Hello

I am dealing with a multiphase problem with air and water as two medium. In addition to that, I am sending some steel balls inside that medium. I am having a doubt to select whether discrete phase model or multiphase model.

Also in injections, how I can specify that this much no of balls to be injected.

Regards Karthick

Hi, How your system {air + water} is flowing? If it is a stratified flow (means with a free surface) you can coupled VOF model and DPM. If your air is behaving as a discrete phase you can both way use DPM (for low volume fraction of air and steel particle) or the Euler-euler model for higher range of volume fraction. For your second quetsion i do not undertsand -> sorry.

Cheers thomas.

Sir I'll tell u the exact problem of mine. We are having a U shaped section with bottom area filled with water and above portion is filled with air for some proportion. In this stage, we are keeping some 20 balls on the free surface of the section and again water is send with some velocity. I want to know how much time the balls are taking to leave the whole section. This is like our western toilet problem in our day to day life where we are sending water to flush the particles inside the closet.

Also regarding to the second part of my problem, as I've already explained, the balls I am sending in the form of injections in DPM. That's why I am asking how I can send 20 nos of balls.

- Karthick

Hi, I found your problem interesting so i am proposing to you a solution which could fit in theory. If i am right you want to simulate the following experiment. A U shaped section tube full of water with on a side of a U 20 balls in stagnation at the free surface of the water. So what I would propose to you is: Design your U tube with a velocity Inlet for the inlet and whatever you want for the outlet. To set the number of balls at 20 you have to simulate you stuff 'UNSTEADY'. Using 'stochastic model', knowing the number of tries and your time step, you can play on the injection time to get the number of particle you want. That can be assimilate to a dirac impulse. You intialize your domain with no velocity. You set your velcotiy inlet to the conditions you want. And then you set your DPM injection time depending on your time step. I would personally make the whole simulation unsteady.

Hope this help. thomas

Hope this help. Thomas

Sorry a correction about how to fix a number of particle in your domain, i have done a mistake. In unsteady simulation, there is 1 particle injected per time step. So in fact to inject a 20 particle dirac impulse you have to set 20 points of injection and fit your mass flow rate.

just on thing: in unsteady simulation, you can inject more than 1 particle per time step, the number depends on your injection type (from surface, point, group...). For exemple if you inject particle from a surface containing 100 mesh facets, you will inject 100 particles per time step.

sir

Is this possible to add particle-particle interaction also in DPM model.

- Karthick

you've secondary atomization models available with dpm: collision, breakup and dynamic drag models. For more details, you can read Fluent6 User's Guide: chapter 24.1.Spray models