Neural network based constitutive model for elastomeric foams

Elastomeric foam materials find wide applications for their excellent energy absorption properties. The mechanical property of elastomeric foams is highly nonlinear and it is essential to implement mathematical constitutive models capable of accurate representation of the stress–strain responses of foams. A novel constitutive modeling method of defining hyperfoam strain energy function by a neural network is presented in this work. The architecture of the artificial neural network is described. The calculation of the strain energy and its derivatives by neural network is explained in detail. The preparation of the neural network training data from foam test data is described. Curve fitting results are given to show the effectiveness and accuracy of the neural network approach. A neural network based elastomeric foam constitutive model is implemented in simulation of a plane-strain foam indentation process to demonstrate the application and efficiency of the neural network approach in finite element analysis. Results indicate that the neural network model provides a better representation of the test data than the commonly used Hyperfoam model.