High-throughput prediction of stress–strain curves of thermoplastic elastomer model block copolymers by combining hierarchical simulation and deep learning

Abstract We achieved high-throughput prediction of the stress–strain (S–S) curves thermoplastic elastomers by combining hierarchical simulation and deep learning. ABA triblock copolymer with a phase-separated structure was used as elastomer model. The S–S copolymers were calculated from self-consistent field theory calculations coarse-grained molecular dynamics simulations. Because such simulations require considerable computational resources, we applied learning technique to accelerate prediction. Sets structures obtained train 3D convolutional neural network. Using trained network, confirmed that predicted untrained accurately reproduced results. These results will enable us design novel polymers desired screening wide variety structures. Graphic abstract