A Conference with Colin Bonatti on ‘Modeling history-dependent material behavior using Recurrent Neural Networks’
Colin Bonatti, post-doctoral fellow with the Chair of Artificial Intelligence in Mechanics and Manufacturing, MAVT Department at ETH Zurich, will give a conference on Thursday 7 April 2022 at 2 pm at Centrale Nantes.
On April 7, 2022 from 14:00 To 16:00
Dr. Colin Bonatti from ETH Zurich will be on campus at the invitation of the Research Institute in Civil and Mechanical Engineering (GeM).
Recent years have seen the development of various Machine-Learning based approaches to material modeling. Among them, Recurrent Neural Networks (RNNs) hold a particular place due to their ability to treat sequential data. Notably, they can construct their own internal state-variables in order to reproduce history-dependent material behavior based purely on stress-strain sequences.
Most of the literature on RNN-based mechanical models relies on established RNN architectures called LSTMs and GRUs. In this presentation, we will show that modifying the RNN architecture can provide several advantages. The proposed architectures provide compact models, can reproduce the state-space of phenomenological models, and are usable in complex explicit finite element simulations.
Due to the combined large data requirements of RNNs and technical difficulties in leveraging experimental results, we will focus on stress-strain sequences drawn from numerical examples. We will detail applications to phenomenological models, crystal plasticity and structural simulations
The conference will be held in Lecture Theatre A at Centrale Nantes.
Conference presentation:
Recent years have seen the development of various Machine-Learning based approaches to material modeling. Among them, Recurrent Neural Networks (RNNs) hold a particular place due to their ability to treat sequential data. Notably, they can construct their own internal state-variables in order to reproduce history-dependent material behavior based purely on stress-strain sequences.Most of the literature on RNN-based mechanical models relies on established RNN architectures called LSTMs and GRUs. In this presentation, we will show that modifying the RNN architecture can provide several advantages. The proposed architectures provide compact models, can reproduce the state-space of phenomenological models, and are usable in complex explicit finite element simulations.
Due to the combined large data requirements of RNNs and technical difficulties in leveraging experimental results, we will focus on stress-strain sequences drawn from numerical examples. We will detail applications to phenomenological models, crystal plasticity and structural simulations
The conference will be held in Lecture Theatre A at Centrale Nantes.
