Non-physical behavior in internal flow (two inlet --> one outlet)

chintan · February 11, 2015, 16:38

I'm solving a film cooling problem with one hole and I'm getting non-physical behaviour inside hole and one of the inlet.

I have been trying to solve these for few months now. It doesn't converge at all. (after 15 hours of simulation on 20 of 2.4 GHz cores, density residual is still -0.9)

Problem:
inlet: -hot gas: mass flux (from yz plane i.e. only x velocity)
-coolant: mass flux (from xz plane ie only y velocity)
outlet: -hot gas+coolant: back pressure (from yz plane)

Issues:
1. Velocity inside hole in extensively large.
2. Bottom inlet has some velocity components going out of boundary.
[Images]
https://drive.google.com/file/d/0BzU...ew?usp=sharing
https://drive.google.com/file/d/0BzU...ew?usp=sharing

[configuration file]
https://drive.google.com/file/d/0BzU...ew?usp=sharing

My questions are: a. Am I doing something wrong? (configuration file linked above) b. If not, can su2 solve this kind of problem? (it seems fairly possible but test cases doesn't have internal flow problems except flat plate with a bump!) I would appreciate any help to address/identify the issue.

fpalacios · February 18, 2015, 14:40

In principle, SU2 should be able to solve this problem.

Could you please share the .su2 mesh?

Thanks!

Francisco Palacios
SU2 lead developer

chintan · February 18, 2015, 14:52

Hi Francisco,
Here is link for the mesh I'm using for test case.
https://drive.google.com/file/d/0BzU...ew?usp=sharing

chintan · February 18, 2015, 15:04

I want to note that I have tried both Total quantities and mass flux as inlet BC.

Also tried simplified cases like no turbulence, Euler equations etc. But I could never achieve convergence (and roughly, based on changes in residuals, it will not converge even after 10 days of computations with noted cores!)

Also, I got density residual about 1e-2 for test case with 1M cells, but then convergence stops. Momentum residuals stay in range of 1e3 to 1e5 and don't reduce.

February 11, 2015, 16:38	Non-physical behavior in internal flow (two inlet --> one outlet)	#1
chintan New Member Chintan Patel Join Date: Oct 2014 Location: USA Posts: 19 Rep Power: 12	I'm solving a film cooling problem with one hole and I'm getting non-physical behaviour inside hole and one of the inlet. I have been trying to solve these for few months now. It doesn't converge at all. (after 15 hours of simulation on 20 of 2.4 GHz cores, density residual is still -0.9) Problem: inlet: -hot gas: mass flux (from yz plane i.e. only x velocity) -coolant: mass flux (from xz plane ie only y velocity) outlet: -hot gas+coolant: back pressure (from yz plane) Issues: 1. Velocity inside hole in extensively large. 2. Bottom inlet has some velocity components going out of boundary. [Images] https://drive.google.com/file/d/0BzU...ew?usp=sharing https://drive.google.com/file/d/0BzU...ew?usp=sharing [configuration file] https://drive.google.com/file/d/0BzU...ew?usp=sharing My questions are: a. Am I doing something wrong? (configuration file linked above) b. If not, can su2 solve this kind of problem? (it seems fairly possible but test cases doesn't have internal flow problems except flat plate with a bump!) I would appreciate any help to address/identify the issue. Last edited by chintan; February 16, 2015 at 02:19. Reason: adding more details

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Internal Flow BCs for Compressible inlet flow	Katie	FLUENT	2	March 3, 2012 18:35
total pressure boundary problem ==> flow from outlet to inlet!!	mrshb4	OpenFOAM	0	November 20, 2010 13:41
ATTENTION! Reliability problems in CFX 5.7	Joseph	CFX	14	April 20, 2010 16:45
Duct Flow with 90 degree inlet and outlet	k	FLUENT	3	August 19, 2007 15:30
what the result is negatif pressure at inlet	chong chee nan	FLUENT	0	December 29, 2001 06:13

February 18, 2015, 14:40		#2
fpalacios Super Moderator Francisco Palacios Join Date: Jan 2013 Location: Long Beach, CA Posts: 404 Rep Power: 15	In principle, SU2 should be able to solve this problem. Could you please share the .su2 mesh? Thanks! Francisco Palacios SU2 lead developer

February 18, 2015, 14:52		#3
chintan New Member Chintan Patel Join Date: Oct 2014 Location: USA Posts: 19 Rep Power: 12	Hi Francisco, Here is link for the mesh I'm using for test case. https://drive.google.com/file/d/0BzU...ew?usp=sharing

February 18, 2015, 15:04		#4
chintan New Member Chintan Patel Join Date: Oct 2014 Location: USA Posts: 19 Rep Power: 12	I want to note that I have tried both Total quantities and mass flux as inlet BC. Also tried simplified cases like no turbulence, Euler equations etc. But I could never achieve convergence (and roughly, based on changes in residuals, it will not converge even after 10 days of computations with noted cores!) Also, I got density residual about 1e-2 for test case with 1M cells, but then convergence stops. Momentum residuals stay in range of 1e3 to 1e5 and don't reduce.