You are on page 1of 11

Cracking and Tension Stiffening Behavior of

High-Strength Concrete Tension Members Subjected


to Axial Load
Gi-Yeol Lee and Woo Kim*
Department of Civil Engineering, Chonnam National University, Buk-gu, Gwangju, Korea 500-757
(Received: 9 November 2007; Received revised form: 30 April 2008; Accepted: 21 May 2008)

Abstract: This paper presents the test results of 35 direct tensile specimens to
investigate the effect of concrete strength on the tension stiffening effect and cracking
response in axially loaded reinforced concrete tensile members. Three concrete
strengths 25, 60 and 80 MPa were included as a major experimental parameter
together with six concrete cover thickness ratios. The results showed that as higher
strength concrete was employed, not only splitting cracks along the reinforcement
more extensive, but also the transverse crack spacing became smaller. Thereby, the
effective tensile stiffness of the high-strength concrete specimens at the stabilized
cracking stage was much smaller than those of normal-strength concrete specimens.
This observation is contrary to the current design provisions, and the reduction in the
tension stiffening effect by employment of high-strength concrete is much greater than that
would be expected. Based on the present results, a modification factor is proposed to
account for the effect of the cover thickness and the concrete strength.

Key words: bond, cover thickness, crack spacing, high-strength concrete, tension stiffening effect.

1. INTRODUCTION
In the conventional methods for designing reinforced
concrete members the concrete tensile resistance in a
section is generally neglected, and the force equilibrium
condition is based on reinforcement tension only. The
recent trend of the employing more accurate analysis
such as strength design method as well as higher
strength materials, has resulted in smaller member sizes
and higher tension stress in reinforcement. Thereby, it
is becoming increasingly necessity to take into account
for the deformation compatibility conditions in
designing reinforced concrete structures.
When a reinforced concrete member is cracked, the
stiffness is drastically decreased and the internal
stresses are redistributed in such a way that in the
cracked cross-section all tensile forces are balanced by
the steel only, while in the zone between adjacent

cracks, a part of tensile force is transmitted from the


steel to the surrounding concrete by bond forces. Thus
the concrete between cracks contribute to the stiffness
of the tensile member, and this effect is normally called
the tension stiffening effect (CEB-FIP 1991).
As shown in Figure 1, the cracking behavior and the
stiffness of a tension member depend mostly on bond
characteristics at the interface between the concrete and
the rebar. When the first crack occurs, the distribution of
strains in the steel and the concrete can be divided into
two parts of the stress disturbed region(D-region) as in
Figure 1(b), where the strains vary, and the stress
continuity region(B-region), where the strains are
uniform along the section. The strains in D-region
depend on the distribution of bond stress within the
transmission length lt. As indicated in Figure 1(d), the
cracking continues until the whole member consists of

*Corresponding author. Email address: wkim@chonnam.ac.kr; Fax: +82-62-530-1659; Tel: +82-62-530-1655.

Advances in Structural Engineering Vol. 11 No. 5 2008

127

Reinforcement

Concrete

(b)

1st crack 2nd crack

2nd crack
s , c

so
s = c

so

sm

so

N
L+

(a)

Axial load, N

Cracking and Tension Stiffening Behavior of High-Strength Concrete Tension Members Subjected to Axial Load

RC member
response

D-region B-region

Tension stiffening effect


f
ct
Es

Ncr

5
(c)

(d)

It

s , c

max
m

max

so

fct .(0.6 + n)
Es

so
s
c

Member strain, m

overlapping D-regions. The smallest possible crack


distance is found just at the end of a D-region. And the
maximum crack distance will be 2lt. Thus the crack
distance varies between lt and 2lt. A summary of
research carried out on tension stiffening and cracking
of structural concrete in direct tension is given by ACI
Committee 224 (2005) and FIB Manual (1999).
According to the design code for tension stiffening of
tension members in MC 90 (CEB-FIB model code
1990), the overall tension stiffening behavior is divided
into 4 loading stages and expressed in terms of the mean
steel strain sm as shown in Figure 2. The average strain
in the reinforcement in the stabilized cracking stage is
given by

sm = so

Uncracked stage
Crack formation stage
Stabilized cracking stage
Yielding of reinforcement
Bare bar

