Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Electrified fluid...
Posted 27 giu 2011, 07:06 GMT-4 Computational Fluid Dynamics (CFD) Version 4.2 12 Replies
Please login with a confirmed email address before reporting spam
Hello,
I am just getting started with Comsol (4.1) and am attempting to model the shape of a fluid droplet as it is deformed by an externally applied electric field (i.e: electrospray).
So far I have managed to calculated the electric field for a chosen geometry (stationary solution) using the Electrostatics physics module. I have also used the laminar two-phase flow, level set physics module to calculate the dynamic shape of a fluid meniscus at the end of a capillary with pressure applied.
The problem I am having is with combining these two types of physics so that I can correctly model my situation.
What is the appropriate method to use in modelling this experiment?
Is it possible to add the force from the electric field into the level set calculation?
I have read that the ALE method may be more appropriate than the level set one because there is always a clearly defined boundary for the electric field to act upon. The problem is that the fluid meniscus is expected to deform into a thin jet. This might be too large a deformation for the ALE method to handle.
What about an iterative method in which the geometry is re-meshed between steps? Is this possible to achieve in Comsol?
Any advice that you can offer will be gratefully received.
Thanks,
Matt
I am just getting started with Comsol (4.1) and am attempting to model the shape of a fluid droplet as it is deformed by an externally applied electric field (i.e: electrospray).
So far I have managed to calculated the electric field for a chosen geometry (stationary solution) using the Electrostatics physics module. I have also used the laminar two-phase flow, level set physics module to calculate the dynamic shape of a fluid meniscus at the end of a capillary with pressure applied.
The problem I am having is with combining these two types of physics so that I can correctly model my situation.
What is the appropriate method to use in modelling this experiment?
Is it possible to add the force from the electric field into the level set calculation?
I have read that the ALE method may be more appropriate than the level set one because there is always a clearly defined boundary for the electric field to act upon. The problem is that the fluid meniscus is expected to deform into a thin jet. This might be too large a deformation for the ALE method to handle.
What about an iterative method in which the geometry is re-meshed between steps? Is this possible to achieve in Comsol?
Any advice that you can offer will be gratefully received.
Thanks,
Matt
12 Replies Last Post 12 dic 2013, 08:52 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
How to couple Two Phase Flow with Electostatics ?
See, for example, "Separation Through Electrocoalescence" in the Model Library.
See, for example, "Separation Through Electrocoalescence" in the Model Library.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks for your rapid reply.
I will look at this example.
Thanks,
Matt
I will look at this example.
Thanks,
Matt
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I have been trying a few things using the model suggested above...
Remove one circle.
Centre the circle in the rectangle.
Set the fluid flow rate to zero.
Now I have the situation of a droplet of water floating in oil between parallel plates:
When I run this simulation the circular droplet deforms to become an ellipse.
However, the droplet is compressed in the direction of the E-field and extended perpendicular to it.
I may be mistaken but I was expecting the droplet to extend in the direction of the electric field.
Any ideas what is going wrong here?
Remove one circle.
Centre the circle in the rectangle.
Set the fluid flow rate to zero.
Now I have the situation of a droplet of water floating in oil between parallel plates:
When I run this simulation the circular droplet deforms to become an ellipse.
However, the droplet is compressed in the direction of the E-field and extended perpendicular to it.
I may be mistaken but I was expecting the droplet to extend in the direction of the electric field.
Any ideas what is going wrong here?
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I think the homogeneous field between parallel plates doesn't deform the droplet, it just accelerates it (if it carries a charge). The deformation must result from the interaction of the moving droplet with the oil around it. This could explain the picture you found.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks for your reply.
The deformation is expected, see this reference: C. E. Rosenkilde 1969
rspa.royalsocietypublishing.org/content/312/1511/473.abstract
Quoting the first few lines of the abstract:
"In this paper an appropriate extension of the virial method developed by Chandrasekhar is
used to systematically re-examine the equilibrium and stability of an incompressible dielectric
fluid drop situated in a uniform electric field. The equilibrium shapes are initially
assumed to be ellipsoidal; and it is shown that only prolate spheroids elongated in the
direction of the applied field are compatible with the moment equations of lowest order."
In the simulation the droplet centre of mass does not move relative to the plates and because I have set the fluid flow rate as zero at the inlet there is no flow of fluid past the droplet either. However the result of the simulation is an oblate ellipse rather than a prolate one.
