Predicting Stress-Strain Behavior of Thermoplastic Elastomer by Theoretical Calculation and Deep Learning

Thermoplastic elastomer (TPE) is a typical industrial product where the microphase separation of block copolymer is utilized. This elastic behavior is one of the examples where the phase separated structure affects the physical properties. However, it is not simple to find the relation between complicate phase separated structure and stress-strain behavior. To tackle the problem, we applied coarse-grained simulation and deep learning technique. Stress-strain curve of various phase separated structures of ABA type triblock copolymers, where A blocks and B blocks form glassy domain and rubbery domain respectively, are investigated by the collaborative simulation of self-consistent field theory and coarse-grained molecular dynamics. Furthermore, we applied deep learning approach to make regression between the phase separated structure and stress-strain (S-S) behavior obtained by the computational simulation. The prediction of S-S behavior using trained deep learning network showed reasonably good results, and was very fast comparing to the computationally intensive simulation.