CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > OpenFOAM Running, Solving & CFD

Stirred tank reactor convergence problem

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   September 29, 2015, 20:32
Default Stirred tank reactor convergence problem
  #1
New Member
 
kevin lea
Join Date: Sep 2015
Posts: 9
Rep Power: 3
kevinlea is on a distinguished road
Hi,

I am working on a stirred tank reactor project using OpenFOAM. From my research, it involves multiple reference frame (one rotating zone enclosing the rotating impeller, and a stationary zone outside the rotating zone).

I have made the mesh for the tank, impeller shaft, impeller, rotating zone, and stationary zone. I also define the interface boundary between the rotating zone and stationary zone as cyclicAMI (please correct me if this is not the right boundary definition). The rest of the boundary are fixed value for velocity and zero gradient for pressure.

The solver I am using is pimpleDymFOAM. I am running into the problem where the Courant number keeps getting larger during the simulation. I hope someone can take a look at what I have done wrong. I have attached my files below. Thank you.

Reactor mesh image:http://imgur.com/4FmRwEx
Reactor image: http://imgur.com/dy1cVub


Kevin Lea
Attached Files
File Type: zip Bioreactor new.zip (11.6 KB, 9 views)
kevinlea is offline   Reply With Quote

Old   September 29, 2015, 20:50
Default
  #2
New Member
 
Join Date: Jul 2015
Posts: 23
Rep Power: 3
nicholas.jones is on a distinguished road
Hey Kevin,

Although I am no expert, I have found that some of my previous convergence problems were due to poor mesh. You might want to analyse your mesh in greater detail, and consider using snappyHexMesh instead of Salome (looks like salome to me, I may be wrong).

More detail on your mesh may be helpful (checkMesh).
nicholas.jones is offline   Reply With Quote

Old   September 29, 2015, 21:14
Default
  #3
Senior Member
 
chegdan's Avatar
 
Daniel P. Combest
Join Date: Mar 2009
Location: St. Louis, USA
Posts: 585
Rep Power: 20
chegdan will become famous soon enoughchegdan will become famous soon enough
Kevin,

if you post your surfaces you are using; some information about your mesh e.g. the cell count you would like; and the simulation conditions then we may be able to help. Your case had no mesh and could not be run. Be more specific and I will see what i can do to help.
__________________
Dan

Find me on twitter @dancombest and LinkedIn
chegdan is offline   Reply With Quote

Old   September 30, 2015, 09:39
Default
  #4
New Member
 
kevin lea
Join Date: Sep 2015
Posts: 9
Rep Power: 3
kevinlea is on a distinguished road
Quote:
Originally Posted by chegdan View Post
Kevin,

if you post your surfaces you are using; some information about your mesh e.g. the cell count you would like; and the simulation conditions then we may be able to help. Your case had no mesh and could not be run. Be more specific and I will see what i can do to help.
Thanks Dan,

I don't know how to upload the mesh files because of file size limit. Can I email you the files?
I meshed with Salome with Netgen, but I think the culprit is how I define my cyclicAMI boundaries.

For the rotating zone enclosing the impeller, I exclude the impeller and impeller shaft volume, so that the rotating zone is a cylindrical donut with hollow impeller inside. Then I define the outer top, side, and bottom surfaces as cyclicAMI.

For the stationary zone, I exclude the rotating zone and impeller shaft volume, so it is a cylindrical donut with hollow rotating zone inside. I define the inside boundary touching the rotating zone also as cyclicAMI.

Is it necessary to exclude their containing elements' volume for both zone for the multiple reference frame setup?

Thanks
Kevin
kevinlea is offline   Reply With Quote

Old   September 30, 2015, 10:07
Default
  #5
Senior Member
 
chegdan's Avatar
 
Daniel P. Combest
Join Date: Mar 2009
Location: St. Louis, USA
Posts: 585
Rep Power: 20
chegdan will become famous soon enoughchegdan will become famous soon enough
If you are using MRF, there is no need to use Cyclic AMI (though it is possible). If you want to use dropbox and provide a public link i can give it a go to mesh your surfaces and get a steady simulation started.
__________________
Dan

Find me on twitter @dancombest and LinkedIn
chegdan is offline   Reply With Quote

Old   September 30, 2015, 11:39
Default
  #6
New Member
 
kevin lea
Join Date: Sep 2015
Posts: 9
Rep Power: 3
kevinlea is on a distinguished road
Quote:
Originally Posted by chegdan View Post
If you are using MRF, there is no need to use Cyclic AMI (though it is possible). If you want to use dropbox and provide a public link i can give it a go to mesh your surfaces and get a steady simulation started.
Dan,

Thanks for your help. I have attached the stl files of the parts. Please read the text file for clarification on my setup.

https://www.dropbox.com/s/zq9ae20kfg...actor.zip?dl=0

Thanks
Kevin
kevinlea is offline   Reply With Quote

Old   October 1, 2015, 03:26
Default
  #7
Senior Member
 
Dongyue Li
Join Date: Jun 2012
Location: Torino, Italy
Posts: 743
Rep Power: 9
sharonyue is on a distinguished road
Hi kevin,

When you are using sliding mesh, you have to make sure that your mesh is indeed sliding. Otherwise this means Courant number keeps getting larger. I suggest you check out your mesh, make a slice on the AMI patch to see if the patch is sliding smoothly, or its fixed.
__________________
Im the translator of OpenFOAM User Guide Chinese Edition! But always newbie on CFD.

http://dyfluid.com/en.html
sharonyue is offline   Reply With Quote

Old   October 2, 2015, 17:18
Default
  #8
New Member
 
kevin lea
Join Date: Sep 2015
Posts: 9
Rep Power: 3
kevinlea is on a distinguished road
Quote:
Originally Posted by sharonyue View Post
Hi kevin,

When you are using sliding mesh, you have to make sure that your mesh is indeed sliding. Otherwise this means Courant number keeps getting larger. I suggest you check out your mesh, make a slice on the AMI patch to see if the patch is sliding smoothly, or its fixed.
Hi Sharon,

Thank you for your feedback. I believe am not using the sliding mesh method. I am trying to implement MRF method to solve the rotation of the impeller. I hope you briefly explain what are the necessary boundaries that I need to define for this setup.

Thanks
Kevin
kevinlea is offline   Reply With Quote

Old   October 12, 2015, 14:34
Default
  #9
New Member
 
kevin lea
Join Date: Sep 2015
Posts: 9
Rep Power: 3
kevinlea is on a distinguished road
Hi All,

I am still working on this project. I hope I can get some clarification on what I should define the boundary conditions for the following elements in the reactor for the MRF approach.

1. Interface boundary between the rotating frame and the stationary frame
2. Impeller boundary
3. Impeller shaft boundary

Thanks
Kevin
kevinlea is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Stirred tank torque validation - understanding the moment report MCGI FLUENT 1 July 21, 2015 17:06
A confusing problem at making a wave tank divoone1985 CFX 3 June 21, 2015 06:30
A complete tutorial for stirred tank reactor wavalaniket Main CFD Forum 0 December 15, 2014 10:26
Stirred Tank Reaktor (transient Simulation) andon6141 FLUENT 7 December 12, 2014 04:36
CFX 5.5.1 stirred tank and LES problem Nishant CFX 2 September 13, 2002 07:11


All times are GMT -4. The time now is 23:58.