[Kwiaci]

I have conducted two simulations of the steady flow in a pipe with some obstacles inside.
First simulation was in steady state and second was in unsteady state with 0.05 second step and long simulation time was used. Flow generally is laminar but gets turbulent in vicinity of obstacles.
The problem is that I get two different solutions. Contours of velocity magnitude are in attachments.

Does anyone know why these solutions are different?

[siara817]

It is hard to decide, but I think the right picture seems that it is not converged very well, since there is high velocity at top left of the picture. I think you need to let it to converge more. What is the reason that you are comparing the steady state and unsteady case?

[Kwiaci]

Thanks,
I will converge it more.

I wanted to check if there are any vortices separating from the edges which disappear before the outlet - but there are not there. This was just an additional analysis.

[Far]

you cannot compare two solutions directly. Take some average for the transient solution and then compare it to steady state.

[Kwiaci]

OK, after it converged more I got almost identical results. Thx

Far, you have mentioned about taking some average for the transient solution. How it can be done in Fluent to obtain, for example, time-averaged heat flux on a wall?