Hyperelastic material modelling using symbolic regression

Recently, data-driven approaches in the field of material modeling have gained significant attention. A major advantage these is direct integration experimental results into models. Nevertheless, artificial neural networks (ANNs) are especially challenging to interpret from a physical point view, since internal processes ANNs difficult understand. In this work new automatic method for generation constitutive models hyperelastic materials introduced. The presented based on symbolic regression, which genetic algorithm. Thereby, mathematical model form an algebraic expression found that fits given data as accurately possible and has compact representation. strain energy density function determined directly invariants. proposed ansatz embedded continuum mechanical framework combining benefits known relations with unbiased optimization approach regression. Benchmark tests generalized Mooney-Rivlin uniaxial, equibiaxial pure shear presented. Finally, procedure tested temperature-dependent dataset thermoplastic polyester elastomer. good agreement between obtained demonstrated.