Calibration and Bias-Correction of the Steel Offshore Jacket Platform Models Using Experimental Data

Abstract

Offshore jacket platforms are of great importance due to their roles in shallow water areas, such as the Persian Gulf. Building infrastructures are always a challenge for engineers because of uncertainties in different aspects. Using an experimental model is an essential technique for reducing uncertainties. However, using experimental data adds other uncertainties to the problems and dealing with them can play a key role in project applicability and success. Uncertainties in every engineering problem can be classified into three distinct groups related to the computer model: (1) numerical uncertainty, (2) parameter uncertainty, and (3) structural uncertainty. Among these categorized uncertainties, the second and the third ones are considered for the case study of this research. In this study, model parameter uncertainties and the structural uncertainties have been considered simultaneously. This approach is applied to a three-dimensional (3D) fixed offshore jacket model (SPD9 located in the Persian Gulf) for the first time in the offshore structure field, both numerically and experimentally. Disagreements between the numerical model predictions and experiments are reduced by considering the calibration of model uncertain parameters and a discrepancy term, which is defined by a mathematical function for a real project. Therefore, the main benefits and innovations of the methodology can be stated as (1) using experimental data through modal analysis of a 3D steel offshore jacket platform for numerical model updating and (2) developing a previously studied method through reducing parametric and structural uncertainties simultaneously. The results express the success of considering parameter uncertainty and structural uncertainty for the previously mentioned case study by decreasing the disagreements between experimental data and numerical simulation. The results also indicate the importance of considering a discrepancy term in the calculation phase as a model form error term due to increasing the applicability of the proposed method for an offshore jacket platform.