Gravity Vector Direction for One-Way Particle Track

FrankS · July 20, 2016, 15:33

Hi All,

I am modelling a machine consisting of a high-speed (Re-independent) rotor enclosed within a vat. The working fluid is water. I have been using the frozen rotor approach to account for rotor rotation. Buoyancy does not affect the flow field so I have not included it yet (it will be important in the next step though and that is why I mention it). I have been tweaking the design of this machine based on a number of flow simulations and I am now quite happy with the flow field.

One of the important functions of this machine is to separate particles suspended in the water. Low density particles float up and high-density particles sink down. Obviously then, I need to include buoyancy if I am to introduce particles to the flow domain. When I include buoyancy, the gravity vector will be perpendicular to the spinning rotor's axis of rotation. This means I can't use the frozen rotor method anymore and now my simulations will go from 2-hours each run to 2-weeks as I must adopt the full transient method.

I only need to inject a small number of small particles into my simulation domain and one-way coupling is plenty good enough. I understand that one way coupling is calculated after the fact as the particles do not influence the flow. Is it possibly to change the gravity vector direction just for the particles in CFD post so that I could use the frozen rotor method to develop the flow? I would even be happy if I could introduce particles into a snapshot of the flow field at a given time and just see where they go based on the static snapshot.

I am looking for a quick way to run simulations so that I can make the final tweaks to my design before I go all in and run a full transient simulation for two to three weeks.

Sorry about the lack of details on the machine itself. It is novel and it should offer a massive performance improvement over existing machines so I am keeping it secret for the moment. Any help is appreciated.

Frank

ghorrocks · July 20, 2016, 22:11

Are you sure you can't use gravity with frozen rotor?

FrankS · July 21, 2016, 15:14

The CFX manual states that gravity can be used with the frozen rotor method only if the gravity vector is parallel to the rotor's axis of rotation. The design of this machine is such that the gravity vector is perpendicular to the rotor axis of rotation. I tried anyways with the frozen rotor and the gravity vector perpendicular and the solver threw an error stating specifically that that is not allowed.

I bet what I want to do is not an option. I am trying to make some really coarse meshes so that if I am limited to a full transient sliding mesh simulation at least it will be a bit quicker for the final CFD design phase for this machine.

ghorrocks · July 21, 2016, 19:21

You are right. The only suggestion I can make is to try some of the options and see what works for you. But that is pretty obvious really

July 20, 2016, 15:33	Gravity Vector Direction for One-Way Particle Track	#1
FrankS New Member Frank Join Date: Oct 2013 Posts: 26 Rep Power: 13	Hi All, I am modelling a machine consisting of a high-speed (Re-independent) rotor enclosed within a vat. The working fluid is water. I have been using the frozen rotor approach to account for rotor rotation. Buoyancy does not affect the flow field so I have not included it yet (it will be important in the next step though and that is why I mention it). I have been tweaking the design of this machine based on a number of flow simulations and I am now quite happy with the flow field. One of the important functions of this machine is to separate particles suspended in the water. Low density particles float up and high-density particles sink down. Obviously then, I need to include buoyancy if I am to introduce particles to the flow domain. When I include buoyancy, the gravity vector will be perpendicular to the spinning rotor's axis of rotation. This means I can't use the frozen rotor method anymore and now my simulations will go from 2-hours each run to 2-weeks as I must adopt the full transient method. I only need to inject a small number of small particles into my simulation domain and one-way coupling is plenty good enough. I understand that one way coupling is calculated after the fact as the particles do not influence the flow. Is it possibly to change the gravity vector direction just for the particles in CFD post so that I could use the frozen rotor method to develop the flow? I would even be happy if I could introduce particles into a snapshot of the flow field at a given time and just see where they go based on the static snapshot. I am looking for a quick way to run simulations so that I can make the final tweaks to my design before I go all in and run a full transient simulation for two to three weeks. Sorry about the lack of details on the machine itself. It is novel and it should offer a massive performance improvement over existing machines so I am keeping it secret for the moment. Any help is appreciated. Frank

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
particles leave domain	Steffen595	CFX	9	March 7, 2016 17:19
injection problem	Mark New	FLUENT	0	August 4, 2013 02:30
Check particle impaction with User Fortran	Julian K.	CFX	3	January 12, 2012 10:46
DPM UDF particle position using the macro P_POS(p)[i]	dm2747	FLUENT	0	April 17, 2009 02:29
Particle track files	Sankalp	CFX	1	August 26, 2004 00:25

July 20, 2016, 22:11		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Are you sure you can't use gravity with frozen rotor?

July 21, 2016, 15:14		#3
FrankS New Member Frank Join Date: Oct 2013 Posts: 26 Rep Power: 13	The CFX manual states that gravity can be used with the frozen rotor method only if the gravity vector is parallel to the rotor's axis of rotation. The design of this machine is such that the gravity vector is perpendicular to the rotor axis of rotation. I tried anyways with the frozen rotor and the gravity vector perpendicular and the solver threw an error stating specifically that that is not allowed. I bet what I want to do is not an option. I am trying to make some really coarse meshes so that if I am limited to a full transient sliding mesh simulation at least it will be a bit quicker for the final CFD design phase for this machine.

July 21, 2016, 19:21		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	You are right. The only suggestion I can make is to try some of the options and see what works for you. But that is pretty obvious really