New preprint on physics-informed neural networks for contact mechanics

1 Dezember 2023

In this new preprint, we explore the ability of physics-informed neural networks (PINNs) to solve forward and inverse problems of contact mechanics for small deformation elasticity. We deploy PINNs in a mixed-variable formulation enhanced by output transformation to enforce Dirichlet and Neumann boundary conditions as hard enforced constraints. We compare three methods to enforce Karush-Kuhn-Tucker type inequality constraints of contact problems as soft constraints: sign-based method, Sigmoid-based method, and Fischer-Burmeister nonlinear complementarity problem function.

We demonstrate the novel application of PINNs for contact mechanics including benchmark examples, e.g. contact between an elastic block and a rigid flat surface, as well as the Hertzian contact problem.

Sahin, T., von Danwitz, M., Popp, A. (2023): Solving Forward and Inverse Problems of Contact Mechanics using Physics-Informed Neural Networks, Preprint, submitted for publication, https://arxiv.org/abs/2308.12716