Figure 2. Simplified load-deformation response for a centrically


reinforced member subjected to tension defined by MC 90

Figure 1. Distribution of strain and bond stress in tension specimen


(CEB-FIP Model Code 1990) (a) Tension member (b) Single crack
stage (c) Bond stress at single crack stage (d) Stabilized crack stage

fct
Es

(1)

Where so is the strain in tension reinforcement at the


cracked section, is an experimental constant
accounting for the effects of tension stiffening with 0.40
for short term loading, and 0.25 for long-term or
repeated loading, fct is the tensile strength of concrete, Es
is the modulus of elasticity of steel, and is the ratio of
reinforcement.
Eqn. 1 is a simple expression based on the work
carried out by Leonhardt (1977) and Stroband (1991).
But their work did not include two major parameters of

128

1
2
3
4
5

It

cover thickness and concrete strength influencing the


tension stiffening effect. The splitting cracks resulting
from a low cover thickness of concrete is a major
influence on the tensile stiffness of axially loaded
members in which transverse cracks and splitting cracks
occur. The experimental investigation carried out by
Abrishami and Mitchell (1996) showed a significant
reduction in the tension stiffness by the splitting cracks
along the reinforcement, and they suggested that the
influence should be taken into account unless the cover
to bar diameter ratio c/db is more than 2.5.
Eqn. 1 also cannot express the effect of the concrete
strength that may alter the bond characteristics. The
work done by Azizinamini et al. (1993) and Hungspreug
(1981) indicated that, as shown in Figure 3, as higher
strength concrete was employed, not only a peak bond
stress become higher, but it also becomes more
concentrated near the loaded end. These characteristics
in bond behavior of high-strength concrete can be
quantitatively drawn from the elastic theory. Since
the elastic modulus of concrete is a function of
compressive strength, while that of steel remains constant,
the composite structural system consisted of
reinforcement and concrete is altered with concrete
strength, so that different stress state in the interface is
resulted in. Moreover, high-strength concrete is
generally more brittle than normal-strength concrete, and
in turn, less stress redistribution can take place at the
ultimate loading stage. These two material properties in
high-strength concrete may lead to making the D-region
shorter, thus, the crack spacing as well as the effective
tensile stiffness of axially loaded members may change.

Advances in Structural Engineering Vol. 12 No. 2 2009

Bond stress

Gi-Yeol Lee and Woo Kim

2. EXPERIMENTAL WORK
The test specimens in this study were, as shown in
Figure 4, axially loaded direct tension members had a
length of 1500 mm, containing single 19 mm bars of
yield stress 430 MPa at the center of the rectangular
section. The sectional area of the specimen was kept
constant of 23250 mm2 in all specimens, but had
rectangular shapes with different aspect ratios in order
to vary the cover thickness of c/db between 1.0 and 3.5
in 0.5 increments.
The concrete mix design is shown in Table 1, with
three types of concrete normal-strength (NSC, 25 MPa),
medium-strength (MSC, 60 MPa), and high-strength
(HSC, 80 MPa). The concrete compressive strength fck
was obtained by testing standard cylinders having a
diameter 100 mm and a height of 200 mm. All
properties of the 35 specimens are listed in Table 2. To
minimize the shrinkage during the curing procedure,
wet curing was provided for 3 days after concrete
casting until demoulding, and afterwards submerged
curing was provided for 4 to 6 weeks until the test.
For the loading a displacement control method was
employed, and the longitudinal elongation between the

N
High-strength
concrete
Normal-strength
concrete

Distance from loading end

Figure 3. Bond stress distribution of pullout test

Hence, in the present study an experimental program


was planned to evaluate the tension stiffness of axially
loaded members of high-strength concrete. Direct
tension tests were performed with concrete strengths
and six cover thickness to bar diameter ratios c/db as
major variables. Using the results obtained from the
experiments, the cracking behavior and the effect on
tension stiffening were examined.

