Modeling of combustion and flow in a single element GH2/GO2 combustor

Abstract

A single element gaseous H2/O2 combustion test case set up by the Pennsylvania State University is investigated using the new Reynolds Averaged Navier–Stokes (RANS) code, Rocflam3. There are considerable numbers of publications on this test case. Some of the publishing authors assume that it is not possible to achieve good results for this test case with simple RANS methods whereas others demonstrate the opposite. This is a reason for discussion which continues in this work. To obtain a reasonable solution for the test case, different simulation settings have been examined. This includes a grid study, a comparison of various $$k{\text{-}} \epsilon$$ type turbulence models and the variation of the turbulent Prandtl and Schmidt numbers. Turbulent combustion is treated via an equilibrium-based, presumed probability density function approach with a mixture fraction formulation. The 2D steady state simulations show good agreement with the experimentally measured heat flux. However, a discrepancy between the computed chamber pressure and the measured one is observed and is discussed incorporating considerations about combustion efficiency. This discussion shows that the experimental data are inconsistent. Computations performed with the chemical equilibrium code by Gordon and McBride are used to support the statements.