Professional Documents
Culture Documents
CONCRETE MEMBERS
ABSTRACT
KEY WORDS
Cracking; crack width and spacing; creep; deformation; shrinkage; restraint;
reinforced concrete; slabs; time-dependent behaviour; laboratory tests.
1
TABLE OF CONTENTS
1. INTRODUCTION 2
2. TIME-DEPENDENT DEFORMATION OF CONCRETE 4
2.1 Background 4
2.2 Instantaneous Strain 6
2.3 Concrete Tensile Strength 7
2.4 Creep Strain 7
2.5 The Creep Coefficient 8
2.6 Shrinkage Strain 9
3. ANALYTICAL MODEL FOR CRACKING IN DIRECT TENSION 10
MEMBER
3.1 Introduction 10
3.2 First Cracking in a Fully-Restrained Direct Tension Member 11
3.3 Determination of Final Crack Spacing and Crack Width in 14
a Fully-Restrained Direct Tension Member
3.4 Final Crack Spacing and Crack Width in Direct Tension 19
Members Suffering Relative Support Displacements
4. EXPERIMENTAL PROGRAM 20
4.1 Slab Specimens and Test Parameters 22
4.2 Construction of Specimens and Test Procedures 23
4.3 Construction and Testing of Companion Specimens 26
5. TEST RESULTS 28
5.1 Material Properties 29
5.2 Crack Width 30
5.3 Crack Spacing 31
5.4 Steel Stress 32
5.5 Concrete Stress 33
6. DISCUSSION OF RESULTS 34
7. CONCLUDING REMARKS 35
8. ACKNOWLEDGEMENT 35
9. REFERENCES 36
APPENDIX 1 37
APPENDIX 2 57
APPENDIX 3 64
2
1 INTRODUCTION
distribution of the reinforcing steel crossing the crack. It also depends on the
bond characteristics between the concrete and the reinforcement bars at and in
the vicinity of the crack. A local breakdown in bond at and immediately adjacent
to a crack complicates the modelling, as does the time-dependent change in the
bond characteristics caused by drying shrinkage.
A great deal of research has been undertaken throughout the world on cracking
and crack control, but to date the time-dependent mechanisms and interactions
are incompletely understood and a reliable and universally accepted design
procedure has not been developed.
Excessively wide cracks in floor systems and walls may often be avoided by the
inclusion of strategically placed contraction (or control) joints, thereby removing
some of the restraint to shrinkage and temperature movements and reducing the
internal tension. When cracking does occur, to ensure that crack widths remain
acceptably small, adequate quantities of well-anchored reinforcement must be
included at every location in the structure where significant tension is expected.
The maximum crack width, which may be considered to be acceptable in a given
situation, depends on the type of structure, the location within the structure, the
environment, and the consequences of excessive cracking. In external
environments, some building codes and specifications recommend that crack
widths do not exceed about 0.3 mm. For members with one or more exposed
surfaces, a maximum crack width of 0.3 mm should also provide visual
acceptability. In corrosive and aggressive environments, the maximum crack
width may need to be considerably smaller to ensure durability. For the sheltered
interior of most buildings, where the concrete is not exposed and aesthetic
requirements are of secondary importance, a larger crack width may be
acceptable (say 0.3 to 0.5).
Most existing techniques for predicting crack widths involve empirical models that
have been calibrated from results obtained in laboratory tests and, as such, often
fail to accurately predict crack widths in actual structures. It is difficult, therefore,
for a structural designer to be confident that the cracks in a structure will in fact
satisfy the maximum crack width requirements of the local building code.
In beams and slabs that have been proportioned to avoid excessive deflection at
service loads and that contain sufficient quantities of reinforcement to provide
adequate ultimate strength and ductility, flexural crack widths are rarely a
problem under normal in-service conditions, provided of course that the
reinforcement bars are not spaced too widely apart. Sensible reinforcement
detailing is the key to flexural crack control. It is not usually critical, therefore, if
flexural crack widths are not specifically checked in design. The introduction of
high strength reinforcement could change this, however, if the typical in-service
steel stress levels at flexural cracks are increased.
When flexural members are also restrained at the supports, shrinkage causes a
build-up of tension in the member, in addition to the bending caused by the
4
external loads. Crack control is still not usually a problem, since shrinkage is
accommodated by small increases in the widths of the numerous flexural cracks.
However, for members not subjected to significant bending in which restraint is
provided to the longitudinal movement caused by shrinkage and temperature
changes, cracks tend to propagate over the full depth of the section. Excessively
wide cracks are not uncommon. Such cracks are commonly called direct tension
cracks, since they are caused by direct tension rather than by flexural tension. In
restrained direct tension members, relatively large amounts of reinforcement are
required to control load-independent cracking.
In this report, the results of an investigation of shrinkage induced cracking in
restrained reinforced concrete members are presented and discussed. The
investigation consisted of both experimental and analytical components.
Eight restrained reinforced concrete slab specimens were monitored for periods
in excess of 120 days to assess the effects of drying shrinkage on the
development of direct tension cracking. The ends of each specimen were fixed in
position, so that longitudinal deformation of the specimen was effectively
prevented. The only load applied to each slab was the longitudinal restraining
force that gradually developed with time due to drying shrinkage. Strains in both
the reinforcement and the concrete were monitored throughout the tests and the
age of the concrete when each crack developed, the crack locations and the
gradual change in crack widths with time were also recorded. The effects of
varying the quantity of reinforcing steel, the bar diameter and the bar spacing
were studied in order to gain a clearer understanding of the mechanisms of direct
tension cracking caused by restrained shrinkage and the factors affecting it.
An analytical model developed previously to study the problem (Gilbert, 1992) is
also described and experimental results and analytical predictions are compared.
2.1 Background
independent and additive. The total concrete strain is thus considered to consist
of a shrinkage component, which is time-dependent but stress-independent, an
instantaneous component, which occurs immediately upon application of stress
but does not subsequently change (unless the stress level changes), and a creep
component, which is both time-dependent and stress-dependent.
where Ec(t) is the elastic modulus of concrete. The value of Ec(t) increases with
time, and is also dependent on the rate of loading. The faster the load is applied,
the larger is the value of Ec(t). However, in practical situation, this variation is
small and is usually ignored. The variation of the elastic modulus with time for
normal strength concrete may be obtained using the well-known expression
developed by Pauw and found in many building codes:
where ρ is the density of concrete (about 2400 kg/m3 for normal weight concrete)
and fc(t) is the mean compressive strength in MPa at the time under
consideration. The value of Ec(t) given by Eqn (3) are applicable for stress levels
up to about 0.4 fc(t) and for stresses applied over a relatively short period of time
(up to about 5 minutes). For stresses applied over longer periods (say up to one
day’s duration), significant increases in deformation occur due to the rapid early
development of creep. Yet in the broad sense, loads of one day’s duration are
usually considered to be short-term and the effects of creep are often ignored.
This may lead to significant error, if short-term deformation is required after say
one day’s loading. In such a case, it is suggested that the elastic modulus given
by Eqn (3) is multiplied by 0.8 (Gilbert, 1988).