LVDT
19 mm Deformed bar

1500 mm

LVDT

19 mm

b
c

db c

a
(See table 2 for dimensions)

Figure 4. Specimen geometry

Table 1. Mixture design of concrete


Design
Strength
(MPa)
25
55
85

Admixture
(kgf/m3)

Unit Weight (kgf/m3)

Slump

Test
Strength

W/C

S/A (%)

(mm)

S.F

M-150

(MPa)

0.60
0.25
0.20

45
37
38

150
150
150

385
550
650

229
138
153

810
662
578

991
1105
950

0
0
98

0
11
13

24.8
60.7
80.4

Note: C: cement, W: water, S: sand, G: gravel, S.F: silica fume, M-150: supper plasticizer

Advances in Structural Engineering Vol. 12 No. 2 2009

129

130
100 230
115 200
135 170
150 155

2.5

3.0

3.5

115 200

2.5

2.0

100 230

2.0

80 290

80 290

1.5

1.5

60 385

1.0

60 385

150 155

3.5

1.0

135 170

3.0

150 155

115 200

2.5

3.5

100 230

2.0

135 170

80 290

1.0

3.0

60 385

1.0

Cross
Sectional
Dimensions
a b (mm)

80.4

60.7

24.8

Compressive
Strength
fck (MPa)

3.53

3.34

1.96

Tensile
Strength
fct (MPa)

Note: In all specimens, a 19 mm bar was placed at the center of the section (fy = 430 MPa, = 0.0124).

N 10-A
N 10-B
N 15-A
N 15-B
N 20-A
N 20-B
N 25-A
N 25-B
N 30-A
N 30-B
N 35-A
N 35-B
M 10-A
M 15-A
M 15-B
M 20-A
M 20-B
M 25-A
M 25-B
M 30-A
M 30-B
M 35-A
M 35-B
H 10-A
H 10-B
H 15-A
H 15-B
H 20-A
H 20-B
H 25-A
H 25-B
H 30-A
H 30-B
H 35-A
H 35-B

Specimens

Cover
Thickness
Ratio
(c/db )

Measured Concrete
Strength

55
54
55
46
51
48
52
47
51
52
45
50
60
76
75
62
71
73
81
71
79
77
77
68
73
71
73
83
81
86
81
94
94
94
97

Transverse
Crack
Ntr (kN)
42
36
39
46
53
62
68
84
84
86
93
93
41
46
37
42
46
69
64
77
80
85
84
50
44
54
48
64
64
86
72
94
94
94
97

Splitting
Crack
Nsp (kN)

Initial Cracking
Load

Table 2. Specimen properties and test results

7
10
8
12
10
9
7
9
6
5
4
6
11
12
15
14
14
11
14
13
9
10
8
11
14
16
16
15
16
12
14
12
9
12
13

Number of
Transverse
Crack
(each)
188
136
167
115
136
150
188
150
214
250
300
214
125
115
940
100
100
125
100
107
150
136
167
125
100
880
880
940
880
115
100
115
150
115
107

Average
Crack
Spacing
(mm)

650(43)
690(46)
720(48)
810(54)
760(51)
590(39)
510(34)
290(19)
350(23)
260(17)
50(3)
180(12)
1100(73)
1010(67)
1160(77)
770(51)
880(59)
840(56)
680(45)
760(51)
680(45)
660(44)
540(36)
1020(68)
1170(78)
1000(67)
1070(72)
1060(71)
1100(73)
850(53)
900(60)
470(31)
500(33)
380(25)
370(25)

Length of
Splitting
Crack
mm (%)
Cracking and Tension Stiffening Behavior of High-Strength Concrete Tension Members Subjected to Axial Load

Advances in Structural Engineering Vol. 12 No. 2 2009

Gi-Yeol Lee and Woo Kim

130

fck = 25 MPa

120

110

110

100

100

90

90
Applied axial load (kN)

Applied axial load (kN)

