Reduction of Early-Age Crack Risks in Concrete Walls

by Using a New Casting Technique

Majid Al-Gburi, Dr., College of Engineering, University of Mosul, Iraq; Jan-Erik Jonasson, Professor Dr., Structural Division,
Lulea University, Lulea, Sweden; Martin Nilsson, Dr, Structural Division, Lulea University, Lulea, Sweden.
Contact: majid.al-gburi@ltu.se
DOI: 10.2749/101686616X14555429843960

Abstract the wall together with the slab, where

the wall portion is called a kicker. In
Volumetric changes in early-age concrete that are restrained might lead to general, a small kicker is used (height
cracks. The degree of restraint is influenced by the casting sequence and the of 0.2 to 0.5 m) as a support for
dimensions of the castings. In the current study, a new casting technique is pro- correct placement of the wall form-
posed to reduce the restraint in newly cast concrete by following a new method work. Here, the word kicker refers to
of arranging the structural joint with the existing old concrete. The proposed the lower part of the wall when it is
technique is valid for the typical wall-on-slab structure using one structural cast simultaneously with the slab as a
joint. According to this casting method the lower part of the wall is cast measure to obtain as low restraint as
together with the slab, which is called a kicker. Thereafter, the behaviour of possible in the upper part of the wall
the structure changes from a typical wall-on-slab case to a typical wall-on-wall for the subsequent casting. Thus, the
case. It has been proven based on the beam theory and demonstrated by investigation of potential height of the
numerical calculations that there is a clear reduction in the restraint from the kicker in this study is from almost zero
slab to the wall by using the kickers. In this paper, different kicker heights are up to nearly full height of the wall.
studied with the aim of determining the minimum restraint in the upper part of The kicker height that gives minimum
the wall cast in contact with the kicker. The technique using kickers is com- restraint in the upper part of the wall
pared with common measures used in the field to avoid cracking, such as using is denoted as the optimum kicker
cooling pipes in the new casting and/or heating cables in the adjoining old con- height. The only reference found in
crete. The presented method is both cost and time effective, as it opens up the the literature about kickers is in Ref.
possibility to use larger structural length for each casting sequence. [10] where it is mentioned that a
kicker height of about 1.5 m reduces
Keywords: restraint, early-age concrete, wall-on-slab, casting sequence, kicker.
the restraint in the upper wall. But,
neither any background nor explana-
tion is provided in the same.
Introduction structures. Therefore, it is very impor-
tant to avoid such cracks. This new technique is theoretically
One important issue in newly cast con- applicable to any building site. It can
crete structures is the tensile stress The most common measures imply using longer casting lengths,
leading to cracking when temperature- undertaken on site to reduce the risk which usually is beneficial for the con-
and shrinkage-induced volumetric of cracking are cooling of the newly tractor, and there is no need for addi-
changes are restrained.1–3 Such cracks cast concrete and/or heating of the tional materials or special equipment.
may cause very early corrosion of adjacent structure.6–8 Both these mea-
The objectives of this study are to:
rebars or penetration of harmful sures need consultants, materials, spe-
cial tools and equipment, which may • propose and present a new casting
liquids or gases into the concrete body. technique to reduce the restraint in
These situations could result in signifi- not be available for all building sites.
the newly cast concrete, which will
cantly increased maintenance cost. Reducing the restraint is one of the lower the risk of through cracking
The cracking originates either from most economical methods to reduce • evaluate the technique and deter-
uneven deformations due to tempera- the risk of cracking during the early mine the optimum kicker height
ture and/or moisture gradients inside ages of concrete. Several ways are sug- • investigate the influence of the slab
the young concrete during the temper- gested to reduce the restraint in fresh and wall dimensions that significantly
ature expansion phase, which is the cast members, for instance opting for influence the height of the kicker
source of surface cracking, or from suitable casting sequences, shortening • compare the kicker method with tra-
restraint in adjacent structures during of the section being cast and ditional site measures, and in addition
the contraction phase, which might arrangements of construction joints.9–13 discuss the possibility of combining it
cause through cracking.4,5 Cracking Reducing the length of the casting with the existing methods to obtain
affects the service live and durability member to control the risk of cracking less restraint in the structure.
and increases maintenance costs of delays the construction work and
The research questions for this
increases the cost. It is also possible to
study are:
reduce the risk of cracking by choosing
a concrete mix with low temperature • Does the use of a kicker decrease
Peer-reviewed by international ex-
perts and accepted for publication development due to hydration.14–16 the restraint and thereby the risk of
by SEI Editorial Board through cracking as compared with
The present study is aimed to decrease not using kickers?
Paper received: November 11, 2015 the restraint in walls cast on slabs by • Can the use of kickers be combined
Paper accepted: February 24, 2016 studying the effects of casting a part of with other measures for reducing

