Thermal Expansion Analysis

[Slagathor]

Are there any software packages out there that do a decent job of predicting deformation or complex shapes due to thermal effects?

The parts I want to analyze will be castings, so material is uniform. I am looking to minimize the angular distortion of mounting points, so as to maintain overall mechanical accuracy when the part is subjected to a moderate, uniform temperature rise.

Better still, would be the ability to assign a temperature gradient, and determine how the shape changes.

This must be similar to what designers of machine tools do...but I can not find a software package that describes this capability. Perhaps I am not using the right search terms, or talking to the right people.

 

[corus]

You'd need FE software to do that unless the shape was simple enough to perform hand calculations.

Tara

http://tinyurl.com/4ydjg7m

 

[Slagathor]

Yes of course...I know that...but what FEA. That is what I am looking for. I have not been able to find one that will predict deformation as a result of an applied thermal profile.

 

[Slagathor]

I was reviewing NEI Nastran NEIWorks 2.1, and I see they introduced a thermal stress/deflection feature in their most recent release.

 

[electricpete]

I’m going to show my ignorance.

Uniform material at uniform temperature does not expand uniformly?


=====================================
(2B)+(2B)' ?

 

[electricpete]

.. or are we talking about a temperuatre gradient within the part?

=====================================
(2B)+(2B)' ?

 

[Slagathor]

Complex closed loop parts, in my experience, do not hold their shape perfectly, even under uniform temperature rise. Think about a complex casing, with ribs, shells, etc.

Of course, in the real world, the temperature is never uniform, so a proper analysis would indeed assume a gradient, or custom temperature mapping....

 

[IDS]

Surely any general purpose FE package would be able to do this? Certainly Strand7 can.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[corus]

With ribs etc. that would locally be cooler, you'd need a program that calculated the temperature profile and then calculated the deflection. Most FE packages can do this.

Tara

http://tinyurl.com/4ydjg7m

 

[electricpete]

Complex closed loop parts, in my experience, do not hold their shape perfectly, even under uniform temperature rise. Think about a complex casing, with ribs, shells, etc.

Is this a result of residual stresses from forming the part?



=====================================
(2B)+(2B)' ?

 

[jhardy1]

@electricpete,

Yes, theoretically, an unconstrained component which is made of a single homogeneous isotropic material and subject to a uniform temperature change should expand / contract uniformly. However, many (most? all?) "real world" scenarios break at least one of those assumptions:

Unconstrained? Probably not - the part is probably attached to something or is sitting on something (most things are!), which may locally constrain its thermal expansion / contraction.

Single homogeneous isotropic material - OK, many components would approximate to this condition

Uniform temperature rise or fall? Probably not - many heating / cooling scenarios will produce some sort of thermal gradient ( a hot side and a cold side), and / or the attachment to another component can act like a "heat sink" leading to local thermal gradients.

As soon as you have thermal gradients, you have a non-trivial condition, and thermal analysis (including possibly FEA) may be warranted.

Cheers!

http://julianh72.blogspot.com