Aerodynamic performance investigation through different chemistry modelling approaches for space re-entry vehicles using the DSMC method

High-speed flows with Mach numbers well above the hypersonic regime pose significant modelling com-plexities due to increased levels of thermal energy, which in turn result in a variety of chemical reactionsthat become dominant and thus have to be accurately modelled. Within Computational Fluid Dynamics(CFD), the Direct Simulation Monte Carlo (DSMC) method is commonly chosen here as it has shownsuperior performance over traditional Navier-Stokes-based solvers due to a breakdown in the continuumhypothesis. Space re-entering vehicles are commonly exposed to high Mach numbers when entering intoearth’s atmosphere and low density so that the mean free path of particles is comparable to the lengthof the vehicle itself. Thus, these types of applications require challenging modelling approaches which isthe subject of this study. We use the open-source CFD solver OpenFOAM in this study, which comesprebuilt with the dsmcFoam solver. This implementation of the DSMC method lacks, however, the abilityto model chemical reactions and thus is not equipped to predict aerodynamic coefficients for high-speedflows. Recently, the dsmcFoam+ solver has been proposed [1] and implemented into OpenFOAM whichfeatures, among other things, the ability to model chemical reactions through the Quantum-Kinetic (QK)model. The aim of this study, then, is threefold; 1) Validate the new dsmcFoam+ solver against availablereference data from the literature and compare it to the default dsmcFoam solver, highlighting the im-portance of chemical modelling, 2) Publish all simulation and setup files through an online repository tofacilitate an easy case setup for researchers wishing to evaluate or adopt the new dsmcFoam+ solver, 3)Provide documentation for the new dsmcFoam+ solver in the context of OpenFOAM where there is littledocumentation available. We investigate the flow of a re-entry vehicle with a freestream Mach number of25.6 at different angle of attacks and find that the chemical modelling approach taken has a significantinfluence over the aerodynamic coefficients which are up to 24% apart. Similar results are obtained forthe heat transfer coefficient, which shows differences of up to 28%. Based on our findings, we advocatethat the dsmcFoam+ solver should be used for aero-thermodynamic calculations as its ability to predictchemical reactions and thus changes in the flow field will significantly affect the overall solution accuracycompared to a non-reacting modelling approach.