120

130

80
70
60
50
Bare bar
40
30

c/db
1.0
1.5
2.0
2.5
3.0
3.5

80
70
60
50
Bare bar
40
30

20

c/db
1.0
1.5
2.0
2.5
3.0
3.5

20
Transeverse crack starting point

10
0
0.0

fck = 80 MPa

Splitting crack starting point


0.3

0.6
0.9
1.2
1.5
1.8
Average steel strain (103)

2.1

2.4

(a) NSC

Transeverse crack starting point

10
0
0.0

Splitting crack starting point


0.3

0.6
0.9
1.2
1.5
1.8
Average steel strain (103)

2.1

2.4

(b) HSC

Figure 5. Tension versus average strain responses of normal-strength and high-strength concrete specimens

ends of each specimen was measured as in Figure 4,


with four LVDTs mounted at each corner. Then the
complete response of each specimen is described by
plotting the applied tension versus the average member
elongation. The cracks observed during the loading
were marked and used in investigating the overall
cracking behavior.
3. RESULTS AND DISCUSSION
3.1. Effect of Cover Thickness
Figures 5(a) and 5(b) show the tension versus average
strain responses of normal-strength and high-strength
concrete specimens respectively. The comparison of the
curves resulted from having different cover thickness
shows that in spite of the same sectional area and steel
ratio the axial stiffness becomes significantly higher
with increasing the cover thickness.
The load at which the first transverse crack was
observed is defined as the transverse cracking load Ntr,
and the load for the first cracking parallel to the rebar is
defined as the splitting cracking load Nsp for each
specimen, and listed in Table 2. These are also marked
on the curves in Figure 5, and show that the axial
deformation after cracking is highly dependent on the
thickness of concrete cover.
From the response of the normal-strength concrete(fck =
25 MPa) specimens in Figure 5(a), it is clear that Ntr
of all 6 specimens was nearly identical, while the

Advances in Structural Engineering Vol. 12 No. 2 2009

splitting cracking load Nsp varied significantly with the


cover thickness. In the specimen having thin cover
thickness (c/db = 1.0 and 1.5), the splitting cracks
appeared before the transverse cracks, and extended
along the bar. Thus, there was larger deformation, and
the effect of tension stiffening was small. On the
contrary, in the specimens with thick cover thickness
(c/db = 3.0 and 3.5) the splitting cracks appeared at a
much higher loading stage, and did not affect the axial
stiffness of the members.
Figure 5(b) shows the tension versus average strain
responses of the specimens with high-strength concrete
(fck = 80 MPa). Comparing Figure 5(b) with Figure 5(a),
it can be observed that the axial strain in high-strength
concrete is larger than those in the normal-strength
concrete at the stabilized cracking stage although
the concrete compressive strength is three times higher.
This result is obviously contrary to that predicted by the
current code provisions of Eqn. 1. Another important
thing that can be observed in Figure 5(b) is that in the
HSC specimens the splitting cracks appeared when
earlier loading stage than the transverse cracking.
The transverse cracking load and the splitting
cracking load observed in each specimen are plotted in
Figure 6. It can be said that the transverse cracking loads
were almost the same regardless of the cover thickness
and increased as the concrete strength became higher.
This was quite natural because of the same sectional

131

Cracking and Tension Stiffening Behavior of High-Strength Concrete Tension Members Subjected to Axial Load

120

120

fck = 25 MPa

fck = 60 MPa
100

Cracking load (kN)

Cracking load (kN)

100

80

60

40

20

0
0.0

1.0 1.5 2.0 2.5 3.0


Cover thickness ratio, c/db

3.5

60

40

20

Transverse cracking load


Splitting cracking load
0.5

80

0
0.0

4.0

Transverse cracking load


Splitting cracking load
0.5

1.0 1.5 2.0 2.5 3.0


Cover thickness ratio, c/db

(a) NSC

3.5

4.0

(b) MSC

120

fck = 80 MPa

Cracking load (kN)

100

80

60

40

20

0
0.0

Transverse cracking load


Splitting cracking load
0.5

1.0 1.5 2.0 2.5 3.0


Cover thickness ratio, c/db

3.5

4.0

(c) HSC

Figure 6. Measured variation of transverse cracking load and split cracking load

area and the increase in concrete tensile strength.