216 Scientific Paper Structural Engineering International Nr. 3/2016

 cracking risks, such as cooling of the plane, implying the strains over representative value for the ratio of
the new concrete or heating of the the height of the structures varying lin- the E-modulus between young and
old concrete? early. The plane-section restraint coef- old concrete may be taken as
• Does the use of kicker affect the ficient γ 0R can be separated into one 1/1.07 = 0.93 when calculating the
casting length? translational part γ 0TR and one rota-
maximum stress ratio for through
tional part γ 0φ , which gives: cracking. Under other circumstances
R
this ratio might be different. Aa, Ak
Stress due to Concrete oφ and Ay are the areas of slab, kicker
γ Ro ðxd ,z1 Þ = 1 − γ oT
R ðxÞ −γ R ðxd ,z1 Þ ð1Þ
Sequence Casting
 and young concrete respectively; γRT
1 is the translational boundary restraint
The casting sequence has an effect on R ðxd Þ =
γ oT
 1 −γ RT ðxd Þ
E28old Aa + Ak
1 + ξE28y Ay coefficient; and γRR is the rotational
the degree of restraint.17 Reducing the boundary restraint coefficient for the
restraint is one of the most effective ð2Þ studied case of self-balancing the
ways to decrease the risk of thermal
cracking in early age concrete. When
casting a concrete wall on an older 2   3
slab, volume changes occur in the wall. −Aað Ha Hy + 2Hk Hy + Hy 2 Þ− Ak Hk Hy + Hy 2 +
1 E28old 6
  7
Ay ξE28y 4 A 2 Ha 2 + 4Ha Hk + Ha Hy + 4Hk 2 + 2Hk Hy + 5
Coupled with the temperature devel-  2a   
A k
2
H k + H H
y k + A A
a k 2H H
a k + H H
a y + 4H k
2
+ 3H H
k y


opment and moisture changes, the γ R ðxd ,z1 Þ = 2
oφ
3 1 −γ RRðxd Þ
 !2 2
restraint from the slab induces stresses E28old 1 Aa + Ak Ay Hy
6 + + 7
in the wall. As is well known, both the 6 ξE E28old A 3 7
6 28y ξE28y y 7
6 7
degree of heat generated in the fresh 6 1 E28old     2 7
6 − Aa Ha + 2Hk + Hy − Ak Hk + Hy + 7
concrete and the restraint are influ- 6 A ξE 7
6


y

28y
  7
6 2 2 7
enced by the shape and size of the 6 1 + E28old Aa + Ak 2 Aa Ha + A H + 2H + H + E28a Ak ðH + H Þ + 7
6 ξE28y a a k y a k 7
structure. The casting sequence also 6 Ay 3 ξE A
28y y 7
6

   2 7
has an effect on the tensile stresses in 4 2 Ak H 2
E28a Aa 5
E28a Aa + Ak
1 + ξE k
+ A k H k + H y − ðH a + H k Þ
the concrete.18,19 The degree of
28y Ay 3 ξE28y Ay
restraint in a concrete wall depends ð3Þ
mainly on the vertical distance from
the slab–wall interface and the length-
to-height ratio, L/H, of the wall.20,21 where z1 is the coordinate from the
structural member. In this case, wall-
centroid of the transformed section to
When contraction of concrete is con- on-slab cast on frictional ground gives
the contact area; E28y, E28old are the
sidered in areas of high restraint from γRT = γRR = 0.
modulus of elasticity for the young
adjoining members, internal axial ∂γ 0
concrete and old concrete respec- The derivative of Eq. (1), ∂HRk , which
forces occur.22 So, both the axial tively; ξ is the time factor for modulus equals zero leads to the optimum
(translation) restraint and the rota- of elasticity = 0.93.26 This relation is kicker height. This can be derived
tional restraint affect the decisive commonly used in civil engineering analytically, but since the procedure
restraint in the wall; equations derived structures in Sweden. It is based on will result in very lengthy formulas,
according to the beam theory are pre- E28old/ E28y as an average value from Eqs. (1)–(3) are applied to draw the
sented in the next section. te ≈ 7d at zero stress point to a maxi- restraint curve depending on the
mum stress ratio when te ≈ 28d equals kicker height variation, and the opti-
Theoretical Derivation ~1.07. If the adjoining structure is cast mum kicker height is thus obtained
of Restraint Formulas with with the same concrete, a for the minimum restraint value.
the Kicker An application using Eqs. (1)–(3) is
Wy shown in Figs. 2–4. The studied struc-
For simple structures, analytical for-
ture is a wall (0.4 m thick and 6 m
mulation can be derived23 for a
high) cast on an existing slab (4 m
restraint coefficient in young concrete
Ay, ξE28y wide and 1 m thick). As can be seen
in terms of the translational and rota-
in Fig. 2, the translation restraint at
tional boundary conditions for a
Hy
the contact area between young con-
length coordinate (x) at the decisive
Htot crete (upper part of the wall) and the
cross-section, where x is denoted as
older concrete (the kicker) is almost
(xd). The young concrete part (Wy ×
linear dependent on the kicker height.
Hy) here is assumed to be rectangular Ak, E28old
On the other hand, in Fig. 3 it can be
shaped, while the old concrete part is Hk
seen that the rotational restraint at
assumed to consist of two rectangular Aa, E28a the contact area is non-linear depend-
shaped areas including the slab (Wa × Ha ent on the kicker height. The real
Ha) and the kicker (Wy × Hk) (see
Wa,eff restraint is a combined effect of trans-
Fig. 1). Htot denotes the total height of
lation and rotational restraint (see
the wall (Htot = Hk + Hy).
Fig. 1: Definition of parameters for the Eq. (1) and Figs. 2 and 3). The com-
The present derivation is based on decisive cross-section (x = xd) containing a bined effect has a minimum value for
section analyses according to the beam young concrete part cast on an older a certain kicker height, denoted here
theory, as in Refs. [23–25]. Plane sec- concrete part consisting of a slab and a as the optimum kicker height (see the
tions are assumed to be remaining on kicker (Units: –) curve for Eq. (1) in Fig. 4).

