BC problem with Eulerian-Eulerian Multiphase Flow

ngoc_tran_bao · September 21, 2016, 22:14

Dear all friends,
I am doing an analysis on powder extinguish system and I have some troubles with Eulerian-Eulerian Multiphase Flow that I hope you guys can help me. In the simulation, I have 2 phases: Air – continuous fluid, and powder-dispersed fluid. In order to make the powder valid, I give it a very small dynamic viscosity of 1E-10 as some aforementioned threads suggested. My question is that:
1. In case of volume fraction of Air/powder=7/3, am I right when choosing E-E approach? I heart that if volume fraction of solid particle is over 10%, we should use E-E Model rather than L-E Model.
2. As for boundary condition, my target is to simulate how the velocity, pressure and volume fraction distribute when the mass flow rate of powder is at 3.5kg/s.
However, I cannot set the mass flow rate at Inlet or Outlet because when I do it, the solver always gets overflow error. I have tried with some relative small flow rate such as 0.1 or 0.01 kg/s but it doesn’t work. If I set a velocity at Inlet and relative pressure at Outlet, the solver runs smoothly but I cannot control the mass flow rate of powder. I hope you can help me shed some light on this situation.
Thank you very much for your time!

ghorrocks · September 22, 2016, 09:06

Above VF of 10% or particles then particle to particle collisions become important. CFX has no particle collision model in the langrangian model. CFX does have some models for this in Eularian models. So you are correct, you can use Eularian models at higher VF than lagrangian models in CFX.

Does your simulation run correctly with the mass flow inlet when you model it single phase? Your question is strange as mass flow boundaries are resolved to velocities, so a mass flow and velocity boundary are really the same thing.

Have you read the section in the CFX manuals about selection of boundary conditions? You need to be aware of what combinations of boundary conditions are numerically possible and stable. This is in the Modelling guide.

ngoc_tran_bao · September 23, 2016, 02:23

Dear Mr. Glenn,
Thank you for your answer of E-E model, I will continue carrying out simulation in this approach. As for BC, I understand that velocity and flow rate have a relative relationship (I calculate inlet velocity from known flow rate). In single flow with only 1 continuous fluid (air), the simulation run smoothly with inlet or outlet flow rate.
I also have already read the robust combinations between BC for inlet and outlet. One thing I wonder is the flow rate in multi-phase simulation, whenever I set inlet/outlet flow rate, the solver always gets overflow error. So if this situation is persistence, can I apply inlet total Pressure and Outlet velocity for my model?
Thank you!

ghorrocks · September 23, 2016, 06:07

For an incompressible simulation with one inlet and one outlet you cannot specify both the inlet and outlet to be mass flow rate. This is badly posed and will crash. One of those boundaries will need to be a pressure boundary of some description.

ngoc_tran_bao · September 24, 2016, 04:43

Hi Glenn,
I am so sorry when make you confused and misunderstand my idea. In my case, Inlet or Outlet flow rate is set but the solver get crash. I can not understand why because Inlet and outlet velocity work well.
If it is possible, could you please give me some suggestions of CFX tutorials of E-E model? Thank you so much!

September 21, 2016, 22:14	BC problem with Eulerian-Eulerian Multiphase Flow	#1
ngoc_tran_bao Member ngoc tran bao Join Date: Jan 2016 Posts: 35 Rep Power: 10	Dear all friends, I am doing an analysis on powder extinguish system and I have some troubles with Eulerian-Eulerian Multiphase Flow that I hope you guys can help me. In the simulation, I have 2 phases: Air – continuous fluid, and powder-dispersed fluid. In order to make the powder valid, I give it a very small dynamic viscosity of 1E-10 as some aforementioned threads suggested. My question is that: 1. In case of volume fraction of Air/powder=7/3, am I right when choosing E-E approach? I heart that if volume fraction of solid particle is over 10%, we should use E-E Model rather than L-E Model. 2. As for boundary condition, my target is to simulate how the velocity, pressure and volume fraction distribute when the mass flow rate of powder is at 3.5kg/s. However, I cannot set the mass flow rate at Inlet or Outlet because when I do it, the solver always gets overflow error. I have tried with some relative small flow rate such as 0.1 or 0.01 kg/s but it doesn’t work. If I set a velocity at Inlet and relative pressure at Outlet, the solver runs smoothly but I cannot control the mass flow rate of powder. I hope you can help me shed some light on this situation. Thank you very much for your time!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem of simulating shape oscillations of Bubble - Multiphase flow	akash	FLUENT	2	January 29, 2013 13:46
Problem with multiphase flow - mixture model	Mat22	FLUENT	2	October 27, 2010 03:07
Multiphase Eulerian Gas-Solid Flow	Ssn	FLUENT	0	April 21, 2008 05:52
problem with multiphase flow	sri	FLUENT	4	July 24, 2007 06:56
Multiphase flow problem	icedou	FLUENT	6	July 10, 2005 02:52

September 22, 2016, 09:06		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Above VF of 10% or particles then particle to particle collisions become important. CFX has no particle collision model in the langrangian model. CFX does have some models for this in Eularian models. So you are correct, you can use Eularian models at higher VF than lagrangian models in CFX. Does your simulation run correctly with the mass flow inlet when you model it single phase? Your question is strange as mass flow boundaries are resolved to velocities, so a mass flow and velocity boundary are really the same thing. Have you read the section in the CFX manuals about selection of boundary conditions? You need to be aware of what combinations of boundary conditions are numerically possible and stable. This is in the Modelling guide.

September 23, 2016, 02:23		#3
ngoc_tran_bao Member ngoc tran bao Join Date: Jan 2016 Posts: 35 Rep Power: 10	Dear Mr. Glenn, Thank you for your answer of E-E model, I will continue carrying out simulation in this approach. As for BC, I understand that velocity and flow rate have a relative relationship (I calculate inlet velocity from known flow rate). In single flow with only 1 continuous fluid (air), the simulation run smoothly with inlet or outlet flow rate. I also have already read the robust combinations between BC for inlet and outlet. One thing I wonder is the flow rate in multi-phase simulation, whenever I set inlet/outlet flow rate, the solver always gets overflow error. So if this situation is persistence, can I apply inlet total Pressure and Outlet velocity for my model? Thank you!

September 23, 2016, 06:07		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	For an incompressible simulation with one inlet and one outlet you cannot specify both the inlet and outlet to be mass flow rate. This is badly posed and will crash. One of those boundaries will need to be a pressure boundary of some description.

September 24, 2016, 04:43		#5
ngoc_tran_bao Member ngoc tran bao Join Date: Jan 2016 Posts: 35 Rep Power: 10	Hi Glenn, I am so sorry when make you confused and misunderstand my idea. In my case, Inlet or Outlet flow rate is set but the solver get crash. I can not understand why because Inlet and outlet velocity work well. If it is possible, could you please give me some suggestions of CFX tutorials of E-E model? Thank you so much!