Air volume fraction in Multiphase flow in Hydrocyclone

RyanBari · August 6, 2017, 19:02

Hi there
I wanna simulate flow within a hydrocyclone and I have a problem when I use free surface method. First of all, I run single phase in steady state as a initial condition for multiphase flow. Then, when I run for instance for 1000 iteration
(for reaching stationary state)in multiphase state, I don't have any continuous air core in the center of hydrocylone. Just it's seen a little air in the outlets( I mean overflow and underflow). Please see the attached photos( the left one is my problem and the right one is my desire result). So what do I do to solve this problem in order to have a continuous air core.
Thanks in advance.

ghorrocks · August 6, 2017, 19:57

FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F

First of all make sure your simulation is accurate. Do a sensitivity study on convergence criteria, mesh size and time step size.

Once you have done that and if the problem still exists have a look at how appropriate your physical models are. Cyclones are a classic case of where 2-eqn RANS turbulence models don't work. You might need to consider RSM models or LES models.

Some background reading into cyclone modelling is recommended. There is lots of stuff in the open literature about it.

RyanBari · August 7, 2017, 05:55

Thank you for replying
First of all, I should to check grid independent. For this reason I've simulated that and I faced that problem I told. I've studied some papers about my thesis and the authors checked some RANS and LES turbulent models. I simulate with nearly all of turbulent model in CFX and there is still that problem. I have a hybrid mesh I mean with structure and unstructure grid. Can it be a problem?
I've completely confused. please help me ASAP

ghorrocks · August 7, 2017, 09:03

Can you confirm whether you have done a grid independence check or not? You comment was not clear.

Once you have checked the grid you then also need to check the time step size and convergence criteria.

And don't forget that a grid which gives grid independent simulations for one turbulence model might not be grid independent for another turbulence model. And if you are doing LES models you need to size your grid based on the turbulence length scale and turbulence spectrums.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Other] mesh airfoil NACA0012	anand_30	OpenFOAM Meshing & Mesh Conversion	13	March 7, 2022 18:22
how to set periodic boundary conditions	Ganesh	FLUENT	15	November 18, 2020 07:09
How to use a UDF to set the volume fraction in the cells next to a wall?	DF15	Fluent UDF and Scheme Programming	33	August 20, 2020 14:36
multiphase flow ，mass and volume fraction imbalance	sope111	CFX	16	September 3, 2018 01:10
[blockMesh] Axisymmetrical mesh	Rasmus Gjesing (Gjesing)	OpenFOAM Meshing & Mesh Conversion	10	April 2, 2007 15:00

August 6, 2017, 19:57		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F First of all make sure your simulation is accurate. Do a sensitivity study on convergence criteria, mesh size and time step size. Once you have done that and if the problem still exists have a look at how appropriate your physical models are. Cyclones are a classic case of where 2-eqn RANS turbulence models don't work. You might need to consider RSM models or LES models. Some background reading into cyclone modelling is recommended. There is lots of stuff in the open literature about it.

August 7, 2017, 05:55		#3
RyanBari New Member Join Date: May 2017 Posts: 14 Rep Power: 9	Thank you for replying First of all, I should to check grid independent. For this reason I've simulated that and I faced that problem I told. I've studied some papers about my thesis and the authors checked some RANS and LES turbulent models. I simulate with nearly all of turbulent model in CFX and there is still that problem. I have a hybrid mesh I mean with structure and unstructure grid. Can it be a problem? I've completely confused. please help me ASAP

August 7, 2017, 09:03		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Can you confirm whether you have done a grid independence check or not? You comment was not clear. Once you have checked the grid you then also need to check the time step size and convergence criteria. And don't forget that a grid which gives grid independent simulations for one turbulence model might not be grid independent for another turbulence model. And if you are doing LES models you need to size your grid based on the turbulence length scale and turbulence spectrums.