Can I read Towards Machine Learning of Crystal Plasticity by Neural Networks on EtoBox?

Towards Machine Learning of Crystal Plasticity by Neural Networks by Christoph Hartmann is a book available to read on EtoBox.

What is Towards Machine Learning of Crystal Plasticity by Neural Networks about?

The use of crystal plasticity models in macroscopic numerical analysis still poses challenges. Component and crystal structures have different scales by several orders of magnitude. For this reason, a discretization of the crystal structure within the framework of the component is not reasonably possible. Therefore, for each integration point its microstructure is represented by a representative volume element. Due to the computational effort associated with this coupling, the use of crystal plasticity models has been limited. They are mainly used in academia and with little application in component and process design. The approach taken here is to decouple computational effort through machine learning by training a neural network that eventually serves as a material model in macro-scale analysis. Based on the deformation gradient and the microstructure, the trained neural network reproduces the resulting stress response, where the conducted investigations also cover consecutive deformation patterns and general stress states.

Author

Christoph Hartmann

Publisher

Springer Nature Switzerland : Imprint: Springer

Published

2023

Language

EN

ISBN

9783031225239

Subjects

Science, Chemistry, Engineering