Bond of Reinforcement in High-Performance Fiber- Reinforced Concrete

peletonwhoopUrban and Civil

Nov 26, 2013 (3 years and 8 months ago)

77 views


1

Bond of Reinforcement in High
-
Performance Fiber
-
Reinforced Concrete

Matthew J. Bandelt and Sarah L. Billington



ABSTRACT

High performance fiber
-
reinforced concrete (HPFRC) materials are being investigated for use in
bridge piers and other structural
systems in seismic regions because of their unique structural
properties. These materials, which are damage tolerant can be used with lower amounts of
transverse steel while still exhibiting superior behavior to ordinary cement
-
based materials.
Understand
ing the interaction between steel reinforcement and these
composites

is essential to
develop accurate modeling tools and design guidelines for their implementation into practice.
On
-
going research on characterizing the bond strength and bond
-
slip behavior

of steel
reinforcement embedded in HPFRC materials will be presented. Beam splice specimens have
been tested with varying amounts of cover and steel confinement to study the bond performance
in these materials relative to traditional concrete. Results t
o date show that the confinement
provided by HPFRC alone can significantly increase the bond strength and
bond
ductility of
reinforcement in HPFRC materials as compared to ordinary concrete. Future research on
modeling approaches, proposed design guideline
s, and
the cyclic bond performance of reinforced
HPFRC components will be discussed.































_____________________________________

Matthew J. Bandelt, Ph.D. Student, NSF Research Fellow
,
Dept. of Civil and Environmental Engineering,
Stanford University
,
439 Panama Mall, Rm 213, Stanford, CA 94305
,
Phone: (650)
-
723
-
4150, Fax: (650)
-
725
-
9755, E
-
Mail:
mbandelt@stanford.edu


Sarah L. Billington, Associate Professor, Associate Chair
,
Dept. of Civil and Environmental Engineering, Stanford U
niversity
,
473 Via Ortega, Rm 285A, Stanford, CA 94305
,
Phone: (650)
-
723
-
4125, Fax: (650)
-
723
-
7514, E
-
Mail: billington@stanford.edu


2


INTRODUCTION


Recent
development
s

of

innovative h
igh
-
performance fiber
-
reinforced
concrete (HPFRC)
materials

have

led to
a
strong interest in using
HPFRCs

for a number of seismic applications
.
High performance fiber
-
reinforced concrete materials exhibit
a pseudo strai
n hardening behavior
in tension, show
limited

spalling in compression, and can often exhibit a hardening behav
ior
when loaded in flexure as well

(Naaman and Reinhardt 2006)
.

Potential applications for HPFRC
materials
have bee
n widely studied and
include hinge regions for self
-
centering bridge piers

(Lee
and Billington 2008)
.

P
roof of concept applications
have
show
n

enhanced performance when these materials
are used instead of ordinary cement based materials
. Guidelines f
or modeling and designing
with
HPFRCs
are now needed
.
In order to develop adequate modeling tools and appropriate
design guidelines for HPFRC materials,
how they
interact with steel reinforcement

and in
particular, how bond
allows transfer of stresses betw
een reinforcement and the surrounding
composite
material
s
,
is being evaluated
.

Previous experimental studies on bond in HFPRC materials have used the direct pull
-
out
test and test results show a substantial increase in bond strength due to the mechanical
properties
of these materials

(Chao et al. 2009)
. However, the pull
-
out test method is recognized as an
inadequate method to fully characterize bond strength due to the simplified stress state
(ACI
Comittee 408 2003)
.
Instead, a beam specimen with lap splices is the preferr
ed experiment since
a more realistic stress state can be obtained. Traditional strain softening fiber reinforced concrete
(FRC) materials have been test
ed using beam splice specimens
(Harajli et al. 2002)
, but there are
no experimental results availa
ble, to date, on
bond of reinforcement in
tensile strain hardening
HPFRC materials.

On
-
going research is being conducted by the authors to characterize bond between a
variety of HPFRCC materials and steel reinforcement for future model development. In this

paper, the results of experiments on beam specimens containing lap splices are presented for two
different mixture types.




EXPERIMENTAL PROGRAM


Materials


Two cement based mixes were used to fabricate the specimens in this study; a normal
weight concre
te and an Engineered Cementitious Composite (ECC).

