[AS_Aero]

Hi

Can someone tell me how can I calculate the droplet velocity of water from a Spray or nozzle where the water is coming out of the nozzle in atomized droplets ?

[taillanm]

Just create a new scalar scene, as input parts your droplets and scalar velocity

[AS_Aero]

I dont have droplets as my parts, its just a VOF computation, and i dont predefine the droplets.

[taillanm]

How are the droplets generated?

[AS_Aero]

Its expected to generate automatically as our spray is atomization spray and due to the swirl inside the spray is generated and i expect by doing URANS and VOF I will get some breakup and droplets. (Which I am not sure though)
But based on my postprocess contour vor Volume fraction of water, I can see some water droplets and thin films of water. So now when I check the velocity, its the mixture velocity, which gives me wrong result. so now am finding a way to get it done.

[ptemp]

URANS and VOF won't create droplets by themselves. It will only change the volume fraction of water of the cells near the waterline (which you can visualize with an isosurface, specifying a certain volume fraction). If you want to analyze actual droplets, you have to run a Lagrangian Multiphase (LMP) simulation and then specify multiple multiphase interactions; namely LMP-Film, Film-VOF_phase1, LMP-VOF_phase1, etc. I don't know your specific case, so you have to determine how the different phases interact (or don't interact for that matter).

Once you have the LMP simulation set up, you must create a particle track file in order to visualize the droplets after the simulation has run. There you can specify multiple field functions on the droplets to visualize them in a scene.

LMP is pretty complicated, especially when you have multiple phases. If at all possible, I would suggest simplifying your simulation or making certain assumptions to get the output you want with the data you currently have.

If you're feeling ambitious, here is how to set up LMP, and here is info on LMP Impingement:

[AS_Aero]

Hi Paul

A small doubt, if I refine my mesh as small as the droplet size of 100 micro meter and then run a simple vof doesnt it work ? And cant i see the droplets and the breakups happening ?
To do a LMP simulation, according to my knowledge we will need to give the details of drplet size distribution and velocity distribution at the injection right ? (I have never done LMP) so am not familiar with it. So these interactions and all, I dont know how it works and why we need. My case is a simple swirl atomizer nozzle. Your link doesnt work, can you send it again, so that I can have a look to it.

[JBeilke]

Originally Posted by AS_Aero

Hi Paul

A small doubt, if I refine my mesh as small as the droplet size of 100 micro meter and then run a simple vof doesnt it work ? And cant i see the droplets and the breakups happening ?
To do a LMP simulation, according to my knowledge we will need to give the details of drplet size distribution and velocity distribution at the injection right ? (I have never done LMP) so am not familiar with it. So these interactions and all, I dont know how it works and why we need. My case is a simple swirl atomizer nozzle. Your link doesnt work, can you send it again, so that I can have a look to it.

Then you have to refine your mesh, so that you can resolve the surface waves for each single droplet. This also means that your timesteps become very small.

[ptemp]

Forgive me if I'm wrong, but if the water is atomized, can't you just make the assumption that the atomized water droplets have the same velocity as the flow? If so, just create a few cut planes to visualize the flow field or create a resampled volume and look at the flow velocity there.

Here are the links I referenced in my previous reply:
How to set up LMP: file:///C:/Program%20Files/CD-adapco/12.04.010/STAR-CCM+12.04.010/doc/en/online/index.html#page/STARCCMP%2FGUID-1B086E21-3898-472C-9559-AD1225574BFC%3Den%3D.html%23

How LMP Impingement Works: file:///C:/Program%20Files/CD-adapco/12.04.010/STAR-CCM+12.04.010/doc/en/online/index.html#page/STARCCMP%2FGUID-7A323811-291C-4503-93BA-840661D954AB%3Den%3D.html

[JBeilke]

Thats right. You might try this approach just for the atomisation itself if you want to avoid any additional breakup models. But you will end up somewhere between DNS and LES resolution . So it is not a practical approach at all.