[afsaneh]

hello
i study on simulation particle tracking in a hydrocyclone to predict particle separation efficiency. in my system particles are exist in water with 2% volume fraction.
i found some method for simulation: 1)first run my fluid simulation without solid (K-epsilon viscous model) then run DPM 2) run eulerian and DPM multiphase model both 3) run mixture and granular multiphase model

with this regrade, i want to now for my case (2% solid) which method is better?

i have one more question: when i choose 1 or 0.1 or 0.01 step size with 1000 iteration i face AVG problem for k or epsilon but if i choose 0.001 step size with 1000, i have no problem in fact it runs just for 1 s flow time!! why this problem occurs ?

i really appreciate you if you can help me

[CeesH]

To answer which is better; that is something you either have to test yourself, or draw from literature. All methods have their pros and cons (regarding the options, I assume you mean DPM OR Eulerian and not and?)

related to your question, you have to select a timestep size appropriate to capture the physics in your problem. It seems this is around 0.001s for your particular case. The fact that you only run for 1 second is logical, 1000 timesteps of 0.001 second adds up to 1 second.

[afsaneh]

Originally Posted by CeesH

To answer which is better; that is something you either have to test yourself, or draw from literature. All methods have their pros and cons (regarding the options, I assume you mean DPM OR Eulerian and not and?)

related to your question, you have to select a timestep size appropriate to capture the physics in your problem. It seems this is around 0.001s for your particular case. The fact that you only run for 1 second is logical, 1000 timesteps of 0.001 second adds up to 1 second.

dear CeesH:
hi
thanks for your help.
yes, i choose these method from literature but i don't know which one is the most suitable for my case because of low solid volume fraction. in ansys 15 i understood that i can use 2 phase eulerian multiphase model which one phase is my continuous phase and one discrete phase mosel.
about my second question, i meant that if my total flow time be just 1 second
isn't it gave me wrong efficiency, because separating particle just in 1 s is not logical