nonlinearBeamColumn/dispBeamColumn

[geoStru]

Hi everyone, for the pushover analysis of a 3D RC frames what is the difference about nonlinearBeamColumn and dispBeamColumn?
When I plot the curves there are many diffence about the results. Why?

[selimgunay]

nonlinearBeamColumn is force-based. The other one is displacement based and considers linear curvature distribution along the member length. Therfore, you can model a single member using one nonlinearBeamColumn, however you need several dispBeamColumn elements to be able to capture the correct curvature distribution. Please see this link for a detailed explanation of the differences.

https://opensees.berkeley.edu/wiki/inde ... ed_Element

[mhscott]

This might help too: https://wp.me/pbejzW-A6

The example is a steel member, but the same ideas Selim mentioned hold true for RC.

[geoStru]

Is there much difference for the results if I use $np 3 instead $np 5 (np=numbers of point integration) for the nonlinearBeamColumn?
I have got a small 3D structure, with columns=4.00 meters and beams=6.00 meters.

When I changed the numbers of the story (2 story to 3 story) the analysis doesn't converge with $np=5 but it is converges with $np=3.

[mhscott]

The force-based elements can be sensitive to the number of integration points, especially for RC structures. You may have localized response somewhere in your model.

[geoStru]

If I would like to use $np=5 what could I do to converge the analysis? or it is no necessary

[mhscott]

If the results look reasonable to you, then it is not necessary. You can try np=4 as well.

[selimgunay]

You can compare your analysis results from np=3 and np=5 until the point of nonconvergence. That could give you some idea if np=3 is good enough