[muhanad]

Dear foamers,

I am new to OpenFoam, and currently have a task to alter the dambreak case into two different cases:-

1- Falling 2D drop from the dam top to hit the empty bottom
2- Falling 2D drop from the dam top to hit a film of water in the bottom.

my problem:-

Defining the alpha1 phase to the spherical/cylindrical droplet, I used the following two expressions which i feel correct in setFieldsDict:-

25 sphereToCell
26 {
27 centre (0.05 0.1 0.5);
28 radius 0.01;
29 fieldValues
30 (
31 volScalarFieldValue alpha1 1

and:-
25 cylinderToCell
26 {
27 p1 (0.05 0.1 0);
28 p2 (0.05 0.1 0.5);
28 radius 0.01;
29 fieldValues
30 (
31 volScalarFieldValue alpha1 1

in the blockMeshDict, I considered having 3 blocks:- one block for the circular droplet where its bottom is tangible to the dam square top from the centreline, and 2 blocks for the dam, left and right to the tangible point centreline (just consider a sphere sitting on a box in the centre).

now i have 4 vertices for the droplet,6 vertices for the dam (one vertex is shared between the three blocks: the tangible point).

well, i defined my vertices, the three blocks..
when it came to the droplet i defined four circular arcs, each arc passes through a point (i used inscribed equal rectangle in the the circle to determine these points).

I had a challenge on boundary part of the blockMeshDict:

I have walls to describe the dam left, right and bottom wall, as well as atmosphere to describe the top part.

how to describe the droplet in this case? and how to describe its type/values in alpha1 file and U, & P in directory (0)?

Would be grateful for your support..Maybe all the above is incorrect??!

Regards
Muhanad

[muhanad]

well, I have maintained everything in the dambreak case and just changed the setFields using cylinderToCell instead of box to cell, and altered the position of the droplet to be inside the domain on the top.

the simulation has captured the droplet fall successfully, just i noticed the slowness of the fall, like if there is a drag preventing it from falling quicker!..

Still need to investigate the situation more closely to find more about the required settings...would appreciate your inputs and discussions on this..

[muhanad]

I have now used the refineMesh and got a finer mesh on my droplet, which is supposed to be smaller, the simulation is getting slower really, not falling on real time

still awaiting some comments/ feedbacks on droplet simulation and case setup from your expereinces..

[arsimons]

Hi


I am currently also trying to simulate a droplet falling through air. I have found so far that it indeed falls faster when the mesh is coarser. I think this is inherent to interFoam, since the Courant number needs to be smaller than 1. My guess can be found in the following paragraph:


Since the Courant number needs to be smaller than 1, the droplet will always fall less than 1 complete cell for each time step. As a result, after the first step, the top part of the cell right below the cell initially containing the droplet now contains the droplet. However, this cell is the smallest part in the simulation and as such, the simulation will only remember that X% was filled with the droplet. In the following time step, these X% start to fall for less than 1 cell again, and the next cell is reached. This process will continue during the entire simulation.


It might be possible to solve this by using an even smaller mesh or smaller maximum value for Courant. However, I did not investigate this yet. Currently, I am trying interIsoFoam, since this might solve the problem. I will keep you informed!