MRFzone problem in ICEM,
 

Hi Foamers:

Now I am facing a real problem.

I need to generate two structured mesh seperately via ICEM, the inner is rotate zone, the outer is stationary zone. both of them are filled with water, expect that a impeller exists in the rotate zone.
the part of stationary is walls,outlet,slide. part of rotate zone is impeller and slide. you can see that slide is actually a interior face.
after I generate it seperately,I check the mesh quality is above 0.7.

Then I open a new project and open the rotate mesh(1.uns), and then open the stationary mesh(2.uns) and select merge, so now I have two meshes. but no geometry. after this I set the solver is fluent_v6, and set the boundary condition of slide is interior. the fluent3DMeshtofoam, but in boundary dictionay there is always a int_SLIDE
{
type patch;
nFaces 9500;
startFace 216163;
}

I have to set bc of int_slide in U and P and else, but actually its an interior face which dont need to set BC.

I have tryied unstructured mesh, set up two bodies called rotate and stationary, that works and dont show the interior face. but in structured mesh its not the same way.

anyone konw how to handle this? look forward to your reply, thanks in advance.

BTW if the slide part is included in the TWO SIDED, just as the image showes, it wont appear in the boundary dictionay.

but I forget how to do this....

up
 

up..........................

I even tried delete the slide mesh , I got this........

woo...I am stucked by this.

okay guys I tried to interpret this in a simple way.

just as the image, I cycinder with a inlet outlet and walls,

after fluent3Dmeshtofoam , the inner appear again....


its a very simple geometry, just using hybrid mesh.................help me plz....because if I use unstructured mesh totally, the solution blows up.
http://www.cfd-online.com/Forums/ope...tml#post396016

Is the interface of your two mesh regions conformal? That is, do each of the mesh faces match up exactly? OpenFOAM requires this unless you apply AMI or something else. I am unfamiliar with ICEM myself though there is a useful example for the same case using gambit
here. In that case, the two regions are joined at the interface and when the mesh is converted it will show up as a single volume with the cellZones already written for the MRF parts.

I have made MRF meshes in star-ccm+ as well--though not in some time. In this case if I recall correctly, the mesh regions convert as separate unconnected volumes which have to be connect post-conversion by using stitchMesh---this will complain if the two interface patches are not exactly conformal.

Hope something here is helpful.

I think you should use "interface" boundary type instead of "interior"!

about this I know how to handle the nodes problem.http://www.cfd-online.com/Forums/ope...tml#post404012. just use AMI. but I am still confused about some other things but I dont know where is going wrong.