HyperSizer Mechanical Limit Factor of Safety for Composites

[HyperSizer User]

What do you guys recommend using for the mechanical limit factor of safety for composite materials? Since composite failures are driven by ultimate load failures, are the limit load failures applicable? Should I put a factor of safety of 1.4 for both the limit and ultimate load limits? I am getting some limit failures for local buckling on some stringer-stiffened designs.

[Phil]

We recommend not using a 1.4 factor on limit loads.    Perhaps limit loads could be a factor of 1.1 (we have seen that), but it should not be the same as ultimate.

The HyperSizer default is to leave composite failure turned off for limit loads.  I can’t think of a reason to turn them on for limit loads.  Did you happen to get them turned on by using our “Analysis Levels”?  We recently noticed that the analysis levels are erroneously turning on limit based composite analysis and that’s something we need to fix. 

If you do turn them on for limit and ultimate, it just means HyperSizer will calculate the composite margin both using ultimate loads and at limit loads, and since it uses the same allowable for both, the ultimate should always control.

I think it is appropriate to use limit loads for local buckling. Here is an exchange I had with another user:

------------
How can I access your documentation on local buckling.
 
Specifically, interested in the reasoning that it is a limit rather than ultimate failure.

------------

If you generate a stress report with sample calculations, there is a section of that report that describes our method… however, I don’t think it describes why we treat it as a limit failure method.  I think that is more a traditional aerospace thing.  One way to look at it is that local buckling is not necessarily a catastrophic (or “ultimate”) failure, therefore it is allowed at a lower level than panel buckling, which is a catastrophic failure.

Keep in mind that we also have the crippling failure mode.  Crippling is basically like looking at the culmination of all local buckling.  For example, the web could local buckle, and additional load would be concentrated in the intersection between web and flange.  Crippling is basically what happens whne the entire cross-section cripples.  We treat that as an ultimate failure.

Also, remember that treating local buckling as a limit method is HyperSizer’s default (and our recommendation), but if the program decides that local buckling should treated as an ultimate failure, it is very easy to turn off it off for limit loads and turn it on for ultimate.  Then, you can apply this change to all components in a group, assembly, entire project, or set it as a database default.

[kshea]

I'm curious why HyperSizer returns an ultimate crippling margin far lower than my minimum ultimate local buckling margin.