CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   help with boundary profile in pressure inlet (http://www.cfd-online.com/Forums/fluent/119439-help-boundary-profile-pressure-inlet.html)

quiqui June 17, 2013 14:57

help with boundary profile in pressure inlet
 
1 Attachment(s)
Hi
I built a 2-d model, using pressure inlet and pressure outlet.
I have the experimental inlet boundary total pressure data(attached). I used these data for a boundary profile.

~~~~~~~~~~~~~~~~~~~~~~ I saved the mesh file, boundary profile and other conditions in a gmail account (fluent111, PIN:1234506789), if anyone interested, please go to have a look.~~~~~~~~~~~~~~~


the other settings are as below:
operating pressure : 0 psi
pressure inlet total pressure : profile
pressure inlet static pressure : 9.819 psi
pressure out static pressure : 14.7 psi
total temperature for all is 300 K

The problem is: when I choose inlet as compute from, it shows "_1.#IND" for x velocity, y velocity, turbulent kinetic energy, and turbulent dissipation rate, and "_1.#INF" for temperature.

I cannot run, it give me AMG error directly.

If I still use the boudanry profile as the pressure inlet total pressure, but input values for x velocity, y velocity, turbulent kinetic energy, and turbulent dissipation rate, and temperature for initialization, I can run it only when X velocity was set below 300 m/s, if it was set above 300m/s, it gave me the AMG divergence directly. But with 300m/s, the results is not right.

Anybody has any idea about this?
Thanks

-mAx- June 18, 2013 01:09

do you have 2 inlets with different settings?

quiqui June 18, 2013 09:37

no, I only have one inlet, one outlet.

quiqui June 18, 2013 09:46

2 Attachment(s)
attached is the sketch and profile data.

-mAx- June 18, 2013 09:54

I know that I always had some unstabilities with pressure-inlet/pressure-outlet.
Are you computing compressible?

RodriguezFatz June 18, 2013 10:06

Are you saying that your expect your result to depend on the initialization? Or what do you mean by "But with 300m/s, the results is not right."?

quiqui June 18, 2013 10:07

Quote:

Originally Posted by -mAx- (Post 434651)
I know that I always had some unstabilities with pressure-inlet/pressure-outlet.
Are you computing compressible?

yes, supersonic flow with Mach number 1.9 @ total pressure 70 psi, and total temperature 300 K.

RodriguezFatz June 18, 2013 10:10

Quote:

Originally Posted by quiqui (Post 434655)
yes, supersonic flow with Mach number 1.9 @ total pressure 70 psi, and total temperature 300 K.

In your initial post you write:
operating pressure : 0 psi
Shouldn't that be 70 then?

-mAx- June 18, 2013 10:11

Quote:

Originally Posted by quiqui (Post 434655)
yes, supersonic flow with Mach number 1.9 @ total pressure 70 psi, and total temperature 300 K.

Ok then try to compute first without profile for checking if your setup is ok.
One stupid question: on your profile, I see you have points with total pressure = 0 psi --> :confused:

quiqui June 18, 2013 10:13

Quote:

Originally Posted by RodriguezFatz (Post 434654)
Are you saying that your expect your result to depend on the initialization? Or what do you mean by "But with 300m/s, the results is not right."?

If I initialized the x velocity with 300 m/s, the computing velocity result cannot reach the actual velocity.
In experiments, with total pressure @ 70 psi, total temperature 300k, static pressure 9.819psi @ inlet, the velocity is M=1.9, which is around 500 m/s.

without importing the data profile, the computing results are similar with experimental, but once importing the boundary total pressure data profile, all the problems I mentioned above appeared.

quiqui June 18, 2013 10:16

Quote:

Originally Posted by -mAx- (Post 434658)
Ok then try to compute first without profile for checking if your setup is ok.
One stupid question: on your profile, I see you have points with total pressure = 0 psi --> :confused:

yes, there is 0 psi, which is @ the wall.

I tried running without profile, it gave the experimental trend, but after some running, I import the profile, It gave me the error immediately.

quiqui June 18, 2013 10:18

Quote:

Originally Posted by RodriguezFatz (Post 434656)
In your initial post you write:
operating pressure : 0 psi
Shouldn't that be 70 then?

in supersonic case, operating pressure is 0.
70 psi is the total pressure.

Far June 18, 2013 10:18

Quote:

Originally Posted by quiqui (Post 434492)
Hi
I built a 2-d model, using pressure inlet and pressure outlet.

I have the experimental inlet boundary total pressure data(attached). I used these data for a boundary profile.

the other settings are as below:
operating pressure : 0 psi
pressure inlet total pressure : profile
pressure inlet static pressure : 9.819 psi
pressure out static pressure : 14.7 psi
total temperature for all is 300 K

The problem is: when I choose inlet as compute from, it shows "_1.#IND" for x velocity, y velocity, turbulent kinetic energy, and turbulent dissipation rate, and "_1.#INF" for temperature.

I cannot run, it give me AMG error directly.

If I still use the boudanry profile as the pressure inlet total pressure, but input values for x velocity, y velocity, turbulent kinetic energy, and turbulent dissipation rate, and temperature for initialization, I can run it only when X velocity was set below 300 m/s, if it was set above 300m/s, it gave me the AMG divergence directly. But with 300m/s, the results is not right.

Anybody has any idea about this?
Thanks

can you share your geometry, mesh and pressure profile?

quiqui June 18, 2013 10:21

Quote:

Originally Posted by Far (Post 434664)
can you share your geometry, mesh and pressure profile?

to capture the shock with separation, the mesh is very fine, so I only posted the geometry.

-mAx- June 18, 2013 10:22

Quote:

Originally Posted by quiqui (Post 434662)
yes, there is 0 psi, which is @ the wall.

I tried running without profile, it gave the experimental trend, but after some running, I import the profile, It gave me the error immediately.

How can you have total pressure 0 psi at wall (at inlet)?
At wall you have no velocity (if stationnary wall), which means no dymamic pressure. But the static pressure should remain. (ptot= pstatic + pdynamic = pstatic + 0 = pstatic )

quiqui June 18, 2013 10:31

Quote:

Originally Posted by -mAx- (Post 434666)
How can you have total pressure 0 psi at wall (at inlet)?
At wall you have no velocity (if stationnary wall), which means no dymamic pressure. But the static pressure should remain. (ptot= pstatic + pdynamic = pstatic + 0 = pstatic )

yes, I notice this. I will remove the o points and try.

quiqui June 18, 2013 10:35

Quote:

Originally Posted by -mAx- (Post 434666)
How can you have total pressure 0 psi at wall (at inlet)?
At wall you have no velocity (if stationnary wall), which means no dymamic pressure. But the static pressure should remain. (ptot= pstatic + pdynamic = pstatic + 0 = pstatic )

I revised the o psi to the static pressure and tried, it gave the same error when compute from inlet.
still cannot run the case

Far June 18, 2013 10:51

If you can share your mesh, then we can try to rectify problem...

quiqui June 18, 2013 11:12

Quote:

Originally Posted by Far (Post 434674)
If you can share your mesh, then we can try to rectify problem...

hi
I saved the mesh file and profile in draft @ a gmail account

user: fluent111
password: 1234506789

thanks

Far June 18, 2013 12:23

Using density based solver (uniform inlet pressure, averaged from pressure profile data) and solver is running without problems.

Few points :

1. Mesh is not good enough

2. Place outlet further downstream.

3. Check boundary conditions


All times are GMT -4. The time now is 16:21.