Engineering Software Research & Development

Traditional pointwise stress assessments often prove inadequate near composite-material interfaces, where steep gradients and singularities complicate the interpretation of finite element results. This technical note explains why nonlocal, volume-averaged failure criteria are better suited for correlating experimental observations with numerical predictions. By reducing sensitivity to mesh-dependent local stress fields, these criteria provide a more robust foundation for developing reliable, decision-grade predictive models.

Reliable strength prediction in laminated composites requires the simultaneous control of discretization and model-form errors. Conventional plate and shell formulations cannot generally provide this control. The hp-version of the finite element method offers a practical solution.

"From Legacy Finite Element Modeling to Explainable Simulation: Technical Requirements for XAI in Computational Mechanics", a technical paper authored by ESRD Co-founder and Chairman Dr. Barna Szabó, was recently published in Computational Mechanics and is available to all as an open-access article. This paper discusses technical requirements posed by Explainable AI (XAI), methods for controlling both model-form and discretization errors in AI-assisted simulation, and model-centric vs. element-centric implementations.