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.

sessile droplet simulation

Please login with a confirmed email address before reporting spam

Hi everybody!

I am a PhD student and using the microfluidics module of COMSOL. I want to simulate a sessile droplet placed on a flat horizontal surface with specified contact angles in axial 2D system. For this I use the phase field method that solves the Cahn-Hilliard equation coupled with the Navier-Stokes equation. First I have to initialize the phase field function in the simulation domain after which I apply the time dependent solver to set the correct contact angles. But during the simulation the integral of the phase field function does not conserve! The longer the solver runs the less the volume of the droplet becomes. And this is a numerical phenomenon not physical! Is there anybody here who experienced similar things? By the way I experience the same when using the level set method instead of phase field method.

I guess it is something wrong with the setting of constants involved in either method, though I followed what the tutorial suggested.

I would really appreciate an answer from an expert.
Daniel

7 Replies Last Post 25.07.2016, 01:11 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.01.2010, 16:55 GMT-5
I don't have experience with phase field method but I am guessing that the method does not necessarily intrinsically conserve volume (similar to volume of fluid method). So the logical approach would be to somehow impose a volume conservation constraint (i.e. total volume remains the same). Not sure how it can be implemented though.

Ozgur
I don't have experience with phase field method but I am guessing that the method does not necessarily intrinsically conserve volume (similar to volume of fluid method). So the logical approach would be to somehow impose a volume conservation constraint (i.e. total volume remains the same). Not sure how it can be implemented though. Ozgur

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.01.2010, 17:41 GMT-5
Thanks for the fast reply. I am just getting into closer acquaintance with this method but I read that the total integral of the phase field function must be constant and for me this is exactly what does not seem to be constant at all. And since the volume is calculated from it the volume does not remain constant either. Insignificant changes would not be a problem, but here more percent discrepancies occur in several dozens of milliseconds of time which is serious.
Thanks for the fast reply. I am just getting into closer acquaintance with this method but I read that the total integral of the phase field function must be constant and for me this is exactly what does not seem to be constant at all. And since the volume is calculated from it the volume does not remain constant either. Insignificant changes would not be a problem, but here more percent discrepancies occur in several dozens of milliseconds of time which is serious.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.02.2010, 03:47 GMT-5
I am in the process of doing a similar simulation. I have to simulate a single droplet on a flat surface so as to determine the final shape and contact angle with the surface in equilibrium. I have faced convergence problems using the level set method. If you could carry out your simulation, please help. I would keep you updated of my progress.

I am in the process of doing a similar simulation. I have to simulate a single droplet on a flat surface so as to determine the final shape and contact angle with the surface in equilibrium. I have faced convergence problems using the level set method. If you could carry out your simulation, please help. I would keep you updated of my progress.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.02.2010, 06:49 GMT-5
Thanks for the comment. My final goal is to simulate the droplet in a silicone oil environment and using air instead of oil was just a preliminary calculation. I have found that the level set method absolutely fails to converge properly and a significant portion of the droplet disappears during the simulation. So I turned to phase field method but experienced the same - though to a less extent. Then I found a paper stating that the phase field does not guarantee the mass (an so volume) conservation of the droplets. (Search for the following expression in Google: "spontaneous shrinkage of drops", and you will find it under the first link.) Especially small droplets are inclined to disappear and this problem is more serious if the density difference between the drop and the ambient medium is big (such as water and air). But water and silicone oil have almost the same density and with these I found almost no reduction in volume.

As for the water-air situation I know no solution for the moment. I think it is possible that there does not exist at all, at least if you insist on using finite element method based tools like level set or phase field. Because these tools do not guarantee mass and volume conservation.

Regards,
Daniel
Thanks for the comment. My final goal is to simulate the droplet in a silicone oil environment and using air instead of oil was just a preliminary calculation. I have found that the level set method absolutely fails to converge properly and a significant portion of the droplet disappears during the simulation. So I turned to phase field method but experienced the same - though to a less extent. Then I found a paper stating that the phase field does not guarantee the mass (an so volume) conservation of the droplets. (Search for the following expression in Google: "spontaneous shrinkage of drops", and you will find it under the first link.) Especially small droplets are inclined to disappear and this problem is more serious if the density difference between the drop and the ambient medium is big (such as water and air). But water and silicone oil have almost the same density and with these I found almost no reduction in volume. As for the water-air situation I know no solution for the moment. I think it is possible that there does not exist at all, at least if you insist on using finite element method based tools like level set or phase field. Because these tools do not guarantee mass and volume conservation. Regards, Daniel

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10.03.2011, 18:38 GMT-5
I have the same problem
I have the same problem

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 19.09.2015, 13:30 GMT-4
Hi, did you solve this problem? I have similar problem and don't know what I can do. Thank you in advance.
Hi, did you solve this problem? I have similar problem and don't know what I can do. Thank you in advance.

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 25.07.2016, 01:11 GMT-4
Hey
Did you finish your project?My project is exactly the same as yours. It would be great if you could guide me.
Hey Did you finish your project?My project is exactly the same as yours. It would be great if you could guide me.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.