Learning Forum

[dhaya400]

Hi,

I am using Windkessel model to apply pressure outlet boundary condition for transient, turbulent flow simulation in Ansys Fluent 2020R2. The geometry is patient specific aorta, which consists of 2 inlets and 7 outlets. At the inlet, mass flow BC is applied and at the outlets, pressure outlet with windkessel udf is applied.

While running the simulation, at few outlets, reverse flow is happening. This reverse flow affects the mass flow rate which I calculated at the end of the each time step for windkessel udf.

So the error in mass flow rate transfer to the outlet pressure value and it cause the simulation to diverge.

Would anyone help me to solve this issue?

with regards,

DK

[Rob]

Have a look at the flow field to understand why you have reverse flow from there you may need to rethink your boundary conditions. A few images of what you're modelling would help.

[dhaya400]

Hi Rob Thank you for your reply. By geometry look like below.
At the inlet 1, applying mass flow inlet which is a periodic pulse.
At the inlet2, applying constant mass flow rate.
At all the outlets, applying pressure outlet condition (attached below)
I am observing the reverse flow at the outlet BCA, LSCA, LCCA (almost 100% reverse flow).

[Rob]

I hadn't realised there was a return flow on the aortic arch, I thought RIA & LIA would be heading towards the kidneys/legs?
Plot the pressure on the artery wall. I suspect the flow patterns are such that the main flow from Inlet1 will want to entrain fluid from the top three outlets.

[dhaya400]

Hi Rob Under normal condition, the blood flow form inlet 1 to all the outlets (including LIA). Here the patient is under ECMO support. So the blood is supplied using external pump via LIA to all outlets.
Contour plot shown below:

[Rob]

OK. check the flow in the regions with reverse flow. What are you basing the back pressure values on?

[dhaya400]

Hi Rob The velocity contours for outlets shown below
The inlet1 mass flow profile shown below

Do you want to see the pressure contour?

[Rob]

I think the clue is in the velocity field. I'm not a medic but suspect if blood is flowing at 125m/s then reverse flow is the least of your worries. Have a proper look at the boundary values, mesh and convergence. I think at least one of those needs some work.

[ALI70]

Hi dhaya400, can I have your UDF? i AM WORKING ON A PROJECT LIKE YOU

[pen m]

Hi, have you found the code for the Windkessel model? Can you give me your UDF code for reference? My email is 972539894@qq.com, thank you.

[pen m]

Hello, I can refer to the code of your Windkessel model. There are some problems in the code I wrote. If you can, please send it to my email 972539894@qq.com, thank you.