However, the splitting cracking loads were hardly
affected by the concrete strength, but increased in
proportion to the cover thickness.
It can be observed in Figure 6(a) that the intersecting
point at which the two lines representing the cracking
loads meet is c/db = 2.0. This implies that in the NSC
specimens with thin cover thickness of c/db smaller than
2.0 splitting crack occurred earlier, and resulted in small
tension stiffness. While in the specimens with c/db > 2.5
the transverse crack formed earlier, and the effect of the
crack on the stiffness was insignificant.
This result coincides well with the results of
Abrishami and Mitchell (1996). However, as the

132

concrete strength increases, the value of c/db at the


intersecting point becomes higher, 2.5 for MSC and 3.0
for HSC as shown in Figures 6(b) and 6(c). This result
indicates that in high-strength concrete needs thicker
cover to prevent splitting cracks, and hence the criterion
of c/db = 2.5 which is usually applied to the normalstrength concrete is not applicable to high-strength
concrete.
3.2. Effect of Concrete Strength
Figure 7 shows a comparison of the measured loaddeformation response with the curves predicted by
Eqn. 1 for the specimens having thick cover. This
comparison clearly shows that the effect of tension

Advances in Structural Engineering Vol. 12 No. 2 2009

Gi-Yeol Lee and Woo Kim

130

130

c/db = 3.0

c/db = 3.5

120

120
MC 90
(HSC)

110

MC 90
(MSC)

100

90
Applied axial load (kN)

Applied axial load (kN)

MC 90
(MSC)

100

90
80
70
Bare bar

60
MC 90
(NSC)

50
40

80
70
Bare bar

60
MC 90
(NSC)

50
40

30

30

NSC (fck = 25 MPa)


MSC (fck = 60 MPa)
HSC (fck = 80 MPa)

20

NSC (fck = 25 MPa)


MSC (fck = 60 MPa)
HSC (fck = 80 MPa)

20

10
0
0.0

MC 90
(HSC)

110

10
0.3

0.6
0.9
1.2
1.5
1.8
Average steel strain (103)

2.1

0
0.0

2.4

0.3

0.6
0.9
1.2
1.5
1.8
Average steel strain (103)

(a) c/db = 3.0

2.1

2.4

(b) c/db = 3.5

stiffening at the stabilized cracking stage decreased


with increasing concrete strength. This result obviously
contradicts to the general knowledge that the effect will
be greater in high-strength concrete than normalstrength concrete because of increase in tensile strength,
as predicted by Eqn. 1.
The relative cracking loads of MSC and HSC with
respect to those in NSC are plotted in Figure 8. As the
compressive strength of the concrete increased by about
3 times from 25 MPa to 80 MPa, the transverse cracking
load increased by approximately 1.9 times, but the
splitting cracking load stayed constant. It is noted that
the tensile strength assumed to increased by 1.78 based
on the commonly used square root rule relating the
tensile strength and the compressive strength. From the
experimental results, the relative transverse cracking
loads in the specimens matches well with the increase in
the tensile strength of the concrete, so that the transverse
cracks are directly affected by the concrete tensile
strength.
However, the splitting cracking strength did not
increase in spite of 1.8 times increase in the concrete
tensile strength. From this result, it can be said that
the splitting cracks are not affected by the concrete
tensile strength but affected by the thickness of the
cover. From the number of transverse cracks formed up

Advances in Structural Engineering Vol. 12 No. 2 2009

Relative ratio of cracking load

Figure 7. Influence of concrete strength on tension stiffening effect

2.0

fck
fck,NSC

1.0

c/db
3.0
3.5
3.0
3.5
0.0

20

Type
Transverse cracking load
Splitting cracking load

40
60
80
Concrete strength, fck (MPa)

100

Figure 8. Relative ratio of cracking loads

to the yield stage in a specimen, the average distance


between adjacent cracks(crack spacing) was calculated
Table 2, and plotted in Figures 9 and 10 shows the final
crack configuration for each specimen.
These results show that the crack spacing is larger
with the increase in cover thickness. This result agrees well
with other results published including Broms (1965a,b).

