Hello, all.
I want to know how the section curvature of displacement-based & force-based beam-column element come from?
Have any reference describe that?
About section curvature...
Moderators: silvia, selimgunay, Moderators
Re: About section curvature...
they come from the element, which determines them from the nodal displacements. for the displacement base element, the displacements are transformed to the elements basic system (simply supported beam with a rollar at one end) resulting in two end rotations and an axial displacement. The curvature for the sections are determined assuming a linear curvature over the element length given the two end rotations.
the force beam is more complicated as it tries to find the section deformations at the gausspoints that given section forces computed for trial section deformations and node displacements will result in equilibrium being satisfied over the element length. See the papers on the forceBeamColumn element that are givn in the command page.
the force beam is more complicated as it tries to find the section deformations at the gausspoints that given section forces computed for trial section deformations and node displacements will result in equilibrium being satisfied over the element length. See the papers on the forceBeamColumn element that are givn in the command page.
Re: About section curvature...
Dear frank, thanks for your reply. I have another question about displacement-based & force-based beam-column element.
1. Why OpenSees provide that two kind of elements for building model? Are they have advantage or disadvantage for building model?
2. If i want to get the moment & curvature distribution along the pile, which kind of element is better for me to build the model?
1. Why OpenSees provide that two kind of elements for building model? Are they have advantage or disadvantage for building model?
2. If i want to get the moment & curvature distribution along the pile, which kind of element is better for me to build the model?
Re: About section curvature...
1. the force is theoretically better than the element, basically the assumption of linear curvature along the element is not correct for nonlinear elements. As a consequence more displacement based elements are required to give theoretically exact response. the force element can suffer from convergence issues as it needs to iterate to achieve convergence. you will find if you subdivide a displacement beam element you will have similar convergence issues, but you can do something about the convergence at the global level in the script. the force element (by default has the gauss points at the end, wheras the displacement based has them in a bit - gauss lobotto versus gauss legendre - though you can change the defaults). for softening systems all bets are off - the beamWitHinges should probably be used as the hinge length comes into play in the response.
2. along a pile - as you are breaking the pile into many elements anyway use the displacement based element.
2. along a pile - as you are breaking the pile into many elements anyway use the displacement based element.