Age of Concrete
1 5 14 28 40 100 365 103 104
(days)
fc(t)/ fc(28) 0.20 0.67 0.9 1.0 1.06 1.16 1.24 1.26 1.27
7
When stress is constant, the instantaneous strain obtained from Eqn. (2)
gradually decreases with time as the elastic modulus increases. However, in
order to simplify structural analysis, it is usual to assume that the instantaneous
strain remains constant with time, and equal to its initial value (calculated using
the value of Ec corresponding to the age at first loading). Furthermore, when
concrete is subjected to a gradually varying stress history, the variation of elastic
modulus with time is often ignored altogether and a constant elastic modulus Ec
is assumed.
When the tensile stress in the concrete reaches the tensile strength of the
concrete ft(t) at a particular point, cracking occurs. The strength of concrete in
tension is of the order of 10% of the compressive strength, with the ratio |ft(t)/fc(t)|
decreasing as the compressive strength increases. AS3600 expresses the
tensile strength in the form
Type of Tensile k1
Strength Lower Mean Upper
Characteristic Characteristic
Direct Tension 0.4 0.5 0.7
Flexural Tension 0.6 0.8 1.1
In Figure 1, the gradual development of creep strain with time was illustrated. In
the period immediately after first loading, the increase in creep strain is rapid, but
after several months, the rate of increase has slowed down dramatically. After
several years under load, the rate of increase in creep strain is very small indeed.
8
The magnitude of creep and its rate of development are influenced by many
factors. Some are properties of the concrete mix, whilst others depend on the
environmental and loading conditions. An increase in concrete strength is
associated with a decrease in creep (ie. higher strength, better quality concrete
creeps less). Creep is also reduced by an increase in the maximum aggregate
size and by the use of a stiffer aggregate type. Creep also depends on the water-
cement ratio and the cement type, insofar as these influence concrete strength.
The magnitude of creep depends on the age of the concrete at the instant of
loading or, more precisely, on the degree of hydration at first loading. Creep
decreases as the age at first loading increases. Concrete is therefore considered
as a time-hardening material, although even in very old concrete the tendency to
creep never entirely disappears. This time-hardening, or aging, complicates the
prediction of creep strain under a time-varying stress history.
Creep is also dependent on the factors, which affect drying. Creep increases as
the environmental humidity decreases, and creep is greater in smaller members
than in larger ones. Creep at the surface of a member generally occurs in a
drying environment and is therefore greater than in regions remote from a
surface. Therefore, thin members with a large surface area to volume ratio, such
as floor slabs, creep more than more massive members with a smaller surface
area to volume ratio. A rise in temperature also increases creep.
Since both the creep and the instantaneous strain components are proportional
to stress, the creep coefficient ϕ(t,τ) is a pure time function and is independent of
the applied stress. As time approaches infinity, the creep coefficient is assumed
to approach a final value ϕ*(τ), where
The creep coefficient for a particular concrete mix can be readily obtained by
testing or estimated using one of the many predictive methods available in the
literature, such as the method contained in Clause 6.1.8 of AS3600-2001. Under
a constant sustained stress σo, the creep strain can be obtained from Eqn 5 as
Shrinkage is affected by all the factors that affect the drying of concrete, in
particular the water content and the water-cement ratio of the mix, the size and
shape of the member and the ambient relative humidity. All else being equal,
shrinkage increases when the water-cement ratio increases, the relative humidity
decreases and the ratio of exposed surface area to volume increases.
Temperature rises accelerate drying and therefore increase shrinkage.
The procedures currently available for estimating shrinkage strains (εsh) are
empirical formulae for the mean shrinkage on a section. Procedures for making
reliable estimates of the variation of shrinkage strain through the depth of a
section are at present unavailable.
10
3.1 Introduction
A rational approach is presented for the determination of the spacing and width
of cracks in a reinforced concrete member that is fully restrained and subjected
only to the axial restraining force caused by shrinkage. The approach is based
only on the principles of mechanics and was developed and presented without
experimental verification by Gilbert (1992) for direct tension members held at
each end by immovable supports. The theoretical results are shown in
subsequent sections of this report to agree well with experimental results
obtained in this study. These experimental results, together with a description of
the test program and test set-up, are also presented subsequently.
As the concrete shrinks, an axial tensile restraining force N(t) develops with time.
When the concrete stress caused by N(t) first reaches the tensile strength of the
concrete, ft(t), at a particular section, cracking occurs. After the first crack occurs,
the magnitude of N(t) and the crack width depend primarily on the amount of
bonded reinforcement crossing the crack.
Consider the fully restrained member shown in Figure 2a. As the concrete
shrinks, the restraining force N(t) gradually increases until the first crack occurs
when N(t) = Ac ft(t). Immediately after first cracking, the restraining force reduces
to Ncr, and the concrete stress away from the crack is less than the tensile
strength of the concrete ft(t). The concrete on either side of the crack shortens
elastically and the crack opens to a width w, as shown in Figure 2b. At the crack,
the steel carries the entire force Ncr and the stress in the concrete is obviously
zero. In the region adjacent to the crack, the concrete and steel stresses vary
considerably, and there exists a region of high bond stress. Immediately adjacent
to the crack a region of partial bond breakdown exists. At some distance so on
each side of the crack, the concrete and steel stresses are no longer influenced
directly by the presence of the crack, as shown in Figure 2c and 2d.
In Region 1, where the distance x from the crack is greater than or equal to so,
the concrete and steel stresses are σc1 and σs1, respectively. Since the steel
12
stress (and hence strain) at the crack is tensile as shown and the overall
elongation of the steel is zero (full restraint), σs1 must be compressive.
Equilibrium requires that the sum of the forces carried by the concrete and the
steel on any cross-section is equal to the restraining force. Therefore with the
force in the steel in Region 1 being compressive, the force carried by the
concrete (σc1 Ac) must be tensile and somewhat greater than the restraining force
(Ncr). In Region 2, where the distance x from the nearest crack is less than so, the
concrete stress varies from zero at the crack to σc1 at x = so. The steel stress
varies from σs2 (tensile) at the crack to σs1 (compressive) at x = so, as shown.
To determine the crack width w and the concrete and steel stresses in Figure 2,
the distance so, over which the concrete and steel stresses vary, needs to be
known and the restraining force Ncr needs to be calculated. Gilbert (1992)
adopted the following simple formula for so, which was earlier proposed by Favre
et al. (1983) for a member containing deformed bars or welded wire mesh:
db
so = (8)
10 ρ
where db is the bar diameter, and ρ is the reinforcement ratio (As/Ac ). Base and
Murray (1982) used a similar expression. However, the results of the analysis are
sensitive to variations in so and the applicability of this simple expression is
examined in Section ??? by reference to results obtained in a laboratory based
experimental program.
σ s1 σ s 2 − σ s1 ⎛ 2 ⎞
L+ ⎜ so + w ⎟ = 0 (9)
Es Es ⎝3 ⎠
Realising that w is very much less than so, Eqn 9 may be rearranged to give
− 2s o
σ s1 = σ s2 (10)
3L − 2 s o
At the crack, the restraining force Ncr is carried entirely by the steel. That is
N cr
σ s2 = (11)
As
By substituting Eqn 11 into Eqn 10, the steel stress away from the crack is
expressed in terms of the unknown restraining force
13
2 s o N cr N
σ s1 = = −C1 cr (12)
3L − 2 s o As As
2so
where C1 = (13)
3L − 2 s o
Prior to cracking, the total concrete strain at any point is zero, since shortening is
prevented. Although the total concrete strain is zero, the individual strain
components of creep, shrinkage, and elastic strain are not. The creep and elastic
strains are tensile (positive) and the shrinkage strain is compressive (negative).
