[KushalJ]

Hello fellow CFD users,

I am a student and fairly new to using CFD analysis using CFX software. I am working on a FSI analysis of a pre-stressed artery with a pulsatile blood flow. After solving the first iteration for both Mechanical & CFX, I am consistently getting very large pressure developed at the outlet leading to a large force transferred to the structure. The large force causes excessive deformation of some elements resulting in non-converged Mechanical solution. The details of my analysis are:
- laminar flow
- artery wall modeled with low modulus material (E = 4 MPa)
- zero reference pressure in CFX
- pulsatile inlet velocity and outlet pressure. I am using a linear ramp-up for the velocity & pressure curves to achieve smooth transition to the pressure. The initial outlet pressure is 12 KPa and I am modeling a ramp such that the outlet pressure increases from 0.001 KPa to 12 KPa in 0.12 seconds. The rest of the pulsatile pressure curve starts at 0.12 seconds.
- The initial inlet velocity is 10 mm/s and I am modeling a ramp such that the inlet velocity increases from 10 mm/s to 24 mm/s in 0.12 seconds. The rest of the pulsatile inlet velocity curve starts at 0.12 seconds.

I have already gone through the relevant forum posts (e.g. http://www.cfd-online.com/Forums/cfx...-fsi-case.html) and have tried varying the parameters like Under Relaxation Factor, Number of Coefficient Loops, but without any luck.

Attached is the typical pressure spike that is generated. The pressure in the plot is in Pa.

I will appreciate a lot if someone can offer me advice to eliminate this issue of developing larger pressure in the fluid domain and the associate large force in structure.

Thank you,
Kushal

[KushalJ]

Guys, does anyone have any inputs on this issue?

Thank you.

[ghorrocks]

Have you tried smaller time step size?

Have you considered whether this is real - in other words, the material model you specify cannot handle the pressure applied to it and so will therefore grow rapidly. In the real world it will then fracture and break, but if you don't have a fracture model it will just keep growing forever.

[KushalJ]

yes, i have tried smaller steps of ~1E-5. The issue is that the force transferred to the structure is very high & is definitely unrealistic from the magnitudes of the pressure and dimensions of the blood vein.

[ghorrocks]

I see you are ramping both the inlet and outlet. Why are you doing that?

Most people would just ramp one and leave the other constant. I would ramp the inlet velocity and leave the outlet pressure constant.