Problem with wall adhesion

MachZero · September 12, 2012, 20:56

Hello, I am encountering a problem with two fluids in a given domain, and I am trying to determine the effects of increasing or decreasing one of the fluids' flow rates on the overall stability of the system. In this simulation, fluid 1 is acting as a lubricant such that fluid 2 does not contact the wall. The fluids are immiscible so I am using the VOF method in fluent. For this model, surface tension is very low, near 0.1 dynes/cm and a wall adhesion angle is modeled at 40 degrees.

In order to get a worst case scenario and to establish a baseline "failure criteria", I turned the first fluid flow rates to zero, and expected to see the second fluid foul up on the wall and contact it. I tried running the simulation as an explicit VOF with a Courant number near 1, but I was then instructed to switch to implicit solving such that I could use a larger timestep (please feel free to comment on this).

I have found that, after running the simulation out for a long time, the best I can do is get the second fluid to come within 1 cell of the wall, but can never get it to contact the wall. Note that the entire simulation is initialized as fluid one, the lubricating fluid. My thought is that this be due to a zero-wall velocity, such that whatever fluid is at the wall can never be pushed away. This obviously causes problems because I do not have a good failure criteria.

I know I have seen simulations before in which the fluid at the wall acts less like a permanent lubricating layer and more like a fluid that can drain off. Am I missing something here? Or is there something special about my case? Here are two examples of this behavior that I have seen but cannot replicate:

http://www.youtube.com/watch?v=-NQeI...eature=related
http://www.youtube.com/watch?v=v5txXX_nkRk

Is it something physical or am I not modeling something correctly? Why can I not remove a fluid from the wall. Please let me know if I am not clear in this description, I would like to have a good conversation about the underlying physics and modeling.

matmat · July 2, 2013, 09:08

Did you find a solution? I have a similar problem

MachZero · July 2, 2013, 09:50

Matmat, I am using fluent, and we discovered that the real problem was the choice of pressure-velocity coupling. We were not getting converged results on each timestep. I switched to a coupled solver and this helped get a wall adhesion.

That being said, I played with the physics in my head for a long time and I attended the Fluent Advanced training, and I could not determine after either if my one fluid ever really should contact the wall or not. It is hard to say, is there a layer of the first fluid ALWAYS there? Or does it scrape it away?

If you have any pictures you can post, I encourage you to do so. I would definitely be interested in discussing more.

September 12, 2012, 20:56	Problem with wall adhesion	#1
MachZero Member Join Date: Aug 2011 Posts: 89 Rep Power: 14	Hello, I am encountering a problem with two fluids in a given domain, and I am trying to determine the effects of increasing or decreasing one of the fluids' flow rates on the overall stability of the system. In this simulation, fluid 1 is acting as a lubricant such that fluid 2 does not contact the wall. The fluids are immiscible so I am using the VOF method in fluent. For this model, surface tension is very low, near 0.1 dynes/cm and a wall adhesion angle is modeled at 40 degrees. In order to get a worst case scenario and to establish a baseline "failure criteria", I turned the first fluid flow rates to zero, and expected to see the second fluid foul up on the wall and contact it. I tried running the simulation as an explicit VOF with a Courant number near 1, but I was then instructed to switch to implicit solving such that I could use a larger timestep (please feel free to comment on this). I have found that, after running the simulation out for a long time, the best I can do is get the second fluid to come within 1 cell of the wall, but can never get it to contact the wall. Note that the entire simulation is initialized as fluid one, the lubricating fluid. My thought is that this be due to a zero-wall velocity, such that whatever fluid is at the wall can never be pushed away. This obviously causes problems because I do not have a good failure criteria. I know I have seen simulations before in which the fluid at the wall acts less like a permanent lubricating layer and more like a fluid that can drain off. Am I missing something here? Or is there something special about my case? Here are two examples of this behavior that I have seen but cannot replicate: http://www.youtube.com/watch?v=-NQeI...eature=related http://www.youtube.com/watch?v=v5txXX_nkRk Is it something physical or am I not modeling something correctly? Why can I not remove a fluid from the wall. Please let me know if I am not clear in this description, I would like to have a good conversation about the underlying physics and modeling.

July 2, 2013, 09:08	Similar problem	#2
matmat New Member mathieu pic Join Date: Jul 2013 Posts: 1 Rep Power: 0	Did you find a solution? I have a similar problem

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] Export ICEM mesh to Gambit / Fluent	romekr	ANSYS Meshing & Geometry	1	November 26, 2011 12:11
Problem with wall function	gregjunqua	OpenFOAM	5	August 25, 2011 23:32
Two Sided wall Problem	Mk1987	ANSYS Meshing & Geometry	0	March 26, 2010 13:37
Problem Interface Solid Fluid with wall velocity Solver v12	hills1	CFX	2	October 12, 2009 05:36
moving wall BC problem	Sourabh	Main CFD Forum	1	August 8, 2006 07:47

July 2, 2013, 09:50		#3
MachZero Member Join Date: Aug 2011 Posts: 89 Rep Power: 14	Matmat, I am using fluent, and we discovered that the real problem was the choice of pressure-velocity coupling. We were not getting converged results on each timestep. I switched to a coupled solver and this helped get a wall adhesion. That being said, I played with the physics in my head for a long time and I attended the Fluent Advanced training, and I could not determine after either if my one fluid ever really should contact the wall or not. It is hard to say, is there a layer of the first fluid ALWAYS there? Or does it scrape it away? If you have any pictures you can post, I encourage you to do so. I would definitely be interested in discussing more.