The sum of the time-dependent creep and shrinkage strain components must be
equal and opposite to the elastic strain component. Immediately before the first
crack occurs when the concrete stress just reaches ft , the sum of the creep and
shrinkage strain components is therefore
ft
ε c + ε sh = − (14)
Ec
where Ec is the elastic modulus of the concrete at the time of first cracking.
Immediately after first cracking, the magnitude of the elastic component of strain
in the uncracked concrete decreases (as the concrete stress decreases), but the
creep and shrinkage strain components are unaltered. Creep and shrinkage
strains only change gradually with time.
In Region 1, at any distance greater than so from the crack, equilibrium requires
that the sum of the forces in the concrete and the steel immediately after first
cracking is equal to Ncr . That is
σ c1 Ac + σ s1 As = N cr (15)
N cr − σ s1 As N cr (1 + C1 )
σ c1 = = (16)
Ac Ac
The compatibility requirement is that the concrete and steel strains in Region 1
are identical, ie.
ε s1 = ε 1 (17)
With the concrete strain ε1 equal to the sum of the elastic, creep, and shrinkage
strain components, Eqn 17 can be re-expressed as
σ s1 σ c1
= + ε c + ε sh (18)
Es Ec
14
Substituting Eq. (12), (14), and (16) into Eq. (18) and solving for Ncr gives
n ρ f t Ac
N cr = (19)
C1 + n ρ (1 + C1 )
where ρ=As/Ac and n = Es /Ec. When Ncr is calculated from Eqn 19, the concrete
and steel stresses immediately after cracking may be obtained from Eq. (11),
(12), and (16).
With the stresses and deformations determined immediately after first cracking,
the subsequent long-term behaviour as the member continues to shrink must
next be determined. Consider again the fully-restrained member shown in Figure
2a and 2b. After first cracking, the concrete is no longer fully restrained, since the
crack width can increase with time as shrinkage continues. A state of partial
restraint therefore exists after first cracking. Subsequent shrinkage will cause
further gradual increases in the restraining force, and hence in the concrete
stress away from the crack, and a second crack may develop. Additional cracks
may occur as the shrinkage strain continues to increase with time. However, as
each new crack forms, the member becomes less stiff and the amount of
shrinkage required to produce each new crack increases. The process continues
until the crack pattern is established, usually in the first few months after the
commencement of drying.
The number of cracks and the final average crack width depend on the length of
the member, the quantity and distribution of reinforcement, the quality of bond
between the concrete and steel, the amount of shrinkage, and the concrete
strength. A typical final shrinkage crack pattern is shown in Figure 3. Let the
number of cracks be m and the final shrinkage-induced restraining force be N(∞).
given by Eqn 8 and previously used in the analysis at first cracking. This
assumption may introduce significant error. Results from the experimental
program presented in the next section indicate that shrinkage will cause a
deterioration in bond at the concrete-steel interface and a gradual increase in so
with time. It is suggested here that for final or long-term calculations the value of
so given by Equation 8 should be multiplied by 1.33.
Figure 4 Final concrete and steel stresses after direct tension cracking
In Region 1, where the distance x from the nearest crack is greater than or equal
to so, the final concrete and steel stresses are σ*c1 and σ*s1, respectively. For the
member containing m cracks and provided the reinforcing steel has not yielded,
the following expressions (similar to Eqns 9 and 10) are obtained by equating the
overall elongation of the steel to zero:
− 2 so m
σ s*1 = σ s*2 (21)
3L − 2s o m
Dividing both the numerator and the denominator on the right-hand side by m
and letting the crack spacing s = L/m, Eqn 21 becomes
− 2 so
σ s*1 = σ s*2 = − C 2σ s*2 (22)
3s − 2 s o
2s o
where C2 = (23)
3s − 2s o
At each crack
N (∞)
σ s2∗ = (24)
As
In Region 1, away from each crack, a typical concrete stress history is shown
diagrammatically in Figure 5. The concrete tensile stress increases gradually with
time and approaches the direct tensile strength of the concrete ft.
When cracking occurs elsewhere in the member, the tensile stress in the
uncracked regions drops suddenly as shown. Although the concrete stress
history is continuously changing, the average concrete stress at any time after
the commencement of drying σav is somewhere between σc1 and ft, as shown in
Figure 5, and the final creep strain in Region 1 may be approximated by
σ av
ε c* = ϕ* (25)
Ec
17
where ϕ * is the final creep coefficient (defined as the ratio of the final creep strain
to elastic strain under the average sustained stress σav ). Gilbert (1992) assumed
that
σ + ft
σ av = c1 (26)
2
The final concrete strain in Region 1 is the sum of the elastic, creep, and
shrinkage components and may be approximated by
σ av σ av
ε 1∗ = ε e + ε c ∗ + ε sh ∗ = + ϕ ∗ + ε sh ∗ (27)
Ec Ec
In Region 1, at any distance from a crack greater than so, equilibrium requires
that the sum of the force in the concrete and the force in the steel is equal to
N(∞). That is
σ c1* Ac + σ s1* As = N (∞)
N (∞) − σ s1 As
*
or σ c1 *
= (30)
Ac
The compatibility requirement is that the concrete and steel strains are identical
( ε s1 = ε 1 ) and using Eqn 28, this becomes
* *
σ s1* σ av
= + ε sh
*
*
(31)
Es Ee
n * As ∗
N (∞) = − (σ av + ε sh E e )
*
(32)
C2
where n* = Es/Ee*.
The average crack spacing s (=L/m) must be first determined to calculate C2 (as
defined in Eqn. 23). Knowing that σ*c1 must be less than the tensile strength of
concrete ft and making use of Eqns. 22 and 24, Eqn. 30 becomes
N (∞)(1 + C 2 )
σ c1* = ≤ ft (33)
Ac
2s o (1 + ξ )
s ≤ (34)
3ξ
where
− n * ρ (σ av + ε sh E e )
* *
ξ= (35)
n * ρ (σ av + ε sh E e ) + f t
* *
The number of cracks m (=L/s) may be taken as the smallest integer that causes
Eqn. 34 to be satisfied. With m thus determined, the restraining force N(∞) can be
calculated using Eqn. 32 and the steel and concrete stresses in the various
regions of the member may be determined from Eqn. 21, 24, and 30.