The normal weight
concrete consisted of Type II/V Portland cement, water, fine aggregate and coarse aggregate.
The fine aggregate had a fineness modulus of 3.1 and the coarse aggregate had a maximum size

of
0.375 inches
.

Engineered Cementitious Composites are one type of HPFRC material that exhibits
tensile strain hardening and deflection hardening. The ECC mix consisted of Type II/V Portland
cement, Class F fly ash, silica sand, a viscosity modifying adm
ixture, super
-
plasticizer and 2.0%
by volume polyvinyl alcohol (PVA) fibers. The silica sand had a 0.
005

inch

particle size and the
PVA fibers were
0.472

inches

long. The proportions for 1 cubic yard of the concrete and ECC
mixes are shown in Table
I
.



3

Table
I.

Proportions for One Cubic Yard of Concrete and ECC Mix
.




Specimen Design


As previously noted, beam specimens with lap splices were fabricated and tested in this
study since they are considered the preferred specimen type for bond experiments.

All
specimens were designed to ensure that a bond failure would occur prior to
yielding of the
reinforcement.
The specimens were designed in this manner so that the maximum
elastic
bond
stress could be measured.

Future work on the inelastic bond behavior
is planned.

The beams were 54 inches in length and had a heigh
t of 9 inches.
Shear stirrups, with a
diameter of 0.375 inches, were placed in the maximum shear region and spaced at 3.5 inches to
guarantee that the beams failed due to insufficient bond stren
gth rather than shear.

Two steel longitudinal reinforcing bars with a diameter (
d
b
) of 0.625 inches were used

throughout the cross section.
At mid
-
span two bars were lap spliced to another another
two bars
as shown in Figure 1.
Two of the bars were
debonded for the remaining length of the beam using
a thin wall PVC pipe so that reinforcement slip could be measured at the end of the beam. The
reinforcement that was debonded was extended outside of the beam so that instrumentation to
measure reinforc
e
ment slip could be installed.
A splice length (
L
s
) of 10
d
b

(6.25 inches) was
chosen to ensure that the specimens would fail due to bond rather than flexure.




Figure 1
.

a) Specimen Cross Section within Splice Region b) Specimen Plan View
.

C
e
me
n
t
FA
Fine
C
oar
s
e
SP
VMA
(l
b
)
(l
b
)
(l
b
)
(l
b
)
(l
b
)
(l
b
)
(l
b
)
(l
b
)
Conc
re
t
e
1
100
-
420
1015
1351
-
-
-
E
CC
922
1
106
526
738
-
4.6
1.0
44
PV
A

Fibers
F
A
= F
l
y
A
s
h;
S
P
= S
upe
rpl
a
s
t
i
c
i
z
e
r;

V
M
A
=
V
i
s
c
oc
i
t
y M
odi
fyi
ng
A
ge
nt
Mix
Bi
n
d
e
r
W
ate
r
A
ggr
e
gate
A
d
mi
xtu
r
e
s
a
)
b)

4

In addition to
the mechanical properties of a material, there are two other important
factors that affect bond behavior between steel reinforcement and a cementitious mix

(Orangun
et al. 1977)
. The first is the ratio of concrete cover to bar diameter (
c/d
b
), w
here the term “cover”
refers to bottom cover (
c
b
), side cover (
c
so
) and cover between bars (
c
si
) as shown in Figure 1a.
The second important factor is the presence of confinement (i.e., stirrups) in areas where bond is
critical.

Four specimen types were

designed with these factors in mind to examine how these two
parameters (
c/d
b

ratio and presence of confinement)
affected bond strength in HPFRC

materials.
Specimens with two different cover to bar diameter ratios were studied, 1.0 and 1.5, and
specimens

with and without confinement in the splice region were tested for each mix type. The
compressive strengths for ECC and concrete were 7,300 psi and 6,500 psi, respectively. The test
matrix and compressive strengths of each
material are outlined in Table
II
.


Table
II
: Test Matrix and Compressive Strengths
.




Test Setup and Procedure


The simply supported beams had a span length of
54 inches

and were loaded in four
-
point bending to create a constant moment region. A
55

kip

universal testing machine applied
force to a load distribution element, which then loaded the beam at the third points. The testing
machine was set to displ
acement control at a rate of 0.02

inches
/min. The test was paused at
certain points during testin
g (i.e., just after first cracking, just before the expected bond failure
load, after bond failure and at other intermittent points) to investigate and map crack patterns.
All specimens were coated in a lime wash so that crack propagation could be easily
identified.
An overview of the test setup can be seen in Figure 2.

