Optimized winding pattern design for composite pressure vessels using multiobjective particle swarm optimization and mandrel profile updating

To address the winding trajectory deviation and local stress concentration caused by fiber stacking in the dome region of composite pressure vessels, a filament-winding pattern design method based on multiobjective particle swarm optimization (MOPSO) and mandrel profile updating is proposed. First, a winding pattern model is established based on the non-geodesic equations and the continued fraction principle. Subsequently, multi-objective optimization of the winding parameters is performed with the objectives of minimizing fiber consumption and maximizing the minimum strength ratio. On this basis, a mandrel profile updating strategy is introduced to periodically correct the fiber trajectories during the multilayer winding process. Finally, the effectiveness of the proposed method is validated through fiber layer thickness analysis and finite element analysis. The results demonstrate that the proposed method can effectively alleviate fiber accumulation in the dome region, reduce the stresses in the hoop layers, and promote a more appropriate load distribution toward the helical winding layers, thereby improving the overall stress distribution. Based on the maximum stress failure criterion, the predicted failure pressure increases from 93 MPa to 97 MPa. The proposed method can therefore improve the structural load-bearing capacity while reducing local fiber accumulation, providing a useful reference for the lightweight design and safety performance enhancement of composite pressure vessels.