Composition optimization of a high-performance epoxy resin based on molecular dynamics and machine learning

• Main Authors: Kai Jin, Hao Luo, Ziyu Wang, Hao Wang, Jie Tao
• Format: Article
• Language: English
• Published: Elsevier 2020-09-01
• Series: Materials & Design
• Subjects: Epoxy resin; Composition optimization; Molecular dynamics simulation; Machine learning; Neural network
• Online Access: http://www.sciencedirect.com/science/article/pii/S0264127520304664

Epoxy resin is a general term for a class of thermosetting polymers containing two or more epoxy groups in the molecule and has an excellent comprehensive performance. The properties of the resin system vary greatly due to the different compositions of the base resin, curing agent, and toughening agent. In this study, an optimization method for the multi-component epoxy resin system was put forward by using molecular dynamics simulations and machine learning methods. An optimized high- performance epoxy resin system considered Young's modulus (E), Ultimate Tensile Strength (UTS), Elongation (δ), and the glass transition temperature (Tg) together was designed by using the proposed method. The influence of each component proportion on mechanical properties can also be obtained automatically. It was found that 4,4′-Diaminodiphenyl Sulfone (DDS) was a better curing agent to improve Tg, E, and δ, compared with Dicyandiamide (DICY). Tetraglycidyl Diamino Diphenylmethane (TGDDM) could ensure high Tg, E and UTS, but the system still needed some Diglycidyl Ether of Bisphenol A (DGEBA) to improve toughness. The toughening agent Polyether Sulfone (PES) improved the toughness of the epoxy resin system significantly. The presented method could be extended to other resin system composition optimization.