Reliability-Based Topology, Shape, and Size Optimization to Find the Optimal Four-Legged Jacket Offshore Platform

This study aims to establish a comprehensive reliability-oriented optimization framework for a four-legged offshore platform jacket without batter, addressing the persistent issue of material overuse in conventional deterministic design. The proposed methodology sequentially integrates topology, shape, and reliability-based size optimization to achieve a lightweight yet safe structural configuration. Topology optimization is first performed using the Solid Isotropic Material with Penalization (SIMP) method to minimize strain energy under prescribed volume constraints, generating an optimal load-transfer mechanism. The obtained layout is refined through shape optimization to improve joint configuration and stress redistribution. Finally, reliability-based design optimization (RBDO) is implemented using a Kriging surrogate model coupled with Sequential Quadratic Programming (SQP) and Monte Carlo simulation to minimize structural weight under probabilistic constraints, including strength, displacement, and target reliability index (β ≥ 3). The optimized jacket achieves a 53.71% weight reduction while satisfying safety requirements with a maximum unity check of 0.77 and a reliability index of β = 3.11. The proposed surrogate-assisted RBDO significantly reduces computational effort compared to direct reliability analysis. The primary novelty lies in the unified integration of topology, shape, and probabilistic size optimization for a non-battered platform jacket, which has not been comprehensively reported in previous offshore structural optimization studies.