Dynamic characterization of lightweight corrugated FRP tubes: Empirical modeling and probabilistic reliability assessment

: This study presents empirical formulas derived from finite element (FE) simulations for the dynamic characterization of lightweight corrugated fibre-reinforced polymer (FRP) tubes, addressing the lack of reliable predictive tools for resonance-prone corrugated structures. Four surrogate models for predicting the first natural frequency are developed, among which the FE-calibrated analytical model and symbolic regression achieve high predictive accuracy ( ). The framework incorporates corrugation geometry and composite lamina architecture within a probabilistic design setting. Pareto trade-off analysis identifies lightweight woven composites as optimal configurations balancing mass efficiency and natural frequency performance. Global Sobol sensitivity analysis and Monte Carlo simulations reveal that convolution count, layer number, and crest radius dominate natural frequency variability, with a pronounced nonlinear interaction between convolution count and layer number governing stiffness-mass coupling effects. In contrast, trough radius, bellows outer radius, and fiber orientation exhibit marginal impact on Sobol sensitivity indices and natural frequency. These findings demonstrate that corrugation geometry exerts a stronger influence on dynamic response uncertainty than further composite tailoring. The proposed formulations provide quantitative guidance for the probabilistic design of resonance-tolerant corrugated FRP tubes, supporting resilient infrastructures in pipeline and aerospace applications.