133

Cracking and Tension Stiffening Behavior of High-Strength Concrete Tension Members Subjected to Axial Load

30

Crack space, s (cm)

25

NSC (fck = 25 MPa)


MSC (fck = 60 MPa)
HSC (fck = 80 MPa)

NSC

20
MSC
15

10

HSC

0
0.0

0.5

1.0
1.5
2.0
2.5
3.0
Cover thickness ratio, c/db

3.5

4.0

Figure 9. Average crack spacing

It can be noted that the wider crack spacing in the


specimen with the thinnest cover(c/db = 1.0) may be
attributed to the excessive splitting cracks at the early
loading stage.
One of the important facts observed in Figures 9 and 10
is that the crack spacing significantly decreases with the
increase in concrete strength. As the concrete
compressive strength was increased from 25 MPa to 80
MPa, the crack spacing decreased by 3050 percent for
all of the cover thicknesses. This result contradicts the
result of Abrishami and Mitchell (1996) that the crack
spacing increases for higher strength concrete.

3.3. Effect of Tension Stiffening Behavior


According to Eqn. 1 for MC 90, the tension stiffening
effect should have been largest in the HSC specimens,
since tensile strength of concrete fct is proportional to the
compressive strength while the elastic modulus of steel
Es and steel ratio stay constant. However, the results
from the experiments showed smaller tension stiffening
as shown in Figure 7. As mentioned previously, this is
clearly attributed to the early splitting cracks and
excessive progress along the rebar with increase in load.
On the concrete splitting cracks along the rebar, the
bond between the bar and the concrete is diminished, so
that the concrete is no longer able to share the tensile
force, in turns resulting in large deformation with small
stiffening effect.
Examining the tension stiffening effect with
variation of cover thickness (refer to Figure 5), it is
known that the effect is proportional to the cover
thickness, and Eqn. 1 is still valid in normal-strength
concrete. However, in high-strength concrete the rate
of increase in stiffening effect is much lower than that
in normal-strength concrete and significantly different
values are predicted by Eqn. 1. Even in thin covered
HSC specimens, there was almost no tension
stiffening effect.
From the above results and discussions, it can be
concluded that neither the effect of splitting cracks in
thin cover thickness nor the material characteristics of
high-strength concrete is not accounted in Eqn. 1.
Therefore, any reduction coefficient for accounting
those influences should be required to modify Eqn. 1.

N10

H10

N15

H15

N20

H20

N25

H25

N30

H30

N35

H35

(a) NSC

(b) HSC

Figure 10. Final crack configuration in normal-strength and high-strength concrete specimens

134

Advances in Structural Engineering Vol. 12 No. 2 2009

Gi-Yeol Lee and Woo Kim

= 0.4 st co 0.4

1.25

c/db = 2.5

1.00
Experimental coefficient st

4. FORMULATIONS
The experimental coefficient in Eqn. 1 is fixed with the
value of 0.4 for the short-term loading. As discussed
previously, the present experimental result shows that
the tension stiffness of the axially loaded members is
highly dependent on the cover thickness and the concrete
strength. It is observed that the coefficient of 0.4 in Eqn. 1
fits appropriately only for the normal strength and for
c/db value of above 2.5 Figures 5 and 7. This observation
indicates that tension stiffness of members after cracking
cannot be sufficiently predicted by the in Eqn. 1, so
that additional experimental coefficients are required to
account for the two parameters above.
Accordingly, two additional modification factors have
been incorporated in Eqn. 1 st and co are intended to
take into account for the effect of fck and c/db respectively,
and each factor has a value of less than unity.

