In this thesis, different approaches are presented for the computation of radiative transfer in enclosures containing non-gray media. An existing finite volume method is chosen as the computational procedure for the solution of the radiative transfer equation in multidimensional geometries. This procedure is first validated with the help of benchmark solutions generated in this work using the Monte Carlo method. The spectral variation of absorption property of gases is considered next, and three approaches to model non-gray gas properties for radiation computations are presented. The objective in these approaches is to provide absorption coefficient as the basic property for the radiative transfer equation solvers, and to compute radiative transfer in a computationally efficient manner. The first method is the narrow band k-distribution method based on the Malkmus narrow band model, developed by researchers in the atmospheric science field. A wide band k-distribution method based on the Wang wide band model was also formulated and tested. The narrow band based method although more accurate, is more computationally intensive. The wide band based method gives reasonably accurate results, and is faster. An expression derived in this work enables the determination of wide band k-distributions simple. The third approach presented here is based on the weighted-sum-of-gray-gases principle. A simple method of calculating gray-gas weights from narrow band cumulative k-distribution expression is presented. This procedure is as accurate as the narrow band k-distribution method, and is the most computationally efficient. The models are compared with each other in terms of accuracy, speed and scope for generality. The third approach combining k-distribution and weighted-sum-of-gray-gases principles is shown to be the best for engineering applications. Results are presented for radiative transfer in multidimensional complex geometries using this approach.
展开▼
