Geometry for FSI in blood vessel

Marten · June 9, 2010, 07:14

Hi everyone,

I´m trying to set up an FSI analysis in a blood vessel. For this I designed 2 parts. one is the arterywall and one the fluid domain. the only connection between those two is the fluid structur interface. Also I´m not sure about setting any support for the vessel (fixed support on both endings???,elastic support, but with which elasticity?).
when I run the simulation like in the tutorial in CFX with the oscillating plate I get an error after the first 10 iterations, because fluid can leak out of the domain, which is why I think I made a mistake with the connection of the two parts.
has someone any help for me?

kind regards
Marten

ahmad95 · May 21, 2016, 08:26

hello and how are you?
I m glad to speak to you my friend.
I m at the start point which you past later.
I m starting to model a vessel with a stenosis and the I want analyse it in ansys-cfx.
can you help me and show me the way.
where you modeled your geometry?
where you meshed it?
how did you import your geometry in cfx?
do you have an example of this process?
which I mean by process is that:
model a geometry in a software like catia then meshing it and at the end analysis it in another software.

Marten · May 21, 2016, 16:01

Hi,
I am fine,thanks. How about you?
This was some time ago, but I try to remember.

I tried different approaches. The blood vessel models were supported in stl format from patient scanned with CT. So the basis was a triangle surface. I tried using ICEM CFD which has probably the best mesh quality at least in comparison to different freeware.
I used prism layers growing from the stl-surface and then filled the rest of the volume with Tetraeder. This was leading to meshes with really high cell numbers.
Better approach was to use different blocks so in the middle of the vessle a square and 4 blocks around this central one (butterfly) with hexahedral elements. This was reducing the number of cells a lot.
You can also use the ANSYS Meshing as starting point. Just to get a result to begin with.
For the fluid structure interaction it was important to get a solid and a fluid region. So I just took parts of the prism layer and defined them as solid depending on how thick the vessel should be. In ANSYS you can then use the structual solver to handle this domain and define the face in between both regions as interface to CFX.

Stranglely for transient calculations there was lots of divergence when using to small time step. So play around with this if you have problems to get to converged solutions.

Maybe start with a simplified case first before you use the patient geometry. This will help you get working setup before increasing the complexity.

I used a simplified transient inlet velocity profile and zero presure outlet and no slip walls.

Wish you good luck with your work.
Marten

June 9, 2010, 07:14	Geometry for FSI in blood vessel	#1
Marten New Member Marten Join Date: Jun 2010 Posts: 7 Rep Power: 15	Hi everyone, I´m trying to set up an FSI analysis in a blood vessel. For this I designed 2 parts. one is the arterywall and one the fluid domain. the only connection between those two is the fluid structur interface. Also I´m not sure about setting any support for the vessel (fixed support on both endings???,elastic support, but with which elasticity?). when I run the simulation like in the tutorial in CFX with the oscillating plate I get an error after the first 10 iterations, because fluid can leak out of the domain, which is why I think I made a mistake with the connection of the two parts. has someone any help for me? kind regards Marten

May 21, 2016, 08:26	blood vessel	#2
ahmad95 New Member ahmad Join Date: Jun 2013 Posts: 11 Rep Power: 12	hello and how are you? I m glad to speak to you my friend. I m at the start point which you past later. I m starting to model a vessel with a stenosis and the I want analyse it in ansys-cfx. can you help me and show me the way. where you modeled your geometry? where you meshed it? how did you import your geometry in cfx? do you have an example of this process? which I mean by process is that: model a geometry in a software like catia then meshing it and at the end analysis it in another software.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
FSI modeling of blood vessel???	smn	CFX	20	June 2, 2019 05:46
[Gmsh] Problem meshing an imported geometry	Alicia	OpenFOAM Meshing & Mesh Conversion	0	March 30, 2010 04:53
ccl for fsi modeling of blood vessel	smn	CFX	0	July 19, 2009 05:36
Simulation of Flow through Complex 3D Geometry	EmersonKB	CFX	5	July 2, 2009 08:17
vitual _ real	deneb	FLUENT	3	January 22, 2007 04:31

May 21, 2016, 16:01		#3
Marten New Member Marten Join Date: Jun 2010 Posts: 7 Rep Power: 15	Hi, I am fine,thanks. How about you? This was some time ago, but I try to remember. I tried different approaches. The blood vessel models were supported in stl format from patient scanned with CT. So the basis was a triangle surface. I tried using ICEM CFD which has probably the best mesh quality at least in comparison to different freeware. I used prism layers growing from the stl-surface and then filled the rest of the volume with Tetraeder. This was leading to meshes with really high cell numbers. Better approach was to use different blocks so in the middle of the vessle a square and 4 blocks around this central one (butterfly) with hexahedral elements. This was reducing the number of cells a lot. You can also use the ANSYS Meshing as starting point. Just to get a result to begin with. For the fluid structure interaction it was important to get a solid and a fluid region. So I just took parts of the prism layer and defined them as solid depending on how thick the vessel should be. In ANSYS you can then use the structual solver to handle this domain and define the face in between both regions as interface to CFX. Stranglely for transient calculations there was lots of divergence when using to small time step. So play around with this if you have problems to get to converged solutions. Maybe start with a simplified case first before you use the patient geometry. This will help you get working setup before increasing the complexity. I used a simplified transient inlet velocity profile and zero presure outlet and no slip walls. Wish you good luck with your work. Marten