Multi-objective optimization of a lean premixed laboratory combustor through CFD-CRN approach

Abstract

In this paper, the effects of geometrical characteristics variations on the pollutant emissions in a laboratory lean premixed combustor are studied numerically and experimentally. For this purpose, the combustion chamber, premixing chamber, and equipment for measuring the flame temperature and combustion product concentrations are designed, fabricated, and assembled for a laboratory atmospheric test stand. An object-oriented optimization algorithm based on computational fluid dynamics (CFD), a chemical reactor network (CRN), and the response surface method (RSM) is presented for optimization of the combustion chamber geometry to obtain minimum pollutant emissions. The numerical approaches for modeling the flow field and predicting the detailed chemistry are validated against the experimental results. Parameters, such as minimizing NOx, CO, and NO emissions, are considered for optimizing the combustion chamber's geometry. Finally, it is shown that the hybrid approach of CFD-CRN can serve as a useful tool in the design process and the optimization of combustion chambers. The optimization process results show that minor changes in the combustor design can minimize the defined objectives to an acceptable level.