P. 1
Bond Characteristics High-Strength Steel Reinforcement

Bond Characteristics High-Strength Steel Reinforcement

|Views: 4|Likes:
Published by Umair Baig
Civil Enginering
Civil Enginering

More info:

Categories:Types, Research
Published by: Umair Baig on Apr 17, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

04/17/2013

pdf

text

original

ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no, 103-878
Bond Characteristics of High-Strength Steel Reinforcement
by Raafat EI-Hacha, Hossam EI-Agroudy, and Sami H, RizkalJa
This paper summarizes an investigation undertaken to study the
bond characteristics oj high-strength steel reinforcement bars
commercially known as microcomposite, multistructural, fonnable
steel (MMFX). The objective a/the investigation is to examine the
applicability of the ACI 318-02 equation and a current proposed
equation by Zuo and Danvin on bond behavior of steel reinforcement
10 the concrete member. The experimental program included two
phases. The jirsfphase of the experimental program consisted oj
testing four beam-end specimens reinforced with MMFX steel
bars, whereas the second phase included testing eight beam-splice
specimens reinforced with MMFX steel bars. The selectedfour factors
considered in this study were bar size, level of confinement,
bonded length, and bar cast pOSition. The bond behavior of the
MMFX steel bars was found to be similar to that of cOIwentional
Grade 420 MPa (60 ksi) steel up to the proportional limit of
550 MPa (80 ksi). The bond strength of the MMFX significantly
changes as the tensile stresses developed in the bar exceed the
proportional limit. The test results indicated that both the ACI 318-02
equation and the current proposed equation by ZliO and Darwin on
bond are adequate and resulted in conservative prediction at low
stress levels up to 550 MPa (80 ksi). At high stress levels, however, the
prediction using both equations is unconservative due to the
nonlinear behavior of the MMFX stress-strain relationship.
Based on the limited number of specimens considered in this
study, modification (0 both the ACI 318-02 equation and the
Zuo and Darwin equation is proposed to predict the bond forces
beyond the proporlionailimitfor MMFX steel bars.
Keywords: bond; confinement; flexure; splice; steel; strength.
INTRODUCTION
Bond behavior of reinforcing steel bars to concrete is one
of the most important mechanisms that should be properly
designed to ensure satisfactory performance of reinforced
concrete structures, The bond strength and mode of bond
failure are affected by many factors, The most important
factors are thickness of the clear concrete covers (bottom
and/or side), clear spacing between bars, nominal bar diameter,
embedment or development and splice length, amount of
transverse steel reinforcement, and concrete compressive
strength, The individual contributions of these factors are
difficult to separate or quantify, Another factor that influences
the bond strength of bars is the depth of fresh concrete below
the bar during casting.
In general, any increase in confinement of the bar by the
surrounding concrete, and/or by transverse reinforcement
increases the bond strength and minimizes splitting.
Confinement by the concrete is dependent on the clear
concrete covers (bottom andlor side) and the bar spacing,
Increasing the development/splice length of a reinforcing bar
increases its bond strength, The bond strength, for a given
length, mobilized by both concrete and transverse reinforcement,
increases as the bar diameter increases. Bond strength of bars
confined by transverse reinforcement increases with the
increase in the relative rib area. Top-cast bars have lower
bond strength than bottom-cast bars, Also, bond strength
ACI Structural Journal/November-December 2006
increases with increasing concrete compressive strength for
bars not confined by transverse reinforcement approximately
with the 1/4 power of the compressive strength (f; 1/4), The
additional bond strength, provided by transverse reinforcement,
increases approximately with the 3/4 of the compressive
strength 3/4), Note that the 14 has been shown to
provide a better representation of the effect of concrete
strength on bond than 112 This point is recognized by
ACI Committee 408 and within ACI 318, which sets an upper
limit on the value of 1/2 for use in design. An increase in the
aggregate quantity and strength results also in an increase in
bond strength, More details on the factors that affect the bond of
reinforcing steel to concrete can be found in ACI 408R-03
1
Bond characteristics of cooventional carbon steel reinforce-
ment and epoxy-coated reinforcement with concrete has
been thoroughly investigated by many researchers
2
-
12
and
addressed in terms of bond or development length, Their
experimental results contributed to the ACI Committee 408
database on "Bond and Development of Straight Reinforcing
Bars in Tension" and were used in formulating the current
equations in both ACI 318-02
13
and ACI 408R-03
1
to
predict the bond force,
The experimental investigation presented in this paper is
designed to study the bond behavior of the high-strength
steel (conunercially known as MMFX steel bar) and included
the effect of bar size, level of confinement with transverse
reinforcement, bonded length, and bar cast position, The
MMFX steel bars exhibit superior mechanical properties
when compared with conventional steel reinforcement,14
and the requirements covering deformation dimensions of
ribs (length, height, and frequency along bars) are the same
as conventional steel bars and confonn to ASTM A 1035-06
15
and ASTM A 615-04,16 The validity of such innovative
reinforcement and its ability to transfer stresses to the
concrete through bond must be considered, To
the best knowledge of the authors, very little information is
available about the bond strength of high-strength steel
reinforcement. Therefore, MMFX steel reinforcement was
considered to provide a database and kTIowledge of the bond
of high-strength steel to concrete and to compare the
behavior to conventional A61S Grade 420 MPa (60 ksi)
carbon steel reinforcement.
RESEARCH SIGNIFICANCE
Typical design code equations attempt to ' predict a
required bonded length that will result in yielding of the bar
before bond failure, The importance of conducting an
experimental investigation to identify the bond characteristics
ACI Structural Journal, V. 103, No.6, November-December 2006.
MS No. 03-378 received July 25. 2005, and reviewed under Institute publication policies.
Copyrighl © 2006, American Concrete Institute. All rights reserved, including the making
of copies unless pennission is obtained from the copyright proprietors. Pertinent discussion
including author's closure, jf any, will be published in the September-October 2007
ACI Structural Journal if the discussion is received by May I, 2007.
771
Table 1-Test matrix of experimental Program 1: beam-end specimens
MMFXbar Concrete dimensions
location
Beam ID n during cast
lb, m.m
L,mm b,mm h.mm
Bl I No. I3 Bottom 356 2032 356 508
B2 I No. I3 Top 356 2032 356 508
B3 I No. 13 Bauam 356 2032 356 508
B4 I No. 2S Bottom 864 2032 356 508
>N""O"'''go' l' :"'\ ............
(211 6 Gnde 60) sea mm
(20 in)
4No.2S Oracle 420 . .