Strain gages were installed to calculate bond s
tresses.
The gages were placed on all four
pieces of reinforcement 6 mm

outside of the splice region.
Linear variable differential
transducer
s (LVDTs) were positioned at mid
-
span to monitor vertical displacement.

LVDTs were also installed at the ends of the two reinforcing bars that were debonded.
These LVDTs, shown in Figure 3, measured displacement (slip) of the reinforcement as the
specim
ens were loaded.


b
c
f'
c
(i
n
.)
(i
n
.)
(p
s
i
)
CO
N
-1-U
C
Conc
re
t
e
5.00
0.625
1.0
-
6500
CO
N
-1-C
Conc
re
t
e
5.00
0.625
1.0
2 #3
6500
CO
N
-1.5-U
C
Conc
re
t
e
6.25
0.938
1.5
-
6500
CO
N
-1.5-C
Conc
re
t
e
6.25
0.938
1.5
1 #3
6500
E
CC-1-U
C
E
CC
5.00
0.625
1.0
-
7300
E
CC-1-C
E
CC
5.00
0.625
1.0
2 #3
7300
E
CC-1.5-U
C
E
CC
6.25
0.938
1.5
-
7300
E
CC-1.5-C
E
CC
6.25
0.938
1.5
1 #3
7300
S
p
e
c
i
me
n
N
ame
Mix
c
/
d
b
C
on
fi
n
e
me
n
t
c
=
c
b
=
c
si
=
c
so
Confi
ne
m
e
nt
= num
be
r of s
t
i
rrups
i
n s
pl
i
c
e
re
gi
on

5



Figure 2
.

Beam Specimen Test Setup
.




Figure 3
.

LVDTs Used for Measuring Reinforcement Slip
.



EXPERIMENTAL RESULTS


Failure Mechanisms and Crack Propagation


Initial cracks were mapped after a change in stiffness was detected during

testing. These
first flexural cracks, shown in Figure 4, were well distributed throughout the constant moment
region. ECC and concrete specimens both showed similar initial crack patterns within the
constant moment region.


50 k
i
p

T
e
s
ti
n
g M
ac
h
i
n
e

Load
D
i
s
tr
i
b
u
ti
on
El
e
me
n
t
S
h
e
ar
S
p
an

18 i
n
.(typ
.)
C
on
s
tan
t M
ome
n
t R
e
gi
on
18 i
n
.
P
i
n
S
u
p
p
or
t
R
ol
l
e
r

S
u
p
p
or
t
L
V
D
T
M
e
as
u
r
i
n
g
R
e
i
n
for
c
e
me
n
t S
l
i
p


6



Figure 4
.

Typical Flexural

Cracks Observed After First Cracking
.



A bond failure was observed when a drop in load carrying capacity occurred. The ECC
specimens exhibited more distributed cracking; however,
bond
cracks localized in the splice
region for both ECC and concrete speci
mens causing a bond failure. The bond cracks were
observed on the vertical and bottom faces of

the beams as shown in Figures 5 and 6
, respectively.

While testing ECC
-
1.5
-
UC, a bond crack on the underside of the beam developed in the
splice region. This c
rack then propagated towards a flexural crack outside of the splice region.
As the experiment continued, the flexural crack widened and a portion of the beam began to
separate o
ut of plane as shown in Figure 7
.
T
his
specimen
was the only
one
observed

to h
ave this
behavior
.




Figure 5
.

Typical Bond Cracks on Vertical Side of Beam
.



S
p
l
i
c
e
R
e
gi
on

C
on
s
tan
t M
ome
n
t R
e
gi
on

19 mm

F
l
e
xu
r
al
C
r
ac
k

M
i
d
-s
p
an
L
V
D
T

19 mm

S
p
l
i
c
e
R
e
gi
on

Bon
d
C
r
ac
k
s


7



Figure 6
.

Typical Bond Cracks on Underside of Beam
.





Figure 7
.

Separating Crack from ECC
-
UC
-
1.5
.