The overall shortening of the concrete is an estimate of the sum of the crack
widths. The final concrete strain at any point in Region 1 of Figure 4 is given by
Eqn 28, and in Region 2, the final concrete strain is
fnσ c*1
ε 2* = *
+ ε sh
*
(36)
Ee
where fn varies between zero at a crack and unity at so from a crack. If a
parabolic variation of stress is assumed in Region 2, the following expression for
the average crack with w is obtained by integrating the concrete strain over the
length of the member
⎡σ c1* 2 * ⎤
w = − ⎢ * ( s − s o ) + ε sh s ⎥ (37)
⎣⎢ E e 3 ⎦⎥
The preceding analysis may be used to determine the number and width of
shrinkage cracks, provided the assumption of linear-elastic behaviour in the steel
is valid. However, if the area of steel As is small, yielding may occur at each
crack and the values of N(∞), calculated from Eqn. 32, will not be correct. In such
a case, σ*s2 is equal to the yield stress of the reinforcement fsy and N(∞) is equal
19
to fsy As . From Eqn. 30, the stress in the concrete away from the crack in region
1 is now
f sy As − σ s1 As
*
σ c1 =
*
(38)
Ac
After the steel at the first crack yields, the tensile concrete stress σc1 increases
only slightly as the compressive steel stress σs1 increases with time and the first
crack opens. Since the restraining force is constant at all times after yielding of
the steel at the first crack (and equal to fsy As ), the tensile stress in the concrete
never again approaches the tensile strength of concrete, and subsequent
cracking does not occur. The width of the initial crack is usually unacceptably
large as the steel at the crack deforms plastically. The crack width w may be
found by ensuring that the overall elongation of the steel is zero. That is
σ s1* f sy − σ s*1 2
( L − w) + ( so ) + w = 0 (39)
Es Es 3
and since w is very much smaller than L , Eqn. 39 can be arranged to give
σ s1* (3L − 2s o ) + 2 so f sy
w=− (40)
3E s
Since the tensile stress in the uncracked concrete does not change significantly
with time, it is reasonable to assume that the average concrete stress σav is given
by Eqn. 38 and the final steel stress in Region 1 may be obtained by substituting
Eqn. 38 into Eqn. 28 and simplifying:
σ s1* f sy As − σ s1 As
*
= + ε sh
*
*
Es Ac Ee
n * ρ f sy + ε sh E s
*
∴ σ s1 =
*
(41)
1 + n* ρ
3.4 Final Crack Spacing and Crack Width in Direct Tension Members
Suffering Relative Support Displacements
It should be remembered that for long term calculations, the value of so may be
taken as
1.33d b
so = (44)
10 ρ
With these modifications and utilising Eqns 24 through 31, the final restraining
force may be expressed as
3 As E s Δu (3L − 2 s o m) n * As ∗
N (∞) = − (σ av + ε sh E e )
*
(45)
2s o m 2s o m
With the restraining force N(∞) calculated using Eqn 45 and the steel stress in
Region 1 obtained from Eqn 43, the final concrete stress in Region 1 is obtained
by enforcing equilibrium and is given by
N (∞) − σ s*1 As
σ c1 *
= (46)
Ac
The number of cracks m is the lowest integer value of m that satisfies Eqn 47.
σ c1* ≤ f t (47)
The direct tensile strength ft in Eqn 47 should be taken as the 28 day value.
Provided the steel at each crack has not yielded, the final average crack spacing
s = L/m and the average crack width may be calculated using Eqn 37.
4. EXPERIMENTAL PROGRAM
A total of eight fully restrained slab specimens with four different reinforcement
layouts were monitored for up to 150 days to measure the effects of drying
shrinkage on the time-dependent development of direct tension cracking due to
restrained deformation. A photograph of the test specimens anchored to the
laboratory strong floor is shown in Figure 6. The steel strains in the vicinity of the
first crack, the steel strains along the full length of the reinforcement, the
21
concrete surface strains, crack widths and crack spacing were recorded
throughout the period of testing.
The compressive and tensile strengths of the concrete at various times were
measures in tests on companion specimens (in the form of concrete cylinders
and prisms), together with the elastic modulus and the creep coefficient. In
addition, unrestrained and unreinforced companion specimens with the same
cross-section as the test specimens were used to measure the magnitude and
rate of development of free shrinkage in the concrete.
(c) To accurately measure the material properties, including the creep and
shrinkage characteristics of the concrete used in the slab specimens.
Details of the cross-section of slabs are provided in Figure 7, and details of the
parameters varied in the tests are given in Table 1. Each specimen consisted of
a 2000 mm long prismatic portion (with the section dimensions shown in Figure
7) which was effectively anchored at its ends by casting each end monolithically
with a 1m by 1m by 0.6m concrete block. These blocks were rigidly clamped to
the reaction floor using 36 mm diameter high strength alloy steel rods tensioned
sufficiently to ensure that the ends of the specimen were effectively held in
position via friction between the end block and the laboratory floor (see Figure 8).
600
As 100 (nominal)
cb = 46 mm
s cs
Note : Slabs S1a and S1b were identical (so too were S2a & S2b, S3a & S3b,and S4a & S4b)
except for small variations in the measured slab thicknesses arising during construction.
23
A plan view of a typical slab is shown in Figure 9, together with the typical
reinforcement layout. The 600 mm wide by 100 mm deep slab specimens were
gradually splayed at each end, as shown, to ensure cracking occurred within the
specimen length and not at the restrained ends. At the mid-length of each
specimen the section was locally reduced in width by 150 mm using thin 75 mm
wide plates attached to the side forms to ensure that first cracking always
occurred at this location (see Figure 11). The bottom surface of each specimen
was supported on smooth supports to ensure negligible bending in the specimen
and no longitudinal restraint, except at the ends of the specimen (see Figure 10).
By using the experience gained from two preliminary specimens, the shape of
the specimens was altered in order to ensure direct tension cracking occurred in
the 2m long prismatic specimen and the first crack occurred at mid-length
(Section AA in Figure 9). In order to measure the steel strains in the vicinity of the
first crack, three strain gauges were attached to each main reinforcement bar,
one at the mid-length and the others 50 mm either side. The strain gauges were
connected to a HBM amplifier. In order to measure steel strains along the full
length of the reinforcement, 12 steel pins were welded to one of the interior bars
at 250mm centres, as shown in Figure 11, and a demec gauge, able to measure
in micro strains, was used to manually measure the deformation between the
pins. Two steel pins were welded to each longitudinal bar (one on each side of
the slab mid-length) in order to measure the deformation of each bar over a
gauge length of 250 mm containing the first crack (see Figure 11).
24
2000
A
B C
1000
A
1000 2660 1000
Plan
B C
600 100
Elevation
36mm alloy steel bar
tensioned to stongfloor
600
As 100 (nominal)
To measure the concrete surface strains, targets were glued adjacent to the pins
on the top surface of the slab (see Figure 12). A demec gauge was also used to
measure the concrete surface strains and a microscope with a magnification
factor of 40 was used for measuring the crack widths. The development,
propagation, extent and width of cracking was observed and recorded throughout
the test. Figure 13 shows the demec gauge, the microscope and the HBM
amplifier used in the experiments for strain and deformation measurements.
Prior to casting each specimen, the inside surface of the mould was cleaned and
thinly coated with a concrete release agent to prevent adhesion of the concrete.
The concrete was placed in the mould in equal layers and compacted by internal
vibration. For each layer, the concrete was vibrated until the surface became
smooth in appearance. Sufficient concrete was placed in the top layer to overfill
the mould when compacted, and the surface was then stripped off and finished
with a steel trowel. Companion specimens were also cast at this time in the form
of concrete cylinders and prisms.
The specimens were undisturbed in their moulds for 3 days and kept
continuously moist by a thick and complete covering of wet hessian to minimise
the loss of moisture from the specimens and thereby to delay the onset of drying
shrinkage. After 3 days the specimens were removed from the mould, the demec
strain targets were glued to the concrete surface and initial strain measurements
were recorded. The blocks at each end of the test specimens were then rigidly
clamped to the reaction floor using 36 mm diameter high strength alloy steel rods
tensioned sufficiently to ensure that the ends of the specimen were effectively
held in position by friction.
