I have got a problem with the measurement of drag and lift forces using grid refinement on the problem of a flow through a cylinder (2D).
When information is transferred from the coarse grid to the fine grid, and there is no transfer from fine grid to coarse grid the solution obtained is accurate. The forces are measured on the fine grid.
But adding the transferring of information from fine grid to the coarse grid got worst results.
Why could this be happening??. It supposed the information of the coarse nodes on the interface between grids are corrected with the fine ones. I’ve heard that there is need to filter the population densities of the fine grid before correct the coarse nodes on the interfaces. Is this the reason?. Any suggestions?
i am using a multi block approach to transfer densities populations.However I have a coarse grid that covers entire domain, and a patch that is over the cylinder, that’s why i can avoid to transfer from fine patch to the global coarse grid.
Sorry for my poor English.