Prediction of offshore structural response extreme values by three different approaches of efficient time simulation technique

Abstract

Offshore structures are exposed to random wave loading in the ocean environment and hence the probability distribution of the extreme values of their response to wave loading is required for their safe and economical design. To this end, the conventional (Monte Carlo) time simulation technique (CTS) is frequently used for predicting the probability distribution of the extreme values of response. However, this technique suffers from excessive sampling variability and hence a large number of simulated extreme responses (hundreds of simulated response records) are required to reduce the sampling variability to acceptable levels. A more efficient version of the time simulation technique (ETS) was recently introduced which takes advantage of the correlation between the extreme surface elevations and their corresponding extreme responses by dividing the simulated extreme responses into a number of groups based on the magnitude of their associated extreme surface elevations. The probability distribution of response extreme values for each group is then calculated individually based on a relatively small number of simulations, and then the total probability theorem is used to derive the probability distribution of response extreme values. The ETS procedure is found to be many times more efficient than the CTS method. However, its efficiency and accuracy reduces for sea states of lower intensity. In this paper, a more efficient version of the ETS technique is introduced which takes advantage of the correlation between extreme values of the nonlinear response and their corresponding linear response extreme values (both quasi-static and dynamic linear responses are considered)). The three different versions of the ETS technique are compared by exposing a test structure to sea states of different intensity. It is demonstrated that the extreme linear response versions of the ETS technique are more accurate and efficient than the version based on surface elevation extreme values.