Integration of Physics-based Models and In-situ Process Monitoring for Predicting Variability Associated with Liquid Composites Molding

Abstract

The focus of this paper is on the integration of physics-based models and in-situ process monitoring for predicting variability associated with liquid composites molding (LCM) manufacturing processes. LCM processes are complex and involve multiple length scales, which introduce variability and uncertainty in the processing-structures-properties-performance of composite parts, the addressment of which represents a key gap in the aerospace industry. Experiments are conducted using high-definition distributed fiber optic sensors and dielectric sensors to measure strain and resin state during LCM manufacturing processes. It is shown that these sensors can be used for in-situ process monitoring to measure resin flow arrival, temperature, viscosity, gelation, degree of cure, glass transition temperature, and strain at different layers within the composite layup. The integrated computational models consists of coupled flow-compaction by linking the commercial computational fluid dynamics (CFD) and finite element analysis (FEA) software, STAR-CCM+ and ABAQUS. The multi-scale modeling framework is implemented via UMATHT and UMAT user-subroutines in ABAQUS decoupled heat transfer and stress analyses. The objective of this study is to use in-situ process monitoring techniques to combine experimental manufacturing data with integrated computational modeling to develop the cyberinfrastructure necessary to facilitate adoption of composite manufacturing technologies in the industry and enable performance-driven manufacturing process design and future integration of artificial intelligence.