finite element question

[joelrob]

Hi all,
I am reading a finite element program listing code (plane stress or strain analysis of solids), and I find a section where the programmer is talking about the 'elimintation of auxiliary degrees of freedom' when implementing the elementary matrices. Can anyone tell what this step
is for ? what do we mean by auxiliary dof's and why their elimination ? any references (books, articles) explaining that. This troubles me since this code gives results that are different form another code I wrote as well as two other cpmmercial FE codes in which I am more confident.

Thanks in advance.

 

[Qshake]

Perhaps they are referring to the off-diagonal terms which are often times difficult to define. I don't really know what the application is for.

 

[Ron]

Fairly often in FEA printouts, you will see the non-critical axes with very small numbers taken out to a bezillion significant figures and in some cases in the critical planes or axes where values would normally be zero (moment at end of cantilever, for example) but end up as a very small number due to matrix manipulation, rounding, and other small errors. This could be the reference.

 

[joelrob]

Yes I agree with you, but what I don't understand is why 2 other commercial FE packages (SAP and ADINA) don't consider this step in the analysis.

 

[Ron]

Likely because it doesn't make any fundamental difference.