Hi,
I have been frequently using the the tetra remeshing function to remesh the deformed meshes created by FE analysis in order to carry out subsequent analyses. However, some parts of the mesh gets finer and finer after every time being remeshed (as shown in the pic). These element concentrations are undesired and they largely increases computation cost. Also, the degree of concentration increases rapidly with the number of remesh process being applied which eventually cause difficulties in carrying out efficient FE analysis.
I wonder if there is any function available to coarsen these small elements? I can accept the lost in boundary geometry accuracy. I have tried limiting the min element size for remeshing, this problem is mitigated but can't be completely solved - extremely small elements still occur after a large amount of remesh processes. I suspect these concentration in elements are created by HM in order to exactly track the deformed mesh - but I can't find any 'feature angle' parameters that can be adjusted to ignore certain geometry features of the deformed meshes. Any alternative solutions?
picture: https://drive.google.com/file/d/0B8pqXnKPKIghWlpMY21MYllKWkk/view?usp=sharing
Many thanks,
Wensong
P.S. Here is another picture showing the remesh result after roughly 20 FE analysis steps with HM remeshing in between of each step.
https://drive.google.com/file/d/0B8pqXnKPKIghb2hfVkduNWZVWkk/view?usp=sharing