[Lilly]

Hello,
I have got a problem with the type of grid I am using for my simulations. It would be really nice, if one of you could help me!
If I am using a structured hexahedral volume mesh my results differ a lot from the results for the case that I am using an unstructured tetrahedral volume mesh. The parameter I am interested in is the area weighted mass fraction of contrast agent in blood. I already tried using grids with a different number of elements (from about 50.000 up to a few million elements) for both cases. The results are converging in each case, but the results differ about two orders of magnitude comparing both grid types! Do you have any idea how I can fix this problem?
Thanks a million,
Lilly