Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

FEA (ANSYS) Pass/Fail Criteria

Status
Not open for further replies.

StrEng007

Structural
Aug 22, 2014
543
I'm new to programs like Solidworks and ANSYS and need some clarification on how other engineers are using them to determine a pass/fail criteria for their materials. I am a structural engineer and all of my designs require that I conform to certain design specifications/manuals as set forth by the governing building code. These manuals provide the ability to determine the available stress (ASD/LRFD) of a material based on its physical properties and geometric (cross-sectional) properties. In many situations, the allowable stresses for structural members are well below the threshold of the material’s yield stress as a result of its cross section (think light gauge slender columns).

I understand that FEA programs will allow you to determine the stresses in your materials based on an element mesh analysis. From that point, you compare these stresses (principal or von mises) to yield or ultimate stress values of the material with an applied factor of safety. Pending the ratio of these stress values, is this the pass/fail criteria you use to satisfy your design? What about your members that are subject to buckling or torsion that would typically fail well below yield? Do these programs have the ability to foresee these limit states?

I currently use structural analysis programs that have the ability to apply loads to structural members. I can determine a pass/fail criteria for each member so long as it's a fairly common cross section that is covered by the scope of that particular design standard. Think of it as though the member was "catalogued" and the design aid within the program is performing those limit state checks for you. My issue here is that I cannot create custom members and have the program perform the same task. Instead, I have to compare either my required force reactions or stresses to allowable member reactions. Does Solidworks or ANSYS offer a faster way around this?
 
Replies continue below

Recommended for you

I would say that making a full 3D solid model of anything to perform a FEA analysis in a high powered FEA program is never faster than a more general, less 'powerful' analysis program. The big FEA programs such as ANSYS can do buckling, non-linear, dynamic, etc analyses if you have the time to compute them. If you have something so irregular that this is required, it involves a lot of input and debugging to get you an answer - and these programs tend to have a steep learning curve.

There are many more basic FEA programs that can handle this - maybe even the one you are currently using - with some additional add ons that can handle a buckling analysis of a unique shape, at least globally. Local buckling analysis is generally best left for the more powerful programs, but you have to know how to find them in some cases, which may require a 'perturbing' force applied to initial buckling.
 
My suggestion to you would be to get the compressive 'loads' in the members from your FEM models (no need for high mesh density or fidelity), then use classical hand calculations from BRUHN or Niu (books used in aerospace) for your buckling and crippling margin of safety.

Stressing Stresslessly!
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor