Analysis of Fuel-Air Mixing in Jet in Crossflow

Abstract

With the growth of energy demand nowadays, environmental concern of NOx emission from gas turbines has been increasing in attention. NOx emission can be lowered by decreasing the local maximum temperature which results from a good premixing of air and fuel. The mixing performance is affected by location of the fuel holes on the fuel pins which can be optimized using computational fluid dynamics. However, the computational mesh is known to be an important parameter which affect solution convergence and accuracy, thus it is significant to be optimized. The main goal was to develop the new mesh strategy and confirm grid independence to achieve higher accuracy with minimum number of cells. This thesis study applied a typical jet in crossflow (JIC) configuration to analyze the mixing of fuel and air in gas turbines related to fuel injection nozzles and Siemens PLM's STAR-CCM+ was chosen as a software. Previously used mesh strategies in SGT-750, SGT-800 and other projects in Siemens Energy AB were investigated in terms of grid independence properties and accuracy compared to the experimental data presented by F. Galeazzo et al. [1]. The results revealed that the mesh required improvement. Model dependence test including realizable k-epsilon, LES, lag EB k-epsilon and SST k-omega was performed. The choice of turbulence model was found to be important to achieve reliable predictions of the flow. Realizable k-epsilon and LES were the most suitable models for JIC simulation. To achieve good flow resolution, evaluated by turbulent kinetic energy and subgrid viscosity ratio, the cell sizes were properly identified by Kolmogorov and Taylor microscales. A successful mesh optimization approach was achieved by combining the best previously used mesh strategy and adaptive mesh refinement with respect to the concentration gradient. The optimized mesh exhibited high grid independency and the velocity and concentration fields agreed well with available experimental data. Therefore, the static optimized mesh is recommended to be used for future optimization of the fuel hole distribution in gas turbines.