MM!'X b&r L .1
NOi l orNo2S {1'1- ) S6mm( 14in)fOrNo. 13 (II4) MMI'X
01118) ' (14 in) Sanded ungth L. e 864 mm (14 in) for No.2S (_8) MMFX
I 2032 mm -----I
(80 in)
Fig. 2-Details of beam-end specimens.
possible conical failure. the flrst 102 mm (4 in.) of the
MMFX steel bar from the concrete surface at the loaded end
were debonded using a plastic tube. as shown in Fig. 2. The
test matrix for the beam-end specimens program is shown in
Table 1. Details of the test setup are shown in Fig. 3.
The MMFX steel bar was tensioned using a hydraulic jack.
and the specimen was held in place using steel beams and
high-strength Dywidag steel bars anchored to the laboralory
floor. as shown in Fig. 3. The applied tension load was
measured using a load cell placed at the jacking side of the
beam-end specimen. Three 6 mm (0.25 in.) 120 ohm electrical
resistance strain gauges. installed on the bonded surface of
each MMFX steel bar. were used to measure the strain distribu-
tion along the bonded length of the bar. lb. at distances of Ib/6.
Ib12. and 5I
b
/6. A 50 mm (2 in.) extensometer was used to
measure the strain of the unbonded loaded Uacking) end of
the MMFX steel bar. Both the loaded and the unloaded
(dead) end slippage were measured using a linear variable
differential transformer (L VDT).
The bearn-end specimens were cast using normal-strength
concrete of 41.4 MPa (6000 psi) supplied by a local concrete
plant. Standard concrete cylinders 102 x 204 mm (4 x 8 in.)
were caSI according to ASTM C 31_00
23
for the purpose of
determining the compressive strength of the concrete. The
concrete cylinders were cured in the same manner as the test
specimens. The average concrete compressive strength for
each specimen, as measured on the day of testing using three
cylinders according to ASTM C 39-01,24 is given in Table 1.
Results of beam-end specimens
Failure mode-Beams BI. B2. and B3. each reinforced
with one No. 13 (No.4) MMFX steel bar. failed by rupture
of the MMFX bar. As the applied load increased. the number
and width of the flexural cracks developed along the bonded
length of the MMFX steel bar increased. The spacing
between these cracks was approximately equal to the spacing
between the stirrups. No signs of bond failure were observed
in these beams. It was concluded that the bonded lengths
were greater than the development lengths required to
develop the ultimate strength of the bar for the levels of
confinement provided in the three beams. For the beam
reinforced with one No. 25 (No.8) MMFX steel bar
(Beam B4). a longitudinal crack developed in the concrele
cover at the loaded end and propagated toward the unloaded
end of the specimen along the steel bar. indicating a typical
ACI Structural Journal/November-December 2006
Concrete cover Stirrup details
d. mm
db. mm C
so
• mm Cb,mm MPa S, mm dst, mm i YI ,mm
464
464
464
457
12.7 172 38 36. 1 229 9.5 420
12.7 172 38 36 229 9.5 420
12.7 172 38 36.06 76 9.5 420
25.4 165 38 36.7 229 9.5 420
"
Fig. 3- Test setup of beam-end specimens.
Fig. 4--Splitting bondfailure mode for beam-end Specimen B4
(top view).
splitting mode of bond failure. as shown in Fig. 4. An
excessive increase in the width of the first flexural crack
near the loaded end was observed. Figure 5 shows the Iensile
stress and the average bond stress developed in the MMFX
steel bar in Beam B4 versus the measured slip at the loaded
and unloaded ends of the bar.
Bond distribution- Figure 6 shows the bond stress distri-
bution Ub; at distance x; from the loaded end along the
bonded length of the MMFX steel bar for Beam B I at
different stress levels based on the linear behavior of MMFX
within the elastic range to determine the stress /;. corresponding
to the measured strain. Figure 6 was developed as follows: at
a given bar stress level, the .corresponding strain readings
along the bonded length. as measured by the three strain
gauges, were determined. Using the stress-strain curve
oblained from the mechanical properties of the MMFX steel
bar. the corresponding stress levels were obtained for each
strain gauge reading along the bar. These stress values were
773
Table 3-Test matrix of Experimental Program 2: beam-splice specimens
Concrete di mensions

