Abstract
In the present study, the FLAME large-scale hydrogen combustion tests F-8 and F-22 performed at Sandia National Laboratories, USA, were simulated using the TONUS CFD code. In test F-8 a configuration without obstacles and no top venting was used. The objective of the calculation of test F-8 was to investigate the effect of two different boundary condition specifications at the test channel exit as well as the effect of mesh refinement. The difference between the tested boundary conditions was observed to be significant. The CREBCOM combustion model was used in the simulation and attention was paid at finding an appropriate model parameter K_0 for the current test. Test F-22 which includes obstacles resulting in a 33% blockage ratio in the test channel was modeled by the k-ε turbulence model with Eddy Break-Up reaction kinetics. Both CREBCOM and EBU models were found to perform qualitatively well for both 2D and 3D models of the system. However, it was observed that the correlations presently available for the key parameter K_0 of the CREBCOM model are not suitable for the FLAME tests. In test F-22, an additional challenge was provided by the unclear conditions near the exit of the channel and it was noticed that high local pressure peaks were formed near the obstacles in the channel.
