[dandes]

I execute the origional case of the tutorials ,it generates 1000 elements (u can investigate CH2.2 in the prommer's guide ). but the numbers of the rsults(include sigmaxx,sigmayy,sigmaxy) are much larger than 1000 . Even the value is unreasonable . I think the functuin of finiteFemFoam is not exact , right ? and can u tell me the arguments of the finiteFemFoam solver , because the origional Fvsolution dosen't have the finiteFemFoam codes(i compared with the stressedFoam case). sincerely ^^

[hjasak]

You have to be very careful - the FEM solver has been implemented by my colleague Dr. Zeljko Tukovic and myself and has been validated in detail in his PhD Thesis. The solver is correct and there are no implementation errors, especially not to the magnitude you describe.

My guess is that you either do not understand the results you are looking at (e.g. the mesh is 3-D) or that you have messed up the boundary conditions or the material properties.

Hrv

[dandes]

thx so much . But can u explain the numbers of the results ? 1000 elements should be versus 1000 values (like 1000 sigmaxx , 1000 sigmaxy ...etc), isn't it ? But my results contain 7212 values for each kind of stress , isn't it ? Hence , i got incorrest results when my answers convert to VTK form (i guess),so i can't present my paragragf on paraview . please provide the thoughts in detail for me , thank you very much .

[mattijs]

The FEM solver decomposes the cells into tetrahedra. You are probably seeing the fields on the decomposed mesh and postprocessing those is probably not yet added to paraFoam/foamToVTK. dxFoam does so see if you can get a 1.2 version of it or use 1.1.