Modelling Aerothermal Heating with Conjugate Heat Transfer

Abstract

Supersonic and hypersonic vehicles experience flows under complex thermodynamic conditions due to the presence of shock waves and large thermophysical gradients. A large thermal load is imparted to the vehicle from the high-speed flow, leading to the aerodynamic heating of the surface. Additionally, the vehicle must sustain heat loads from within its structure, such as the propulsion system. The accurate modelling of heat transfer at these highly non-adiabatic wall conditions is critical in designing optimal thermal protection systems for hypersonic vehicles. This thesis aims to assess the predictive capabilities of computational fluid dynamic solvers in modelling aerothermal heating in supersonic and hypersonic flows using Conjugate Heat Transfer (CHT) methodologies, and investigate the impact of internal heating sources on the thermal and aerodynamic loads on aerospace vehicles. The predictive capabilities of near-wall turbulence modelling at high-speed, non-adiabatic flow conditions are first assessed for two commercial and two open-source CFD solvers: OpenFOAM (open-source), SU2 (open-source), Star-CCM+ (commercial), and ANSYS CFX (commercial). The overall error and uncertainty that can be attributed to solver selection at these complex conditions is quantified. SU2 and Star-CCM+ are assessed on their ability to model aerodynamic heating using CHT, with comparisons to hypersonic experimental studies and prior numerical investigations. The results from the code validations are used as a basis to conduct a CHT analysis on a simplified model of a supersonic vehicle to investigate the impact of internal heating on the thermal boundary layer of the external flow. The results show notable variations between the solvers in the kinematic and thermodynamic profiles of the high-speed non-adiabatic boundary layers, which are quantified. Furthermore, the treatment of the boundary condition at the wall plays a significant role in the variation of wall properties, particularly with the wall temperature prediction. Moreover, CHT validation studies show that aerothermal heating predictions of current commercial and open-source CHT solvers agree well with experimental and numerical data, but significant prediction errors occur in regions of Shockwave Boundary Layer Interaction (SWBLI) and stagnation points. The addition of internal heating on the CHT simulations of the generic high-speed vehicle results in a reversal of the wall heat flux vector as the freestream Mach Number is increased, where the heated wall case at low supersonic speeds transforms to a cooled wall case at hypersonic speeds. This thesis work provides a solid groundwork for conducting CHT simulations of high-speed wall-bounded flows with internal heating, using RANS solvers.