CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Stable Boundary Conditions (

blackbody May 15, 2010 02:07

Stable Boundary Conditions
i'm trying to simulate a 1-1/5 stage turbine.
I want to use a profile BC at inlet.

But when I use T_tot, velocity x, velocity r, velocity theta, k and epsilon as an inlet profile and average static pressure as an outlet, the solver crashes:

"Fatal bounds error detected
variable: absolute pressure"

how can i use the velocities? because when i use them, i cant specify total pressure at inlet (which i would also have as a profiel data) anymore...

what is the way if i want to use T_tot, Velocity -x,-r,-theta, k and epsilon as an inlet profile BC??

Thank you very much!!!

ghorrocks May 15, 2010 07:01

Have you set the correct reference pressure? Is your boundary pressure correct relative to the reference pressure?

blackbody May 15, 2010 11:15

I set the reference pressure of the domain to "0 atm".
For the inlet I can't specify any pressure (i'm using velocity u, r, theta). There is no option where I can specify for egsample total pressure at inlet.
If I use unit velocity direction, than I can specify the pressure at inlet. but not withe the absolut values of the velocity....

At outlet I set 1.1 bar average static pressure...

but the solver crashes due to the BC...

What can I do?

Attesz May 16, 2010 12:37


the referece pressure should be set near your operating pressures. I think 0 atm is low for a turbine. You should set at least 1 atm. The relative pressure will be (1.1bar-1atm) of course.

And I think, you can not set total pressure and velocity at the inlet together, because the equations will be overconstrained.

blackbody May 17, 2010 09:35

so is this a correct way of setting BC:

inlet: velocity -u, -r, -theta, T_tot, k, epsilon
outlet: average static pressure

but when i use this, the solver crashes.....
what is wrong?

Attesz May 17, 2010 09:38

The crash can caused by many other problems! what is the error message?

blackbody May 17, 2010 09:45

i tried a lot of different BC setups with the same model, and they all worked..
now i wanted to use velocity components (from measurement) for the inlet.

when i use velocity components, no pressure at inlet can be deffined anymore...

1st try:

inlet: velocity components, t_tot, k, epsilon (profile BC)
outlet: massflow

-> error!

2nd try:
inlet: velocity components, t_tot, k, epsilon (profile BC)
outlet: average static pressure

-> same error message:

"Fatal bounds error detected
variable: absolute pressure"

what can i do if i want to use velocity componets at inlet?

thank you

Attesz May 17, 2010 09:49

The BC setting seems to be OK. Check your values, units etc. I have no more idea.

ghorrocks May 17, 2010 18:46

The error message is simply saying you are trying to get a negative absolute pressure somehow. Based on the sketchy information you have provided I have no idea where, but somewhere in your setup you are asking for a negative absolute pressure.

blackbody May 18, 2010 13:03

I set the reference pressure now to 1bar and the outlet average static pressure to 1.1-1=0.1 bar and it worked...

but don't really know why...

anyhow, thank you guys!

ghorrocks May 18, 2010 19:14

Numerical round-off. If you ran using double precision it probably would have worked. But the better solution is to set a reference pressure more representative of the flow average pressure, which is what you have done.

ghorrocks May 18, 2010 19:23

Trust me, the problem is caused by numerical round off leading to a convergence problem.

All times are GMT -4. The time now is 11:36.