[StormedStar]

Hi to all,

I am trying to simulate a water droplet in a filament (the filament is modelled as a wall), I have simulated all for a inlet of air of 1m/s only activating the VOF model (water-air as water as the principal) and laminar.

The problem is that in the first 0.3 seconds, most of the water escapes from the droplet and difusses through the air to the outlet and inside the droplet you can only find air.

How can I mantain inside the droplet the water?


Thanks for all

[jpina]

I would try to compare your setup with the one in this manual:

https://www.researchgate.net/file.Po...01441967624980

[StormedStar]

Thanks for the fast reply :

I have tried this model before, but the main thing is:

I have a wall in form of a filament (rectangular with a circular end) that is connected to the droplet and the main differences to the other model are:

1.- I use explicit in VOF because the courant number makes me use a very low step size and my simulation is for 40 seconds, so I don´t want to wait a lot for analyse the results. (It would be nice to be able to use the implicit with low computation time)

2.- I use air as primary because I enter air constantly from the inlet, to put the water I use patch and I specify water as 0 vol fraction and for air 1

3.- I don´t use any velocity distribution function, I only specify in the inlet a velocity of 1 m/s

Thanks for all

[jpina]

I would appreciate some screenshot... It's difficult for me to imagine your problem...

[jpina]

Btw, with explicit solver you won't get a sharp interface. Implicit solvers are recommended when the interest is focused on the interface between phases.

I would recommend you to use implicit solver with transient time step, fix the Courant number to 2 and try if it converges and goes fast.

[StormedStar]

Dear jpina,

With Implicit I cannot adjust the courant number, I attach you the geometry of the problem, the inlet is the bottom part and the outlet the upper part.

Water inside the droplet is escaping and finally at time 1 (very fast) there´s only air inside, furthermore, I have put the surface tension in the model and it´s the same.


Thanks for your help

[jpina]

Excuse me, I meant that you need to use explicit scheme.

You can use Variable time Step (in the Run Calculation part) and there fix a global courant number of around 2. Let the minimum time step size be very low (1e-10).

Try this and see if it converges and is fast enough.

[StormedStar]

Dear jpina,

I tried that, but the time step size is always small and it takes a lot of time, but for the time that I have observed, I have seen that the water still goes out of the drop, so I think that is a problem of the configuration, so I will tell which it is:

Models:

VOF (default values) with water as principal and air as secondary

Laminar Flow

Phases:

In interaction I specify surface tension with wall adhesion of 0.0735 N/m

Boundary Conditions:

Walls by default and inlet I specify 1 m/s and 1 atm of gauge pressure, also I specify vol fraction of air of 1 because air enters in the system. For the outlet I put 1 atm of gauge and backflow vol fraction of 1.

Finally the solution is as you suggested, and after making the initial solution, I patch a vol fraction of 1 in the air system and 0 in the drop system.

Then the error of water leaving appears in the first seconds...


Thanks for all

[jpina]

Mmm... In order to assure you don't have something wrong related to pressures (operating pressure, etc.).

Look at the "absolute pressure" values and see if those are the ones that should be.

[StormedStar]

Dear jpina,

From the observation of the results I have seen that the pressure increases dramatically from time 0 (pressure of 0 according to the pressure plot, because I put 0 of gauge in the inlet but 101325 Pa in operating conditions) to the last time.

I don´t know what can cause this dramatical increase of Pressure


Thanks for all