An Automated Sequential Computational Screening Workflow for Hydrogen Diffusion, Structural Response, and Fatigue Analysis of EPDM O-Rings Under High-Pressure Hydrogen Exposure

Hydrogen transport and cyclic loading complicate O-ring assessment. This study developed an automated sequential screening workflow for ethylene–propylene–diene monomer (EPDM) O-rings, linking Abaqus mass diffusion, one-way concentration-field transfer to nonlinear structural/contact analysis with prescribed isotropic swelling, and fe-safe/Rubber fatigue post-processing. EPDM 60A and 90A were evaluated at sampled pressures of 3, 7, 14, and 21 MPa and circumferential installation stretches of 1–5%, with radial squeeze fixed at 20%. Structural analyses used third-order Reduced Polynomial models with built-in Mullins effects. Material-specific diffusivities characterized at approximately 1 bar were held constant across pressures; hydrogen-content and post-decompression free-specimen volume-change data informed model inputs. Fatigue inputs were characterized in air at approximately 23 °C using non-equivalent bases for the two compounds. At 1% stretch, fatigue-life outputs decreased from 6.877 × 107 to 12.923 repeats for 60A and from 3.155 × 109 to 3.896 repeats for 90A between 7 and 14 MPa. At 21 MPa, outputs were 0.712 and 0.041 repeats, respectively; these fractional values are numerical screening indicators without assignment to a within-cycle event or damage mode. The outputs support only input-conditional comparison of axisymmetric cases and experimental prioritization; they do not support decisions on component durability, leakage, sealing performance, or operating pressure.