Structural design of anion exchange membrane electrolyzer using finite element methods

Abstract Bolt preload plays a critical role in the mechanical integrity and sealing performance of large-scale rectangular anion exchange membrane (AEM) electrolyzers. However, conventional uniform preload strategies often lead to severe stress concentration and non-uniform deformation because of the large aspect ratio of the stack. In this study, a finite element model was established to systematically investigate the effects of bolt quantity, bolt specification, and preload distribution on the mechanical behavior of a rectangular AEM electrolyzer. Based on the stress distribution characteristics, a position-dependent differential preload strategy was proposed to improve load transfer and structural uniformity. The results show that increasing the bolt number from 22 to 30 significantly improves the structural response while maintaining reasonable manufacturing cost, whereas further increasing the bolt number provides only marginal benefits. Compared with the conventional uniform preload, the proposed preload strategy reduces the inter-bolt stress variation by approximately 6 MPa, improves the cover plate stress uniformity by approximately 40%, decreases the maximum stress concentration factor by approximately 35%, and reduces the deformation of the most critical bolts by approximately 0.05 mm, resulting in more coordinated structural deformation. In addition, the proposed position-dependent differential preload strategy increases the minimum contact pressure at the cover plate–bipolar plate interface and produces a more uniform contact pressure distribution, indicating improved sealing capability. Engineering-scale leakage tests further verify the effectiveness of the proposed preload strategy. The optimized configuration employs 30 M18 bolts with position-dependent preloads, providing practical guidance for the preload design and structural optimization of large-scale rectangular AEM electrolyzers with similar structural configurations.