Thermal Stress

[NL8]

Outside of COSMOS can anybody recommend a reasonable priced program that can do thermal stress analysis on SW assemblies and parts.

Joe Hasik, CSWP/SMTL
SW 08 x64, SP 4.0
SW 09 x64, BETA 1
Dell T3400
Intel Core2 Quad
Q6700 2.66 GHz
3.93 GB RAM
NVIDIA Quadro FX 4600

 

[ctopher]

I don't know of any that are reasonably priced. What is a reasonable price range for you?
I have not used anything else with SolidWorks other than COSMOS.

http://www.solidworks.com/pages/products/analysis.html

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[GBor]

What do you consider "reasonably priced"?

SolidWorks will output so many different file formats that can be read back in by a variety of software packages. Are you looking for the more automated integration from something like NEiWorks or Algor? Or are you willing to export as an IGES file and read it in to an outside solver?

 

[NL8]

I'm not sure what management would consider reasonable yet, I guess I'll settle for something costed around a seat of COSMOS professional or less, and I would like to avoid exporting to one of the generic file formats if at all possible, we haven't had the best of luck regarding that.

Joe Hasik, CSWP/SMTL
SW 08 x64, SP 4.0
SW 09 x64, BETA 1
Dell T3400
Intel Core2 Quad
Q6700 2.66 GHz
3.93 GB RAM
NVIDIA Quadro FX 4600

 

[GBor]

Really? I have a couple of software packages that integrate with SolidWorks, but I generally don't use that functionality. I prefer to save my computer the operating overhead, export the file in IGES or STEP, shut down SW, open one of the FEA packages and import. Sometimes IGES works better than STEP and others are the opposite. Some packages prefer trimmed surface IGES files (iges entity 143), other prefer B-REP solids (144)...anyway. It may be cost efficient to find a low-priced package and learn the best settings for importing...especially if you are just doing thermal analysis.

In the category of COSMOS, I generally throw Algor (integrated or variety of file import formats), NENastran (integrated or variety of import file formats), and AMPS (uses the .sat file format - ACIS kernal exportable from SW). Any one of these should work for thermal stress, but COSMOS should do the trick, too, so why are you turning away from COSMOS?

 

[ctopher]

It would probably be cheaper to have the SolidWorks package including COSMOS than purchasing a separate FEA app. You could also use the SolidWorks materials.

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[NL8]

GBor - I'm not necessarily turning away from COSMOS, I knew it could, so I was turning to the community to get a few suggestions for other packages, allowing me to do some competitive shopping.

Chris - Got my quote from my VAR about ten minutes ago. Including the subscription fee it would cost me about $8000 to tack COSMOS Professional on to my current SW Premium license.

Joe Hasik, CSWP/SMTL
SW 08 x64, SP 4.0
SW 09 x64, BETA 1
Dell T3400
Intel Core2 Quad
Q6700 2.66 GHz
3.93 GB RAM
NVIDIA Quadro FX 4600

 

[ctopher]

Remove the subscription fee!

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[GBor]

"Including the subscription fee", that's probably not a bad deal. With any other package, you will have to have linear static stress analysis capability and a thermal processor unless you are just applying known steady state temperatures to certain parts and trying to see how they stress under expansion. Some linear stress processors will allow you to apply a known temperature as an initial condition.

For FEA embedded in SW, I say stick with the COSMOS path...and I'm not a big COSMOS fan (due primarily to lack of experience with it and my disdain for COSMOS Express that I think is dangerous).

 

[NL8]

I've done my time with both Express and regular COSMOS, Express is like handing a shot gun to a twelve year old, actual COSMOS at leasat allows for more real would condition to be simulated...even if it does treat the Steel stress/strain curve as being perfectly linear.

Chris,

I'd love to remove my subscription fees, but as a company and department we've gotten a little too well known for trying to stay on the cutting edge of technology. Besides, take away the fees and I won't have any expenses to approve anymore.

Joe Hasik, CSWP/SMTL
SW 08 x64, SP 4.0
SW 09 x64, BETA 1
Dell T3400
Intel Core2 Quad
Q6700 2.66 GHz
3.93 GB RAM
NVIDIA Quadro FX 4600

 

[GBor]

Love the shotgun analogy...

 

[kellnerp]

NEiWorks might deserve a look. It is Nastran based, but has a user interface with SW that is very similar to Cosmos. It has a couple strong points.

1. It solves outside of SW so on a long problem you can continue with CAD.
2. It creates a text file with a record of the whole analysis so you have an unchangeable document of what you did. In other words you are better insulated from changing things in SW after the analysis is over.
3. It's FEA model whether thermal or stress will directly translate into it's big brother Pre/Post processor FEMAP and into NENastran should you grow out of the integrated system.

So you have the advantages of quick model setup in SW with some additional advantages to boot.

There is also a Nastran forum on Eng-Tips with some real experts from time to time.



TOP

http://www.engtran.com