[macfly]

I'm using ansys fluent 12.1.2.

Is there a way to refine the mesh without any adaption? Refine the mesh over the entire domain independently of the solution ?

Why would I want to do that? I'm modeling turbulence in a complex geometry and I'm having a hard time with the y+ being in the acceptable range everywhere. I want to start with a coarse mesh, refine it in Fluent, this way I can coarsen back the mesh if my y+ are too low in some places.

[jack1980]

There is. But this method is limited to a integer factor of refinement (2 is default). Do the following:

- let your model converge on the coarse mesh
- adapt > region : enter a very large region that includes your complete domain (like x from -100 to 100 etc) and click 'Adapt'
- continu your calcs on the now 2x refined mesh
- check your y+

hope it helps, good luck!

[macfly]

thanks Jack,

It works just fine for refining the mesh. And it is possible to coarse the mesh back.

[echelon]

Hi all,

I am a new user to the world of CFD and FLUENT and am trying to refine my mesh using the grid adaption feature in FLUENT based on region.

My model consists of an ellipse (representing a idealised wing) surrounded by a boundary layer consisting of quadrilateral cells, which in turn, is surrounded by triangular cells that form an O-shaped grid. A picture of the initial mesh is attached.

Currently I have arround 60,000 cells but want to refine my mesh for grid-independence studies. As established above, I am attempting to use the grid-adaption feature within FLUENT based on region. Specifically I:

• Imported the mesh into FLUENT
• Selected adapt->region->
• selected 'circle' radio button under shapes
• Inputted X center=Y center=0 and radius of 0.3 (which extends to the far field of my mesh) as the input coordinates
• selected the 'controls' button -> selected 'refine' and selected the 'maximum level of refine' as 2 (later I tried 3,4 etc as described below)
• Then selected 'ok' and then 'adapt'

This increased the number of cells to roughly 200,000 cells and the mesh is more refined. However, no matter how I change the 'maximum level of refine' value thereafter, I always end up with the same number of refined cells (roughly 200,000).

From a grid independence study point of view, where I should gradually increase the mesh resolution and see how much the solution (I am measuring lift and drag values generated by the wing btw) changes, I clearly need a few scenarios with differing mesh resolutions. At present I am at a loss on how to achieve this.

Any advise or comments on this would be greatly appreciated.

Thanks and Regards
Arnab