Vehicle thermal design traditionally depends heavily on prototype testing in thermal wind tunnels. This testing process is very expensive, time-consuming, and inflexible. Testing involves thermocouple instrumentation that requires test engineers to estimate a priori where thermal problems might occur — but the highly turbulent nature of convective underhood flows makes this very difficult or impossible to predict. Relying on redesign and retesting is an expensive hit-or-miss process that often ultimately fails to identify the highest temperature locations.
Heat soak is an inherently transient problem where the temperature quickly rises then gradually falls through convective cooling. This is almost impossible to visualize in a wind tunnel, yet these complex structures must be understood in order to optimally locate and protect components. In addition, temperature is a function of the complex interaction between conduction, radiation, and convection in the surrounding fluid, especially for very hot components. Accurately predicting this is extremely challenging.
Given that there is increasing pressure from the marketplace to speed up and improve the vehicle development process, it is clear that a more effective method is required to address soak early in the vehicle design process. A high-fidelity simulation is necessary to capture the relevant physics. However, simulation of a transient problem like soak is challenging and time-consuming with traditional Navier-Stokes-based fluid simulation codes because they are not inherently transient.