Received: 2 July 2020 Revised: 14 November 2020 Accepted: 19 November 2020

DOI: 10.1002/suco.202000403

TECHNICAL PAPER

Experimental and analytical study on tension lap splices

in nonconventional concrete using different techniques

Tarek Fawzy1 | Abdel-Hakim Khalil1 | Ahmed Atta1 | Hamdy M. Afefy2 |

Mohamed Ellithy1

1
Department of Structural Engineering,
Faculty of Engineering, Tanta University,
Abstract
Tanta, Egypt An experimental and analytical study on 15 specimens of lap splices embedded
2
Construction Engineering and in nonconventional concrete without confining reinforcement was performed
Management Department, Faculty of
under direct tension. The experimental program involved two groups of
Engineering, Pharos University,
Alexandria, Egypt straight-ended and anchored-ended bars; three different techniques were
employed with different splice lengths. For straight-ended bars, conventional
Correspondence
lap splice was studied as a benchmark. The anchored-ended steel bars were
Hamdy M. Afefy, Construction
Engineering and Management hooks, hooks intersecting with cross bars, and plate-end bars. It was found
Department, Faculty of Engineering, that, the used techniques not only achieved higher tensile stress at failure, but
Pharos University, Alexandria, On leave
from Tanta University, Egypt.
also different modes of failure were observed. Besides, a splice length of five
Email: [email protected], hamdy. times bar diameter was found to be sufficient to achieve the nominal yield
[email protected] stress of the reinforcing steel bar in the case of anchored-ended bars. Finally,
an analytical model was proposed in order to predict the ultimate tensile stress
of the straight-ended spliced bars. The accuracy of the proposed model was
verified against the test results of 137 existing specimens from previous
research. The comparison showed good agreement between the results of the
proposed model and the test results.

KEYWORDS
anchored bars, direct tension, lap splice, nonconventional concrete, plate-end bars

1 | INTRODUCTION previous studies revealed that the main governing param-

eters affecting on the splice behavior are concrete type/
The overall structural performance of reinforced concrete strength, concrete cover, confinement state of the
elements is mainly dependent on the bond characteristics spliced/developed bars and splicing techniques.
between reinforcing steel bars and the surrounding con- Since the type/strength of concrete is a main control-
crete. Therefore, the development length as well as lap ling parameter, using conventional/modern types of con-
splice of reinforcing steel bars have to be fully under- crete affects on the bond behavior. Based on the outcomes
stood. Thus, this area of research has attracted the inter- of several experimental studies, many researchers found
est of many researchers worldwide. Broad overview on that the bond strength is directly proportional to the
square root of the compressive strength of concrete.1–5 On
the other hand, for higher compressive strength of con-
Discussion on this paper must be submitted within two months of the
print publication. The discussion will then be published in print, along
crete (fc0 > 55 MPa), the average bond strength at failure
with the authors’ closure, if any, approximately nine months after the becomes less proportional to the square root of the com-
print publication. pressive strength of concrete. This rate decreases more

Structural Concrete. 2020;1–15. wileyonlinelibrary.com/journal/suco © 2020 fib. International Federation for Structural Concrete 1
2 FAWZY ET AL.

obviously as splice length increases due to the nonuniform congestion and improves concrete consolidation, it
distribution of bond stress along the splice length.6,7 requires special labor and equipment, which entails addi-
For the last decades, adding fibers to different types of tional cost.18 One of the most effective methods for splic-
concrete generally affects on the mechanical properties of ing bars is to use the mechanical splices but due to its
concrete and its effect is mainly dependent on the fibers' higher expenses, it is considered a convenient method
type and content. Small amount of Polypropylene fibers only for large diameters. Headed bar can transfer the
(volume fraction less than 0.50%) added to concrete led to a bar's full strength to concrete with only 4db length of
slight decrease in workability but it increases its resistance splice.19 Inspite that headed bar could be considered as
for dry shrinkage as well as enhances its durability.8,9In an effective way to minimize the development length, it
order to increase the dry shrinkage resistance simultaneous still not practical method in narrow elements with adja-
with good workability for high performance fiber reinforced cent spliced/developed bars.
concrete, the optimum volumetric ratio of the polypropyl- As previously mentioned, welding, mechanical
ene fibers ranges between 0.25 and 0.50%.10 For seeking an splices, and headed bars have several drawbacks in prac-
obvious influence in the toughness and the tensile strength tice, thus, in the current study hooks in addition to plate-
of the concrete, researchers used other types of fibers (poly- end bars were utilized in order to improve the conven-
ethylene and steel fibers) with higher volumetric ratios up tional lap splices. These techniques were developed to
to 4%. Fiber reinforced cement composites showed that the boost the splice strength of steel bars embedded in NCC
fibers induced bridging effect after cracking can effec- specimens. The current research studied these splicing
tively support the postcracking tensile capacity of the techniques in NCC owing to its widely spread in several
concrete cover, limit splitting cracks width and enhance long-span structures at which it is more predictable to
the wedging action induced by the bar's ribs, thereby, have spliced reinforcement. The NCC in this study was
leading to enhanced bond resistance.11–13 These supe- intentionally selected with low content of Polypropylene
rior types of cement composites may suffer from low fiber to consider the workability, durability, and cost to
workability, thus, additional precautions are considered have a convenient mix to be used in massive wide-span
to avoid this low workability such as a restricted content structures. In this study, two groups of spliced specimens
of coarse aggregate and using several types of admix- were carried out; a reference group of straight-end bars
tures. These types of cement composites are more con- as well as a group of anchored-end bars for enhancement
venient for repair and strengthening purposes because of the splice strength. In addition, analytical equations
of its very high cost in case of massive quantities. were proposed then compared to the results of experi-
The bond of steel reinforcement is also dependent on mentally tested specimens.
the concrete cover as well as the spacing between the
spliced bars. Not only the bond strength is affected by
increasing the concrete cover and bar spacing, but also 2 | EXPERIMENTAL
mode of failure can be altered.4,5,14,15 Pullout failure is INVESTIGATION
recorded for cases with large concrete cover and bar spac-
ing; however, splitting tensile failure takes place for 2.1 | Test specimens and studied
smaller concrete cover and bar spacing. In addition, pull- parameters
out failures can take place with slightly observed splitting
cracks in case of spliced bars with a proper confinement In the current study, an experimental program involved
(ACI 408R-03). Transverse reinforcement contributes to 15 lap spliced bars embedded in NCC prisms without
restrict the progression of splitting cracks, which plays an confining reinforcement was carried out. The specimens
important role for increasing the bond strength of consisted of two pairs of lap spliced 12 mm diameter bars
spliced/developed bars.4,5,16,17 By increasing the amount embedded in NCC prisms without transverse reinforce-
of transverse reinforcement, the bond strength increases ment. The test specimens were loaded by axial tension
and a splitting failure can be switched to pullout failure. force-controlled system. Specimens were categorized into
Providing additional confinement more than that is two groups of straight-ended and anchored-ended bars;
needed, does not gain higher bond strength.4 three techniques were used with different splice lengths.
Several techniques were used to improve the perfor- The first group (Group I) is a reference group and studied
mance of the conventional spliced/developed bars such the conventional lap spliced bars as depicted in Figure 1,
as welding, mechanical splices, headed bars, and hooks. while the second group (Group II) studied the anchored-
The main idea of hooks and headed bars is to use anchors ended steel bars with miscellaneous configurations as
at the end of steel bars, which provides higher resistance shown in Figures 2–4. The spliced bars of Group II were
against slippage. Although welded splicing reduces rebar anchored by (1) steel plates, (2) standard 180 hooks of
FAWZY ET AL. 3

