A multi-load joint distribution model to estimate environmental design parameters for floating structures. (1st December 2020)

• Record Type: Journal Article
• Authors: Zhao, Yuliang; Dong, Sheng
• Is Part Of: Ocean engineering. Volume 217(2020)
• Publication Date: 2020-12-01
• Subjects: Floating structures -- Environmental design parameters -- Single extreme factor -- Multivariate joint model -- Conditional probability model; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2020.107818
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: Floating structure designs require an environmental condition simulation and corresponding design parameter determination. Typically, the design values are defined by individually calculating the N-yr return period; however, this not only results in overestimation, but it also reduces the accuracy of the structural reliability analysis. Therefore, a new environmental design parameter estimation method which is based on the joint distribution of multiple loads and conditional joint probability models was developed. A multivariate joint model for the metoceanic parameters of the South China Sea based on 30 yrs of annual extremum hindcast data and conditional probability models with one predominant factor were established; the most likely pair of other variables were also considered because of the complexity and variability of the environmental conditions. Environmental parameters with N-yr joint return period, as well as the structural response surface, corresponding to the most unsafe conditions were determined to derive design parameters. The possible conservatism of the conventional design standard was assessed by applying this joint design criterion to a semi-submersible platform example. The results indicate the design response is much more conservative than those obtained using proposed criteria, neither multivariate nor conditional joint models. The proposed design criteria can significantly reduce construction cost. Highlights: Environmental design parameters estimation based on joint multi-loads is proposed. A multivariate joint model based on 30-years hindcast data is constructed using copula. Conditional probability model with one dominating factor is presented. N-year joint design parameters is determined combined with response surface. Possible conservatism of traditional design standard is assessed.