Companion cylinders, prisms and unreinforced block specimens were cast at the
same time as each slab specimen in accordance with the relevant Australian
Standard. The companion specimens were used to measure the properties of the
concrete throughout the period of the testing. Measured properties included the
27
compressive strength, the elastic modulus, the flexural and direct tensile
strengths, and the creep coefficient. The time-dependent development of drying
shrinkage strain was also measured on unrestrained specimens of similar
dimension to the slab specimens.
The companion specimens were of standard size. The cylinders were either
150mm or 100 mm in diameter and 300 mm or 200 mm in length, respectively.
The prisms had dimensions of 100mm × 100mm × 500mm and the dimensions of
the shrinkage blocks were 600mm × 600mm × 100mm. The companion
specimens were exposed to the same environmental, curing and drying
conditions as the slab specimens. Testing of companion specimens was carried
out in accordance with the relevant parts of the Australian Standard AS1012.
Cylinder and prism strength and elastic modulus tests were carried out at ages 3
days, 7 days, 14 days, 21 days, and 28 days.
Creep tests for each batch of concrete were carried out using standard creep
rigs, each containing two or three 150 mm diameter cylinders loaded with a
constant sustained stress of 5 MPa first applied at age 3 days. Other unloaded
companion cylinders, with the same dimensions as the creep cylinders, were
placed beside the creep rig and used to measure drying shrinkage strains. The
creep strain was determined by subtracting the measured shrinkage strain and
the instantaneous strain from the total strain measured on the creep cylinders.
The creep coefficient at any time was determined as the measured creep strain
at that time divided by the measured instantaneous strain.
5. TEST RESULTS
The experimental results taken from the 8 slab specimens are presented in this
section (and in Appendix I and III). All slabs were anchored to the laboratory floor
at age 3 days and drying also commenced at age 3 days. First cracking, due to
direct longitudinal tension caused by restraint to shrinkage, occurred within the
first week and subsequent cracks developed over the next 60 days. Crack widths
gradually increased with time, with the change being relatively small after about
90 days. All cracks observed during the tests penetrated through the full depth of
the specimen’s cross-section. Steel strains in the vicinity of the first crack and
along the full length of the reinforcement, concrete surface strains, crack widths
and crack spacings were recorded.
After 150 days, the finest final crack width (w = 0.18 mm) and the lowest final
steel stress (σ*s2 =190 MPa) measured during the experimental program were in
slab S1b (the slab containing the most longitudinal reinforcement - 3 N 12 bars).
The measured widest final crack (w = 0.84 mm) and highest final steel stress
(σ*s2 =532 MPa ) were in slab S3a (the slab containing the least longitudinal
reinforcement - 2 N 10). Slab thicknesses were measured across the slab width
in the region of first cracking to determine the average value and to account for
any small variations between specimens arising during construction. The
average slab thickness for each specimen was given in Table 1. The material
properties, steel reinforcement layouts, cross-sectional dimensions and details of
the final crack pattern are presented for each slab in Appendix I.
By using the mathematical model presented in Section 3, the final average crack
width (w) and crack spacing (s), the final steel stress at the crack (σ*s2), and the
final steel and concrete stresses away from the crack (σ*s1 and σ*c1) were
calculated for each slab, together with the values immediately after first cracking.
These quantities have also been determined from the test data and comparisons
are made between theoretical and experimental results for each slab in Appendix
I. Full calculations for slab S1a and slab 2b are presented in Appendix II. The
measured final elongations Δu (see Table 2) were used in the calculations. Raw
data measured throughout the tests are presented in Appendix III.
29
Two different concrete batches were used throughout the test - concrete batch I
(for slabs S1a, S1b, S2a, S3a, S3b, S4a and S4b) and concrete batch II (for slab
S2b). The measured concrete compressive strength, the flexural tensile strength,
the indirect tensile strength and modulus of elasticity at ages 3, 7, 14, 21 and 28
days are presented in Table 3.
Age (days)
Concrete Batch 1 Material Property 3 7 14 21 28
Compressive Strength (MPa) 8.17 13.7 20.7 22.9 24.3
Flexural Tensile Strength(Modulus of Rupture)(MPa) 1.91 3.15 3.43 3.77 3.98
Indirect Tensile Strength (Brazil test) (MPa) 1.55 1.97
Modulus of Elasticity (MPa) 13240 17130 21080 22150 22810
Age (days)
Concrete Batch II Material Property 3 7 14 21 28
Compressive Strength (MPa) 10.7 17.4 25.0 27.5 28.4
Flexural Tensile Strength(Modulus of Rupture)(MPa) 2.47 3.10 3.77 3.97 4.04
Indirect Tensile Strength (Brazil test) (MPa) 1.60 2.1
Modulus of Elasticity (MPa) 16130 18940 21750 22840 23210
The measured creep coefficient and shrinkage strain for concrete batch I and II
are presented in Table 4.
Table 4 Creep coefficient and shrinkage strain for concrete batches I and II.
Batch 1:
Age(days) 7 14 21 28 36 43 53 77 100 122
ϕcc 0.38 0.60 0.68 0.69 0.73 0.84 0.86 0.93 0.97 0.98
εsh (×10-6) 66 115 154 208 244 313 327 342 421 457
Batch II:
Age(days) 7 13 21 27 33 51 62 79 127 150
ϕcc 0.40 0.60 0.82 0.87 0.85 1.04 1.00 1.15 1.07 1.16
εsh (×10-6) 72 183 277 258 331 381.2 456.7 463 469.4 495
30
For each crack on each specimen, the crack width is taken as the average of the
measured crack width at five locations across the slab. The average crack width
for each slab is the sum of all the crack widths divided by the number of cracks.
For all slabs, first cracking occurred within the first week, except for slab S1b,
where first cracking occurred within 10 days. The development, extent and width
of cracks were observed and measured carefully using a microscope. The
magnitude of crack width depends primarily on the amount of bonded
reinforcement crossing the crack. In addition, the width of a crack in a restrained
member depends on the degree of restraint, the quality of bond between the
concrete and the steel, the size and distribution of the individual reinforcement
bars and the concrete quality. Table 5 provides a comparison between the
calculated final crack width (obtained using the analytical model described
earlier) and average final crack width measured in the laboratory. Final average
crack width versus steel area is presented in Figure 14.
0.8
0.6 Theoretical
0.4 Experimental
0.2
0
0 100 200 300 400
0.6
0.5
0.4
(mm)
Theoretical
0.3
Experimental
0.2
0.1
0
0 100 200 300 400
2
Steel area (mm )
Table 5 Theoretical and experimental values for final average crack width.
The distance between cracks was measured every 50 mm across the slab width
and averaged to obtain the average crack spacing for each slab (s in Figures 3
and 4). In Table 6, the observed average crack spacing is compared with the
predicted average crack spacing. In general, an increase in the steel area causes
a reduction in the average crack spacing. In Figure 15, the crack spacing versus
steel area is plotted. An increase in the steel area causes a reduction in the
average crack spacing.
1200
1000
800
600 Experimental
400
200
0
0 50 100 150 200 250 300 350 400
2
Steel Area (mm )
At each crack, the steel carries the full restraining force (N) and the stress in the
concrete is zero. Using the measured final steel strains in the vicinity of the first
crack, the maximum stress in each steel bar crossing the crack was determined
(σ*s2 in Figure 4). The restraining force was thus determined from the
experiments (N = σ*s2 As). Comparisons between theoretical and experimental
results for the final steel stress at each crack (σ*s2) are presented in Table 7 and
Figure 16 shows plots of the theoretical and experimental vales of σ*s2 versus
steel area.
600
500
400
Theoretical
300
Experimental
200
100
0
Series ‘a’
0 50 100 150 200 250 300 350 400
2
Steel Area (mm )
600
Steel Stress at the Crack
500
400
(MPa)
Theoretical
300
Experimental
200
100
0
Series ‘b’
0 100 200 300 400
Steel Area (mm2)
In the region immediately adjacent to the crack, the concrete and steel stresses
vary considerably with distance from the crack (within Region 2 in Figure 4). At
any distance greater than so from a crack (ie. within Region 1 in Figure 4) at any
time instant, the concrete and steel stresses (σc1 and σs1, respectively) remain
constant until the next crack is approached (Region1 in Figure 4). From the
measured steel strains in Region 1 and obtaining the average value for these
regions along the steel bar, the steel stress σs1 was calculated and, by enforcing
equilibrium (N = σc1Ac + σs1As), the concrete stress away from the crack σc1 was
determined. Comparisons between theoretical and experimental results for the
final concrete stress (σ*c1) are made in Table 8. Concrete stress, σ*c1, versus
steel area is plotted in Figure 17.
2
1.8
Concrete Stress (MPa)
1.6
1.4
1.2
Theoretical
1
0.8 Experimental
0.6
0.4 Series ‘a’
0.2
0
0 100 200 300 400
1.8
Concrete Stress (MPa)
1.6
1.4
1.2
1 Theoretical
0.8 Experimental
0.6
0.4
0.2
Series ‘b’
0
0 100 200 300 400
2
Steel Area (mm )
6. DISCUSSION OF RESULTS
With an increase in the steel area, the loss of stiffness at first cracking
reduces and, hence the restraining force after cracking is greater, but the steel
stress at each crack decreases (see Table 7 and Figure 16). With a larger
restraining force, the concrete stress in regions remote from a crack tends to be
higher and hence further cracking is more likely.
35
Comparing the behavior after 150 days of drying for slabs S3b (As = 157
mm2, ρ = 0.00262) and S1a (As = 339 mm2, ρ = 0.00565), for example, the final
measured average crack widths are 0.50 mm and 0.21 mm, respectively. The
measured average crack spacings are 997 mm and 670 mm, respectively; the
final measured steel stress at the first crack are 467 MPa and 273 MPa,
respectively; and the concrete stress away from each crack is 1.31 MPa and 1.77
MPa, respectively.
7. CONCLUDING REMARKS
A simple and rational analytical procedure for the determination of the stresses
and deformation after shrinkage cracking in a restrained direct tension member is
presented and a series of long-term tests on eight restrained reinforced concrete
slab specimens has been described. The experimental program also involved
testing numerous companion specimens to determine the material properties,
creep and shrinkage characteristics of the concrete used in the slabs, including
the compressive strength, the flexural tensile strength, the elastic modulus, the
creep coefficient and the shrinkage strain.
The measured width of shrinkage cracks, steel stresses and concrete stresses,
agree well with the results of the analytical model.
8. ACKNOWLEDGEMENT
The funding for this research was provided by the Australian Research Council
(ARC).
36
9. REFERENCES
3. Base, G.D. & Murray, M.H. 1982. “New Look at Shrinkage Cracking”, Civil
Engineering Transactions, Institution of Engineers Australia, V. CE24 No. 2
171 pp.
APPENDIX I
A
2000
1000
A
1000 2660 1000
600
100
75 40 185 185 40 75
SECTION A-A
Centre Line
441 mm
1057 946 mm
Final average crack width = 0.21 mm
Strain (microstrain)
1500
1250 2
1000
3
750
500 4
250 5
0
6
-250
-500 7
0 10 20 30 40 50 60 70 80 90 100 110 120
8
Age (days)
1500
14
1250
1000 15
750 16
500 17
250 18
0 19
-250 20
-500
21
0 10 20 30 40 50 60 70 80 90 100 110 120
22
Age (days) 23
24
11
Strain (microstrain)
2000
1750 12
1500 13
1250
1000 14
750 16
500
250 17
0 18
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Age (dayes)
Figure AI.3 Steel strain and concrete surface strain versus time curves for slab
specimen S1a
41
AI.2 SLAB S1b (Reinforcement layout as for S1a - see Figure AI.1)
Table AI.3 Geometrical details and material properties for slab S1b
Centre Line
w=0.1 mm w=0.21 mm
w=0.13 mm
w=0.1 mm w=0.12 mm
254 361 mm
847 mm 449 mm
Final average crack width=0.18 mm
2
2000
1750 3
Strain (microstrain)
1500
1250 4
1000
750 5
500
250 6
0
-250 7
-500
8
-750
0 10 20 30 40 50 60 70 80 90 100 110 120 130 9
Age (days)
1400 15
1200 16
Strain (microstrain)
1000 17
800
18
600
400 19
200 20
0 21
-200
22
-400
-600 23
0 10 20 30 40 50 60 70 80 90 100 110 120 130 24
Age (days)
1050 1
900
2
750
3
600
450 4
300 5
150 6
0
7
0 10 20 30 40 50 60 70 80 90 100 110 120 130
8
Age (days)
Figure AI.5 Steel strain and concrete surface strain versus time curves for slab
specimen S1b
43
AI.3 SLAB S2a (Reinforcement layout as for S1a - see Figure AI.1)
Table AI.5 Geometrical details and material properties for slab S2a
Centre Line
w=0.22 mm
w=0.25 mm
w=0.43 mm
957 mm 391 mm
Final average crack width=0.3 mm
1
Steel Strain by Demec (slab # 3 - 3Y10 )
2
1750 3
1500 4
Strain (microstrain)
1250 5
1000 6
750
7
500
250 8
0 9
-250 10
-500 11
0 10 20 30 40 50 60 70 80 90 100 110 120 130 12
Age (days) 13
1
Steel strain by demec (S2a - 3Y10 )
2
1750 3
1500 4
Strain (microstrain)
1250 5
1000 6
750
7
500
250 8
0 9
-250 10
-500 11
0 10 20 30 40 50 60 70 80 90 100 110 120 130 12
Age (days) 13
2250
Strain (microstrain)
2000
1750
10
1500
1250 11
1000 13
750
14
500
250 15
0 16
0 10 20 30 40 50 60 70 80 90 10 11 12 13 17
0 0 0 0
Age (days)
Figure AI.7 Steel strain and concrete surface strain versus time curves for slab
