Development of an Unstructured CFD Solver for External Aerothermodynamics and Nano/Micro Flows

Computational aerothermodynamics is the branch of science which focuses on the computation of the effect of thermodynamic and transport models on aerodynamics and heating. They are widely used for external ow cases. On the other hand, the computation of heat and stress in the design of Nano/Micro Electronic Mechanical Systems from the point of view of a Fluid mechanics engineer is also an important area of study. A generalized computational tool which can simulate the low and high speed flows at both the macro and micro levels is desirable from the perspective of industry, academics and research. For a developing nation, it is extremely important to have such a solver developed indigenously to create self-sufficiency and self-reliance. In this work, a robust three-dimensional density-based general purpose computational fluid dynamics solver was developed in house by our research group. The cell centred finite volume discretization method is used on an unstructured grid, which is more desirable for computation on a complex geometry from the perspective of pre-processing (meshing). Compressible ow solutions obtained from density-based solvers usually do not work well at low speeds where the ow is close to incompressible, unless special schemes and/or special treatments are used. An all-speed algorithm was incorporated using two different methods: (a) preconditioning of the governing equations or (b) through the use of the recently developed SLAU2 all speed convective scheme. The time-stepping discretization is done implicitly, using the lower-upper symmetric-Gauss-Seidel method, which allows us to take a high CFL number during computations. Throughout this work, we have used a second-order accurate reconstruction with limiters to accurately capture the shocks without dispersive error. Turbulence modelling is done using Favre- and Reynolds- Averaged Navier-Stokes equations using the Spalart Allmaras turbulence model. The developed solver is used to solve external ow problems at low and high speeds (hypersonic regimes). In these problems, the thesis focus is on the implementation and testing of an automatic wall function treatment for the Spalart-Allmaras turbulence model. The applicability of the solver is extended to rarefied gas ow regimes in the following manner. Thermal non-equilibrium which exists in the rarefied ow regime is tackled using non-equilibrium boundary conditions in the slip ow regime. The use of non-equilibrium boundary conditions allows the applicability of the Navier-Stokes equation to be extended beyond the continuum to the slip regime. This approach is used to solve problems of hypersonic rarefied flows and nano/micro flows; and for testing and validation of several recently proposed boundary conditions for several problems in the slip ow regime. The main focus of this work is in developing newer numerical methods and on testing and improving other recently proposed numerical techniques that are used for solving the problems covered in this thesis. In the following paragraphs we present the major outcomes of the thesis. The Spalart-Allmaras (SA) is one of the most popular turbulence models in the aerospace CFD community. In its original (low-Reynolds number) formulation it requires a very tight grid (with y+ ' 1) spacing near the wall to resolve the high ow gradients. The use of _ne grids increases the computational cost of the solutions. However, the use of wall functions with an automatic feature of switching from the wall function to the low-Reynolds number approach is an effective solution to this problem. We have extended Menter's automatic wall treatment (AWT), devised for the K