Structural Engineering International Nr. 3/2016 Scientific Paper 217

 6 6

5 5
Kicker height, m

Kicker height, m
4 4

3 3

2 2

1 1
Translation restraint Rotation restraint
0 0
0 0.1 0.2 0.3 0.4 –0.2 –0.1 0 0.1 0.2 0.3 0.4

F ig . 2: Translation restraint variation with kicker height, F i g. 3: Rotational restraint variation with kicker height,
see Eq. (2) see Eq. (3)

6 6
Eq. 1 Beam theory, Eq. 1
5 FE. mesh 0.5 × 0.5 5.5 Mesh 0.25 × 0.25
FE. mesh 0.25 × 0.25
Mesh 0.5 × 0.5
Kicker height, m

Wall height, m
5 Mesh 0.125 × 0.125

3 4.5

2 4

1 3.5
Restraint Restraint
0 3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 –0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

F i g. 4: Restraint variation at contact area with different kicker F i g. 5: Restraint variation with mesh element using optimum
heights kicker height 3 m

The beam theory calculations are
compared with the finite element
(FE) calculations for a structure with
total length of 40 m and L/H equal to
6.67. The beam theory is usually
regarded as applicable for L/H > 5.
The restraint curves in Fig. 4 appear
to be different, but the FE results only
show a mesh-dependent effect for the
restraint at the contact area. This can
be seen in Fig. 5 for wall heights of
6 m using the optimum kicker height
of 3 m. The decisive position for crack
risk design is located approximately
one wall thickness above the construc-
tion joint,23 in this case 0.4 m.
From Fig. 5 using the optimum kicker
height of 3 m it is evident that the
beam theory and FE calculations with
different element meshes give practi-
cally the same decisive restraint and
restraint distribution. Therefore it is
acceptable in this case to use FE
results with mesh 0.25 × 0.25 m to
save calculation time.
As can be seen in Fig. 4, the use of
kickers significantly reduces the
restraint, from about 0.85 without
using kickers (kicker height = 0) to
0.5 for the optimum kicker height of
approximately 3 m.
Numerical Estimation of the
Restraint for Different
Kicker Heights
Here, the structure and the numerical
estimations are the same as in
previous section, and additionally the
distribution of restraint values for the
whole structure is investigated for
both horizontal and vertical restraint
values. The total structure is analyzed
elastically by FE software. The
restraint values in Table 1 are calcu-
lated using the C3D8R element with
mesh size 0.25 × 0.25 m. The kickers
vary between 0 m and 4 m (see
Table 1). The restraint γ R is evaluated
from FE calculations defined as:4
σ ui
γR =
ð −Δε0 EC Þ
where σ ui = resulting stress from the
elastic calculation; i = a chosen direc-
tion in the concrete body; u = uniaxial
coordinate in i direction; Δε0 = the
homogenous contraction in the
concrete; and EC = Young’s modulus
in the newly cast concrete. When the
lower part of the wall is cast together
with the slab, that part is called as
kicker. There is no restraint in the
joint between the kicker and the slab.
Hereby, the behaviour of the structure
changes from a typical wall-on-slab
case to a typical wall-on-wall case.
As shown in Table 1, the horizontal
restraint γR33in the newly cast concrete
wall is decreased by 21 % (from 0.88
to 0.69) when using a 1 m high kicker
compared with no use of kicker, and
for a kicker equals 2 m the reduction is
36 % (γR33 about 0.56). The lowest
restraint equals 0.53 at kicker height
3 m, i.e. the optimum kicker height is
3 m. In Fig. 6, the distribution of hori-
zontal (S33) and vertical stresses (S22)
(here stresses mean restraint values,
γR33 and γR22) is presented for cases
ð4Þ 1–4 in Table 1.
The vertical restraint γR22 in the
kicker too offers an interesting analy-
sis, with an increased local risk of hor-
izontal cracking near the ends of the
kickers. As can be seen in Fig. 6b,
there are very small vertical restraints

218 Scientific Paper Structural Engineering International Nr. 3/2016

 Case Kicker height, m Newly cast wall γR33 γR22
height (m)
1 0 6 0.88 0.19
2 1 5 0.69 0.27
3 2 4 0.56 0.33
4 3 3 0.53 0.45
5 4 2 0.54 0.52
Table 1: Example of effects on the restraint in typical wall-on-slab case using kickers

