rhoSimpleFoam with MRF (in rotation) outlet instability

42SilentRunner · January 18, 2020, 07:09

Hi Foamers,
I am a relative newbie to this forum and would be interested in advice from others to help me fix my problem.

I have a problem when running my mesh with rhoSimpleFoam using MRF (rotation). The mesh is a symmetrical fuselage with 4 blades placed around that fuselage. I have modelled the system as an annular duct (open on the inlet, outlet and circumference) and with the mesh quartered given the symmetry. The blade rotates at 1046rad/sec (200 metres/sec at the tip) about the centre-line of that duct fully within that duct, 0.6 metres from the inlet (0.3 metres from the outlet). The duct length is 0.9metres.

When I run with the inlet and internal initial conditions at 0m/sec, the solution seems to be stablizing (at least to 33000 iterations) and the outlet distribution seems reasonable. However, if I set the same conditions to 30metres/sec the solution crashes at around 3500 iterations.

Checkmesh shows some non-orthogonality, but the mesh passed (see attached).

Examination of the mesh reveals that points on the outlet mesh have unexpected spikes in the distribution for the 30metres/sec case. This seems to start at around 3000 iterations (see U-3000.jpg). The spikes appear to have originated at or near the outlet (not the blade upstream). The spikes continue to expand with time until the crash.

Checkmesh shows some non-orthogonality, but the mesh passed. The non-orthogonality appears to be no-where near the problem mesh (see Non-orthogonal.jpg)

I have tried alternative U, P, etc outlet conditions, fvSchemes, fvSolutions etc but without finding a solution that matches the current stability (please check as I could easily have overseen something).

Please can someone help identify where I am going wrong? My thoughts suggest I need to increase the length of the duct behind the blade to allow the flow to re-distribute better, but I am guessing.

Thanks for your help.

The attached files:
conditions.docx - contains all files apart from the mesh to describe the test.
jpg's - the U distribution at 3000 iterations, and the non-orthogonality at the outlet from inside the mesh.

U-3000.jpg

Non-orthogonal.jpg

checkmesh_results.txt

Results.txt

Conditions.docx

42SilentRunner · January 28, 2020, 13:49

Hi all.
Has anyone had any thoughts on this issue?
Thanks.

ybapat · January 29, 2020, 00:37

Non-orthogonality is high (87). You can try improving mesh. You can also try to use non-orthogonal correctors.

January 18, 2020, 07:09	rhoSimpleFoam with MRF (in rotation) outlet instability	#1
42SilentRunner New Member Join Date: Feb 2017 Posts: 2 Rep Power: 0	Hi Foamers, I am a relative newbie to this forum and would be interested in advice from others to help me fix my problem. I have a problem when running my mesh with rhoSimpleFoam using MRF (rotation). The mesh is a symmetrical fuselage with 4 blades placed around that fuselage. I have modelled the system as an annular duct (open on the inlet, outlet and circumference) and with the mesh quartered given the symmetry. The blade rotates at 1046rad/sec (200 metres/sec at the tip) about the centre-line of that duct fully within that duct, 0.6 metres from the inlet (0.3 metres from the outlet). The duct length is 0.9metres. When I run with the inlet and internal initial conditions at 0m/sec, the solution seems to be stablizing (at least to 33000 iterations) and the outlet distribution seems reasonable. However, if I set the same conditions to 30metres/sec the solution crashes at around 3500 iterations. Checkmesh shows some non-orthogonality, but the mesh passed (see attached). Examination of the mesh reveals that points on the outlet mesh have unexpected spikes in the distribution for the 30metres/sec case. This seems to start at around 3000 iterations (see U-3000.jpg). The spikes appear to have originated at or near the outlet (not the blade upstream). The spikes continue to expand with time until the crash. Checkmesh shows some non-orthogonality, but the mesh passed. The non-orthogonality appears to be no-where near the problem mesh (see Non-orthogonal.jpg) I have tried alternative U, P, etc outlet conditions, fvSchemes, fvSolutions etc but without finding a solution that matches the current stability (please check as I could easily have overseen something). Please can someone help identify where I am going wrong? My thoughts suggest I need to increase the length of the duct behind the blade to allow the flow to re-distribute better, but I am guessing. Thanks for your help. The attached files: conditions.docx - contains all files apart from the mesh to describe the test. jpg's - the U distribution at 3000 iterations, and the non-orthogonality at the outlet from inside the mesh. U-3000.jpg Non-orthogonal.jpg checkmesh_results.txt Results.txt Conditions.docx

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
rhoSimpleFoam doubts on BC and fvSchemes (external aero)	giovanni.medici	OpenFOAM Running, Solving & CFD	3	October 27, 2021 05:06
Volume fraction at outlet not known	Bisht	FLUENT	8	September 5, 2017 02:38
ATTENTION! Reliability problems in CFX 5.7	Joseph	CFX	14	April 20, 2010 15:45
VOF Outlet boundary condition in cfd - ace	JM	Main CFD Forum	0	December 15, 2006 08:07
Impinging Jet. Outlet Boundary Condition	Vincent	Main CFD Forum	0	December 16, 2003 10:49

January 28, 2020, 13:49		#2
42SilentRunner New Member Join Date: Feb 2017 Posts: 2 Rep Power: 0	Hi all. Has anyone had any thoughts on this issue? Thanks.

January 29, 2020, 00:37		#3
ybapat Senior Member Yogesh Bapat Join Date: Oct 2010 Posts: 102 Rep Power: 15	Non-orthogonality is high (87). You can try improving mesh. You can also try to use non-orthogonal correctors.