Skip to main content

Finite Element Method

Submitted by Zhigang Suo on

The notes are attached. Return to the outline of the course.

Attachment Size
FEM 2008 10 26.pdf 86.56 KB

In the FE formulation of beam element, generally, nodal displacement (v) and slope (θ) are taken as primary unknowns. What will be the obstacles if the nutral axis curvature (κ) is used instead of slope as primary unknown, along with nodal displacement?

Tue, 01/29/2008 - 09:14 Permalink

In developing weak forms for beams, the second derivation of weight function appears. So it is concluded that we need a C1 continuity element for beams. In order to use the C1 continuity between elements, the displacements and the derivatives of the displacement at the nodes must be degrees of freedom. So we should use slope (instead of curvature, which is second derivation of displacement). Also the loads imposed to the structure are considered force (that products displacement) and moment (that products rotation-slop) but we don't have such physical vision about what products curvature. So if we use curvature instead of rotation, we will face the problem of choosing load for that degree of freedom:d=[u1 θ1 u2 θ1]T : f=[V1 m1 V2 m2]Td=[u1 κ1 u2 κ1]T : f=[V1 ?1 V2 ?2]   

Fri, 03/02/2012 - 19:03 Permalink