(a) Case1- stress33 (c) Case2- stress33

S, S33 S, S33
(Avg: 100%) (Avg: 100%)
+8.838e–01 +6.920e–01
+5.300e–01 +5.300e–01
+4.858e–01 +4.858e–01
+4.417e–01 +4.417e–01
+3.975e–01 +3.975e–01
+3.533e–01 (b) Case3- stress22 (d) Case2- stress22 +3.533e–01
+3.092e–01 +3.092e–01
+2.650e–01 +2.650e–01
+2.208e–01 +2.208e–01
+1.767e–01 +1.767e–01
+1.325e–01 +1.325e–01
+8.833e–02 +8.833e–02
+4.417e–02 +4.417e–02
+0.000e+00 +0.000e+00
–6.144e–01 –4.943e–01

(e) Case3- stress33 (g) Case4- stress33

S, S33
(Avg: 100%)
+5.571e–01
+5.300e–01 S, S33
+4.858e–01
(Avg: 100%)
+4.417e–01
+3.975e–01 +5.347e–01
+3.533e–01 +5.300e–01
+3.092e–01 +4.858e–01
+2.650e–01 +4.417e–01
+2.208e–01 +3.975e–01
+1.767e–01 (f) Case3- stress22 (h) Case4- stress22 +3.533e–01
+1.325e–01 +3.092e–01
+8.833e–02 +2.650e–01
+4.417e–02 +2.208e–01
+0.000e+00 +1.767e–01
–5.739e–01 +1.325e–01
+8.833e–02
2 +4.417e–02
+0.000e+00
–6.495e–01

3
1

Fi g. 6: Resulting restraints horizontally (γ R33) and vertically (γ R22) for different kicker heights, see values in Table 1 (Units: –)

without kickers, but it has been
shown27 that very high local vertical
restraints exist in the wall near the
ends. These local restraints have to be
counteracted by the use of reinforce-
ment bars, but the local vertical
restraint values are rather small in
Fig. 6, only about 0.45.
It can be seen that for the wall-on-slab
case not using kickers, the increased
height of the wall always led to smaller
restraint and vice versa.23,28,29 From
Fig. 6 and Table 1 it is clear that an
increase of the kicker height from zero
to the optimum height, i.e. a decrease
of the upper part of the wall, will
decrease the restraint monotonically.
The decrease of the upper wall should
not be confused with the decrease of
the total wall height, as the effects on
the restraint are reversed.
The reduction of the restraint by
increasing the kicker height up to the
optimum height, i.e. the decrease in
the height of the upper part of the
wall, shows that using the kicker
changes the situation from a wall-on-
slab to a wall-on-wall behaviour. The
restraint reflects a balance between
the new concrete volume and the
existing adjacent structure.29 These
effects are interesting searching for
the optimum height of the kicker in
the next section.
Comparative Estimations
Using Optimum Kicker
Heights and Other Site
Measures
High-strength concrete (with low water
to cement ratio) increases the volumet-
ric changes of concrete at early age.3,5
Several methods are proposed to
reduce the risk of cracking such as

Structural Engineering International Nr. 3/2016 Scientific Paper 219

 6 4.5
No kicker No kicker
4
5 0.25 m kicker 0.5 m kicker
0.5 m kicker 3.5 1.0 m kicker
1.0 m kicker 1.5 m kicker
4 3
Wall Height (m)

Wall Height (m)

1.5 m kicker
1.75 m kicker 2.5
3 2.0 m kicker
0.926
2
0.60 0.935
2 1.5
0.596
0.538 1
1.08
1 0.562
0.641 0.5
1.24
0.702
0 0.772 0
–0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.2 0.4 0.6 0.8 1 1.2 1.4
Restraint Strain ratio

F i g. 7: Effect of kicker height on restraint for casting F ig . 8: Effect of kicker height on cracking risk for casting
length = 6 m length = 6 m

controlling of material mixtures, 4.5

No Kicker
decreasing the generated temperature, EQM
4
using higher water-to-cement ratios, 1.5 m Kicker
optimizing the construction methods, 3.5 No Kicker+ 3 cooling pipes
No Kicker+ 12 heating cables
and controlling the highest tempera- 3 1.5 m Kicker+ 3 cooling pipes
ture that can be reached and the tem- 1.5 m Kicker+ 6 heating cables
Wall Height (m)

