Mesh convergence for turbulent combustion

Our central result is a methodology for predicting mesh convergence for three dimensional (3D) turbulent combustion simulations, based on less expensive one dimensional (1D) and two dimensional (2D) simulations. We verify the prediction by comparison to a 3D finite rate chemistry simulation based on a reduced reaction mechanism, and we further verify it by comparison to a completely independent simulation of the same problem. We validate our simulation by comparison to experiment. Additionally, we assess grid requirements for finite rate chemistry with more detailed chemical reaction mechanism. In both cases, the test problem is an engineering scale study of a model scramjet combustor designed by Gamba et al. We find that the mesh requirements are not feasible for finite rate chemistry simulations of engineering scale problems with detailed reaction mechanism, as expected, but these criteria are less severe than the Kolmogorov scale.