Stochastic Sensitivity Analysis of Nonlinear Crack Detection Using Quadratic Teager-Kaiser Energy

This study investigates the influence of uncertainties in materials on nonlinear crack detection, particularly in cases where responses exhibit slight nonlinearity.A stochastic sensitivity analysis approach is conducted on a novel indicator for nonlinear crack detection, known as the quadratic Teager-Kaiser energy, and its performance is compared with the harmonic indicator derived from the frequency spectrum.The method considers each input parameter as a random variable individually, as well as all input parameters collectively as random variables, by incorporating the partial sensitivity factor of each input and the coefficient of variation of all inputs.The findings indicate that, among all material parameters discussed, quadratic Teager-Kaiser energy exhibits relatively greater sensitivity than frequency spectrum responses, with elastic modulus being the most influential parameter in both cases.These results suggest that considering the uncertainties in input parameters is crucial when utilizing harmonic indicators for crack detection, especially for slight cracks.