[tusin111]

hello! I'm new to using openFoam!

I have one question for you.

In the figure, convergence occurs when the cylinder at the top is used as the entrance and the speed is given downward.

however

If the cylinder on the top side is used as the same entrance and the air is released by speeding upward, convergence will not occur.

Do I need to modify the initial conditions or anything else in this process??

[calim_cfd]

Hi.
Try setting the initial velocity within the fluid either the up or down direction, according to the problem.. this should aid convergence. If it still crashes, try checking your outflow boundaries, as well as gravity terms, if they matter for you problem... if it still crashes, then you have a mesh issue...




good luck!

[tusin111]

Originally Posted by calim_cfd

Hi.
Try setting the initial velocity within the fluid either the up or down direction, according to the problem.. this should aid convergence. If it still crashes, try checking your outflow boundaries, as well as gravity terms, if they matter for you problem... if it still crashes, then you have a mesh issue...




good luck!

thank you for the reply

Now my U and P are in the same condition as the picture

Solved with kOmegaSST. but can't converge

I never considered the gravitational term. How do I check the gravity term?

It is currently in its initial state

There are U, P, nut, k, and omega.

[tusin111]

Originally Posted by calim_cfd

Hi.
Try setting the initial velocity within the fluid either the up or down direction, according to the problem.. this should aid convergence. If it still crashes, try checking your outflow boundaries, as well as gravity terms, if they matter for you problem... if it still crashes, then you have a mesh issue...




good luck!

I have one question!!!

I converged

SIMPLE
{
nNonOrthogonalCorrectors 0;
consistent yes;

residualControl
{
p 1e-4;
U 1e-4;
"(k|omega)" 1e-4;
}
}

relaxationFactors
{
equations
{
U 0.001;
k 0.004;
omega 0.005;
}
}


Relaxation coefficients higher than that did not converge until 1.e-4.

What are the problems with using very small relaxation coefficients?
Or is it okay to interpret it this way?

[calim_cfd]

Originally Posted by tusin111

I have one question!!!

I converged

SIMPLE
{
nNonOrthogonalCorrectors 0;
consistent yes;

residualControl
{
p 1e-4;
U 1e-4;
"(k|omega)" 1e-4;
}
}

relaxationFactors
{
equations
{
U 0.001;
k 0.004;
omega 0.005;
}
}


Relaxation coefficients higher than that did not converge until 1.e-4.

What are the problems with using very small relaxation coefficients?
Or is it okay to interpret it this way?

You seem a bit lost overall... to relax your solution this much usually indicates a problem or a potential one.... but sometimes this is allowed... but lowering all factors is danger because you may "finish you simulation" but it might not have converged to the actual solution because you damped the solution so much that it eventually stopped and thus you get an intermediate result.... to see this.. check the solution data and see whether the problem actually ended.


regards,