perature differences between and 2.5

within the concrete members.30
In spite of using some material related
actions, the risk of cracking is still
present;29 therefore, on-site measures
are needed in many situations. For
example, the use of cooling pipes
and/or heating cables has been highly
efficient in reducing the risk of
cracking.
In the present study, a specific
software31 and the equivalent restraint
method, EQM,4 are used to compare
some of these measures with the pro-
posed technique of using kickers.
The strain ratio η is calculated accord-
ing to:
εm
η= ð5Þ
ε1
where εm = stress related strain, which
may also be denoted “material” strain
and ε1 = strain associated with the ten-
sile strength for a linear σ − εcurve.31
The formulation for concrete cracking
risk can be expressed as stress or strain;
here, the accuracy of the results is not
considered, rather the understanding of
the formulation from the user’s point
view. A stress formulation may be easy
to understand because the failure crite-
rion is related to strength while a for-
mulation of strain criterion better
describes the deformation that actually
occurs in the concrete element during
the hardening process.26
2 0.72 0.896
1.5
1
0.5
0
–0.5
0 0.2
F i g. 9 : Comparison between using the kicker technique and common on-site measures
The study uses a slab with dimensions
(width × thickness) 4 × 1 m and a wall
with dimensions (thickness × height)
0.4 × 6 m. The length of the structure
is 6 m. The resulting restraint curves
shown in Fig. 6 are arrived at by using
FE calculation with mesh size
0.25 × 0.25 m. Figure 7 shows that the
restraint in the newly cast wall-on-slab
without a kicker is 0.77, and that the
restraint decreases with increasing
kicker height. The optimum kicker
height is 1.5 m with a restraint equal-
ling 0.54. Kicker heights larger than
and smaller than 1.5 m increase the
restraint.
Figure 8 shows that the cracking risk
(strain ratio) in the newly cast wall-
on-slab without a kicker is 1.24, and
that the cracking risk decreases with
increasing kicker height. The opti-
mum kicker height is 1.5 m with
cracking risk equal to 0.926. The
0.926
0.853 1.24
1.08
1.24
0.4 0.6 0.8 1 1.2 1.4
Strain ratio
critical strain ratio for through crack-
ing is normally highest late in the
cooling phase. Cracking will probably
only occur theoretically if the calcu-
lated strain ratio exceeds 1.0.
Figure 9 shows the risk of cracking
when using kickers with and without
usage of cooling pipes in the wall or
heating cables in the slab. Using three
cooling pipes of steel (25 mm),
reduces the cracking risk to 0.85 from
1.24. When using 12 heating cables the
cracking risk decreases to 1.08. The
combination of three cooling pipes
with 1.5 m kicker height reduces the
cracking risk to 0.72, while a combina-
tion of six heating cables with a 1.5 m
kicker height reduces the cracking risk
to 0.90. These results demonstrate the
efficiency of using the kicker tech-
nique to reduce the cracking at early
ages of concrete. The next section pre-
sents the effects of the structural

220 Scientific Paper Structural Engineering International Nr. 3/2016

 Parameter Denotation Values No. of values Unit
Slab width Ba 2, 4, 8 3 m
Slab thickness Ha 0.4, 1, 1.4 3 m
Wall thickness Bc 0.3, 0.6, 1 3 m

Hc
Bc
Wall height Hc 3, 6, 8 3 m
Length of the L 3, 5, 10, 18 4 m
structure

Ha
L
*Summaryof 324 cases.
Ba
Table 2: List of parameters and their values used in the FE calculations* γRR

6
Casting length 3 m Study of Parameters
5
Casting length 5 m Influencing the Optimum
Casting length 10 m
Casting length 18 m Kicker Height
Wall height (m)

4
3 weights proposed in Ref. [33] and
2 study to determine the relative impor-
1 As shown in Fig. 15, the major param-
0 the length (L) followed by the wall
–0.2 –0.1
Fi g. 10: Restraint variations for different casting lengths using the optimum kicker height
dimensions of the wall and the slab on
the optimum kicker height.
Effects of Wall and Slab
Dimensions on the Optimum
Height of the Kicker
In all, 324 cases were used in the cal-
culations of the restraint in the walls
(see Table 2). The chosen spatial
dimensions are in accordance with
typical dimensions in real structures,
although some combinations might
not be good design in practice. Each
of these calculations has been ana-
lyzed elastically applying the three-
dimensional FE method (3D FE)
using suitable software.
On reviewing the calculations, signifi-
cant results were obtained concerning
variation both of the structural length
and the optimum kicker height. Most
studies in the literature recommend
decreasing the casting length to reduce
the restraint,9–11,32 which results in
more number of casting joints and
increase in the construction cost.
Figure 10 shows a wall of dimensions
0.6 × 6 m cast on a slab with dimen-
sions 1 × 4 m. The lengths of the cast-
ings are 3, 5, 10 and 18 m, and the
corresponding optimum kicker heights
are 1, 1.75, 3.25 and 3.25 m
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Restraint
respectively. The restraint is almost the
same, about 0.63, for all cases. This
shows that, by using the kicker tech-
nique, the restraint level can be main-
tained even when the length of the
casting is increased.
In the case study in Fig. 10, the
restraint has been examined by
varying the kicker height until the low-
est restraint is reached in each case,
i.e. by using the optimum kicker height
for each casting length. The restraint is
calculated as the average wall thick-
ness, which is regarded to be valid for
analyses of the risks of through crack-
ing. The outcome has been processed
using artificial neural network (ANN)
tools, see e.g. Ref. [29], where 90% of
the results were used for training the
ANN in two steps. The first step used
the data from the database to estimate
the optimum kicker height (see
Fig. 11). The second step comprised
calculations of the restraint at the base
of the upper part of the wall using
optimum kicker heights (see Fig. 12).
The remaining 10% of data was used
for testing of the ANN after the
training. The test showed good accu-
racy compatible with the FE calcula-
tions. This compatibility reached up to
93% for the kicker heights and 96 %
for restraint calculations; see Figs. 13
and 14 respectively.
The method of the partitioning
adopted in Ref. [34] is used in this
tance of the various input parameters.
eter influencing the kicker height is
height (Hc). Then the wall thickness
(Bc) and slab thickness (Ha) follow.
The width of the slab (Ba) has the
least influence. The ANN tool is used
in the parametric study in Fig. 15.
Effect of Wall Height
The major factors affecting the kicker
height are the length and height of the
wall. In the typical wall-on-slab case, it
is seen that the increase in the length-
to-height ratio of the wall leads to
greater restraint.1,26,35 In addition,
increased wall height too reduces the
restraint.1,3,28,29,36 Therefore, increased
kicker height helps to introduce a bal-
ance between old and new concrete.
As shown in Fig. 16, a bigger wall
height will raise the optimum kicker
height. When varying the wall height
(Hc) between 3 and 7 m, the optimum
kicker height is practically constant
from a certain casting length of the
structure. After further calculations, it
can be concluded that the optimum
kicker height is about (0.54–0.58) × Hc,
for length-to-height ratios, L/Hc ≥ 2.2.
Further, increase in length of the struc-
ture increases the optimum kicker
height for L/Hc ≤ 2.2.
Effect of Wall Thickness
Increased wall thickness reduces the
restraint in the wall for a typical wall-
on-slab case.29 Hence, this may lead
to a conflict in the balance between
old and new concrete. The wall thick-
ness in Fig. 17 was varied between 0.3
and 1.0 m, and the wall height was

