For hexahedral meshing, it is quite simple to divide one edge in a particular number of division. But for tetrahedral meshing it is not that simple. If we are using a complicated domain for aerodynamic flow and if it is difficult for doing the hexahedral meshing, then how to divide the edge in element in that fashion and what thing must be kept in mind while doing the grid convergence study.
I would like to suggest a good lecture for edge detection by Prof Dr Muubarak Shah: https://www.youtube.com/watch?v=lC-IrZsdTrw
See also https://www.crcv.ucf.edu/wp-content/uploads/2019/03/CAP5415_Fall2013_Lecture-3-EdgeDetection.pdf
Dear Prasenjit,
Actually it is important to increase the number of elements and check the results as well, the variation of results based on the increased number of elements should be minimized, then the lower number of elements can be used in order to reduce the computational volume.
Milad Ghalamboran Dear sir, is it logical to distribute and divide the edges in the ratio of their length. Sometimes dividing all the side by ratio of their length also doesn't give proper results. Rather dividing the edges in their length ratio and then adding some more elements in the shorter edges is proved out to be beneficial. So is there any logic or works behind this for tetrahedral meshing??
Basically it depends on the lowest length and it is better to choose equally elements or even finer I mean smaller division, which will increase your computational volume.
Mr.Prasenjit, i feel for correcting your problem it is better to go with smaller size element with fine mesh. Also, your problem could be corrected with aspect ratio 1, many researchers mentioned that aspect ratio 1 would give better results with close converging with exact results. I mean to say instead of refining longer or shorter direction it is better to use length to width ratio of element is 1 for getting close results..ok all the best in your research..
- Badges
- Science topic