Prediction and Optimization of the Long-Term Fatigue Life of a Composite Hydrogen Storage Vessel Under Random Vibration

A composite hydrogen storage vessel (CHSV) is one key component of the hydrogen fuel cell vehicle, which always suffers random vibration during transportation, resulting in fatigue failure and a reduction in service life. In this paper, firstly, the free and constrained modes of CHSV are experimentally studied and numerically simulated. Subsequently, the random vibration simulation of CHSV is carried out to predict the stress distribution, while Steinberg's method and Dirlik's method are used to predict the fatigue life of CHSV based on the results of stress distribution. In the end, the optimization of ply parameters of the composite winding layer was conducted to improve the stress distribution and fatigue life of CHSV. The results show that the vibration pattern and frequency of the free and constrained modes of CHSV obtained from the experiment tests and the numerical predictions show a good agreement. The maximum difference in the value of the vibration frequency of the free and constrained modes of CHSV from the FEA and experiment tests are, respectively, 8.9% and 8.0%, verifying the accuracy of the finite element model of CHSV. There is no obvious difference between the fatigue life of the winding layer and the inner liner calculated by Steinberg's method and Dirlik's method, indicating the accuracy of FEA of fatigue life in the software Fe-safe. Without the optimization, the maximum stresses of the winding layer and the inner liner are found to be near the head section by 469.4 MPa and 173.0 MPa, respectively, and the numbers of life cycles of the winding layer and the inner liner obtained based on the Dirlik's method are around 1.66 × 10⁶ and 3.06 × 10⁶, respectively. Through the optimization of ply parameters of the composite winding layer, the maximum stresses of the winding layer and the inner liner are reduced by 66% and 85%, respectively, while the numbers of life cycles of the winding layer and the inner liner both are increased to 1 × 10⁷ (high cycle fatigue life standard). The results of the study provide theoretical guidance for the design and optimization of CHSV under random vibration.