Tailoring Failure Interactions for Validation Testing of Composite Laminate Progressive Failure Models

Efficient validation of numerical models for predicting failures in composite structures is imperative due to the intricate nature of such failures. The multifaceted aspects, including multimodal, multiscale, and interactive factors, pose challenges in the calibration and validation of models. Constraints in cost and time often limit the comprehensive testing required for model validation across diverse failure response scenarios. This study proposes an optimal test selection approach, building upon previous insights that optimizing measures derived from progressive failure responses can unveil failure mode interactions sensitive to stochastic variations. Calibration designs should minimize interaction while maximizing sensitivity to calibration parameters, whereas validation models must encompass varied interactions to understand predictability limits. However, optimizing progressive failure analyses for composites is computationally intensive. A straightforward design of experiments exploration is demonstrated here, focusing on efficiently identifying designs sensitive to stochastic effects, which reveal mode interactions and suitability for calibration/validation. This method is exemplified using computational models for composite bolted joints subjected to pin-bearing loads, showcasing their intricate failure modes. Designs displaying high sensitivity to parameter changes offer insights into failure mechanisms, aiding in model validation and enhancement. Thus, employing such critical designs as test cases can refine numerical models, resulting in more accurate representations of complex failure behaviors in diverse composite components.