0.75

st =

25

fck

1.0

0.50

0.25

0.00

25
50
75
Concrete strength, fck (MPa)

100

Figure 11. st - concrete strength relationship

(2)
1.50

E
= s ( so sm )
fct

(3)

Where, Es = 200,000 MPa, and = 0.0124. For defining


fct the following equation in FIB Manual (1999) is
employed.

fck = 25 MPa
1.25
Experimental coefficient co

The results of the present tests have been utilized to


evaluate the value of the modification factors in Eqn. 2. To
derive the value of the factors, Eqn. 1 is rearranged for .

co =

c/db
2.5

1.00

0.75

0.50
Proposed
0.25

f
fct = fcto ln 1 + ck
fcko

(4)

In which fcto = 2.12 MPa, fcko = 10 MPa, and fck is the


compressive strength. Then, by putting the mean sm
measured at the stabilized cracking stage in each
specimen into Eqn. 3, can be obtained.
It should be noted here that in every specimen a
significant shrinkage strain had been induced during the
curing, and resulted in early cracking in the concrete and
pre-compressive strain in the reinforcement. For this
reason, the deformations of some specimens in Figure 5
became larger than that of the bare bar. To account for this,
the gross shrinkage strain was assumed to be 200 10-6 for
every specimen, and every measured strain sm was
adjusted by this amount. As the result of the adjustment,
the for the specimen with fck = 25 MPa and c/db of 2.5
becomes 0.4 equal to the value given by Eqn. 1.
Using the process above, the relationship between the
concrete strength and the relative value of st is

Advances in Structural Engineering Vol. 12 No. 2 2009

1.0

0.00
0.0

0.5

Abrishami and
Mitchell (1996)
co = 0.8c/db 1

1.0
1.5
2.0
2.5
3.0
Cover thickness ratio, c/db

3.5

4.0

Figure 12. co - cover thickness relationship

determined by a regression analysis of the experimental


results as shown in Figure 11.
It can be seen that the effect of the concrete strength
on tension stiffening behavior is directly expressed in a
simple form as follows:
25
st = 1.0
fck

(5)

By the same manner, the effect of the cover thickness


was determined from Figure 12.

co =

c /db
1.0
2.5

(6)

135

Cracking and Tension Stiffening Behavior of High-Strength Concrete Tension Members Subjected to Axial Load

130

120

110

110

100

100

90

90
Applied axial load (kN)

Applied axial load (kN)

120

130

c/db = 1.5

80
Bare bar

70
60
50

NSC (fck = 25 MPa)

40

MSC (fck = 60 MPa)


HSC (fck = 80 MPa)

30

80
Bare bar

70
60
50

NSC (fck = 25 MPa)

40

MSC (fck = 60 MPa)


HSC (fck = 80 MPa)

30

NSC Predicted

20

c/db = 2.5

NSC Predicted

20

MSC Predicted
10
0
0.3 0.0

MSC Predicted
10

HSC Predicted
0.3 0.6 0.9 1.2 1.5 1.8
Average steel strain (103)

2.1

2.4

HSC Predicted

0
0.3 0.0

0.3 0.6 0.9 1.2 1.5 1.8


Average steel strain (103)

(a) c/db = 1.5

2.1

2.4

(b) c/db = 2.5

130
120

c/db = 3.5

110
100
Applied axial load (kN)

90
80
Bare bar

70
60
50

NSC (fck = 25 MPa)

40

MSC (fck = 60 MPa)


HSC (fck = 80 MPa)

30

NSC Predicted

20

MSC Predicted
10
0
0.3 0.0

HSC Predicted
0.3 0.6 0.9 1.2 1.5 1.8
Average steel strain (103)

2.1

2.4

(c) c/db = 3.5

Figure 13. Comparison between test results and predicted response

This expression is valid for the practical range of c/db


between 1.0 and 2.5, that is in which splitting cracks
are probable. As the relation is also compared with
the equation suggested by Abrishami and Mitchell
(1996) that may overestimate the effect of the cover
thickness.
In summary, a new experimental coefficient to
simultaneously account for the effect of the concrete

