[quarterlbr]

I need some input. I've been working on a problem where I have steam flowing through a pipe and I spray water into the pipe via a nozzle (so, multiphase with evap/LaGrangian). I have data for the nozzle that tells me the nozzle characteristics, e.g., droplet size distribution, vs flowrate. This data was developed in a lab under still ambient air conditions spraying water.

If I have the droplet size distribution obtained as described above, do I still need to implement the secondary breakup droplet model (TAB, R&D, etc.), or do I have everything I need to model the spray evaporating in the steam pipe? I do get different model results when I do (temperature distributions), but I wonder if I am wrong in incorporating secondary breakup models.

Any help is appreciated.

Thanks,
Charlie

[Edeluc]

Dear Charlie,

Breakup models (primary and secondary) give a mathematical formulation for the breakup of the liquid into droplets - Therefore only should be used i you do not have any information on the droplet size distribution. One thing to make sure first is to check, how/where the droplet size distribution has been measured. If it is far enough downstream of the nozzle where all the breakup mechanisms have already taken place, then specifying a droplet spectrum suffices (in my opinion).

If you switch on the breakup model and also specify a droplet spectrum, your results will definitely be different. Your specified droplets are then broken up into even smaller droplets which changes their mass and therefore you influence penetration lengths (liquid and vapor) as well as the evaporation behaviour (which you will probably see in your temperature distributions as you said).

Best regards,
Lucas

[quarterlbr]

Thank you, Lucas.

Yes, I have noticed the results are very different with the secondary breakup turned on. I beileve the measurements were taken far enough away from the nozzle that would indicate the break up was complete, however, these measurements were taken in a still room, not in a moving fluid, so I didnt know if that should impact my modeling or not, which does include injecting into a bulk fluid that moves up to 45 m/s. This is my conundrum - whether or not the relative velocity of the fluid and droplets would cause further breakup of the droplets.

Given this fact, would you still ignore the secondary breakup model from the cfd code and just use the droplet distribution, or would you include it?

Thanks very much for your reply.

Charlie

[Edeluc]

Dear Charlie,

Aerodynamic breakup will most certainly change the droplet breakup.

As far as I understand it you do have measurements of the droplet distribution but nothing else?

If you have any further measurements, try to do several simulations (breakup switched on/off, droplet spectrum specifed/not specified) and see, which results fit best. This "brute force" method should give some result basis for further conclusions.

If you do not have any further measurements, I would suggest that you find out which breakup mechanism will influence your liquid phase the most. If this breakup mechanism is captured by the droplet spectrum and others are negligible, you can proceed with the droplet spectrum specification. If it is not, then you have to activate a breakup model that captures this effect. In your specific case I would suggest to include the secondary breakup, as you do have relatively high bulk flow velocities.


Regards,
Lucas