FIGURE 3 Configuration of sub-group G2-HC, standard 180

ACI-318 hook with cross-bars

FIGURE 1 Configuration of sub-group G1-O, conventional

of bars was connected to the bearing plate by welding to
lap splice
the plate's rear face in order to avoid the heating effects
on the bars within splice zone, while the other pair is
passing throw holes in the same plate. The plate-welded
connection is shown in Figure 6. In the lab, all bars were
inserted through the plate holes then, the connected bars
were bended prior to welding at the rear face of plate.
Note that, the feasibility of this technique in site condi-
tions needs further feedback from site engineers after
application.
Table 1 summarizes the concrete dimensions of the
NCC prisms and the lap splice lengths for each sub-
group. All dimensions in Table 1 were obtained regard-
ing to the nominal 12 mm diameter of the reinforcing
bar. The adopted notation system of specimens shows
the various testing parameters; where for specimen Gx-
y-zd; “x” indicates the group number (1, 2), “y” refers to
the sub-group (O, straight-ended bars; H, hooked-ended
bars; HC, hooked-ended bars intersecting with cross
bar; P, plate-end bars), and “z” is the splice length in
term of bar diameter (5, 10, 20, and 30db). The back of
the anchors was covered by concrete with thickness of
three times bar's diameter (3db), along this zone the
FIGURE 2 Configuration of sub-group G2-H, standard 180
bars were isolated by bond-breakers in order to main-
ACI-318 hook
tain the same splice length between the anchors for all
specimens.
ACI-31830, and (3) standard hooks intersecting with cross Although pullout test is commonly used owing to its
bars. The configuration of the standard 180 hook of ease of fabrication and the simplicity of the test, the stress
ACI-318 is shown in Figure 5. The cross bar intersected fields inside the specimen do not match with the actual
perpendicularly with the hook was selected with 16 mm construction. Since the bar is under tension, the sur-
diameter and 150 mm length. For plate-end bars, a Pair rounding concrete is under compression. In addition, the
4 FAWZY ET AL.

which allows testing of full-length lap splice without exe-

cution of full-scale beam splice specimen. The bond
strength obtained by this configuration showed variation
within 10% of the test results obtained by beam splice
specimens with similar conditions.20–22 In this research,
the minimum concrete cover was intentionally selected
as 3db for all test specimens, which is common in practice
for RC beams in the Middle East region.

2.2 | Steel bars properties

The nominal yield stress of the used reinforcing bars

was 400 MPa as commonly used in the Middle East
region. To estimate the average mechanical properties
of steel bars, three steel samples were collected then
tested in accordance with ASTM A370.33 The average
cross sectional area of the bar's (Ab) was calculated by
dividing the bar's weight by the density of steel then
FIGURE 4 Configuration of sub-group G2-P, plate-end splices inferring its volume and diameter. The average mechan-
ical properties and geometries of the tested samples are
listed in Table 2.
4 d b or 2.5 in.
which is greater

