The Effect of Boundary and Geometry Simplification on the Numerical Simulation of Front-End Cooling

The maturity of Computational Fluid Dynamics methods and the increasing computational power of today's computers has allowed the automotive industry to integrate CFD into the mainstream design process in many areas. As the application of CFD technology is moving from the component level analysis to the system level, the complexity and the size of the models increases continuously. A successful simulation requires synergy between CAD, grid generation, solvers and post-processing so that a timely solution can be obtained that influences the design directions.

The complexity of the simulations introduces several issues that affect the acceptance of these methods including but not limited to the accuracy, grid independence, influence of boundary conditions, level of geometry detail. The investigation of these issues is the purpose of the current work. The flow around a family sedan has been investigated using a commercial code, and solutions have been obtained for different test conditions. The grid generation was based on a tetrahedral unstructured technology. This approach provides the means for fast and efficient generation of volume meshes for complex geometries. The influence of the geometry detail on the results was investigated by comparing two models, one with all the major underhood components included and a second with a simplified engine and most of the underhood components removed. The flow rate through the cooling package was used as the influence gauge. Two tunnel configurations were used to investigate the effect of the tunnel inlet conditions on the cooling flow rate. Again the flow rate through the cooling package is used to judge the influence of the boundary conditions on the solution. The influence of the grid on the solution was also investigated by applying local refinement in the regions of high gradients. Numerical results for the flow rate through the cooling package are presented and compared with the available experimental data. The usefulness of the approach and implications of the different factors in the successful implementation of the method is also discussed.