New publication in Advances and Challenges in Computational Mechanics

2 Januar 2026

We are happy to announce the publication of the article "Physics-Informed Neural Networks for Solving Contact Problems in Three Dimensions" in the special issue Advances and Challenges in Computational Mechanics.

In this paper we explore the application of physics-informed neural networks (PINNs) to tackle forward problems in 3D contact mechanics, focusing on small deformation elasticity. We consider two examples: A patch test, where the contact surface is known a-priori and the Hertzian contact of a cylinder where the contact area is determined during the solution process. Adopting a Hellinger-Reissner-like two-field formulation for the PINN as well as a problem-specific output transformation layer, our models are able to learn the governing physics, resulting in full-field predictions for displacements and stresses that are in good agreement with the analytical solutions.

Sahin, T., Wolff, D., & Popp, A. (2025). Physics-Informed Neural Networks for Solving Contact Problems in Three Dimensions. In Advances and Challenges in Computational Mechanics (pp. 419–431). Springer Nature Switzerland. :10.1007/978-3-031-93213-7_33