화학공학소재연구정보센터
Combustion and Flame, Vol.131, No.3, 329-348, 2002
Release of chemical energy by combustion in a supersonic mixing layer of hydrogen and air
The process involved in chemical energy release by combustion in a supersonic, constant pressure, hydrogen-air laminar mixing layer was studied computationally, with a chemical kinetics model involving nineteen reactions and eight species. To try to find out the physical reason for the different trends of the pressure curves observed in an experimental supersonic combustor at two different initial air stream temperatures. Two initial air stream temperatures corresponding to the two experimental cases are chosen such that the higher temperature yielded a shorter ignition distance, and the lower temperature yielded a longer ignition distance. For both cases the stream wise rate of energy release rises rapidly to a peak after ignition then falls to a post-ignition value which decreases very slowly with distance. A single premixed flame occurs at ignition for both cases, but then develops into a triple flame structure in the high temperature case, and a flame with only two branches in the low temperature case. The flames move from the airside to hydrogen side consuming the oxygen as they go, until the post-ignition phase is reached. There the dominant energy release arises from the formation of a diffusion flame. In the high temperature case a narrow lean premixed flame accompanies this diffusion flame on the airside. The flame structure, but not the energy release, is effected by the initial temperature distribution across the mixing layer, which is found to be influenced by the velocity difference between the faster air stream and the slower hydrogen stream. Increasing the concentration of oxygen atoms in the oncoming air stream was found to cause substantial reduction in the ignition distance, but did not significantly effect the flame structure, or the rate of heat release in the post-ignition phase. Finally, the different trends of pressure curves observed in the experiment can be reconstructed when pressure variation was considered in this model. Thus we can conclude that the difference in the trends of the pressure curves is caused by the difference in the initial air stream temperature.