- Posts: 14
- Joined: Mon Sep 23, 2013 9:47 pm
- Location: Indian Institute of Technology, Delhi
I'm working on non-linear analysis of SCBF structures. I am a bit confused regarding checking for plastic hinge formation and rotation at plastic hinge after reading some of the earlier posts. I have used nonlinearBeamColumn elements with 3 integration points with fiber section for modelling beams and columns. I learnt that this formulation is capable of simulating P-M interaction. My questions are
1. For checking formation and rotation of plastic hinges, (a) do I need to change to nonlinearBeamColumn/ BeamWithHinges elements with defined plastic hinge lengths, record curvatures at two integration points within the hinge and multiply with hinge length (am I correct the way I understood?) to compare with hand calculated yield- or ultimate- curvatures, or
(b) section curvatures obtained at single integration points (span ends or center) using basicDeformation commands in the original fiber section NIP formulation can be compared with hand-calculated section yield- or ultimate-curvatures. Yield-curvature calculated as eps(yield)/(d/2) (W-section) (correct?).
2. If 1(a) is the correct method, integration points for recording curvature will be 1 & 2 and 5 & 6 for HingeRadau object. Correct? Can you suggest any reference for recommended plastic length values (empirical formulas)?
Appreciating in advance for your time and efforts to be devoted!
Please pardon me if I'm asking very basic questions or missing any obvious point.
You really don't need to change your elements! I personally prefer to use nonlinearBeamColumn elements rather than BeamWithHinges for your case. Anyway, if you want to find the location of plastic hinges, there are different commands such as "basicDeformation", "ChordDeformation" ... that you can use them to capture "rotations" at your elements. However, as Dr. Vesna has told me, it's preferable to capture "curvatures" instead of "rotations" and then calculate the rotation according to the curvature you obtain (by multiplying it by the plastic hinge length). what I think you need to do is to use more integration points to better distinguish the location of plastic hinges. In other words, Since you will need to record the curvature at two integration points that approximately encompass the plastic hinge region at the end of your elements, it's a better idea to use 5 to 7 integration points, then choose the points that embrace the plastic hinge zones. Moreover, you will need to write a script in MATLAB (or even Excel!) to compare the results you obtain from opensees with limitations, and see weather the rotations are big enough to form a plastic hinge. Finally, to sum up:
1) Use the following recorders to capture the curvatures at the end of your elements:
recorder Element -time -file ele.out -ele $elenum section $intPoint deformation
2) Multiply the results by Lp (Plastic hinge length)
3) Compare the results with limitations to figure out weather the rotations are big enough to form a plastic hinge or not (using MATLAB)
It should be noted, since my MS thesis is defined on RC frames, I have no comment on the plastic length values or empirical formulas implemented for steel frames. In RC frames there are lots of formulas that calculate plastic hinge rotations, length, and suggest standards that show the length in which the plastic hinges might take place logically. I think it should be the same for steel frames. Please let me know if you found the formulas for steel frames.
Colorado State University
Civil & Environmental Engineering
Fort Collins, CO 80523, USA