19 mm

Bon
d
C
r
ac
k
s

19 mm

M
i
d
-s
p
an

S
p
l
i
c
e
R
e
gi
on

C
r
ac
k
S
e
p
ar
ati
n
g
O
u
t of P
l
an
e


8

Bond Strength


Bond strength was calculated from the strain data collected during testing using Equation
1; where u is the bond stress,
T
s

is the tension force in the steel reinforcement,
d
b

is the diameter
of the steel reinforcement,
L
s

is the splice length,
A
b

is the a
rea of the reinforcement,
E
s

is the
modulus of elasticity of the reinforcing steel and

s

is the strain in the reinforcing steel. It should
be noted that the average strain from the four gages was used.


u

T
s

d
b
L
s

A
b
E
s

s

d
b
L
s






(1)



The peak
bond strengths from all of the experiments are summarized in Table III. The
ECC specimens had higher maximum bond strengths than their concrete specimen counterparts.
The increase in bond strength from ECC ranged from 22 to 64 percent.

There was an incre
ase in bond strength by changing the
c
/
d
b

value from 1.0 to 1.5 for the
concrete specimens, but little to no increase was observed for the ECC specimens. A substantial
increase in bond strength was observed for the
c
/
d
b

value of 1.5 when confinement was a
dded to
the concrete specimens; however, there was little to no increase in bond strength by confinement
otherwise.


Table III.

Maximum Bond Strengths in Concrete and ECC Specimens
.




Bond
-
Slip Behavior


The bond stress vs. slip responses of the unconfin
ed and confined specimens are shown in
Figures

8 and 9
. In general, the bond
-
slip curves have three distinct
segments
:


1.

An initial branch where bond strength and reinforcement slip increase until the maximum
bond stress is reached

2.

A descending branch,
after the maximum bond stress is reached where bond stress
decreases as reinforcement slip increases

3.

A final branch where bond stress remains relatively constant while reinforcement slip
increases


Concerning the initial branch of the bond
-
slip curve, the ascending stiffness was similar for
both materials. However, as noted in
the previous section,

the peak bond strengths for the ECC
specimens were much higher.

After the maximum bond stress was reac
hed, the bond strength of the ECC specimens
decreased at a lower slip rate than the concrete specimens. When confinement was added to the
concrete specimens, the descending portion of the bond
-
slip slip curve was less brittle than when
confinement was not

present. This was also true for the ECC specimens, but to a lesser degree.

The ECC residual bond strengths were substantially higher than the residual concrete bond
strengths. Additionally the ratio of residual bond strength to maximum bond strength was

higher
u
m
ax,Con
c
r
e
t
e
u
m
ax,ECC
(p
s
i
)
(p
s
i
)
U
nc
onfi
ne
d
1.0
776
1249
61%
U
nc
onfi
ne
d
1.5
848
1202
42%
Confi
ne
d
1.0
747
1222
64%
Confi
ne
d
1.5
1014
1237
22%
C
on
fi
n
e
me
n
t
c
/
d
b
P
e
r
c
e
n
t I
n
c
r
e
as
e

9

for ECC specimens.

One important observation is that the residual bond strength of ECC
-
UC
-
1.5 was lower
than that of ECC
-
UC
-
1.0. This was unexpected since bond strengths are expected to increase
with increasing
c
/
d
b

ratios. However, one explanati
on for the decrease in residual strength is that
the separating crack that developed in ECC
-
UC
-
1.5, as previously discussed and shown in Figure
7,

resulted in a lower residual bond capacity. Additional specimens are planned to further
investigate the dist
inct behavior of ECC
-
UC
-
1.5.




Figure 8.

Bond Stress vs. Reinforcement Slip for Unconfined Specimens
.




Figure 9.

Bond Stress vs. Reinforcement Slip for Confined Specimens
.

0
3
6
9
12
15
0
1
2
3
4
5
6
7
8
9
0
200
400
600
800
1,000
1,200
1,400
0.000
0.125
0.250
0.375
0.500
0.625
R
e
i
n
for
c
e
me
n
t S
l
i
p
,
s
(mm)
B
o
n
d

S
t
r
e
s
s
,

u

(
M
P
a
)

B
o
n
d

S
t
r
e
s
s
,

u

(
p
s
i
)

R
e
i
n
for
c
e
me
n
t S
l
i
p
,
s
(i
n
.)
CO
N
-U
C-1.0
E
CC-U
C-1.0
CO
N
-U
C-1.5
E
CC-U
C-1.5
0
3
6
9
12
15
0
1
2
3
4
5
6
7
8
9
0
200
400
600
800
1,000
1,200
1,400
0.000
0.125
0.250
0.375
0.500
0.625
R
e
i
n
for
c
e
me
n
t S
l
i
p
,
s
(mm)
B
o
n
d