Structural Engineering International Nr. 3/2016 Scientific Paper 221

 Outputs vs. targets, R = 0.96111 Outputs vs. targets, R = 0.98251
5 0.75
Data Points Data Points
Best Linear Fit Best Linear Fit
Outputs Y, Linear Fit: Y = (1)T + (0.0077)

OutputsY, Linear Fit:Y = (1)T + (–0.0018)

Y = T 0.70 Y = T

3 0.65

2 0.60

1 0.55

0 0.50

–1 0.45
–1 0 1 2 3 4 5 0.5 0.55 0.6 0.65 0.7
Targets T Targets T
F ig . 11 : Training results of ANN for optimum kicker height F ig . 12 : Training results of ANN for restraint calculation
calculation (Units: –) (Units: –)

Outputs vs. targets, R = 0.93069 Outputs vs. targets, R = 0.96874

4.5 0.70
Data Points Data Points
4.0 Best Linear Fit Best Linear Fit
OutputsY, Linear Fit: Y = (0.93)T + (0.044)
OutputsY, Linear Fit:Y = (0.98)T + (0.085)

Y = T Y = T

3.5 0.65

3.0

2.5 0.60

2.0

1.5 0.55

1.0

0.5 0.50
–1 0 1 2 3 4 5 0.50 0.55 0.60 0.65 0.70
Targets T Targets T

F i g. 13: Comparison between FE and ANN calculations for F i g. 1 4: Comparison between FE and ANN calculations for
kicker height calculation (Units: –) restraint calculations (Units: –)

5 m for all cases. As can be seen in Effect of Slab Thickness between Figs. 17 and 18 shows that the
the figure, the optimum kicker height optimum kicker height is quite similar
Increased slab thickness, i.e. growth of
is almost constant, about 3 m, for
stiffness of the structure, raises the in both figures. This shows that the
L/Hc ≥ 2.0 (L ≥ 10 m in the figure), thickness of the slab has no influence
and the optimum kicker height is restraint in a wall of typical wall-on-
slab case.29 The degree of restraint in on the optimum kicker height.
found increasing with the length of
structures for L/Hc ≤ 2.0. In addition, a concrete wall depends mainly on the
vertical distance from the slab–wall Effect of Slab Width
the optimum kicker height is slightly
increased when increasing the wall interface and the length-to-height ratio The results in Fig. 15 show that the
thickness. L/H of the wall.20,21 A comparison width of the slab has only a small

222 Scientific Paper Structural Engineering International Nr. 3/2016

 35 33.37 shown in “Comparative estimation”
section and in Fig. 9.
30 29.00 • For the same casting length, the
use of the optimum kicker height
25
always reduces the restraint (see
% Importance
20
Figs. 7 and 8). This answers the
16.22 research question, and is seen in
15 14.22 Fig. 10, where approximately the
same restraint can be reached for
10 casting lengths between 3 and 18 m
7.169
using optimum kicker heights,
5 depending on the casting length.
0
Slab Slab Wall Wall Length
width thickness thickness height Conclusion
Fi g. 15: Relative importance of input parameters using optimum kicker heights (Units: –)
In general, the restraint in concrete is
affected by the dimensions of both
influence on the optimum kicker • The kicker technique can decrease the newly cast structure and the
height. Generally, increasing the slab the restraint as compared with adjoining structure. The proposed
width means bigger restraint.3,37,38 that of not using kickers as shown technique of casting the bottom part
Increasing the width of the slab in the “Numerical estimation” of the wall together with the slab, i.-
increases the optimum kicker height section and in Table 1. Thereby, e. using the kicker technique, can
for all cases of structural lengths (see the risk of through cracking can change the situation significantly, from
Fig. 19). be reduced as shown in Figs. 7 a typical wall-on-slab case to a typical
and 8. wall-on-wall case behaviour. This
Answers to Research • The use of kickers can be combined change is accompanied by a decrease
with other measures, such as cool- in the restraint and reduces the risk of
Questions ing of the new concrete and/or cracking in the upper part of the wall.
Based on the study in this paper the heating of the old concrete, for Using the kicker technique can
answers of the research questions are: reduction of cracking risks, as increase the casting length while

