Microscopic residual stresses analysis and multi-objective optimization for 3D woven composites

Abstract

To balance the conflict between low cost and high quality of 3D woven composites (3DWC), we develop a multi-objective optimization methodology. A multiscale process simulation model is proposed to calculate microscopic residual stresses in 3DWC, and the model is validated by experimental measurement using fiber Bragg grating sensors. To address the enormous computational amount, a surrogate model is constructed through three sequential sampling algorithms called Monte Carlo based space reduction, cross-validation Voronoi and greedy algorithm. An interface is developed to integrate the process simulation, sequential sampling and optimization. The multi-objective optimization is conducted to minimize residual stresses and process time utilizing a genetic algorithm, and the Pareto front is obtained. The global sensitivity analysis is conducted based on the surrogate model, and the key parameters are identified. The results show that the effects of process parameters on objectives are coupled. The overall cure cycle should be considered in process optimization.