2.3 | Material properties of NCC

8 db 6d b

For all tested specimens, the same NCC mix was used
where all specimens were cast at the same time. The
concrete mix design used in the current research was the
FIGURE 5 ACI-318 standard hook of 180
same mix adopted in a previous research.23 The non-
conventional concrete mix consisted of the following:
(1) ordinary Portland cement of grade 52.5, (2) dolomite
coarse aggregates, (3) fine aggregates, (4) polypropylene
fibers, (5) silica fume with average diameters ranged
30 mm
between 0.1 and 0.3 μm, (6) superplasticizers, and (7) clean
water. Regarding to the manufacture's datasheet, the poly-
propylene fibers (Fibermesh 300) length is about
12–19 mm length with tensile strength of about 386 MPa
30 mm and elongation of 13.2%. The ingredients of concrete mix
per cubic meter (kg/m3) are summarized in Table 3. The
low content of the Polypropylene fibers added to the used
FIGURE 6 Steel plate in which the bar is welded to the concrete mix (0.2% volumetric ratio) can only reduce the
rear face effect of dry shrinkage and increase the durability of con-
crete as previously demonstrated in Section 1. Precautions
were taken into consideration in order to ensure bars
resulting compressive struts between the support points alignment inside the wooden forms during casting. The
and the reinforcing bar compress on the bar's surface current NCC mix showed satisfactory consolidation and
(ACI 408R-03). This state is significantly dissimilar to was poured with moderate vibration. After casting, speci-
most of reinforced concrete members at which both the mens were covered by wet textile then curing process
bar and the surrounding concrete are in tension. The began and continued for six days.
adopted configuration of test specimens in the current Concrete compressive strength fc' was determined
study at which both of concrete and steel are subjected to from testing of 150 × 300 mm standard cylinders
tension was selected in several previous researches, according to ASTM.32 Brazilian test was performed in
FAWZY ET AL. 5

TABLE 1 Test matrix

Prism dimensions (mm)

Group Sub-group Splice length Specimen notation Length Width Depth

(G1) O 10db G1-O-10d 120 180 84
Straight bars 20db G1-O-20d 240 180 84
30db G1-O-30d 360 180 84
(G2) H 05db G2-H-05d 130 180 170
Anchored bars 10db G2-H-10d 190 180 170
20db G2-H-20d 310 180 170
30db G2-H-30d 430 180 170
HC 05db G2-HC-05d 130 180 170
10db G2-HC-10d 190 180 170
20db G2-HC-20d 310 180 170
30db G2-HC-30d 430 180 170
P 05db G2-P-05d 130 180 84
10db G2-P-10d 190 180 84
20db G2-P-20d 310 180 84
30db G2-P-30d 430 180 84

Abbreviations: H, Hooked-end spliced bars; HC, hooked-end bars intersecting with cross bar; O, straight-end spliced bars; P, plate-ended bars.

TABLE 2 Steel bars geometries and mechanical properties 2.4 | Test setup
Property Splice bar
Splice bars were fixed to a rigid steel cap system through
Nominal diameter, db (mm) 12 steel grips and wedges to ensure that both lapped bars
Core diameter, D (mm) 12 were subjected to the same displacement. The rigid cap-
Average rib depth, hr (mm) 0.9 ping system was tested against bare bars, the cap defor-
Rib spacing, sr (mm) 8.5 mation at its midpoint was found inferior to 0.3 mm. In

Rib face angle, θ ( ) 37 order to assure the best fitting of the bars with grips, the
specimens were preloaded to 5 kN then released before
Relative rib area, Rr 0.1125
2
testing. The cap system was attached to the thick plate of
Ab (mm ) 114.7
the hydraulic actuator with rigid connections. The load
Average Young modulus, Es (GPa) 207 was applied with a force-controlled system, the force
Average yield strength, fy (MPa) 463 acted on the upper pair of bars with a constant loading
Average ultimate strength, fu (MPa) 641 rate of about 5 kN/min until failure. During the test,
Maximum strain (%) 12.7% actuator output data was exported to a data- acquisition
system with sampling frequency of 2 Hz. Figure 7 shows
the test set-up and the instrumentation.

order to obtain the indirect tensile strength ft of NCC.

Three standard cylindrical samples per each test were 2.5 | Instrumentation
prepared for each sub-group in order to determine the
compressive and tensile strengths at the testing day of For each spliced bar, one strain gauge outside the speci-
the splice specimens. To ensure early gained strength, men was mounted in order to ensure equal stress in both
three cylinders were tested; the average 7-day compres- bars during the test. A total of 4LVDTs (Linear Variable
sive strength was about 40.1 MPa. Table 4 reports NCC Differential Transformer) having 50 mm gauge length
main characteristics pertaining to each sub-group. The were installed on the concrete surface on the specimen's
tensile strength of the used NCC with low fiber content front face, perpendicular to the axis of the splice as
lies within the normal ranges comparing to its compres- depicted in Figure 7. The faces were sprayed with white
sive strength. paint in order to have a clearer view of the observed
6 FAWZY ET AL.

TABLE 3 Concrete mix ingredients of NCC (kg/m3)

Cement Water Water-to-cement ratio Sand Dolomite Silica fume Super plasticizer Polypropylene fibers
450 135 0.3 535 1,279 45 4.5 1.8

TABLE 4 NCC material properties

3 | RESULTS A ND DISCUSSION
Subgroup Age (days) fc0 (MPa) ft (MPa)
G1-O 28 54 4.7 3.1 | Overall behavior
G2-H 40 54 4.9
G2-HC 43 56 4.9 Table 5 summarizes the test results corresponding to maxi-
G2-P 54 58 5.1 mum load for all tested splice specimens. The maximum
steel stress at failure (fs) represents the average stress devel-
oped in the reinforcing bars. Bar's stress is the output of
dividing the maximum force by the average area of bar's
cross section (Ab). To allow all sub-groups to be compared
with the reference sub-group G1-O, the ratio between √fc' of
G1-O and √fc' of each sub-group was used to normalize the
obtained stress of steel. The normalized stress of steel (fsn) of
a specimen from any sub-group can be calculated as follows:
h i
f snðspecimenÞ = √f 0 cðG1 − OÞ =√f 0cðsub − groupÞ × f sðspecimenÞ ð1Þ

The maximum normalized steel stress was compared

to the actual yield stress of steel in term of (fsn/fy) to cal-
culate the efficiency of splice Ef. Assuming that the bond
stress is distributed uniformly along the splice length, the
normalized bond stress per unit length at failure (Un) is
estimated. To calculate the normalized bond stress, the
bar force at failure was calculated at the normalized ten-
sile stress level and then divided by the surface area of
the bar within the splice zone as follows:

U n = Ab f sn =π db Ls ð2Þ

where Ls = splice length. Eventually the modes of failure

were also reported in Table 5. For Group I specimens,
the splices with lengths of 10, 20, and 30db achieved a
normalized tensile stress of 265, 517, and 576 MPa,
respectively. G1-O-10d with splice length of 10db was
not able to achieve the design stress of the steel bar,
while, the other specimens of Group I recorded higher
tensile stresses compared to the yield stress of steel
bars. Providing ACI 180 standard hooks to the end of
FIGURE 7 Test setup and instrumentation the straight bars led to achieve the nominal yield stress
with just 5db of splice length. On the other hand, the
cracks on the surface of the specimens. Hydraulic actua- minimum length of the conventional splice of G1-O
tor was used to apply tension force where its outputs needs at least 20db length to achieve the yield stress.
(load and displacement) were recorded. The failure of Standard 180 ACI hooks enhanced the splice capacity
specimens is expected to be sudden and violent, thus, of 10, 20, and 30db splices by about 95, 11, and 1%,
tests were recorded by slow-motion high definition video respectively, compared to those of the conventional lab
camera of 120 frames per second. splices of G1-O.
FAWZY ET AL. 7

TABLE 5 Test results of lap splice specimens

Sub- Specimen Max. fs fsn Un Ef Mode of

Group group notation load (kN) (MPa) (MPa) (MPa) (fsn/fy) failure
G1 O G1-O-10d 60.9 265 265 6.71 0.58 P
G1-O-20d 118.7 517 517 6.55 1.12 YP
G1-O-30d 132.1 576 576 4.86 1.25 YP
G2 H G2-H-05d 99.5 434 434 – 0.94 SH
G2-H-10d 118.7 517 517 – 1.12 SH
G2-H-20d 132 575 575 – 1.25 YPC
G2-H-30d 133.2 581 581 – 1.26 YPC
HC G2-HC-05d 109.8 479 470 – 1.02 YPC
G2-HC-10d 143.3 625 614 – 1.33 YPC
G2-HC-20d 136.1 593 582 – 1.27 R
G2-HC-30d 139 606 595 – 1.29 R
P G2-P-05d 145.5 634 612 – 1.33 R
G2-P-10d 145.2 633 611 – 1.33 R
G2-P-20d 138.6 604 583 – 1.27 R
G2-P-30d 142.4 621 599 – 1.3 R

Abbreviations: Ef, the efficiency of splice; fs, the average stress developed in the reinforcing bars; fsn, the maximum normalized steel stress; fy, actual yield stress
of steel; P, Pullout of reinforcing bars; R, Rupture of reinforcing bars; SH, Direct shear on the plane of intersection of hooked lapped bars; Un, the normalized
bond stress per unit length at failure; YP, Yielding of steel followed by pullout of bars; YPC, Yielding of steel followed by pullout of bars then concrete
crushing.

It is clearly obvious that the anchoring techniques

adopted in specimens of G2, enhanced their capacities than
those of the same lap lengths of G1. Using cross-bar with
hooked reinforcement in G2-HC showed a noticeable effect
not only on the efficiency but also on the mode of failure.
Specimens of G2-HC recorded a normalized tensile stress of
470, 614, 582, and 595 MPa for splice lengths of 5, 10,
20, and 30db, respectively. Although the recorded stresses of
G2-HC-20d and G2-HC-30d were slightly less than that of
G2-HC-10d, these specimens failed out of splice due to rup-
ture of reinforcing bars. This was attributed to the slight dif-
ferences in the steel strength of the used splice bars. The
cross bar of G2-HC enhanced the splice strength than that
of G2-H by 9, 19, 2, and 2% for splice lengths of 5, 10,
20, and 30db, respectively. The contribution of cross bar in
the enhancement of the splice strength of short splices
G2-HC-05d and G2-HC-10d compared to other specimens
of G2-HC was significant because of different mode of fail-
ure was exhibited. All specimens of G2-P achieved the max- FIGURE 8 Bars pullout of G1-O-20d
imum potential strength of steel and exhibited rupture of
reinforcing bars.
in Figure 8. The failure of G1-O-10d took place before
achieving the yielding of steel bars, while the other speci-
3.2 | Modes of failure mens of G1-O failed beyond reaching to yield stress. Split-
ting cracks were not obvious during testing which is
Shear crack and/or local concrete crushing due to bar attributed to the effect of concrete cover thickness (3db).
pullout were noticed for all specimens of G1-O as shown The anchoring techniques used for G2 changed the
8 FAWZY ET AL.