4.0 3.5

3.5
Optimum kicker height (m)

3.0
Optimum kicker height (m)

3.0
2.5
2.5
2.0
2.0
1.5
1.5 Ba = 5 Hc = 3 Ba = 5 Bc = 0.3
Ha = 0,9 Hc = 4 Ha = 0,9 Bc = 0.5
Bc = 0,65 Hc = 5 Hc = 5 Bc = 0.7
1.0 Hc = 6
1.0
Bc = 0.9
Hc =7 Bc = 1.0
0.5 0.5
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Length of the structures (m) Length of the structures (m)

Fi g. 16: Effects of wall height and length on optimum kicker Fi g. 17: Effects of wall thickness and length on optimum kicker
height; see Table 2 for definition and units of the parameters height; see Table 2 for definition and units of the parameters

3.5 3.5

3.0
Optimum kicker height (m)

3.0
Optimum kicker height (m)

2.5 2.5

2.0 2.0

1.5 1.5
Ba = 5 Ha = 0.4 Ha = 0,9 Ba = 2
Bc = 0,65 Ha = 0.6 Bc = 0,65 Ba = 3
1.0 Hc = 5 Ha = 0.8 Hc = 5 Ba = 5
1.0
Ha = 1.0 Ba = 6
Ha = 1.2 Ba = 7
0.5 0.5
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Length of the structures (m) Length of the structures (m)

Fi g. 18: Effects of slab thickness and length on optimum kicker Fi g. 19: Effects of slab width and length on optimum kicker
height; see Table 2 for definition and units of the parameters height; see Table 2 for definition and units of the parameters

