Hi

I'm working on a CFD code for which I'm trying to implement the VOF technique, in order to handle free surface flows. The code is partly working, but I need to validate the VOF results, and would like to know if anyone has any suggestions as to where to find litterature on validating VOF results. Any idears how to approach the subject are welcome.

Kristian

Hi,

Martin and Moyce (1952) contains a very nice experimental study of the collapse of liquid columns and broken dams. It is considered to be a classical bechmarrk for free-surface flows (the original reports of SOLA-VOF, SMAC, MAC used it as a reference). The full reference is:

Martin, J. C., Moyce, W.J., ' An Experimental study of the collapse of liquid columns on a rigid horizontal plane', Philos. Trans. R. R. Soc. London A, no. 244, pp. 312-324, 1952.

Another good one is Chandra and Avedisian (1991) which contains some nice pictures of a impinging droplet on a solid flat surface (if you are interested on impact simulations). The reference is:

Chandra, S. Avedisian, C. T.' On the collison of a droplet with a solid surface' Proc. Royal Soc. London A, 432, pp 13-41, 1991.

If you do manage to get a good agreement with these two, I believe your code is more the validated.

Cheers,

Neyval

Have a look in the following Los Alamos reports:

SOLA-VOF: A Solution Algorithm for Transient Fluid Flow with Multiple Boundaries, Rep. LA-8355, 1980

RIPPLE: A Computer Program for Incompressible Flows with Free Surfaces, Rep. LA-12007-MS, 1991

You will find some standard (2D) cases computed there (broken dam, tank reorientation, etc.).

If you want a difficult test for your implementation of VOF, perform the following 2D test:

1. Setup your 2D domain, e.g. 100 by 100 elements.

2. Arrange velocity field to form a circular flow pattern. The flow field is set once and remains constant.

3. Arrange a rectangular shape somewhere in your domain. For this, set VOF=1.0 for say 10by10 elements, the rest of the elements have VOF=0.0

4. Rotate your rectangular shape by 360 deg. (assume no diffusion) and see what remained from your initial shape. Your initial and final shapes should ideally be the same, but they very seldom are.

good luck

DML

I have quite a bit of experience in this area and suggest a few steps for code validation.

1. As per John's response, you may want to test your code on a suite of increasing complex advection cases, Ryder & Kothe (1995), AIAA paper. This includes standard translation and rotation of a circular body, single vortex and complex deformation fields.

2. If you want to include surface tension effects, then I would recommend you simulate an initially circular droplet in a periodic domain without gravity. The results will probably surprise you by showing the generation of spurious currents for a problem which should remain as a circle with no velocity field.

3. Next turn to real problem such as the classic dam break problem, etc.

Good luck.