Creating reversed flow at a supersonic outlet

January 1, 2003, 20:04

I am trying to create reversed flow in a sudden expansion geometry case. In FLUENT when the outlet is locally supersonic it extrapolates the data in the interior of the geometry and creates new values at the outlet. What can I do to force the outlet boundary conditions to be enforced so that I can create an upstream interaction in the isolator region before the sudden expansion? Is there a udf that needs to be created? My inlet boundary condition is supersonic. Any help will be appreciated.

January 2, 2003, 15:39

I have solved the problem. I simply initialized the case from the outlet and everything has worked out.

January 3, 2003, 11:52

Hi Stephen,

For a supersonic fully compressible flow, outlet boundary conditions should be ignored just like Fluent is doing otherwise your problem will be illposed.

The treatment of compressible flows is dictated by the theory of characteristice. You can find a discussion of this in "Numerical Computational of Internal and External Flows" Volume 2 (chapter 16 I think) by Charles Hirsch.

January 1, 2003, 20:04	Creating reversed flow at a supersonic outlet	#1
Stephen Guest Posts: n/a	I am trying to create reversed flow in a sudden expansion geometry case. In FLUENT when the outlet is locally supersonic it extrapolates the data in the interior of the geometry and creates new values at the outlet. What can I do to force the outlet boundary conditions to be enforced so that I can create an upstream interaction in the isolator region before the sudden expansion? Is there a udf that needs to be created? My inlet boundary condition is supersonic. Any help will be appreciated.

January 2, 2003, 15:39	Re: Creating reversed flow at a supersonic outlet	#2
Stephen Guest Posts: n/a	I have solved the problem. I simply initialized the case from the outlet and everything has worked out.

January 3, 2003, 11:52	Re: Creating reversed flow at a supersonic outlet	#3
Amadou Sowe Guest Posts: n/a	Hi Stephen, For a supersonic fully compressible flow, outlet boundary conditions should be ignored just like Fluent is doing otherwise your problem will be illposed. The treatment of compressible flows is dictated by the theory of characteristice. You can find a discussion of this in "Numerical Computational of Internal and External Flows" Volume 2 (chapter 16 I think) by Charles Hirsch.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pressure outlet in two-phase flow in horizontal 2D channel	AlmostSurelyRob	Main CFD Forum	0	November 17, 2010 08:32
ATTENTION! Reliability problems in CFX 5.7	Joseph	CFX	14	April 20, 2010 16:45
reverse flow at outlet	Neil Forsythe	Main CFD Forum	12	June 17, 2005 19:22
Inlet and outlet flow rate	Neser	CFX	1	March 2, 2004 17:02
Outlet BC for subsonic flow in pipe	andrea panizza	FLUENT	6	May 10, 2003 08:44