S
t
r
e
s
s
,

u

(
M
P
a
)

B
o
n
d

S
t
r
e
s
s
,

u

(
p
s
i
)

R
e
i
n
for
c
e
me
n
t S
l
i
p
,
s
(i
n
.)
CO
N
-C-1.0
E
CC-C-1.0
CO
N
-C-1.5
E
CC-C-1.5

10


CONCLUSIONS AND FUTURE WORK


Conclusions


Four variations of beam splice
specimens were studied to examine the bond behavior in
Engineered Cementitious Com
posites (ECC), one type of HPFR
C,
and a

traditional normal
weight concrete. Eight beam specimens were tested in four
-
point bending with constant moment
and shear regions. T
he parameters
that were
varied
include

cover to bar diameter ratios and
amount of confinement in the splice region. Bond stresses were calculated from strain data, and
reinforcement slip was measured using LVDTs attached to the reinforcement.

The ECC spec
imens had higher bond strengths than the concrete specimens in all of the
experiments. Additionally, the concrete specimen failures were significantly more brittle since
there was a more sudden drop in bond strength after the maximum bond stress was reach
ed.
Residual bond strengths were higher in ECC than concrete. Bond
-
slip data also showed that
after the maximum bond strength was reached, bond strengths decreased at a lower slip rate for
ECC than concrete.


Future Work


Additional specimens are being
f
abricated
and tested
using two other HPFR
C mix types;
a Hybrid Fiber Reinforced Concrete (HyFRC)

(Blunt and Ostertag 2009)
,

which utilizes steel
fibers and PVA fibers, and a High
-
Performance Fiber
-
Reinforced Concrete which uses
hooked
steel fibers. These three mixes (ECC, HyFRC and HPFR
C) represent a range of emerging
materials for which characterizing bond with reinforcing steel is needed
.

As
HPFR
C materials
are good candidates

for earthquake
-
resistant construction, future
experiments are planned to examine bond strength det
erioration u
nder cyclic loads. Additional
experiments are also planned to investigate the bond
-
slip behavior in the inelastic regime.

The results of these experiments will be used to develop theoretical bond
-
slip curves for
different HPFR
C materials. These bond
-
slip
curves will then be implemented in finite element
programs to develop ro
bust analytical models for HPFR
C materials.



ACKNOWLEDGMENTS

T
he authors gratefully acknowledge funding from the John A. Blume Earthquake Engineering
Center and the National Science F
oundation (NSF) Graduate Research Fe
llowship Program
(GRFP)
. The assistance of undergraduate researcher Albert Alix, Stanford University, is
gratefully acknowledged.



REFERENCES

ACI Comittee 408, . (2003).
Bon
d and development of straight reinforcing bars in tension
.
Farmington Hills, Michigan, 49.

Blunt, J., and Ostertag, C. P. (2009). “Performance
-
Based Approach for the Design of a
Deflection Hardened Hybrid Fiber
-
Reinforced Concrete.”
Journal of Engineering
Mechanics
, 135(9), 978

986.


11

Chao, S.
-
H., Naaman, A. E., and Parra
-
Montesinos, G. J. (2009). “Bond behavior of reinforcing
bars in tensile strain
-
hardening fiber
-
reinforced cement composites.”
ACI Structural
Journal
, 106(6), 897

906.

Harajli, M. H., Hamad, B. S., and Karam, K. (2002). “Bond
-
slip response of reinforcing bars
embedded in plain and fiber concrete.”
Journal of Materials in Civil Engineering
, 14(6),
503

511.

Lee, W. K., and Billington, S. L. (2008). “Simulation of self
-
cent
ring fibre
-
reinforced concrete
columns.”
Proceedings of the ICE
-

Engineering and Computational Mechanics
, 161(2),
77

84.

Naaman, A. E., and Reinhardt, H. W. (2006). “Proposed classification of HPFRC composites
based on their tensile response.”
Materials a
nd Structures
, 39(5), 1

13.

Orangun, C. O., Jirsa, J. O., and Breen, J. E. (1977). “A Reevaluation of Test Data on
Development Length and Splices.”
Journal of the American Concrete Institute
, 74(3), 114

122.