Numerical simulations of high-pressure mixing and combustion

This work aims at highlighting differences between low-pressure and high-pressure flows. A simple temporal mixing layer of oxygen and methane is adopted as start test case. From the numerical point of view, this simulation is a good test to check the robustness of the integration schemes adopted, in relation to the problem of spurious wiggles originating from the density gradients and the multi-species approach in a fully compressible formulation. Two non-reacting simulations are initially performed at 1.5 and 150 bar. The differences between the low-pressure regime in ideal gas conditions and the high-pressure regime with real gas thermodynamics are highlighted. Then the high-pressure shear-layer is simulated igniting the mixture. Numerical data from both non-reacting and reacting shear-layers are analysed to highlight the relative importance of mass diffusion mechanisms and to show how this affects the distribution of the Schmidt and Lewis numbers. Prandtl number distribution is also investigated. New effective (or actual) diffusion times (derived from Navier-Stokes equations) and characteristic non-dimensional numbers (assumed as ratio of effective times) are defined and compared to the standard ones. Finally, the robustness of the numerical schemes is tested by simulating the mixing between a central trans-critical liquid nitrogen jet and a coaxial gaseous hydrogen jet owing into a chamber filled in with super-critical nitrogen. Real gas Effects are investigated. Turbulence production mechanisms relevant at such conditions are also identified.