In CFD, is there any numerical disadvantage of using very fines meshes ? Can someone provide more insight on numerical dispersion and dissipation in this regard ?
Why i am getting fluctuations in, for example, heat transfer or pressure drop(between inlet outlet in a channel flow for conjugate heat transfer ) with iterations in steady flow for a fine mesh but no such fluctuations when the mesh is coarse .
Thanks*
Because your artificial diffusion kills off any oscillations and probably destroys all accuracy on the coarse mesh. If a scheme does not work on a fine mesh, it does not work on a coarse mesh either. (Producing a nice non-oscillating colour plot does not count as "workiing".) A scheme has to be stable and accurate in order to provide useful answers. In CFD, boundary conditions are often the culprit.
I think the problem is with your boundary conditions not probably meshing (fine/coarse). Furthermore please check with wall function, relaxation factors, tolerance
Because your artificial diffusion kills off any oscillations and probably destroys all accuracy on the coarse mesh. If a scheme does not work on a fine mesh, it does not work on a coarse mesh either. (Producing a nice non-oscillating colour plot does not count as "workiing".) A scheme has to be stable and accurate in order to provide useful answers. In CFD, boundary conditions are often the culprit.
Thanks all for your inputs. Since my problem involves vortex shedding, i am trying unsteady simulations for the same problem. My steady simulations are behaving like quasi steady and will change the relaxation factors in the steady simulation to see the effect.
- Badges
- Science topic