FIGURE 9 The plane of direct shear failure of G2-H-10d

FIGURE 11 Compression force acting on concrete

the cross bar prevents the separation of concrete on the

plane of the direct shear.
For all specimens of sub-group G2-P, the new devel-
oped technique of plate-end bars was effective and led to
a similar mode of failure. Specimens of G2-P failed away
of splice and showed rupture of reinforcing bars at the
maximum potential strength of steel. During the test of
G2-P specimens, the movable plates converted the ten-
sion force of bars into compression force acting on the
FIGURE 10 Concrete block formed during failure of G2-HC-10d concrete in front of the plates. This action played a major
role in suppressing the direct tension cracks of concrete
manner of specimens and consequently the modes of fail- during testing especially in short lap splice specimens
ure were affected. For 5 and 10db splice lengths of G2-H, (G2-P-5d and G2-P-10d). These short splice specimens
a unique mode of failure was noticed; direct shear failure remained without cracking till the complete failure.
on the planes formed by the adjacent lapped hooked bars Figure 11 depicts the compression force created on the
as shown in Figure 9. For 20 and 30db splice lengths of front face of bearing plates.
G2-H, the area of this formed shear plane was much big-
ger, thus a different mode of failure was noticed, pullout
of bars followed by crushing of concrete. Using cross bars 4 | ANALY TIC AL STUDY
in G2-HC perpendicular to the adjacent lapped hooked
bars prevented creation of the direct shear plane. This Codes and design provisions consider a uniform bond
behavior caused by the cross bars that intersected the stress distribution along developed/spliced reinforcing
formed shear plane may be attributed to the dowel bars. This may underestimate the bond strength gained
action, which resisted the direct shear. Figure 9a shows by long development/splice lengths. Based on the non-
the front face of G2-H-10d, while the plane of direct shear uniform bond stress distribution for elastic bars, Hwang
from the side view of the marked lower segment is shown et al.24 established an analytical model to predict the
in Figure 9b. Figure 10 shows the failure of G2-HC-10d at maximum tensile stress of spliced/developed bars. In
postultimate level at which a concrete block was formed, order to overcome the yielding of steel bars, Hwang
FAWZY ET AL. 9

qffiffiffiffiffiffi
τu = 0:91 αd f c 0 ðin MPaÞ ð3Þ

qffiffiffiffiffiffiffiffiffiffiffiffiffi
S1 = 0:30 f c 0 =30 ðin MPa and mmÞ ð4Þ

where αd = the bar's diameter coefficient; (1.1 for db

≤19 mm, 1.0 for 22 ≤ db ≤ 29 mm, and 0.9 for db
0
≥ 32 mm); fc = standard cylinder compressive
strength.
According to previous research, the bond stress
affected by the yielding of bars at the loaded end which
known as yield penetration.26 The bond stress at the
damaged region becomes less than that in the case of
elastic bar; consequently, the lab splices with lengths
exceed 20db did not reach a significant gain in capacity
than that of the shorter splices.26 The distribution of bond
FIGURE 12 Simplified bond stress distribution along the bar
length stress is dependent on the equation of deformation,
which relies on the bar's strain. On the other hand, the
tensile strain distribution at the damaged part will be
et al.24 considered that the maximum estimated tensile affected due to the decrease in the modulus of elasticity
strength of a bar does not exceed the yield stress, thus, at the yielded part. In the current proposed model, for
Hwang model cannot be judged in terms of over- simplicity, the stress–strain relationship of steel can be
estimating the bond strength. In the current study, a new considered as a bilinear relationship. In bilinear stress–
proposed model was developed taking into consideration strain relationship, after yielding stress, the value of
the state of the yielded bars. Comparisons were carried young's modulus can be defined as the inclination of the
out between the predicted and the experimental results line between the yielding point and the ultimate strength
of 137 existing splice specimens to validate the proposed point. The young's modulus after yield point (EP) can be
model. In addition, equations of ACI-318 and others assumed as 0.02 of the elastic modulus (ES). Subse-
obtained from previous research were applied to assess quently, the bar's strain relationship in the damaged
the failure characteristics of the short hooked splices region of the bonded length will be affected as shown in
(L <12db at which the adjacent lapped hooked bars form Figure 13.
a direct shear plane) and the plate-end bars. At the plastic stage of the steel bar, the stress, and
strain of the bar are not linearly proportional but the
bilinear stress–strain relationship can be applied.
4.1 | Proposed model for predicting the Thus, the following equations describe the strain at
ultimate tensile stress of straight-ended any point located at distance x along the damaged/
spliced bars undamaged regions. The undamaged zone strain is ε1
and the damaged zone strain is ε2 . For bars that do
Based on the nonuniform bond stress distribution,24 a sim- not reach the tensile yield stress, the young's modulus
plified bond stress distribution along an embedded length of plastic behavior (E P) will be replaced by (E S) in
(L) of a steel bar under tensile pullout force (P) is shown Equation (6).
in Figure 12. Using the ratio (α), the embedded length is
divided into two zones (the loaded/unloaded zones). At 4 τ1 x
ε1 = ) ð0 ! α LÞ ð5Þ
the bar's loaded end (x = α L ! x = L), the relative defor- E s db
mation exceeds the deformation related to the maximum
bond stress, which refers to a local bond damage leading 4 τ1 αL 4 τ2 ðx −αLÞ
ε2 = + ) ðαL ! LÞ ð6Þ
to a reduction in bond stress near to the loaded end. In the E s db Ep db
unloaded embedded part of the bar (x = 0 ! x = α L), as
long as the relative deformation rises, the bond stress As shown in Figure 13, the absolute deformation of
increases. The peak bond strength (τu) and the accompa- reinforcing bar (ΔS) along X axis can be estimated by
nied slip (S1) are defined25 as follows: integrating the strain equations as following:
10 FAWZY ET AL.

At the loaded end, the relative deformation of steel

exceeds the maximum slip (S1) related to the ultimate
bond strength (τu) which indicates a softening behavior
in bond stress, subsequently, the bond stress decreases
along the loaded end. Owing to the effect of the
decreased bond stress, the strain and deformation of con-
crete along the loaded end is almost constant. Therefore,
to simplify the equations, the values of ΔCαL and ΔC was
assumed to be equal. Then, the relative displacement (S)
at the end of the bar (x = L) can be obtained as follows:

L2   L2  
S = S1 + 4α−4α2 τ1 + 2 −4α + 2α2 τ2 ð9Þ
db E s db E p

2 2
Considering C1 = db Es ðL1 − S1 Þ and C2 = db Ep ðL1 − S1 Þ.