136

strength as well as the cover thickness can be expressed


as follows:
25 c / d b
= 0.4
0.4
fck 2.5

(7)

Eqn. 7 has been examined by comparing with the


experimental load-deformation curves as illustrated in

Advances in Structural Engineering Vol. 12 No. 2 2009

Gi-Yeol Lee and Woo Kim

Figure 13, and it can be seen that the new coefficient


provides good predictions at the stabilized cracking
stage.
5. CONCLUSION
In the present study, a total of 35 direct tension tests
were performed to investigate the basic tension
stiffening behavior of high strength concrete. The
strength of concrete and the cover thickness were
selected as the main variables to examine the tension
stiffening effect. It is observed that the tension stiffening
effect is highly dependent on the cover thickness and the
concrete strength. As the concrete strength becomes
higher, the tension stiffening effect becomes smaller for
the practical range of c/db less than 2.5. This is attributed
to it being more probable that splitting cracks along the
rebar is coincided with transverse cracks in highstrength concrete. The crack spacing between the
adjacent transverse cracks becomes narrower as higher
strength concrete is used, a reduction in the crack
spacing of 3050 percent as the concrete compressive
strength varied from 25 MPa to 80 MPa. A new
experimental coefficient that can be incorporated with
into the equation in MC 90 is proposed. The new factor
provides good predictions for the stiffness at the post
cracking stage in axially tensioned reinforced concrete
members.
ACKNOWLEDGEMENTS
This work is a part of a research project supported by
Korea Ministry of Construction & Transportation
(MOCT) through Korea Bridge Design & Engineering
Research Center at Seoul National University. The authors
wish to express their gratitude for the financial support.
REFERENCES
Abrishami, H.H. and Mitchell, D. (1996). Influence of splitting
cracks on tension stiffening, ACI Structural Journal, Vol. 93,
No. 6, pp.703710.
ACI Committee 224. (2005). ACI Manual of Concrete Practice Part 2: Control of Cracking in Concrete Structures, American
Concrete Institute, Detroit, USA.

Advances in Structural Engineering Vol. 12 No. 2 2009

Azizinamini, A., Stark, M., Roller, J.J. and Ghosh, S.K. (1993).
Bond performance of reinforcing bars embeded in high-strength
concrete, ACI Structural Journal, Vol. 90, No. 5, pp. 554561.
Broms, B.B. (1965). Stress distribution in reinforced concrete
member with tension cracks, ACI Journal, Vol. 62, No. 9,
pp. 10951108.
Broms, B.B. (1965). Crack width and crack spacing in reinforced
concrete members, ACI Journal, Vol. 62, No. 9, pp. 12371256.
CEB-FIP. (1991). CEB-FIP Model Code 1990, Comite EuroInternational Du Beton, Paris, France.
FIB(CEB-FIP). (1999). Structural Concrete: Textbook on Behavior,
Design and Performance - Volume 1, International Federation for
Structural Concrete, Lausanne, Switzerland.
Hungspreug, S. (1981). Local Bond Between a Reinforced Bar and
Concrete under High Intensity Cyclic Load, Structural
Engineering Report No. 816, Cornell University, Ithaca, NY,
USA.
Leonhardt, F. (1977). Crack Control in Concrete Structures, IABSE
Surveys No. S-4/77, IABSE Periodical 3/1977, International
Association for Bridge and Structural Engineering, Zurich,
Switzerland.
Stroband, J. (1991). Experimental Research into the Bond Behavior
of Reinforcing Bars in Light weight and Normal Weight
Concrete, Report 25.5-91-3/VF C, Delft University of
Technology, Netherlands.

NOTATION
c
cover thickness
db
bar diameter
Es
modulus of elasticity of steel
fck
compressive strength of concrete
fct
tensile strength of concrete
lt
transmission length
Nsp splitting cracking load
Ntr transverse cracking load

tension stiffening factor


co experimental coefficient of cover thickness
st experimental coefficient of concrete strength
sm average steel strain
so steel strain at the cracked section

reinforcement ratio

bond stress

137

You might also like