“…Blair and Annis [37] used the Choleski decomposition to simulate blade thickness variations, assuming that the thickness variation is a homogeneous and isotropic field. Other researchers, including Ghiocel [38,39], Griffiths and Tschopp [40], Cassenti [41], and Brown and Grandhi [35], have used the KL expansion (or POD, PCA) to simulate more complex stochastic blade thickness variations due to manufacturing. Ghiocel [36] applied the KL expansion, both in the original, nonGaussian stochastic space and transformed Gaussian space.…”