Multidimensional discrete ordinates solutions to combined mode radiation heat transfer problems and their application to a free-falling particle, direct absorption solar receiver

Multi-dimensional radiative transfer in combined mode heat transfer problems was investigated with emphasis on the analysis and characterization of a free-falling particle cloud, direct absorption solar central receiver. A model was developed to calculate the relevant distributions in the curtain while a concentrated solar beam is impinging on the front face of the medium. The discrete ordinates approximation was applied to allow the spectral equation of transfer (EOT) to be modeled as a PDE.; Model verification tests were conducted to determine the accuracy of the model. One- and two-dimensional results showed that the discrete ordinates model provides satisfactory estimates of the radiant intensity, the heat flux and the temperature distributions for ordinate sets above {dollar}Ssb4{dollar} (12-flux approximation) for both the black and gray cases.; An experimental program was conducted to provide data on the performance of the free-falling particle receiver and to compare the results with model predictions. The extinction coefficient and the curtain porosities along with the transmitted fluxes and the exit temperatures were measured.; The boundary condition for the front face of the curtain, which was described in terms of a Fredholm integral problem, was determined through the use of angular heat flux data and parameter estimation techniques. A check of the accuracy of these calculations was performed by integrating the intensity to determine the boundary fluxes. Results showed reasonable flux distributions with a significant improvement from the 12- to the 24-flux model.; Comparisons of the exit temperature and the transmitted flux distributions were made with the model. Results showed satisfactory agreement with errors within 25% being observed at most points.; The results of the temperature predictions also showed reasonable agreement with the measured data. Errors ranged form 8% to 52%. Small temperature increases were thought to be the primary cause of the relatively large errors. This indicates that a larger temperature increase through the receiver would probably result in more accurate readings.; Indications are that the model performs satisfactorily for the higher order sets and that, when applied to a full scale receiver, reasonably small errors can be expected.