Duct flow

December 2, 1999, 21:45

Hi,

I am modeling a duct flow with one inlet and four outlets. The duct cross section is non circular and very irregualr. The obejective of the simulation is to predict the mass flow distribution at the outlets for a given mass inlet. The flow is incompressible and turbulent. I am imposing a velocity inlet and a pressure outlet ( outlets were placed 10 diameters downstream to avoid reversed flow).

I was able to obtain a solution where the residuals reduced to 1.e-5 smoothly, but the mass flow rates at the outlets were oscillating. In other words, outlet mass flow rates are a function of iteration. The solution I got shows that there are a lot of secondary flows and a swirl flow at the outlets ( physical outlets). This is somewhat similar to the "swirl back flow" message posted by Jonas Larsson.

I am using Fluent solver and tried different boundary conditions and different turbulent models with no success. Please advice .....

Thanks

jlehew · February 23, 2020, 22:33

Given the problem you described, it sounds like one of the following might be true:

1) The problem is naturally unsteady and you'll need to run an unsteady simulation and calculate the time averaged mass flow rate
2) By extending the domain past the physical outlets of your system, you've created a different boundary condition than you would see in the real system. If you were to extend your domain in a way that better matched the physical system you may obtain a steady solution.
3) You may need to use a different solver, solver options, or a more refined mesh to get a steady flow in the region of interest.

If the channel is very irregular I could imagine you having either a fixed recirculation region or vortex shedding occurring where the physical system outlets should be leading to the variation in mass flow rate with iteration that you are seeing. In this case I would recommend trying to run an unsteady simulation and seeing what you get.

If this doesn't work or you don't think there should have any recirculation at the exit, you can test out options to address items 2 and 3 above.

Let me know if you need any clarification or would like to share any images of the flow domain to better illustrate the problem.

Best regards.

December 2, 1999, 21:45	Duct flow	#1
Omar Mohamud Guest Posts: n/a	Hi, I am modeling a duct flow with one inlet and four outlets. The duct cross section is non circular and very irregualr. The obejective of the simulation is to predict the mass flow distribution at the outlets for a given mass inlet. The flow is incompressible and turbulent. I am imposing a velocity inlet and a pressure outlet ( outlets were placed 10 diameters downstream to avoid reversed flow). I was able to obtain a solution where the residuals reduced to 1.e-5 smoothly, but the mass flow rates at the outlets were oscillating. In other words, outlet mass flow rates are a function of iteration. The solution I got shows that there are a lot of secondary flows and a swirl flow at the outlets ( physical outlets). This is somewhat similar to the "swirl back flow" message posted by Jonas Larsson. I am using Fluent solver and tried different boundary conditions and different turbulent models with no success. Please advice ..... Thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
on fully developed duct flow	yfyap	Main CFD Forum	6	November 7, 2012 10:59
convergent duct water flow	kkiitm	CFX	0	January 2, 2011 12:11
Duct Flow Analysis	benbru	Siemens	1	September 26, 2007 16:53
Turbulence flow in a rectangular duct	Watchapon	Main CFD Forum	0	April 7, 2007 08:34
Question about flow in ducted fan in another duct	Jane	FLUENT	0	June 21, 2004 14:57

February 23, 2020, 22:33		#2
jlehew New Member Join Date: Feb 2020 Posts: 2 Rep Power: 0	Given the problem you described, it sounds like one of the following might be true: 1) The problem is naturally unsteady and you'll need to run an unsteady simulation and calculate the time averaged mass flow rate 2) By extending the domain past the physical outlets of your system, you've created a different boundary condition than you would see in the real system. If you were to extend your domain in a way that better matched the physical system you may obtain a steady solution. 3) You may need to use a different solver, solver options, or a more refined mesh to get a steady flow in the region of interest. If the channel is very irregular I could imagine you having either a fixed recirculation region or vortex shedding occurring where the physical system outlets should be leading to the variation in mass flow rate with iteration that you are seeing. In this case I would recommend trying to run an unsteady simulation and seeing what you get. If this doesn't work or you don't think there should have any recirculation at the exit, you can test out options to address items 2 and 3 above. Let me know if you need any clarification or would like to share any images of the flow domain to better illustrate the problem. Best regards.