We propose a method for automizing the discovery process of hyperelastic material models from experimental data. To this end, we introduce a large set of physically admissible candidate material models and employ sparse regression to automatically select a model that can describe the given data and is at the same time expressed by a concise mathematical expression, increasing the physical interpretability of the model. The method is validated by applying it to data generated through mechanical testing of human brain tissue under uniaxial tension and compression as well as simple torsion.

Download the slides for this talk.Download ( PDF, 1196.42 MB)