Discussion Forum

I just found a video on youtube.
I'm trying to simulate what this user has already done.
I'm trying to get what happens in the video from 0:12 to 0:20 along with the droplet spreading on the surface.
Any help will be appreciated.

Sorry.
Here's the link:
www.youtube.com/user/COMSOLmultiphysics?ytsession=owXhWC9v2xOVocgiwSWUnM9AS9193dej-aqPfcbkS04Z5s6wZxTg59wv5aiyvOU-5JJB-P5wqm2pX3r5lPSyvBfnyPxYWTYpYQdFyfjgNZK_djInEir4qQL1WWiKHGCmFLmkN8IV9pzG7g78oWedH5lSma2GpIChyBf41vCR5V_vmILu5kaJS0qA5JPFNOSCekSQZv9YQgeRRzf6ZADnC4FZ2oXqcu4Cds-5jeQfYmjGtpvNSiYVF7sPHKPXOZFcgv01qqTKu1tcfzux6n4FChDuslTzJJVL46A6SyELo3w#p/c/418E0AC69302BB16/4/EyYagXQPFV8.

Or you can just type "two phase model using comsol 3.5a" in the search space on youtube.

Thank you.

you should use level set method,

Thank you for your reply. But i have used level set method only.
Any other suggestions?

I have noticed that you changed the parameters controlling the interface in Level set method, keep the default ones. Epsilon is related to the maximum mesh elemnt size, so do not change it from default.

epsilon=tpf.hmax/2

and also the reinitialization parameter

Hi,

I'm trying to simulate something similar...a water droplet falling through air and hitting the surface of a reservoir full of water and achieved some success. Note that I've used COMSOL 4.2, the two-phase flow level set method. Here are a few things I noted and applied to my model from the 'rising bubble model' available at the COMSOL online gallery:-

* If the initial domain where the water droplet resides isn't 'viscous' enough or 'dense' enough, (ex: regular air) there might be some problem with the water droplet being blown out of the domain by air eddies forming and rising up the wall. So unless you can bound the water droplet within the domain, try entering a manual value for the density and viscosity of air. I entered values that were 10 times as big just to see if it works or not.

* The gravity has to be increased as well if your droplet is too small compared to the reservoir, otherwise your risk blowing the droplet out of the domain, so experiment with different droplet sizes and different increased gravity values...then come to a conclusion which gives you the optimum results.

* When you are setting up the boundary conditions under the "Level Set"-physics node, you can do like I did, set pressure point constraints on the two sides of the water droplet to keep it from deforming, or if you want you can do without it and see the droplet deform with time and space.

* This one is a little tricky, but you have to modify the default solver if you are running the Phase Initialization+Time Dependent solution (which will be the case most of the time). You have to modify the "Dependent Variable 2" settings under "Solver Configuration". Follow the instructions in the attached tutorial where you have to modify the 'scaling' of the variable which is recommended by COMSOL for time dependent solutions.


Hope that works. If you are still having trouble converging, play around with the mesh size. Try to stick as close to the 'rising bubble' tutorial as possible. That helped me a lot.



Attachment:

# Water droplet hitting water (dl.dropbox.com/u/6219893/demo_col.mph)
# Rising bubble tutorial, both for COMSOL 4.2

Hi

looks nice, to allow us to have the model on the Forum, do a Edit clear Solutions, Edit clear Mesh, File reset history, then save in a new name and upload here

few comments: to better show the falling drop add a few step until the droplet is in contact.
Also to get a more realistic p=0 pressure point I would use the two top points at the top air edge, or even the full top edge via a weak constraint

--
Good luck
Ivar

Hi Ivar,

I'm also trying to quantify the 'perturbation profile' of the surface of the water reservoir after the droplet hits. Can you suggest how this can be done? I've tried specifying an Edge-2D node under Data Sets in the results/post-processing section, then added a 'maximum' evaluation node and afterwards added a 1-D line graph. But I don't know what to plot to show, say the maximum height of the reservoir surface at different time-t. Should I plot the y-coordinate under the mesh/geometry ? But I don't think the "Edge-2D"-line prove is representing the continuously deforming mesh.

Please help.

Thanks.

Hi

if you are using ALE method your edge/boundary deforms, but with level se t/ phase method you have no "edge". I do not know, but there must be a way, havw you checked the model library examples ?

The question to answer is probably: how to get the isocontour Phi = 0.5 ? as a field from the coordinates location

--
Good luck
Ivar

what is the procedure to simulate the water droplet falling through air on a structured surface?
please send me the step by step information of that which will be very useful to my work of nano surface coating.I am very much thankful to you.

Is there a model file for "water droplet falling through air" that I can download.
I have been working similar models for years.

Hi,
Even I'm in need of such a model.Could anyone tell me where I can find one?