Discussion Groups

Simulation CFD

Reply
New Member
dcimarkh
Posts: 2
Registered: ‎08-01-2012

Pulsation in convergence plot

262 Views, 6 Replies
11-18-2013 02:56 PM

I have run an analysis of a small but complicated flow path using air as the medium.  The convergence plot settles down at around 200 iterations.  After that all of the velocity values are pulsing at the same rate.  the other values are smooth and steady.  I have a jpeg image of the plot attached.  What would cause this?  Does this mean the flow is pulsating due to some resonance or something?  Can I trust the flow and pressure values that I am getting?  Should it be run as transient instead of steady state?

Please use plain text.
Product Support
wildej
Posts: 728
Registered: ‎08-25-2011

Re: Pulsation in convergence plot

11-18-2013 10:58 PM in reply to: dcimarkh

Hi,

 

It is possible that this is really a transient solution but I would suggest it is more likely down to setup or lack of mesh.

 

Could you post an image of the overall setup with Boundary Conditions?

Also, what does your mesh look like (on a cut-plane)?

 

If the setup is OK, my first step would be to refine the mesh by 0.7 and re-run (an uninformed guess at this stage).

 

Kind regards

Jonn



Jon Wilde
Please use plain text.
New Member
dcimarkh
Posts: 2
Registered: ‎08-01-2012

Re: Pulsation in convergence plot

11-19-2013 11:26 AM in reply to: dcimarkh

The circuit is pretty complicated so it is not easy to show.  However I will take your advice and try refining the mesh to see if that gives an improvement.

Thanks for your help.

Please use plain text.
Product Support
wildej
Posts: 728
Registered: ‎08-25-2011

Re: Pulsation in convergence plot

11-19-2013 01:55 PM in reply to: dcimarkh

Hi,

 

Sounds like a plan, please let us know how you get on.

 

What air speeds are you seeing? It just crossed my mind that perhaps you are reching potential compressible flowrates.

 

Kind regards,

Jon



Jon Wilde
Please use plain text.
Member
scott.biba
Posts: 4
Registered: ‎12-03-2013

Re: Pulsation in convergence plot

12-03-2013 03:46 PM in reply to: wildej

I was previously logged in as dcimarkh, but now, finally, correctly logged in as me.  Same computer, same analysis model.

I tried using the support from the program menu but it says that I don't have subscription service.  When i try to login it asks for my contract number and I have no idea what that is.

 

I tried refining the mesh using regions this time, instead of selecting surfaces, as I did before.  I thought it would make it easier for the program to remember where I had control meshes.  Before, when I just selected surfaces, if I had a geometry change and it had to resize the mesh, all of my controls would go away and I would have to recreate them from scratch.  Using regions helped with keeping track of locations, but for some reason with a geometry change, it loses track of the mesh size setting, which kind of sucks.  So I have to update the definition of all of them each time I make a change.

But that is not my primary issue at this time.

Along with defining the regions to refine the mesh, I also updated some other aspects of the design.  I can generate the mesh, but that is as far as the analysis will go before I get this message: "The Analysis has stopped because the Solver has exited unexpectedly".  Sometimes it pops up by itself and sometimes the analysis hangs (for a very long time) before I stop it and the message comes up then.

I know that in the documentation that it suggest 4 things.

  1. memory problems - I have 16 GB of ram and we are working in the cloud, so this shouldn't be an issue
  2. poorly posed setup parameters - I have reviewed all of them and don't see any problems here.  Though sometimes the setup procedure is vague for creating restriction, etc. and there appears to be no way to have it show you what it thinks it is doing.  When you create a restriction it asks for a surface, OK, but I am assuming that it knows where the opposite surface is.  It doesn't say and there is no graphic representation to make it show you what it just did. 
  3. file I/O errors - I have no idea
  4. Incompatible scalar diffusion coefficient value - the general scalar formulation was off.  I tried turning it on and using the recommended value, but no change in the error.

I have tried varous things for over a week to get this resolved and could use some assistance.

 

I have created a share file that I could send if someone can help.

 

Thank you.

 

Please use plain text.
Member
scott.biba
Posts: 4
Registered: ‎12-03-2013

Re: Pulsation in convergence plot

12-03-2013 09:06 PM in reply to: scott.biba

So apparently you replied by email to Mark.    As I stated in my post, I was using Mark's login before and am now using mine.  He is not working on this problem, but keeps having to forward me the messages that he gets.

My email is scott.biba@design-concepts.com

 

The instructions were as follows.

 

To manage your subscriptions or to change your subscription options.

  1. Log      on to Autodesk Community.
  2. Click      My Settings.
  3. Click      the Subscriptions & Bookmarks tab.
  4. Change      your settings and click Save Changes.

Thanks for being a Autodesk Community member.

Your Autodesk Community Team

 

 

But you didn't say what settings to change.  I don't see anything relevant in the options.

I would really like to get the analysis problem resolved.  How do I get in touch with you or someone else from support?

Thank you.

 

Scott

Please use plain text.
Product Support
wildej
Posts: 728
Registered: ‎08-25-2011

Re: Pulsation in convergence plot

12-04-2013 01:54 AM in reply to: dcimarkh

Hi Scott,

 

I think we are just replying here, the updates might just be linked to the wrong person due to the login. I hope that you will see this reply.

 

It looks to me like you have a support subscription for Sim 360. I will pm you your serial and contract numbers. Please let me know if you are unabe to log a case though.

 

Kind regards,

Jon



Jon Wilde
Please use plain text.