fluent详细新功能.ppt

* * This slide shows another mesh picture of the same engine in the valve area. The pure structured hex mesh in this area is all automatically generated by ic3m. * * For this Subaru engine, a comparison is made between Fluent’s native approach using remeshing and IC3M approach using pure layering. In the highlighted area, one can see that the jet coming from the valve is more diffusive for the remeshing approach and thus attached to the walls. In the same locations, the results from ic3m showed less diffusive solution and a detached jet. * The solution difference observed from the previous slide can perhaps be explained by different meshes in the region. In the remeshing approach, a tet mesh is used in the jet area, while hex is used in the ic3m layering approach. Please also note the cell count difference. Even with the exhaust port included, ic3m mesh has only 412K at this crank angle compared with 493K in the remeshing case without the exhaust port. * * As for robustness of IC3M approach, since ic3m replies on ICEM for all meshing, the bottleneck is ICEM in this approach as Fluent solver is quite robust for layering. So the robustness of ic3m approach is as good or as bad as ICEM. With all the benefits of ic3m approach, it does come with an Archil’s heel. This approach in general does not handle complex geometry very well. A good quality mesh becomes very difficult when the geometry becomes complex. * As for robustness of IC3M approach, since ic3m replies on ICEM for all meshing, the bottleneck is ICEM in this approach as Fluent solver is quite robust for layering. So the robustness of ic3m approach is as good or as bad as ICEM. With all the benefits of ic3m approach, it does come with an Archil’s heel. This approach in general does not handle complex geometry very well. A good quality mesh becomes very difficult when the geometry becomes complex. * As for robustness of IC3M approach, since ic3m replies on ICEM for all meshing, the bottleneck is ICEM in