[sajjan]

Hello Fluent users,

I am trying to simulate a heat transfer phenomenon in a pipe. I apply velocity-inlet and pressure-outlet as the inlet and outlet boundary conditions respectively.

Nature of problem is that the velocity and the temperature at the inlet changes with time. It is not periodic either. I use a transient-table to define inlet velocity and tempetature profile.

I tried with different solution schemes available in fluent, changing relaxation factors, using small time-steps, etc. But the continuity diverges all the time. Even with small change in inlet velocity has a huge impact in continuity residuals.

I wonder if anyone has gone through a similar situation before.

Many thanks!
Sajjan

[AlexanderZ]

there are too many reasons which could lead to divergence
most likely your mesh is not good enough
check material properties and models you are using
check other boundary conditions

are you able to get converged solution in case of steady state boundary conditions?

[sajjan]

Originally Posted by AlexanderZ

there are too many reasons which could lead to divergence
most likely your mesh is not good enough
check material properties and models you are using
check other boundary conditions

are you able to get converged solution in case of steady state boundary conditions?

Hi Alex,

Thank you so much for the reply. My mesh quality is very good. Its a pipe flow and I generated the mesh using ICEM with O-grid. Even a worst element has a quality above 0.7. In addition, flow is perfectly aligned with the mesh. And yes, the flow gets converged with fixed inlet velocity and temperature. So, I believe, there should not be a problem with material properties or other boundary conditions.
The problem is, continuity residuals increase alot and ultimately the solution diverges with floating point error, even with a small change in the inlet velocity.
Thank you.

[AlexanderZ]

check your mesh size, are you sure you are using proper units?

small changes in BC condition should not affect that match in case of flows in pipe

definitely you have problem in setup (check material properties, if they are not default)