[llalra]

I constructed CFD transient simulations of human blood vessel to estimate the wall shear stress with turbulence models (k-omega) in CFX. However, no matter how I adjusted the mesh or modified the boundary conditions, convergence failed all the time, and the residual graph is shown in the figure below. How to solve this situation?
Detailed information:
1. The quality of meshes in the vessel were higher than 0.5 and the number of the mesh was about 2 million.
2. The inlet (velocity) and outlet (pressure) condition were based on the normal human measured value in a cardiac cycle, including density and viscosity of blood (constant). The wall was vessel wall which was no-slip.
3. Time duration was a cardiac cycle (0.8s), and timestep was 0.008s.
4. RMS<10e-6
Figures:
rp.png : The residual plot
vessel.png: the model of vessel (reconstructed by CTA)

[FrankW]

Your vessel size is a little to large. I do not know any creature with 25m vessel diameter . Maybe the right size could solve your problem.

[Opaque]

Good catch!!!

Just for context, the largest Francis runner in the world is @11 m in diameter. Not sure about the diameter for the draft tube of such a machine, but I doubt you can fit two runners side by side.

[llalra]

Originally Posted by FrankW

Your vessel size is a little to large. I do not know any creature with 25m vessel diameter . Maybe the right size could solve your problem.

Thank you for your point out my defects. The problem has been solved.
Kind regard,
llalra

[llalra]

Originally Posted by Opaque

Good catch!!!

Just for context, the largest Francis runner in the world is @11 m in diameter. Not sure about the diameter for the draft tube of such a machine, but I doubt you can fit two runners side by side.

Thank you. I forgot to change my unit. The problem has been solved.
Kind Regards,
llalra