specimen S2a
45
AI.4 SLAB S2b (Reinforcement layout as for S1a - see Figure AI.1)
Table AI.7 Geometrical details and material properties for slab S2b
Centre Line
533 mm
1035 mm
Final average crack width=0.31 mm
750
4
500 5
250 6
0 7
-250 8
-500 9
-750
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
10
Age (days) 11
1250 4
1000 5
750
500 6
250
7
0
-250 8
-500
9
-750
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 10
Age (days) 11
3000 2
2500 4
Strain (microstrain)
2000 5
1500
7
1000
8
500
9
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
6
Age (days)
Figure AI.9 Steel strain and concrete surface strain versus time curves for
slab specimen S2b
47
A
2000
1000
A
1000 2660 1000
600
100
75 75 300 75 75
SECTION A-A
Table AI.9 Geometrical details and material properties for slab S3a
Centre Line
w=0.84 mm
3000 3
Strain (microstrain)
4
2400
5
1800
6
1200
7
600
8
0 9
-600 10
0 10 20 30 40 50 60 70 80 90 100 110 120 130 11
Age (days) 12
3000 15
2400 16
1800 17
1200 18
600 19
0 20
-600 21
0 10 20 30 40 50 60 70 80 90 100 110 120 130 22
Age (days)
3000
2700 22
2400 23
2100
1800 24
1500 25
1200
900 26
600
300
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Age (days)
Figure AI.12 Steel strain and concrete surface strain versus time curves for slab
specimen S3a
50
AI.6 SLAB S3b (Reinforcement layout as for S3a - see Figure AI.10)
Table AI.11 Geometrical details and material properties for slab S3b
Centre Line
w=0.78 mm w=0.22 mm
997 mm
Final average crack width=0.5 mm
Figure AI.13 The final crack pattern for slab S3b
Strain (microstrain)
2100
4
1800
1500 5
1200
900 6
600 7
300
0 8
-300 9
-600
0 10 20 30 40 50 60 70 80 90 10 11 12 13 10
0 0 0 0 11
Age (days) 23
3900 12
3600 13
3300
Strain (microstrain)
3000 14
2700
2400 15
2100 16
1800
1500 17
1200
900 18
600
300 19
0 20
-300
-600 21
0 10 20 30 40 50 60 70 80 90 100 110 120 130
22
Age (dys)
3200 21
2800
22
2400
2000 23
1600 24
1200 25
800
400
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Age (days)
Figure AI.14 Steel strain and concrete surface strain versus time curves for slab
specimen S3b
52
A
2000
1000
A
1000 2660 1000
600
100
75 45 120 120 75
SECTION A-A
Table AI.13 Geometrical details and material properties for slab S4a
Centre Line
w=0.18 mm
w=0.28 mm w=0.29 mm
w=0.18 mm
562 mm
1004 mm 1001 mm
Final average crack width=0.23 mm
Strain (microstrain)
4
600 5
400 6
7
200
8
0
9
-200 10
-400 11
0 10 20 30 40 50 60 70 80 90 100 110 120 130 12
Age (days) 13
14
800 18
600 19
400
20
200
0 21
-200 22
-400 23
-600
24
0 10 20 30 40 50 60 70 80 90 10 11 12 13
25
0 0 0 0
Age (days)
1600
Strain (microstrain)
1400
1200 27
1000 28
800 31
600 32
400 33
200
34
0
36
0 10 20 30 40 50 60 70 80 90 10 11 12 13
0 0 0 0 37
Age (days)
Figure AI.17 Steel strain and concrete surface strain versus time curves for slab
specimen S4a
55
AI.8 SLAB S4b (Reinforcement layout as for S4a - see Figure AI.15)
Table AI.15 Geometrical details and material properties for slab S4b
Centre Line
w=0.32 mm w=0.16 mm
w=0.26 mm
995 mm 1016 mm
Final average crack width=0.25 mm
Strain (microstrain)
800 4
600 5
6
400
7
200
8
0 9
-200 10
-400 11
0 10 20 30 40 50 60 70 80 90 100 110 120 130 12
Age (days) 13
14
1000 17
800 18
600 19
400 20
200 21
0 22
-200 23
-400 24
0 10 20 30 40 50 60 70 80 90 100 110 120 130
25
Age (days)
1400 31
1200
32
1000
33
800
600 35
400 36
200 37
0 38
0 10 20 30 40 50 60 70 80 90 100 110 120 130 39
Age (days)
Figure AI.19 Steel strain and concrete surface strain versus time curves for slab
specimen S4b
57
APPENDIX II
The modular ratio is n = Es /Ec(3) = 200000/13240 = 15.1 and the distance so,
over which the concrete and steel stresses vary, is given by Eqn 8:
db
so = = 216 mm
10 ρ
The final effective modulus (Eqn 29) is
E c (3) 13240
E *e = = = 6690 MPa
1 + φ * 1 + 0.98
and the corresponding effective modular ratio is
Es 2 × 10 5
n = * =
*
= 29.9 .
E e 6690
The ratio C1 is given by Eqn 13:
2so 2 × 216
C1 = = = 0.0776
3L − 2 s o 3 × 2000 − 2 × 216
and the restraining force immediately after first cracking is obtained from Eqn 19
(where ft is taken as the 7 day value as cracking generally occurs within the first
week of drying):
The concrete stress away from the crack immediately after first cracking is given
by Eqn 16:
59
N cr (1 + C1 ) 47240(1 + 0.0776)
σ c1 = = = 0.83 MPa
Ac 60980
and the estimate of the average concrete stress in the period after first cracking
is given by Eqn 26:
σ c1 + f t 0.83 + 1.55
σ av = = = 1.19 MPa
2 2
For long-term calculations, the final value for so, over which the concrete and
steel stresses vary, is given by Eqn 44:
1.33d b
so = = 287 mm
10 ρ
By trial and error, it is assumed that the number of cracks m = 3. This is
demonstrated subsequently. From Eqn 45, the final restraining force is
3 × 339 × 200000 × 0.305 (3 × 2000 − 2 × 287 × 3) × 29.9 × 339
N (∞ ) = − (1.19 − 457 × 10 − 6 × 6690)
2 × 287 × 3 2 × 287 × 3
= 82930 N
N (∞ ) 82930
From Eqn 24, the final steel stress at the cracks is σ s*2 = = = 245 MPa
As 339
and the final steel stress in Region 1 is obtained using Eqn 43:
which is less than the final tensile strength taken as the 28 day value (ft(28) =
1.97 MPa). If it is initially assumed that m = 2, then σ c*1 ≥ ft(28) and so the initial
assumption that m = 3 is correct.
⎡ 1.67 2 ⎤
w = −⎢ (667 − × 287) − 457 × 10 − 6 × 667 ⎥ = 0.19 mm.
⎣ 6690 3 ⎦
60
The final steel stress at the crack (σ*s2), the final steel stress away from the crack
(σ*s1) and the final concrete stress away from the crack (σ*c1) are determined
from the experimental data as follows:
The average final steel strain, measured by strain gauges at the first crack, is
ε (ave )s 2 = 1367 με
The average final steel strain, measured by the demec targets away from the
crack, is
and the average final steel stress away from the crack is therefore
The steel compressive force at any section away from the crack is
From equilibrium considerations, the concrete tensile force at any section away
from the crack is
and the final concrete stress away from the crack is therefore
Tc 108910
σ c*1 = = = 1.78 MPa.
