A multi-analysis framework for uncertainty quantification and data-driven simulation of design allowables in laminated composites

This study combines various computational methods informed by statistical data from experimental Compact Tension (CT) testing to simulate progressive damage evolution in composites, including their uncertainty. First, efficient Finite Element Analysis (FEA) is employed to generate large datasets. Second, a global sensitivity analysis identifies influential FEA input parameters. Third, a simplified surrogate model based on polynomial regression is developed. This rapid model enables the coupling with Bayesian parameter estimation in form of Markov Chain Monte Carlo to determine the probability distribution of the FEA input parameters, thereby representing the measured uncertainty in CT tests. The capability of this multi-analysis framework is demonstrated through the simulation of A- and B-Basis design allowables in various open-hole tension tests. The results indicate that B-Basis values are consistent between experiments and simulations, with differences ranging from 2% and 9%. Conversely, the calculations of A-Basis allowables vary significantly due to the enhanced data requirements for their correct calculation.