[kopaka]

Hello everyone,

I am currently carrying a simulation of a 3D Poiseuille flow in a square tube with additional source terms. My simulation needs to have a fully developed Poiseuille as initial conditions. To have these initial condition, I tried to run the simulation without my additional source terms and waited for the steady state.

The trouble is that after several hours and 20 000 iterations, the convergence is still not reached. (see picture 1 attached).

The velocity profiles on different sections should be exactly the same if the steady state is reached am I right ? But here I still have different profiles in different sections. (see picture 2).

So that's why I was wondering if there is a way to speed up the simulation to have a steady state more quickly, or if I could implement a steady state inlet condition, but I think it is quite complicated for a 3D square tube Poiseuille because of the complexity of the analytical solution.

I have a 1 million nodes mesh, I use SIMPLE algorithm and velocity-inlet, pressure-outlet BCs.

Thank you by advance

Simon

[raghulvr]

I think your solution is converged because the residuals are showing E-12.which is very good in cfd. you using simple algorithm,then what are the values of pressure correction factors and other related flow parameters?
sometimes making the mesh too fine can be a result of this but i recommend you to go for SIMPLEC algorithm as it can give better accuracy in tems of solutions

[kopaka]

I solved my problem by applying the outlet velocity profile after 10 000 iteration to my inlet profile. Now I have a uniform field (begginer mistake)

Thank you !