Ultrasonic evaluation of CFRP bonding quality using stiffness and tensile prediction

This study aims to develop a non-destructive evaluation framework for bonding quality in carbon fiber-reinforced polymer (CFRP) adhesive joints by integrating ultrasonic transmission spectroscopy with finite element modeling (FEM), addressing the critical gap in detecting interfacial weakening in anisotropic composite structures.

Ultrasonic transmission coefficients were measured using a water immersion system for CFRP specimens with controlled interfacial roughness (600# to 60# grit). A hybrid Particle Swarm Optimization-based Simulated Annealing (PSO-b-SA) algorithm was developed to inversely determine interfacial tangential stiffness (K_T) by minimizing discrepancies between experimental and theoretical spectra. A cohesive zone model (CZM)-based FEM incorporating inverted K_T and experimentally measured mode-I/II fracture energies (G_Ic, G_IIc) was established to predict tensile strength.

The transmission coefficient spectrum shifted toward lower frequencies as interfacial roughness increased, correlating with a 92 % reduction in K_T (from 1.77 × 10¹⁴ to 1.26 × 10¹³ N/m³). The FEM predictions aligned with experimental tensile strengths within 7 % error, demonstrating robustness across varying bonding conditions. The PSO-b-SA algorithm achieved >98 % correlation between theoretical and measured spectra, outperforming traditional single-parameter optimization methods in convergence speed and stability. This work provides a systematic, non-destructive framework for evaluating CFRP bonding quality by linking ultrasonic metrics to mechanical performance.