Tests of Zipper Frames

1
Tests of Zipper Frames
Chuang-Sheng Yang1,Macarena Schachter2, Tony
Yang3, Andreas Stavridis4, Roberto T. Leon1
Georgia Institute of Technology1 University
at Buffalo2 University of California
Berkeley3 University of Colorado at
Boulder/UCSD4 Florida AM
2
Past Braced Frame Performance
Buckling of compression brace leading to collapse
3
Inverted-V-braced frames

• Localization of deformation cannot be avoided
• Beam designed for large axial force, shear, and
  moment

4
Solution 1 Full-height zipper mechanism

• Better distribution of energy dissipation.
• Instability and collapse can occur
• Khatib, Mahin and Pister, 1988.

5
Research Approach
Simultaneous testing Substructuring
Makola M. AbdullahFlorida State
P. Benson Shing U of Colorado
Andrei M. Reinhorn U_at_Buffalo
Quasi-Static R. Leon and R. DesRoches Georgia
Tech
Shake Table
B.Stojadinovic and J.Moehle UC-
Berkeley Substructure/ Hybrid
6
Solution 2 Partial-height zipper mechanism

• Top-story bracing members are designed to remain
  elastic.
• Beams can be designed to be flexible.

7
Model frame (SAC 3-Story, LA)
Prototype
1/3-Scale Model (shake table)
8
Experimental Work

• Three quasi-static tests at Georgia Tech
• Pushover
• LA22 (Kobe 1995, near field)
• Llolleo (Chile) 1985 (far field)
• Three shake table tests at Buffalo
• LA22 (Kobe 1995, near field)
• Hybrid simulations at UCB and CUB
• LA22 (Kobe 1995, near field)

9
Quasi-Static Tests Conducted at Georgia
Institute of Technology
10
Pushover Experiment
11
Pushover curve
(Tri-linear curve)
12
New Configuration for Gusset Plates
Designed to new AISC recommendation of 2t
13
Cyclic Test LA22 (75)
14
Comparisons for 50 of LA22

• First story left brace

15
Overall Load-Deflection Behavior (75)
16
Comparisons for 75 Case

• Axial force in the 2nd-story zipper strut

17
Cyclic Test LA22 (100)

• Video of 100 of LA22 (1st-story Back view)

18
Overall Load-Deflection Behavior (100)
19
Photograph of 100 of LA22 (2nd-story Back View)
20
Comparisons for 100 Case
Axial force in the 3rd-story zipper strut
21
Conclusions

• The reduced-scale zipper frame designed in
  compliance with the capacity design procedure
  exhibited great strength, energy dissipation and
  ductility, and remained stable up to the roof
  drift ratio of 3.6.
• The theoretical load path was validated in these
  tests. Once buckling had occurred in the braces,
  the zipper strut performed as a tension member,
  providing support at mid-span of the beam and
  transmitting the unbalanced vertical forces
  upwards to mobilize the unbuckled braces.
• Design guidelines are under development and will
  be available in Fall 2006

22
Shake Table Tests conducted atUniversity at
Buffalo
23
Zipper Frame test setup.
Test 1
Test 2
Test 3
24
Video.Test 3LA22yy80
25
Shaking table test main findings

• Test 1.
• Out of plane buckling of braces.
• Torsion transmitted to beam yielded in
  torsion.
• No readings at zipper columns.
• Test 2.
• Restrained beam zipper performed as
  expected.
• Test 3.
• Unrestrained beam and new brace to beam gusset
  plate zipper performed as expected and no
  torsion observed.
• All tests
• 3D movement of braces was recorded by Krypton
  tracking system.

26
Test 1
27
Test 1.
Braces at the end of 50 test.
Base of columns at the end of 50 test.
28
Test 2 Krypton Data
Out of plane displacements
Height (in.)
Height (in.)
East brace (in.)
West brace (in.)
Axial load (kips)
Axial load (kips)
Axial Deformation (in.)
Axial Deformation (in.)
29
Test 3.

• Unrestrained beam.
• Different gusset plate.
• Test Sequence 15, 40, 50, 60, 70, 80.
• No rotation of beam observed.
• Fracture of west brace at 80.

30
Conclusions (2)

• The frame behaved as expected. Buckling/ yielding
  at the second floor followed buckling/ yielding
  at the first floor.
• For this mechanism to happen, the beam has to
  remain straight.
• A flexible brace to beam gusset plate is
  preferable when braces buckle out of plane.
• The behavior is highly nonlinear and 3D
  analytical model must consider out of plane
  buckling, correct torsional properties of the
  beam, correct modeling of the gusset plate
  behavior.

31
Hybrid Simulation Tests conducted at UC Berkeley
32
Experimental set up at UC Berkeley
33
Hysteretic Behavior of the Braces (50 of LA22)
34
Time Histories of the Zipper Struts
35
Time History of Roof Drift Ratio
36
Hybrid Simulation Tests conducted at University
of Colorado at Boulder
37
Experimental Substructure in CU Lab

• 2 degrees of freedom are controlled with 3
  actuators

38
Test Summary

• Specimen 1
• 7.5 , 15 , model calibration
• 40
• Specimen 2
• Distributed Tests (UCB-CU)
• 15 , 40 , 50
• Specimen 3
• 15 , 40 , 50
• Specimen 4
• Refined OpenSees model
• 15 , 40 , 100

39
Failure Mechanisms

• Specimen 2

• Specimen 3

• Specimen 4

40
Specimen 3 highlights

• Discrepancy due to
• Inadequate modeling of the gusset plate influence
• OpenSees model
• assumes pinned connection
• is calibrated with a different sequence of GM
• Discrepancy reduces for 50

41
Specimen 4 15 and 40 Tests

• Slight difference in analytical and experimental
  time histories
• Experimental buckling loads differ for two first
  story braces due to a cycle in tension of Brace 1
  (consistent with other sites)
• Consistent buckling strength in analysis between
  two braces

42
Specimen 4 100 Test

• Brace fractured
• in tension
• in a brittle manner (maximum recorded tensile
  strain between 0.004 and 0.008)
• close to the weld of slotted end
• due to initial stress concentration and
• Large rotation required at the brace end
• Displacement dfr2.8 in
• Inter-story drift ratio afr5.3

43
Conclusions

• The results of the quasi-static tests performed
  at Georgia Tech validated the design procedure
  for zipper frames, and showed the zipper frame
  exhibiting great strength, energy dissipation and
  ductility.
• The shaking table tests carried out at y Buffalo
  investigated the seismic response of the zipper
  frame. The observations demonstrated the
  partial-height zipper mechanism.

44
Conclusions

• The local and global hybrid simulation tests
  conducted between the UC berkeley and University
  of Colorado at Boulder not only provided the
  precise brace model, but also verified the
  overall behavior of the zipper frame.
• Design guidelines are under development and will
  be available in Fall 2006

45
The main researcher
46
Thank you for your attention. Questions?