[agungwp]

hi guys,
we know that there are some various grid topology we could used. O, C, and H type of grid topologies. but untill now, i have met the advantage or disadvantage not yet. could you describe to me what is advantage and disadvantage each grid topologies above?

thanks,
nidji

[agungwp]

anyone could help me?

[PSYMN]

In what situation…?

These are all tools in a tool box, the right one depends on your job. You may even want a combination of these to really optimize your mesh.

The H grid is often fine if your model has all sharp corners (like a box), because all the block corners can project to geometry corners and give you good quality.

But if you apply an H Grid to something without clear corners for the domain corners, your domain corners and their adjancent nodes will all be projecting to the same surface (think a block mapped to a circle…) The elements in the “corners” have internal angles approaching 180 degrees and it only gets worse with refinement.

The choice of O Grid, C Grid, Y Grid, etc. all depends on how many block corners need to project to geometry without a corner…

Maybe that is not the clearest answer, but it was a pretty vague question.

[PSYMN]

Sorry, I missed that this was a Gambit question…

I was thinking ICEM CFD when I went to grab a couple illustrative pics, but I guess the same logic applies.

The first shows an Hgrid applied to a circle so you can see the problem in the corners… Then how the OGrid solves the problem. A C-Grid would be half the ogrid (Half the circle, two sharp corners). A Y Grid would be one quarter of the Ogrid (Quarter circle, three sharp corners).

The second pic shows ICEM CFD Hexa with an automatic CGrid around this real geometry… It also shows that ogrids can be mixed and matched and intersected, note the CGrid coming thru the Mercedes logo… ICEM CFD Hexa is very flexible.