Numerical simulation of thermobuoyant flow with large temperature variation

The use of the classical Boussinesq approximation is a straightforward strategy for taking into account the buoyancy effect in incompressible solvers. This strategy is highly effective if density variation is low. However, ignoring the importance of density variation in highly thermobuoyant flow fields can cause considerable deviation from the correct prediction of fluid flow behavior and the accurate estimation of heat transfer rate. In this study, an incompressible algorithm is suitably extended to solve high-density-variation fields caused by strong natural-convection influence. The key point in this research is the way that an ordinary incompressible algorithm is extended to non-Boussinesq-regime applications. The extension results in a unified algorithm capable of solving thermobuoyant flow fields in either a pure incompressible algorithm incorporated with the Boussinesq approximation or an entirely compressible algorithm where the density field is affected by both temperature and pressure fields. The extended algorithm is then verified by solving the benchmark convecting cavity problem at Rayleigh 10(6) and a temperature range of epsilon = 0.01-0.6. The results show that the method can vigorously solve thermobuoyant flow fields with extreme density variation.