TRIUMPH - Research project

Dry friction is at the heart of several technological and societal concerns: a large amount of energy is needed to overcome it (24% of the energy produced globally), and it produces wear particles which have an impact on human health. In order to improve our understanding and our control of mechanical contacts, the TRIUMPH project will be devoted to the construction of two parallel and independent tools for the prediction of friction based on surface morphologies acquired by microscopic imaging (SEM). The first one will be purely empirical (interpretable Deep Learning model trained on a database of experimental images), and the second one will be physical (dialog between experimental images and particle-based simulations though topographic descriptors). The combination of both approaches will make it possible to improve both our understanding and our prediction of dry friction.