Uncertainty analysis in the design of Type-IV composite pressure vessels for hydrogen storage

Abstract

This study focuses on uncertainty quantification (UQ) and global sensitivity analysis (GSA) for the burst pressure (BP) in Type-IV hydrogen composite pressure vessels. Key uncertain parameters, including elastic properties, composite strengths, ply thicknesses, and fiber orientations, were considered. Latin Hypercube Sampling (LHS) efficiently explored the uncertainty space, while Polynomial Chaos Expansion (PCE) modeled BP responses, with Sparse PCE reducing computational costs by selecting influential polynomial terms. Sobol’ indices were used to assess the direct and total influence of the uncertain parameters on the BP variability, guiding optimization in composite pressure vessel design. The development and analysis of the tank model used conventional shell elements, starting from the liner’s inner dimensions and incorporating filament winding via the Abaqus Composite Layup feature. Critical design aspects, such as ply thickness, material properties and fiber orientation, were employed. Failure analysis, driven by internal pressure, evaluated burst pressure in cylindrical and dome sections. Damage progression was assessed using the Hashin failure criterion. The study explored uncertainty propagation in tank designs across four scenarios, including low-pressure 12-ply tanks and high-pressure 52-ply configurations, incorporating 15 and 37 uncertain parameters. Fiber tensile strength and ply thickness emerged as the dominant factors affecting the BP. Fiber strength and ply thickness consistently influenced stiffness and failure mechanisms, emphasizing their critical role in the hydrogen tank design.