Little is known about the intrinsic instability of oil paintings. It can become manifest by visual damage, a possible warning sign for more dramatic deterioration in the future. Damage such a protrusions or discolorations is often caused by the formation of metal soaps, compounds of metals with fatty acids. Oil paintings are essentially thermodynamically unstable systems developing towards more stable, but aesthetically less attractive, states. Analyzed paint samples prove that metal is displaced across the paint film in this process, but the exact role of the linseed oil networks used as binding medium is unclear. This project aims to better understand the behavior of metal in linseed oil networks by gaining more information about the microstructure of the latter. It will use multi-scale mathematical modeling to predict this microstructure in the various states of an oil painting, from the process of applying the paint through the drying stage until highly degraded conditions. The models will answer questions about the extent to which higher temperatures and relative humidity and the ranges of fluctuations in both promote degradation of pigments in oil paintings. This information is crucial in making informed decisions about safe display and storage conditions.