Ansys Fluent analysis is used to analysis , a type of mesh plays a crucial role in the accuracy of the solution.
Forget the mesh type, stick to mesh metrics like orthogonality. As long as getting a good value, the mesh suits the simulation, disregard trimmer mesh always works.
It is not the mesh type, which ultimately determines the accuracy of a fluid simulation. But it is the mesh density dx/dy/dz. In the mathematical limit of dx-->0 all meshes of all different mesh types should lead to the same CFD solution, if the underlying CFD solver is implemented correctly. And a mesh refinement study should be used, to have a control about the involved discretization error of the obtained solution on the finest used mesh. And only well converged solutions should be analyzed. Not converged solutions need to be discarded.
Nevertheless different mesh types and in particular the mesh quality can make a great difference in terms of solver convergence behavior, solver robustness, speed of convergence (due to numerical disturbances coming into play because of e.g. bad mesh quality). Mesh quality can be well controlled by the user for certain mesh types (e.g. structured hex meshes) and cannot be well controlled for others (e.g. quasi-automatically produced tet/prism meshes). That is, what makes essentially the difference between mesh types.
Usually polyhedral meshes (with boundary prism layers) and hexahedral meshes with a good boundary layer resolution and resolved fluid gradients are doing a better job then tet/prism meshes or even such things like cut-cell or hex-dominant meshes with hanging nodes.
Regards,
Dr. Th. Frank.
mesh type to be used depends alot on the geometry in consideration. with that being said, it is wise to go for reasonably good mesh statistics (orthogonal quality, aspect ratio, skewness). this will largely ensure a good solution.
Use StarCCM and over set meshing that is much easier and manageable compared to Ansys.
All answers are relevant. But @Zahid Ahmed has pointed out important parameters. Based on my own experience, when I get good values of orthogonal quality and skweness, my simulations are more stable and I get more accurate results. There is a chart, I thing google images will show you one, with recommended values of these parameters. Do not forget to carry out a grid independence test, that is really important as well. Good luck.
its quite generic to answer. a good mesh depends on the actual process you want to simulate. some indicators like ortho quality, skewness and aspect ratio may give you a hint but dont guarantee anything. eventually, the areas of high slopes need to be meshed finer and far stream areas can be left relatively coarse.
- Badges
- Science topic