Condition prediction for liquid hydrogen containers is crucial to their safety and efficient utilization. An explainable multi-fidelity machine learning model was established to predict the ventless hold time of 500 L vehicle liquid hydrogen cylinder, and the effects of machine learning algorithms and the proportion of high-fidelity data on model performance were analyzed. The results showed that ensemble learning algorithms exhibited superior accuracy and lower computational overhead compared to the multilayer perceptron. In terms of prediction accuracy, the random forest algorithm achieved the lowest root mean square and mean absolute error on the test set, with values of 4.04 and 2.10. The enhanced low-fidelity data strategy effectively reduced the discrepancies between data from different fidelity levels, while the light gradient boosting machine was the optimal choice to conduct multi-fidelity modeling. The accuracy of the enhanced multi-fidelity model greatly outperformed the multi-fidelity model without enhancement. As the proportion of high-fidelity data in the training set increased from 30% to 35%, the root mean square error of the enhanced multi-fidelity model decreased by 4.7%. In addition, an explainable analysis revealed that pressure increment and heat leakage were dominant promoters affecting ventless hold time. This multi-fidelity modeling study provided a method for predicting ventless hold time that combines accuracy with explainability, and offered insights into the state prediction of liquid hydrogen containers.

As sustainable development becomes increasingly important, technologies for liquid hydrogen (LH₂) storage and transportation are advancing rapidly. Safety concerns regarding LH₂ tanks in fire accidents require further attention. In this study, a one-dimensional thermodynamic model was developed based on layer-by-layer analysis to assess the heat transfer performance of the insulation structure in LH₂ tanks under fire conditions. Through the transformation of the solving target and iteration rules, a novel and efficient solution method was proposed for such thermodynamic problems. The thermodynamic performance of the insulation structure coupled with spray-on foam and variable-density multilayer under normal temperature (NT) and standardized fire conditions (863.15 K) was analyzed, and the effects of insulation structure parameters and environmental factors were evaluated. A case study of a 500 L vehicle LH₂ tank was conducted using the software package BoilFAST, with the total heat leakage as the key input, to analyze the evolution of internal pressure and temperature. Results show that within the insulation structure, temperature decreases rapidly by 80.35% and 89.55% under fire and NT conditions, respectively. Spray-on foam insulation thickness, layer density, residual gas pressure, and hydrogen temperature exert minor effects, while the temperature of the external environment and the number of layers significantly affect the heat flux under the fire condition. Under the NT condition, heat leakage is primarily attributed to support structures and accessory pipelines, whereas under the fire condition, heat leakage from the insulation structure becomes the main source, accounting for 63%. This study provides a reference for heat transfer assessment of LH₂ tanks in fire scenarios.