A computational study on modelling and simulation of non-premixed flames

Abstract

A computational study on two-dimensional laminar and turbulent non-premixed combustion has been conducted. Firstly, the laminar regime and then the turbulent regime were discussed. OH* chemiluminescence emission phenomena in laminar co-flow methane-oxygen flames were studied under different operating conditions in a confined concentric tube burner and injector flow was analysed by simulating the flame in a GCH4-GO2 thrust chamber with various combustion models and chemical kinetic schemes numerically. Results for the co-flow laminar flame was compared and validated with experimental work. It was shown that the reduced mechanism over-predicted the maximum temperature, as well as slightly under-predicted the flame height. Several problems were investigated with pure methane-oxygen flame and nitrogen diluted methane-air flame to identify variation in OH* chemiluminescence distribution. It was found that the nitrogen diluted methane-air flame became thinner and longer when compared with the pure methane-oxygen flame. Moreover, further neglecting the nitrogen-related species in the oxidizer benefited curbing pollutant formation. Then, the radiation effect in the flame was analysed using the discrete ordinates (DO) radiation model. From the OH* distribution, a sheen-less OH* intense region was observed in the case with the no-radiation model, which was not the case when the radiation model was included in all kinetic mechanisms. Furthermore, increasing the equivalence ratio drove the flames to move higher and become thin, and the flames became contorted near the flame front. From the CH* chemiluminescence distribution, it is concluded that only OH* chemiluminescence can be utilized for the combustion process, and CH* chemiluminescence may be inappropriate. Overall, the detailed chemical kinetic schemes predicted the flame and OH* emissions effectively. OH* chemiluminescence was well observed under various operating conditions, and the modified detailed mechanisms can be recommended for reasonable and economical prediction of oxy-rich methane flames. The turbulent combustion of GCH4-GO2 was modelled at relevant engine conditions using two combustion models (Non-premixed combustion model (PDF-based steady diffusion flamelet and chemical equilibrium) and Eddy Dissipation Concept combustion model). For all the approaches used, we performed a comparison with experimental data. And a good agreement is obtained between the computed results and experimental data in terms of OH* chemiluminescence emission. Combustion kinetics plays an important role in turbulent combustion modelling; the computational cost depends on the number of species, number of chemical reactions, and grid size considered in the respective domain. In this work, GCH4-GO2 jet flames have been simulated by implementing simple chemistry up to comprehensive multi-step chemistry, as reported in many literatures. And we observed that the results obtained using the simple kinetic mechanisms in the non-premixed model itself provide a good coherence with experimental results. But reduced kinetic schemes were unable to predict some intermediate species emissions in the combustion engine. Additionally, the effect of various chemical kinetic models on the combustion process was analysed and compared with one another. And the mass fraction changes of some chemistry species for various kinetic models during the combustion process were studied. The primary intention of this work is to study the effects of various combustion models and chemical kinetic schemes in the GCH4-GO2 injector and quantify these effects.