hi.

I'm simulating 3d duct flow. You can see the geometry through these links:

http://bp0.blogger.com/_bLa3MZKMUtw/...00-h/test2.bmp

http://bp1.blogger.com/_bLa3MZKMUtw/...00-h/test3.bmp

I have three volume. The duct (meshed in the picture), small volume inside the duct and big volume covering the duct. I simulate only half since the duct is symmetric.

I tried to use size funtion. I couldn't create the transition from fine grid near the duct's wall to coarser grid at z+ direction.

Any idea is greatly appreciated....

Thanks

Hey

Since the two sides of the duct are not meshed, it is impossible to use the size function in this case. Finish meshing all the sides of the duct. Then mesh the volume covering the duct with the size function, starting the mesh from all the three faces of the duct.

If this doesn't help, sub-divide the large volume into three volumes (slices), with the common two faces coinsiding with those of the small duct. Then use the size function.

Regards AAA

hi, thanks for your answer.

This picture is only to show the geometry. I tried to mesh the three sides of the duct, but the result is fine grid at the Y+ only and not to z+.

Can you explain me more about the slice. I already tried to divide the volume by adding to small volume at the right and left side of the inner volume (small volume inside the duct), but it still difficult to mesh the big volume.

thanks

only a mesh size function requires pre meshing. Other size functions like curvature, proximity and fixed - no pre meshing is required.

Hi

Meshing the large volume with a size function starting from the faces of the duct outwards should be easy, given that you specify the proper growth factor (<1.2). Try using a growth factor closer to 1 (say 1.05) and see if you get a fine mesh in the z direction. What are the values you are using?

I've encountered problems in the past while using a size function and having to start the meshing process from two faces that make an angle larger than 180 deg. This is why I suggested to mesh the faces of the duct (which has a 270 deg angle between the faces). I hope it will work for you too.

Regards

AAA

Hi. thanks again for the reply.

Right now I'm using growth ratio=1.2 and max size 20. i tried to decrease them but i'm afraid that my computer can't run out of memory.

I meshed the duct (outer, inner wall) and also the small volume below the duct. then i used the right+left side of the small volume+ outer wall of the duct to mesh the big volume using meshed size function.

the result is only good for y+ and not z+ direction.

thanks

Hi. thanks again for the reply.

Right now I'm using growth ratio=1.2 and max size 20. i tried to decrease them but i'm afraid that my computer will run out of memory.

I meshed the duct (outer, inner wall) and also the small volume below the duct. then i used the right+left side of the small volume+ outer wall of the duct to mesh the big volume using meshed size function.

the result is only good for y+ and not z+ direction.

What if i use other size function?

thanks

Hi

Based on your explaination, the only thing I can think of to have better control over the mesh in the Z-direction would be to sub-divide the domain.

Regards

AAA