The term (τ2) can be expressed as a function of (τ1)

and (τu) as follows:

1 −2 C1 α ð1 −αÞτ1 τu
τ2 = ≥ ð10Þ
1=τu + C2 ð1 −αÞ2 2

The bond stress (τ1) can be expressed as given in

FIGURE 13 Concrete/steel strain distribution along the bar length Equation (11)25:

Sð1  0:40  1:40

ð
αL ðL 1 S τu 1 − Δf=S1
ΔS = Δf + ε1 dx + ε2 dx τ1 = τu ds = ×   ≤ τu
S1 −Δf S1 1:40 1 − Δf=S1
Δf
0 αL
    ð11Þ
4 τ1 x 2 αL 4 τ1 αLx 4 τ2 x 2 4 τ2 α Lx L
= Δf +   + + − 
2 E s d b 0  E s db E p db E p d b α L
L2   L2   From Equation (8), by dividing the equation by S1,
= Δf + 4α −2α2 τ1 + 2 −4α + 2α2 τ2
db E s db E p the term Δf/S1 can be expressed as follows:
ð7Þ
L2  2  ΔC
Δf=S = 1− 2α τ1 + ≥0 ð12Þ
where Δf is the absolute deformation of the steel bar at 1
db S1 E s S1
zero strain (x = 0).
The relative displacement at (x = α L) where the max- Finally, the ultimate tensile stress (fs) of the steel bar
imum bond stress (SαL = S1) can be expressed by sub- under pullout force can be expressed as given in
tracting the absolute deformations of steel (ΔSα L) and Equation (13):
concrete (ΔCα L) as follows:
4L 
Sα L = S1 = ΔSαL −ΔCαL fs = ½ατ1 + ð1 −αÞτ2  ) fy ≤fs ≤fu ð13Þ
ð
αL db
= Δf + ε1 dx − ΔCαL
Where fy and fu are the yield and ultimate strengths
0
  ð8Þ of the embedded steel bar, respectively. For cases at
 4 τ 1 x 2 α L

= Δf +   −ΔCαL which the bars do not attain the tensile yield stress, the
2db E s 0 term “C2” can be replaced by “C1” in Equation (10).
L2  2  Based on the test results collected from several
= Δf + 2α τ1 −ΔCαL
db E s researches, assumptions were made to limit the variations
FAWZY ET AL. 11

between the predicted and the experimental results. Statisti- branches cross the splice zone; fyt = yield strength of
cal analysis was applied on the results of 137 tested speci- transverse reinforcement; s = spacing of transverse rein-
mens in this research and 534 tested specimens regarding forcement; n = number of bars being developed or
to.24 It was found that the length ratio of undamaged part spliced; Rr = relative rib area of the reinforcement.
(α = 0.75) gives the best predicted results comparing to the In the short splices, the bond stress distribution along
available test data. The concrete displacement (Δc) can be the bonded length tends to be uniform and the unit
estimated by the definite integration of the concrete strain length bond stress at failure (Un) achieves an acceptable
along the bonded length, which approximately equal to accuracy comparing to the bond strength (τu). For the
αLεc/2, where εc is the maximum concrete strain at x = 0. application of the proposed model, the value of unit
The concrete strain can be approximately considered as a length bond stress at failure could be utilized instead of
constant value and equals to 0.001.24 τu if it was provided by the experimental results of short
spliced/developed bars.

4.1.1 | Proposed model summary

4.1.2 | Validation of the proposed model
In case of insufficient thickness of cover concrete and/or
the confining reinforcement, the splitting failure takes Although the proposed model was based on anchorage
place prematurely to pullout failure with a decrease in behavior, it can be also applied to several experimental
bond strength. The bond strength calculated by Equa- works containing lap splices. Several reports/codes such
tion (3) was not considering the splitting failure, thus, the fib Bulletin 10,34 Model Code 2010,35 and ACI 408-0336
modification factor of ACI 408R-03 is used to reduce the consider that the anchorages and lap splices have the
bond strength due to splitting failure and lack of trans- same basic behavior. The deference between the anchor-
verse reinforcement. ages and laps was also discussed analytically in previous
research work. Tastani et al.27 studied analytically the
qffiffiffiffiffiffi
ðcw + k t r Þ=db state of bond along the contact lap splices in order to fig-
τu = 0:91 αd f c0 ðN and mmÞ
2:5 ure out the fundamental relationship between the bond
ð14Þ conditions of lap splices versus anchorages. To under-
stand the relationship between the two apparently differ-
ent stress conditions, the mechanics of lap-splices are
0
where αd = the bar's diameter coefficient; fc = stan- explored from first principles. The differential equation
dard cylinder compressive strength; c = cmin + 0.50db; of bond is solved for this problem, using an elastic stress–
w = 0.10(cmax/cmin) + 0.90 ≤ 1.25; strain law for the reinforcement, and a nonlinear local
k tr = At rsnCR ð0:0283db + 0:28Þ ) bond-slip law. According to Tastani et al.,27 the main
ðACI 408R-03Þ ðcw + k tr Þ=db ≤ 4:00 ; CR = 44 + 330 (Rr– derived conclusion is that the confirmation of the empiri-
0.10); cmin = minimum [cs, cb]; csi = 1/2 of the bar clear cal rule given by the code design procedures, namely that
spacing; cso = side concrete cover for reinforcing bar; bond in lap splice regions behaves exactly the same as in
cs = minimum [cso, csi + 0.25 in.]; cb = bottom concrete the anchorage. The proposed model was applied to
cover; cmax = maximum (cb, cs); Atr = area of stirrups 137 existing specimens from previous research works.
The specimens involve widely varied parameters, con-
0
crete cylinder compressive strengths (fc ) between 26 and
110 MPa, steel yield stress between 453 and 555 MPa, rel-
ative rib areas between 0.06 and 0.14, bars' diameters
between 12 and 36 mm and splice lengths between
10 and 35db. The experimental results of the current
research in addition to the results of previous work were
used to validate the proposed model. Eighty-three beams
splice specimens of Darwin et al.,14 70 beams splice speci-
mens of Azizinamimi et al.,28 and three direct tension
splices of the current research were used. Nineteen
beams splice specimens were eliminated from verification
because they failed away of splice due to other reasons.
F I G U R E 1 4 Ultimate tensile predicted/test stress ratio of As shown in Figure 14, the data of the predicted (fs)
beam splice specimens and the tested (ftest) tensile strengths of the 137 splice
12 FAWZY ET AL.

