Turbulence Modelling on Supercritical Jet Injection

As space exploration enters a new era where not only public but also private enterprises begin to flourish, financial viability becomes a preponderant factor. It is hard to estimate how far the contribution of this sector to the development of the resources available thus far will be, but the link between engineering technology and cost reduction is unquestionable. In a Liquid Rocket Engine, higher combustion efficiency comes at the cost of both the fuel and/or the oxidiser surpassing their critical points and entering the domain of supercritical fluids. While these conditions have been present in rocket engines combustion chambers for decades, they are still not fully understood and tools to properly simulate such conditions are still in the development stages. The highly non-linear behaviour of the thermophysical properties of a fluid in this regime increases the difficulty of any attempt to run numerical simulations. The ideal gas law is no longer valid and must be replaced by more accurate models. The Peng-Robinson and the Soave-Redlich-Kwong Equations of State (EoS) are compared to a reference EoS for nitrogen to gain some insight on the magnitude of error that simpler cubic Equations of State incur in the simulation results. The output of the multi-parameter EoS is obtained from the real gas library REFPROPv9.1 thus reducing computational costs while still achieving a level of accuracy that would otherwise not be possible. A similar treatment must be given to transport and caloric properties that have a significant impact on the flow structure. To deal with the incompressible but variable density conditions inside the combustion chamber, the standard time-averaging method is replaced by the Favre averaging procedure and the system of equations is closed with different turbulence models, the main focus of this study. The performance of said models, designed and calibrated to run in subcritical conditions, is then studied and their validity for the supercritical regime is assessed. Ultimately, computational ...