Multidimensional nonstationary Gaidai reliability assessment for UIKKU tanker operational safety, by novel 4-parameter Weibull estimator

This benchmark case study introduces recently developed state-of-the-art general purpose multi-modal nonstationary reliability assessment method for marine, offshore, and naval structures, suitable for physically measured multidimensional dynamic systems. Existing reliability methodologies predominantly focus on univariate or bivariate systems, frequently failing to adequately address the interdependencies among the key components of high dimensional systems. The principal research aim of this case study was to illustrate the technical potential and adaptability of the proposed recently developed multimodal reliability approach through its application to the onboard dynamic measurements collected during Arctic operation of Finnish UIKKU oil/chemical tanker. The hull of the UIKKU tanker encountered ice loads throughout its polar expedition, making it pertinent to evaluate potential risks of structural hull damage. The UIKKU's vessel hull was outfitted to monitor ice-induced strains and loadings in real-time at multiple crucial (hot-spot) he hull plating locations. The overall structural behavior of the vessel during intermittent collisions with surrounding ice ridges was examined by measuring longitudinal bending stresses and vertical accelerations. This case study's findings may enhance the International Association of Classification Societies (IACS) standards and standardization for commercial polar vessels. Advocated multivariate reliability concept is aimed at aiding vessel captains in case reliability factors are analyzed in real time using live streaming sensor data, according to DNV requirements. Novelty of this case study lies in cross-validation of the proposed multidimensional reliability concept, employing comprehensive unfiltered onboard sensor-measured data as input, combined with a novel 4-parameter Weibull maximum likelihood estimator with semi-analytical model parameter solution. The proposed generic multidimensional structural reliability concept facilitates precise design prognostics for nonstationary multivariate systems, characterized by multiple failure modes and nonlinear interdependencies among critical components or dimensions of the system.