No abstract

No abstract

No abstract

The interactions and effects of turbulent mixing, pseudo-boiling phenomena, and chemical reaction heat release on the combustion of cryogenic liquid oxygen and gaseous hydrogen under supercritical pressure conditions are investigated using RANS simulations. Comparisons of the present numerical simulation results with available experimental data reveal a reasonably good prediction of a supercritical axial shear hydrogen-oxygen flame using the standard k-ε turbulence model and the eddy dissipation concept combustion model with a 23 reaction steps kinetics for H2-O2 reaction. The present simulation qualitatively reproduced oxygen injection and its reaction with the co-flowing hydrogen, which is characterized by rapid flame expansion, downstream flame propagation, and expansion induced flow recirculation. Several turbulence models were used for numerical simulations. It is shown that the selection of an appropriate turbulence model for transcritical reacting flows is crucial and far more important than for subcritical reacting flows. It is indicated that the pseudo-boiling phenomena is the main reason for the considerable differences between the turbulence models in a transcritical flame. Also, it is demonstrated that the liquid oxygen core disappears faster in a non-reacting flow than in a reacting flow. The shear layer in the non-reacting flow is much stronger than reacting case; providing a large transfer of energy from the outer layer to the inner layer. At the supercritical injection conditions, the difference between the turbulence models is much less than the transcritical injection conditions.

A numerical simulation of swirling methane/air non-premixed flame (TECFLAME swirl burner) in a two-dimensional model combustion chamber is carried out to assess the influence of entrance flow swirl number on temperature distribution, flow behavior and NO pollutant formation. A Finite Volume staggered grid approach is employed to solve the governing equations. The second-order upwind scheme is applied for the space derivatives of the advection terms in all transport equations. The eddy dissipation-finite rate model is employed to predict the heat release and the Reynolds stress turbulence model is applied to simulate the flow behavior. NO formation is modeled as a post-processing solution. NO formation prediction has reasonable agreement with experiments for smaller swirl numbers but with the increase of swirl number, deviations between numerical results and the experimental data increase. It may be due to incorrect prediction of temperature distribution in higher swirl numbers. With the increase of swirl number, maximum temperature of chamber decreases from 1900 (K) to 1650 (K). With temperature decline, NO concentration in the exhaust decreases from 27 (ppm) at swirl number of 0.7, to 4 (ppm) at swirl number of 2. On the other hand, with increase of swirl number, ratio of prompt NO formation to thermal NO increase rapidly. In another word, with decrease of flame temperature, prompt NO roles increase noticeably.