[harry123]

Dear all,

I am starting to do simulation of water jet from nozzle into ambient atmosphere.

It would be nice if people doing similiar simulation get together and discuss the problems in such simulation and little tricks that may help for getting better results.

So, if you are someone stuck up in water jets, please comment below anything and everything you feel others should know about your cool ideas..!!

Regards,
Harry

[Rob87]

Hi harry123,

is it possible to get more details? What do you wanna achieve? Which kind of physics you wanna use?

[harry123]

Hi,

I have to study the characteristics of water jet coming out from a nozzle of dia around 2 mm ( Variable, but normally in that range).

The physics models that I was planning to use in Star CCM was : Eulerian Multiphase - Implicit unsteady - Volume of Fluid Method with water as primary phase. Regarding the turbulence model, I haven't yet decided, I might first use a k-epsilon RANS model and then LES later maybe.

As this is my first time with star CCM, I'm also not sure if the above models are the best suited for this problem.

Also, at the moment, I'm trying to get a proper mesh. Any idea what should be the smallest size if I'm trying to simulate a jet with 2 mm dia ?

Harry

[Rob87]

Okay. And with which velocity range (Reynoldsnumber) do you wanna simulate? Currently I simulate a similar thing. In my experience you should use LES if you wanna see a primary breakup. In literature you can find that turbulence is a key factor for that occurrence and no turbulence model can provide you with the needed physics.
For LES you need a very fine grid to solve the energy directly.

best regards

[harry123]

I would be using velocity range of 80-150 m/s, so for sure in turbulence regime..!!

But before I venture into turbulence, is it a good idea to keep the velocity in laminar range (so that it won't take long times) and test my mesh to make sure there is convergence ?

Also, which boundary conditions you use ? my domain is a rectangular box of dimensions 200x60x60 mm. The longer side being along the x direction and jet path. A circular inlet is placed on left, dia 2mm. The far right boundary and also the face on which the inlet is placed are of wall type. The rest 4 boundaries are pressure outlets with fixed atmospheric pressure.

[Rob87]

Have a look to this presentation:
http://www.first-fp7project.eu/media...mont_CORIA.pdf
Here you can get a feeling about your grid size. You could test your mesh with a laminar flow, but most problems appear with specific dissipation rates in the prism layers.
Your boundary conditions sounds alright.

[cwl]

Unfortunately i'm stuck with the same problem.
Direct simulation of drops formation and break-up is impossible due to required mesh size.
In Star-CD there's a model ELSA - in which detached droplets are modeled as Lagrangian particles, while the core of the jet is treated using Euler approach.
http://www.cd-adapco.com/products/st...spray-modeling

But - unfortunately i can't see anything similar in Star-CCM+.

[harry123]

What mesh size are you using now ? relative to your jet diameter size

[Rob87]

My cell size is 1/70 of the nozzle diameter.

[harry123]

Thanks for the info, actually I was using much higher mesh size than that, around 1/15 th of jet diameter.

I was thinking along the line that, my interest are primarily on the average flow values like pressure on the traget plate, variation of pressure, velocity and volume fraction in radial direction etc. Hence, I can stick to a larger mesh size. But I'm not sure if that sounds correct.

My major concern is about dependence of VOF method on mesh size, as I vary mesh size or time step, I have quite different results !!

[richengle]

has any more work been done on water jets?