Bifurcation explode at intersection
 

Hi everyone,

I am simulation the blood flow in the bifurcation( I used Gmsh to mesh the bifurcation and import it to Openfoam). When I icoFoam it, the courant number always explode at the intesecation which you can view it one the picutures below.

It seems like the velocity is reflected by the sharp corner and pierce the wall then cause it explode.

I am quite new to Openfoam and not sure whether it is because my mesh or because of other reasons.

Can anyone point out what's going wrong?

Attachment 8189
Attachment 8189

Attachment 8190

My p and U file
 

p file
dimensions [0 2 -2 0 0 0 0];
internalField uniform 0;
boundaryField
{
wall
{
type zeroGradient;
}
inlet
{
type zeroGradient;
}
bent_outlet
{
type fixedValue;
value uniform 0;
}
straight_outlet
{
type fixedValue;
value uniform 0;
}
}
U file
dimensions [0 1 -1 0 0 0 0];

internalField uniform (0 0 0);

boundaryField
{
wall
{
type fixedValue;
value uniform (0 0 0);
}
inlet
{
type fixedValue;
value uniform (0 0 0.05);
}
bent_outlet
{
type zeroGradient;
}
straight_outlet
{
type zeroGradient;
}
}

Can you check whether there's any "funny" cell in that region?
I would open the mesh in paraview (or whichever software you prefer) and do some slices...

Hi lore,

Thanks for replying my post.

I tried to use paraview to slice it. First of all , there is no hole in that region and I didn't find any wired cells in that region.

Could other reasons lead to explode?

Can you please post a picture of that region? just the mesh.

Attachment 8195

Attachment 8196

Attachment 8197

Attachment 8198

Attachment 8199

Attachment 8197Hi lore,
Here are some pictures of the mesh, please let me know if you need more.
thanks

Attachment 8200
and this thanks

Could you please post a slice of the region where you're having the blown up?

Attachment 8202

Attachment 8203

Attachment 8204

Attachment 8205

Hi lore,

It is quite hard to slice the exact point where it blow up, but here are some mesh slice that is the blow up region. hope it helps .

If you want, I can send you my file which is too large to upload here.

Cheers

Guifan, you should slice the pipe with a zx plane, ie a plane wich lies in the flow direction.

Hi lore, Sorry for the misunderstanding.

Attachment 8206

Attachment 8207

Attachment 8208

Mmm, your mesh seems to be ok...Maybe you should wait for somone with more experience than me...

lore, thanks for that anyway

Have you tried changing the time step?

Hi lore,

I tried to change the time step from 0.0001 to 0.000001 and it works. But now I reduced the element sizes from 150,000 to around 60,000 with time step 0.0001. It works as well.

Sorry for the late reply.

Yes, that's what I meant: in both cases you changed the Courant number. I only mentioned the time step and not the cell size because I thought you wanted to keep the mesh the same.