News: HyperSizer Pro V7.3.24 just Released!  Contact us to upgrade your software!

Author Topic: Different results with new version  (Read 1704 times)

cfriedland5

  • Client
  • **
  • Posts: 17
    •  
Different results with new version
« on: February 14, 2014, 09:29:44 AM »
I've recently switched from using version 6 to version 7, and I'm getting some different results for a wingbox model that I import. In particular, the upper skin seems to have increased in weight by about 5%. Other groups have also changed, but not nearly as much and it is probably related to the increased load traveling through the upper skin. So far I've checked that the same failure analyses are being applied (the upper skin is mostly failing from crippling in both versions) and that the panels have the same settings (zero-out FEA moments, curvature, etc.). Do you know what might have caused this difference? Are there any default settings that were changed that I could have overlooked?

James

  • Administrator
  • *****
  • Posts: 230
    •  
Re: Different results with new version
« Reply #1 on: February 27, 2014, 08:28:11 AM »
I would check your backdoor data settings. See help topic:
http://hypersizer.com/help_7.0/#BackdoorData/bd-00-jump.php%3FTocPath%3DSoftware%20Forms%7CBackdoor%20Data%7C_____0

Do you remember what specific HyperSizer V6 you were using? 6.2, 6.3 or 6.4?

There was a change in v6.3 that was also applied to v6.4 and v7.0 that could be causing this increase in mass. The FEA stastical loading method was updated to treat Nxy, Mxy, Qx and Qy differently by using the max +/- Nxy rather than pairing +Nxy with +Nx.  There is a backdoor setting that will allow you to turn this off and analyze project the old way (however using the old way of treating Nxy is too under-conservative).

Backdoor flag:
Use Max Positive-Negative Values of Shear Load Components (default = False)
By setting value = True, it will revert to the old method of processing Nxy.

There is another backdoor setting you may want to check. This flag will enforce that the 2-sigma loads don't exceed the peak element loads.

Backdoor flag:
FEA Statistical Limit Component Load to Element Max_Min (default = true)

I would keep this flag = true, otherwise you could be too conservative.

-James