Finite deformation frame with plasticity

[watkinrt]

Some of the work I do involves modeling elasto-plastic beams. According to the v7.5 manual (the version of FEAP that I am using), there is the option for elasto-plastic, finite deformation, frame elements (see Table Table 6.3:); however, the plasticity portion does not appear to be functioning correctly. Attached are two basic files exhibiting the issues that I am having. Here is an outline of the two files:

In both files, an initially flat 2mm diameter steel rod (E = 200e3 MPa) is examined. A total of 71 nodes and 70 elements make up the specimen which is 70 mm long. I've setup the material model for finite deformation elasto-plastic behavior with a von Mises yield stress of 400 MPa. The file Ih000001.txt is a pure bending setup for the beam (see the attached figure Ih000001_schematic.pdf where α is driven to an angle of 60º)  and the file Ih000002.txt is a simple tension of a flat beam. In both simulations, the beam is loaded well beyond the yielding, yet the mechanical response still exhibits perfectly elastic behavior throughout.

Can I use an elasto-plastic material model with frame elements? Is there something that I am missing? Thank you so much for your time!

[FEAP_Admin]

Does it work if you define your cross-section with say SECTion RECTangle?

[FEAP_Admin]

Sorry for the misinformation in the last message.

In 2D you will need to use the SECTion LAYEr specification for the plasticity.  The plasticity is integration through the thickness so it needs more information than just A and I.

Attached is a sample input -- base on your file but with a beam of length 100; fixed at the left and rotated at the right.  The yield moment is 1800.

NB: This was run with the current release of the program not 7.5 (but the 7.5 manual describes section layers so it should work).

[watkinrt]

That was what I was missing. Thanks so much for the help!

[watkinrt]

Page 90 of the FEAP users manual (for v7.5) discusses the use of the SECTion command for both 2D and 3D frames. In the case of 3D framer, different cross-sectional profiles can be specified, such as circle, rectangle, etc., while in the case of 2D frames, it appears that only a build up of rectangular layers can be used to specify the cross-section using the SECTion LAYEr command. However, this is inconvenient when dealing with profiles that are circular and takes away from the overall accuracy of the simulation. Is there a way to specify a circular cross-sectional geometry for 2D frames as can be specified for 3D frames? If not, is there a particular reason why it is setup this way?

Thanks for your help!   

[FEAP_Admin]

The feature was just never programed.  If you have a look at the code in the elements/frame folder you should be able to 'copy' the 3D code over to the 2D case.