Bubble column Substracting the drag on the liquid phase

sarrazs · October 2, 2019, 11:47

Hi,

I'm setting up a multiphase simulation using CFX for a bubble column. I set the water as the continuous Fluid and the bubbles as the dispersed fluid. I want to set the interfacial forces acting on the liquid phase to zero. I know it is not accurate but it is for a verification purpose. I put the drag = 0 and I added a drag force by a user expression as a Momentum source for the bubbles, the simulation didn't work. Than I kept the drag and I substract it by a user function Momentum source on the liquid. The simulation worked but it the liquid didn't stay stagnant. I may be doing it wrong.
Is there any other way to do it??

Opaque · October 2, 2019, 13:06

How did you add the momentum source?

Can you show your setup for the momentum source only?

sarrazs · October 2, 2019, 13:57

I added a subdomain, I put the location as the whole domain, then in the fluid Sources, I selected Water - Momentum Source - General Momentum Source and in x,y,z component I put the drag force components as for the x component -(3/4)*air.vf*water.density*sqrt(UrU^2+UrV^2+UrW^2)*CD/air.mean particle diameter *(air.Velocity u-water.Velocity u)

UrU = (air.Velocity u-water.Velocity u)…
and CD is equal to 0.44 because in the fluid pair model I put the drag coefficient = 0.44 for simplification .

Opaque · October 2, 2019, 14:07

How much source/linearization coefficient are you using?

sarrazs · October 2, 2019, 15:03

Quote:

Originally Posted by Opaque

How much source/linearization coefficient are you using?

I did not use a momentum linearization coefficient. Is there any default value used by CFX for the drag?? and what is the best value? or how can I choose the right value?

Thank you very much

Opaque · October 3, 2019, 09:25

You need to linearize such a source term; otherwise, it will only converge for a very small timestep (if it ever converges).

The linearization coefficient is the derivative of the source term respect to the variable being solved for, say velocity u/v/w, in the equation.

In the case of the momentum equation, it should be a vector, but since a single linearization entry is provided (isotropic linearization), you must pick the most conservative value, that is, the maximum between any of the 3 possible derivatives.

October 2, 2019, 11:47	Bubble column Substracting the drag on the liquid phase	#1
sarrazs Member cfxtwophaseflow Join Date: Aug 2017 Posts: 46 Rep Power: 9	Hi, I'm setting up a multiphase simulation using CFX for a bubble column. I set the water as the continuous Fluid and the bubbles as the dispersed fluid. I want to set the interfacial forces acting on the liquid phase to zero. I know it is not accurate but it is for a verification purpose. I put the drag = 0 and I added a drag force by a user expression as a Momentum source for the bubbles, the simulation didn't work. Than I kept the drag and I substract it by a user function Momentum source on the liquid. The simulation worked but it the liquid didn't stay stagnant. I may be doing it wrong. Is there any other way to do it??

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
twoPhaseEulerFoam bubble column crashes due to problems at outlet region	hester	OpenFOAM Running, Solving & CFD	5	July 21, 2024 09:19
Monitoring Volume Average of Liquid Velocity Magnitude in a Bubble Column / Airlift	kevinlipps	OpenFOAM Post-Processing	4	May 19, 2016 06:59
twoPhaseEulerFoam: turbulence on produces instability of drag phase	florian_krause	OpenFOAM	1	October 11, 2013 10:02
Bubble Column	Glen	Main CFD Forum	0	January 24, 2006 01:56
compressible two phase flow in CFX4.4	youngan	CFX	0	July 2, 2003 00:32

October 2, 2019, 13:06		#2
Opaque Senior Member Join Date: Jun 2009 Posts: 1,880 Rep Power: 33	How did you add the momentum source? Can you show your setup for the momentum source only?

October 2, 2019, 13:57		#3
sarrazs Member cfxtwophaseflow Join Date: Aug 2017 Posts: 46 Rep Power: 9	I added a subdomain, I put the location as the whole domain, then in the fluid Sources, I selected Water - Momentum Source - General Momentum Source and in x,y,z component I put the drag force components as for the x component -(3/4)air.vfwater.densitysqrt(UrU^2+UrV^2+UrW^2)CD/air.mean particle diameter *(air.Velocity u-water.Velocity u) UrU = (air.Velocity u-water.Velocity u)… and CD is equal to 0.44 because in the fluid pair model I put the drag coefficient = 0.44 for simplification .

October 2, 2019, 14:07		#4
Opaque Senior Member Join Date: Jun 2009 Posts: 1,880 Rep Power: 33	How much source/linearization coefficient are you using?

October 3, 2019, 09:25		#6
Opaque Senior Member Join Date: Jun 2009 Posts: 1,880 Rep Power: 33	You need to linearize such a source term; otherwise, it will only converge for a very small timestep (if it ever converges). The linearization coefficient is the derivative of the source term respect to the variable being solved for, say velocity u/v/w, in the equation. In the case of the momentum equation, it should be a vector, but since a single linearization entry is provided (isotropic linearization), you must pick the most conservative value, that is, the maximum between any of the 3 possible derivatives.