Simulation issues in ST5. FEA with assemblies

[erikisolator]

Hi all,

Long time SE user here (since version 2). I just upgraded my seat of SE to ST5 premium with the intention of using the more powerful Simulation package to do more of our FEA work in house. I previously only ever had access to Simulation Express (FEMAP express before that) and was aware of the limitations, the biggest for us being the lack of ability to analyze assemblies.

So I understand that Simulation is sort of in between the full version of FEMAP and FEMAP (simulation) Express. It is not quite as simplified as Express, but much simpler to use than FEMAP. Is this a correct assumption? That is what I was lead to believe.

So I have had some issues with Simulation. I kept getting errors of various types while trying to analyze fairly simple parts. Gtac support has be quite helpful and courteous, however, I have some concerns:

I intend to use Simuation to analyze all up assemblies of several tens of parts. I am trying to run a static loading test on just a few plates that are bolted together with covers attached. (like a big box with thicker top and bottom) The plates have some holes that are needed to make the structure work. Simulation is meshing them with incredible density, such that it takes almost 45 min to just mesh. It has tried to solve for over 3 hours and no results. Turning down the subjective mesh size slider has no effect. Is the only option to simplify the part to the point of having no holes? Are there settings in there that would help this? This seems counterproductive.

I also need to be able to analyze flexures that have thick and thin sections. Comparing a uniform fixed/free cantilever with hand calcs and Simulation checks out nicely. However, comparing a cantilever beam with a thin section near the fixed end produced differences of up to 5 times. This is troubling. Increasing the mesh density to something like 8 or 9 helps get the value closer, but it never quite matches. Is there some trick to doing this? Is Simulation not up to the task? There have even been instances where inceasing the thickness of the thin section has made the deflection under load go UP. This is backwards. Not too confidence inspiring.

Is anyone here actually using Simulation on assemblies (large or small) that contain real-world levels of detail and getting good results? Is there a better forum for me to be asking this within?

Thanks in advance,

Erik

 

[BlasMolero]

Dear Erik,
ST5 SIMULATION is a FEA package with a long run of evolution, so keep going, new version of ST6 SIMULATION is really, really powerful, including optimization!!.
Regarding the problem you have meshing assemblies, the trick is to prescribe mesh density in solids, not to use the default mesh size. See the dimensions of you parts, and use BODY SIZE to prescribe a reasonable mesh size to your part.
Please note in case you have THIN & THICK parts in you assembly then you must use SHELL elements to mesh thin parts and TET elements to mesh solid thick components, SIMULATION allows you to mix Shell & Solids tetrahedral eements in the same FE model, the trick is to create a study of type "Mixed & General Bodies". Please note that if you plan to mesh a thin walled part you need to use minimum TWO ELEMENTS in the tkickness to cature the stress gradient .... you will realize that will arrive to millions of nodes & elements!!. Then no practical at all, the SHELL element is by far the most used FE element between us professional engineers & analist.

The most important with SIMULATION is to receive a training class from an expert, tou will realize the powerful & easy to use that could be the tool for the SOLID EDGE ST designer, OK?.

Best regards,
Blas.

~~~~~~~~~~~~~~~~~~~~~~
Blas Molero Hidalgo
Ingeniero Industrial
Director

IBERISA
48011 BILBAO (SPAIN)
WEB:

http://www.iberisa.com

Blog de FEMAP & NX Nastran:

http://iberisa.wordpress.com/