specimens were compared considering the splice length prism. The concrete resisting forces can be classified into
to bar diameter ratio (Ls/db). The average predicted to test two main forces. The first one is formed by the direct ten-
results ratio was about 1.11 with standard deviation of sion strength, which is distributed on the projection of
about 0.151 and coefficient of variance (COV) of about the cross sectional area of specimen. The second one is
.135. In the current research, the unit bond strength was created by the shear resistance of the planes formed by
obtained by the direct tension test of G1-O-05d, it was the adjacent lapped hooks. Figure 15 shows the planes of
considered as 6.71 MPa. The model was applied to three direct shear/tension of specimen G2-H-05d. The dimen-
specimens of sub-group G1-O. The predicted ultimate sions shown in Figure 15 are measured considering the
tensile stress of the bars was found in acceptable limits bars' ribs, which made the 12 mm core diameter to be
compared to the test results. The predicted and test 14 mm in application.
results of specimens of G1-O are listed in Table 6. Figure 15a depicts the planes at which a direct shear
failure occurred, while, Figure 15b shows the projected
plane at which the failure due to direct tension took
4.2 | Hooked-ended splices place. The horizontal projection of the arched planes
shown in Figure 9a can be considered as tension planes.
For the short spliced hooked bars (splice length less than Taking into account, the horizontally projected area of
12 times bar diameter, in the current study), the adjacent the hooked steel bars should be subtracted from the total
lapped hooked bars formed a direct shear plane at which area of the direct tension plane. To predict the ultimate
the failure took place thereafter. After applying the direct strength, the shear, and tension components should be
tension force to the bars, it transfers to the concrete considered. According to ACI 224.2R-92,31 the direct ten-
sion strength (ft) can be estimated based on the cylinder
compressive strength of concrete as follows:

TABLE 6 Bar's predicted/test ultimate tensile stress ratio of qffiffiffiffiffiffiffiffiffiffiffiffi

G1-O specimens f t = gt wc f c 0 ðMPa Þ ð15Þ

Property G1-O-10db G1-O-20db G1-O-30db

τu 6.71 MPa (unit length bond stress Un) where wc = unit weight of concrete (kg/m3);
fc0 = concrete compressive cylinder strength in MPa;
Δf/s1 1.0314 0.9021 0.612
gt = 0.007.
τ1 6.7519 6.5759 6.1410
The direct shear strength (V) of concrete29 can be esti-
C1 0.01 0.04 0.09 mated as follows:
C2 0.010 2.008 4.519
qffiffiffiffi
τ2 6.512 3.355 3.355 0
V = 0:9961 Ak f c ðMPaÞ ð16Þ
fs (MPa) 267.68 461.66 631
ftest (MPa) 265 517 576
Where; Ak = area of the failure plane in mm2;
fs/ftest 1.01 0.89 1.09 0
fc = concrete compressive cylinder strength in MPa.

F I G U R E 1 5 Shear/tension
planes of failure of short hooked
spliced length, specimen G2-H-05d
FAWZY ET AL. 13

