124 G.N. Stevens et al. Fig. 10.23 Standard deviation of stochastic thickness estimates Fig. 10.24 Lower (left) and upper (right) level bounds of stochastic thickness estimate when considering predictions within one standard deviation of the posterior mean through non-destructive techniques. Many existing methods, however, are unable to adequately capture the damage due to insensitivity of global response metrics to small, localized damage. Acoustic Wavenumber Spectroscopy (AWS) has been shown to be able to detect these local damages because of the local nature of measurements and high sensitivity of wavenumber to damage. As such, AWS offers a real potential to prevent failure of a system due to continued undetected damage. The main drawback is that AWS currently is unable to quantify the level of damage in a structure. This paper introduced a method for incorporating parametric and experimental uncertainty in AWS inverse problems through the use of Stochastic Wavenumber Estimation. Inverse analysis applied to AWS measurements can not only detect, but also quantify the severity of local damage in thin walled structures. Uncertainties, however, pose a problem for the method when there exists uncertainties in material properties or spurious peaks in experimental data. The methodology has been demonstrated herein through a case study in which thinning of an aluminum plate was detected and its severity quantified. The proposed approach successfully mitigates the degrading effects of parametric uncertainty, demonstrating the ability to
RkJQdWJsaXNoZXIy MTMzNzEzMQ==