Group Specimen
no. lD
(,. mm
L. mrn
b
w
• mm Bf , mm
h. mrn
I B-6-12 305 4877 305
-
311
11 B-6-24 610 4877 311
-
362
III B-6-36 914 4877 305 61 3 464
IV B-6-60 1524 6096 305 121 9 457
I B-8- 12 305 4877 305 - 305
11 B-8-24 610 4877 305 356
III B-8-48 1219 4877 311 616 470
IV B-8-n 1829 6096 305 1219 470
.
DimenSIOns of specimens were measured after castmg.
specimens allows only rotation of the section without deflection
of the beam at the location of bonded length. Deflection of
flexural beams simulated by the beam-splice specimens
allows for deflection of the beam, which reduces the bond
strength of the spli ced bars as the deflection forces the
spliced bar to exert additional outward pressure on the
bottom concrete cover, causing a premature failure of the
splice. TIris behavior is supported by Ferguson and Thompson.
27
Phase II: beam-splice specimens
A total of eight large-scale concrete beams reinforced with
MMFX steel bars splicea at the midspan were tested. Four
specimens were each reinforced with two No. 19 (No.6)
MMFX steel bars and the other four with one No. 25 (No.8)
MMFX steel bar. The beam-splice specimens were divided
into four groups, as shown in Table 3. Each group consisted
of two specimens with identical concrete dimensions but had
different amounts (one or two) and sizes (No. 19 or No. 25
[No.6 or No.8]) of the reinforcing MMFX steel bars. The
specimens in Groups I and II were rectangular in cross
section, whereas those in Groups III and IV had T-shaped
cross sections. The spliced lengths varied from one specimen
to another and ranged from 305 to 1829 mm (1.0 to 6.0 tt),
as shown in Table 3. To minimize the effect of the applied
loads on the spliced length, the distance between the end of
the splice length and the center of the applied load was
always more than 305 rum (1.0 tt) . For the specimens reinforced
with No. 25 (No.8) MMFX steel bars. double-legged closed
stirrups were evenly di stributed along the splice length to
provide the required level of confinement around the spliced
bars. To prevent possible premature shear failure, shear
reinforcement was provided using No. 10 (No.3) Grade 60
double-legged closed stirrups spaced at 127 rum (5 in.) along
the shear span for all tested beam specimens. Compression
reinforcement was provided by two No. 13 (two No.4)
Grade 420 MPa (60 ksi) steel bars as top reinforcement. The
variation in tbe beam's dimensions was selected to achieve
different stress levels in the MMFX steel bar length at failure.
The bottom and side concrete covers and the transverse spacing
between the spliced bars were kept constant for all the beams
reinforced with two No. 19 (two No.6) MMFX steel bars.
The selected values were 1. 8d
b
, 3d
b
, and 6d
b
• where db is the
bar diameter. For beams reinforced with one No. 25 (No.8)
MMFX steel bar, the bottom and side concrete covers were
al so kept constant at values of 1.375d
b
and 5d
b
, respectively.
Table 3 shows the actual measured dimensions of the specimens
after casting.
The test matrix for the beam-splice test program is shown
in Table 3. The first letter of the beam designation "B" stands
for Beam; the middle number identifies the bar size- No. 19
ACI Structural Journal/November-December 2006
Cover and spacing Stirrup details along spli ce length
d. mrn Csa.mm Csi, mm Cb,mm
N S. mrn ds" mm fyl' MPa
264
314
419
143
254
305
422
410
~


J
~

0
m
~
0
~
~
~
~
E

~
60.3 54.0 38 0 - - -
54.0 70.0 38 0
- - -
54.0 60.3 35 0 - -
57.2 57.2 35 0 - -
-
123.8 - 38 I 305 9.53 420
123.8 38 2 305 9.53 420
114.3 - 35 5 245 9.53 420
120.7 - 35 7 260.4 9.53 420
0.9
m LImit confinment level I
IiiI Actual conflnrnent level
0.'
0.'
0.8
0.5
0.'
0.3
81, No. 13 (/II4) MMFX B2, No.13(/II4) MMFX 83, NQ.13 (II4) MMFX B<I .No.25(II8) w.t=X
Zuo and Darwin (2000) Equation
Fig. 7- Experimentallpredicted ratio for bonded length for
beam-end specimens using Zuo and Dmwin12 equation.
£
c

~
~
~

0
m
~
0
~


j
~
f
~
~

~
0.'
0. '
0.'
0.5
0.'
0.3
0.2
0.'
0
III Urnlt confinrnent level :
iii Actual confinment level
81 , No.13(#4)MMFX 82, No. 13 (AI4)MMFX 83. No.13 (#4) MMF"X 84, No.25 (/l8)MMFX
ACI318-02 Code Equation
Fig. 8-Experimentallpredicted ratio for bonded length f or
beam-end specimens using ACI 318-02
13
equatioll.
and No. 25 (No.6 and No. 8); whereas the last number represents
the spliced length of the bar, in inches. Standard concrete
cylinders, 102 x 204 rum (4 x 8 in.), were cast according to
ASTM C 31_00
23
for the purpose of determining the compres-
sive strength of the concrete. The concrete cylinders were
cured in the same manner as the test specimens. The average
concrete compressive strength, determined using three
cylinders according to ASTM C 39-01 ,24 at the age of 28 days
was 41.8 MPa (6071 psi). Table 4 shows the concrete
compressive strengths as measured on the day of testing.
All beams were simply supported loaded in four-point
bending. The load was applied using an MTS actuator operated
775
Fig. 12-Failure of beam-splice Specimen B-8-48.
just before failure, calculated using the moment curvature
analysis, are given in Table 4.
Evaluation of bond strength and splice length
The experimental results of the bond force of No. 19 and
No. 25 (No.6 and No. 8) MMFX steel bars were compared
with the rredictions from the proposed by Zuo and
Darwin 1 and the ACI 318-02 3 equation. The bond force
was calculated from the bar stress detennined using the
experimental stress-strain curve of the MMFX bar (Fig. I)
for the corresponding measured strain reading in the bar, as
measured by the strain gauges attached to the MMFX steel bars.
Zuo and Darwinl2 equarion-The bond capacity of the
No. 19 and No. 25 (No.6 and No.8) MMFX steel bars was
evaluated using the equation proposed by Zuo and Darwin,I2
as shown in Fig. 13 and 14, re';j'ectively. Test results indi-
cated that the Zuo and Darwin 1 equation provided conser-
vative prediction of the bond capacity of No. 19 (No.6)
MMFX steel bars at stress levels up to the proportional
strength of 579 MPa (84 ksi), and provided a very close
prediction to the bond capacity of No. 25 (No.8) MMFX
steel bars up to the proportional strength of 607 MPa (88
ksi). At higher stress levels; however, the equation resulted
in unconservative prediction for both No. 19 and No. 25
(No.6 and No.8) MMFX steel bars. Test results suggested
that there is a need to modify the Zuo and Darwinl 2 equation
to include a higher stress level. The stress limitation imposed
by ACI 318-02
13
for the maximum allowed design yielding
strength of 550 MPa (80 ksi) has been selected as an upper
boundary for using the equation by Zuo and Darwin12
Therefore, modification of the equation has been proposed
beyond the stress level of 550 MPa (80 ksi).
Bond equations generally relate bar stresses to bond
lengths. The equation proposed by Zuo and Darwinl2 for bond
strength of bars not confined with transverse reinforcement
(Tb = Tc and Ts = 0), in terms of bond force T
b
, as is the case
for Beams B6-12, B6-24, B6-36, and B6-60, is given by
SI units
Imperial units
ACI Structural Journal/November-December 2006
Splice Length (In)
"
20 30 40 50 60 70
::, ,,[1.43/,k_ +O.Sd.l+S6.ZA.(O.JC- +0.9) (Sf uMirs) 80000
J. c ...
"'0000
60000
L'wTent Luo and OllWln (WUU) cqva\1OII
TCSlRaults
No. 19 (N6)
MMFXBar
_ .
Modified ZUG . nd Dlrwin (2000) equl lion
;'" -[! I.(c_ .. O.ScI,)+ 120A, I 0.1:: +0.9) (Sf tlMIIJ)

o 250 51)0 750 1000 1250 1500 1750
Splice Length (mm)
""00
60000
g

' 0000
<
o
"""0 ·
20000
"x,oo
Fig. 13-Bond force versus splice length for beams reillforced
with No. 19 (No.6) MMFX steel bars using Zuo and
Darwin
J2
equation.
_ '00000
Z
. """""
"
& """"'"
21 250000
o
• 200000
"'''00
,-
"""'"
o 10
Splice length (In)
20 30 ., 50 70

o 200 500 150 1000 1250 1500 1150
Splice length (mm)
10000
Fig. 14-Bolldforce versus splice length for beams reinforced
with No. 25 (No.8) MMFX steel bars using Zuo and
Darwin12 equation.
To limit the applicability of Eq. (2a) and (2b) to cases in
which a splitting failure governs
SI orimperial units
Equations (2a) or (2b) were modified by changing only
the numerical constants to ensure conservative predic-
tions at any stress level between 550 and 831 MPa (80 and
120.5 ksi). The proposed modification of the Zuo and
Darwin 12 equation for MMFX beyond the stress level of
550 MPa (80 ksi) is
SI units
(3a)
777
Imperial units
Equations (6a) and (6b) can be rewritten in terms of Id
as follows
SI units
Id
!ilL
I
=
db 10 J1[ (" ;";n + 0.5db)
db
(7a)
Imperial units
Id
llL
I
=
db 40 J1[ (C ;";n + O.5db)
db
(7b)
Equations (7a) and (7b) were written in terms of the yield
stress. To determine the development length of the bar in the
tested beams. however. this stress should be taken as the actual
measured stress in the MMFX steel bars.
Based on the tested results. Eg. (6a) or (6b) was modified by
changing the numerical constants and addition of a term func-
tion of d/. The proposed modification of the ACI 318_02
13
equation for MMFX beyond stress level of 550 MPa (80 ksi) is
SI units
Tb T, 1. 2
- = -= - ldn( e . + O.5d
b
) + 40d
b
9 min
Imperial units
(8a)
(8b)
Equations (8a) and (8b) can be rewritten in terms of I
d
• as
follows
SI units
lL- sl
Id
=
Jj;
db
;";n + O.5db)
(9a)
9 db
Imperial units
lL - 637
Id
=
Jj;
(9b)
db 52(C ;";n + O.Sdb)
10 db
ACI Structural Journal/November-December 2006
Splice Length (in)
0
" "
30
" "
60
" 600000
T. 5 ( . ) ,
"0000
550000
,; • IS '," '- +O.Sd, .. IS"' IOJ.4, ..
(SI
" 0000
500000
ACI JI8-02 Codecq\ll.Uon 1'0000
"0000 700000
_ 400000
Test Results No. 25 90000
(*8) MMFX Bar
80000
GI 350000 .
"
"'000
"
& 300000 Limiw"", IOttwSInM
&
2SOOOO
- . - . - - - (AClJis-:ii2i - . 60000

<
0
60000
0
= """
=
"'"
"0000
r, e' (' OSd) 4OJ'J - . A..I"
30000

700000
.1:- 9,trc ... + .• + ," SOOOm
'DODO
50000
70000
0
0 250 500 750 7000 1250 1500 1150
Splice Length (mm)
Fig. 16-Bondforce versus splice length for beams reinforced
with No. 25 (No.8) MMFX steel bars using ACI 318-02
13
equation.
Using Eq. (8a) or (8b) to predict the bond capacity of No. 19
(No.6) MMFX steel bars for stress levels higher than 550 MPa
(80 ksi). the maximum test/predict ratio was approximately
1.19. as can be seen in Fig. 15.
The current ACI 318-02
13
equation. for bars confmed with
transverse reinforcement (Tb = T, + T,). in terms of bond force
T
b
• as is the case for Beams B8-l2. B8-24. B8-48. and B8-72. is
given by
SI units
Tb = T, + T,= 2. l
dn
(c'. + 0 5d
b
) + 5 nldA"/,, (lOa)
/;112 18 ,",". 181O.34sn
Imperial units
(lOb)
Equations (lOa) and (lOb) can be rewritten in terms of Id as
follows
SI units
Imperial units
and k = A,,!,,
" ISOOsn
(lla)
779
, ."
900
..
l
100
!.

""

g
""


m
. 00



300

'"
'00
0
0
CIlo.2>.... '*2.9
__ __ ____ _
lim;la1ioft fur bu strtu l8Oksi)
-_._ . _._- -_._._- --- ---_ . _ -----
\0
(ACI 3I B_0l)
C"'-2SIHl
aZ
.
8
20
_ Experlmantal for No.19 (116) MMFX Steel Bin
_Experimentill for No.25 (118) MMFX Steel Bars
C : Level of Confinment
30 50 60 10 00
SptJce Length/Bar Diameter (L .ld b)
'"
130
120
'"
100 ';i
"-
90
j
00
10

60
m


"


"
30
20
"
90
Fig. 19-5teel bar stress versus splice length/bar diameter
for beams reinforced with No. 19 (No.6) and No. 25 (No. 8)
MMFX steel bars.
Midspan Deflection (in)
0 0.'
,.,
2
225
50
200 ' 6-&60 (p =0.45%)
45
'"
B-{J·72 (p -0.40%)
40

35
'i:
"-
125
30


"
25 .9
1
100
'ii
20


75
'ii
<

"
<
50
,.
25 ,
0 0
0
,.
20 30 40 50 60
Midspan Deflection (mm)
Fig. 20-Load-midspan deflection for beams reinforced
with No. 19 (No.6) and No. 25 (No.8) MMFX steel bars.
is nearly linearly related, but not proportional, to the splice
length to the bar diameter ratio (L,ld
b
) up to the minimum
yield strength for No. 19 and No. 25 (No.6 and No.8)
MMFX steel bars. The relationship also suggests that a
splice length of 30d
b
can be safely used to achieve the
maximum yield strength of 550 MPa (80 ksi) limited by
ACI 318-02.
13
As shown in Fig. 19, a splice length of 45d
b
can be used to achieve the yield strength of758 MPa (110 ksi)
for MMFX steel bars. The linear, not proportional, relationships
extend to a stress of 831 MPa (120.5 ksi), which corresponds
to a splice length of 50d
b
. Beyond the yield strength, the
relationship becomes highly nonlinear and significant splice
length is required to achieve higher stress levels, which
could be impractical to use for typical applications.
Load-midspan deflection-The load-midspan deflection
behavior for the beam-splice specimens reinforced with two
No. 19 (No.6) MMFX bars and specimens reinforced with
one No. 25 (No.8) MMFX bar with different splice lengths
ranging from 305 to 1829 mm (12 to 72 in.) are shown in
Fig. 20. Two important points are worth noting. As can be
seen, for the beams reinforced with two No. 19 (No.6)
MMFX bars or with one No. 25 (No.8) MMFX bar, the
behavior is affected by the splice length in which increasing
the splice length increased the deflection at ultimate. By
comparing the beams with tbe same splice length (such as
ACI Structural Journal/November-December 2006
B-6-12 and B-8-12) throughout the post-cracked portion of
the load-midspan deflection, and at any given load level, the
deflection for the beams with a higher reinforcement ratio is
always less than that for beams with a lower reinforcement
ratio. This behavior is due to the higher stiffness as a result
of the higher reinforcement ratio. Throughout the post-
cracking behavior, the stiffnesses presented by the slopes of
the curves were also different. For beams with a lower
reinforcement ratio, the slope was less than for beams with a
higher reinforcement ratio. The difference was found to be
closely matching the difference in the reinforcement ratio
between the beams with higher reinforcement ratios and
the beams with lower reinforcement ratios. The ultimate
load-carrying capacity is bigher for beams with higher
reinforcement ratios than beams with lower reinforcement
ratios. Note that the results of Beams B-6-36 and B-8-48 are
not shown in Fig. 20 due to problems that occurred during
the test in measuring the deflection at midspan.
CONCLUSIONS
The following conclusions were made based on the limited
number of tested specimens:
1. Bond behavior of the MMFX steel bars is similar to that
of the conventional Grade 420 MPa (60 ksi) carbon steel up
to the stress level corresronding to the proportional limit,
imposed by ACI 318-02, 3 of 550 MPa (80 ksi). At higher
stress levels, bond failure changed from the typical sudden
and brittle failure, normally observed for conventional steel,
to a gradual and ductile failure due to the nonlinear behavior
of the MMFX steel bars in this range;
2. The nonlinear ductile response of the MMFX bars at
high stress levels beyond proportional limit strength bas a
strong influence in reducing the bond strength of the MMFX
bars compared with Grade 420 MPa (60 ksi) steel;
3. The current eguations proposed by Zuo and Darwin
l 2
and ACI 318_02
13
for bond force provided conservative
prediction of the bond capacity for No. 19 and No. 25 (No.6
and No.8) MMFX steel bars up to 550 MPa (80 ksi). For
stress levels exceeding 550 MPa (80 ksi) and up to a stress
level of 831 MPa (120.5 ksi) for No. 19 (No.6) MMFX bars,
and 955 MPa (138.5 ksi) for No. 25 (No. 8) MMFX bars,
respectively, both equations were modified to provide better
prediction of the bond force capacity; and
4. The splice length to bar diameter ratio is nearly linearly, but
not proportionally, related to the induced stress in the MMFX
steel bar up to yield strength. The relationship becomes highly
nonlinear beyond a stress level of758 MPa (llO ksi).
ACKNOWLEDGMENTS
The authors would like to thank the technical staff and graduate students
at the Constructed Facilities Laboratory at N.C. State Uni versity for their
help with the laboratory work. The authors are grateful to the support
provided by MMFX Tcchnologies Corp. for donating the steel materials.
NOTATION
A
ty
area of each stirrup or tie crossing potential plane of splitting
adjacent to reinforcemenl being developed or spliced
Bf width of flange of beam
b
w
width of web of beam
Cb thickness of clear bottom concrete cover
Cm/U = maximum of cb or C
s
minimum of concrete covers surrounding bar or half clear spacing
between bars, minimum of csi and (cb or cso)
cmill minimum of cb or C
s
C
s
minimum of e
so
or csi + 6.35 mm (Csi + 0.25 in.)
esi half clear spacing between spliced bars
C so thickness of clear side concrete cover
781

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->