Structural Engineering International Nr. 3/2016 Scientific Paper 223

controlling the restraint to smallest
possible value. By comparing estima-
tions using the kicker technique with
estimations using more common mea-
sures, such as cooling pipes and/or
heating cables, the use of kickers has
been shown to be sufficient to reduce
the cracking risks at early ages of con-
crete. Therefore, the proposed
method is useful to reduce the risk of
cracking and to minimize the cost of
the structure.
References
[1] ACI Committee 207. Effect of restraint,
Volume Change, and Reinforcement on Crack-
ing of Massive Concrete. ACI Committee 207
ACI207.2R-95. Reapproved 2002, 26.
[2] Emborg M, Bernander S. Assessment of the
risk of thermal cracking in hardening concrete.
J. Struct. Eng. 1994;120(10):2893–2912.
[3] Mihashi H, Leite JP. State of the art report
on control of cracking in early age concrete.
J. Adv. Concr. Technol. 2004;2(2):141–154.
[4] Al-Gburi M, Jonasson JE, Nilsson M,
Hedlund H, Hösthagen A. Simplified methods
for crack risk analyses of early age concrete
part 1: development of equivalent restraint
method. Nordic Concrete Research Publication
No.46, 2012, 17–38.
[5] Sato R, Shimomura T, Maruyama I,
Nakarai K. Durability mechanics of concrete
and concrete structures re-definition and a new
approach. Committee Reports of JCI, 8th Inter-
national Conference on Creep, Shrinkage and
Durability of Concrete and Concrete Structures
(CONCREEP8), Ise-Shima, Japan, 2008, 1073–
1098.
[6] Emborg M, Bernander S. Avoidance of
early age thermal cracking in concrete struc-
tures – predesign, measures, follow-up. In Proc-
eedings of the RILEM International Symposium
‘Thermal Cracking in Concrete at Early Ages’,
Springenschmid R (ed) E & FN Spon: London,
UK, 1994; 409–416.
[7] ACI 207.4R-93. Cooling and insulating sys-
tems for mass concrete. ACI Manual of Con-
crete Practice, Part 1–1992: Materials and
General Properties of Concrete, Detroit, MI,
1994, 22.
[8] Uuckfeldt J, Duddeck, H, and Ahrens, H,
Numerical Simulation of Crack avoiding
measures, Proceedings of the RILEM Interna-
tional Symposium ‘Thermal Cracking in Con-
crete at Early Ages’, R Springenschmid, E & FN
Spon, London, UK, 1994; 255–265.
[9] JSCE. English Version of Standard Specifi-
cation for Concrete Structures 2007, Japan Soci-
ety of Civil Engineer: Tokyo, Japan, 2010,
503 pp.
224 Scientific Paper
[10] Bamforth PB. Early age thermal crack
control in concrete. In CIRIA 2007, Construc-
tion Industry Research and Information
Association: London, UK, 2007; 112, RP722.
ISBN 978-8-86107-660-2.
[11] Wu S, Huang D, Lin B, Zhao H, Wang P.
Estimation of cracking risk of concrete at early
age based on thermal stress analysis. J. Therm.
Anal. Calorim. 2001;105(1):171–186.
[12] ACI 224R-90. Control of cracking in con-
crete structures. ACIA4 Manual of Concrete
Practice, Part 3, American Concrete Institute,
Detroit, MI, 1992.
[13] Kheder GF. A new look at the control of
volume change cracking of base restrained con-
crete walls. ACI Struct. J. 1997;94(3):262–271.
[14] Utsi S, Jonasson J-E. Estimation of the risk
for early thermal cracking for SCC containing
fly ash. Mater. Struct. 2012;45(1–2):153–169.
[15] Schindler AK, McCullough BF. The
importance of concrete temperature control
during concrete pavement construction in hot
weather conditions. J. Transport. Res. Board
2002;1813:3–10.
[16] Bentz DP. A review of early-age proper-
ties of cement-based materials. Cement Concr.
Res. 2008;38:196–204.
[17] Harrison, TA. Early-age Thermal Crack
Control in Concrete. CIRIA Report No. 91,
Construction Industry Research and Informa-
tion Association, 1992, London, UK.
[18] Lin F, Song X, Gu X, Peng B, Yang L.
Crackinganalysis of massive concrete walls with
cracking control techniques. Construct. Build.
Mater. 2012;31:12–21.
[19] Lebet JP, Ducret JM. Early concrete
cracking of composite bridges during construc-
tion. In Composite Construction in Steel and
Concrete, vol. IV ASCE: Reston, VA, 2002;
13–24.
[20] Zhou J, Chen X, Zhang J. Early-age
temperature and straining basement concrete
walls: field monitoring and numerical modeling.
J. Perform. Construct. Facil. 2012;26:254–265.
[21] Cusson D, Repette W. Early-age cracking
in reconstructed concrete bridge barrier walls.
ACI Mater. J. 2000;97(4):438–446.
[22] Kwak HG, Seo YJ, Jung CM. Effects of
the slab casting sequences and the drying
shrinkage of concrete slabs on the short-term
and long-term behavior of composite steel box
girder bridges. Part 1. Eng. Struct.
2000;23:1453–1466.
[23] Nilsson M. Thermal Cracking of Young
Concrete- Partial Coefficients, Restraint Effects
and Influence of Casting Joints. Licentiate The-
sis, Division of Structural Engineering, Luleå
University of Technology, Luleå, Sweden, 2000,
27, 267 pp.
[24] Collins MP, Mitchell D. Prestressed Con-
crete Structures, Prentice-Hall Inc: Englewood
Cliffs, NJ, 1991, 766 pp. ISBN 0-13-691635-X.
Structural Engineering International Nr. 3/2016
[25] JCI. A Proposal of a Method of Calculat-
ing Crack Width due to Thermal Stress. JCI
Committee Report, Japan Concrete Institute,
Committee on Thermal Stress of Massive Con-
crete Structures, Tokyo, Japan, 1992, 106 pp.
[26] Larson M. Thermal Crack Estimation in
Early Age Concrete Models and Methods for
Practical Application. Doctoral Thesis, Division
of Structural Engineering, Luleå University of
Technology, Luleå, Sweden, 2003.
[27] Bernander, S. Practical measurement to
avoiding early age thermal cracking in concrete
structures. In Prevention of Thermal Cracking
in Concrete at Early Ages, RILEM Report
No. 15. State of the Art Report by RILEM Tech-
nical Committee 119, Prevention of Thermal
Cracking in Concrete at Early Ages,
Springenschmid R (ed), E & FNSpon: London,
UK, 1998, 255–315.
[28] Kheder GF, Al-Rawi RS, Al-Dhahi JK. A
study of the behavior of volume change crack-
ing in base restrained concrete walls. Mater.
Struct. 1994;27:383–392.
[29] Al-Gburi M, Jonasson JE, Yousif ST,
Nilsson M. Simplified methods for crack risk
analyses of early age concrete part 2: restraint
factors for typical case wall-on-slab. Nordic
Concrete Research 2012; 46: 39–58.
[30] Folliard KJ, Juenger M, Schindler A,
Riding K, Poole J, Kallivokas LF, Slatnick S,
Jared W, Meadows JL. Prediction Model for
Concrete Behavior Texas Department of Trans-
portation: Austin, TX, 2008, 78 pp.
[31] CTS Pro. User’s Manual – Program for
Temperature and Stress Calculations in
Concrete. Developed by JEJMS Concrete AB
in Co-operation with Luleå University of Tech-
nology, Cementa AB and Peab AB, Luleå,
Sweden, 2003, 198 pp.
[32] Choktaweekarn P, Tangtermsirikul S.
Effect of aggregate type, casting, thickness and
curing condition on restrained strain of mass
concrete. Songklanakarin J. Sci. Technol.
2010;32(4):391–402.
[33] Garson GD. Interpreting neural network
connection weights. Artif. Intell. 1991;6:47–51.
[34] Goh ATC. Back-propagation neural net-
works for modeling complex systems. Artifi.
Intell. Eng. 1995;9(3):143–151.
[35] Kim SC. Effects of a Lift height on the
thermal cracking in wall structures. KCI Concr
J. 2000;12(1):47–56.
[36] Emborg M. Thermal Stresses in Concrete
Structures at Early Ages. Doctoral Thesis, Div-
ison of Structural Engineering, Luleå University
of Technology, 1989, 73D, 280 pp.
[37] Nagy A. Parameter study of thermal crack-
ing in HPC and NPC Structures. Nordic Con-
crete Research, 2001, 26.
[38] Kwak HG, Ha SJ. Non-structural cracking
in RC walls: part II. Quantitative prediction
model. Cement Concr. Res. 2006;36:761–775.