Answers are grounded in this paper and cite their sources.
Finite-element modelling has been widely used to analyse the influence of damage on structural dynamic responses. Existing studies have predominantly focused on traditional isotropic materials or idealized damage geometries, while its application to complex additively manufactured structures, including three-dimensional (3D)-printed polymer composites with inherent anisotropy and geometric complexity, remains limited. Recent efforts have incorporated thermomechanical coupling and experimental validation to improve prediction accuracy, yet many models still overlook material anisotropy and geometric intricacies specific to additive manufacturing. This study systematically investigates the effects of crack length, angle and position on the dynamic response of 3D-printed acrylonitrile butadiene styrene/polylactic acid composite plates using finite-element simulation, experimental testing and machine learning. A high-fidelity parametric model was developed in COMSOL Multiphysics® under cantilever boundary conditions. The results show that increasing crack length monotonically reduces natural frequencies, with higher-order modes being more sensitive. Crack angles between 0° and 90° cause nonlinear stiffness degradation associated with the alignment between the crack and vibration stress direction, while crack-location effects depend on proximity to regions of high strain energy. Experimental validation shows trends consistent with the simulations, with errors controlled within 10%. In addition, a Deep Q-Network model was used to predict frequency changes caused by crack parameters, achieving 99.9% accuracy. The study provides a reproducible workflow for damage modelling of printed composites, clarifies the influence of crack geometry on dynamic behaviour, supports vibration-based damage assessment and integrates finite-element analysis with reinforcement learning for structural parameter regression, with theoretical and practical value.
No takes yet. Share an insight, caveat, or question.
Yang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: