic standards do not provide a quantitative, object-specific criterion for deciding whether an observed deformation is mechanically significant. We present an open, modular pipeline—a sequence of independent, replaceable modules in which the output of each feeds the next—that converts environmental and deformometric records into a conservation-relevant quantity: strain at the painted surface. The workflow comprises four modules: calculation of cupping from deformometric-kit data, estimation of equilibrium moisture content from temperature and relative humidity, local-polynomial pre-processing, and a reduced-order hygro-mechanical model. The method is applied to three poplar panel paintings from the Opificio delle Pietre Dure exposed to the same conservation environment. The model reproduces measured cupping with training R2 values of 0.90–0.95. Expressing the response as painted-surface strain reveals object-specific differences; under a conservative worst-case reference, all three paintings remain below the lower bound of the cracking-strain range reported for gesso grounds, although with different margins. Crucially, these strain levels cannot be inferred from cupping amplitude alone because the cupping-to-strain relationship varies among panels. The pipeline is not intended to replace finite-element analysis, but to provide a transparent, reproducible framework for turning routine monitoring into a physically interpretable screening metric.