when does a finite element analysis become labeled "nonlinear"

[OUe]

I tend to think of a nonlinear finite element analysis as "nonlinear" if it is a time-history analysis or dynamic analysis. I am a structural plan reviewer for a west-coast jurisdiction. I am the only engineer. The plans I'm reviewing are of a 2nd order magnified moment analysis of columns and braces - aka 2nd order analysis with P-delta effects (the old way of doing 2nd order). Is this considered "nonlinear" I believe the engineer is basically doing a linear elastic analysis and then applying the magnification. I know it's splitting hairs, but it stuck out to me because of its repetitive use of terminology.

Thanks anyone who can respond kindly.

OUe

 

[IDS]

I'm not sure what you mean by "the old way of doing 2nd order". If the analysis adjusts the results to account for deflections then it is non-linear. If it doesn't, and a factor is applied to the output actions, then it is linear.

Static analyses can certainly be non-linear, both for geometric and material property effects, and dynamic analyses can be linear.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[whitwas]

I would call P-delta a poor man's nonlinear analysis. I think it technically is nonlinear because the displacements aren't proportional to the loads but it's perhaps not what people first think of when you say nonlinear.

Dynamic analysis doesn't have to be nonlinear. Doing a nonlinear dynamic analysis is much less common in my experience than a linear one.

 

[Mustaine3]

There are 3 classicals type of nonlinearities:
1. Nonlinear boundary conditions: Usually a change in the contact situation; Friction
2. Nonlinear material
3. Geometric nonlinearity: Large strains; large rotations; preloading effects; load stiffening effects; update of load directions; etc.

 

[wannabeSE]

From a building analysis perspective, time history and dynamic analysis are not necessarily nonlinear. Many engineers consider 2nd order analysis (P-little delta) as nonlinear because they account for the geometric nonlinearity. They are technically correct. However, nonlinear analysis is more often used describe material non-linearity. For instance ASCE 41 has provisions for nonlinear static and nonlinear dynamic procedures (NSP and NDP) and these account for material nonlinearity. A pushover analysis is a nonlinear static procedure.

The FEA guys in this forum may scoff at the nonlinear analysis used by building structural engineers. Commonly used building analysis programs with nonlinear capabilities use non-linear hinges or fibers that the user defines. The model still includes a bunch of non-linear sticks (frame elements) and shells. For building analysis techniques, the "Structural engineering other technical topics" forum would receive more responses.