Ac 60980
61
The modular ratio is n = Es /Ec(3) = 200000/16130 = 12.4 and the distance so,
over which the concrete and steel stresses vary, is given by Eqn 8:
db
so = = 249 mm
10 ρ
The final effective modulus (Eqn 29) is
E c (3) 16130
E *e = = = 7470 MPa
1 + φ * 1 + 1.16
and the corresponding effective modular ratio is
Es 2 × 10 5
n* = = = 26.8 .
E *e 7470
The ratio C1 is given by Eqn 13:
2so 2 × 249
C1 = = = 0.0905
3L − 2s o 3 × 2000 − 2 × 249
and the restraining force immediately after first cracking is obtained from Eqn 19
(where ft is taken as the 7 day value as cracking generally occurs within the first
week of drying):
The concrete stress away from the crack immediately after first cracking is given
by Eqn 16:
N (1 + C1 ) 32330(1 + 0.0905)
σ c1 = cr = = 0.60 MPa
Ac 58740
and the estimate of the average concrete stress in the period after first cracking
is given by Eqn 26:
σ c1 + f t 0.60 + 1.60
σ av = = = 1.10 MPa
2 2
For long-term calculations, the final value for so, over which the concrete and
steel stresses vary, is given by Eqn 44:
1.33d b
so = = 331 mm
10 ρ
By trial and error, it is assumed that the number of cracks m = 3. This is
demonstrated subsequently. From Eqn 45, the final restraining force is
3 × 236 × 200000 × 0.315 (3 × 2000 − 2 × 331 × 3) × 26.8 × 236
N (∞ ) = − (1.10 − 495 × 10 −6 × 7470)
2 × 331 × 3 2 × 331 × 3
= 55610 N
N (∞ ) 55610
From Eqn 24, the final steel stress at the cracks is σ s*2 = = = 236 MPa
As 236
and the final steel stress in Region 1 is obtained using Eqn 43:
⎡ 1.23 2 ⎤
w = −⎢ (667 − × 331) − 457 × 10 −6 × 667⎥ = 0.26 mm.
⎣ 7470 3 ⎦
63
The final steel stress at the crack (σ*s2), the final steel stress away from the crack
(σ*s1) and the final concrete stress away from the crack (σ*c1) are determined
from the experimental data as follows:
The average final steel strain, measured by strain gauges at the first crack, is
ε (ave )s 2 = 1450 με
The average final steel strain, measured by the demec targets away from the
crack, is
and the average final steel stress away from the crack is therefore
The steel compressive force at any section away from the crack is
From equilibrium considerations, the concrete tensile force at any section away
from the crack is
and the final concrete stress away from the crack is therefore
Tc 86140
σ c*1 = = = 1.47 MPa.
Ac 58740
64
APPENDIX III
EXPERIMENTAL DATA
65
s la b # 1 ( 3 Y 1 2 ) S T E E L S T R A IN B Y S T R A IN G A UG E S
Date Temp. A ge 10 11 12 13 14 15 16 17 18
12/7/01 16 3
12/7/01 16 3 21 19 12 13 21 19 23
13/7/01 15 4 37 36 25 21 44 31 46
14/7/01 15 5 45 41 29 21 49 38 52
15/7/01 14 6 73 61 29 19 75 73 76
16/7/01 15 7 451 433 430 438 431 439 420
17/7/01 15 8 534 521 525 535 537 540 515
18/7/01 14 9 603 584 585 598 601 605 572
20/7/01 15 11 708 685 692 710 714 754 675
22/7/01 14 13 782 755 765 788 789 864 744
23/7/01 14 14 779 755 767 792 789 870 745
25/7/01 16 16 830 805 836 853 851 936 804
27/7/01 16 18 867 835 867 886 886 985 838
30/7/01 15 21 991 960 999 1021 1022 1129 964
1/8/01 15 23 1148 1113 1156 1182 1186 1300 1117
3/8/01 15 25 1173 1127 1174 1201 1206 1316 1134
6/8/01 15 28 1342 1297 1338 1377 1371 1496 1288
8/8/01 16 30 1277 1252 1219 1260 1126 1223 1058
10/8/01 16 32 1291 1274 1238 1279 1117 1236 1066
14/8/01 17 36 1229 1197 1197 1240 1065 1197 1020
17/8/01 16 39 1413 1392 1354 1406 1214 1384 1166
21/8/01 16 43 1385 1320 1381 1456 1340 1528 1280
24/8/01 16 46 1251 1187 1341 1394 1332 1515 1294
28/8/01 15 50 1213 1143 1324 1379 1305 1472 1265
31/8/01 15 53 1159 1113 1312 1355 1267 1442 1229
7/9/01 16 60 1225 1178 1391 1429 1329 1549 1324
13/9/01 18 66 1251 1230 1446 1512 1393 1614 1395
24/9/01 20 77 1191 1183 1476 1510
4/10/01 19 87 1205 1185 1467 1581
17/10/01 21 100 1251 1233 1632 1670
8/11/01 18 122 1172 1171 1491 1635 Wires have damaged
72
s la b # 7 ( 4 Y 1 0 ) S T E E L S T R A IN B Y S T R A IN G A UG ES
Date Temp. A ge 26 27 28 29 30 31 32 33 34 35 36 37
12/7/01 16 3
12/7/01 16 3 8 9 8 108 0 8 29 9
13/7/01 15 4 2 2 4 113 5 4 196 5
14/7/01 15 5 4 2 10 74 6 3 197 6
15/7/01 14 6 329 311 318 147 294 281 485 328
16/7/01 15 7 423 402 408 250 393 382 572 431
17/7/01 15 8 520 499 495 369 507 495 669 553
18/7/01 14 9 569 549 546 431 562 553 711 618
20/7/01 15 11 681 659 645 585 706 937 794 743
22/7/01 14 13 681 716 705 647 866 985 849 803
23/7/01 14 14 741 719 712 616 788 780 849 797
25/7/01 16 16 788 763 744 662 845 821 877 865
27/7/01 16 18 826 794 777 701 861 846 898 905
30/7/01 15 21 948 912 894 820 1018 966 989 1040
1/8/01 15 23 1073 1050 1036 961 1209 1121 1102 1192
3/8/01 15 25 1055 1055 1043 973 1251 1151 1097 1195
6/8/01 15 28 912 918 934 815 1147 1027 949 1046
8/8/01 16 30 948 949 968 889 1151 1047 966 1095
10/8/01 16 32 994 996 1014 940 1162 1103 995 1133
14/8/01 17 36 947 947 966 865 1241 1054 949 1080
17/8/01 16 39 1126 1127 1151 1048 1395 1298 1111 1280
21/8/01 16 43 1158 1160 1171 1101 1491 1321 1144 1334
24/8/01 16 46 943 967 1046 733 1111 984 859 1044
28/8/01 15 50 926 951 1015 717 1113 971 843 1037
31/8/01 15 53 900 928 982 704 1017 897 823 1018
7/9/01 16 60 981 1011 1063 856 1118 1040 906 1127
13/9/01 18 66 1056 1064 1120 945 1264 1100 955 1199
24/9/01 20 77 1095 1067 1073 1017 1271 1148 980 1228
4/10/01 19 87 1166 1136 1168 1156 1325 1282 1042 1354
17/10/01 21 100 1158 1157 1149 1151 1341 1292 1043 1368
8/11/01 18 122 1092 1109 1032 1104 1361 1345 996 1342
90