News: Contact us to upgrade your software!

Author Topic: Analyzing a known structure for Element Based Sizing vs Optimization  (Read 20830 times)

HyperSizer User

  • Client
  • **
  • Posts: 47
    •  
What is the quickest/best way to do a non-optimized stress analysis? I kind of used HS to do this before

I know that the "element based" option basically does this---the design has to be "frozen" to do this, correct?    Or, I can set the variable bounds/permutations set to 0--then I can use HS to do a stress analysis for each element (no optimizing) and output the results (element #/stress,etc).  Does this sound correct?

Phil

  • Administrator
  • *****
  • Posts: 218
    • HyperSizer Structural Sizing Software
    •  
Re: Analyzing a known structure for Element Based Sizing vs Optimization
« Reply #1 on: January 19, 2009, 03:21:14 PM »
You actually CAN use the element based option when doing optimization, just be aware that it will be much slower than using the statistical option.  Our process is usually to use the statistical option (e.g. 2 sigma) for sizing and then when we are happy with the optimized result, then freeze the design and turn on element based to do a final margin check.  If there are concentrations when cause negative margins at this point, they can be separated out into separate components for additional sizing (for example, adding padups in corners, etc.)

You can get element based margins, controlling failure modes, load cases, etc. by going to “Sizing Element Results” in the Graphics form.

For the stress reports, you probably don’t want results for all elements, rather you just want them for the 10 or 20 elements with the minimum margins of safety.  This will speed up the generation a lot.  Very large tables are really what slows the report generation down more than anything else.  I would set the stress report options for elements to this when generating the report: