Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Next generation seismic resistant structural elements using high-performance materials
(USC Thesis Other)
Next generation seismic resistant structural elements using high-performance materials
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
© 2017 By Seyyed Farshid Hosseini Khasheh Heiran
All rights reserved
Next Generation Seismic Resistant Structural Elements Using
High-Performance Materials
BY
Seyyed Farshid Hosseini Khasheh Heiran
DISSERTATION
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy in Civil Engineering
in the Viterbi School of Engineering at the
University of Southern California, December 2017
Los Angeles, CA
Doctoral Committee:
Assistant Professor Bora Gencturk, Chair
Professor Erik A. Johnson
Professor Steven Nutt
i
ABSTRACT
In the recent past, earthquakes have proven that reinforced concrete (RC) structures remain highly
vulnerable to lateral loads. The yielding of longitudinal reinforcement as the main source of energy
absorption, in conjunction with the cracking and spalling of the concrete, leads to severe damage
and permanent deformations, which jeopardizes the post-earthquake functionality of these
structures. A swift response and recovery in the aftermath of a major disaster strongly relies on the
serviceability of key infrastructure, including bridges and other vital service structures such as
hospitals and government buildings. Recently, the use of high-performance materials such as
Engineered Cementitious Composites (ECC) and Superelastic Alloys (SEA) have been considered
as alternative materials for conventional concrete and steel reinforcing bars (rebar), attracting
heightened attention to improve the seismic performance of RC structures. ECC refers to a special
class of high-performance fiber-reinforced cementitious composites (HPFRCC) that exhibit
superior tensile ductility, energy absorption, bond characteristics and shear resistance. SEAs are
innately capable of recovering large inelastic deformations upon stress removal.
The primary endeavor of this study is to improve the performance of bridge columns and beam-
column joints of special moment frames (SMF) using high-performance materials, particularly,
ECC and Cu-Al-Mn SEA bars. To this end, an innovative bridge column design is first introduced
and examined through an experimental framework. The column design comprises of a
prefabricated reinforced ECC (RECC) hollow section that is embedded in a RC foundation and
filled with conventional concrete. Additionally, the longitudinal reinforcement at the potential
plastic hinge region is totally or partially replaced using the recently developed Cu-Al-Mn SEA
bars. The proposed approach utilizes the deformability of the ECC in order to enhance damage
ii
tolerance of the bridge columns and large strains recovery capability of Cu-Al-Mn SEA to reduce
permanent deformations. Following the completion of experimental work, a finite element
approach is developed to numerically investigate the performance of the bridge columns designed
on the basis of this proposed approach. The developed finite element approach is verified using
the results ascertained from the tested columns and used to carry out a parametric study to obtain
an optimal design strategy for the proposed design concept. Furthermore, in an attempt to improve
the performance of corner and exterior beam-column joints in SMFs, conventional RC in 3D beam-
column subassemblies is substituted with reinforced ECC (RECC) that extends from the panel
zone area into the adjacent beams and columns in order to cover the potential plastic hinge regions.
The 3D RECC beam-column subassemblies are then subjected to complex loading scenarios,
including torsion, to investigate their performance due to more realistic seismic loads in 3D
structures.
The results from the experimental works indicated that ECC can significantly improve damage
tolerance of bridge columns and beam-column joints subjected to extensive seismic loads and shift
the damage mode from concrete spalling to fine distributed cracks. Incorporating ECC in panel
zone of corner and exterior beam-column subassemblies sufficiently tolerated complex loading
combinations even in absence of panel zone transverse reinforcement, due to superior shear
strength of ECC compared to conventional concrete. Additionally, replacing longitudinal
reinforcement with Cu-Al-Mn SEA bars in plastic hinge region of bridge columns, considerably
decreases the permanent deformations of RECC bridge columns compared to the conventional RC
ones. Furthermore, conducting the numerical and parametric studies revealed that: (i) mechanical
properties of ECC can be successfully simulated by introducing fibers as smeared reinforcement
in concrete; (ii) cyclic behavior of bridge columns incorporating ECC and Cu-Al-Mn SEA bars
iii
can be accurately captured using numerical models; (iii) hollow RECC sections with maximum
hollow ratio of 35% have comparable performance as solid RECC bridge column; (iv) ultimate
strain hardening capacity of ECC has limited effect on performance of bridge columns; (v) design
of the bridge columns can be optimized by implementing hollow sections and partial replacement
of the longitudinal reinforcement with Cu-Al-Mn SEA bars.
iv
ACKNOWLEDGEMENTS
I would like to first express my genuine gratitude to my academic advisor, Dr. Bora Gencturk, for
his extensive support, advise, and guidance throughout my PhD program. My graduate program
under supervision of Dr. Bora Gencturk has been not only an incredible intellectual experience but
also a priceless life session. Beside his keen scientific insight, encouragement for engaging in new
ideas, and demand for a high-quality work, he flourished creativity in me and taught the value of
being hard working, ambitious, wise planning, and spreading kindness.
A very special thanks goes to my qualification exam and final defense committee members
Professors Erik A. Johnson, Steven Nutt, Vincent W. Lee, and Qiming Wang for their interest and
significant contribution in my work. Their insights and guidance have been astonishing
enlightenments throughout this work.
I would like to extent my thanks to all the people who contributed in the work presented in this
thesis and helped to accomplish it. With special mention to Sirvan Lahour Pour, David Ibague Gil,
Gustavo Cadaval, Hadi Aryan, and amazing Gerald McTigret for their unfailing support and
assistance. Additionally, the support of the Civil and Environmental Engineering Department of
the Cullen College of Engineering at the University of Houston, where majority of the
experimental work was conducted, is greatly appreciated. My time at the University of Houston
left great memories in me through my interactions with incredible students, staff, and faculty.
I would like to acknowledge the support of the United States National Science Foundation for
funding this research under the awards 1723393 and 1642488, and providing an opportunity to
complete my graduate studies. Additionally, thanks to Dr. Yoshikazu Araki from Kyoto
v
University, Japan, for providing Cu-Al-Mn SEA bars, US Silica company for providing silica sand,
and Boral Material Technologies (BMT) for providing fly ash used in this research.
Last but not least, my sincere gratitude to my parents, Hatam and Manizheh, for their unyielding
support and unconditional love. Your strength during this long separation is admirable and your
love is the panacea for any struggle and hardship. My deepest love and gratitude to my siblings,
Saleh, Majid, and Saeid, my lovely grandparents, Hasan and Golsim, alongside other family
members and friends for their inspiration, encouragement, and support.
vi
To my beloved family, for their patience, support, encouragement, and
unwavering love…
vii
TABLE OF CONTENTS
LIST OF FIGURES ...................................................................................................................... xii
LIST OF TABLES ........................................................................................................................ xx
LIST OF ABBREVIATION ....................................................................................................... xxii
OBJECTIVE AND SCOPE .................................................................................. 1
BACKGROUND AND RESEARCH SIGNIFICANCE ...................................... 6
2.1 Engineered Cementitious Composites (ECC) ..................................................................... 6
2.2 Cu-Al-Mn Superelastic Alloy (SEA) ............................................................................... 13
2.3 Application of ECC and SEA in Bridge Columns ........................................................... 17
2.4 Application of Fiber-Reinforced Concrete (FRC) in Beam-Column Joints .................... 20
2.5 Research Significance ...................................................................................................... 22
EXPERIMENTAL PROGRAM- BRIDGE COLUMNS .................................... 25
3.1 Innovative Column Design Concept ................................................................................ 25
3.2 Geometry and Test Matrix ............................................................................................... 27
3.3 Construction Process ...................................................................................................... 30
3.3.1 Formwork ................................................................................................................... 30
3.3.2 Rebar cages ................................................................................................................ 31
3.3.3 RECC hollow section ................................................................................................. 33
viii
3.3.4 Casting concrete ......................................................................................................... 34
3.3.5 Preparing the specimens ............................................................................................ 35
3.4 Test Setup and Instrumentation ........................................................................................ 36
3.5 Loading Protocol ............................................................................................................. 38
3.6 Material Properties ........................................................................................................... 40
3.6.1 Concrete ..................................................................................................................... 40
3.6.2 ECC ............................................................................................................................ 43
3.6.3 Reinforcing steel ........................................................................................................ 50
3.6.4 Cu-Al-Mn SEA bars .................................................................................................. 53
3.7 Results and Discussion .................................................................................................... 59
3.7.1 Crack distribution...................................................................................................... 59
3.7.2 Load-drift hysteresis curves ...................................................................................... 61
3.7.3 Quantitative comparison ........................................................................................... 63
NUMERICAL STUDY OF TESTED BRIDGE COLUMNS ............................ 69
4.1. Material Constitutive Models ......................................................................................... 70
4.1.1 Concrete .................................................................................................................. 70
4.1.2 Reinforcing steel ..................................................................................................... 74
4.1.3 ECC ......................................................................................................................... 75
ix
4.1.4 Cu-Al-Mn SEA bars ............................................................................................... 82
4.1.5 Bond characteristics ................................................................................................ 85
4.2. Geometry of the Numerical Models ............................................................................... 87
4.3. Boundary Conditions...................................................................................................... 91
4.4. Loading........................................................................................................................... 92
4.5. Meshing and Solution Parameters .................................................................................. 93
4.6. Results and Discussion ................................................................................................... 95
NUMERICAL PARAMETRIC STUDY OF BRIDGE COLUMNS ............... 100
5.1. Geometry of the Columns ............................................................................................ 100
5.2. Parametric Study Matrix .............................................................................................. 103
5.3. Material Constitutive Models ....................................................................................... 108
5.3.1 Concrete ................................................................................................................ 108
5.3.2 Reinforcing steel ................................................................................................... 109
5.3.3 ECC ....................................................................................................................... 110
5.3.4 Cu-Al-Mn SEA ..................................................................................................... 113
5.3.5 Bond characteristics .............................................................................................. 113
5.4. Meshing, Boundary Conditions and Loading Protocol ................................................ 114
5.5. Results and Discussion ................................................................................................. 117
x
5.5.1 Validation of the control case ............................................................................... 117
5.5.2 Effect of ECC quality............................................................................................ 118
5.5.3 Load-deformation diagrams .................................................................................. 121
5.5.4 Quantitative evaluation ......................................................................................... 124
5.6. Optimal Design ............................................................................................................ 141
EXPERIMENTAL PROGRAM OF BEAM-COLUMN JOINTS .................... 143
6.1. Beam-Column Joints Design Concept ......................................................................... 143
6.2. Geometry and Test Matrix ........................................................................................... 144
6.3. Testing the Specimens .................................................................................................. 148
6.3.1 Test setup and instrumentation ............................................................................. 148
6.3.2 Loading protocol ................................................................................................... 152
6.4. Concrete and ECC Material Properties ........................................................................ 154
6.5. Construction Process .................................................................................................... 157
6.5.1 Formwork .............................................................................................................. 157
6.5.2 Rebar cages ........................................................................................................... 158
6.5.3 Casting ECC sections ............................................................................................ 159
6.5.4 Steel end-plates ..................................................................................................... 161
6.5.5 Casting conventional concrete .............................................................................. 162
xi
6.5.6 Curing and preparing the specimens ..................................................................... 163
6.6. Experimental Results and Discussion .......................................................................... 164
6.6.1 Cyclic behavior ..................................................................................................... 164
6.6.2 Crack patterns ....................................................................................................... 168
6.6.3 Torsion .................................................................................................................. 171
6.6.4 Shear deformation ................................................................................................. 172
6.6.5 Reactions at boundaries ........................................................................................ 174
6.6.6 Deformations in beams and columns .................................................................... 180
CLOSURE ......................................................................................................... 201
7.1. Conclusions .................................................................................................................. 201
7.2. Recommendations for Future Research ....................................................................... 207
REFERENCES ........................................................................................................................... 209
xii
LIST OF FIGURES
Figure 2.1: Schematic tensile behavior of ECC. ............................................................................. 7
Figure 2.2: Typical σ-δ curve,, and for strain-hardening composites. ................................ 8
Figure 2.3: Accelerated corrosion test on: (a) ECC for 300 hours, (b) Mortar for 75 hours [17]. 12
Figure 2.4: Comparing stress-strain diagrams of NiTi [32] and Cu-Al-Mn SEAs [33] at different
temperatures (This pictures was reproduced using the data from the provided in the references).
....................................................................................................................................................... 14
Figure 3.1: Proposed design concept for bridge columns. ............................................................ 26
Figure 3.2:Geometry of bridge columns (in mm): (A) as built specimens in the test setup, (B)
RECC prefabricated hollow sections with conventional concrete core, and (C) detail of plastic
hinge rebar replacement with SEA. .............................................................................................. 28
Figure 3.3: Test matrix: section A-A and B-B are indicated in Figure 3.2................................... 29
Figure 3.4: Formwork to construct the bridge columns: (A) formwork of the foundation and the
top cap as well as the frame, (B) the base of the top cap and the fixtures to hold the PVC pipes in
place, (C) PVC pipes to create the whole through holes. ............................................................. 31
Figure 3.5: Rebar cages of the bridge columns: (A )rebar cage of RC-Column and HFA-Column,
(B) rebar cage of column of HFA-Tube and M45-Tube, (C) machined Cu-Al-Mn SEA bars, (D)
threaded steel reinforcement, (E) rebar cage of column of HFA-Tube-SEA with debonded Cu-Al-
Mn SEA bars in the potential plastic hinge region. ...................................................................... 33
Figure 3.6: Prefabricated RECC hollow section: (A) placing the rebar cage and inside form, (B)
placing the outside form, (C) pouring ECC between the two forms, (D) prefabricated RECC hollow
section after removing the forms. ................................................................................................. 34
Figure 3.7: Casting concrete: (A) placing RECC hollow section inside the foundation, (B)
preparing the formwork and reinforcement of the top cap, (C) placing and fixing the PVC tubes,
(D) casting the foundation, (E) casting the top cap and vibrating. ............................................... 35
Figure 3.8: Preparing the specimens: (A) Curing, (B) removing the formwork, (C) preparing the
upper surface of the top cap for applying self-leveler, (D) removing extra portion of the PVC pipes,
(E) cleaning and pained the specimen. ......................................................................................... 36
Figure 3.9: Test setup. ................................................................................................................... 38
Figure 3.10: Loading protocol. ..................................................................................................... 40
Figure 3.11: Test setup of the uniaxial compression test. ............................................................. 42
xiii
Figure 3.12: Some of the results from uniaxial compression test on conventional concrete: (A) RC-
Column, (B) core of HFA-Tube, (C) core of M45-Tube. ............................................................. 42
Figure 3.13: Results of uniaxial compression test on ECC cylinder specimens: (A) HFA-Column,
(B) HFA-Tube, (C) HFA-Tube-SEA and HFA-Tube-PSEA. ...................................................... 47
Figure 3.14: (A) Dimensions of the dog-bone specimens, (B) tensile test setup.......................... 49
Figure 3.15: Results of the direst tensile test on the dog-bone ECC specimens(different curves are
different tests on the same material). ............................................................................................ 49
Figure 3.16: Uniaxial direct tensile test setup for reinforcing steel. ............................................. 51
Figure 3.17: Stress-strain diagrams of 9.5 mm (#3) reinforcing steel. ......................................... 52
Figure 3.18: Stress-strain diagrams of 12.7 mm (#4) reinforcing steel. ....................................... 52
Figure 3.19: (A) Machined Cu-Al-Mn SEA bars, (B) test setup for training Cu-Al-Mn SEA bars.
....................................................................................................................................................... 54
Figure 3.20: Training of Cu-Al-Mn SEA bars. ............................................................................. 55
Figure 3.21: (A) Extracting Cu-Al-Mn SEA bars from the columns, (B) locations and numbering
of Cu-Al-Mn SEA bars, (C) extracted Cu-Al-Mn SEA bars (not all bars are shown), and (D) Cu-
Al-Mn SEA bars ruptured during testing. ..................................................................................... 56
Figure 3.22:Cyclic stress-strain behavior of Cu-Al-Mn SEA bars: (A) Bar 2, (B) Bar 6, (C), Bar 1,
and (D) Bar 3. Test were conducted some posteriori to column tests. ......................................... 58
Figure 3.23: Observed damage in test specimens at various drift levels. ..................................... 60
Figure 3.24: Load-drift hysteresis curves: (A) RC-Column, (B) HFA-Column, (C) HFA-Tube, (D)
M45-Tube, (E) HFA-Tube-SEA, and (F) HFA-Tube-PSEA. In (E) and (F), circles mark the Cu-
Al-Mn SEA bars ruptured during testing. ..................................................................................... 61
Figure 3.25: Definition of (A) initial stiffness, E0, yield drift, δy, yield strength, Vy, maximum drift,
δm, maximum strength, Vm, ultimate drift, δu, and ultimate strength, Vu, and (B) permanent drift,
δp, and energy absorption, Ea. In presentation of the permanent drift and energy absorption
definitions, 7% maximum experienced drift is used for comparison purposes. Permanent drift is
obtained as the average of drift in positive and negative directions, only positive direction
permanent drift is shown for clarity in presentation. .................................................................... 64
Figure 3.26: Permanent drift as a function of maximum experienced drift. ................................. 68
Figure 3.27: energy absorption as a function of maximum experienced drift. ............................. 68
Figure 4.1:General behavior CC3DNonLinCementitious2 of material. ....................................... 70
xiv
Figure 4.2: Single brick element. .................................................................................................. 71
Figure 4.3: Average stress-strain behavior of concrete brick model: (A) tension and compression,
(B) tension. .................................................................................................................................... 72
Figure 4.4: One dimensional model for “Cyclic Reinforcement” material: (A) developed single bar
model, (B)Stress-strain diagram. .................................................................................................. 75
Figure 4.5: Modeled dog-bone specimen. ..................................................................................... 77
Figure 4.6: Multilinear form of “Cyclic Reinforcement” material. .............................................. 78
Figure 4.7: Stress-strain diagrams from modeled dog-bone specimen. ........................................ 81
Figure 4.8: Different cracking stages: (A) initial cracking, (B) crack opening, (C) crack
localization. ................................................................................................................................... 81
Figure 4.9: ECC stress-strain diagram in tension and compression: (A) HFA-Column, (B) HFA-
Tube, (C) HFA-Tube-SEA & HFA-Tube PSEA. ......................................................................... 82
Figure 4.10: Parameters of the SEA model. ................................................................................. 83
Figure 4.11: Schematic cyclic behavior of the constitutive model for SEA. ................................ 84
Figure 4.12: Comparing experimental and numerical results for SEA. ........................................ 85
Figure 4.13: General behavior of “Memory Bond” material. ....................................................... 86
Figure 4.14: Bond models for the longitudinal reinforcement. .................................................... 87
Figure 4.15: Generating geometry of the numerical models: (A) RC-Column and HFA-Column,
(B) HFA-Tube, HFA-Tube-SEA, HFA-Tube-PSEA.................................................................... 88
Figure 4.16: Modeling longitudinal reinforcement :(A) normal steel reinforcement in RC-Column,
HFA-Colum, and HFA-Tube and neutral axis reinforcement of HFA-Tube-SEA and HFA-Tube-
PSEA , (B) debonded steel reinforcement in HFA-Tube-PSEA , (C) longitudinal reinforcement
with Cu-Al-Mn SEA bar in HFA-Tube-SEA and HFA-Tube-PSEA. .......................................... 90
Figure 4.17: (A) Boundary condition and (B) analogous rigid body motion. .............................. 91
Figure 4.18: Loading: (A) Axial and lateral loading, (B) Shrinkage. ........................................... 93
Figure 4.19: Meshing: (A) linear tetrahedral element with four nodes, (B) meshed model. ........ 94
Figure 4.20: Load-deformation diagrams of the numerical model: (A) RC-Column, (B) HFA-
Column, (C)HFA-Tube, (D) HFA-Tube-SEA, (E)HFA-Tube-PSEA. ......................................... 96
Figure 5.1: Geometry of columns in the parametric study (dimensions are in mm). ................. 102
xv
Figure 5.2: Different cross-sections considered in the parametric study: (A) RC, (B) RC-H, (C)
RECC, (D) RECC-H, and (E) RECC-H-F. ................................................................................. 103
Figure 5.3: Different wall thicknesses for RECC-H cross-section in the parametric study
(dimensions in mm). ................................................................................................................... 104
Figure 5.4: Bridge columns with: (A) no SEA, (B) four SEA, (C) eight SEA, (D) 12 SEA, and (E)
16 SEA bars. ............................................................................................................................... 105
Figure 5.5: Different ECC qualities in parametric study. ........................................................... 105
Figure 5.6: Stress-strain diagrams for different ECC properties. ............................................... 112
Figure 5.7: Cyclic tension-compression behavior of ECC with 0.5% ultimate strain-hardening.
..................................................................................................................................................... 113
Figure 5.8: Bond-slip model used in the parametric study. ........................................................ 114
Figure 5.9: Developed numerical model for parametric study (RECC-H-F sections): (A) geometry
and loading, (B) reinforcement and SEA bars, (C) boundary condition and meshing. .............. 116
Figure 5.10: Comparing the numerical results with the idealized monotonic behavior for the
control case. ................................................................................................................................ 118
Figure 5.11: Effect of ECC quality in cyclic behavior of the bridge columns. .......................... 119
Figure 5.12: Rebar strain at 2% drift: (A) RECC/0.5/NA/00, (B) RECC/1/NA/00, and (C)
RECC/2/NA/00. .......................................................................................................................... 120
Figure 5.13: Strains in Plastic hinge region at 2% drift: (A) RECC/0.5/NA00, (B) RECC/1/NA/00,
and (C) RECC/2/NA/00 (cracks over 1 mm are shown). ........................................................... 121
Figure 5.14: Load-deformation diagrams: (A) RC/NA/NA/##, (B) RECC/NA/NA/##, (C) RC-
H/NA/NA/##. .............................................................................................................................. 122
Figure 5.15: Load-deformation diagrams: (A) RECC-H/2/254/##, (B) RECC-H-F/2/254/##,(C)
RECC-H/2/305/##, (D) RECC-H-F/2/305/##, (E) RECC-H/2/381/##, (F) RECC-H-F/2/381/##.
..................................................................................................................................................... 123
Figure 5.16: Definition of maximum strength, Vm, maximum drift, δm, ultimate strength, Vu,
ultimate drift, δu, yield strength, Vy, yield drift, δy, initial stiffness, E0, permanent drift, δp, and
energy absorption, E0, as metric for quantitate comparison of the parametric study results. ..... 125
Figure 5.17: Normalized maximum lateral strengths, Vm. ......................................................... 129
Figure 5.18: Stresses at maximum drift in the section located at the distance equal to the diameter
of the columns above the foundation: (A) RC/NA/NA/00, (B) RECC/2/NA/00. ...................... 130
xvi
Figure 5.19: Normalized maximum drifts, δm. ........................................................................... 131
Figure 5.20: Normalized yield strengths, Vy. ............................................................................. 132
Figure 5.21: Normalized yield drifts, δy. .................................................................................... 133
Figure 5.22: Normalized ultimate drifts, δu. ............................................................................... 134
Figure 5.23: Normalized initial stiffness, E0. .............................................................................. 135
Figure 5.24: Normalized permanent deformations, δp. ............................................................... 136
Figure 5.25: Stresses in the longitudinal reinforcement at 7% drift: (A) RC/NA/NA/08, (B)
RC/NA/NA/12. ........................................................................................................................... 137
Figure 5.26: Normalized absorbed energies, Ea. ......................................................................... 139
Figure 5.27: 0.5 mm and larger cracks at 7% drift: (A) RC/NA/NA/00, (B) RECC/2/NA/00, (C)
RECC/2/NA/04, (D) RECC/2/NA/16. ........................................................................................ 140
Figure 5.28: Optimized bridge design (highlighted ones are the optimal design) ...................... 142
Figure 6.1: Proposed beam-column joint concept in RC special moment resisting frames. ...... 144
Figure 6.2: Beam-Column subassemblies (shaded areas show the location of ECC, dimensions are
in mm): (A) out-of-plane view of the corner and exterior joint subassemblies, (B) in-plane view of
corner subassembly, (C) in-plane view of the exterior subassembly.......................................... 146
Figure 6.3: Boundary condition of the beam-column subassemblies: (A) exterior, (B) corner. 148
Figure 6.4: Illustration of a test setup for beam-column joints: (A) boundary conditions and
instrumentation, (B) detail of torsional deformation restrainer, (C) instrumentation to measure
shear deformation in the panel zone. .......................................................................................... 151
Figure 6.5: Approximate location of the LEDs on: (A) corner, (B) exterior beam-column joints
(dimensions are in mm). ............................................................................................................. 151
Figure 6.6: Bidirectional bending loading protocol: (A) position of the top of the column in X-Y
plane, (B) column deformation and drift in X-direction, (C) column deformation and drift in Y-
direction. ..................................................................................................................................... 153
Figure 6.7: Stress-strain diagrams from uniaxial compression test: (A) conventional concrete, (B)
ECC. ............................................................................................................................................ 156
Figure 6.8: Assembled formwork of the corner beam-column subassemblies. .......................... 158
Figure 6.9: Preparing the rebar cages: (A) tying the rebar cages, (B) trimming the longitudinal
reinforcements, (C) placing the rebar cages inside the formwork. ............................................. 159
xvii
Figure 6.10: Casting the horizontal ECC Section: (A) wooden dividers and braces, (B) mixing dry
materials for ECC, (C) uniform matrix of ECC, (D) adding fiber to ECC matrix, (E) casting
horizontal ECC section ............................................................................................................... 160
Figure 6.11: Casting the vertical ECC section: (A) placing and sealing the formwork, (B) pouring
ECC, (C) covering and curing the cast ECC sections. ................................................................ 161
Figure 6.12: Preparing and installing the end-plates: (A) top end-plate with the corresponding hole
pattern, (B) bottom end-plate with the corresponding hole pattern, (C) preparation to weld the high
strength bolts, (D) end-plate with welded bolts, (E) centering the end-plate and drilling holes in
the formwork, (F) installed end-plate. ........................................................................................ 162
Figure 6.13: Casting the conventional concrete sections: (A) the prepared specimens, (B) pouring
concrete and vibrating, (C) horizontal section of the control specimen for exterior beam-column
joints. ........................................................................................................................................... 163
Figure 6.14: Preparing the specimens: (A) curing, (B) flipping and mounting on the test setup.
..................................................................................................................................................... 164
Figure 6.15: Beam-column subassemblies load-deformation hysteresis: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ..................................... 165
Figure 6.16:Crack pattern of the tested beam-column subassemblies: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ..................................... 170
Figure 6.17: Torsion: (A) C/RECC/ABBT/RS, (B) E/RECC/ABBT/RS. .................................. 171
Figure 6.18: Position of the potentiometers. ............................................................................... 172
Figure 6.19: Shear deformation: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS,
(D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS,
(H) E/RECC/ABBT/RS. ............................................................................................................. 173
Figure 6.20: Shears, moments, and axial loads in the corner beam-column joints. .................... 174
Figure 6.21: Shear, moments, and axial load in the exterior beam-column joints. .................... 175
Figure 6.22: Reactions at zero moment location of the in-plane beams at the pin supports: (A)
C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E)
C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. 176
Figure 6.23: Reactions at zero moment location of the out-of-plane beams at the pin supports: (A)
C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E)
C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. 177
xviii
Figure 6.24: Reaction at the fixed supports: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ........................................................................... 179
Figure 6.25: Expected deformations in (A) out-of-plane beam of the exterior and in-plane and out-
of-plane beam of the corner beam-column subassemblies, (B) in-plane beams of the exterior beam-
column subassemblies. ................................................................................................................ 180
Figure 6.26: Schematic illustration of the defined parameters for the deformations (in naming, “C”
refers to column, “IPB” refers to in-plane-beam, and “OPB” refers to out-of-plane beam): (A)
C\dX, (B) C\dY, (C) IPB\dX, (D) IPB\dZ, (E) OPB\dY, (F) OPB\dZ ....................................... 181
Figure 6.27: Schematic illustration of the defined tangents to the deformation in the beams and
columns (in naming, “C” refers to column, “IPB” refers to in-plane-beam, and “OPB” refers to
out-of-plane beam): (A) C\θX, (B) C\θY, (C) IPB\θX, (D) IPB\θZ, (E) OPB\θY, (F) OPB\θZ. .... 183
Figure 6.28: An example of the movements of the interest points and the distance between them.
..................................................................................................................................................... 184
Figure 6.29: Maximum and minimum measured C\dX (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 186
Figure 6.30: Maximum and minimum measured C\dY (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 187
Figure 6.31: Maximum and minimum measured IPB\dX (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 190
Figure 6.32: Maximum and minimum measured IPB\dZ (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 191
Figure 6.33: Maximum and minimum measured OPB\dY (shown in (G) and (H) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 192
xix
Figure 6.34: Maximum and minimum measured OPB\dZ (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 193
Figure 6.35: Maximum and minimum measured C\θX (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ........................................................................... 194
Figure 6.36: Maximum and minimum measured C\θY (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS (note that Figure 6.36 (G) has different scale that
others). ........................................................................................................................................ 195
Figure 6.37: Maximum and minimum measured IPB\θX (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 197
Figure 6.38: Maximum and minimum measured IBP\θZ (shown in (G) and (H) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 198
Figure 6.39: Maximum and minimum measured OPB\θY (shown in (A) and (B) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 199
Figure 6.40: Maximum and minimum measured OPB\θZ (shown in (G) and (H) for corner and
exterior subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS,(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS. ...................................... 200
xx
LIST OF TABLES
Table 3.1: Conventional concrete mixture design. ....................................................................... 41
Table 3.2: ECC mixture designs. .................................................................................................. 44
Table 3.3: Properties of PVA fibers [68]. ..................................................................................... 44
Table 3.4: Properties of type F fly ash [69]. ................................................................................. 45
Table 3.5: Properties of silica sand [70]. ...................................................................................... 46
Table 3.6: Mechanical properties of reinforcing steel. ................................................................. 53
Table 3.7: Results of uniaxial cyclic test on removed SEA bars from tested columns. ............... 58
Table 3.8: Quantitative comparison of the performance of test columns. Values in parenthesis
represent the differences in percentage from the control RC-Column. Negative values indicate a
decrease and vice versa. See Figure 3.25 for definitions of the parameters in the first column of
the table. Permanent drift and energy absorption values pertain to a maximum experienced drift of
7%. ................................................................................................................................................ 66
Table 4.1: Material constitutive model properties for concrete to model RC sections and concrete
core. ............................................................................................................................................... 73
Table 4.2: Material constitutive model properties for reinforcing steel. ...................................... 74
Table 4.3: Vectors defining direction of smeared reinforcement in ECC. ................................... 77
Table 4.4: Material constitutive model properties for smeared reinforcement in ECC. ............... 79
Table 4.5: Material constitutive model properties for concrete, CC3DNonLinCementitious2, in
ECC. .............................................................................................................................................. 80
Table 4.6: Material constitutive model properties for SEA. ......................................................... 85
Table 4.7: Solution parameters. .................................................................................................... 95
Table 4.8: A comparison of experimental and numerical results for bridge columns. ................. 98
Table 5.1: Parametric study matrix. ............................................................................................ 107
Table 5.2: Parameters of concrete constitutive model, CC3DNonLinCementitious2, in the
parametric study. ......................................................................................................................... 109
Table 5.3: Parameters of reinforcement constitutive model, “Cyclic Reinforcement”, in the
parametric study. ......................................................................................................................... 110
xxi
Table 5.4: Smeared reinforcement constitutive model to simulate ECC in the parametric study.
..................................................................................................................................................... 111
Table 5.5: Concrete constitutive model, 3D Nonlinear Cementitious2, to simulate ECC.......... 111
Table 5.6: Quantitative evaluation of the parametric study. ....................................................... 126
Table 5.7: Change in the quantitative metrics with respect to the control case. ......................... 127
Table 6.1: Test matrix for beam-column joints. ......................................................................... 147
Table 6.2: Properties of oil-coated PVA fibers [85]. .................................................................. 154
Table 6.3: ECC mixture design for beam-column joints. ........................................................... 155
Table 6.4: Compressive strength of ECC and concrete mixtures. .............................................. 155
Table 6.5: Maximum lateral strength of beam-column subassemblies ....................................... 167
xxii
LIST OF ABBREVIATION
D
DFRCC Ductile Fiber Reinforced Cement Based Composites
DOF Degree OF Freedom
E
ECC Engineered Cementitious Composites
F
FEM Finite Element Modeling
FRC Fiber-Reinforced Concrete
H
HPFRC High-Performance Fiber-Reinforced Concrete
HPMC Hydroxypropyl Methylcellulose
HRWR High Range Water Reducer
L
LED Light Emitting Diode
P
PVA Polyvinyl Alcohol
R
RC Reinforced Concrete
RECC Reinforced ECC
S
SEA Superelastic Alloy
SFRC Steel Fiber-Reinforced Concrete
SMA Shape Memory Alloy
1
CHAPTER 1 - OBJECTIVE AND SCOPE
In the aftermath of a major disaster, a functional transportation network supplemented by other
key infrastructure is essential in order to support quick response and recovery for reducing direct
and indirect losses, in addition to safeguarding the lives of civilians. In the existing design
regulations for seismic regions, the yielding of longitudinal reinforcement is the main mechanism
to absorb seismic energy in bridge priers. Even though this design approach is well tailored to
prevent the superstructure from collapsing, it leads to large permanent deformations and damage
in bridge columns in most cases, rendering the entire bridge dysfunctional. Consequently, the
response and recovery efforts in the aftermath of a major earthquake are impeded and the bridge
is often demolished despite light damage to most of the structural components. Similarly, beam-
column joints in earthquake resistant reinforced concrete (RC) buildings generally experience
large shear deformations during earthquake induced lateral displacements. In addition to shear,
corner and exterior beam-column joints are subjected to torsion and bidirectional bending which
can lead to severe damage in the building, given that conventional concrete is susceptible to these
actions especially under reversed loading cycles.
The main objective of this study is to improve the seismic damage tolerance of bridge priers and
beam-column joints of buildings using high-performance materials, particularly, engineered
cementitious composites (ECC) and superelastic alloys (SEA). Even though a significant amount
of work has been conducted in this regard, the novelty of the work in this thesis is in the
development and understanding of an innovative design concept for the bridge columns and more
realistic simulation of the beam-column joints under seismic loads both of which employ high-
performance materials. For the bridge columns, an innovative design concept incorporating
2
recently-developed Cu-Al-Mn SEA bars and ECC has been introduced. The column design
consists of a prefabricated reinforced ECC (RECC) hollow section embedded in a RC foundation
and filled with conventional concrete. Additionally, the longitudinal reinforcement at the potential
plastic hinge region is totally or partially replaced with recently developed Cu-Al-Mn SEA bars.
The proposed design considers permeability issues in addition to the seismic concerns within these
bridge columns and addresses the relatively high construction costs because of implementing the
high-performance materials. Furthermore, a state-of-the-art testing unit is leveraged to investigate
the performance of exterior and corner beam-column subassemblies in 3D under more realistic
loading scenarios including torsion. Investigating the behavior of 3D beam-column subassemblies
itself is quite challenging given the predicaments in applying bidirectional loading and
corresponding boundary conditions.
To attain the objectives of this research, first, the performance of quarter-scaled bridge columns
constructed based on the proposed design concept is investigated through experimental work.
Following the experimental work, a finite element approach is developed in order to numerically
investigate the performance of the bridge columns. Behavior of ECC is simulated by implementing
the available material constitutive models for concrete and steel; a new SEA constitutive model is
accordingly implemented based on the literature to resemble the behavior of Cu-Al-Mn SEA bars.
The developed finite element approach is verified using the experimental data and then
implemented to conduct a parametric study to optimally use high-performance yet expensive
materials for bridge columns. Finally, the performance of eight quarter scaled 3D beam-column
joints is investigated under combined axial load, bidirectional bending, and torsional loads. The
objective of this testing was to ascertain the efficiency of ECC in terms of enhancing damage
3
tolerance and reducing shear reinforcement when incorporated within corner and exterior beam-
column joints.
Chapter 2 is dedicated to literature review. In this section, the advantages of ECC are discussed
and the conditions for obtaining a strain-hardening behavior are explained. Then, Cu-Al-Mn SEA
is compared with other types of SEAs and its characteristics are described. Subsequently, some
studies incorporating similar materials or methodology to improve the seismic performance of
bridge columns and beam-column joints are discussed. The methodologies and outcomes of each
study are explained briefly to illustrate the contribution of the high-performance materials from
different viewpoints and distinguish the differences between this research and those available in
previous literature.
Chapter 3 focuses on the experimental work conducted on the bridge columns. The scaled bridge
columns are designed based on the proposed concept and their geometry is detailed. In totality, six
scaled columns are considered including a control RC specimen. Number of rebar replaced with
Cu-Al-Mn SEA bars, ECC mixture design, and the ratio of the concrete core area-to-the total
column cross-sectional area were the variables that are considered in developing the test program.
The construction and testing processes are explained and the resulting load-deformation diagrams
are presented along with the imposed damage to each specimen at several drift levels. Finally,
several measures were defined and corresponding values were derived from the load-deformation
diagrams in order to make a quantitative comparison of the experimental results. Additionally,
several material tests were considered to extract the mechanical properties of the consumed
materials in the construction process.
4
Chapter 4 addresses finite element modeling to numerically investigate the performance of the
bridge columns. At the material level, the behavior of ECC and Cu-Al-Mn SEA bars are simulated
and matched with the experimental data. Subsequently, the tested bridge columns in Chapter 3 are
modeled. Establishing the finite element models is explained in a step-by-step process, including
loading, boundary condition, and meshing/solution parameters. After analyzing the models, the
numerical and experimental results are compared using the same measures that are defined in
Chapter 3 so as to evaluate the accuracy of the developed finite element approach.
Chapter 5 deals with a parametric numerical study in order to optimize use of ECC and Cu-Al-Mn
in the bridge columns. A full-scale bridge column designed for high seismic regions was
considered as the control specimen. Different parameters, such as the quality of ECC, number of
Cu-Al-Mn SEA bars in the potential plastic hinge regions, and different cross-sections, including
hollow sections, are considered. The same methodology used in Chapter 4 is implemented to
develop the numerical models within this section. The models are subjected to one cycle of 7%
drift and their load-deformation diagrams are recorded. Several measures are extracted from the
load-deformation diagrams and compared between different cases.
Chapter 6 discusses about the experimental investigation of RECC 3D corner and exterior beam-
column joints using the combination of axial load, biaxial bending, and torsion. In this section, the
design and construction process of eight approximately quarter-scaled corner and exterior (four
each) beam-column subassemblies are explained. While designing the test matrix, the construction
material of the panel zone area, number of shear reinforcement within the panel zone area, as well
as the loading protocols were considered. Subsequently, the test setup and loading protocols are
detailed and the experimental results are summarized accordingly. The experimental results
5
include load-deformation diagrams of the tested specimens, shear deformations in the panel zone
areas, reactions within the boundaries, and deformations in the beams and columns.
The last chapter of this thesis is dedicated to the main findings of each chapter and proposes some
recommendations for future researches.
6
CHAPTER 2 - BACKGROUND AND RESEARCH SIGNIFICANCE
In the following, the literature is first introduced as it relates to the research in thesis. Then the
significance of this research is presented.
2.1 Engineered Cementitious Composites (ECC)
ECC is a special type of high-performance fiber-reinforced concrete (HPFRC) introduced in the
late 90s as an alternative for conventional concrete by Li and co-workers [1-3]. ECC comprises of
Portland cement (Type I and II), fly ash or silica fume, water, silica sand, high range water reducer
(HRWR), and randomly distributed polymeric fibers. ECC mix design is engineered using
micromechanical models in order to reach optimized microstructures, resulting in high ductility
and deformability in tension, with an ultimate tensile strain capacity as high as 5% - equivalent to
around 300-500 times of the corresponding value for concrete. Additionally, plain ECC depicts
multiple micro-cracks with crack widths limited to 100 µm in tension, unlike conventional concrete
for which the tensile failure occurs at a localized crack [4]. Figure 2.1 schematically presents the
tensile behavior of ECC and compares it to tensile behavior of conventional concrete and
conventional fiber-reinforced concrete (FRC).
7
Figure 2.1: Schematic tensile behavior of ECC.
So as to ensure pseudo strain-hardening behavior in ECC, two fundamental requirements need to
be met; otherwise, normal tensile strain-softening behavior occurs as in FRC. First, the maximum
crack bridging stress,
, must be higher than the first tensile cracking strength,
. Second, the
crack tip toughness,
, should be less than the complementary energy,
[4]. Equations(1) and
(2) indicate the two conditions.
<
0
(1)
≤
′
(2)
The first condition simply postulates that after the first crack, bridging effect of fibers should be
strong enough to transfer load through the crack and cause steady state and multiple cracking. In
Equation (1),
is a function of fiber volume, fiber interfacial frictional bond, matrix modulus of
elasticity, and fiber properties, such as length, diameter, and modulus of elasticity. Additionally, a
number of factors such as maximum fabrication flaw size, fiber bridging effect, and matrix
toughness affect the maximum crack bridging stress,
[2].
8
In the second condition, the complimentary energy is calculated using the relationship between the
bridging stress of fibers, , and crack opening, . The relationship between bridging stress of fibers
and crack opening is derived based on fracture mechanics principles and statistics to consider
random orientation and embedment length of the fibers. Figure 2.2 shows a typical − curve,
, and
for strain-hardening composites. In this diagram,
represents the first crack stress
and
is the steady state first crack width [4].
Figure 2.2: Typical σ-δ curve,
, and
for strain-hardening composites.
Equation (3) represents the strain-hardening condition.
≤
0
0
− ≡
′
0
0
(3)
is related to crack tip toughness,
, which is a function of the fabrication void size and
bridging effect of fibers along with the modulus of elasticity of the matrix, !
.
can be calculated
using Equation (4) [4].
σ
δ
0
0
""
""
′
′
9
=
$
2
!
(4)
During the direct tensile test, the number of fine cracks starts to grow in the strain-hardening
branch until the localization point, where one of the cracks starts opening up and causes the failure
of the specimen. The number of tensile cracks and ultimate tensile strain-hardening capacity of
ECC is predicated on how well the two fundamental requirements - Equations (1) and (2) - are
satisfied. As the difference between the left- and right-hand sides of the Equations (1) and (2)
increases, ECC shows a higher strain-hardening capacity with a larger number of tensile cracks.
Parameters such as fiber dispersion uniformity, size distribution of the trapped air, and bond
properties at the interface of the matrix and fiber are important in characterizing the behavior of
ECC within tension and ultimate tensile hardening capacity. Different factors, such as mixer type,
mixing speed and time, sequence of adding the ingredients, experience of personnel, and the
properties of admixtures can strongly affect these parameters. It was shown that an ECC mortar
with low viscosity and high workability generates low shear forces during mixing and results in
poorly distributed fibers. On the other hand, an ECC mortar with a considerably high viscosity
yields a larger size of entrapped air pores. Therefore, the workability and viscosity of the fresh
ECC mortar should be adjusted optimally to ensure proper tensile hardening properties [5].
Through their experimental work, Maalej et al. [6] demonstrated that hybrid-fiber ECCs exhibit
significant strain-hardening capacity and multiple cracking when subjected to high strain rates as
10
-1
s
-1
. The specimens showed a similar behavior in terms of ultimate strain-hardening capacity,
strain-softening, average width of the cracks, and multiple cracking owing to dynamic and quasi-
static loads. However, the ultimate tensile strength was seen to increase drastically (almost 100%)
under high strain rate dynamic loads. According to Kesner et al. [7], the compressive behavior of
10
the highly ductile fiber-reinforced cement based composites (DFRCC) is largely similar to
conventional concrete under cyclic loads. Additionally, cyclic tension-compression loading lead
to reduction in tensile strength and tensile strain capacity of DFRCC due to crack localization in
regions, which get damaged under compressive loads. The results from monotonic tension and
monotonic compression tests of DFRCC were similar to the envelope curves obtained from cyclic
tension-compression tests in terms of modulus of elasticity, peak stresses, and peak strains in the
absence of any compressive damage during the cyclic loading (i.e., the compressive strength is not
exceeded). However, the tensile response of the material should be modified in case the material
compressive strength is exceeded. Even though the tensile properties of ECC have been
investigated extensively, most of these tests were conducted on relatively small specimens without
reinforcement, which disregards the realistic distribution of the fibers and tension stiffening effect
of the continued reinforcement. It was shown that large scale specimens with a continuous
longitudinal reinforcement yield considerably higher strain-hardening capacity (almost doubled)
and extensive multiple cracks in HPFRC [8]. The mechanical properties of ECC in uniaxial tensile
test is a function of fiber orientation. In a comparison between the tensile properties of ECC from
dog-bone specimens using different cross-sections, the ultimate tensile capacity lowered as the
thickness of the specimens increased even though the specimens reached similar tensile strength
values. This phenomenon was related to a better alignment of the fibers in the thinner specimens
[9]. A set of uniaxial tensile tests showed that the tensile strain-hardening capacity of ECC
deteriorates when the specimens were subjected to elevated temperatures, and disappears in
temperatures over 200
o
C. An increment in matrix toughness and the degradation of fiber/matrix
interfacial properties were observed in elevated temperatures up to 200
o
C [10]. Therefore, special
11
considerations are necessitated for the application of ECC under high temperatures such as power
plants and places where there is a high possibility of fire incidents.
In addition to improved mechanical properties of ECC, which makes it suitable for seismic
applications, ECC also exhibits superior durability (i.e., lower permeability) than normal concrete
in both un-cracked and cracked conditions. Freeze-thaw tests that simulate temperature changes in
cold regions lead to a durability factor of 100, which is calculated according to ASTM C666-2013
[11], for ECC compared to 10 for normal concrete. The ECC specimens could undergo over 300
cycles of freezing and thawing with no major deterioration in dynamic modulus and tensile strain
capacity; on the other hand, severe deterioration was observed in normal concrete specimens after
100 cycles [12]. Even though the strain capacity of ECC dropped considerably when exposed to
hot and humid environments in experimental conditions equivalent to 70 years of natural
weathering, it is still significantly higher (over 250 times) than the tensile strain capacity of
conventional concrete [13]. Using single fiber pull out test, Kabele et al. [14] demonstrated that
the chemical bonding between the matrix and fibers reduces in high chloride concentrations such
as marine environments. A reduction in the interface bonding property increases the average width
of the cracks by around two times and brings down the tensile strength by around 10% [15].
The durability of RC structures closely depends on the concrete permeability, which is the rate of
water and corrosive agent penetration. In reinforced structural elements, replacing conventional
concrete with ECC is known to considerably improve the durability of structures. In restrained RC
structural elements, the presence of cracks is inevitable because of shrinkage, structural loads,
thermal deformations, and deterioration through chemical reactions, which modifies the
permeability of cementitious composites. ECC is resistant to cover spalling and pitting formation
on reinforcement owing to high tensile ductility and distributed microcracking [16]. Figure 2.3
12
presents a comparison between the corrosion of steel reinforcement embedded in ECC as well as
conventional concrete.
Figure 2.3: Accelerated corrosion test on: (a) ECC for 300 hours, (b) Mortar for 75 hours [17].
ECC specimens subjected to sustained tensile loads for a long period of time (16 days) showed a
substantial increase in deformation due to the widening of cracks and the emergence of new cracks.
The average crack width was more dependent of the sustained lading time than the level of creep
load [18,19]. However, it was reported that the creep and shrinkage of ECC is significantly lower
as compared to an identical cementitious mixture without fibers. Adding fiber to the mixture
reduced the shrinkage by up to 60% [20]. Additionally, superior shear properties of ECC over
conventional concrete, FRC, and wire mesh RC were observed under intense shear loading. ECC
showed continued straining unlike conventional concrete and improved both the cracking behavior
and the load capacity [21]. In another study, ECC demonstrated shear strength as high as 4.1 MPa
and multiple cracking phenomenon under pure shear loads [22].
13
2.2 Cu-Al-Mn Superelastic Alloy (SEA)
Shape Memory Alloys (SMAs) are a special type of material that exhibit shape memory or
superelastic effect, depending on the crystal structure of the alloy under operating conditions. The
former refers to recovery of plastic strains with heating, while the latter refers to the recovery of
inelastic strains upon stress removal. Martensitic phase transformation between austenite and
martensite phases is known to yield these unique properties. Martensite entails a low-symmetry
crystal structure and typically remains stable at high stresses and low temperatures, whereas
austenite has a high-symmetry crystal structure with stability at low stresses and high temperatures.
Phase transformation between martensite and austenite can be occurred through either temperature
or stress change. At temperatures above the austenite finish temperature, Af , where all crystals are
in austenite form, superelastic behavior is observed [23]. The shape memory effect was initially
observed in 1932 when Chang and Read [24] investigated the resistivity changes as well as
reversibility of transformations in AuCd. Thence, several investigations were conducted to find
new SMA compositions. Nowadays, SMAs are widely renowned as a high strength material,
compared to reinforcing steel, with the ability to undergo large deformations, high damping
capacity, and fatigue and corrosion resistance [25-27]. SMAs that have a superelastic effect under
operating temperatures are called SEAs. Most previous studies related to structural applications of
SMA have made use of NiTi SMA composition. However, the cost of NiTi SMAs is prohibitively
high for structural applications. Additionally, NiTi SMA composition is plagued with
machinability issues, particularly those related to drilling and threading, which stymy its
application in structures. Meanwhile the lowest service temperature for bridges is -51
o
C, according
to ASTM 709 [28], at which NiTi SMAs lose their superelasticity considerably [29]. Over the past
few years, copper (Cu) based SMAs have been developed as an alternative to address the
14
shortcomings of NiTi SMAs. Although Cu-Al-Be is the most widely studied alternate composition,
it exhibits inferior superelasticity as compared to NiTi, and the toxicity of beryllium limits the
practical usage of this composition [30,31]. Cu-Al-Mn SEA, on the other hand, has undergone
significant development over the past few years, yielding significantly stable superelastic behavior
at low and high temperatures, with recovery and rupture strains reaching 12% and 25%,
respectively. This alloy composition is also more easily machineable, and costs less as compared
to NiTi SMAs; it also does not contain toxic compounds [30,32,33]. Figure 2.4 illustrates the
mechanical behavior of NiTi and Cu-Al-Mn SEAs at different temperatures.
Figure 2.4: Comparing stress-strain diagrams of NiTi [32] and Cu-Al-Mn SEAs [33] at different
temperatures (This pictures was reproduced using the data from the provided in the references).
It is reported that the superelastic behavior of Cu-Al-Mn SEA bars is strongly dependent on the
grain size. Like other SEAs, the superelasticty of Cu-Al-Mn is attributed to a reversible phase
transformation between the two main material phases - austenite and martensite. Due to the
0 1 2 3 4 5 6
0
100
200
300
400
500
600
Strain (%)
Stress (MPa)
NiTi at 25 C
NiTi at -25 C
Cu-Al-Mn at 25 C
Cu-Al-Mn at -25 C
Cu-Al-Mn at Room Temp.
15
crystallographic anisotropy in the structure of Cu-Al-Mn, small grain sizes result in a strong
deformation constraint during loading and unloading. This strong constraint makes the plastic
deformations more dominant than reversible deformations through the phase transformation and
degrades the superelastcity [30]. Uniaxial cyclic tensile tests on Cu-Al-Mn SEA wires with
different grain size, d, to wire diameter, D, ratio demonstrated a negative correlation between d/D
ratio and yield stress, post yield stiffness, and ultimate recoverable tensile strain of the wires.
Superelastic behavior was observed in specimens with d/D ratio of 0.6 whereas the specimens with
d/D ratio of 0.217 and 0.06 showed inelastic deformations. The yield strength dropped from 250
MPa to 100 MPa by increasing the relative grain size from 0.217 to 0.6, whereas the corresponding
ultimate reversible tensile strain increased from 2% to 10% [34]. Therefore, developing large
section SEA bars with coarse grains is paramount for SEA and hence for seismic applications such
as bridge columns. Araki et al. [30] tested 8 mm and 4 mm Cu-Al-Mn SEA bars using a cyclic
tensile test. The SEA bars were produced by hot forging and cold-drawing procedures, followed
by substantial aging process at 150
o
C. The aging process is required to stabilize the superelastic
property. The production process resulted in the grain sizes ranging from µm to several tens of mm
along with the presence of grains bigger than the diameter of the specimens. The maximum
recovery strains for 4 mm and 8 mm specimens were reported as 12% and 9%, respectively, with
both specimens failing at 18% tensile strain. Omori et al. [35] produced considerably large Cu-Al-
Mn SEA grains using a crystal growth method based on cyclic heat treatment. This method led to
Cu-Al-Mn grains bigger than several centimeters through the formation of a subgrain structure and
paved the way for producing large diameter Cu-Al-Mn SEA bars. This method is simpler than
conventional strain-anneal methods for attaining abnormal grain growth and also avoids
macroscopic deformation in Cu-Al-Mn to enlarge the grains. Using the cyclic heat treatment
16
procedure, Cu-Al-Mn SEA bars with diameter of 15 mm and 30 mm, and excellent superelasticty
were produced successfully.
Additionally, Cu-Al-Mn SEA bars demonstrated stable superelastic behavior under both quasi-
static and dynamic cyclic loads. Conducting cyclic loading tests on 8 mm Cu-Al-Mn SEA bars led
to a less than 5% increase in stress due to 0.001, 0.5, and 1 Hz loading rates, which stood at around
25% of the corresponding value for NiTi. The stress growth occurs during the forward
transformation process [36]. In another study, the behavior of large diameter Cu-Al-Mn SEA bars
in high loading frequencies reflecting an extreme earthquake event was investigated. In this
experiment, 8 mm, 12 mm, and 16 mm Cu-Al-Mn SEA bars were subjected to cyclic loading tests
with loading rates up to 15 Hz. The results revealed a stable superelasticity in strains of up to 8%
under several loading rates up to 15 Hz. Negligible differences were observed between the
hysteresis loops as the loading rate increased [33]. These negligible differences indicate that the
behavior of the Cu-Al-Mn SEA bars is almost rate independent in loading rates of up to 15 Hz
which ease off the development of constitutive material models for seismic application. Unlike
Cu-Al-Mn SEA bars, the behavior of NiTi SEA bars is function of loading rate [36]. Therefore,
developing a constitutive model for NiTi poses more challenges. To investigate the response of
Cu-Al-Mn SEA bars to fatigue loads, 11 mm diameter bars were subjected to 1,000 cycles of
tensile loads with 6-7% strain amplitudes. The specimens showed a full strain recovery up to 50-
100 cycles, with insignificant degradation in forward transformation stress. Subsequently, the
stress plateau decreased gradually along with a gradual permanent strain up to 200-400 cycles.
Finally, the specimens displayed a nearly linear response. However, no fatigue fracture was
observed in the specimens through the existence of grain sizes larger than the diameter of these
specimens and a uniform distribution of the crystallographic orientation [37].
17
2.3 Application of ECC and SEA in Bridge Columns
Over the past twenty years or so, an emerging technique has attracted the attention of the
researchers to safeguard the bridge columns during strong ground motions using high-performance
materials. Several studies have been conducted about the application of SEA bars and ECC in
bridge columns to mitigate intensive damage and permanent deformations in the aftermath of a
major earthquake. Longitudinal steel rebar was replaced with SEA bars in potential plastic hinge
regions, which, in some cases, were constructed with ECC, so as to enhance damage tolerance in
addition to reducing the residual deformations. Some other relevant features of the SEA and ECC
were also presented in previous sections. This section includes some studies that implemented
these features so as to safeguard bridge columns in regions with high seismic activates are
reviewed.
Saiidi and coworkers reported that the combination of ECC with NiTi SEA in the potential plastic
hinge area of bridge columns considerably decreases the permanent deformations (up to 83%) and
improves the damage tolerance of bridge columns when subjected to incremental biaxial
earthquake excitations [38], incremental cyclic loads [39], and two horizontal components of an
earthquake [40].The average residual to maximum displacement was reduced by 83% and 67% in
circular bridge columns with conventional concrete and ECC, respectively, using the same
arrangement of NiTi in the plastic hinge region. This indicated that ECC contributes towards both
enhancing damage tolerance and mitigating permanent deformations. Through biaxial earthquake
excitation, bridge columns with ECC and NiTi SEA bars showed less damage as compared to the
bridge columns with posttensioning tendons and embedded elastomeric pads, which deployed
other protective methods for bridge columns. In another study, ¼-scale bridge columns with NiTi
18
SEA bars in the plastic hinge region were repaired with ECC after being damaged through a series
of earthquake motions. For repairing, the spalled concrete at the plastic hinge region was replaced
with ECC. The repaired specimens showed significantly lower permanent deformations as
compared to the control RC specimen and the extent of apparent damage reduced considerably
even at large amplitude ground motions [41]. A quasi-static cyclic test on 1/5-scalded bridge
columns incorporating NiTi SEA bars and ECC in the plastic hinge regions demonstrated that this
combination leads to a cosmetic damage to the column after an extensive 10% lateral drift. Even
though the longitudinal SEA bars experienced large strains, the surrounding ECC were able to
accommodate the strains unlike conventional concrete [39].
Nakashoji and Saiidi [42] tested square bridge columns scaled to 30% of the original dimensions
using NiTi SEA bars and ECC at the plastic hinge regions. Two identical specimens with different
NiTi SEA bar length, three-quarters and one time of the cross-section diameter, were considered.
The objective of this research was to evaluate the performance of these columns of the first bridge
to be constructed in Seattle, Washington by incorporating these high-performance materials. The
longitudinal reinforcement had a circular formation, whereas the SEA bars had different lengths.
Up-set headed couplers were used to connect the longitudinal mild steel and NiTi bars without
machining the SEA bars into a dog-bone shape. The columns were subjected to a quasi-static cyclic
loading and their performance/damage level was compared with the control RC column. On
average, the columns with ECC and SEA showed 85%, less permanent deformation as compared
to the control RC bridge columns; the damage was limited to just one major crack. It was found
that using shorter SEA bars has a negligible effect on the self-centering capability of these columns
and the couplers are able to effectively connect NiTi and steel bars, eliminating the machining
costs.
19
Most of the experimental work has been conducted on the application of SEAs in bridge columns
concentrated on NiTi SEA bars, since application of other types of SEAs were unfeasible in bridge
columns. Specifically, there is a limited number of studies in literature regarding the application
of Cu-Al-Mn SEA bars in bridge columns [43-45].
In addition to the experimental work, several numerical studies have been conducted to evaluate
the resiliency of bridge columns with SEAs. Bipin and Hao [46] numerically investigated the
performance of multiple and single-pier bents with different reinforcement ratios. It was observed
that replacing the longitudinal rebar with SEA bars yielded comparable results with RC bents in
terms of peak drift and damage parameters, such as concrete crushing and reinforcement yielding.
However, implementing SEA bars decreased the residual drifts significantly (70-90%), especially
in multiple-pier bents with high reinforcement ratios. A multivariate parametric study illustrated
that the plastic hinge length of bridge columns reinforced with SMA (SMA-RC pier) increases as
the yield strength of SMA, the axial load, and the column length-to-diameter (aspect) ratio
increases. On the contrary, it decreases when the concrete compressive strength and reinforcement
ratio increases [47]. The performance of SMA bars in improving the energy absorption capacity
of self-centering precast segmental bridge columns was investigated through 3D finite element
models. These columns suffer from low energy absorption capacity, which imposes a high lateral
seismic demand. To neutralize this deficiency while maintaining the re-centering property of the
precast segmental bridge columns, supplementary SEA bars were utilized as starter bars. The SMA
starter bars were able to improve the energy absorption capacity of these bridge columns in high
seismicity zones [48].
A number of experimental and numerical studies confirmed the efficiency of SMA-RC piers in
mitigating residual deformations, when subjected to seismic loads. However, most of them
20
concentrated on replacing the longitudinal reinforcement completely with SMA bars. In addition,
in the wake of considerable costs associated with SEA and corresponding construction difficulties,
SEA-RC pier showed a lower energy absorption capacity as compared to an RC pier due to the
inherently low energy absorption capacity of SEAs. Additionally, the materials surrounding the
SEA bars required a higher damage tolerance capacity like ECC, due to large strains in the rebar
[44].
Several experimental and numerical studies have demonstrated the efficiency of incorporating
ECC and SEAs in bridge columns to mitigate the extensive damage and permanent deformations
caused by a strong earthquake. However, the relatively high cost of ECC and SEAs as compared
to conventional concrete and mild steel, respectively, impedes their application. The cost of ECC
is around 2-3 times higher than conventional concrete. Although the price of Cu-Al-Mn SEA bars
is a fraction of NiTi SEA bars, it is substantially higher than that of mild steel. On the other hand,
considering the long term performance of these bridge columns as well as the effectiveness of
high-performance materials in reducing repair cost, the application of SEA bars and ECC is an
attractive proposition in scenarios, which necessitate a resilient infrastructure.
2.4 Application of Fiber-Reinforced Concrete (FRC) in Beam-Column Joints
RC frames are very common structural systems for residential, commercial, and governmental
buildings, in addition to key social service facilities such as hospitals. Recent strong earthquakes
have exposed the susceptibility of RC beam-column joints to extensive damage, resulting in the
collapse of buildings in some cases. In the case of extremely intense earthquakes, RC buildings
with highly damage tolerant joints can have a considerable impact on the resiliency of communities
by reducing the casualties and repair cost whilst ensuring suitable aftermath recovery efforts.
21
Beam-column joints in RC frames generally experience large shear stresses during earthquake-
induced lateral displacements [49]. The main focus in the existing design recommendations for
RC beam-column connections is to address the minimum requirements in order to prevent the
collapse of structure and loss of lives. Some of the requirements include ensuring a strong column-
weak beam behavior, providing adequate confinement in the joint and critical regions of the
adjoining members, and delivering sufficient shear strength as well as anchorage length to
longitudinal rebar passing through the joint [50]. It is important to note that meeting these
requirements does not prevent the formation of large diagonal cracks during strong earthquakes
and causes reinforcement congestion in the joint region. Using FRC in constructing beam-column
joints is just one of the options to improve their rotational capacity while minimizing post-
earthquake repair needs. Shakya et al. [51] demonstrated that using 1.5% of steel fibers in concrete
improves the flexural and shear capacity of beam-column joints and reduces the beam
longitudinal/joint transverse steel rebar requirements.
In addition to steel fiber-reinforced concrete (SFRC), the implementation of HPFRC has been
considered to improve the seismic performance of beam-column joints. Several studies have been
conducted to investigate the performance of beam-column joints constructed with HPFRC and
were subjected to reversal cyclic loading. The high shear strength and flowability characteristics
of ECC makes it an ideal choice for seismic applications. Generally, ECC is only used in the joint
panel zone and potential plastic hinge region of adjoining elements owing to its relatively high
cost as compared to conventional concrete. Qudah and Maalej [52] reported that the use of ECC
can increase the energy dissipation capacity of interior beam-column joints by up to 20%. The
specimens preserved their structural integrity, and no major crack or spalling was observed under
considerable inelastic deformations. Zhang et al. [53] pointed out that replacing panel zone
22
transverse reinforcement with polypropylene based ECC maintains the flexural failure mode of
the connection, suggesting that ECC can be a feasible alternative for stirrups in the joint panel
zones. Parra-Montesions et al. [49] developed a highly damage tolerant interior beam-column joint
by applying HPFRCC to the joint panel zone and potential hinge regions of adjacent beams. The
specimens showed a significant deformation capacity, and no bond deterioration was observed in
the beam longitudinal rebar.
Although significant effort has been devoted to study the behavior of beam-column joint, most of
them have focused on in-plane elements which do not reflect the complex loading conditions of
an earthquake. In real situations, earthquakes apply two horizontal components to the building;
thereby, the connections of RC moment resisting frame structures are subject to bidirectional
bending [54]. Consequently, rapid stiffness degradation can occur in beam-column joints, given
that the favorable confinement effect of stirrups diminishes [55,56]. Furthermore, bidirectional
bending is aggravated by a concurrent torsion. Additional shear stresses formed by torsion inflicts
asymmetric damage to the joint panel zone, which leads to a lower energy dissipation capacity
owing to the pinched type load-deformation hysteresis diagrams [57,58].
2.5 Research Significance
Environmental aging and earthquakes signify the two main challenges for bridge infrastructure
worldwide. In the United States, nearly one quarter of over 600,000 bridges either do not meet the
requirements of current codes and standards (functionally obsolete), or are in extensive need of
repair and maintenance (structurally deficient) [59]. The average age of a U.S.-based bridges is 42
years, and most bridges have a design life of 50 years. Similar statistics hold true for other parts
of the world. A significant investment will be made to repair or replace these existing bridges, not
23
to mention the addition of new ones to meet the demands of an increasing population. A new
column design is proposed in this research in order to address the seismic resiliency and durability
objectives for bridge infrastructure using ECC and Cu-Al-Mn SEA bars. These high-performance
materials are used in an innovative way to maximize the performance while minimizing costs. The
proposed concept is experimentally and numerically evaluated in terms of mechanical performance
under simulated seismic loading, using state-of-the-art multi-axial testing units. The results suggest
that the proposed concept can almost eliminate permanent deformations, thereby, meeting
immediate occupancy objective up to drift levels of 7%. A numerical parametric study is conducted
to optimize the use of high-performance materials with a view to lower construction costs in of the
wake of financial issues regarding large infrastructures.
Additionally, despite significant advances in structural materials, catastrophic beam-column joint
failures in RC buildings continue to be a problem during extreme events, such as earthquakes. The
behavior and failure mechanisms in RC joints have been extensively studied in the past and are
well understood under simple (planar) loading conditions. Experiments have been conducted on
cruciform and T-joints wherein the beams and columns have been subjected to in-plane (uniaxial)
bending moments. Current practice entails an increase in the joint dimensions or the amount of
reinforcing steel to enhance the capacity of these joints and prevent premature failure. However,
this approach is definitely not sustainable. Additionally, as observed in recent earthquakes, it does
not always prevent catastrophic failures in the joint region, which leads to the conclusion that the
behavior of joints under complex (multi-axial) loading conditions needs to be better understood.
Therefore, there is a need and opportunity to employ promising advancements in material science
to ensure a smooth transition of the design and construction of beam-column joints from being
inherently vulnerable and unsustainable to inherently forgiving and sustainable. HPFRC with high
24
ductility, confinement, and energy absorption characteristics is an attractive material to enhance
the mechanical performance and improves the sustainability of RC buildings within the long term.
Experimental investigations support our understandings about the performance of new materials
and create a benchmark for numerical studies. There is very scarce experimental data on the
applicability of HPFRC in order to improve the seismic resistance of beam-column joints. This is
particularly true for 3D joint configurations under complex loading conditions.
25
CHAPTER 3 - EXPERIMENTAL PROGRAM- BRIDGE COLUMNS
The main objective of this thesis is improving seismic performance of structural elements using
high-performance materials as a transition toward next generation seismic resistant structural
elements. Because of time and resources limitations, two structural elements, bridge columns and
beam-column joints, were chosen to investigate through their impact on resiliency of
infrastructures. In this section performance of an innovative bridge column design incorporating
ECC and Cu-Al-Mn SEA bars is investigated experimentally while experimental program of
beam-column joints is explained in Chapter 6. The proposed columns feature a precast reinforced
RECC hollow section that encompasses the longitudinal and transverse reinforcement. The
intended purpose of the hollow section is to delay the ingress of corrosive agents to the
cementitious matrix with the low permeability of ECC both in uncraked and cracked conditions.
The plastic hinge longitudinal rebar is totally or partially replaced with Cu-Al-Mn SEA bars to
absorb energy while reducing the permanent deflection of columns. It should be mentioned that
mechanical performance of the proposed approach under simulated seismic loads was investigated
without considering the durability aspects.
3.1 Innovative Column Design Concept
A new bridge column design is introduced to enhance the damage tolerance of bridge piers and
minimize permanent deformations caused by lateral loads. Figure 3.1 displays the proposed design
concept for bridge columns. A concrete-filled prefabricated RECC hollow section that consist of
both longitudinal and transverse reinforcement was proposed. In this column design, a portion or
the entire longitudinal rebar within the potential plastic hinge region is replaced using Cu–Al–Mn
SEA bars. The RECC hollow section is embedded inside a conventional RC footing, up to a depth
26
that approximately equals the diameter of the column, in order to prevent cold joint formation at
the highest moment location. Then, the footing and bent are cast simultaneously as the hollow
section is filled with conventional concrete.
Figure 3.1: Proposed design concept for bridge columns.
This design approach is expected to improve the performance of bridge columns in terms of lateral
strength and ductility (due to improved tensile stress-strain behavior of ECC), minimize the extra
costs associated with construction by minimizing the use of more expensive ECC material, and
improve durability as a result of significantly lower permeability of RECC hollow section, when
compared to conventional concrete, servings as a protective layer for rebar. Additionally, material
quality control could be easily achieved using a precast hollow section; the total cost of
27
construction could be reduced further by reducing the amount of transverse reinforcement (due to
higher shear resistance of ECC), and forming and workmanship (attributed to RECC hollow
section serving as the formwork and vertical support for the bent).
3.2 Geometry and Test Matrix
A bridge column was designed based on the current provisions for seismic design of highway
bridges [60] and scaled in accordance to the displacement and force capacities of the testing
equipment used for this research, which resulted in a scale factor of approximately 1/4. Five
reduced-scale bridge columns were considered so as to investigate their flexural behavior and
compare with that of a conventional RC column. Figure 3.2 illustrates the geometry of the
specimens. All the six specimens were seen to have a similar geometry with a 914.4 mm clear
height and a 203.2 mm diameter (circular cross-section). The inner diameter of the hollow sections
in the specimens constructed on the basis of the proposed approach was 127 mm. To prevent the
formation of cold joints, the precast RECC hollow section was inserted 152.4 mm and 203.2 mm
into the top end cap and footing, respectively. In order to transfer the load from the testing
equipment into the columns, a 508 x 508 mm square end cap with 254 mm thickness was poured
atop each column. A 393.7 mm deep footing with plan dimensions of 635 x 635 mm was
constructed in order to connect the columns to the strong floor (see Figure 3.2). Both the end cap
and the footing were constructed using conventional concrete and were heavily reinforced to
prevent damage in these regions during the test.
28
Figure 3.2:Geometry of bridge columns (in mm): (A) as built specimens in the test setup, (B) RECC
prefabricated hollow sections with conventional concrete core, and (C) detail of plastic hinge rebar
replacement with SEA.
Figure 3.3 presents the test matrix. A conventional RC column was constructed as a control
specimen and was labeled 'RC-Column'. Other specimens included one column that was
constructed entirely from RECC ('HFA-Column'); two columns were constructed using the
proposed concept, but no Cu–Al–Mn SEA bars in the plastic hinge area with two different ECC
mixtures ('HFA-Tube' and 'M45-Tube'), whereas two columns were constructed based on the
29
proposed concept with a different number of Cu–Al–Mn SEA bars across the plastic hinge area
('HFA-Tube-SEA' and 'HFA-Tube-PSEA'). In labeling these specimens, HFA refers to an ECC
mixture with a high fly ash content, and M45 is a commonly used ECC [61-63] with a higher
cement content. 'PSEA' is used to indicate that the column is partially reinforced using Cu–Al–Mn
SEA bars in the plastic hinge region.
Figure 3.3: Test matrix: section A-A and B-B are indicated in Figure 3.2.
Eight Grade 60 rebar, including two 12.7 mm (#4 in U.S. designation) diameter placed on the
section's neutral axis and three 9.5 mm (#3 in U.S. designation) diameter on either side of the
neutral axis, were placed evenly in a circular pattern to achieve a longitudinal reinforcement ratio
of 2% in each specimen. A clear cover of 10 mm was used for all the specimens. In the HFA-Tube-
SEA and HFA-Tube-PSEA specimens, a 10.2 mm rebar was used to connect the Cu–Al–Mn SEA
bars with the longitudinal reinforcement to ensure yielding of Cu–Al–Mn SEA bars prior to the
yielding of the steel longitudinal rebar. The Cu–Al–Mn SEA bars had a diameter of 11.2 mm and
9.5 mm at both ends as well as in the gage region, respectively, with a total length of 249.9 mm.
30
The Cu–Al–Mn SEA bars were threaded at both ends and connected to the threaded rebar using
custom-made mechanical couplers with U.S. designation 3/8-16 thread at one (for connecting to
the Cu–Al–Mn SEA bars) and U.S. designation 1/2-13 thread on the other end (for connecting to
the steel rebar). The Cu–Al–Mn SEA bars were embedded into the foundation 66 mm. Both the
Cu–Al–Mn SEA bars and steel rebar (in the plastic hinge region) were debonded from the
surrounding ECC matrix in the HFA-Tube-SEA as well as HFA-Tube-PSEA specimens. A #9
gage galvanized smooth steel bar with 3.89 mm diameter was utilized in spiral form with a pitch
of 31.75 mm in order to meet the requirements for transverse reinforcement in existing provisions
for a seismic design of highway bridges [64]. This arrangement resulted in a transverse
reinforcement ratio of 0.64%.
To connect the specimens to the loading unit as well as strong floor, 12 and 8 through holes with
25.4 mm diameter were considered on the top cap and the foundation, respectively. Center-to-
center distance between the holes of the top cap were 127 mm following the hole pattern on the
testing unit. The holes of the foundation were located based on the hole pattern of the strong floor,
which resulted in 203.2 mm center-to-center distance.
3.3 Construction Process
3.3.1 Formwork
Two wooden formworks that can be disassembled and reused were designed to construct the
specimens. Figure 3.4 shows the formworks. Each set consisted of a base plate placed atop of a
wooden pallet, formwork of the foundation, formwork of the top cap, and a frame in order to hold
the formwork of the top cap. A wooden plate was placed on top of the frame as a base for the
formwork of the top cap. The formwork was constructed in four distinct pieces from 19 mm
31
plywood and strengthened across different locations. The pieces were fastened to each other using
threaded rods and nuts. PVC pipes with 25.4 mm inner diameter were used to create the through
holes in the top cap and the foundation. Wooden pieces connected to the base plate and wooden
plate consisting of the hole patterns were used to hold the PVC pipes in place. A PVC tube with
an inner diameter of 203.2 mm was used to cast the circular column. Several wood ties were used
to retain the sections and restrain their deformation with respect to each other.
Figure 3.4: Formwork to construct the bridge columns: (A) formwork of the foundation and the top cap
as well as the frame, (B) the base of the top cap and the fixtures to hold the PVC pipes in place, (C) PVC
pipes to create the whole through holes.
3.3.2 Rebar cages
For consistency, the entire steel reinforcement of each rebar size was selected from the same heat
of material. 12.7 mm (#4 in U.S. designation) and 9.5 mm (#3 in U.S. designation) steel bars were
32
used to reinforce the foundation and the top cap, respectively. Figure 3.5 illustrates the construction
of the rebar cages. The longitudinal reinforcement was cut, bent and assembled with the
reinforcement of the foundation along with the top cap inside the formwork for RC-Column and
HFA-Column specimens. The rebar cage of the column for HFA-Tube, M45-Tube, HFA-Tube-
SEA, and HFA-Tube-PSEA was assembled individually to construct the prefabricated RECC
hollow section first. In HFA-Tube-SEA and HFA-Tube-PSEA specimens, the longitudinal
reinforcement with Cu-Al-Mn SEA bars was constructed in three distinct parts (two steel
reinforcement parts and one Cu-Al-Mn SEA bar). The steel reinforcement parts were threaded
using tap and die and linked to the machined Cu-Al-Mn SEA bars using custom made mechanical
couplers. Several wooden fixers were used to keep the longitudinal reinforcement in place after
which, the transverse reinforcement was attached to the longitudinal reinforcement. In HFA-Tube-
SEA and HFA-Tube-PSEA specimens, the entire longitudinal reinforcement in the potential
plastic hinge region was wrapped with tape to debond them from the surrounding ECC matrix
prior to attaching the transverse reinforcement. The SEAs bars were debonded so that the
deformations and cracks got concentrated within the plastic hinge region with limited damage to
other portions of the column. Additionally, debonding helps to reduce permanent deformation in
the presence of SEA bars following the load removal by eliminating any interaction with the
surrounding material.
33
Figure 3.5: Rebar cages of the bridge columns: (A )rebar cage of RC-Column and HFA-Column, (B)
rebar cage of column of HFA-Tube and M45-Tube, (C) machined Cu-Al-Mn SEA bars, (D) threaded steel
reinforcement, (E) rebar cage of column of HFA-Tube-SEA with debonded Cu-Al-Mn SEA bars in the
potential plastic hinge region.
3.3.3 RECC hollow section
To construct the prefabricated RECC hollow section in HFA-Tube, M45-Tube, HFA-SEA, and
HFA-PSEA specimens, an open-ended bucket was initially passed through the rebar cage of the
column and the inside form; see Figure 3.6(A). Second, the bucket was filled with sand up to the
beginning of the RECC hollow section and stabilized with cement grout. Third, the outside form
was fixed in place by bridging the gap between the bucket and the outside form; see Figure 3.6(B).
To ensure the clear cover, several wooden blocks were used to retain the longitudinal
reinforcement in a specific distance from both inside and outside forms. Next, ECC was prepared
and poured between the two forms; see Figure 3.6(C). No consolidation of the ECC was
necessitated here. Finally, both the inside and outside formwork were removed after 24 hours of
curing to obtain the prefabricated RECC hollow sections; see Figure 3.6(D). It is noted that the
34
forms should be water-proof and greased to prevent any water absorption from the ECC mix and
ease their removal. The prefabricated RECC hollow sections were wrapped using plastic sheets to
prevent water evaporation from ECC for curing purposes.
Figure 3.6: Prefabricated RECC hollow section: (A) placing the rebar cage and inside form, (B) placing
the outside form, (C) pouring ECC between the two forms, (D) prefabricated RECC hollow section after
removing the forms.
3.3.4 Casting concrete
After preparing the RECC hollow sections, each one was placed inside a foundation formwork;
the foundation rebar was assembled around the hollow sections; and the wooden frame was
assembled to support the top cap, see Figure 3.7 (A). Subsequently, the formwork of the top cap
was assembled and the prepared reinforcement was placed, see Figure 3.7 (B). The PVC tubes
35
used to create the through holes in the foundation and the top cap were fixed prior to casting the
concrete, see Figure 3.7(C). Finally, conventional concrete was used to cast the foundation and the
top cap simultaneously with the core of the RECC hollow section. The casted concrete was
vibrated and tapped to release the trapped air and improve the final quality; see Figure 3.7(D) and
Figure 3.7(E). In RC-Column and HFA-Column specimens, the foundation, the column and the
top cap were casted separately with a time difference of 24 hours to allow the previous cast to gain
sufficient strength.
Figure 3.7: Casting concrete: (A) placing RECC hollow section inside the foundation, (B) preparing the
formwork and reinforcement of the top cap, (C) placing and fixing the PVC tubes, (D) casting the
foundation, (E) casting the top cap and vibrating.
3.3.5 Preparing the specimens
To prepare the specimens for testing, the cast specimens were initially covered with plastic sheets
and kept wet, by spraying water frequently, for at least 4 days for curing purposes, see Figure
36
3.8(A). Second, the specimen was moved inside the lab and the formwork was removed, see Figure
3.8(B). Third, the topmost surface of the top cap was smoothed using a thin layer of self-leveler
mixture after applying a proper adhesive to prepare a flat surface for loading and prevent stress
concentrations, see Figure 3.8(C). Fourth, the extra portion of the PVC pipes was cut and ground
to remove any bumps, see Figure 3.8(D). Finally, the surface of the specimen was cleaned and
painted for better visibility of the cracks during the testing, see Figure 3.8(E).
Figure 3.8: Preparing the specimens: (A) Curing, (B) removing the formwork, (C) preparing the upper
surface of the top cap for applying self-leveler, (D) removing extra portion of the PVC pipes, (E) cleaning
and pained the specimen.
3.4 Test Setup and Instrumentation
A state-of-the-art multi-axial testing unit was used to test the specimens under combined horizontal
and axial loading, which can be seen in Figure 3.9. These testing units are capable of imposing
any combination of forces (or moments) and displacement (or rotations) in all six degrees-of-
37
freedom (DOF). An advanced controlling system enables the operator to control several DOFs in
load or displacement control mode simultaneously and gauge the feedbacks. After preparing the
specimens, twelve post-tensioned high strength rods were used to transfer the load from the platen
of the multi-axial testing unit into the specimens. These specimens were placed atop three steel
rectangular tubes as spacers and facilitators so as to move the specimens underneath the loading
unit. Eight high strength post-tensioned threaded rods as well as nine high strength post-tensioned
bolts were used to connect the steel tubes to these specimens and strong floor respectively, see
Figure 3.2(B). This configuration was designed to remain elastic for the highest loads expected
during testing. A redundant set of six external linear potentiometers were used to record the
deformations of the specimens, in addition to the internal sensors of loading units that measured
the displacements as well as forces in the actuator space, which were subsequently converted into
displacements, rotations, forces and moments in the six DOF Cartesian space. The crack
propagation and damage evolution during these tests were recorded using three high resolution
digital cameras synchronized with the loading protocol in order to capture one image at each
loading step.
38
Figure 3.9: Test setup.
3.5 Loading Protocol
In order to investigate the performance of the proposed column design under earthquake loads, the
columns were tested in a cantilevered condition with a planar loading. In other words, the in-plane
rotation was released at the top of the columns and all the loading (horizontal and vertical) was
contained in the same plane. A quasi-static reversed cyclic displacement was applied horizontally
on the top of these specimens. Quasi-static testing method is appropriate for structural elements
incorporating materials with rate-independent response, such as steel and concrete [65]. Cyclic
behavior of Cu-Al-Mn SEA bars are independent of loading rate [33]and ECC mechanical
39
characteristics of HPFRC, except tensile strength, remain unchanged by increasing loading rate
[6]; therefore, quasi-static testing method is a suitable approach to investigate seismic
performance of the innovative bridge design concept introduced in this thesis. However, some of
the results might be a lower limit for high frequency earthquakes since the tensile strength of
HPFRC improves as loading rate increases [6,66].
The horizontal reversed cyclic displacements were repeated twice at each drift level from 0.25%
to 9.5%, which can be seen in Figure 3.10. It must be noted that the drift was calculated as the ratio
of the horizontal displacement-to-the total height of the column, which denotes the distance
between the top of the footing to the top of the top cap where the horizontal displacement was
applied. In addition to the horizontal cyclic displacement, a constant 100 kN axial load was applied
on the top cap to simulate superstructure weight and corresponding live loads, see Figure 3.9. The
applied axial load is approximately equal to 7.5% of the squash capacity of the control RC column,
which is a typical axial load level for bridge columns subjected to earthquakes, according to
seismic design regulations.
40
Figure 3.10: Loading protocol.
3.6 Material Properties
This section explains the mechanical properties of the materials used in constructing the bridge
columns.
3.6.1 Concrete
The concrete mix design included a typical type I/II Portland cement, gravel, sand, and water
without any superplasticizer. Table 3.1 shows the mixture proportions of the concrete. The cement,
sand, and gravel were purchased from local suppliers and mixed using a gravity type mixer. A
maximum aggregate size of 10 mm was used in this concrete mixture, which had a target
compressive strength of 35 MPa.
0 1 3 5 7 9 11 13 15 17 19
-100
-80
-60
-40
-20
0
20
40
60
80
100
Loading Cycle
Displacement (mm)
0 1 3 5 7 11 13 15 17 18
-9.5
-7
-5
-4
-3
-2
-1
0
1
2
3
4
5
7
9.5
Drift(%)
41
Table 3.1: Conventional concrete mixture design.
Specimen Cement Gravel Sand Water
RC-Colum 1 2.1 2.3 0.51
HFA-Tube 1 2.1 2.3 0.45
M45-Tube 1 2.1 2.3 0.51
HFA-Tube-SEA
& PSEA
1 2.1 2.3 0.45
In order to conduct material tests, cylinder specimens with a height of 203.2 mm and a diameter
of 101.6 mm were prepared at the time of casting the columns. Figure 3.11 displays the setup to
the uniaxial compression test. The uniaxial compression test was performed on the specimens
using a Forney machine with 2224 kN load capacity at a constant loading rate of 0.3 mm min
-1
,
according to ASTM C39/C39M
[67]. Both ends of these specimens were ground to provide a
smooth loading surface and prevent stress concentrations. The vertical strain was measured
constantly during the test using a linear strain conversion transducer.
The uniaxial compression test resulted in an average compressive strength of 49.2 MPa, 30.1 MPa,
35.7 MPa, 40.7 MPa, and 35.4 MPa for RC-Column, HFA-Tube, M45-Tube, and HFA-Tube-SEA
and HFA-Tube-PSEA (four separate batches), respectively. The difference between the
compressive strengths resulted from different water-to-cement ratio and testing ages which were
over 80, 34, 55, and 29 days respectively. Figure 3.12 presents the results from the uniaxial
compression test on cylinder specimens. It should be noted that the ultimate compressive strength
was measured in all the tested specimens while the stress-strain diagram was captured in some of
them.
42
Figure 3.11: Test setup of the uniaxial compression test.
Figure 3.12: Some of the results from uniaxial compression test on conventional concrete: (A) RC-
Column, (B) core of HFA-Tube, (C) core of M45-Tube.
43
3.6.2 ECC
The ECC mix design included typical type I/II Portland cement, type F fly ash, silica sand, water,
HRWR, and polyvinyl alcohol (PVA) fibers. Two different ECC mixtures (HFA and M45) were
used as explained before. The only difference between these two mixtures was the amount of
Portland cement that was replaced with fly ash. In both HFA and M45 mixtures, a 2% volume
fraction of PVA fibers was distributed randomly. Table 3.2 provides the proportions of the ECC
mixtures. The fibers (Nycon-PVA RECS15) were obtained from Nycon Corporation [68]. Table
3.3 summarizes the properties of these PVA fibers. The fly ash was obtained from the Rockdale
plant of Boral Material Technologies [69] within Texas and entailed the properties shown in Table
3.4. The silica sand, under the commercial name of T-60, was obtained from the Kosse plant of
US Silica company [70] based in Texas. Table 3.5 summarizes the properties of the silica sand.
44
Table 3.2: ECC mixture designs.
Fraction by weight
HFA M45
Ingredient
Portland cement, c 1 1
Fly ash, fa 2 1.2
Silica sand, ss 0.8 0.8
Water, w 0.57 0.55
HRWR
a
0.01 0.01
PVA
b
fiber 2%
c
2%
c
Ratios
w/c 0.57 0.55
w/(c+fa) 0.19 0.25
w/solids
d
0.15 0.18
(c+fa)/solids 0.78 0.72
Mech. Prop.
Target compressive strength
(MPa)
55 55
a
High range water reducer
b
Polyvinyl alcohol
c
By volume
d
Solids=c + fa + ss
Table 3.3: Properties of PVA fibers [68].
Length (mm) Diameter (µm) Specific gravity
Tensile strength
(MPa)
Alkali resistance
8 38 1.3 1600 Excellent
45
Table 3.4: Properties of type F fly ash [69].
Test type Tested item Results
ASTM C
618 class
F/C
AASHTOO
295 class
F/C
Chemical
tests
Silicon dioxide, SiO2 (%) 57.61
Aluminum oxide, Al2O3 (%) 22.79
Iron oxide, Fe2O3 (%) 4.38
Sum of SiO2, Al2O3, Fe2O3 (%) 84.78 70/50 min. 70/50 min.
Calcium oxide, CaO (%) 9.65
Magnesium oxide, MgO (%) 1.86
Sulfur trioxide, SO3 (%) 0.42 5 max. 5 max.
Sodium oxide, Na2O (%) 0.2
Potassium oxide, K2O (%) 0.92
Total alkalies, asNa2O (%) 0.81
Physical
tests
Moisture content (%) 0.04 3.0 max. 3.0 max.
Loss on ignition (%) 0.41 6.0 max. 5.0 max.
Amount retained on No. 325 sieve
(%)
25.0 34 max. 34 max.
Specific gravity 2.3
Autoclave Soundness (%) -0.02 0.8 max. 0.8 max.
Strength activity index with
Portland cement at 7 days (% of
control)
82.4 75 min. 75 min.
Strength activity index with
Portland cement at 28 days (% of
control)
99.6 75 min 75 min
Water required, % of control 95.0 105 max 105 max
Loose bulk density (lbs/ft
3
) 79.8
46
Table 3.5: Properties of silica sand [70].
Test type Tested item Results
Chemical
analysis
Silicon dioxide, SiO2 (%) 99.7
Iron oxide, Fe2O3 (%) 0.05
Aluminum oxide, Al2O3 (%) 0.39
Titanium dioxide, TiO2 (%) 0.04
Calcium oxide, CaO (%) <0.01
Magnesium oxide, MgO (%) 0.01
Sodium oxide, Na2O (%) <0.01
Potassium oxide, K2O (%) 0.04
Loss on ignition, LOI (%) 0.2
Physical
properties
AFS
a
acid demand (@pH 7) 0.6
AFS grain fineness 55
Color White
Grain shape Subangular
Hardness (Mohs) 7
Melting point (F
o
) 3100
Mineral Quartz
Moisture content (%) <0.1
pH 7
Specific gravity 2.65
a
American Foundrymen’s Society
A shear type mixer was used for mixing the ECC. The mixing process consisted of mixing all the
dry materials for three to five minutes in order to create a uniform blend, adding 50% of the water
and HRWR and then mixing it for three to five minutes, before gradually pouring the rest into the
mix while the mixer was continuously working. The HRWR was in liquid form and gradually
added to the mixture after pouring the first water portion. Finally, the fibers were gradually added
to the mixture while the content was being mixed; the final mixture was mixed for three to four
minutes to reach a homogeneous mixture.
47
Figure 3.13 provides the uniaxial compressive test results. The uniaxial compression test on the
101.6 x 203.2 mm cylinder specimens resulted in an average compressive strength of 51.5 MPa,
47.7 MPa, 66.1 MPa, and 31.4 MPa for HFA-Column, HFA-Tube, M45-Tube, and HFA-Tube-
SEA and HFA-Tube-PSEA (four separate batches), respectively. In the testing process, the
ultimate compressive strength was measured in all the tested specimens while the stress-strain
diagram was captured in some of them.
Figure 3.13: Results of uniaxial compression test on ECC cylinder specimens: (A) HFA-Column, (B)
HFA-Tube, (C) HFA-Tube-SEA and HFA-Tube-PSEA.
48
It should be noted that the difference between the uniaxial compression test results is attributed to
the testing age since the cylinder specimens were tested at the time of testing the corresponding
bridge column. Even though the same ECC mix design was used for all the specimens, in the case
of M45-Tube, the testing ages were different. The testing date for the specimens of HFA-Column,
HFA-Tube, M45-Tube, and HFA-Tube-SEA and HFA-Tube-PSEA were 26, 48, 54, and 40 days,
respectively. Additionally, considerably large voids were observed in the specimens of HFA-Tube-
SEA and HFA-Tube-PSEA, due to low workability of the mixture, which could lead to the low
compressive strength of the ECC mixture.
To obtain the tensile properties of these ECC mixtures, dog-bone shaped specimens with 30 x 12.7
mm of cross-sectional dimensions in gage length (see Figure 3.14 (A)) were prepared from the
same batch used to construct the columns and tested in uniaxial tension. A special test setup was
designed to eliminate any unintentional shears or moments that may develop in the specimens due
to misalignment, see Figure 3.14 (B). Having a pure uniaxial testing condition is of great
significance for tensile testing of cementitious materials due to their brittle behavior in tension.
Prior to testing, these specimens were removed from the molds 24 hours after pouring and
submerged in water for 28 days (for curing). The direct uniaxial tensile test was carried out in
displacement control with a strain rate of 0.001 min
-1
. In order to measure the tensile strains, a
high precision linear potentiometer with a gage length of 140 mm was attached to either side of
the specimens. Figure 3.15 presents the results of the direct tensile test on dog-bone specimens.
49
Figure 3.14: (A) Dimensions of the dog-bone specimens, (B) tensile test setup.
Figure 3.15: Results of the direst tensile test on the dog-bone ECC specimens(different curves are
different tests on the same material).
50
The tensile strength of the mixes stood at approximately 3 MPa, with an ultimate tensile strain of
0.3-0.5%. HFA and M45 mixes demonstrated a very similar performance in the direct tensile tests.
3.6.3 Reinforcing steel
In order to obtain the mechanical properties of the steel rebar, monotonic direct uniaxial tension
test was performed on the three specimens of each diameter (9.5 mm and 12.7 mm). Figure 3.16
displays the setup for the direct tensile test. It entailed the use of 490 kN MTS loading frame with
hydraulic grips to conduct the direct uniaxial tensile tests. These specimens were loaded in
displacement control with two different displacement rates of 0.254 mm min
-1
and 2.54 mm min
-1
in the elastic and plastic regions, respectively, in accordance with ASTM A1034/A1034M [71].
An extensometer with a 50.8 mm gage length was used to measure the deformations continuously
during the test.
51
Figure 3.16: Uniaxial direct tensile test setup for reinforcing steel.
Figure 3.17 and Figure 3.18 show the stress-strain diagrams of 9.5 mm (#3) and 12.7 mm (#4)
reinforcing steel bars from uniaxial tensile tests, respectively. Table 3.6 summarizes the key
parameters of the reinforcing steel obtained from the test results.
52
Figure 3.17: Stress-strain diagrams of 9.5 mm (#3) reinforcing steel.
Figure 3.18: Stress-strain diagrams of 12.7 mm (#4) reinforcing steel.
0 2 4 6 8 10 12 14 16 18 20
0
100
200
300
400
500
600
700
800
Strain (%)
Stress (MPa)
0 2 4 6 8 10 12 14 16 18 20 22
0
100
200
300
400
500
600
700
800
Strain (%)
Stress (MPa)
53
Table 3.6: Mechanical properties of reinforcing steel.
Parameter 9.5 mm (US #3) rebar 12.7 mm (US #4) rebar
Elastic modulus (GPa) 179.3 190.9
Yield stress (MPa) 422.9 423.9
Ultimate stress (MPa) 677.6 677.5
Ultimate elongation (%) 17.9 17.7
3.6.4 Cu-Al-Mn SEA bars
During the initial cycles, Cu-Al-Mn SEA bars show an increasing residual strain and a downward
shift of stress-strain hysteresis loops induced by microstructural slips, which can be removed
through a process called training. After machining the alloys (see Figure 3.19(A)), the Cu-Al-Mn
SEA bars were subjected to three tensile strain cycles before being placed inside the columns to
decrease the effect of initial microstructural slips further to properly characterize their behavior.
The Cu-Al-Mn SEA bars were loaded in displacement control with the same 490 kN MTS loading
frame up to around 3% strain before being unloaded in load control mode to zero load. This process
was repeated as many as three times. The deformation was measured in a 203.2 mm gage length
(see Figure 3.19 (B)). An extensometer with larger gage length was used in testing of the Cu-Al-
Mn SEA bars, compared to testing of the reinforcing steel bars, since they were debonded in the
tested columns and deformations expected to occur through the length with reduced section, see
Figure 3.2. To protect the threads at both ends of the specimens, they were gripped through the
custom-made mechanical couplers inside of the hydraulic grips.
54
Figure 3.19: (A) Machined Cu-Al-Mn SEA bars, (B) test setup for training Cu-Al-Mn SEA bars.
Figure 3.20 exhibits the results from the training of the Cu-Al-Mn SEA bars. Unfortunately, the
data for two of the Cu-Al-Mn SEA bars was unpresentable due to slip of the extensometer. The
average elastic modulus, post-yield stiffness, lower transformation stress and upper transformation
stress obtained from the training of Cu-Al-Mn SEA bars were 33.9 GPa, 2.1 GPa, 210 MPa, and
259.3 MPa, respectively.
55
Figure 3.20: Training of Cu-Al-Mn SEA bars.
56
In order to investigate the stress-strain behavior of Cu-Al-Mn SEA bars up to failure, the bars were
removed from the columns following the structural tests and tested in uniaxial tension, similar to
the training procedure that was mentioned previously. Figure 3.21(A) demonstrates the procedure
for removing the bars. Figure 3.21(B) provides a schematic that displays the locations, numbering
and rupture points of the Cu-Al-Mn SEA bars. Figure 3.21(C) shows the bars retrieved from the
columns. The ruptured bars are separately shown in Figure 3.21(D).
Figure 3.21: (A) Extracting Cu-Al-Mn SEA bars from the columns, (B) locations and numbering of Cu-
Al-Mn SEA bars, (C) extracted Cu-Al-Mn SEA bars (not all bars are shown), and (D) Cu-Al-Mn SEA bars
ruptured during testing.
57
The Cu-Al-Mn SEA bars that have not ruptured during the column tests were first cycled as many
as three times up to approximately 3% of the strain (similar to the training procedure).
Subsequently, the amplitude of tensile strain cycles was gradually increased in increments of 1%
and two cycles were conducted at each strain level. A displacement rate of 6.35 mm/min
(0.035/min strain rate) was used for the purpose of loading. Unloading to zero force, in force
control, was completed within one minute. One of the Cu-Al-Mn SEA bars, [bar 2 according to
Figure 3.22(B)], was tested up to failure with monotonically increasing strain after the first three
cycles of 3% strain in order to make comparisons between the cyclic and monotonic behavior, see
Figure 3.22(A) for results of bar 2, which ruptured at around the 9% strain. Bar 6 exhibited a
similar rupture strain after being subjected to increasing cycles of tensile strains; however, it was
observed that the recovery strain of this bar was in the range of 7%, and permanent deformations
were observed past this strain level, as can be seen in Figure 3.22(B). Bars 1 and 3 failed at around
5% strain before reaching the recovery strain limit, see Figure 3.22(C and D). It is important to
note that all these bars were subjected to prior tensile and compressive strains during the column
tests. One end of the Bar 7 was damaged during the retrieval from the columns and could not be
tested under uniaxial tension as against those that were. Table 3.7 summarizes the results from
uniaxial cyclic tensile tests conducted on the removed bars from the tested columns.
58
Figure 3.22:Cyclic stress-strain behavior of Cu-Al-Mn SEA bars: (A) Bar 2, (B) Bar 6, (C), Bar 1, and
(D) Bar 3. Test were conducted some posteriori to column tests.
Table 3.7: Results of uniaxial cyclic test on removed SEA bars from tested columns.
Bar #
Elastic
modulus (GPa)
Post-yield
stiffness (GPa)
Ultimate stress
(MPa)
Ultimate strain
(%)
2 35.1 1.1 356.3 9.56
6 37.3 1.6 385.6 9.46
1 40.3 1.8 356.5 5.40
3 54.9 1.8 418.7 5.72
59
3.7 Results and Discussion
3.7.1 Crack distribution
Figure 3.23 demonstrates the condition of the specimens and the crack distribution at the various
stages of testing. A comparison of the observed crack patterns revealed that significantly fewer
cracks occurred in the specimens with Cu-Al-Mn SEA bars due to unbonded plastic hinge
reinforcement. Localized, large cracks were observed at drift levels of 5% and higher. This was a
desired characteristic to reduce the area required to be patched following an earthquake so as to
protect the steel transverse reinforcement from corrosion. Otherwise, the columns did not require
any post-earthquake repair. The concrete filled RECC hollow sections demonstrated finer and
more distributed cracks as compared to others. When no Cu-Al-Mn SEA bars were used in the
plastic hinge region, it was seen that ECC is capable of limiting the widths of cracks that develop
during an earthquake, which is desirable for the durability of the columns. The HFA-Column
specimen entirely comprised of RECC showed a similar behavior to that of HFA-tube specimen
in terms of cracking and damage distribution. This interesting finding indicated that building the
entire cross-section using ECC, which is a relatively expensive material, has limited effect on the
sustained damage. The control specimen, RC-Column, exhibited the highest level of damage at all
drift levels. In addition, concrete spalling was observed only in the RC-Column starting at around
5% drift. The responses of all the specimens were flexural dominated and no crushing of concrete
or ECC was observed during the testing.
60
Figure 3.23: Observed damage in test specimens at various drift levels.
61
3.7.2 Load-drift hysteresis curves
Figure 3.24 presents the load-drift hysteresis curves for all the specimens.
Figure 3.24: Load-drift hysteresis curves: (A) RC-Column, (B) HFA-Column, (C) HFA-Tube, (D) M45-
Tube, (E) HFA-Tube-SEA, and (F) HFA-Tube-PSEA. In (E) and (F), circles mark the Cu-Al-Mn SEA bars
ruptured during testing.
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
(A) (B)
(C)
(E)
(D)
(F)
62
A very similar hysteretic behavior was observed for RC-Column, HFA-Column, HFA-Tube and
M45-Tube specimens. The maximum lateral force developed in the last three specimens using
ECC was similar and slightly higher as compared to the control column. No buckling or rupture
of steel rebar was observed in these four specimens. Rupture of one Cu-Al-Mn SEA bar was
observed at 7.0% drift followed by the rupture of another one at 7.25% drift in the HFA-Tube-
SEA specimen. Similarly, one Cu-Al-Mn SEA bar was ruptured at 5% drift within the HFA-Tube-
PSEA specimen. The bar ruptures have resulted in a sudden decline in the load carrying capacity
of these columns. Since rupture of the SEA bars occurred at large drift ratios and after the peak
load, it can be related to the material performance. The ruptured Cu-Al-Mn SEA bars were located
at the maximum distance from the neutral axis where the maximum tensile strains appear.
Additionally, cyclic tensile test of the removed bars, removed after testing the columns, indicated
that the implemented Cu-Al-Mn bars were unable to tolerate strains over 10%. This behavior could
be attributed to grain size-to-bar diameter ratio as discussed in Section 2.2.
Owing to the lower yield strength of the Cu-Al-Mn SEA bars, a lower maximum lateral force was
observed for HFA-Tube-SEA and HFA-Tube-PSEA specimens. With regard to the rupture of Cu-
Al-Mn SEA bars in the HFA-Tube-SEA specimen, the desired behavior of almost zero permanent
drift was achieved (it must be noted that the unloading curves pass through origin). Obviously, this
has resulted in a reduced area within the hysteresis loops, which is equal to the energy absorption
of the columns. This trait was improved if plastic hinge steel rebar was only partially replaced with
Cu-Al-Mn SEA bars in the HFA-Tube-PSEA column. Some energy dissipation was also achieved
by yielding the steel rebar in expense of increased permanent deformations.
63
3.7.3 Quantitative comparison
To enable a quantitative comparison of the performance of the six columns, the initial stiffness,
yield, maximum and ultimate points, permanent deformation and energy absorption were defined,
as shown in Figure 3.25, and used as metrics. The initial stiffness, E0, was defined as the chord
stiffness between 40% and 70% of the peak load. The yield drift, δy, was assumed to occur at the
yield point of an equivalent elasto-plastic system with reduced stiffness, as the secant stiffness
stood at 75% of the maximum lateral load of the real system [72]. The yield force, Fy, was obtained
as the force on the force-drift envelope at the yield drift. The maximum point corresponded to the
maximum load carrying capacity of the columns, with δm and Fm being the corresponding drift and
force values, respectively. The ultimate point was found on the softening branch of the force-drift
curves at a reduction of 15% in load carrying capacity, with δu and Fu being the corresponding
drift and force values, respectively. These drifts and forces were found on the average of the
envelope curves in push and pull directions that are also shown in Figure 3.24 alongside the
hysteresis loops. The curve shown in Figure 3.25(A), as an example, belongs to the RC-Column
tested in this study. Additionally, the permanent drift, δp, and energy absorption, Ea, were defined
using the hysteretic response of these columns. Energy absorption was considered equal to the area
inside the hysteresis loops up to the completion of drift cycles with increasing peak values. The
permanent drift was defined as the average of the absolute values of drifts corresponding to zero
force on the unloading curves after the attainment of the drift peaks in the push and pull directions.
As an example, the definitions for energy absorption and permanent drift corresponding to a peak
drift of 7% are shown in Figure 3.25(B).
Table 3.8 displays the results of this quantitative evaluation. It must be noted that the values for
permanent drift and energy absorption in Table 3.8 pertain to a peak drift of 7% meaning that
64
average of the absolute values of permanent deformations in push and pull directions and the area
inside the hysteresis curves up to completion of 7% drift were used as a measure to compare the
results in terms of permanent deformation and energy absorption capacity, respectively. The values
inside the parenthesis of each cell indicate the differences (in percentage) with respect to the RC-
Column (control specimen). A negative value within the parenthesis means a reduction while a
positive value implies an increase.
Figure 3.25: Definition of (A) initial stiffness, E 0, yield drift, δ y, yield strength, V y, maximum drift, δ m,
maximum strength, V m, ultimate drift, δ u, and ultimate strength, V u, and (B) permanent drift, δ p, and
energy absorption, E a. In presentation of the permanent drift and energy absorption definitions, 7%
maximum experienced drift is used for comparison purposes. Permanent drift is obtained as the average
of drift in positive and negative directions, only positive direction permanent drift is shown for clarity in
presentation.
It can be seen in Table 3.8 that the columns using ECC had a lower stiffness as compared to the
control column. The stiffness reduction was higher for HFA-Tube-SEA and HFA-Tube-PSEA
specimens (approximately 30% less when compared to RC-Column) exacerbated by the lower
stiffness of the Cu-Al-Mn SEA bars. Owing to the lower stiffness of ECC and Cu-Al-Mn SEA
bars and increased deformability, the yield drift increased in comparison to RC-Column for all the
specimens ranging from approximately 30-70%. The HFA-Column had an approximately 16%
0 2 4 6 8 10
0
5
10
15
20
25
30
Drift (%)
Force (kN)
Maximum
15%
-10 -8 -6 -4 -2 0 2 4 6 8 10
-40
-30
-20
-10
0
10
20
30
40
Drift (%)
Force (kN)
Yield
Ultimate
Energy
Absorption, E
a
Permanent
Drift
7% Drift
δ
y
δ
m
δ
u
V
u
V
m
V
y
E
0
Initial
Stiffness
(A) (B)
δ
p
65
higher strength when compared with RC-Column, while this value was slightly less
(approximately 12%) for the HFA-Tube and M45-Tube specimens. Thus, it can be concluded in
terms of peak load carrying capacity, making the entire cross-section from ECC results in a
negligible increase in strength compared to only using the ECC partially, which is the proposed
approach here.
66
Table 3.8: Quantitative comparison of the performance of test columns. Values in parenthesis represent the differences in percentage from the
control RC-Column. Negative values indicate a decrease and vice versa. See Figure 3.25 for definitions of the parameters in the first column of
the table. Permanent drift and energy absorption values pertain to a maximum experienced drift of 7%.
Specimen RC-Column HFA-Column HFA-Tube M45-Tube HFA-Tube-SEA HFA-Tube-PSEA
Initial Stiffness, E 0 (kN/m) 1288 (0.00) 1129 (-12.38) 949 (-26.32) 1030 (-20.01) 865 (-32.88) 853 (-33.81)
Drift at Yield, δ y (%) 1.75 (0.00) 2.30 (32.00) 2.94 (68.26) 2.53 (44.93) 2.28 (30.47) 2.50 (43.41)
Yield Force, F y (kN) 24.38 (0.00) 27.95 (14.65) 27.20 (11.55) 27.45 (12.58) 20.10 (-17.55) 22.67 (-7.03)
Drift at Maximum, δ m (%) 2.97 (0.00) 3.99 (34.51) 4.01 (35.12) 3.97 (33.99) 4.00 (34.93) 3.73 (25.83)
Maximum Force, F m (kN) 27.35 (0.00) 31.76 (16.12) 30.26 (10.64) 30.95 (13.18) 22.83 (-16.52) 24.94 (-8.79)
Drift at Ultimate, δ u (%) 7.09 (0.00) 7.15 (0.96) 7.15 (0.85) 7.11 (0.42) 7.01 (-1.13) 4.99 (-29.59)
Ultimate Force, F u (kN) 23.25 (0.00) 26.99 (16.12) 25.72 (10.64) 26.31 (13.18) 19.41 (-16.52) 21.20 (-8.79)
Permanent Drift, δ p (%) 4.25 (0.00) 3.94 (-7.25) 3.63 (-14.62) 3.94 (-7.33) 0.37 (-91.18) 2.84 (-33.05)
Energy Absorption, E a (kN-m) 9.51 (0.00) 9.67 (1.67) 8.36 (-12.12) 8.93 (-6.17) 3.43 (-63.97) 5.65 (-40.63)
67
The relatively low strength gain achieved using ECC in HFA-Column, HFA-Tube and M45-Tube
specimens was attributed to the flexural dominant behavior of the specimens, wherein the flexural
strength was dominated by the tensile force in the steel rebar or Cu-Al-Mn SEA bars. The
maximum force was reduced by about 18% and 7% for the HFA-Tube-SEA and HFA-Tube-PSEA
specimens, respectively, as compared to RC-Column due to lower yield (upper transformation)
stress of the Cu-Al-Mn SEA bars. The observations for drift at maximum and force at yield were
similar to those for drift at yield and force at maximum, respectively. With regard to the ultimate
point, similar drift levels were observed for all the specimens, except for HFA-Tube-PSEA, which
witnessed a sudden reduction in the load carrying capacity due to the rupture of the Cu-Al-Mn
SEA bars. Force levels were higher at the ultimate point for HFA-Column, HFA-Tube and M45-
Tube specimens, and slightly lower for the two columns with Cu-Al-Mn SEA bars. In terms of the
permanent drift following the 7% drift peak, a slight reduction of 7-14% was observed for the
specimens using ECC, but not the Cu-Al-Mn SEA bars. The HFA-Tube-SEA column resulted in
a remarkable reduction of 91% (from 4.25% to 0.37% drift) in comparison with the control RC-
Column. HFA-Tube-PSEA had also reduced the permanent drift to 2.84%, which corresponded to
a reduction of about 33% as compared to the RC-Column. Figure 3.26 illustrates the permanent
drift as a function of maximum experienced drift. As seen in this figure, the permanent drift for
the HFA-Tube-SEA specimen showed an approximately linear increase with a maximum
experienced drift, which was distinct from the remaining specimens. The energy absorption of up
to 7% drift peak was similar in all specimens, except for those that used Cu-Al-Mn SEA bars.
Considering the performance of the tested bridge columns up to 7% lateral drift, the HFA-Tube-
SEA absorbed about 64% less energy due to the superelastic nature of the plastic hinge
reinforcement; this improved to 41% lesser energy absorption in the case of HFA-Tube-PSEA
68
specimen due to steel plastic hinge reinforcement; albeit at the expense of permanent drift. Figure
3.27 shows the energy absorption as a function of maximum experienced drift.
Figure 3.26: Permanent drift as a function of maximum experienced drift.
Figure 3.27: energy absorption as a function of maximum experienced drift.
1 2 3 4 5 6 7
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Drift (%)
Permanent Drift (%)
RC-Column
HFA-Column
HFA-Tube
M45-Tube
HFA-Tube-SEA
HFA-Tube-PSEA
1 2 3 4 5 6 7
0
1
2
3
4
5
6
7
8
9
10
Drift (%)
Energy Absorption (kN-m)
RC-Column
HFA-Column
HFA-Tube
M45-Tube
HFA-Tube-SEA
HFA-Tube-PSEA
69
CHAPTER 4 - NUMERICAL STUDY OF TESTED BRIDGE COLUMNS
Experimental studies are generally time-consuming and costly since the process entails purchasing
and preparing materials, sensors and equipment. Additionally, the limitations in the equipment
capacity, lab, budget, and time usually imposes a scaling of the specimens, reducing the number
of tests, and neglecting some of the necessary variables in experimental work. Furthermore, the
challenges in controlling material quality, reaching specific mechanical behavior, and providing
the desired testing environment and loading scenarios are other obstacles of experimental work.
Correspondingly, the numerical methods present a promising approach to overcome these
limitations. However, the numerical models should be validated to ensure an accuracy of the
models in producing reliable data. Modeling an experimental work and verifying the numerical
results with the corresponding experimental data is a legitimate approach before investigating
configurations that fall outside of the experimental data. In this section, the tested bridge columns
were modeled and analyzed numerically to obtain a satisfactory finite element modeling (FEM)
approach. Since material properties and response of HFA-Column and M45-Column under lateral
cyclic loading were almost identical, M45-Column was excluded from the numerical study section.
Subsequently, in Chapter 5 the developed FEM method is employed to conduct a parametric study
to determine the optimal column design with ECC and Cu-Al-Mn SEA bars. ATENA finite
element package [73] was used to conduct the numerical studies in this thesis. ATENA [73] was
developed specifically for RC structures and supports the modern FRC compositions such as ECC,
which makes it an ideal choice for this study.
70
4.1. Material Constitutive Models
4.1.1 Concrete
The CC3DNonLinCementitious2 material model from ATENA [73] material library was
employed to simulate the inelastic cyclic behavior of concrete in the columns. Figure 4.1 illustrates
the general behavior of CC3DNonLinCementitious2 material. In this material, fracture and
plasticity models are combined, which can also be formulated separately. This unique property
allows the material to handle a situation when the failure surfaces of the both models are engaged,
thus allowing a concurrent simulation of the concrete crushing, cracking, and crack closure [74].
Figure 4.1:General behavior CC3DNonLinCementitious2 of material.
Separate concrete material models were defined for the RC column and the core concrete in RECC
specimens. Since the software takes the confinement effect into account automatically, the
constitutive model was defined based on the compression test results, which is the main input of
the material model. The CC3DNonLinCementitious2 material model takes the compressive
strength from cube specimens,
&'
, and generates default values for the other parameters.
(
(
)
*
/ℓ
-
(
71
Equation (5) presents the relationship between the input value and the compressive strength of
cylinder specimens, fc.
=
&'
0.85
(5)
A single eight node brick element, which was subsequently subjected to cyclic incremental
displacements, see Figure 4.2, was considered to investigate the cyclic behavior of the concrete
material. The brick element was restrained at four nodes (1
2
to 1
3
) and subjected to uniaxial cyclic
displacements at the other four nodes.
Figure 4.2: Single brick element.
Figure 4.3 displays an average stress-strain diagram from the single brick element model while
distinguishing the last loading cycle with ABCDEF pass. In this diagram, the average strain was
calculated by dividing the applied displacement, , to dimension of the element, 4 , while the
average stress was calculated through dividing summation of the reactions in the restrained nodes
to the cross-sectional area, 5. The single brick element revealed that the
72
CC3DNonLinCementitious2 material model shows convergence problems under large tensile
strains where changing load direction leads to an almost zero stiffness in the stress-strain diagrams
(line BC in Figure 4.3 (A)). It was found that this behavior can be improved by assigning larger
values to critical compressive displacement, )
*
, and tension stiffening, -
, parameters in the
material model (see Figure 4.1 for definition of )
*
and -
).
Figure 4.3: Average stress-strain behavior of concrete brick model: (A) tension and compression, (B)
tension.
Table 4.1 summarizes the corresponding values for different parameters in the
CC3DNonLinCementitious2 material model. These values were incorporated to simulate behavior
of concrete in the numerical model, i. e., to model RC-Column, foundations, top caps, and concrete
cores.
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2
-50
-40
-30
-20
-10
0
Stress (MPa)
Strain (%)
0.0 0.1 0.2
0
1
2
3
Stress (MPa)
Strain (%)
(A) (B)
6 7 -
8
!
9
73
Table 4.1: Material constitutive model properties for concrete to model RC sections and concrete core.
Section Material Property
Value
RC-
Column
& Caps
HFA-
Tube
HFA-
Tube-
SEA
HFA-
Tube-
PSEA
Basic
Elastic modulus, !
(MPa) 38790 32440 36380 34560
Poisson’s ratio,: 0.2
Compressive strength,
(MPa) 49.1 30.1 40.7 35.4
Tensile strength,
(MPa) 3.57 2.59 3.17 2.88
Tensile
Specific fracture energy, ;
<
(MN/m) 8.96E-05 6.47E-05 7.91E-05 7.21E-05
Tension stiffening, -
0.1
Unloading factor
0.0
Compressive
Critical compressive displacement, )
*
(m) -0.0025 -0.0025 -0.0055 -0.0055
Plastic strain at compressive strength, (
-1.27 E-3 -9.28E-4 -1.12E-3 -1.03E-3
Reduction of strength due to cracks, =
,>
0.8
Shear
Crack shear stiffness factor, ?
<
200
Aggregate size in aggregate interlock, @-9
Unchecked
Miscellaneous
Failure surface eccentricity 0.52
Multiplier for the specific flow direction, A 0.0
Specific material weight, B
CD
E
F
0.023
Fixed crack model coefficient 1.0
74
4.1.2 Reinforcing steel
The “Cyclic Reinforcement” material from ATENA [73] material library was used in a bilinear
form to simulate the behavior of both longitudinal and transverse reinforcement. The material
model is applicable to one-dimensional elements and considers a well-recognized uniaxial
constitutive hysteretic model, nonlinear model of Menegotto and Pinto, and Bauschinger effect
[74]. Since no material test was conducted on the transverse reinforcements, same material
properties from the direct tensile test of the longitudinal reinforcement was considered for the
transverse reinforcement. Moreover, suggested default values were applied to the parameters for
the nonlinear model of Menegotto and Pinto, and Bauschinger effect. Table 4.2 summarizes the
corresponding values for different parameters in “Cyclic Reinforcement” material model.
Table 4.2: Material constitutive model properties for reinforcing steel.
Section
Material Property Value
Basic
Elastic modulus, !
(MPA) 180,000
Yield stress, σ
H
(MPa)
423
M-P
Bauschinger effect exponent, R 20
Menegotto-Pinto parameter, -
2
18.5
Menegotto-Pinto parameter, -
I
0.15
To achieve a better understanding of the behavior of “Cyclic Reinforcement” material, a single bar
was modeled with the material properties as given in Table 4.2 and subjected to incremental
uniaxial cyclic loading. Figure 4.4 presents the developed single bar model and the corresponding
stress-strain diagram. In the numerical model, two elastic cuboids were considered at the top and
bottom of the single bar to apply cyclic loading and calculate the responses, respectively. The top
cuboid was restrained in two perpendicular directions to the load direction while the bottom cuboid
was restrained in all three directions. The cyclic loading was applied in displacement control to
the top cuboid and the average stress and strain values were calculated utilizing the reactions at
75
the bottom cuboid and the applied displacement to the top cuboid, respectively. It should be
mentioned here that the modeled bar would not buckle in compression since a truss element
without any node in the middle portion was used.
Figure 4.4: One dimensional model for “Cyclic Reinforcement” material: (A) developed single bar
model, (B)Stress-strain diagram.
The diagram clearly presents that the material constitutive model is in accordance with the defined
elastic-perfectly plastic behavior.
4.1.3 ECC
The behavior of ECC with PVA fibers was simulated using the CC3DNonLinCementitious2 and
“Cyclic Reinforcement” materials. The ductile behavior of ECC in tension is captured using
smeared reinforcement, defined by “Cyclic Reinforcement” material, in a matrix of
76
CC3DNonLinCementitious2 material. Further, to determine the material properties, the dog-bone
specimens were modeled and subjected to a monotonic tensile load in displacement control as in
the testing procedure. Figure 4.5 illustrates the modeled dog-bone specimen. The material
properties were considered to be satisfactory when the average stress-strain diagram resulted from
the gage length of the numerical model (see Figure 4.5) was close to the averaged stress-strain
diagram of the tested specimens. It was found that the concrete tensile strength and specific fracture
energy, the ratio of smeared reinforcement and its direction, yield and ultimate stresses of the
smeared reinforcement, and reinforcement ultimate strain capacity mainly define the behavior of
ECC in this approach. In the numerical model, 0.8% smeared reinforcement was applied in six
directions, defined by perpendicular vectors to six planes in the global coordinates, as listed in
Table 4.3.
77
Figure 4.5: Modeled dog-bone specimen.
Table 4.3: Vectors defining direction of smeared reinforcement in ECC.
Direction
Vector
X Y Z
1 1 0 0
2 0 1 0
3 0 0 0
4 1 1 0
5 1 0 1
6 0 1 1
Multilinear form of “Cyclic Reinforcement” material was employed to capture strain-hardening
and strain-softening branches in ECC stress-strain diagram. Figure 4.6 demonstrates the
multilinear form and the intended parameters to define the material behavior.
78
Figure 4.6: Multilinear form of “Cyclic Reinforcement” material.
Table 4.4 summarizes the corresponding values for the defined parameters in Figure 4.6.
Additionally, CC3DNonLinCementitious2 material constitutive model properties used in
simulating the behavior of ECC are summarized in Table 4.5. It should be noted that for each
bridge column, the ECC constitutive model properties were determined individually to match the
compressive properties from the material tests while returning similar tensile properties from the
dog-bone numerical model. Figure 4.7 shows the resulting tensile stress-strain diagrams from the
dog-bone model using the provided values for the material constitutive model parameters.
Evidently, the proposed approach to simulate the behavior of ECC can capture tensile strain-
hardening and strain-softening branches as observed in the experiments. Additionally, crack
opening and localization stages were observed in the modeled dog-bone specimen with the
utilization of the proposed approach (see Figure 4.8).
To ensure proper behavior of the proposed method under cyclic loading, single ECC brick element
was modeled with the method explained previously for concrete in Section 4.1.1. Figure 4.9
presents the cyclic response of the single ECC cube. The results illustrate that the developed ECC
2
(
I
J
(
2
(
I
(
J
79
model captures tensile strain-hardening behavior under cyclic loads and shows stable cycle
behavior under both tension and compression conditions. Moreover, the ultimate strength of ECC
slightly increases compared to the base material when smeared reinforcement is added. This
phenomenon was considered in defining the material constitutive model parameters of the plain
concrete to match the ultimate compressive strength of ECC with the corresponding values from
compression test of the cylinders.
Table 4.4: Material constitutive model properties for smeared reinforcement in ECC.
Section Material Property
Value
HFA-
Column
HFA-
Tube
HFA-
Tube-SEA
& PSEA
Basic
ε
2
(%) 0.075
ε
I
(%) 1.10 1.20 1.10
ε
J
(%) 1.11 1.21 1.11
σ
2
(MPa) 150
σ
I
(MPa) 170
σ
J
(MPa) 0
Active in compression Unchecked
M-P
Bauschinger effect exponent, R 20
Menegotto-Pinto parameter, -
2
18.5
Menegotto-Pinto parameter, -
I
0.15
Miscellaneous Specific material weight, B
CD
E
F
0.0785
80
Table 4.5: Material constitutive model properties for concrete, CC3DNonLinCementitious2, in ECC.
Section Material Property
Value
HFA-
Column
HFA-Tube
HFA-
Tube-SEA
& PSEA
Basic
Elastic modulus, !
(MPa) 39390 36480 33010
Poisson’s ratio,: 0.2
Compressive strength,
(MPa) 47.16 41.00 31.4
Tensile strength,
(MPa) 2.5 2.5 3.5
Tensile
Specific fracture energy, ;
<
(MN/m) 8.8E-5 9.53E-5 9.53E-5
Tension stiffening, -
0.4
Unloading factor 0
Compressive
Critical compressive displacement, )
*
(m) -0.0035 -0.0055 -0.011
Plastic strain at compressive strength, (
-1.173E-03 -1.124E-03 -9.519E-04
Reduction of strength due to cracks, =
,>
0.8
Shear
Crack shear stiffness factor, ?
<
200
Aggregate size in aggregate interlock, @-9 Unchecked
Miscellaneous
Failure surface eccentricity 0.52
Multiplier for the specific flow direction, A 0.0
Specific material weight, B
CD
E
F
0.023
Fixed crack model coefficient 1.0
81
Figure 4.7: Stress-strain diagrams from modeled dog-bone specimen.
Figure 4.8: Different cracking stages: (A) initial cracking, (B) crack opening, (C) crack localization.
0.0 0.1 0.2 0.3 0.4 0.5 0.6
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Averaged Experimental Results
HFA-Column
HFA-Tube
HFA-Tube-SEA & SPEA
Stress (MPa)
Strain (%)
82
Figure 4.9: ECC stress-strain diagram in tension and compression: (A) HFA-Column, (B) HFA-Tube, (C)
HFA-Tube-SEA & HFA-Tube PSEA.
4.1.4 Cu-Al-Mn SEA bars
Since the material library of ATENA [73] does not include a proper constitutive model to simulate
behavior of SEA bars, a multilinear one-dimensional constitutive model was implemented in
ATENA FEM package [73] based on the work of Motahari et al. [75]. Figure 4.10 illustrates the
input parameters for SEA constitutive model. Martensite phase finish temperature, @
, martensite
phase start temperature, @
, austenite phase start temperature 6", austenite phase finish
temperature, 6
, operating temperature, L, slope of austenite to martensite transformation, -
M
, and
83
martensite to austenite transformation, -
N
, in critical stress-temperature diagram, full austenite
elastic modulus, !
N
, full martensite elastic modulus, !
M
, and maximum residual strain of the
material in temperatures lower than 6
, (
O
, are inputs to the model to generate stress-strain
behavior of SEA.
Figure 4.10: Parameters of the SEA model.
M
M
N
N
(
M
(
N
(
MP
(
NP
(
!
M
!
N
Q
L
L
@
@
6
6
-
N
-
M
(
O
84
Figure 4.11 shows the schematic cyclic behavior of the SEA model.
Figure 4.11: Schematic cyclic behavior of the constitutive model for SEA.
Since the actual values for the parameters were unavailable, a single bar element was developed
numerically and the SEA constitutive model was assigned. To obtain the material properties of the
numerical model, a match between the numerical response and the experimental results from a
direct cyclic tensile test of Cu-Al-Mn SEA bars was sought. The corresponding values for
parameters of the constitutive model were obtained for several tested SEA bars presented in
Section 3.6.4 and then averaged for subsequent use in the numerical models. Table 4.6 summarizes
the averaged values of the parameters of the SEA constitutive model and Figure 4.12 provides a
comparison of the experimental and numerical results for one of the SEA bar tests. As evident, the
implemented SEA constitutive model can capture the superelastic behavior of Cu-Al-Mn SEA bars
through repeated cyclic loads.
M
M
N
N
(
M
(
&
(
RS
(
N
(
MP
(′
MP
(
NP
(′
NP
6′
6
7
-
8
7′
-′
8′
85
Table 4.6: Material constitutive model properties for SEA.
Parameter Value Parameter Value
@
(
o
C) 11.54 !
N
(MPa) 35,800
@
(
o
C) 3.43 !
M
(MPa) 12,500
A
U
(
o
C) 11.14 C
W
22.10
A
X
(
o
C) 15.43 C
C
20.62
T (
o
C) 25.0 ε
Z
0.05
Figure 4.12: Comparing experimental and numerical results for SEA.
4.1.5 Bond characteristics
“Memory Bond” material with the bond model by Bigaj [76] was used to simulate the bond
behavior of rebar to the surrounding cementitious matrix (concrete or ECC). The bond material
captures the behavior under cyclic loads. The model generates a new unloading pass instead of
following the loading pass based on a parameter ([
2
) as defined in the material constitutive model
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
0
50
100
150
200
250
300
350
Strain (%)
Stress (MPa)
SMA Material Model
Experimental Results
86
[74], as seen in Figure 4.13. The bond model was developed based on the diameter of the
reinforcing bar, the compressive strength of surrounding concrete, and the bond quality.
Figure 4.13: General behavior of “Memory Bond” material.
It was found that the bond properties minimally affect the response of the bridge columns.
Therefore, constant bond properties were used for all the numerical models in alignment with the
compressive strength of the concrete in RC-Column specimen with good bond quality for the main
longitudinal reinforcement (#3 rebar).
Since applying completely debonded rebar is not possible for one dimensional steel elements in
ATENA [73], significantly weak bond properties were used. Figure 4.14 shows and compares the
bond-slip diagrams of regular and weak bonds. [
2
was equal to 4.42 MPa and 0.48 MPa in the
bond model for the longitudinal reinforcement and the SEA bars, respectively.
?
[
[
2
−[
2
87
Figure 4.14: Bond models for the longitudinal reinforcement.
Additionally, perfect bond condition was considered for the transverse reinforcement since
applying bond to the stirrups altered the final results insignificantly while led to convergence
problems in the numerical models.
4.2. Geometry of the Numerical Models
The geometry of the numerical models was specifically created to a degree that ensured an accurate
representation of the actual tested specimens. Figure 4.15 shows the geometry of the numerical
models. RC-Column and HFA-Column were modeled in three separate parts: (i) the top cap, (ii)
the foundation, and (iii) the column, which had perfect contacts with each other at the
corresponding interaction (common) surfaces. The limitation of ATENA [73] in generating
cylindrical geometry, required the creation of circular columns using multi-sided column option
with 12 sides. Furthermore, in HFA-Tube, HFA-Tube-SEA, and HFA-Tube-PSEA specimens, the
RECC hollow sections were created using a multi-sided opening in the multi-sided column. This
was followed by a separate creation of the concrete core, which was then connected to the RECC
0 2 4 6 8 10 12 14 16
0
2
4
6
8
10
12
14
Bond Stress (MPa)
Slip (mm)
Steel Bars
SEA Bars
88
hollow section assuming displacement compatibility at the shared nodes, i.e., perfect connection.
Assumption of the perfect connection between the concrete core and the RECC hollow section
was result of the construction process as well as lack of experimental evidence to indicate
otherwise. In HFA-Tube, HFA-Tube-SEA, and HFA-Tube-PSEA models, to a certain depth,
multi-sided openings were created at the end caps to embed the column as in the tested specimens.
Elastic steel plates were placed above the top cap and underneath of the foundation to distribute
the axial load uniformly and simulate the strong floor, respectively.
Figure 4.15: Generating geometry of the numerical models: (A) RC-Column and HFA-Column, (B) HFA-
Tube, HFA-Tube-SEA, HFA-Tube-PSEA.
89
Figure 4.16 illustrates the modeling approach for the longitudinal reinforcement. The longitudinal
reinforcement was modeled with one-dimensional truss elements in one segment. In HFA-Tube-
SEA and HFA-Tube-PSEA specimens, to simulate the debonding at the plastic hinge region, the
longitudinal bars except those in the neutral axis were divided into three segments to apply the
SEA material as well as the weak bond. Since the bond constitutive model in ATENA [73] is
unable to maintain continuity between two different bond models next to each other, the perfect
bond condition was assigned to the top and the bottom segments and the slip was restrained at the
ends of the middle segment.
The transverse reinforcements with perfect bond to the surrounding concrete were modeled by 36
segments to form a hexagonal shape close to a circle, due to non-feasibility of circular geometry
in ATENA [73]. For simplicity, the transverse reinforcement was modeled using individual hoops
instead of spiral reinforcement as used in the tested specimens yet ensuring equal transverse
reinforcement ratios. Additionally, the pitch between the hoops was doubled and the bar cross-
section was increased correspondingly to decrease the number of elements in the numerical models
in favor of the required memory and the computational cost. For the same reason, smeared
reinforcement was used in the foundation and the top cap instead of discrete reinforcing bars.
Specifically, 3% smeared reinforcement was applied to both the foundation and the top cap in three
main global coordinates (X, Y, and Z). Increasing the pitch while maintaining the lateral
reinforcement ratio is not an issue in flexural dominant elements as practiced in construction of
beams and columns.
90
Figure 4.16: Modeling longitudinal reinforcement :(A) normal steel reinforcement in RC-Column, HFA-
Colum, and HFA-Tube and neutral axis reinforcement of HFA-Tube-SEA and HFA-Tube-PSEA , (B)
debonded steel reinforcement in HFA-Tube-PSEA , (C) longitudinal reinforcement with Cu-Al-Mn SEA
bar in HFA-Tube-SEA and HFA-Tube-PSEA.
91
4.3. Boundary Conditions
Figure 4.17 shows the boundary condition and the analogous rigid body motion. To simulate the
effect of anchorage to the strong floor as well as the observed deformations in the test setup, two
elastic spring elements with 20 mm width and 500 mm length were placed at the two edges of the
foundation. The springs were placed perpendicular to the lateral loading direction to provide rigid
body motion and adjust the lateral stiffness. The center line of the foundation was fixed in three
directions resembling the interaction between the strong floor and the foundation. Additionally,
the out-of-plane deformations were restrained at the top cap to ensure in-plane deformation of the
column and prevent any torsional loads in the system.
Figure 4.17: (A) Boundary condition and (B) analogous rigid body motion.
92
4.4. Loading
Figure 4.18 illustrates the applied loads in the numerical models. The distributed axial load was
applied in two steps to the elastic plate above the top cap resulting in 50 kN in each step. The
lateral cyclic displacement load was applied at surface center points of the top elastic plate. A
0.1524 mm displacement was considered for each lateral loading step leading to a 0.0167% drift
ratio. Small loading steps are needed to overcome convergence issues in ATENA [73] when a RC
numerical model is subjected to cyclic loads. Additionally, the effect of shrinkage was simulated
by assuming an initial -0.0003 shrinkage strain to the column as a recommended approach to
consider shrinkage related issues in RC structural elements [74]. So called cross-sectional
approach is utilized in ATENA [73] to analyze shrinkage problems. In this method, an initial
stress-independent shrinkage strain is linearly added to the calculated strain of concrete in the
cross-section [74].
93
Figure 4.18: Loading: (A) Axial and lateral loading, (B) Shrinkage.
4.5. Meshing and Solution Parameters
Figure 4.19 presents the employed meshing element along with the meshed model. Linear
tetrahedral elements with four nodes and reduced integration points were used to mesh the whole
column, including the foundation and the top cap, and the top elastic plate. Brick elements were
used to mesh the bottom elastic plate. The maximum mesh size for the column, the end caps and
the top elastic plate, and the bottom elastic plate were 0.04 m, 0.1 m, and 0.07 m, respectively,
resulting in a finer mesh for the column.
94
Figure 4.19: Meshing: (A) linear tetrahedral element with four nodes, (B) meshed model.
The Newton-Raphson solution method with line search was used to analyze the numerical models.
The method updated tangent stiffness in each step up to 300 iterations. The objective of the line
search method is to optimize the required work for to reduce the out-of-balance forces by adjusting
the displacement increment towards. Furthermore, the Sloan Algorithm [77] was used to optimize
node numbering in generating a reduced global matrix, which lowers the required CPU and RAM
resources to analyze the numerical models [74]. Table 4.7 summarizes the error tolerances for the
Newton-Raphson solution method, parameters of the line search method, and the conditional break
criteria.
95
Table 4.7: Solution parameters.
Section Parameter Value
General
Displacement error tolerance 0.01
Residual error tolerance 0.01
Absolute residual error tolerance 0.01
Energy error tolerance 0.0001
Line search
method
Solution Method
With
iterations
Unbalanced energy limit 0.8
Limit of line search iteration 2
Minimum line search limit 0.01
Maximum line search limit 1.0
Conditional break
criteria
Displacement error
multiple
Break immediately 10000
Break after step 10
Residual error multiple
Break immediately 10000
Break after step 10
Abs. residual error
multiple
Break immediately 10000
Break after step 10
Energy error multiple
Break immediately 1000000
Break after step 1000
4.6. Results and Discussion
Figure 4.20 shows the resulted load-deformation curves and compares them with the experimental
data.
96
Figure 4.20: Load-deformation diagrams of the numerical model: (A) RC-Column, (B) HFA-
Column, (C)HFA-Tube, (D) HFA-Tube-SEA, (E)HFA-Tube-PSEA.
97
The load-deformation diagrams demonstrated that the numerical models could sufficiently predict
the complex cyclic behavior of the tested columns in terms of maximum strength, permanent
deformations, and post-peak behavior.
The measures, defined in Section 3.7.3, were used to evaluate the numerical results. Table 4.8
compares the corresponding values from the experimental and numerical programs. Specifically,
a negative value for the error means reduction while a positive value indicates an increase in the
resulted value from the numerical program compared to the experimental program. It was observed
that the numerical models could predict the measures related to the force better than those related
to deformation. This is attributed to the concrete constitutive model in ATENA FEM package [73].
It was found that the concrete constitutive model returns convergence issues when subjected to
multiple large cyclic deformations. Hence, to overcome these convergence issues, default values
of -
and )
*
parameters in the concrete constitutive model were increased until convergence was
achieved, which have introduced some error in to the numerical results. A closer look at the load-
deformation diagrams in Figure 4.20 illustrates the errors at large lateral drift ratios, especially in
the unloading branches.
98
Table 4.8: A comparison of experimental and numerical results for bridge columns.
Specimen
RC-Column HFA-Column HFA-Tube HFA-Tube-SEA HFA-Tube-PSEA
Exp. FEM
Error
(%)
Exp. FEM
Error
(%)
Exp. FEM
Error
(%)
Exp. FEM
Error
(%)
Exp. FEM
Error
(%)
Initial Stiffness, E 0 (kN/m) 1288 1198 -6.99 1129 1393 23.38 949 1132 19.28 865 1284 48.44 853 958 12.31
Drift at Yield, δ y (%) 1.75 2.33 33.14 2.3 2.26 -1.74 2.94 2.72 -7.48 2.28 1.99 -12.7 2.5 2.98 19.20
Yield Force, F y (kN) 24.38 28.12 15.34 27.95 30.33 8.52 27.2 28.69 5.48 20.1 22.96 14.23 22.67 27.66 22.01
Drift at Maximum, δ m (%) 2.97 3.94 32.66 3.99 2.99 -25.1 4.01 3.97 -1.00 4.00 2.99 -25.3 3.73 3.94 5.63
Maximum Force, F m (kN) 27.35 31.59 15.50 31.76 33.55 5.64 30.26 30.77 1.69 22.83 26.76 17.21 24.94 29.83 19.61
Drift at Ultimate, δ u (%) 7.09 5.02 -29.20 7.15 NA NA 7.15 5.02 -29.8 7.01 5 -28.7 4.99 5.03 0.80
Ultimate Force, F u (kN) 23.25 26.86 15.53 26.99 NA NA 25.72 26.16 1.71 19.41 22.75 17.21 21.2 25.36 19.62
Permanent Drift, δ p (%) 4.25 4.61 8.47 3.94 4.81 22.08 3.63 4.37 20.39 0.37 0.19 -48.7 2.84 3.59 26.41
Energy Abs., E a (kN-m) 9.51 9.87 3.79 9.67 11.09 14.68 8.36 9.31 11.36 3.43 2.16 -37.0 5.65 5.19 -8.14
99
The yield force was predicted with a difference in range of 5.48% to 22.01% compared to the
experimental data while the maximum and ultimate forces showed 1.69-19.61% and 1.71-19.61%
difference, respectively. Interestingly, the corresponding errors to these three measures were
similar to each other in each numerical model, minimum in HFA-Tube and maximum in HFA-
Tube-PSEA, indicating that the error was transferred from predicting the yield force to predicting
the maximum and the ultimate forces. The numerical models returned the initial stiffness with
maximum and minimum absolute error of 6.99% and 48.44%, respectively. Excluding HFA-Tube-
SEA, the maximum error was limited to 23.38%. Uncertainty in the source of the observed
deformations in the test setup and simplifying it with reaction of the two line springs, contributed
in the deformation related errors especially, in the initial stiffness. Additionally, the numerical
models predicted the yield drift, the maximum drift, the ultimate drift, and the permanent drift (see
definition in Section 3.7.3) with errors in range of -12.7% to 33.14%, -25.3% to 32.66%, -9.8% to
0.8%, and -48.7% to 26.41%, respectively. Since the numerical model for HFA-Tube-SEA and
HFA-Tube-PSEA were unable to simulate the rupture of Cu-Al-Mn bars, which occurred in the
experimental program, these numerical models returned the maximum error in predicting the
permanent drift. Excluding HFA-Tube-SEA which had two ruptures in the SEA bars during the
test, the absorbed energy was captured with maximum error of 14.68% in the numerical models.
Finally, the behavior of HFA-Column and HFA-Tube was predicted with good accuracy, in
general, indicating the sufficiency of the implemented ECC modeling approach for flexural
dominant elements.
100
CHAPTER 5 - NUMERICAL PARAMETRIC STUDY OF BRIDGE
COLUMNS
Subsequent to developing a FEM approach, which satisfactory predicts cyclic behavior of RECC
columns with SEA bars replaced with the longitudinal reinforcement in the potential plastic hinge
region, a parametric study was conducted to investigate the effect of different variables on the
performance of bridge columns constructed with ECC and Cu-Al-Mn SEA bars. A full-size bridge
column, which was previously studied in the literature (RC case only) was considered as the
control specimen. The relatively high costs of both ECC and Cu-Al-Mn SEAs compared to
conventional concrete and steel rebar, constituted the underlying reason for this parametric study.
Therefore, the amount of ECC and the number of SEA bars would affect significantly the final
cost of large projects. The objective of this parametric study was to deduce an effective bridge
design with ECC and Cu-Al-Mn bars while minimizing the construction cost and retaining the
desired general behavior.
5.1. Geometry of the Columns
A previously tested [78] full-size bridge column with 1.2 m diameter and 7.32 m height was
considered as the control case in the parametric study. Eighteen 35.8 mm diameter (#11) evenly
distributed steel rebars formed the longitudinal reinforcement yielding a reinforcement ratio of
1.55%. The transverse reinforcement consisted of 152 mm spaced double hoops made of 15.9 mm
diameter (#5) rebar. This configuration represents the typical single-column bent system in
California and satisfies seismic design requirements for bridge priers in the region, Caltrans Bridge
Design Specifications [79] and Seismic Design Criteria [80] . The column was placed on a 5486.4
101
x 1828.8 x 1219.2 mm RC foundation and subjected to 2530 kN axial load, approximately 5% of
the ultimate capacity of the column, through a superstructure mass [78].
Figure 5.1 presents the geometry of the columns in the parametric study. To conduct the parametric
study, the geometry of the control specimen was modified slightly to resemble the experimentally
studied bridge columns in Chapter 3. A 1828.8 x 1828.8 x 1219.2 mm RC cap was placed on top
of the column to distribute axial loads uniformly and prevent stress concentration during the
analysis. The longitudinal and transverse reinforcements were embedded 1143 mm inside the end
cap and the foundation to create reinforcement connectivity between the column, foundation, and
the top cap.
In case of incorporating Cu-Al-Mn SEA bars as reinforcement at the potential plastic hinge region,
the SEA bars were embedded 355.6 mm in the foundation and extended 1372.1 mm above the
foundation, see Figure 5.1, following the same aspect ratios in the experimental program presented
in Chapter 3. The SEA bars were fully connected to the adjacent rebars and their cross-sections
were reduced around 20% compared to the longitudinal rebar. This configuration matches the
relative dimensions used in the tested bridge columns.
102
Figure 5.1: Geometry of columns in the parametric study (dimensions are in mm).
103
5.2. Parametric Study Matrix
As mentioned before, ECC is a relatively expensive material compared to the conventional
concrete. Therefore, it was proposed to reduce the ECC usage in construction by considering
hollow RECC sections, RECC-H, and hollow RECC sections filled with concrete, RECC-H-F. In
the parametric study, to characterize the effectiveness of these configurations and compare to
conventional construction, full RC cross-section, hollow RC cross-section, and full RECC cross-
sections were considered in addition to the RECC-H and RECC-H-F cross-sections. Figure 5.2
presents the five different cross-sections.
Figure 5.2: Different cross-sections considered in the parametric study: (A) RC, (B) RC-H, (C) RECC,
(D) RECC-H, and (E) RECC-H-F.
To investigate the effect of the wall thickness in the RC-H, RECC-H, and RECC-H-F cross-
sections, three different hollow sections with 711 mm ,610 mm, and 475 mm inner diameters were
considered. Figure 5.3 shows the analogous wall thicknesses for RECC-H cross-section. These
wall thicknesses resulted in the hollow ratio (ratio of A2-to-A1 in Figure 5.3) of 35%, 25%, and
15%, respectively. The cross-sectional analysis, further confirmed that the location of the neutral
104
axis remains inside the wall thickness, see Figure 5.3, at the peak point of moment-curvature
diagram which made the inner-layer longitudinal reinforcement unessential.
Figure 5.3: Different wall thicknesses for RECC-H cross-section in the parametric study (dimensions in
mm).
In addition to the type of the cross-section and the wall thickness, number of Cu-Al-Mn SEA bars
in the cross-section were included in the variables of the parametric study. The results from HFA-
Tube-PSEA, see Chapter 3, showed that partial replacement of the longitudinal reinforcement with
Cu-Al-Mn SEA bars in the potential plastic hinge region considerably reduces permeant
deformations under lateral cyclic loads and increases the absorbed energy. Therefore, with lower
number of SEA bars, the permeant deformations can be reduced considerably, which is in interest
of construction cost reduction since SEA bars are relatively expensive compared to steel
reinforcement. Four, eight, 12, and 16 Cu-Al-Mn bars were considered in the potential plastic
hinge region of the bridge columns to investigate the effect of number of SEA bars in the cross-
section. Figure 5.4 presents location of the SEA bars.
105
Figure 5.4: Bridge columns with: (A) no SEA, (B) four SEA, (C) eight SEA, (D) 12 SEA, and (E) 16 SEA
bars.
Moreover, with reference to the realistic situations, producing large quantities of ECC while
achieving a high tensile ductility is quite challenging [81], despite its feasibility in a controlled lab
environment. Mechanical properties of ECC are strongly related to fiber distribution, void size,
and water-to-binder ratio, which are hard to control in large-scale construction. To consider ECC
mixing quality effect, three different ECC mixtures with 0.5%, 1%, and 2% ultimate strain-
hardening capacity, see Figure 5.5, were considered in the parametric study.
Figure 5.5: Different ECC qualities in parametric study.
Tensile Stress
Tensile Strain
(%)
0.5
1.0
2.0
106
Table 5.1 summarizes the resulted 74 cases defined using all the variables introduced in this
section. All the cases are labeled according to “section type/q/t/xx” where “section type” is the
type of section according to Figure 5.2, “q” indicates ECC quality according to Figure 5.5, “t” is
the wall thickness in mm according to Figure 5.3, and “xx” shows number of Cu-Al-Mn bars in
the section according to Figure 5.4.
107
Table 5.1: Parametric study matrix.
Case Name
Section
Type
ECC
Wall Thickness
(mm)
# of
SEA
ars
# of Steel
Bars
RC/NA/NA/00
RC
NA
a
NA
- 18
RC/NA/NA/04 4 14
RC/NA/NA/08 8 10
RC/NA/NA/12 12 6
RC/NA/NA/16 16 2
RC-T/NA/305/00
RC-H
NA NA
- 18
RC-T/NA/305/04 4 14
RC-T/NA/305/08 8 10
RC-T/NA/305/12 12 6
RC-T/NA/305/16 16 2
RECC/0.5/NA/00
RECC
0.5% NA - 18
RECC/1/NA/00 1% NA - 18
RECC/2/NA/00
2% NA
- 18
RECC/2/NA/04 4 14
RECC/2/NA/08 8 10
RECC/2/NA/12 12 6
RECC/2/NA/16 16 2
RECC-H/2/254/00
RECC-H
2%
254
- 18
RECC-H/2/254/04 4 14
RECC-H/2/254/08 8 10
RECC-H/2/254/12 12 6
RECC-H/2/254/16 16 2
RECC-H/2/305/00
305
- 18
RECC-H/2/305/04 4 14
RECC-H/2/305/08 8 10
RECC-H/2/305/12 12 6
RECC-H/2/305/16 16 2
RECC-H/2/381/00
381
- 18
RECC-H/2/381/04 4 14
RECC-H/2/381/08 8 10
RECC-H/2/381/12 12 6
RECC-H/2/381/16 16 2
RECC-H-F/2/254/00
RECC-H-F
2%
254
- 18
RECC-H-F/2/254/04 4 14
RECC-H-F/2/254/08 8 10
RECC-H-F/2/254/12 12 6
RECC-H-F/2/254/16 16 2
RECC-H-F/2/305/00
305
- 18
RECC-H-F/2/305/04 4 14
RECC-H-F/2/305/08 8 10
RECC-H-F/2/305/12 12 6
RECC-H-F/2/305/16 16 2
RECC-H-F/2/381/00
381
- 18
RECC-H-F/2/381/04 4 14
RECC-H-F/2/381/08 8 10
RECC-H-F/2/381/12 12 6
RECC-H-F/2/381/16 16 2
a
Not applicable
108
5.3. Material Constitutive Models
The material constitutive models used in the numerical study of the tested columns, Section 4.1,
were implemented in the parametric study as well.
5.3.1 Concrete
Identical values were used for parameters of the concrete constitutive model to model concrete in
the columns of RC and RC-H section, the concrete core of RECC-H-F sections, and the foundation
and the top cap of all the columns. In the constitutive model, the compressive strength,
, was
matched with the provided value from compressive test of cylinder specimens at the testing time
as reported in the literature. The uniaxial compression test was conducted on six 152 x 305 mm
cylinder specimens which taken from different concrete batches used to construct the column. The
cylinder specimens had 42 days of age at the time of testing and resulted in 40.9 MPa strength due
to monotonic compressive loads [78]. In addition, default values calculated by ATENA [73]
software were used for the other parameters. Table 5.2 summarizes the corresponding values for
parameters of the concrete constitutive model.
109
Table 5.2: Parameters of concrete constitutive model, CC3DNonLinCementitious2, in the parametric
study.
Section Material Property Value
Basic
Elastic modulus, !
(MPa) 36400
Poisson’s ratio, μ 0.2
Compressive strength,f
^
(MPa) 40.73
Tensile strength, f
_
(MPa) 3.166
Tensile
Specific fracture energy, ;
<
(MN/m) 7.916E-5
Tension stiffening, -
0.1
Unloading factor 0.0
Compressive
Critical compressive displacement, )
*
(m) -0.0025
Plastic strain at compressive strength, (
-1.119E-03
Reduction of strength due to cracks, =
,>
0.8
Shear
Crack shear stiffness factor, ?
<
20.0
Aggregate size in aggregate interlock, @-9 0.02
Miscellaneous
Failure surface eccentricity 0.52
Multiplier for the specific flow direction, A 0.0
Specific material weight, B
CD
E
F
0.023
Fixed crack model coefficient 1.0
5.3.2 Reinforcing steel
“Cyclic Reinforcement” constitutive model in bilinear and bilinear with hardening forms were
used to simulate behavior of the longitudinal and transverse reinforcements, respectively. The
yield strengths were matched with the provided experimental results from direct tensile test of the
reinforcing bars in the control case. The specimens were prepared and tested based on ASTM A
706 standard [82]. Three samples of longitudinal reinforcement were tested which resulted in
average yield strength, ultimate strength, and modulus of elasticity equal to 519 MPa, 707 MPa,
and 196,000 MPa, respectively. Additionally, uniaxial tensile test on five samples taken from the
transverse reinforcement resulted in yield strength and ultimate strength equal to 338 MPa and 592
MPa [78]. Through lack of yield plateau in the stress-strain diagram of the transverse
110
reinforcement, it was converted to bilinear behavior with hardening considering equivalent area
underneath of the stress-strain diagram up to 15% strain. Table 5.3 summarizes the corresponding
values of the reinforcement constitutive models.
Table 5.3: Parameters of reinforcement constitutive model, “Cyclic Reinforcement”, in the parametric
study.
Section Material Property
Value
Longitudinal Transverse
Basic
Elastic modulus, !
(MPa) 200,000
Yield stress, σ
H
(MPa) 517 550
Ultimate Stress, σ
_
(MPa) NA
a
578
Ultimate Strain, (
>
NA 0.15
M-P
Bauschinger effect exponent, R 20
Menegotto-Pinto parameter, -
2
18.5
Menegotto-Pinto parameter, -
I
0.15
a
Not applicable
5.3.3 ECC
The developed approach in the numerical modeling of the tested columns which presented in
Section 4.1.3, concrete constitutive models with 0.8% smeared reinforcement in six directions,
was also implemented to simulate behavior of ECC in the parametric study. Table 5.4 and Table
5.5 list the corresponding values for the smeared reinforcement and concrete constitutive models,
respectively. Figure 5.6 demonstrates the resulted stress-strain diagrams from the numerical
modeling of the dog-bone specimens using the provided values in Table 5.4 and Table 5.5.
111
Table 5.4: Smeared reinforcement constitutive model to simulate ECC in the parametric study.
Section Material Property
Value
0.5% 1% 2%
Basic
ε
2
(%) 0.075
ε
I
(%) 0.5 1.1 2.5
ε
J
(%) 1.5 2.1 3.1
σ
2
(MPa) 150
σ
I
(MPa) 170
σ
J
(MPa) 30
Active in compression Unchecked
M-P
Bauschinger effect exponent, R 20
Menegotto-Pinto parameter, -
2
18.5
Menegotto-Pinto parameter, -
I
0.15
Miscellaneous Specific material weight, B (MN/m
3
) 0.0785
Table 5.5: Concrete constitutive model, 3D Nonlinear Cementitious2, to simulate ECC
Section Material Property
Value
0.5% 1% 2%
Basic
Elastic modulus, !
(MPa) 36450
Poisson’s ratio, μ 0.2
Compressive strength,f
^
(MPa) 40.9
Tensile strength,
(MPa) 3
Tensile
Specific fracture energy, ;
<
(MN/m) 3.938E-5 3.138E-5
Tension stiffening, -
0.4
Unloading factor 0.0
Compressive
Critical compressive displacement, )
*
(m) -0.005
Plastic strain at compressive strength, (
-1.122E-3
Reduction of strength due to cracks, =
,>
0.8
Shear
Crack shear stiffness factor, ?
<
G
a
20
Aggregate size in aggregate interlock, @-9 0.02
Miscellaneous
Failure surface eccentricity 0.52
Multiplier for the specific flow direction, A 0.0
Specific material weight, B
CD
E
F
0.023
Fixed crack model coefficient 1.0
112
Figure 5.6: Stress-strain diagrams for different ECC properties.
For the different ECC properties, the single brick element returned very similar cyclic tension-
compression behavior since the compressive strength was identical in the three considered ECC
qualities. It worth to mention that the compressive strength of ECC is strongly insensitive to the
mixing methods/quality unlike the tensile ductility. Figure 5.7 presents the resulted cyclic tension-
compression behavior for the ECC with 0.5% ultimate strain-hardening.
0.0 0.5 1.0 1.5 2.0 2.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
ECC 0.5%
ECC 1%
ECC 2%
Stress (MPa)
Strain (%)
113
Figure 5.7: Cyclic tension-compression behavior of ECC with 0.5% ultimate strain-hardening.
5.3.4 Cu-Al-Mn SEA
The implemented SEA constitutive model, explained in Section 4.1.4, with identical values as
incorporated in the numerical modeling of the tested columns was used to simulate behavior of
Cu-Al-Mn SEA bars in the parametric study.
5.3.5 Bond characteristics
The same weak bond model developed in Section 4.1.5 and used to simulate debonding in HFA-
Tube-SEA and HFA-Tube-PSEA columns was implemented in the parametric study for the
debonding purposes. However, a new bond material model was generated for the longitudinal
reinforcement in cases without Cu-Al-Mn SEA bars. Figure 5.8 shows the bond-slip diagram. The
bond material model was output of the software for good bond quality of 35.8 mm (#11) rebar in
concrete with 40.73 MPa compressive strength, which matches the concert compressive strength
in the control specimen.
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2
-50
-40
-30
-20
-10
0
Stress (MPa)
Strain (%)
0.0 0.1 0.2
0
1
2
3
4
5
Stress (MPa)
Strain (%)
114
Figure 5.8: Bond-slip model used in the parametric study.
5.4. Meshing, Boundary Conditions and Loading Protocol
In general, the same approach as mentioned in the numerical modeling of the tested columns,
explained in Chapter 4, was used to develop the numerical models for the parametric study as well.
Figure 5.9 illustrates the developed models for the parametric study. The RC-H, RECC-H, and
RECC-H-F columns were connected directly to the foundation and the top cap unlike the
numerical modeling of the tested columns where the HFA-Tube, HFA-Tube-SEA, HFA-Tube-
PSEA columns were embedded in the foundation and the top cap. Since the column is connected
to the end caps with the use of perfect contacts in the both approaches, the overall behavior of the
bridge column is negligibly affected by embedding the columns in the end caps. Additionally,
since there was no mention of foundation flexibility in the original reference [78], the line springs
were removed from the boundary condition and the foundation was fixed in all directions.
The spacing of the transverse reinforcement was doubled, compared to the actual spacing in the
experimental values, and the corresponding area was increased to satisfy the transverse
0 2 4 6 8 10 12 14 16 18 20
0
2
4
6
8
10
12
14
Bond Stress (MPa)
Slip (mm)
τ=0.58 MPa
115
reinforcement ratio in the experimental work (1.55% [78]). A distributed axial load of 758.4 kN/m
2
was assigned to the top cap in the gravity direction in two steps.
Considering large numbers of the cases, time requirements for analyzing numerical models under
incremental cyclic loads, and the results from the numerical modeling of the tested columns in
Chapter 4, it was decided to perform the parametric study with applying one cycle of 7% drift to
each case. The lateral load was applied in displacement control and with a step size of 0.017% drift
up to approximately 0.5% lateral drift and thereafter with 0.087% drift steps to achieve better
convergence. Similarly, the unloading started with sixteen 0.087% drift load steps, up to 5.6%
drift, and then reduced to 0.044%. Linear tetrahedral elements with a maximum mesh size of 20
mm and 33 mm were used to mesh the column and rest of the geometry, respectively. Lastly, the
same solution parameters as used in the numerical modeling of the tested columns were used to
analyze the models.
116
Figure 5.9: Developed numerical model for parametric study (RECC-H-F sections): (A) geometry and loading, (B) reinforcement and SEA bars,
(C) boundary condition and meshing.
117
5.5. Results and Discussion
In this section, the results from the computational models are presented with a discussion on the
effect of each of the variables in the parametric study on the behavior of bridge columns
constructed with ECC and Cu-Al-Mn SEA bars.
5.5.1 Validation of the control case
The numerical modeling approach was validated with the results from the experimental work
conducted on the scaled bridge columns as presented in Chapter 4. However, for the parametric
study, the full-scale bridge column, which is available in the literature was used as the control case.
The control case was subjected to earthquake loads in the experimental work unlike the quasi-
static cyclic loading applied to the models in the parametric study. Accordingly, a direct
comparison of the numerical and experimental results of the control case, for validating purposes,
was unattainable. Therefore, idealized monotonic behavior of the RC column provided by the
authors [78] was used for the comparison. Figure 5.10 provides the comparison between the
idealized monotonic behavior and the numerical results for the control case. As evident, the
numerical model could predict the ultimate lateral strength and the post-peak behavior with
sufficient accuracy.
118
Figure 5.10: Comparing the numerical results with the idealized monotonic behavior for the control case.
5.5.2 Effect of ECC quality
Figure 5.11 presents the load-deformation diagrams of RECC/0.5/NA/00, RECC/1/NA/00, and
RECC/2/NA/00 cases. As evident, the ECC quality had negligible effect on the cyclic response of
the models. The ultimate lateral strength was slightly enhanced as the ultimate tensile strain-
hardening capacity of the mixture was increased but the permanent deformation changed
insignificantly. In addition, all the cases showed similar behavior in low and high drift ratios even
though their responses were different in 2 to 4% drift range.
0 1 2 3 4 5 6 7 8
-600
-400
-200
0
200
400
600
800
1000
RC/NA/NA/00
Idealized Monotonic Behavior
Load (kN)
Drift (%)
119
Figure 5.11: Effect of ECC quality in cyclic behavior of the bridge columns.
Since the considered columns were flexural dominant elements, the cyclic behavior of the column
was mainly defined by the longitudinal reinforcement. Figure 5.12 shows the rebar strain at 2%
drift. The rebar strain at 2% drift (around the maximum lateral strength drift) illustrated that the
longitudinal reinforcement experienced the same strain condition in all three cases. Furthermore,
ECC had similar strain contours in the plastic hinge regions for the all three cases (see Figure
5.13). However, more large cracks developed in RECC/1/NA/00 and RECC/2/NA/00 cases
compared to RECC/0.5/NA/00 case (see Figure 5.13).
0 1 2 3 4 5 6 7 8
-600
-400
-200
0
200
400
600
800
1000
1200
RECC/0.5/NA/00
RECC/1/NA/00
RECC/2/NA/00
Force (kN)
Drift (%)
120
Figure 5.12: Rebar strain at 2% drift: (A) RECC/0.5/NA/00, (B) RECC/1/NA/00, and (C) RECC/2/NA/00.
121
Figure 5.13: Strains in Plastic hinge region at 2% drift: (A) RECC/0.5/NA00, (B) RECC/1/NA/00, and
(C) RECC/2/NA/00 (cracks over 1 mm are shown).
The similarity between strains led to comparable corresponding stresses which resulted in similar
load-drift diagrams in cases with different ECC quality as presented in Figure 5.11. A closer look
at the plastic hinge strains of the columns with different ECC quality, see Figure 5.13, reveals that
the strains were lower than 2% at 2% drift, which made the RECC/2/NA/00 case to maintain its
ultimate strength to higher drift ratios, see Figure 5.11 .
5.5.3 Load-deformation diagrams
Figure 5.14 and Figure 5.15 present the load-deformation diagrams of the parametric study cases.
122
Figure 5.14: Load-deformation diagrams: (A) RC/NA/NA/##, (B) RECC/NA/NA/##, (C) RC-H/NA/NA/##.
123
Figure 5.15: Load-deformation diagrams: (A) RECC-H/2/254/##, (B) RECC-H-F/2/254/##,(C) RECC-
H/2/305/##, (D) RECC-H-F/2/305/##, (E) RECC-H/2/381/##, (F) RECC-H-F/2/381/##.
124
It can be seen that the area inside the load-deformation curves in all the cross-sections shrunk as
the number of the SEA bars at the potential plastic hinge region increases. This phenomenon can
be related to lower yield and ultimate strength of Cu-Al-Mn SEA compared to the reinforcing steel
in addition to its superplastic behavior. Interestingly, RC-H/NA/305/00 showed identical cyclic
load-deformation behavior as RC/NA/NA/00, thereby indicating that the hollow bridge columns
with neutral axis within the wall thickness can return the same cyclic response as regular bridge
columns with lower material consumption. Additionally, all the cases without Cu-Al-Mn SEA bars
had similar load at the end of the unloading branch, i.e., zero drift. The developed ECC model
primarily followed the base concrete constitutive model in terms of compressive behavior and the
corresponding damage equations under cyclic loads. Although generally, ECC has higher
compressive strength compared to conventional concrete because of higher binder-to-water ratio,
in the parametric study, the compressive strength of ECC was considered equal to the compressive
strength of the control case to eliminate effect of the material properties in compression. Similar
unloading behaviors were expected in the cases with equal number of the longitudinal
reinforcement and Cu-Al-Mn SEA bars because of flexural dominant behavior of the bridge
columns, presence of the neutral axis within the wall thickness of the hollow sections, and identical
compressive strength of the concrete and ECC.
5.5.4 Quantitative evaluation
For a quantitative comparison of the performance of the studied cases in the parametric study,
several metrics including maximum, ultimate, and yield points, initial stiffness, permanent
deformation, and energy absorption were defined, as shown in Figure 5.16. The definition of the
maximum, ultimate, and yield points were the same as those defined in Section 3.7.3. However,
the initial stiffness, !
, was defined as the slope of the load-deformation diagram up to 0.05% drift,
125
up to which elastic behavior was observed in all the cases. In the experimental work presented in
Chapter 3, the initial stiffness was defined as the cord stiffness between 40% and 70% of the peak
load because of the applied incremental cyclic loads and difficulties in controlling loads in low
deformations, unlike numerical analyzing. Additionally, energy absorption was taken equal to the
area inside the load-deformation diagram up to permanent drift point, which was defined as the
corresponding drift to zero force in the unloading curve. Table 5.6 and Table 5.7 summarize the
derived values of the defined metrics and their difference with respect to the control case,
respectively.
Figure 5.16: Definition of maximum strength, V m, maximum drift, δ m, ultimate strength, V u, ultimate drift,
δ u, yield strength, V y, yield drift, δ y, initial stiffness, E 0, permanent drift, δ p, and energy absorption, E 0, as
metric for quantitate comparison of the parametric study results.
0 1 2 3 4 5 6 7 8
-600
-400
-200
0
200
400
600
800
1000
1200
Load (kN)
Drift (%)
b
c
Initial
Stiffness
d
d
'
'
15%
Maximum
Ultimate
Permanent Drift
25%
Yield
d
e
e
Energy
Absorption, Ea
126
Table 5.6: Quantitative evaluation of the parametric study.
Case Name
Section
Type
d
(kN)
d
e
(kN)
!
(kN/m)
!
R
(kN.m)
(%)
e
(%)
'
(%)
(%)
RC/NA/NA/00
RC
816.82 720.88 31443 327.92 1.54 1.08 5.84 5.71
RC/NA/NA/04 720.88 628.29 30760 286.60 1.54 1.08 5.66 5.57
RC/NA/NA/08 640.74 552.81 30272 247.71 1.63 1.12 5.24 5.24
RC/NA/NA/12 576.33 494.40 29996 197.57 1.72 1.16 5.19 0.53
RC/NA/NA/16 504.93 421.94 29914 130.74 1.72 1.08 5.25 0.18
RC-T/NA/305/00
RC-H
809.06 713.68 29641 322.78 1.63 1.12 5.73 5.69
RC-T/NA/305/04 709.86 619.65 28966 280.25 1.63 1.11 5.54 5.58
RC-T/NA/305/08 630.04 542.81 28477 242.03 1.72 1.14 5.09 5.26
RC-T/NA/305/12 568.16 489.37 28203 194.70 1.72 1.21 5.07 0.65
RC-T/NA/305/16 495.45 418.85 28118 131.26 1.72 1.15 4.99 0.21
RECC/0.5/NA/00
RECC
1010.23 897.35 31765 410.20 1.37 1.09 6.90 5.66
RECC/1/NA/00 1046.63 920.81 31765 413.18 1.54 1.15 6.48 5.62
RECC/2/NA/00 1065.08 932.26 31765 419.38 1.80 1.18 6.10 5.59
RECC/2/NA/04 963.42 835.67 31081 375.47 1.80 1.19 6.00 5.40
RECC/2/NA/08 880.52 756.16 30594 338.33 1.80 1.21 6.02 4.90
RECC/2/NA/12 821.91 699.05 30319 281.35 1.89 1.25 5.76 0.64
RECC/2/NA/16 758.07 632.12 30236 202.67 1.89 1.18 4.88 0.11
RECC-H/2/254/00
RECC-H
951.23 834.27 28338 375.86 1.63 1.13 5.55 5.69
RECC-H/2/254/04 849.30 736.54 27670 333.53 1.63 1.13 5.66 5.53
RECC-H/2/254/08 768.44 660.36 27181 295.24 1.72 1.15 5.39 5.18
RECC-H/2/254/12 707.67 602.19 26903 245.46 1.80 1.20 5.37 0.72
RECC-H/2/254/16 642.23 533.15 26821 176.93 1.80 1.12 4.98 0.41
RECC-H/2/305/00 1006.38 881.23 29855 398.74 1.80 1.18 5.93 5.63
RECC-H/2/305/04 903.96 782.80 29184 354.72 1.72 1.18 6.33 5.43
RECC-H/2/305/08 824.31 706.52 28695 316.87 1.80 1.21 6.00 5.03
RECC-H/2/305/12 763.99 652.60 28418 261.93 1.89 1.27 5.81 0.62
RECC-H/2/305/16 696.01 581.76 28331 179.21 1.89 1.20 5.10 0.13
RECC-H/2/381/00 1046.77 914.74 31651 416.53 1.71 1.16 6.20 5.61
RECC-H/2/381/04 944.29 817.59 30970 372.22 1.80 1.16 6.19 5.44
RECC-H/2/381/08 863.91 739.05 30490 333.43 1.80 1.19 5.87 5.02
RECC-H/2/381/12 804.49 684.63 30219 278.37 1.89 1.24 5.74 0.64
RECC-H/2/381/16 736.95 612.72 30132 205.83 1.89 1.16 5.47 0.20
RECC-H-F/2/254/00
RECC-H-F
973.52 852.92 31915 387.93 1.63 1.11 5.80 5.69
RECC-H-F/2/254/04 872.71 756.75 31231 344.37 1.63 1.11 5.78 5.52
RECC-H-F/2/254/08 792.27 680.21 30738 306.25 1.72 1.13 5.69 5.10
RECC-H-F/2/254/12 731.57 621.72 30461 254.08 1.80 1.17 5.59 0.71
RECC-H-F/2/254/16 665.59 552.57 30376 183.75 1.80 1.09 5.26 0.22
RECC-H-F/2/305/00 1022.96 893.43 31981 406.37 1.71 1.16 6.11 5.61
RECC-H-F/2/305/04 920.49 796.58 31296 362.00 1.72 1.16 6.26 5.44
RECC-H-F/2/305/08 837.58 717.01 30805 321.74 1.80 1.19 6.08 5.01
RECC-H-F/2/305/12 778.55 663.89 30528 267.21 1.89 1.25 5.91 0.64
RECC-H-F/2/305/16 711.41 593.13 30445 191.80 1.89 1.18 5.32 0.15
RECC-H-F/2/381/00 1055.08 921.98 32476 421.70 1.71 1.16 6.48 5.60
RECC-H-F/2/381/04 954.14 826.89 31786 375.09 1.72 1.16 5.90 5.45
RECC-H-F/2/381/08 873.72 748.62 31301 335.55 1.80 1.19 5.59 5.00
RECC-H-F/2/381/12 813.86 692.91 31028 280.32 1.89 1.23 5.46 0.65
RECC-H-F/2/381/16 746.68 621.42 30948 206.64 1.89 1.15 5.00 0.21
127
Table 5.7: Change in the quantitative metrics with respect to the control case.
Case Name
Section
Type
d
(%)
d
e
(%)
!
(%)
!
R
(%)
(%)
e
(%)
'
(%)
(%)
RC/NA/NA/00
RC
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
RC/NA/NA/04 -11.75 -12.84 -2.17 -12.60 0.04 0.00 -3.08 -2.45
RC/NA/NA/08 -21.56 -23.31 -3.72 -24.46 5.63 3.70 -10.27 -8.23
RC/NA/NA/12 -29.44 -31.42 -4.60 -39.75 11.20 7.41 -11.13 -90.72
RC/NA/NA/16 -38.18 -41.47 -4.86 -60.13 11.23 0.00 -10.10 -96.85
RC-T/NA/305/00
RC-H
-0.95 -1.00 -5.73 -1.57 5.54 3.70 -1.88 -0.35
RC-T/NA/305/04 -13.09 -14.04 -7.88 -14.54 5.59 2.78 -5.14 -2.28
RC-T/NA/305/08 -22.87 -24.70 -9.43 -26.19 11.17 5.56 -12.84 -7.88
RC-T/NA/305/12 -30.44 -32.11 -10.30 -40.63 11.20 12.04 -13.18 -88.62
RC-T/NA/305/16 -39.34 -41.90 -10.57 -59.97 11.24 6.48 -14.55 -96.32
RECC/0.5/NA/00
RECC
23.68 24.48 1.02 25.09 -11.11 0.93 18.15 -0.88
RECC/1/NA/00 28.13 27.73 1.02 26.00 -0.02 6.48 10.96 -1.58
RECC/2/NA/00 30.39 29.32 1.02 27.89 16.61 9.26 4.45 -2.10
RECC/2/NA/04 17.95 15.92 -1.15 14.50 16.65 10.19 2.74 -5.43
RECC/2/NA/08 7.80 4.89 -2.70 3.17 16.68 12.04 3.08 -14.19
RECC/2/NA/12 0.62 -3.03 -3.57 -14.20 22.25 15.74 -1.37 -88.79
RECC/2/NA/16 -7.19 -12.31 -3.84 -38.20 22.27 9.26 -16.44 -98.07
RECC-H/2/254/00
RECC-H
16.46 15.73 -9.88 14.62 5.54 4.63 -4.97 -0.35
RECC-H/2/254/04 3.98 2.17 -12.00 1.71 5.58 4.63 -3.08 -3.15
RECC-H/2/254/08 -5.92 -8.40 -13.55 -9.97 11.16 6.48 -7.71 -9.28
RECC-H/2/254/12 -13.36 -16.46 -14.44 -25.15 16.73 11.11 -8.05 -87.39
RECC-H/2/254/16 -21.37 -26.04 -14.70 -46.04 16.75 3.70 -14.73 -92.82
RECC-H/2/305/00 23.21 22.24 -5.05 21.60 16.62 9.26 1.54 -1.40
RECC-H/2/305/04 10.67 8.59 -7.18 8.17 11.12 9.26 8.39 -4.90
RECC-H/2/305/08 0.92 -1.99 -8.74 -3.37 16.69 12.04 2.74 -11.91
RECC-H/2/305/12 -6.47 -9.47 -9.62 -20.12 22.26 17.59 -0.51 -89.14
RECC-H/2/305/16 -14.79 -19.30 -9.90 -45.35 22.28 11.11 -12.67 -97.72
RECC-H/2/381/00 28.15 26.89 0.66 27.02 11.08 7.41 6.16 -1.75
RECC-H/2/381/04 15.61 13.42 -1.50 13.51 16.66 7.41 5.99 -4.73
RECC-H/2/381/08 5.77 2.52 -3.03 1.68 16.69 10.19 0.51 -12.08
RECC-H/2/381/12 -1.51 -5.03 -3.89 -15.11 22.26 14.81 -1.71 -88.79
RECC-H/2/381/16 -9.78 -15.00 -4.17 -37.23 22.28 7.41 -6.34 -96.50
RECC-H-F/2/254/00
RECC-H-F
19.18 18.32 1.50 18.30 5.54 2.78 -0.68 -0.35
RECC-H-F/2/254/04 6.84 4.98 -0.67 5.02 5.58 2.78 -1.03 -3.33
RECC-H-F/2/254/08 -3.01 -5.64 -2.24 -6.61 11.15 4.63 -2.57 -10.68
RECC-H-F/2/254/12 -10.44 -13.76 -3.12 -22.52 16.72 8.33 -4.28 -87.57
RECC-H-F/2/254/16 -18.51 -23.35 -3.39 -43.96 16.75 0.93 -9.93 -96.15
RECC-H-F/2/305/00 25.24 23.94 1.71 23.92 11.07 7.41 4.62 -1.75
RECC-H-F/2/305/04 12.69 10.50 -0.47 10.39 11.12 7.41 7.19 -4.73
RECC-H-F/2/305/08 2.54 -0.54 -2.03 -1.88 16.69 10.19 4.11 -12.26
RECC-H-F/2/305/12 -4.69 -7.91 -2.91 -18.51 22.26 15.74 1.20 -88.79
RECC-H-F/2/305/16 -12.90 -17.72 -3.17 -41.51 22.28 9.26 -8.90 -97.37
RECC-H-F/2/381/00 29.17 27.90 3.29 28.60 11.07 7.41 10.96 -1.93
RECC-H-F/2/381/04 16.81 14.71 1.09 14.38 11.12 7.41 1.03 -4.55
RECC-H-F/2/381/08 6.97 3.85 -0.45 2.33 16.69 10.19 -4.28 -12.43
RECC-H-F/2/381/12 -0.36 -3.88 -1.32 -14.52 22.25 13.89 -6.51 -88.62
RECC-H-F/2/381/16 -8.59 -13.80 -1.57 -36.98 22.28 6.48 -14.38 -96.32
128
The values in Table 5.7 present the difference of each quantitative metric with respect to
RC/NA/NA/00 (the control case). Specifically, a negative value inside the cells means reduction
while a positive value indicates an increase.
To better understand the effect of each variable in the parametric study, the corresponding values
of the defined metrics were normalized with respect to the results from the control case to be
presented in bar-chart format. Figure 5.17 shows the normalized lateral maximum strengths, d
.
The maximum lateral strength gain of 30.4%, compared to the control case, was observed in
RECC/2/NA/00. The illustrations revealed that the ultimate lateral strength reduced in all the
section types as the number of the Cu-Al-Mn SEA bars increased in the potential plastic hinge
region. This reduction was because of lower yield strength of Cu-Al-Mn SEA bars compared to
the reinforcing steel.
Interestingly, in RC and RC-H sections with 16 Cu-Al-Mn SEA bars, the ultimate lateral strength
dropped to approximately 60% of the sections without Cu-Al-Mn SEA bars. It should be noted
that, application of SEA bars in bridge columns is generally to mitigate permanent deformations
due to strong ground motions. In low and medium intensity earthquakes, it is essential to ensure
adequate lateral strength in bridge columns. Therefore, the column cross-section should be
enlarged to provide adequate lateral strength when large ratio of longitudinal reinforcement is
replaced with Cu-Al-Mn SEA bars in the potential plastic hinge region of RC and RC-H sections.
Incorporating ECC instead of conventional concrete, increasing diameter of the Cu-Al-Mn SEA
bars or implementing more SEA bars, and using RECC hollow sections with low hollow ratio are
some of the other options to overcome this issue.
129
Figure 5.17: Normalized maximum lateral strengths, V m.
The maximum strength of the bridge column was increased with incorporating ECC due to its
strain-hardening behavior. ECC maintains its tensile strength under higher strains, unlike concrete
in which the tensile strength drops to zero after a small tensile strain. Stresses at the maximum
drift, δm, in the section located at the distance equal to the diameter of the columns above the
foundation, (see Figure 5.18) revealed higher tensile stresses in RECC/2/NA/00 compared to
RC/NA/NA/00. Therefore, higher tensile stresses led to higher maximum lateral strength in the
cases with ECC.
130
Figure 5.18: Stresses at maximum drift in the section located at the distance equal to the diameter of the
columns above the foundation: (A) RC/NA/NA/00, (B) RECC/2/NA/00.
As expected, the maximum lateral strength of RC-H/NA/305/## and RECC-H/2/381/## sections
were found to be slightly lower (2.59% of the control case at most) than the corresponding
RC/NA/NA/##, and RECC/2/NA/## sections, respectively, since the neutral axis of the hollow
sections was within the wall thickness. Correspondingly, maximum lateral strength increased
slightly (2.92% at most) by filling the hollow section with concrete since the filled area had an
insignificant contribution in the overall behavior of the designed bridge columns. Finally,
increasing wall thickness in RECC-H and RECC-H-F sections increased the maximum strength
through contribution of larger cross-sectional area of ECC in the response of the bridge column.
131
Figure 5.19 illustrates the normalized maximum drifts, δm. As it can be seen, the maximum drift
increased as number of the Cu-Al-Mn SEA bars in the cross-section and volume of ECC in the
bridge column increased. This trend can be related to the lower modulus of elasticity of Cu-Al-Mn
SEA compared to the reinforcing steel and strain-hardening behavior of the ECC mixture,
respectively, which led to maximum increment of 22.3% in the maximum drift. Additionally, RC-
H sections showed higher maximum drifts than RC sections because of lower section modulus.
However, the increment was insignificant (5.6% at most) specially in cases with 12 and 16 Cu-Al-
Mn SEA bars in the plastic hinge region.
Figure 5.19: Normalized maximum drifts, δ m.
132
Figure 5.20 shows the normalized yield strengths, d
e
. The yield strength followed the same trend
as the maximum strength for the same reasons explained above. The yield strength varied between
41.9% reduction (in RC-H/NA/305/16) to 29.32% enhancement (in RECC/2/NA/00) compared to
the control case. It should be noted that the difference between the yield strengths of the Cu-Al-
Mn SEA and reinforcing steel mainly affected the yield strength of each section type as the number
of Cu-Al-Mn SEA bars increased in the plastic hinge region through flexural dominant behavior
of the column. The yield strength could be improved by either increasing diameter of the Cu-Al-
Mn SEA bars or incorporating more SEA bars in the section.
Figure 5.20: Normalized yield strengths, V y.
133
Figure 5.21 exhibits the normalized yield drifts,
e
. As evident, incorporating four Cu-Al-Mn in
the plastic hinge region had limited effect on the yield drift. Considering the methodology used to
extract the yield points, equal yield drifts indicates that 0.75 of the maximum lateral strength
occurred at the same drift level. Additionally, increasing number of the Cu-Al-Mn SEA bars up to
12 bars increased the yield drift because of lower modulus of elasticity and higher yield drift of
Cu-Al-Mn compared to reinforcing steel. However, replacing 16 of the longitudinal reinforcing
bars with Cu-Al-Mn SEA bars led to the lowest yield drift in each section type meaning that 0.75
of the maximum lateral strength occurred in a lower drift compared to other cases because of larger
strains than (
M
(see Figure 4.10) in the SEA bars.
Figure 5.21: Normalized yield drifts, δ y.
134
Figure 5.22 presents the normalized ultimate drifts,
'
. The ultimate drift represents post-peak
degradation rate. The ultimate drift decreased by increasing number of Cu-Al-Mn SEA bars in the
section through concentration of damage in the surrounding material as observed in the
experimental program, Chapter 3. Additionally, incorporating ECC improved the ultimate drift
because of strain-hardening characteristics of ECC.
Figure 5.22: Normalized ultimate drifts, δ u.
135
Figure 5.23 presents the normalized initial stiffness, !
. The initial stiffness is a function of the
modulus of elasticity of the materials and the cross-sectional area. It can be observed that the initial
stiffness decreased with increase in the number of the Cu-Al-Mn SEA bars due to lower modulus
of elasticity of Cu-Al-Mn (35,800 MPa) as compared to the steel reinforcement (200,000 MPa).
The loss of initial stiffness through increasing number of the SEA bars was identical in each section
type with a constant wall thickness, if applicable. As an example, incorporating 16 Cu-Al-Mn SEA
bars in each cross-section led to reduction equal to 4.8% of the initial stiffness of the control case
compared to the same cross-section without the SEA bars.
Figure 5.23: Normalized initial stiffness, E 0.
Additionally, the elastic modulus of ECC in the proposed approach was higher than the base
concrete because of the smeared reinforcement. Since the base concrete in ECC and the concrete
136
used in the RC, and RC-H sections had a similar modulus of elasticity, RECC sections showed
slightly higher, around 1%, initial stiffness as compared to RC sections. The initial stiffness was
increased by increasing wall thickness in RECC-H sections because of the increase in the moment
of inertia of the sections. In these sections, the initial stiffness was increased around 5% of the
initial stiffness of the control case for every 10% deduction in the hollow ratio. However, the initial
stiffness remained almost constant by changing the wall thickness in the RECC-H-F cases. Figure
5.24 shows the normalized permanent deformations,
f
.
Figure 5.24: Normalized permanent deformations, δ p.
As evident, the permanent deformation decreased as number of Cu-Al-Mn SEA bars increased in
the cross-section. However, increasing the number of Cu-Al-Mn SEA bars in the potential plastic
hinge reduced the permanent deformations inconsiderably up to eight SEA bars. The permanent
137
deformation was decreased 7.88-14.19% by implementing eight Cu-Al-Mn SEA bars in the cross-
sections while it was reduced 87.39-90.72% with 12 Cu-Al-Mn SEA bars. This phenomenon can
be attributed to the stresses and strains in the longitudinal reinforcing. Figure 5.25 presents a
comparison between stresses in RC/NA/NA/08 and RC/NA/NA/12 at 7% drift.
Figure 5.25: Stresses in the longitudinal reinforcement at 7% drift: (A) RC/NA/NA/08, (B) RC/NA/NA/12.
Inspecting the stresses of the longitudinal bars at the 7% drift revealed that some of the longitudinal
steel bars yielded in cases with up to eight Cu-Al-Mn SEA bars. However, the longitudinal steel
reinforcement remained elastic in cases with 12 and 16 Cu-Al-Mn SEA bars. The inelastic
deformations in the longitudinal steel reinforcement caused unrecoverable deformations, which
were observed in cases with up to eight Cu-Al-Mn SEA bars in the potential plastic hinge region.
Furthermore, around 96% reduction in permanent deformations was observed by replacing 16 of
the longitudinal steel bars with Cu-Al-Mn SEA bars in the plastic hinge region. In general,
138
replacing more than 67% of the longitudinal reinforcement in the potential plastic hinge region led
to considerable reduction in the permanent deformations compared to the control case.
Figure 5.26 shows the normalized absorbed energies. As observed in the experiment work, Chapter
3, the absorbed energy showed a decrease in all types of the bridge columns by increasing the
number of Cu-Al-Mn SEA bars in the potential plastic hinge region because of their superelastic
behavior. Additionally, applying ECC to the bridge columns increased the absorbed energy since
the maximum strength, Vm, was increased. The absorbed energy in cases with ECC and 12 or 16
Cu-Al-Mn SEA bars, which showed a considerable reduction in the permanent deformations, was
found to be always lower than that of the control case. This demonstrated that specific attention is
needed in designing bridge columns with a high number of Cu-Al-Mn SEA bars to have enough
energy absorption capacity during strong earthquakes. Furthermore, increasing the wall thickness
in the RECC-H type bridge columns enhanced the absorbed energy since there was higher volume
of ECC in the cross-section. However, filling the RECC-H type bridge columns with conventional
concrete minimally effected the absorbed energy, thereby confirming that the inner volume had
negligible effect on the cyclic behavior of the hollow section bridge columns with neutral axis
within the wall thickness.
139
Figure 5.26: Normalized absorbed energies, E a.
Interestingly, RECC/2/NA/##, RECC-H/2/381/##, and RECC-H-F/2/381/## bridge columns had
almost identical absorbed energies indicating that specific energy absorption capacity can be
reached with a lower volume of material in the bridge columns. Additionally, the crack
distribution at 7% drift showed a higher number of relatively large (larger than 0.5mm) cracks in
cases with ECC than the conventional concrete (see Figure 5.27). The number of cracks, which is
140
related to the absorbed energy, decreased by replacing the longitudinal steel reinforcement with
Cu-Al-Mn SEA bars.
Figure 5.27: 0.5 mm and larger cracks at 7% drift: (A) RC/NA/NA/00, (B) RECC/2/NA/00, (C)
RECC/2/NA/04, (D) RECC/2/NA/16.
141
5.6. Optimal Design
As discussed in the previous section, replacing longitudinal steel reinforcement in the potential
plastic hinge region affects the maximum lateral strength and energy absorption capacity of the
bridge columns in addition to the permanent deformations. These characteristics are important in
low and medium level earthquakes which are more common than those impose large deformations
to bridge columns. Incorporating ECC improved the maximum lateral strength and the energy
absorption capacity of the bridge columns compared to the control RC case which can be taken to
account to compensate the losses through implementing Cu-Al-Mn SEA bars. Increasing the cross-
sectional area, diameter of the Cu-Al-Mn SEA bars, and number of the longitudinal bars are other
options that can be considered to enhance the maximum lateral and energy absorption capacity of
RC bridge columns incorporating Cu-Al-Mn SEA bars in the plastic hinge region. Amount of the
high-performance yet expensive materials used in the design approach is another parameter the
should be consider in order to reduce the construction cost of the final project. It was found that
RECC-H and RECC-H-F sections with neutral axis located in the wall thickness return comparable
load-deformation diagrams to RECC sections. Therefore, a desired behavior can be reached with
lower volume of ECC using RECC hollow sections. Additionally, hollow sections can be
prefabricated which ensures material and fabrication quality control. Figure 5.28 illustrates the
optimal design for the cases studied in the parametric study giving same importance to lateral
strength, reduced permanent deformation, and energy absorption capacity.
142
Figure 5.28: Optimized bridge design (highlighted ones are the optimal design)
143
CHAPTER 6 - EXPERIMENTAL PROGRAM OF BEAM-COLUMN
JOINTS
The main objective of this section is implementing ECC in beam-column joints to improve their
seismic damage tolerance when subjected to more realistic loading. To realise this objective, eight
approximately quarter scaled 3D beam-column joints were constructed and tested under a
combination of axial load, bidirectional bending, and torsion. The variables of the experimental
work included location of the beam-column joints in a building (corner or exterior), construction
material of the panel zone, loading combination, and number of stirrups in the panel zone.
6.1. Beam-Column Joints Design Concept
Replacement of the critical regions of the frame (i.e., the beam-column joints and bases of
columns) with ECC was proposed, to enhance the seismic performance of SMFs. Figure 6.1
schematically illustrates the design concept. Improvement in damage tolerance and energy
absorption capacity of the frames is expected because of superior characteristics of ECC in terms
of shear resistance, bonding with reinforcement, and tensile properties compared to the
conventional concrete. As ECC is more expensive than conventional concrete, the rest of the
structure would be constructed with ordinary concrete.
144
Figure 6.1: Proposed beam-column joint concept in RC special moment resisting frames.
To investigate the behavior of the suggested beam-column joints in 3D form, the beam-column
subassemblies were considered to extend from the midpoint of the panel zone to the adjacent
beams and column where approximately zero moments happen due to lateral loads (see Figure
6.1). These zero moment points were simulated as pin supports in the experimental work.
6.2. Geometry and Test Matrix
To investigate the performance of exterior and corner beam-column joints representing the
behavior of the connections in actual buildings, the joint subassemblies from the first story of a
four story-three bay archetype RC special moment frame were considered. The frame met the
minimum requirements for stiffness, strength, and capacity design as well as strong column-weak
beam design criteria. It was designed according to all the governing code requirements of a high
seismicity location [83].
145
The subassemblies were scaled with an approximately 1/4-scale factor based on the displacement
and force capacities of the testing equipment used for this research. Figure 6.2 demonstrates the
geometry of the beam-column subassemblies. Flexural and shear reinforcement ratios in the beams
and columns corresponded to those in the original frame. Grade 60 hot rolled rebar with 9.5 mm
diameter (#3 U.S. designation rebar) was used for all the longitudinal reinforcements while 6.4
mm diameter deformed grade 60 bars were used as rectangular stirrups. The stirrup spacing was
calculated according to requirements for SMFs in ACI-318 [84]. To maintain symmetry in the
beam-column subassemblies, longitudinal reinforcement was equally distributed on the top and
bottom of the beams and in a square pattern in the columns. This reinforcement configuration
resulted in longitudinal reinforcement ratios of 1.3% and 1.7% in the beams and column,
respectively. Meanwhile, the shear reinforcement ratio in the joint panel zone and the potential
plastic hinge regions of the adjoining beams and columns was maintained at 0.3%. Hoop spacing
was increased in the middle portion of the adjoining elements due to the lower shear demands in
these regions. However, to prevent undesirable failures at the location of boundary conditions,
hoop spacing was decreased to strengthen concrete with additional confinement. Except for the
control specimens, the panel zone region of the beam-column joints was constructed with ECC
and extended 203.2 mm and 304.8 mm to the adjoining columns and beams, respectively. A 457.2
x 457.2 x 12.7 mm steel plate was placed on top of the column to transfer loads from the testing
unit to the specimen. Additionally, A 304.8 x 304.8 x 12.7 mm steel plate was used at the bottom
of the column to connect the specimen to the universal joint at the base. Four high strength bolts
with 152.4 mm length were welded to each of the steel plates to prevent slipping between the steel
plates and the specimen as well as to achieve load transfer. All beams and columns had a 12.7 mm
clear cover.
146
Figure 6.2: Beam-Column subassemblies (shaded areas show the location of ECC, dimensions are in
mm): (A) out-of-plane view of the corner and exterior joint subassemblies, (B) in-plane view of corner
subassembly, (C) in-plane view of the exterior subassembly.
Table 6.1 presents the test matrix. Different variables were considered in the test matrix, including
loading protocol and presence of shear reinforcement in the panel zone area. In each corner and
exterior beam-column joint, three subassemblies with a RECC joint were considered in addition
147
to the control specimen. Axial load and biaxial bending were included in the loading scenarios of
the four specimens. Furthermore, the panel zone shear reinforcement was removed in one of the
subassemblies to investigate the effectiveness of superior shear resistance of ECC in reducing
reinforcement congestion in the panel zone area. Since the perimeter joints are usually subjected
to biaxial bending and torsion during an earthquake, one specimen was tested under torsional
loading simultaneously with biaxial bending.
Table 6.1: Test matrix for beam-column joints.
Subassembly
type
Name
Joint
material
Axial
load
Biaxial
bending
Torsion
Shear
reinforcement
in panel zone
Corner
C/RC/ABB/RS RC ×
C/RECC/ABB/RS RECC ×
C/RECC/ABB/NS RECC × ×
C/RECC/ABBT/RS RECC
Exterior
E/RC/ABB/RS RC ×
E/RECC/ABB/RS RECC ×
E/RECC/ABB/NS RECC × ×
E/RECC/ABBT/RS RECC
In naming the specimens, the first part shows the subassembly type (C for corner and E for
exterior), the second part shows the joint material (RC for conventional reinforced concrete, RECC
for reinforced ECC), the third part shows column loading (A for axial, BB for biaxial bending, and
T for torsion), and the last part presents shear reinforcement type in the panel zone area (RS for
regular stirrups and NS for no stirrups).
148
6.3. Testing the Specimens
6.3.1 Test setup and instrumentation
Figure 6.3 presents the boundary conditions for the exterior and interior beam-column
subassemblies. As mentioned before, the beam-column subassemblies were taken to extend from
the midpoint of the adjoining beams and columns to the panel zone. Since the corresponding
moments from lateral loads are almost zero at these points, they are typically modeled and tested
as pin or roller supports.
Figure 6.3: Boundary condition of the beam-column subassemblies: (A) exterior, (B) corner.
149
Figure 6.4(A) presents the test setup and instrumentation. A universal joint was used at the bottom
of the column to satisfy the condition of zero moment about all three axes. The universal joint was
connected to a 508 x 508 x 57 mm steel plate using four 12.7 mm high strength bolts. Then, the
steel plate was connected to the strong floor by eight high-strength threaded rods. A 305 x 305 x
12.7 mm steel plate was welded to top of the universal joint to facilitate connecting the specimens.
Similarly, 12 high strength post-tensioned bolts were used to transfer the loads from the testing
unit to the specimen. A rigid link with universal joints was employed at the free end of each beam
to release the moments. Additionally, top and bottom of the column were confined externally using
steel plates clamped together to apply external confinement to prevent concrete failure under stress
concentration. It is noted that these regions are expected to remain elastic due to zero moments
and not critical to the test results; however, the concern is that the concrete might crack at the steel-
concrete interface due to axial loads. Finally, a restraint was added to each end of the beams as
shown in Figure 6.4(B), to prevent the out-of-plane movement of the beams and rigid body motion
of the subassembly due to torsional loads. These restraints were achieved by placing a cantilever
steel column mounted to the strong floor. A half-spherical surface was used at the contact point of
the beam and the cantilever steel column allowing rotation of the beam around the Y-axis in in-
plane beam/s and around the X-axis in the out-of-plane beam. A 6.4 mm steel plate, which was
glued to the beam at the contact point, distributed the load and prevented localized crushing of
concrete at the contact points.
A tension-compression load cell was utilized in the middle of each rigid link to continuously
measure the reaction at the supports during the test. Additionally, a donut-shaped compression-
only load cell was considered between each fixed support and the corresponding beam to measure
the lateral reaction during the test. Several string pots (18 and 15 for exterior and corner
150
subassemblies, respectively) were attached to the specimen to measure its global deformation at
different locations. To measure the shear deformation of the joint panel zone, four linear pots were
used in a rectangular form (see Figure 6.4(C)). Additionally, a non-contact measurement system
was used to measure the deformations of specific points in 3D space. The non-contact
measurement system tracks the movement of light emitting diodes (LEDs) using three cameras
with known distance and orientation with respect to each other. The results can be used for
measuring global and relative displacements of the targets in 3D space and with respect to each
other. Figure 6.5 displays approximate location of the LEDs on the corner and exterior beam-
column subassemblies. It should be noted that the out-of-plane beam of the exterior beam-column
subassemblies had a LED pattern same as Figure 6.5(A). This configuration resulted in 5 LEDs in
the in-plane beam, 5 LEDs in the out-of-plane beam, 10 LEDs in the top column, and 10 LEDs in
the bottom column.
151
Figure 6.4: Illustration of a test setup for beam-column joints: (A) boundary conditions and
instrumentation, (B) detail of torsional deformation restrainer, (C) instrumentation to measure shear
deformation in the panel zone.
Figure 6.5: Approximate location of the LEDs on: (A) corner, (B) exterior beam-column joints
(dimensions are in mm).
51
38
76
203
203
64 102
305 203
203
102
305
In-Plane/Out-of-Plane Beam
In-Plane Beam 1
ECC
Top Column
Concrete
Bottom Column
Top Column Bottom Column
(A) (B)
In-Plane Beam 2
64
51
38
76
51
38
76
51
38
76
64 64
152
6.3.2 Loading protocol
A biaxial lateral loading was applied at the top of the column, following a cloverleaf loading path
in displacement control. The cloverleaf displacement path, which is shown in Figure 6.6(A), is
described by Equation (6) in the polar coordinate system where R is the maximum target radial
displacement at an angle of 45
o
in the X-Y plane, see Figure 6.3 for global X and Y directions.
=g=h"12g (6)
The specimen was subjected to one complete cycle of the clover-shape at each specific drift level.
The drift level was calculated by dividing R from Equation (6) by the height of the column, where
the height of the column is taken from the center of the pin support at the base to the loading point
on top. During each complete cycle, the specimen was exposed to one excursion into each area of
the 2-D coordinate system. Figure 6.6(B) and Figure 6.6(C) represent the loading protocol in the
X- and Y- directions (the directions are shown in Figure 6.3), respectively.
153
Figure 6.6: Bidirectional bending loading protocol: (A) position of the top of the column in X-Y plane,
(B) column deformation and drift in X-direction, (C) column deformation and drift in Y-direction.
In addition to the lateral displacement, 89 kN constant axial load was applied to the top of the
column, resulting in an axial load approximately equal to 7.5% of squash capacity of the RC
control column. To apply the torsional load, the corresponding degree of the freedom (θz) was
restrained, which led to a variable torque during the test.
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
-60
-40
-20
0
20
40
60
Step
Y-Deformation (mm)
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
-4.60 (-6.53)
-3.30 (-4.67)
-2.60 (-3.73)
-1.98 (-2.8)
-1.32 (-1.87)
-0.99 (-1.4)
-0.32 (-0.45)
0
0.32 (0.45)
0.99 (1.4)
1.98 (2.8)
2.64 (3.73)
3.30 (4.7)
4.62 (6.53)
Y-Drift (Total-Drift) (%)
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
-60
-40
-20
0
20
40
60
X-Deformation (mm)
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
-4.60 (-6.53)
-3.30 (-4.67)
-2.60 (-3.73)
-1.98 (-2.80)
-1.32 (-1.87)
-0.99 (-1.40)
-0.32 (-0.45)
0
0.32 (0.45)
0.99 (1.40)
1.98 (2.80)
2.64 (3.73)
3.30 (4.70)
4.62 (6.53)
Step
X-Drift (Total-Drift) (% )
-5 -4 -3 -2 -1 0 1 2 3 4 5
-5
-4
-3
-2
-1
0
1
2
3
4
5
X-Drift (%)
Y-Drift (%)
R
r
θ
(A)
(B)
(C)
154
6.4. Concrete and ECC Material Properties
As mentioned before, the concrete was ordered from a ready-mix concrete plant. The maximum
aggregate size and targeted 28 days compressive strength of the mixture were 9.5 mm and 31 MPa,
respectively. Additionally, the standard slump test of the mixture showed 76.2 mm slump at the
time of casting.
The same fly ash and silica sand used in the experimental program of the bridge columns were
adopted in the mixture of ECC for the experimental program of the beam-column joints. However,
oil-coated Kuraray PVA fibers (RECS15) [85] were employed, instead of regular PVA fibers, to
improve the ductility of the mixture. Table 6.2 list properties of the oil-coated PVA fibers.
Table 6.2: Properties of oil-coated PVA fibers [85].
Length
(mm)
Diameter
(µm)
Specific
gravity
Tensile
strength
(MPa)
Elongation
Young’s
modulus
(GPa)
8 40 1.3 1600 6% 41
Additionally, Hydroxypropyl Methylcellulose (HPMC) was added to the mix design to increase
the viscosity of the fresh mixture and improve the homogeneous distribution of the fibers. Table
6.3 provides the mixture proportions of the ECC mixture.
155
Table 6.3: ECC mixture design for beam-column joints.
Fraction by weight
Ingredient
Portland cement, c 1
Fly ash, fa 2.5
Silica sand, ss 1.16
Water, w 0.87
HRWR
a
0.02
PVA
b
fiber 2%
c
HPMC
d
0.001
Ratios
w/c 0.87
w/(c+fa) 0.25
w/solids
e
0.18
(c+fa)/solids 0.75
a
High range water reducer
b
Polyvinyl alcohol
c
By volume
d
Hypromellose Hydroxypropyl Methylcellulose
e
Solids=c + fa + ss
Table 6.4 summarizes the averaged results from the uniaxial compressing tests on 101.6 x 203.2
mm cylinders at the time of testing the beam-column subassemblies. The testing age is provided
inside parenthesis in days. Figure 6.7 presents the resulted stress-strain diagrams from the uniaxial
compression test.
Table 6.4: Compressive strength of ECC and concrete mixtures.
Subassembly
Type
Specimen
Columns and In-Plane
Beams
Out-of-Plane Beams
Concrete
(MPa)
ECC
(MPa)
Concrete
(MPa)
ECC
(MPa)
Corner
C/RC/ABB/RS 40.0 (52) N. A.
a
38.3 (52) N. A.
C/RECC/ABB/RS 35.8 (17) 17.4 (34) 35.8 (17) 31.7 (49)
C/RECC/ABB/NS 31.4 (21) 29.9 (38) 31.4 (21) 31.7 (49)
C/RECC/ABBT/RS 34.9 (34) 32.9 (49) 34.9 (34) 31.7 (49)
Exterior
E/RC/ABB/RS 33.6 (74) N. A. 49.2 (67) N. A.
E/RECC/ABB/RS 29.8 (46) 36.8 (56) 29.8 (46) N. A.
E/RECC/ABB/NS N.A. 49.7 (64) N. A. N. A.
E/RECC/ABBT/RS N.A. 40.6 (77) N. A. N. A.
a
Not available or not applicable
Numbers inside the parenthesis show the testing age in days
156
Figure 6.7: Stress-strain diagrams from uniaxial compression test: (A) conventional concrete, (B) ECC.
(A)
(B)
0.0 0.5 1.0 1.5 2.0 2.5
0
10
20
30
40
Stress (MPa)
Strain (%)
0.0 0.5 1.0 1.5 2.0 2.5
0
10
20
30
40
Stress (MPa)
Strain (%)
157
6.5. Construction Process
Special effort was made to maintain consistency in constructing the specimens. Additionally, high-
quality materials were used, along with employing professional workers to ensure excellence in
the final specimens. This section describes the construction process step-by-step.
6.5.1 Formwork
To help with the construction process and increase the final quality of the joint subassemblies, it
was decided to cast the columns and in-plane beam (or beams for the exterior joints) in a horizontal
configuration while casting the out-of-plane beam vertically. Figure 6.8 shows the assembled
formworks for corner beam-column subassemblies. The formworks were custom designed and the
fabrication was made at a commercial wood shop. In the fabrication of the formworks, high quality
coated plywood was used to prevent water absorption from concrete and ease the demolding
process. After preparing all the pieces, the formworks of horizontal elements were assembled on
custom-made wooden pallets covered with plastic sheets to improve mobility. Then, the
formworks were fastened to the wooden pallet using several wood ties. It is worth mentioning that
the formwork for the out-of-plane beam was assembled after casting the horizontal sections.
158
Figure 6.8: Assembled formwork of the corner beam-column subassemblies.
6.5.2 Rebar cages
All the 9.5 mm rebar (#3 U.S. designation rebar) was purchased from the same heat of steel for
consistency in the material properties. Based on the designed configurations, both the longitudinal
reinforcement and the stirrups were purchased from a rebar supplier in bent form whenever
applicable. Then, all the reinforcement was shipped to a concrete precast plant and the rebar cages
were tied by professional ironworkers following the industry practice, see Figure 6.9(A). After
delivering the rebar cages to the lab, the longitudinal reinforcements were trimmed to remove extra
parts, see Figure 6.9(B). Then the rebar cages were placed inside the formworks with 12.7 mm
spacers underneath to provide the cover, see Figure 6.9(C).
159
Figure 6.9: Preparing the rebar cages: (A) tying the rebar cages, (B) trimming the longitudinal
reinforcements, (C) placing the rebar cages inside the formwork.
6.5.3 Casting ECC sections
Custom made wooden dividers were fabricated and coated with oil based paint to prevent water
absorption from ECC mix. Then, the dividers were fastened to the formwork with the help of
screws and wood ties and sealed to prepare a closed area to cast the ECC section. Subsequently,
the formworks were strengthened against out-of-plane deformations using several wooden braces
and sealed by caulk and duct tape. To prepare the ECC mix, at first, all the dry materials were
mixed for around three minutes in a shear type mixer. Then, about one-third of the water was
gradually added to the dry material mix, after which, the superplasticizer and the remaining water
was gradually added to prepare a uniform matrix. Finally, the fibers were added gradually to the
mix and the mixer was kept running during the mixing procedure to prevent material balling. Upon
preparing the ECC mix, the ECC section was cast without any vibration. The cast region was
hammered by a plastic hammer to release trapped air bubbles and covered with plastic sheets for
160
the curing process at room temperature inside the lab. Figure 6.10 (A) shows the wooden dividers
and the braces, Figure 6.10 (B) to Figure 6.10 (D) illustrates preparation of the ECC mixture, and
Figure 6.10 (D) presents the final casted horizontal ECC section.
Figure 6.10: Casting the horizontal ECC Section: (A) wooden dividers and braces, (B) mixing dry
materials for ECC, (C) uniform matrix of ECC, (D) adding fiber to ECC matrix, (E) casting horizontal
ECC section
The vertical ECC section was cast one day after casting the horizontal section when a hardened
ECC was formed underneath. To cast the vertical ECC section, first, the corresponding formwork
was placed and fastened to the horizontal formwork using several wood ties, see Figure 6.11(A).
Then, the ECC mixture was cautiously poured inside the formwork, to reach the desired height,
see Figure 6.11 (B). Finally, the trapped air bubbles were released by hammering the formwork
and plastic sheets were used to prevent water evaporation from the mix, see Figure 6.11(C).
161
Figure 6.11: Casting the vertical ECC section: (A) placing and sealing the formwork, (B) pouring ECC,
(C) covering and curing the cast ECC sections.
6.5.4 Steel end-plates
The steel plates were considered at the top and bottom of the column to connect the beam-column
subassembly to the testing unit and the base universal joint, respectively. All the end- plates were
fabricated from 12.7 mm thickness normal steel plates. After cutting the plates into the proper
sizes, the hole patterns were drilled to match the corresponding hole patterns on either the testing
unit or the universal joint, see Figure 6.12 (A) and (B). Then, four 19 mm high strength bolts were
welded to each plate as studs, see Figure 6.12 (C) and (D). The distance between the bolts was
calculated based on the minimum requirements for studs as well as by considering the location of
the longitudinal and transverse reinforcement in the section. Each end-plate was centered and
connected to the formwork using four 9.5 mm normal bolts that passed through the holes designed
for this purpose, see Figure 6.12 (E). Figure 6.12 (F) shows an example of the installed end-plate.
162
Figure 6.12: Preparing and installing the end-plates: (A) top end-plate with the corresponding hole
pattern, (B) bottom end-plate with the corresponding hole pattern, (C) preparation to weld the high
strength bolts, (D) end-plate with welded bolts, (E) centering the end-plate and drilling holes in the
formwork, (F) installed end-plate.
6.5.5 Casting conventional concrete
To cast the rest of the columns and beams with conventional concrete, the wooden dividers were
removed and the surface of the ECC section was cleaned to maximize the bonding between the
conventional concrete and ECC. Figure 6.13 displays casting the specimens with conventional
concrete. In view of the endplates at the top and bottom of the column, the open ends of the in-
plane beam/beams were blocked with wooden pieces to form a closed area for casting. Then, the
perimeter of the formwork and all the blocked ends were sealed using caulk and duct tape to
prevent water loss from concrete. Additional wooden braces were implemented to strengthen the
formwork against lateral deformations during casting. Following the preparation of the formworks,
all the specimens were moved outside of the lab and cast simultaneously using the same batch of
163
the concrete from a ready-mix concrete truck. For each section, concrete was poured in several
layers and vibrated using vibrators to increase the quality of the cast concrete. It should be noted
that only the horizontal sections of the control RC specimens, were simultaneously cast with the
concrete sections of the RECC specimens. The out-of-plane beams of the control specimens were
cast two days later with a concrete mix prepared at the lab.
Figure 6.13: Casting the conventional concrete sections: (A) the prepared specimens, (B) pouring
concrete and vibrating, (C) horizontal section of the control specimen for exterior beam-column joints.
6.5.6 Curing and preparing the specimens
The exposed areas of the cast specimens were covered with waterproofed membranes immediately
after finishing the casting process and the specimens were moved inside the lab or under a covered
area. The surfaces of the specimens were kept wet for few hours by pouring water and then covered
with several layers of burlap. The burlap layers were continuously immersed in water and covered
with plastic sheets to improve curing process, see Figure 6.14(A). The specimens were kept inside
the formwork and under the described curing process for more than two weeks. Finally, the
formwork was removed and the surface of the specimens was cleaned and painted as the last
164
preparation step before testing. For testing, each specimen was flipped to make the columns
vertical and to mount to the test setup, see Figure 6.14(B).
Figure 6.14: Preparing the specimens: (A) curing, (B) flipping and mounting on the test setup.
6.6. Experimental Results and Discussion
6.6.1 Cyclic behavior
Figure 6.15 presents the load-drift hysteresis curves of the tested beam-column subassemblies. As
evident, the corner beam-column subassemblies returned comparable results in X and Y directions
through symmetric boundary condition.
165
Figure 6.15: Beam-column subassemblies load-deformation hysteresis: (A) C/RC/ABB/RS, (B)
E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,
(G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
166
A comparison between the results shows that the general behavior of the beam-column
subassemblies remained the same by replacing concrete with ECC in the panel zone and plastic
hinge region of the adjoining beams and columns. As mentioned before, yielding of the
longitudinal reinforcement mainly contributes towards the permanent deformation of the RC or
RECC beam-column joints. Since the longitudinal reinforcement ratio was constant, the
corresponding permanent deformations after each cycle remained close to each other in all the
specimens. Interestingly, load-deformation diagrams of C/RECC/ABB/RS, C/RECC/ABB/NS,
and C/RECC/ABBT/RS were very similar to each other, which can be attributed to the high shear
resistance of ECC. The same trend was observed in the exterior beam-column subassemblies
incorporated RECC in the panel zone and plastic hinge region of the adjacent beams and columns,
i.e. between E/RECC/ABB/RS, E/RECC/ABB/NS, and E/RECC/ABBT/RS specimens. As
mentioned before, torsional load increases the shear demand on the column and the joint; since
ECC had considerable shear strength, the beam-column subassemblies with RECC joints showed
similar behavior when subjected to the loading scenarios with and without torque. Additionally,
removing shear reinforcement in the panel zone area had a negligible effect on the overall behavior
and failure mode of the corner or exterior beam-column subassemblies.
Table 6.5 summarizes the maximum lateral strength of the beam-column subassemblies I both X
and Y-directions. The values inside the parenthesis in each cell indicate differences (in percentage)
with respect to the control specimen in each subassembly type, corner and exterior. A negative
value inside the parenthesis means reduction while a positive value indicates an increase.
167
Table 6.5: Maximum lateral strength of beam-column subassemblies
Specimen
X - Direction Y-Direction
Max. in +X (kN) Max. in -X (kN) Max. in +Y (kN) Max. in -Y (kN)
C/RC/ABB/RS 19.93 (0.00) 21.75 (0.00) 20.60 (0.00) 21.93 (0.00)
C/RECC/ABB/RS 20.19 (1.30) 23.00 (5.75) 23.84 (15.73) 25.18 (14.82)
C/RECC/ABB/NS 19.87 (-0.30) 23.76 (9.24) 23.16 (12.43) 25.50 (16.28)
C/RECC/ABBT/RS 24.06 (20.72) 25.40 (16.78) 26.82 (30.19) 25.89 (18.06)
E/RC/ABB/RS 21.07 (0.00) 34.99 (0.00) 34.00 (0.00) 35.83 (0.00)
E/RECC/ABB/RS 21.20 (0.62) 34.08 (2.60) 37.76 (11.06) 37.54 (4.77)
E/RECC/ABB/NS 22.34 (6.03) 33.70 (3.69) 37.85 (11.32) 40.82 (13.93)
E/RECC/ABBT/RS 23.33 (10.73) 35.50 (1.46) 42.12 (23.88) 42.15 (17.64)
It can be observed that by replacing the conventional concrete with ECC in the panel zone area,
the ultimate strength capacity of the beam-column subassemblies was increased. In the specimens
without torsional loads, strength enhancement was observed in both corner (maximum of 9.24%
and 16.28% in X and Y-directions, respectively) and exterior joints (maximum of 6.03% and
13.93% in X and Y-directions, respectively). By applying torque to the specimens, higher strength
improvement (30.19% and 23.88% for corner and exterior beam-column subassemblies,
respectively) was observed since the rigidity of the specimen increased through restricting the twist
and the loading system contributed in carrying the applied lateral force.
168
6.6.2 Crack patterns
Figure 6.16 provides the crack patterns of the tested specimens. In general, the damage in the
exterior beam-column joints was comparatively less than those in the corner ones because of the
superior confinement in the panel zone area, which resulted from the additional in-plane beam. A
comparison of the observed damage at the final stage revealed that the specimens constructed with
RECC showed finer and more distributed cracks compared to the control RC specimens. Unlike
the RECC beam-column joints, large main cracks were observed in the control RC specimens
where damage was concentrated. Interestingly, the damage level of the RECC joint without shear
reinforcement in the panel zone was significantly moderate than the control RC specimen in both
corner and exterior beam-column subassemblies indicating that ECC could superiorly resist the
shear forces even without transverse reinforcement. Additionally, no major difference was
observed in the damage pattern of the RECC specimens with and without shear reinforcement in
the panel zone area. Usually, panel zone area of RC beam-column joints has high reinforcement
ratios to satisfy seismic requirements and prevent shear failure of the joint. The results from the
experimental work illustrated that shear resistance of ECC was adequate in preventing shear failure
of beam-column joints in the absence of any shear reinforcement, at least for the configurations
studied in this thesis. This characteristic is particularly important in easing the repair process
aftermath of a major earthquake in addition to facilitating the construction process by reducing the
reinforcement conjunction in the panel zone area.
Diagonal shear cracks were formed in the panel zone areas while flexural cracks were observed in
the beams and columns about both weak and strong axes. More cracks appeared in the beams than
the columns, confirming the strong column-weak beam condition in the subassemblies. As evident,
the cracks were extended a longer distance in the beams than in the columns since the longitudinal
169
reinforcement was distributed only at the top and bottom of the beams which created a weak axis.
This distribution of the longitudinal reinforcement in the beams decreased the cracking moment
about the weak axis which could be satisfied in a closer distance to the fixed supports where the
beams were loaded in cantilever form about the weak axis. Unfortunately, the paint was wet at the
testing time of the E/RC/ABB/RS specimen, see Figure 6.16(B), which concealed some of the
cracks. Unlike the control specimens, no spalling was observed in the RECC specimens despite
more complex loading scenarios. The RECC panel zone suffered little damage under biaxial
bending and torsion, which is desired to minimize the damage in perimeter joints of buildings due
to strong earthquakes as discussed in Section 2.5. Some inclined cracks due to shear resulting from
applied torque occurred in the RC sections of C/RECC/ABBT/RS, see Figure 6.16(G), while the
RECC portions showed flexural cracks, which can be related to the superior shear resistance of
ECC compared to conventional concrete. Formation of cracks in the RC section of the RECC
specimens revealed that the bonding between ECC and conventional concrete was strong enough
to diminish the effect of the cold joints formed during the construction process.
170
Figure 6.16:Crack pattern of the tested beam-column subassemblies: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D)
E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
171
6.6.3 Torsion
Figure 6.17 presents the applied torque to C/RECC/ABBT/RS and E/RECC/ABBT/RS specimens.
Considerably larger torsion was applied to C/RECC/ABBT/RS than E/RECC/ABBT/RS because
of the symmetric geometry and boundary condition of E/RECC/ABBT/RS specimen and the
additional resisting element, i. e., the additional in-plane beam.
Figure 6.17: Torsion: (A) C/RECC/ABBT/RS, (B) E/RECC/ABBT/RS.
0
250
500
750
1000
1250
1500
1750
2000
2250
-6
-4
-2
0
2
4
6
6.5%
4.7%
3.7%
Torque (kN.m)
Step
1.4%
2.8%
(A)
(B)
0
250
500
750
1000
1250
1500
1750
2000
2250
-6
-4
-2
0
2
4
6
6.5%
4.7%
3.7%
Torque (kN.m)
Step
1.4%
2.8%
172
6.6.4 Shear deformation
The shear deformation in the panel zone area of the tested specimens was determined using the
data obtained from the linear potentiometers placed in a square form, see Figure 6.4(C). To
calculate the shear deformations, it was assumed that the configuration of the potentiometers
formed a parallelogram during the test. Considering the parallelogram, see Figure 6.18, shear
deformation was calculated as the variation of g where b, d1, and d2 were measured using the
potentiometers and g was calculated using Equation (7) and Equation (8).
Figure 6.18: Position of the potentiometers.
5=
i2
2
I
+2
I
I
−4
I
2
(7)
g=cos
o2
2
I
−5
I
−
I
25
(8)
Figure 6.19 presents the shear deformations in the panel zone of the tested beam-column
subassemblies. The shear deformation was measured in Y-direction, therefore it followed the
corresponding applied lateral deformation. In general, the exterior beam-column subassemblies
showed larger shear deformations than the corner beam-column subassemblies because of the
additional pin support connected to the second in-plane beam.
173
Figure 6.19: Shear deformation: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D)
E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H)
E/RECC/ABBT/RS.
174
As a consequence of the larger cracks in the panel zone area, the shear deformations were
considerably larger in E/RC/ABB/RS compared to E/RECC/ABB/RS, illustrating the superior
shear resistance of ECC. Furthermore, the shear deformations in C/RECC/ABB/NS and
E/RECC/ABB/NS were significantly larger than C/RECC/ABB/RS and E/RECC/ABB/RS,
respectively. Removing the stirrups from the panel zone area of the studied beam-column joints
significantly increased the shear deformation, despite the negligible effect on the global load-
deformation behavior. Since torsion appears as shear forces in the beam-column joints, applying
torque to the tested beam-column subassemblies increased the shear forces and the analogous shear
deformations, see Figure 6.19 (G).
6.6.5 Reactions at boundaries
Figure 6.20 and Figure 6.21 present the resulted forces in the joint of corner and exterior beam-
column subassemblies, respectively, due to the boundary condition presented in Figure 6.3.
Figure 6.20: Shears, moments, and axial loads in the corner beam-column joints.
175
Figure 6.21: Shear, moments, and axial load in the exterior beam-column joints.
Figure 6.22 and Figure 6.23 provides reaction(s) of the pin support(s) in Y-direction (fs1 in Figure
6.20 and fs1 and fs3 in Figure 6.21) and X-direction (fs2 in Figure 6.20 and Figure 6.21), respectively.
The reactions followed the corresponding lateral displacement applied to the top columns because
of the relationship with the shear forces at the top and bottom columns, which are provided in
Equation (9) and Equation (10).
2
4
2
+
J
4
J
=9
e
p
2
+9
e
p
I
(9)
I
4
I
=9
S
p
2
+9
S
p
I
(10)
Since fs1 and fs3 created moments with the same sign, fs1 showed lower values in the exterior beam-
column joints than the corner ones. Even though the reactions of the pin support increased up to
step 750 (2.8% total drift), they remained almost constant for the rest of the test conveying that the
ultimate capacity of the beams was reached when the column was subjected to 2.8% lateral drift.
176
Figure 6.22: Reactions at zero moment location of the in-plane beams at the pin supports: (A)
C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
Force (kN)
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
f
s3
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
f
s3
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
f
s3
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
Force (kN)
(A) (B)
(C) (D)
(E) (F)
(G) (H)
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
Force (kN)
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
Force (kN)
Step
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s1
f
s3
Step
177
Figure 6.23: Reactions at zero moment location of the out-of-plane beams at the pin supports: (A)
C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F)
E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
Step
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
Force (kN)
Step
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
Force (kN)
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
Force (kN)
0
250
500
750
1000
1250
1500
1750
2000
2250
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
f
s2
Force (kN)
(A) (B)
(C) (D)
(E) (F)
(G) (H)
178
Figure 6.24 presents the recorded reactions by the donut load cells mounted to the fixed supports,
fd1 and fd2 in Figure 6.20 and fd1 and fd3 in Figure 6.21. The main purpose of the fixed supports was
preventing any global instability in the system due to the applied torque. The relationship between
fd1, fd2, and fd3 with the applied torque to the subassemblies are provided in Equation (11) and
Equation (12) for the corner and exterior beam-column subassemblies, respectively.
*2
4
2
+L=
*I
4
I
+L′ (11)
*2
4
2
+L=
*J
4
J
+L′ (12)
These supports restrained the free motion of the in-plane beam in negative X-direction and the out-
of-plane beam in positive Y-direction, in the corner beam-column subassemblies. Accordingly,
the free motion of the in-plane beams was restrained in negative X-direction in the exterior beam-
column subassemblies. The main peak points in the corresponding diagrams occured when both
of the fixed supports were restraining the free motion of the beams simultaneously. Considering
the cloverleaf loading path at top of the column, see Figure 6.6, this condition occurred once in the
corner subassemblies and twice in the exterior subassemblies. Reactions from the exterior
subassemblies were larger than the reaction in the corner ones since both the restrictions were in
the same direction, which made the subassembly stiffer. By applying torque to the system, the
second peak appeared in C/RECC/ABBT/RS and the peaks in E/RECC/ABBT/RS increased
revealing that the maximum torsion mostly followed the movement of the top column in negative
X-direction. Additionally, it can be observed that the second peak at each drift cycle has a smaller
amplitude than the first peak in the exterior beam-column subassemblies because of the damage
in the beams which occurred during the first cycle.
179
Figure 6.24: Reaction at the fixed supports: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C) C/RECC/ABB/RS,
(D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS, (G) C/RECC/ABBT/RS, (H)
E/RECC/ABBT/RS.
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d2
6.5%
4.7%
3.7%
Force (kN)
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d3
6.5%
4.7%
3.7%
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d2
6.5%
4.7%
3.7%
Force (kN)
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d3
6.5%
4.7%
3.7%
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d2
6.5%
4.7%
3.7%
Force (kN)
1.4%
2.8%
(A) (B)
(C) (D)
(E) (F)
(G) (H)
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d3
6.5%
4.7%
3.7%
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d2
6.5%
4.7%
3.7%
Force (kN)
Step
1.4%
2.8%
0
250
500
750
1000
1250
1500
1750
2000
2250
0
2
4
6
8
10
12
14
16
18
f
d1
f
d3
6.5%
4.7%
3.7%
Step
1.4%
2.8%
180
6.6.6 Deformations in beams and columns
The data from the non-contact measuring device was used to determine the movement of each
point with respect to the original position (hence linear deformation), and the corresponding angle
with respect to the neighboring points (hence rotational deformation). As mentioned before, the
non-contact measuring device could locate the position of the LEDs, which were attached to the
points of interest, in space during testing. Figure 6.25 illustrates the expected deformations in the
beams and columns of the subassemblies considering the boundary conditions.
Figure 6.25: Expected deformations in (A) out-of-plane beam of the exterior and in-plane and out-of-
plane beam of the corner beam-column subassemblies, (B) in-plane beams of the exterior beam-column
subassemblies.
To better understand the behavior of the tested beam-column subassemblies, several parameters
were defined and the corresponding maximum and minimum values were measured at the interest
points, i.e. location of LEDs. Figure 6.26 schematically illustrates the defined parameters for
deformations in the beams and columns. In naming the parameters, “C” refers to column, “IPB”
In-Plane Beam 1
In-Plane Beam 2
Top Column Bottom Column
In-Plane/Out-of-Plane
Beam
Top Column Bottom Column
Original Shape
Rigid Body Movement
Expected Deformation
(A) (B)
181
refers to in-plane-beam, and “OPB” refers to out-of-plane beam. The defined parameters for the
deformations are: deformation of the columns in X-direction, C\dX, and Y-direction, C\dY;
deformation of the in-plane beam in X-direction, IPB\dX, and Z-direction, IPB\dZ; and
deformation of the out-of-plane beam in Y-direction, OPB\dY, and Z-direction, OPB\dZ.
Figure 6.26: Schematic illustration of the defined parameters for the deformations (in naming, “C” refers
to column, “IPB” refers to in-plane-beam, and “OPB” refers to out-of-plane beam): (A) C\dX, (B) C\dY,
(C) IPB\dX, (D) IPB\dZ, (E) OPB\dY, (F) OPB\dZ.
The global distortions in the beams and columns were a combination of the rigid body motion and
the deformations in the beams and columns due to the applied loads. To eliminate the effect of the
182
rigid body motion, the slope between the two adjacent interest points were calculated. Comparing
the relevant deformation between the points is a measure to capture the level of damage in the
element. Several parameters were defined to measure the relevant deformation between two
interest points. Figure 6.27 schematically illustrates the defined parameters. In naming the
parameters, “C” refers to column, “IPB” refers to in-plane-beam, and “OPB” refers to out-of-plane
beam. The defined parameters for the relevant deformation between two interest points are: tangent
to the in-plane deformation of the column, C\θX, tangent to the out-of-plane deformation of the
column, C\θY, tangent to the in-plane deformation of the in-plane beam, IPB\θX, tangent to the out-
of-plane deformation of the in-plane beam, IPB\θZ, tangent to the in-plane deformation of the out-
of-plane beam, OPB\θY, and tangent to the out-of-plane deformation of the out-of-plane beam,
OPB\θZ.
183
Figure 6.27: Schematic illustration of the defined tangents to the deformation in the beams and columns
(in naming, “C” refers to column, “IPB” refers to in-plane-beam, and “OPB” refers to out-of-plane
beam): (A) C\θ X, (B) C\θ Y, (C) IPB\θ X, (D) IPB\θ Z, (E) OPB\θ Y, (F) OPB\θ Z.
Considering the distance between the interest points and the movement of each point in space
(refer Figure 6.28 for an example), the defined parameters for relevant deformation of two adjacent
points were calculated using Equation (13) to Equation (18).
184
-\g
r
=tan
o2
-\v
w2
−-\v
5
(13)
-\g
x
=tan
o2
-\y
w2
−-\y
5
(14)
z{7\g
r
=tan
o2
z{7\|
w2
−z{7\|
(15)
z{7\g
}
=tan
o2
z{7\v
w2
−z{7\v
(16)
~{7\g
x
=tan
o2
~{7\|
w2
−~{7\|
(17)
~{7\g
}
=tan
o2
~{7\y
w2
−~{7\y
(18)
Figure 6.28: An example of the movements of the interest points and the distance between them.
z{7/|
-/v
5
Z
X
Y
-/v
w2
z{7/|
w2
185
Figure 6.29 and Figure 6.30 present the maximum and minimum measured C\dX and C\dY,
respectively. It is to be noted that the zero point of the coordinate system was considered at the
center of the top column where the lateral displacements were applied. As it can be seen, the
column acted as a cantilever column, as expected, and showed a higher deflection at the top than
the bottom. In the corner subassemblies, the absolute value of the minimum C\dX was higher than
maximum C\dX because of the opening at the cold joint of the out-of-plane beam, which reduced
its resistance. Nevertheless, the fixed supports in the exterior subassemblies restricted deformation
of the joint in negative X-direction. The measured C\dX and C\dY are the combinations of the
deformations in the column and the movement of the pin supports. As the RECC specimens
restored their rigidity at higher drift ratios, higher movement in the pin supports was observed
compared to RC specimens, which led to higher C\dX and C\dY in these specimens. Generally,
exterior subassemblies showed more symmetric behavior than corner subassemblies because of
the symmetric geometry and boundary condition. Since there was limited restrains against rotation
of the corner specimens in the second leaf of the cloverleaf loading protocol (positive X and
negative Y), C/RECC/ABBT/RS could rotate due to torsion, which led to smaller deformations in
higher drift ratios.
186
Figure 6.29: Maximum and minimum measured C\dX (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
Z
X
Y
Z
X
Y
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
max 2.8%
min 2.8%
max 3.7%
min 3.7%
max 4.7%
min 4.7%
max 6.5%
min 6.5%
(A) (B)
(C) (D)
(E) (F)
(G)
(H)
187
Figure 6.30: Maximum and minimum measured C\dY (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
Z
X
Y
Z
X
Y
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
400
600
800
1000
Deformation (mm)
Location (mm)
max 2.8%
min 2.8%
max 3.7%
min 3.7%
max 4.7%
min 4.7%
max 6.5%
min 6.5%
(A)
(B)
(C)
(D)
(E) (F)
(G)
(H)
188
Figure 6.31 and Figure 6.32 present the maximum and minimum measured IPB\dX and IPB\dZ,
respectively. IPB\dX was further related to the movement of the pin support at zero moment
location of the out-of-plane beam and the reaction of the fixed support. Since there was no
restriction in movement of the exterior beam-column subassemblies in positive X-direction, the
measured maximum IBP\dX was higher than the absolute value of the measured minimum IPB\dX
in these specimens. Additionally, the in-plane-beams in the exterior beam-column subassemblies
showed almost a rigid body motion in positive X-direction. However, the in-plane-beam was
twisted in the corner subassemblies because of the asymmetry in the boundary condition.
C/RC/ABB/RS showed almost zero minimum IBP\dX at 6.5% drift, indicating that the concrete
spalling shifted the joint reaction to an upper portion of the top column.
IBP\dZ was a combination of the deformations in the in-plane beam and movement of the pin
supports. The vertical component of the pin support decreased as the column moves in the positive
Y-direction and pushed the end of the in-plane beam downward. Therefore, IBP\Z decreased as
the distance from the column increases. The exterior beam-column subassemblies presented
symmetric IPB\dZ due to the symmetric geometry and boundary condition in Y-direction, unlike
the corner specimens. C/RECC/ABB/NS displayed the highest IPB\dZ between the specimens
since a large crack was developed at the interface of the in-plane beam and the column, which
eased the corresponding deformations. C/RECC/ABBT/RS presented smaller absolute IPB\dZ
than C/RECC/ABB/RS, which can be related to the out-of-plane motion of the pin support because
of the rigid body motion of the in-plane beam due to the applied torque, which further decreased
the vertical component of the pin support. Interestingly, the movement of the last interest point in
the in-plane beam, which was located after the cold joint, was in line with the rest of the points,
189
indicating that the cold joint between ECC and concrete had a negligible effect on the performance
of the in-plane beam.
Figure 6.33 and Figure 6.34 show the maximum and minimum measured OPB\dY and OPB\dZ,
respectively. The exterior beam-column subassemblies presented symmetric OPB\dY since there
was no restriction to limit deformation of the out-of-plane beam in Y-direction. However, the
corner beam-column subassemblies exhibited larger values in negative Y-direction than positive
Y-direction, as expected. Replacing conventional concrete with ECC had a negligible effect on
OPB\dY because of the reaction of the out-of-plane beam, with respect to the weak axis of the
section, which caused comparable damage to both conventional concrete and ECC. Nearly straight
lines of OPB\dZ in different drift levels indicated that RECC section of the out-of-plane beam had
a rigid body motion. This behavior can be related to either transferring the plastic hinge to the RC
section of the out-of-plane beam or opening of the cold joints. There were two cold joints through
the out-of-plane beam of the RECC subassemblies: (i) at the interface of the out-of-plane beam
and the column, and (ii) between the RC and RECC sections. Unsymmetrical diagram of OPB\dZ
in C/RC/ABB/RS strengthen the later assumption since there was only one cold joint along the
out-of-plane beam in this column.
190
Figure 6.31: Maximum and minimum measured IPB\dX (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
191
Figure 6.32: Maximum and minimum measured IPB\dZ (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
192
Figure 6.33: Maximum and minimum measured OPB\dY (shown in (G) and (H) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
193
Figure 6.34: Maximum and minimum measured OPB\dZ (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
194
Figure 6.35 and Figure 6.36 present the maximum and minimum measured C\θX and C\θY,
respectively.
Figure 6.35: Maximum and minimum measured C\θ X (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
Z
X
Y
Z
X
Y
-25 -20 -15 -10 -5 0 5 10 15 20 25
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
(A)
(B)
(C) (D)
(E) (F)
(G) (H)
195
Figure 6.36: Maximum and minimum measured C\θ Y (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS (note that Figure 6.36 (G) has different scale that others).
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
Z
X
Y
Z
X
Y
-15 -12 -9 -6 -3 0 3 6 9 12 15
400
600
800
1000
Rotation (degree)
Location (mm)
-5 -4 -3 -2 -1 0 1 2 3 4 5
400
600
800
1000
Rotation (degree)
Location (mm)
(A)
(B)
(C)
(D)
(E)
(F)
(G)
(H)
196
In exterior beam-column subassemblies, the damage to the top and bottom columns were similar
in X-direction while the loading and unloading damage differed in Y-direction because of the
considerable shear and moment in this direction. C\θX and C\θY increased at 6.5% lateral drift ratio
due to the lack of stirrups in the panel zone area, which indicated that ECC could sufficiently
tolerate the shear loads up to 4.7% without shear reinforcement. Applying torque to
C/RECC/ABBT/RS shifted the deformations to the top column and caused considerably larger
C\θX and C\θY, compared to the other specimens.
Figure 6.37 and Figure 6.38 demonstrate the maximum and minimum measured IPB\θX and
IPB\θZ, respectively. It can be observed that most of the deformation was concentrated at the length
equal to the depth of the in-plane beam from the edge of the column. Therefore, the plastic hinge
was successfully formed in the RECC section, in an adequate length. In Figure 6.37 and Figure
6.38, the value of the last point (around 350 mm) is the angle between two LEDs at the two sides
of the cold joint in the in-plane beam. The corresponding value was almost constant at all the drift
levels illustrating the integrity of the connection between the RECC and RC sections. IPB\θZ
showed nearly constant values in different drift levels, representing a rigid body with respect to
the weak axis of the beam. However, an inconsistency was observed in regions close to the column
when torsion was applied to the subassemblies (Figure 6.38 (G) and (H)). This inconsistency can
be related to the concentration of deformations in the column rather than the beam in the presence
of torsional loads. Figure and Figure 6.40 present the maximum and minimum measured OPB\θY
and OPB\θZ, respectively.
197
Figure 6.37: Maximum and minimum measured IPB\θ X (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Location (mm)
Rotation (Degree)
Z
X
Y
Z
X
Y
(A) (B)
(C)
(D)
(E) (F)
(G) (H)
198
Figure 6.38: Maximum and minimum measured IBP\θ Z (shown in (G) and (H) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
-400 -350 -300 -250 -200 -150 -100
-5
-4
-3
-2
-1
0
1
2
3
4
5
Location (mm)
Rotation (Degree)
teta-x along the InplaneBeam-C/RECC/ABB/RS
-400 -350 -300 -250 -200 -150 -100
-5
-4
-3
-2
-1
0
1
2
3
4
5
Location (mm)
Rotation (Degree)
Z
X
Y
Z
X
Y
-400 -350 -300 -250 -200 -150 -100
-2
-1
0
1
2
Location (mm)
Rotation (Degree)
(A) (B)
(C)
(D)
(E) (F)
(G)
(H)
199
Figure 6.39: Maximum and minimum measured OPB\θ Y (shown in (A) and (B) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
200
Figure 6.40: Maximum and minimum measured OPB\θ Z (shown in (G) and (H) for corner and exterior
subassemblies, respectively) at different drift levels: (A) C/RC/ABB/RS, (B) E/RC/ABB/RS, (C)
C/RECC/ABB/RS, (D) E/RECC/ABB/RS, (E) C/RECC/ABB/NS, (F) E/RECC/ABB/NS,(G)
C/RECC/ABBT/RS, (H) E/RECC/ABBT/RS.
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
100 150 200 250 300 350 400
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
Location (mm)
Rotation (Degree)
Z
X
Y
Z
X
Y
max 2.8%
min 2.8%
max 3.7%
min 3.7%
max 4.7%
min 4.7%
max 6.5%
min 6.5%
(A) (B)
(C) (D)
(E) (F)
(G) (H)
201
CHAPTER 7 - CLOSURE
Improving resiliency of infrastructures against natural hazards is essential to protect the life of
civilians in addition to ease the recovery process and decrease the corresponding repair costs.
Earthquakes are a risk in a considerable portion of the world, which threatens human life and the
welfare of societies. Utilizing advanced materials has been in the consideration of researchers to
enhance the performance of structures and mitigate damage due to major seismic events. The focus
of the presented research is enhancing the seismic performance of bridge columns and beam-
column joints using high-performance materials, particularly, ECC and Cu-Al-Mn SEAs. To fulfill
these objectives, an innovative bridge design concept was developed and studied experimentally
and numerically. Additionally, the performance of corner and exterior beam-column joints
incorporating ECC was investigated through experimental work. Each part of the study was
deliberated extensively in previous sections. This chapter abridges the primary finding of the
research as bullet points and provides proposals for future research.
7.1. Conclusions
Chapter 3-Experimental program of bridge columns
• It has been demonstrated that the proposed innovative column design with a prefabricated
RECC hollow section and Cu-Al-Mn SEA bars in the plastic hinge region could lead to
improved bridge seismic performance by minimizing the permanent drift and repair needs,
thereby, could result in a functional transportation network in the aftermath of an
earthquake, and lower overall life-cycle costs.
• If all the steel rebar in the plastic hinge region of the columns are replaced with Cu-Al-Mn
SEA bars, the columns will be able to recover a remarkable 91% of the permanent
202
deformations that the conventional RC column showed after being subjected to a peak drift
of 7%. This resulted in a reduced energy absorption of approximately 36% of the control
specimen.
• If the plastic hinge steel rebar was only partially replaced with Cu-Al-Mn SEA bars, the
permanent drift reduced by approximately 33%; while, the energy absorption increased to
60% of the control specimen.
• The use of ECC, either on the entire cross-section or partially (without the Cu-Al-Mn SEA
bars) resulted in an increase in lateral strength of approximately 15%. Owing to lower
upper transformation (or yield) stress of Cu-Al-Mn SEA bars, the columns with the
complete and partial replacement of steel rebar with the Cu-Al-Mn SEA bars had a lower
maximum lateral force capacity of approximately 17% and 9%, respectively, compared to
the control RC column.
• The stiffness of all the columns using ECC, Cu-Al-Mn SEA bars, or both, was lower
compared to the conventional RC column due to the lower elastic modulus of both ECC
and Cu-Al-Mn SEA bars, compared to conventional concrete and steel rebar, respectively.
• The RECC hollow sections filled with conventional concrete disclosed a similar
performance in all metrics to the columns made entirely of RECC. This result suggests
that, with the proposed approach, the relatively expensive material ECC could be used in
an efficient way for the flexural behavior of concrete columns.
• The unbonding of the plastic hinge reinforcement was effective in terms of limiting the
cracking and deformations in the plastic hinge region of the columns.
203
Chapter 4-Numerical Modeling of the Tested Column
• The behavior of ECC was simulated by introducing smeared reinforcement to concrete
constitutive model in six directions. The stress-strain behavior of the smeared
reinforcement was adjusted to result in strain-hardening and strain-softening behavior of
ECC.
• Crack initialization, crack opening and crack localization stages were observed in the
modeled ECC dog-bone specimen subjected to uniaxial tension. Additionally, the ECC
model returned strain-hardening behavior under cyclic loading.
• One-dimensional SEA constitutive model was developed, based on the theory provided in
the literature, and implemented in the finite element package used in this research, to model
the tested bridge columns. The developed model accurately captured the superelastic
behavior of Cu-Al-Mn SEA bars, as observed in the experiments.
• The numerical models could accurately predict ultimate strength, post-peak behavior,
absorbed energy, permanent deformations, and hysteric cyclic behavior of the bridge
columns up to 7% lateral drift.
• The developed finite element modeling approach could capture force related measures
better than displacement related measures, especially in high lateral drift ratios.
Chapter 5-Numerical Parametric Study of Bridge Columns
• A numerical parametric study was conducted to obtain the optimal design strategy for the
proposed bridge column design considering ECC quality, number of Cu-Al-Mn SEA bars
in the plastic hinge region, and different cross-sections including hollow sections, as the
204
variables. In total, 47 different cases were developed and subjected to one cycle of 7%
lateral drift.
• In general, the maximum strength of the bridge columns increased by incorporating ECC
and the area under load-deformation diagrams were shrunk as the number of Cu-Al-Mn
SEA bars increased in the plastic hinge region.
• Changing the mechanical properties of ECC in tension had an insignificant effect on the
overall response of the bridge columns. Solid bridge columns with 0.5%, 1%, and 2% strain
hardening capacity of ECC returned similar values for the initial stiffness, the permeant
deformation, and the yield drift. However, the maximum strength, the yield strength, and
energy absorption capacity increased slightly (5.5% maximum).
• A considerable reduction (more than 87%) was observed in the permanent deformation by
replacing the 65%, and higher, of the longitudinal reinforcement with Cu-Al-Mn SEA bars.
The reduction of the permanent deformation was limited to 14.2% in cases with Cu-Al-Mn
SEA bars less than 65% of the longitudinal reinforcement.
• Using ECC instead of conventional concrete increased the lateral strength almost 30%
while replacing 65% of the longitudinal reinforcement with Cu-Al-Mn SEA bars reduced
the energy absorption capacity by around 14%.
• Hollow RECC sections with and without concrete core showed lower maximum strength
and energy absorption capacity, compared to the solid RECC sections. The cases with 15%
hollow ratio showed similar behavior as the solid RECC sections with and without Cu-Al-
Mn SEA bars as longitudinal reinforcement.
• Filling the hollow RECC sections with concrete had limited effect on the response of the
bridge columns with a maximum hollow ratio of 35%.
205
• Considering maximum strength and energy absorption capacity, RECC/2/NA/12, RECC-
H/2/381/12, and RECC-H-F/2/381/12 cases had comparable performance as the control
RC case while reducing the permanent deformations significantly.
Chapter 6-Experimental Program-Beam-Column Joints
• The damage to the corner RC beam-column subassembly (C/RC/ABB/RS) was severe
than the exterior one (E/RC/ABB/RS) due to biaxial loading. Considerably large cracks
were observed in this specimen along with extensive spalling of concrete at the panel zone
area.
• Replacing conventional concrete with ECC in the panel zone area of scaled beam-column
subassemblies and extending it to the plastic hinge region of the adjoining beams and
columns (RECC beam-column subassemblies) significantly reduced the apparent damage
compared to the control RC specimen.
• Fine diagonal distributed cracks were observed in the panel zone area of the RECC beam-
column subassemblies. Flexural cracks propagated in the in-plane and out-of-plane beams
through the axis of strong and weak bending. However, flexural cracks were limited to
less than the depth of the columns in the columns. By applying torsion to RECC beam-
column subassemblies, diagonal cracks appeared beyond RECC section of the columns,
indicating that the plastic hinge region was extended.
• Load-deformation diagram of the corner and exterior RECC beam-column subassemblies
were affected insignificantly by applying torsion and removal of the lateral reinforcement
in the pane zone area.
206
• Removal of the lateral reinforcement from the panel zone area of RECC specimens
increased the shear deformations but the load-deformation diagram revealed that the ECC
could resist large shear forces through its superior shear characteristics, compared to
conventional concrete.
• The construction cold joints between the RC and RECC sections had limited effect on the
consistency of the beams and columns.
• Most of the damage to the beams and columns were concentrated to a length equal to the
depth of the element.
207
7.2. Recommendations for Future Research
Considering the findings of this study, the future research needs are identified as follows:
• The superelastic behavior of SEA bars is strongly dependent on the grain size to bar
diameter ratio. Therefore, the larger grain size is required to develop a suitable SEA bar
for structural applications. Even though researchers could develop large diameter SEA
bars, their mechanical properties have not been completely understood, which creates a
great opportunity for future research.
• Most of the experimental work on bridge columns incorporating SEA bars have been
conducted on scaled specimens. More full-scale experimental studies are needed to confirm
the efficiency of SEA bars, specially Cu-Al-Mn SEA bars, in reducing permanent
deformation of bridge columns that occur due to strong seismic loads.
• Conducting life cycle assessment on bridges, incorporating the proposed bridge column
design concept, to evaluate the long-term benefits of the proposed design concept.
• Supplementary experimental work on seismic performance of bridge columns constructed
based on the proposed design concept and aged in a corrosive environment is needed. The
proposed design concept tackles the corrosion related issues of bridge columns in addition
to improving their seismic performance. However, the efficiency of the design concept in
corrosive environments is unknown.
• Conducting experimental and numerical studies on the performance of bridge columns with
other high-performance materials to improve their seismic and permeability characteristics
would be helpful.
208
• Development of independent multi-dimensional constitutive modes for ECC by defining
specific tensile, compressive, and shear functions for cyclic loads considering crack
opening and closure is needed.
• Investigation of the performance of hollow bridge columns incorporating ECC and SEA
bars through experimental work is needed.
• Investigation of the effect of temperature on the performance of the bridge columns
incorporating ECC and SEA bars since the superelastic behavior of SEAs is a function of
temperature is needed. It should be mentioned that the developed constitutive model for
SEA bars can consider the effect of temperature on the cyclic behavior of SEA bars.
• Conducting numerical studies on the exterior and corner beam-column joints considering
different loading scenarios and boundary conditions would be helpful.
• Development of simplified joint elements to simulate the behavior of 3D RECC joints for
structural level simulations would be needed.
• Investigation of the progressive collapse of the designed multistory building with RECC
beam-column joints would be helpful.
• Experimental investigation of RECC beam-column joints affected by high temperatures to
realize their performance in extreme situations such as fire would be helpful since fires
following earthquakes are not uncommon.
• Developing an experimental program to investigate effect of slab on seismic performance
of RECC beam-column subassemblies is be needed.
• Developing a precast beam by incorporating ECC in the panel zone and removing the
corresponding transverse reinforcement in order to facilitate the construction process
would be an interesting research subject.
209
REFERENCES
[1]
V. C. Li, "Performance driven design of fiber reinforced cementitious composites," in Fiber
reinforced Cement and Concrete, Proceeding of the Fourth International Symposium Held by
Rilem, University of Sheffield, 1992.
[2]
V. C. Li and C. K. Leung, "Steady-state and multiple cracking of short random fiber composites,"
Journal of Engineering Mechanics, vol. 118, no. 11, pp. 2246-2264, 1992.
[3]
V. C. Li and H.-C. Wu, "Conditions for pseudo strain-hardening in fiber reinforced brittle matrix
composites," Applied Mechanics Reviews, vol. 45, no. 8, pp. 390-398, 1992.
[4]
V. C. Li, S. Wang and C. Wu, "Tensile strain-hardening behavior of polyvinyl alcohol engineered
cementitious composite (PVA-ECC)," Materials Journal, vol. 98, no. 6, pp. 483-492, 2001.
[5]
M. Li and V. C. Li, "Rheology, fiber dispersion, and robust properties of engineered cementitious
composites," Materials and Structures, vol. 46, pp. 405-420, 2013.
[6]
M. Maalej, S. T. Quek and J. Zhang, "Behavior of hybrid-fiber engineered cementitious composites
subjected to dynamic tensile loading and projectile impact," Journal of Materials in Civil
Engineering, vol. 17, no. 2, pp. 143-152, 2005.
[7]
K. E. Kesner, S. L. Billington and K. S. Douglas, "Cyclic response of highly ductile fiber-reinforced
cement-based composites," ACI Materials Journal, vol. 100, no. 5, pp. 381-390, 2003.
[8]
S.-H. Chao, W.-C. Liao, T. Wongtanakitcharoen and A. E. Naaman, "Large scale tensile tests of
high performance fiber reinforced cement composites," in 5th Int. RILEM Workshop on High
Performance Fiber Reinforced Cement Composites (HPFRCC5), Mainz, Germany, 2007.
[9]
S. Boughanem, M. J. Jesson, P. A. Smith, C. Eddie, S. Psomas and M. Rimes, "Tensile
characterisation of thick sections of engineered cement composite (ECC) materials," Journal of
materials science, vol. 50, no. 2, pp. 882-897, 2015.
[10]
P. S. Bhat, V. Chang and M. Li, "Effect of elevated temperature on strain-hardening engineered
cementitious composites," Construction and Building Materials, vol. 69, pp. 370-380, 2014.
[11]
ASTM, ASTM C666/C666M-15 Standard test method for resistance of concrete to rapid freezing
and thawing, West Conshohocken, PA: ASTM International, 2003.
210
[12]
V. Li, G. Fischer, Y. Kim, M. Lepech, S. Qian, M. Weimann and S. Wang, "Durable link slabs for
jointless bridge decks based on strain-hardening cementitious composits, Research Report RC-
1438," Michigan Department of Transportation, Ann Arbor, 2003.
[13]
V. Li, T. Horikoshi, A. Ogawa, S. Torigoe and T. Saito, "Micromechanics based durability study of
polyvinyl alcohol-engineered cementitious composite (PVA-ECC)," ACI Materials Journal, vol.
101, no. 3, pp. 242-248, 2004.
[14]
P. Kabele, L. Novak, J. Nemecek and J. Pekar, "Multiscale experimental investigation of
deterioration of fiber-cementitious composites in aggressive environment," in MHM 2007:
Modelling of Heterogeneous Materials, with Applications in Construction and Biomedical
Engineering, Prague, Czech Republic, 2007.
[15]
M. Li, M. Şahmaran and V. Li, "Effect of cracking and healing on durability of engineered
cementitious composites under marine environment," in High Performance Fiber Reinforced
Cement Composites (HPFRCC5), Mainz, Germany, 2007.
[16]
V. C. Li, "Engineered cementitious composites (ECC) – material, structural, and durability
performance," in Concrete Construction Engineering Handbook, CRC Press, 2008, p. Chapter 24.
[17]
M. Şahmaran, V. Li and C. Andrade, "Corrosion resistance performance of steel-reinforced
engineered cementitious composites beams," ACI Materials Journal, vol. 105, no. 3, pp. 243-250,
2008.
[18]
W. P. Boshoff, "Cracking behavior of strain-hardening cement-based composites subjected to
sustained tensile loading," ACI Materials Journal, vol. 111, no. 5, 2014.
[19]
W. P. Boshoff and C. J. Adendorff, "Effect of sustained tensile loading on SHCC crack widths,"
Cement & Concrete Composites, vol. 37, pp. 119-125, 2013.
[20]
J. M. Rouse and S. L. Billington, "Creep and shrinkage of high-performance fiber-reinforced
cementitious composites," ACI Materials Journal, vol. 104, no. 2, pp. 129-136, 2007.
[21]
V. C. Li, D. K. Mishra, A. E. Naaman, J. K. Wight, J. M. LaFave, H.-C. Wu and Y. Inada, "On the
shear behavior of engineered cementitious composites," Advanced Cement Based Materials, vol. 1,
no. 3, pp. 142-149, 1994.
[22]
P. Gideon and v. Zijl, "Improved mechanical performance: shear behaviour of strain-hardening
cement-based composites (SHCC)," Cement and Concrete Research, vol. 37, pp. 1241-1247, 2007.
211
[23]
R. DesRoches, J. McCormick and M. Delemont, "Cyclic properties of superelastic shape memory
alloy wires and bars," Journal of Structural Engineering, vol. 130, no. 1, pp. 38-46, 2004.
[24]
R. Desroches and B. Smith, "Shape memory alloy in seismic resistant design and retrofit: A critical
review of their," Journal of Earthquake Engineering, vol. 8, no. 3, pp. 415-429, 2004.
[25]
O. Ozbulut, S. Hurlembaus and R. Desroches, "Seismic response control using shape memory
alloys: A review," Intelligent Material Systems and Structures, vol. 22, no. 4, pp. 1531-1549, 2011.
[26]
S. G. L. Hurlebaus, "Smart structure dynamics," Mechanical Systems and Signal Processing, vol.
20, no. 2, pp. 225-281, 2006.
[27]
K. N. Melton and J. Harrison, "Corrosion of NiTi based shape memory alloys," in 1st International
Conference on Shape Memory and Superelastic Technologies, Pacific Grove, CA, 1994.
[28]
"ASTM 2006 Standard specification for structural steel for bridges, A 709/A 709M-05 Annual Book
of ASTM Standard," West Conshohocken, PA: ASTM, pp. pp 372-9.
[29]
Y. Zhang, J. A. Camilleri and S. Zhu, "Mechanical properties of superelastic Cu–Al–Be wires at
cold temperatures for the seismic protection of bridges," Smart Materials and Structures, vol. 17,
no. 2, 2008.
[30]
Y. Araki, T. Endo, T. Omori, Y. Sutou, Y. Koetaka, R. Kainuma and K. Ishida, "Potential of
superelastic Cu–Al–Mn alloy bars for seismic applications," Earthquake Engineering and
Structural Dynamics, vol. 40, no. 1, pp. 107-115, 2011.
[31]
IARC, IARC monographs on the evaluation of carcinogenic risks to humans. Beryllium, Cadmium,
Mercury and exposures in glass manufacturing industry, 1994: Lyon, IARC.
[32]
M. Ishibashi, T. Nobuko, S. Takaki, O. Toshihiro, S. Yuji, K. Ryosuke, Y. Kiyoshi and I. Kiyohito,
"A simple method to treat an ingrowing toenail with a shape‐memory alloy device," Journal of
Dermatological Treatment, vol. 19, no. 5, pp. 291-292, 2008.
[33]
B. Gencturk, Y. Araki, T. Kusama, T. Omori, R. Kainuma and M. F, "Loading rate and temperature
dependency of superelastic Cu-Al-Mn alloys," Construction and Building Materials, vol. 53, pp.
555-560, 2014.
[34]
Y. Sutou, T. Omori, K. Yamauchi, N. Ono, R. Kainuma and K. Ishida, "Effect of grain size and
texture on pseudoelasticity in Cu–Al–Mn-based shape memory wire," Acta Materialia, vol. 53, pp.
4121-4133, 2005.
212
[35]
T. Omori, T. Kusama, S. Kawata, I. Ohnuma, Y. Sutou, Y. Araki, K. Ishida and R. Kainuma,
"Abnormal grain growth induced by cyclic heat treatment," Science, vol. 341, no. 6153, pp. 1500-
1502, 2013.
[36]
Y. Araki, N. Maekawa, T. Omori, Y. Sutou, R. Kainuma and K. Ishida, "Rate-dependent response
of superelastic Cu–Al–Mn alloy rods to tensile cyclic loads," Smart Materials and Structures, vol.
21, no. 3, p. 032002, 2012.
[37]
K. C. Shrestha, Y. Araki, T. Kusama, T. Omori and R. Kainuma, "Functional fatigue of
polycrystalline Cu-Al-Mn superelastic alloy bars under cyclic tension," Journal of Materials in Civil
Engineering, vol. 28, no. 5, p. 04015194, 2015.
[38]
C. Noguez, C. A. and M. S. Saiidi, "Shake-table studies of a four-span bridge model with advanced
materials," Journal of Structural Engineering, vol. 138, no. 2, pp. 183-192, 2011.
[39]
M. S. Saiidi, M. O'Brien and M. Sadrossadat-Zadeh, "Cyclic response of concrete bridge columns
using superelastic Nitinol and bendable concrete," ACI Structural Journal, vol. 106, no. 1, p. 69,
2009.
[40]
C. A. Cruz Noguez and M. S. Saiidi, "Performance of advanced materials during earthquake loading
tests of a bridge system," Journal of Structural Engineering, vol. 139, no. 1, pp. 144-154, 2012.
[41]
M. S. Saiidi and H. Wang, "Exploratory study of seismic response of concrete columns with shape
memory alloys reinforcement," ACI structural journal, vol. 103, no. 3, p. 436, 2006.
[42]
B. A. Nakashoji and S. Saiidi, "Seismic performance of square Nickel-Titanium reinforced ECC
columns with headed couplers," Center for Civil Engineering Earthquake Research Report CCEER-
14-05, Univ. of Nevada, Reno, NV, 2014.
[43]
S. Varela and S. Saiidi, "Dynamic performance of novel bridge columns with superelastic CuAlMn
shape memory alloy and ECC," International Journal of Bridge Engineering (IJBE), vol. 2, no. 3,
pp. 29-58, 2014.
[44]
F. Hosseini, B. Gencturk, S. Lahpour and D. Ibague Gil, "An experimental investigation of
innovative bridge columns with engineered cementitious composites and Cu–Al–Mn super-elastic
alloys," Smart Materials and Structures, vol. 24, no. 8, p. 085029, 2016.
[45]
B. Gencturk and S. F. Hosseini, "Use of Cu-Based superelastic alloys for innovative design of
reinforced concrete columns," in Tenth U.S. National Conference on Earthquake Engineering
(NCEE), Anchorage, Alaska, 2014.
213
[46]
B. Shrestha and H. Hao, "Parametric study of seismic performance of super-elastic shape memory
alloy-reinforced bridge piers," Structure and Infrastructure Engineering, vol. 12, no. 9, pp. 1076-
1089, 2016.
[47]
A. M. Billah and M. S. Alam, "Plastic hinge length of shape memory alloy (SMA) reinforced
concrete bridge pier," Engineering Structures, vol. 117, pp. 321-331, 2016.
[48]
E. Nikbakht, K. Rashid, F. Hejazi and S. A. Osman, "Application of shape memory alloy bars in
self-centring precast segmental columns as seismic resistance," Structure and Infrastructure
Engineering, vol. 11, no. 3, pp. 297-309, 2015.
[49]
G. J. Parra-Montesinos, S. W. Peterfreund and S.-H. Chao, "Highly damage-tolerant beam-column
joints through use of high-performance fiber-reinforced cement composites," ACI Structural
Journal, vol. 102, no. 3, pp. 487-495, 2005.
[50]
A. C. 318, Building Code Requirements for Structural Concrete (ACI 318-02) and Commentary
(318R-02), Farmington Hills, Mich: American Concrete Institute, 2002.
[51]
K. Shakya, K. Watanabe, K. Matsumoto and J. Niwa, "Application of steel fibers in beam–column
joints of rigid-framed railway bridges," Construction and Building Materials, vol. 27, pp. 482-489,
2012.
[52]
S. Qudah and M. Maalej, "Application of engineered cementitious composites (ECC) in interior
beam–column connections for enhanced seismic resistance," Engineering Structures, vol. 69, pp.
235-245, 2014.
[53]
R. Zhang, K. Matsumoto, T. Hirata, Y. Ishizeki and J. Niwa, "Application of PP-ECC in beam–
column joint connections of rigid-framed railway bridges to reduce transverse reinforcements,"
Engineering Structures, vol. 86, pp. 146-156, 2015.
[54]
U. Akguze and S. Pampanin, "Effects of variation of axial load and bidirectional loading on seismic
performance of GFRP retrofitted reinforced concrete exterior beam-column joints," Journal of
Composites for Construction, vol. 14, no. 1, pp. 94-104, 2010.
[55]
U. Akguzel, Seismic performance of FRP retrofitted exterior RC beam-column joints under varying
axial and bidirectional loading, University of Canterbury: Christchurch, New Zealand, 2011.
[56]
B. B. Canbolat and J. K. Wight, "Experimental investigation on seismic behavior of eccentric
reinforced concrete beam-column-slab connections," ACI Structural Journal, vol. 105, no. 2, p. 154,
2008.
214
[57]
G. S. Raffaelle and J. K. and Wight, "Reinforced concrete eccentric beam-column connections
subjected to earthquake-type loading," ACI Structural Journal, vol. 92, no. 1, pp. 45-55, 1995.
[58]
G. M. Lawrence, J. H. Beattie and D. H. and Jacks, "Cyclic load performance of an eccentric beam-
column joint,Central Laboratories Report 91-25126," Central Laboratories, Lower Hutt, New
Zealand, 1991.
[59]
A. 2013, "Report Card for America's Infrastructure," American Society of Civil Engineers (ASCE),
2013. [Online]. [Accessed 3 July 213].
[60]
I. Friedland and R. Mayes, Recommended LRFD guidelines for the seismic design of highway
bridges, MECCR highway project 094, task F3-1, State University of New York at Buffalo, 2001.
[61]
G. Fischer and V. C. Li, "Effect of matrix ductility on deformation behavior of steel-reinforced ECC
flexural members under reversed cyclic loading conditions," Structural Journal, vol. 99, no. 6, pp.
781-790, 2002.
[62]
C.-G. Cho, Y. Yong Kim, L. Feo and D. Hui, "Cyclic responses of reinforced concrete composite
columns strengthened in the plastic hinge region by HPFRC mortar," Composite Structures, vol. 94,
no. 7, pp. 2246-2253, 2012.
[63]
A. Aviram, B. Stojadinovic and G. J. Parra-Montesinos, "High-performance fiber-reinforced
concrete bridge columns under bidirectional cyclic loading," ACI Structural Journal, vol. 111, no.
2, p. 303, 2014.
[64]
ATC/MCEER, "NCHRP Report 472 : Comprehensive specification for the seismic design of
bridges," National Cooperative Highway Research Program (NCHRP), Washington, DC, 2002.
[65]
R. T. Leon and G. G. Deierlein, "Considerations for the use of quasi-static testing," Earthquake
spectra, vol. 12, no. 1, pp. 87-109, 1996.
[66]
j. K. Dong, S. El-Tawil and A. E. Naaman, "Rate-dependent tensile behavior of high performance
fiber reinforced cementitious composites," Materials and Structures, vol. 42, no. 3, pp. 399-414,
2009.
[67]
ASTM, Standard test method for compressive strength of cylindrical concrete specimens, ASTM
C39/C39M, West Conshohocken, PA: American Society of Testing Materials (ASTM), 2014.
[68]
"NYcon Corporation," [Online]. Available: http://nycon.com/nycon-pva-recs15/. [Accessed 18
March 2015].
215
[69]
"Boral Material Technologies LLC.," [Online]. Available: http://boralamerica.com/fly-ash.
[Accessed 30 August 2017].
[70]
"US Silica," [Online]. Available: http://www.ussilica.com/. [Accessed 30 August 2017].
[71]
ASTM, Standard test methods for testing mechanical splices for steel reinforcing bars, ASTM
A1034/A1034M, West Conshohocken, PA: American Society of Testing Materials (ASTM), 2010.
[72]
R. Park, "Ductility evaluation from laboratory and analytical testing," in the 9th world conference
on earthquake engineering, Tokyo-Kyoto, Japan, 1988.
[73]
Č. C. s.r.o., ATENA vesrion 5.3.4o.14242, Na Hrebenkach: Czech Republic.
[74]
V. Červenka, L. Jendele and J. Červenka, ATENA program documentation, part 1: Theory, Prague:
Červenka Consulting s.r.o., 2016.
[75]
S. A. Motahari and M. Ghassemieh, "Multilinear one-dimensional shape memory material model
for use in structural engineering applications," Engineering Structures, p. 904–913, 2007.
[76]
A. J. Bigaj, "Structural dependence of rotation capacity of plastic hinges in RC beams and slabs,"
Delft University of Technology, 1999.
[77]
S. W. Sloan and M. F. Randolf, "Automatic element reordening for finite element analysis with
frontal solution schemes," International Journal for Numerical Methods in engineering, vol. 19, pp.
1153-1181, 1983.
[78]
M. Schoettler, J. Restrepo, G. Guerrini, D. Duck and F. Carrea, "A full-scale, single-column bridge
bent tested by shake-table excitation," Center for Civil Engineering Earthquake Research,
Department of Civil Engineering, University of Nevada, Las Vegas, NV, 2012.
[79]
Caltrans, Bridge design specificatio, Sacramento, CA: California Department of Transportation,
2004.
[80]
C. (2006a), Seismic design criteria, Sacramento, CA: California Department of Transportation,
2006.
[81]
G. Fischer, S. Wang and V. C. Li, in Brittle matrix composites 7, Warsaw, 2003.
216
[82]
"ASTM A 706/ A 706M Standard specification for low-alloy steel deformed and plain bars for
concrete reinforcement," West Conshohocken, PA, American Society for Testing and Materials.
[83]
C. B. Haselton and G. G. Deierlein, "Assesment seismic collapse safety of modern reinforced
concrete moment frame buildings, Report No. 156".
[84]
A. C. 318, Building Code Requirements for Structural Concrete (ACI 318-02) and Commentary
(318R-02), Farmington Hills, Mich: American Concrete Institute, 2002.
[85]
"KURARAY Co., LTD," [Online]. Available: http://kuralon-frc.kuraray.com/product-
application/for-mortar/recs. [Accessed 15 09 2017].
Abstract (if available)
Abstract
In the recent past, earthquakes have proven that reinforced concrete (RC) structures remain highly vulnerable to lateral loads. The yielding of longitudinal reinforcement as the main source of energy absorption, in conjunction with the cracking and spalling of the concrete, leads to severe damage and permanent deformations, which jeopardizes the post-earthquake functionality of these structures. A swift response and recovery in the aftermath of a major disaster strongly relies on the serviceability of key infrastructure, including bridges and other vital service structures such as hospitals and government buildings. Recently, the use of high-performance materials such as Engineered Cementitious Composites (ECC) and Superelastic Alloys (SEA) have been considered as alternative materials for conventional concrete and steel reinforcing bars (rebar), attracting heightened attention to improve the seismic performance of RC structures. ECC refers to a special class of high-performance fiber-reinforced cementitious composites (HPFRCC) that exhibit superior tensile ductility, energy absorption, bond characteristics and shear resistance. SEAs are innately capable of recovering large inelastic deformations upon stress removal. ❧ The primary endeavor of this study is to improve the performance of bridge columns and beam-column joints of special moment frames (SMF) using high-performance materials, particularly, ECC and Cu-Al-Mn SEA bars. To this end, an innovative bridge column design is first introduced and examined through an experimental framework. The column design comprises of a prefabricated reinforced ECC (RECC) hollow section that is embedded in a RC foundation and filled with conventional concrete. Additionally, the longitudinal reinforcement at the potential plastic hinge region is totally or partially replaced using the recently developed Cu-Al-Mn SEA bars. The proposed approach utilizes the deformability of the ECC in order to enhance damage tolerance of the bridge columns and large strains recovery capability of Cu-Al-Mn SEA to reduce permanent deformations. Following the completion of experimental work, a finite element approach is developed to numerically investigate the performance of the bridge columns designed on the basis of this proposed approach. The developed finite element approach is verified using the results ascertained from the tested columns and used to carry out a parametric study to obtain an optimal design strategy for the proposed design concept. Furthermore, in an attempt to improve the performance of corner and exterior beam-column joints in SMFs, conventional RC in 3D beam-column subassemblies is substituted with reinforced ECC (RECC) that extends from the panel zone area into the adjacent beams and columns in order to cover the potential plastic hinge regions. The 3D RECC beam-column subassemblies are then subjected to complex loading scenarios, including torsion, to investigate their performance due to more realistic seismic loads in 3D structures. ❧ The results from the experimental works indicated that ECC can significantly improve damage tolerance of bridge columns and beam-column joints subjected to extensive seismic loads and shift the damage mode from concrete spalling to fine distributed cracks. Incorporating ECC in panel zone of corner and exterior beam-column subassemblies sufficiently tolerated complex loading combinations even in absence of panel zone transverse reinforcement, due to superior shear strength of ECC compared to conventional concrete. Additionally, replacing longitudinal reinforcement with Cu-Al-Mn SEA bars in plastic hinge region of bridge columns, considerably decreases the permanent deformations of RECC bridge columns compared to the conventional RC ones. Furthermore, conducting the numerical and parametric studies revealed that: (i) mechanical properties of ECC can be successfully simulated by introducing fibers as smeared reinforcement in concrete
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Material characterization of next generation shape memory alloys (Cu-Al-Mn, Ni-Ti-Co and Fe-Mn-Si) for use in bridges in seismic regions
Asset Metadata
Creator
Hosseini Khasheh Heiran, Seyyed Farshid
(author)
Core Title
Next generation seismic resistant structural elements using high-performance materials
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Civil Engineering
Publication Date
12/11/2018
Defense Date
10/24/2017
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
3D beam-column joints,bridge columns,Cu-Al-Mn superelastic alloy,engineered cementitious composites,high-performance materials,numerical modeling,OAI-PMH Harvest,parametric study,seismic performance
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Gencturk, Bora (
committee chair
), Johnson, Erik A. (
committee member
), Nutt, Steven (
committee member
)
Creator Email
hosseini.seyyedfarshid@gmail.com,Hosseini.Seyyedfarshid@yahoo.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-461651
Unique identifier
UC11266935
Identifier
etd-HosseiniKh-5950.pdf (filename),usctheses-c40-461651 (legacy record id)
Legacy Identifier
etd-HosseiniKh-5950.pdf
Dmrecord
461651
Document Type
Dissertation
Rights
Hosseini Khasheh Heiran, Seyyed Farshid
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
3D beam-column joints
bridge columns
Cu-Al-Mn superelastic alloy
engineered cementitious composites
high-performance materials
numerical modeling
parametric study
seismic performance