Large Eddy Simulation Of Flat Wall Jet Film Cooling Using OpenFOAM

Film cooling technology is very important for the development of gas turbine. However, direct research on the film cooling efficiency of turbine blade is very complex. In this paper, to simplify the problem the flat wall jet film cooling case was simulated. The pisoFoam solver in an open source software OpenFOAM was used. Reynolds averaged Navier-Stokes Equatio-ns (RANS) and Large-eddy-simulation (LES) methods were used. The main content includes the following several aspects:First of all, the numerical models used in the simulations were deduced. Then the pisoF-oam solver was tested. By simulating the plane channel case and backward-facing step flow case, the pisoFoam solver is verified that it can give out good numerical solution, witch close to experimental data. What’s more, the modified pisoFoam solver by adding temperature equ-ation, which is renamed new_pisoFoam is reliable.Finally, the flat wall jet film cooling case was simulated by using the new_pisoFoam solver, RANS method and LES method respectively. The standard k-ε two equations model and Launder-Sharma low Reynolds number k-ε model were used in the RANS simulation. And the Smagorinsky model and the dynamic Smagorinsky model compiled by myself were used in the Large-eddy-simulation. By comparing the numerical results with experimental data, the following main conclusions are got:The numerical solutions from the RANS method match the experimental data. However, this method can’t provide instantaneous turbulent inf-or nation, so it cannot give more detailed flow field information. While the study shows that numerical results, extracted from the simulation using LES method, have a much better consi-stence with the experimental data than RANS method in the aspect of velocity, turbulent velocity and shear stress. Especially that the dynamic Smagorinsky eddy viscosity model offers more accurate flow field information near the wall. However, at the respect of cooling efficiency, the result form LES method is worse than that of RANS method. The main reason is that the added temperature-equation is not suitable to LES method, it’s not same with the temperature-equation in the sub-grid model in LES. In addition, based on the open nature and parallelism of OpenFOAM, users can modify and compile solvers to meet computing needs according to the specific requirements. With enough computing resources, using parallel com-puting to solve cases using the large eddy simulation can get a fully developed turbulent flow field in a short time. And large eddy simulation results tend to be more close to the experimental data and are more valuable for theoretical analysis.