Discussion Groups

Simulation Mechanical and Multiphysics

Reply
Contributor
chikhlas
Posts: 14
Registered: ‎08-13-2011

Re: steady state heat transfer

09-01-2011 06:44 AM in reply to: INACTIVE_AstroJohn

where can i send this file? can you confirm me any e-mail address, for support team ... because it is not possible to upload file here in this forum due to size limit

Please use plain text.
Employee
hupn
Posts: 44
Registered: ‎04-15-2011

Re: steady state heat transfer

09-13-2011 08:43 AM in reply to: chikhlas

Hi,

 

I feel that it is better for you to setup a smaller model to test your issue. It will help you to identify the problem too. If you still observe the similar behavior, then you can attach the smaller model so that we can take a look.

 

hupe

Please use plain text.
Member
ceckel
Posts: 5
Registered: ‎07-03-2012

Re: steady state heat transfer

07-03-2012 09:16 AM in reply to: chikhlas

I see this thread has not been updated in a while but I am having essentially the same problem. I simplified it down to a beam with an applied temperature on each end of the beam (different temperatures). I seem to get the 0°C part temperature when my mesh size is too small or conversely I have too many elements. I have not confirmed if it is one or the other. It seems my threshold is approximately 100,000 elements. I saw from the log in this thread that the OP had over 200,000 elements. Is there a resolution to this problem?

Please use plain text.
Active Contributor tim
Active Contributor
tim
Posts: 37
Registered: ‎10-10-2012

Re: steady state heat transfer

04-09-2013 05:40 AM in reply to: ceckel

John,

I'm like the previous author (cekel).  I've only been using the thermal facility of ASM for about 24 hours so I haven't tried everything.  What I've been doing is applying several convection and radiant loads to a flat steel plate (600mmx600mmx10mm thick).  The ambient temperature within both of these dialog boxes has been set to 40 deg C.  On a number of my runs, I've noticed that parts of the model were 0 deg C as well??

 

I'm using 2013SP2.

 

One thing I haven't done in every cas is to apply convection/radiation loads to every surface.  Some times, I've left the thin edges without anything.  Not sure if this is a problem.  I assumed it means they become insulated.

 

Tonight, I revisited my model.  I had over 200 000 elements with 4 elements thru thickness.  I dropped this to about 30 000 elements (2 thru thickness).  No difference!  I got a minimum of -78 deg C.  Should have been close to 40 deg C.

 

Next i switched off midside nodes, no other change.  Got minimum of +91 deg C which is probably ok.

 

Your thoughts?

 

Tim

Please use plain text.
Active Contributor tim
Active Contributor
tim
Posts: 37
Registered: ‎10-10-2012

Re: steady state heat transfer

04-09-2013 05:42 AM in reply to: tim

John, Sorry, the last result was using ASM 2014 but I seem to have similar problems with ASM 2013.

Tim

 

Please use plain text.
Member
ceckel
Posts: 5
Registered: ‎07-03-2012

Re: steady state heat transfer

04-09-2013 05:48 AM in reply to: tim

I had this problem addressed in another thread. The suggestion in that thread was:

 

My guess is that you can get accurate results by reducing the "Stiffness" of the Controlled Temperatures. According to the documentation, the stiffness should be "(2-3 orders of magnitude higher than the conductivity of the materials in your model)".

 

This worked for me. Good luck

Please use plain text.
Active Contributor tim
Active Contributor
tim
Posts: 37
Registered: ‎10-10-2012

Re: steady state heat transfer

04-09-2013 09:44 PM in reply to: ceckel

cekel,

thanks for this.  I couldn't check this because I'm not using any controlled temperatures.  I still have the problem.

 

Any suggestions?

Please use plain text.
Valued Mentor
AstroJohnPE
Posts: 494
Registered: ‎08-30-2012

Re: steady state heat transfer

04-12-2013 09:14 AM in reply to: tim

Hi,

 

We need to know a whole lot more about your model. For example,

  1. Oops. I see from the subject line that you are performing steady state. Sorry. Are you performing steady state or transient heat transfer?
  2. If the ambient conditions are 40 degrees, wouldn't the solution be 40 degrees? Oh, you have other loads that are causing heat? :smileywink: What are the other loads?
  3. What is the input for each of the loads?
  4. Since you have radiation in the analysis, the solution is nonlinear and requires iterations to converge. Is it converging? Does it reach the maximum number of iterations that you specified and just continues with a bogus answer? What is your input for the convergence? ("Setup > Model Setup > Parameters > Advanced" tab if I remember correctly.)

As Cekel indicated, extreme input can be a cause of the problem, such as a large stiffness with a controlled temperature, or a large convection coefficient used to "hold" the temperature at some location.

 

Another problem that can occur is when heat enters a face of an element and flows out a face that is perpendicular to the inlet face. (Normally, some heat passes out the face that is opposite from the inlet face.) The discussion around Figure 2 on the page "Thermal Results Menu" in the documentation talks about the heat rate issue. This heat issue can lead to strange temperature results like you are seeing. If this is the culprit, either

  • ignore those elements,
  • adjust the heat loads so that the heat is not trying to pass through perpendicular faces. For example, instead of extending the heating load over the entire area, split the surface so that you can apply it to the inner 90% of the surface, 
  • try a finer mesh in the area. This may lead to even higher temperatures which indicates that you have a mathematical "singularity".
  • Is there some way to "help" the heat travel in a smooth path? If the problem is like Figure 2 that I referenced, would a rounded corner (and finer mesh) help the heat to flow better?

 

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

16 years experience with Simulation Mechanical
Please use plain text.