Mechanica thermal problem

[frantzstanford]

I have an interesting thermal FEA problem.

Stated in a generic way: Is it possible in Mechanica to look at the
transient thermal response of two bodies of different materials that
are brought together with different initial temperatures. Heat
transfer from outside the system is negligible.

More specifically, I need to evaluate the temperature of an aluminum
part as it is encapsulated with nylon by an injection molding process.
The aluminum part is initially at room temperature and the nylon is
injected around the aluminum at roughly 480 degrees F.

The heat flow between the nylon and aluminum will constantly vary as
the temperature difference between the two parts changes. This is
preventing me from simplifying the problem to just an analysis of the
aluminum part with a constant heat flow into it.

This seems to be an elementary problem, but running through the menu
system I couldn't figure it out and ptc tech support wasn't able to
help me either...

Any insights would be appreciated. Also, if someone has done this
problem in a different package, I would be interested in hearing about
that.

Frantz Stanford

 

[GBor]

I haven't done this precise problem in any package, but I can't imagine it would be too difficult. It would require some set-up to get the temperature of the nylon to "travel" around, but I can think of a few ways to get some results:

1) Model the aluminum and the encapsulating nylon. Give the nylon an initial temperature of 480 and the aluminum a temp of approximately 70 F. Run the analysis.

2) Model the aluminum and the encapsulating nylon. Tie the nylon temperature to a curve that ramps from room temp to 480.

3) Model the aluminum and the encapsulating nylon. Provide a heat flux on one side of the nylon into a few of the elements so that it slowly wraps through the nylon.

Regardless of how you choose to do this, I'm not sure I would ignore the heat transfer from the nylon to the surrounding environment.

Garland

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

 

[cbrn]

Hi,
either I'm missing something, or under the hypothesis of external adiabaticity ("heat exchange with the environment is negligible") the calculation of the final temperature of the Alu part can be done without any FEM, simply with the basic laws of heat exchange...
If you're interested in the transient temperature rate-of-change, it may be different (though, as far as you're interested in average temperatures in the whole Alu block, I believe there's nothing you can't do with simple analytics...).

Regards

 

[frantzstanford]

Thanks for the responses.

My approach matched what Garland suggested in step 1. However the problem I ran into is that (according to PTC) I cannot set the initial temperatures of these bodies in Mechanica and then just allow the heat to flow. I was hoping that someone here knew how to do this in Mechanica.

Cbrn - I need to determine the peak transient temperature in different areas of the aluminum body. The concern is that if temperatures get too high then stress relieving may occur that will negate a shot peening process.

Frantz

 

[GBor]

So if you ignore the heat exchange with the surrounding environment, you get a "worst case", but it will suggest temperatures higher than what you will see in reality. If you are barely out of the range for effective shot peening, you should be comfortable that you are probably able to shot peen considering the conservative nature of the analysis.

As for Mechanica...it really can't do this? I have two other software packages that you simple set the initial temperature for the part. I've used a least two others that are similar. I typically use h-element codes, not p-element like Mechanica, but I have a hard time believing something as simple as this would be so dramatically impacted by the convergence routines.

Eh...what do I know?

Good Luck!

 

[torpen]

Hello,

If you cannot set the initial temperatures of these bodies in Mechanica and then just allow the heat to flow (it seems to me very strange that it is not possible), you can apply (for example) a convective heat flux on the nylon with a very low heat transfer coefficient.
And perform your heat transfer analysis with this boundary condition.

Regards,

Torpen.

 

[frantzstanford]

Well in the end I created a finite difference matrix in a spreadsheet to examine the transient response of the simplified case of a flat aluminum plate with a nylon casing in .025 second time increments. It turns out that the aluminum distributes the heat so quickly (about 468 times faster than the nylon) that the aluminum maintains an even temperature within 1 degree C even within the first second after it comes into contact with heated nylon.

Thanks all for the suggestions. I also agree with the comments that it's hard to believe that Mechanica can't handle this, it's probably in there somewhere...

Frantz