The deformation is expected, see this reference: C. E. Rosenkilde 1969
rspa.royalsocietypublishing.org/content/312/1511/473.abstract
Quoting the first few lines of the abstract:
"In this paper an appropriate extension of the virial method developed by Chandrasekhar is
used to systematically re-examine the equilibrium and stability of an incompressible dielectric
fluid drop situated in a uniform electric field. The equilibrium shapes are initially
assumed to be ellipsoidal; and it is shown that only prolate spheroids elongated in the
direction of the applied field are compatible with the moment equations of lowest order."
In the simulation the droplet centre of mass does not move relative to the plates and because I have set the fluid flow rate as zero at the inlet there is no flow of fluid past the droplet either. However the result of the simulation is an oblate ellipse rather than a prolate one.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
If I change the definitions for the force components from:
Fx = d(Tem11,x)+d(Tem12,y)
Fy = d(Tem21,x)+d(Tem22,y)
to
Fx = -d(Tem11,x)-d(Tem12,y)
Fy = -d(Tem21,x)-d(Tem22,y)
Then I get the deformation in the expected direction.
Can anyone confirm that the definition used in the original example was in error?
Thanks
Matt
Fx = d(Tem11,x)+d(Tem12,y)
Fy = d(Tem21,x)+d(Tem22,y)
to
Fx = -d(Tem11,x)-d(Tem12,y)
Fy = -d(Tem21,x)-d(Tem22,y)
Then I get the deformation in the expected direction.
Can anyone confirm that the definition used in the original example was in error?
Thanks
Matt
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
you should check the discretization of electric module, use second order in Advanced setting of it. I mean quadratic instead of linear one.
also the oblate deformation can happen in some cases depending on the relative viscosity and other parameters
also the oblate deformation can happen in some cases depending on the relative viscosity and other parameters
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks for the reply,
I changed the discretization to quadratic and put the force definition back to its initial setting.
This does indeed give me the expected prolate deformation.
I need to do some reading to try to understand exactly what the discretization setting does but coupling the Maxwell stress to the level set physics looks like a promising approach.
Thanks,
Matt
I changed the discretization to quadratic and put the force definition back to its initial setting.
This does indeed give me the expected prolate deformation.
I need to do some reading to try to understand exactly what the discretization setting does but coupling the Maxwell stress to the level set physics looks like a promising approach.
Thanks,
Matt
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
hi Matt,
do you know how can we reduce the interface thickness in level set method ?
the interface thickness depends on the maximum element size in level set method and when I change this dependency, I have some problems with convergence.
I do not know how to reduce the interface thickness in this method, the only way it to use finer mesh and I have memory problems with that.
best
do you know how can we reduce the interface thickness in level set method ?
the interface thickness depends on the maximum element size in level set method and when I change this dependency, I have some problems with convergence.
I do not know how to reduce the interface thickness in this method, the only way it to use finer mesh and I have memory problems with that.
best
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Osameh,
I think the interface thickness is controlled by the parameter "\epsilon_{pf}"
If you look in the settings menu:
Laminar Two-Phase Flow, Phase Field (tpf) > Fluid Properties 1 > Phase Field Parameters
There is a setting called: "Parameter controlling interface thickness"
"epsilon_{pf}" by default this seems to be tpf.hmax/2.
This might be the setting you are looking for but I'm no expert. I would have thought that it would be a good idea to reduce the element size at the same time as reducing this parameter.
Hope this helps.
Matt
I think the interface thickness is controlled by the parameter "\epsilon_{pf}"
If you look in the settings menu:
Laminar Two-Phase Flow, Phase Field (tpf) > Fluid Properties 1 > Phase Field Parameters
There is a setting called: "Parameter controlling interface thickness"
"epsilon_{pf}" by default this seems to be tpf.hmax/2.
This might be the setting you are looking for but I'm no expert. I would have thought that it would be a good idea to reduce the element size at the same time as reducing this parameter.
Hope this helps.
Matt
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Mat,
Can u please tell me what values u have used for initialization parameters in Level set method and the mesh size and time step size as well coz I am trying to run a level set simulation but not getting anything right.
Kindest regards.
Can u please tell me what values u have used for initialization parameters in Level set method and the mesh size and time step size as well coz I am trying to run a level set simulation but not getting anything right.
Kindest regards.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Matt
how to measure the droplet centre of mass ?
Dear Fucheng
how to measure the droplet centre of mass ?
Dear Fucheng