Experimental and machine learning studies of thermal impinging flow under ceiling induced by hydrogen-blended methane jet fire: Temperature distribution and flame extension characteristics

Hydrogen-blended natural gas has a winged used in such thermal and combustible systems, the ceiling thermal characteristics have important parameters of impinging jet configurations. In this paper, a series of experiments were carried out to study the ceiling flame extension characteristics and temperature profile in a thermal impinging flow of methane jet diffusion flame with hydrogen addition. It was found that: for the given volume flow rate of methane fuel, the ceiling temperature increases although ceiling flame extension area (length) de-creases, with the volume flow of hydrogen mixing ratios increasing. Then, a physical and mathematical model, combined influence of flame shape hypothesis based on physically the unburnt fuel, hydrogen addition, heat release rate, the effects of source-ceiling height and nozzle diameter is estimated to predict the flame extension area of methane jet diffusion flames with different hydrogen mixing ratios. Moreover, the fire plume temperature between buoyant plume and intermittent flame increased with the increase in hydrogen addition due to the fire plume temperature increased. The temperature profile models of the ceiling jet are proposed for methane diffusion flames with different hydrogen mixing ratios. Furthermore, to perform the calculation quickly and facilitate rapid thermal engineering design, this paper used machine learning methods (GABP neural network) to derive the volume flow rate of gas leak from the temperature data of the ceiling detected by thermocouples. The results show that the AI prediction for the heat release rates has a good correlation with the verification experimental value. The results of this paper may demonstrate the potential for industrial combustion systems and fire safety engineering applications of methane jet diffusion flame with hydrogen addition.