Computational fluid dynamics model based artificial neural network prediction of flammable vapor clouds formed by liquid hydrogen releases

Abstract

Summary The flammable vapor clouds formed by liquid hydrogen release potentially leads to explosions and serious fires. In this paper, an artificial neural network (ANN) model based on the back propagation algorithm was developed to predict the horizontal and vertical hydrogen diffusion distances (the maximum distance of flammable vapor cloud with a hydrogen volume concentration of 4% to the release source in downwind and height direction). For this purpose, liquid hydrogen releases were modeled using a computational fluid dynamics (CFD) model, which was validated by experimental data. The single hidden layer ANN model used the wind speed, ground temperature, leakage duration, and leakage mass flow rate as input parameters. Then, the ANN model was trained and verified using 50 sets of sample data derived from the CFD simulations selected via the orthogonal experiment method. The results show that the ANN model well predicts the hydrogen diffusion distances, with the percentage error of the horizontal diffusion distance within 10% and the prediction error of the vertical diffusion distance less than 1.5 m for 90% of test points. The overall correlation coefficient between ANN predicted output and target results reaches 0.99329. Furthermore, the ANN model predicts four orders of magnitude faster than the CFD simulation and reduces the root means square error and the maximum relative error by more than 23% compared to predictions by the empirical correlation fitted from the CFD data. Then, the effects of the four input parameters on the diffusion distances were analyzed using the ANN model.

Study on Liquid Hydrogen Leakage and Diffusion Behavior in a Hydrogen Production Station

“…Therefore, this study involves a two-phase flow problem, where the gas phase is a mixture of air and hydrogen, and the liquid phase is liquid hydrogen, for which we use the mixture model. The mass conservation equation and momentum conservation equation [19] are as follows:…”

Prediction of Electron Beam Welding Penetration Depth Using Machine Learning-Enhanced Computational Fluid Dynamics Modelling

“…These challenges often involve either the unavailability of extensive experimental data or complexities influenced by various factors. Examples include predicting the size of flammable vapour clouds caused by liquid hydrogen release [24], solving problems related to natural convection and conjugate heat transfer [25], predicting solid particle erosion in the oil and gas industry [26], and enhancing the design of centrifugal pumps [27], etc. Based on these studies, the integrated model has shown promising potential in guiding manufacturing processes, explaining physical rules, and enhancing industrial safety.…”

Optimized model‐based diagnosis approach for hydrogen leakage in hydrogen supply system of fuel cell truck

“…The pressure of pipelines of hydrogen supply system exceeds 0.15 MPa. Considering the degradation of seals due to vibration or crash, hydrogen leaks from hydrogen supply system, leading to energy loss or even safety accidents such as fire and explosion 4,5 . So the diagnose of hydrogen leakage is critical to save hydrogen energy and improve the reliability of hydrogen supply system of FCV.…”