Numerical investigation of turbulent hydrogen/air diffusion flames and turbulence radiation interactions

Abstract

An investigation of the flame structure and radiation properties of turbulent hydrogen/air diffusion flames is reported. The laminar flamelet-conserved scalar probability density function approach is used to predict the scalar distributions throughout a laboratory-scale axisymmetric buoyant hydrogen diffusion flame. Predictions are compared with published measurements of mean and root mean square (RMS) temperatures and species concentrations based on the laminar flamelet concept. Predictions of spectral intensity and received heat flux are made with a narrow-band radiation model using mean properties, stochastic and mean emission methods to evaluate the effects of turbulence radiation interactions (TRI) and modelling TRI to predict the received radiant heat flux was very important. The predictions were, on the whole in good agreement with published measured data available in the open literature. Present study centres on the development of novel numerical models to predict TRI in turbulent hydrogen flames, implemented in a sophisticated way using enthalpy perturbation equation to account for radiative heat loss. This thesis highlights novel accomplishments in areas such as modelling lifted hydrogen jet flames, flame structures and external radiation fields where significant findings are reported. Firstly, successful extension of the lift-off model to hydrogen jet flames using strain rate as stretch parameter to accurately predict the lift-off height and affirming the smallscale strain rate model is better than the large-scale strain rate model which is different to methane lifted jet flames. Secondly, different jet flames were investigated using two different probability density functions (PDFs) and two transport equations taking into account fluctuations of temperature T ¢2 and water vapours 2 2 H O X¢ . The new Truncated Gaussian PDF was confirmed to give better predictions than other methods. Lastly, of the three approaches considered in modelling TRI the stochastic method proved the most accurate to predict the spectral intensity distribution and radiative heat flux distribution.