Rapid configuration aerothermo evaluation

Astron has an in-house tool capable of rapidly predicting and evaluating hypersonic aerodynamic forces and heating in minutes, allowing configuration changes to be investigated rapidly without waiting for slow and expensive high fidelity CFD results.

Levels of fidelity (continuum methods):

• Surface inclination method with streamline tracing. This approach uses a surface inclination method (eg. modified newtonian) flow-field to calculate in seconds the approximate aerodynamic forces on the vehicle or capsule and pressure distribution. The same pressure distribution is then used with a streamline tracing algorithm and a mass-balancing entropy layer method to model the boundary layer and rapidly predict windward surface heating with typically good accuracy especially around areas usually of the highest heat flux, eg. the stagnation point, attachment lines, and around the vehicle nose. This method however does not model viscous effects and does not give good predictions for leeward side vehicle heating.
• Inviscid CFD coupled with streamline tracing. This approach uses a separate CFD program to solve the 3D Euler equations, typically at a coarse fidelity, and simulate the inviscid flow-field around the vehicle. The resulting flow-field is then used with a streamline tracing algorithm to model the boundary layer and rapidly predict surface heating. This method is computationally much lighter than more detailed CFD runs and can run in minutes for most geometries and give more accurate windward heating away from the vehicle nose, and also gives an initial estimate of leeward side heating. This inviscid method however still does not model viscous effects like separation and shock-boundary-layer-interactions (SBLIs), which require higher fidelty CFD and are important for example for accurate leeward side heating on many geometries.
• Detailed CFD. We also have experience using conventional commercial CFD solvers for hypersonic and reacting-flow problems. Simulating individual cases via this method can take many hours and be numerically challenging to converge.