For short hooked specimens (G2-H-05d and G2-H- 5 | CONCLUSION

10d), the tensile components were calculated then added
to the shear components. The area exposed to direct ten- The main objective of this paper is to evaluate the ade-
sion was calculated for both specimens as the cross sec- quacy of different techniques used to enhance the perfor-
tional area of the specimens after subtraction of the mance of tension lap splice in NCC. The test specimens
horizontally projected area of the hooked bars. The area consisted of two pairs of spliced bars in NCC prisms with-
of direct tension for both specimens was calculated out confining reinforcement. Three different techniques
as (170 × 180 − 28 × 2 × 100 = 25,000 mm2). The area with four splice lengths were performed to develop
corresponding to the direct shear was a small accurate practical methods for enhancement of splice strength.
area, thus, it should be obtained precisely. To avoid the Based on the experimental findings of the current study
mistakes in calculations of shear area that may be caused as well as the analytical study, the following notes can be
by misfits in the formation of ribbed steel bars and/or drawn.
the relative slips of bars during casting, the actual
shear surface of the failed specimen is lifted then mea- 1. Providing 180 standard hooks of ACI to the end of
sured to obtain a precise shear area. The area exposed to the straight bars led to achieve the nominal yield
direct shear was obtained and it was found to be stress with just 5db splice length, however, the con-
1,540 × 2 mm2 for G2-H-05d and 4,620 × 2 mm2 for ventional splice of G1-O needs at least 20db splice
G2-H-10d. The predicted capacities of specimens G2-H- length to achieve the yield stress.
05d and G2-H-10d was found 83.05 and 128.65 kN, 2. Using 180 standard hooks of ACI enhanced the splice
respectively. The ratios of predicted to test capacities capacity with lengths 10, 20, and 30db by about 95, 11,
were 0.84 and 1.08, respectively. Although the predicted and 1%, respectively, comparing to the straight-ended
capacities strengthen the prospect of that the failure of splices of Group I.
the short spliced hooked bars was induced by direct 3. For short splices of hooked bars (5db and 10db of
shear between the adjacent hooks, these results are too splice length), the adjacent lapped hooked bars cre-
limited and need further assessment in a future research ated a plane at which the direct shear failure was trig-
works. This kind of tension splices has a lack of the gered thereafter.
experimental studies, the hooked bar technique has a 4. In case of 5 and 10db of splice lengths, using cross bars
great potential and needs further studies with a wider perpendicular to the hooks raised the capacity by
variety of parameters. about 9 and 19%, respectively, than those of G2-H and
changed their mode of failure. The cross bars were
perpendicular to the spliced hooked bars intersecting
4.3 | Plate-ended splices with the plane of direct shear, which may resist its
creation by dowel action.
All plate-ended specimens failed away of the connection 5. For longer splices of 20 and 30db lengths, specimens
between the bars and the plate, thus, the failure force of G2-H failed by pullout followed by concrete
should be related to two main components of resisting crushing, subsequently, the enhancement of splice
forces. The first one is the bond strength of the straight strength of G2-HC than that of G2-H was not signifi-
part, while the second one is the bearing force resisted by cant compared to the shorter splices.
the steel plate. Regardless to the bond resistance of 6. Rupture of reinforcing bars was attained by all plate-
straight part, the bearing force resisted by the steel plate end specimens due to the significant capacity of the
can be obtained by using the equations of ACI-318. The bearing force gained by the utilized plates.
bearing force resisted by the plate (0.85 × plate area ×fc0 ) 7. For plate-end specimens, splices of lengths 5 and 10db
was found to be 448.8 kN. The plate's bearing force of remained without noticeable tension cracks during
G2-P is much greater than the maximum tensile strength testing owing to the effect of the plates, which
of the bare tensile bars. The plate's bearing resistance is converted the tension force of bars into compression
in accordance with the experimental results of G2-P at force acting on concrete.
which all specimens failed due to rupture of steel bars 8. Based on the nonuniform stress distribution, a modifi-
away from the splice. This innovative technique is effec- cation of bond stress at the loaded end (τ2) in case of
tive even with short splices, but it still needs further yielded bars was proposed. The proposed model accu-
research works in the future including different parame- racy was verified against 137 existing specimens. The
ters in order to evaluate the impact of the plate's dimen- predicted to test results showed an acceptable
sions and configuration. variation.
14 FAWZY ET AL.

DATA AVAILABILITY STATEMENT 16. Darwin D, Graham K. Effect of deformation height and spacing on
The data that support the findings of this study are available bond strength of reinforcing bars. ACI Struct J. 1993a;90(6):646–657.
from the corresponding author upon reasonable request. 17. Darwin D, Graham K. Effect of deformation height and spacing
on bond strength of reinforcing bars. SL Report 93-1. Lawrence,
Kans: University of Kansas Center for Research, 1993b.
ORCID 18. Issa CA, Nasr A. An experimental study of welded splices of
Hamdy M. Afefy https://orcid.org/0000-0003-0914-8485 reinforcing bars. Build Environ. 2006;41:1394–1405. https://doi.
org/10.1016/j.buildenv.2005.05.025.
R EF E RE N C E S 19. Vella JP, Vollum RL, Jackson A. Investigation of headed bar
1. Darwin D, McCabe L, Idun K, Schoenekase P. Development joints between precast concrete panels. Eng Struct. 2017;138:
length criteria: Bars not confined by transverse reinforcement. 351–366. https://doi.org/10.1016/j.engstruct.2017.02.018.
ACI Struct J. 1992;89(6):709–720. 20. Burkhardt C. Behavior of lap splices in high strength concrete.
2. Esfahani R, Vijaya V. Local bond strength of reinforcing bars [PhD thesis], Aachen, Germany: RWTH Aachen University;
in normal strength and high-strength concrete (HSC). ACI 2000; p. 190 [in German].
Struct J. 1998a;95(2):96–106. 21. Lagier F, Massicotte B, Charron J. Bond strength of tension lap
3. Esfahani R, Vijaya V. Bond between normal strength and high- splice specimens in UHPFRC. Construct Build Mater. 2015;93:
strength concrete (HSC) and reinforcing bars in splices in 84–94. https://doi.org/10.1016/j.conbuildmat.2015.05.009.
beams. ACI Struct J. 1998b;95(3):272–280. 22. Richter B. A new perspective on the tensile strength of lap
4. Orangun O, Jirsa O, Breen E. Reevaluation of test data on splices in reinforced concrete members. [MSc thesis], West
development length and splices. ACI J. 1977;74(3):114–122. Lafayette, USA Purdue University; 2012; p. 165.
5. Tepfers R. A theory of bond applied to overlapped tensile rein- 23. Horszczaruk EK. Hydro-abrasive erosion of high performance
forcement splices for deformed bars. Goteborg, Sweden: Divi- fiber-reinforced concrete. Wear. 2009;267:110–115. https://doi.
sion of Concrete Structures, Chalmers University of org/10.1016/j.wear.2008.11.010.
Technology, 1973;p. 328. 24. Hwang H, Park H, Yi W. Nonuniformbond stress distribution
6. Azizinamini A, Stark M, Roller J, Ghosh K. Bond performance model for evaluation of bar development length. ACI Struct J.
of reinforcing bars embedded in high-strength concrete. ACI 2017;114(4):839–849. https://doi.org/10.14359/51689446.
Struct J. 1993;90(5):554–561. 25. Eligehausen R, Popov P, Bertero V. Local bond stress-slip rela-
7. Azizinamini A, Chisala M, Ghosh K. Tension development tionships of deformed bars under generalized excitations.
length of reinforcing bars embedded in high-strength concrete. Berkeley, CA: Earthquake Engineering Research Council
Eng Struct. 1995;17(7):512–522. Report No. 82/23, University of California; 1983; 169 pp.
8. Aire C, Mendoza C, Davila P. Polypropylene fibers reinforced 26. Tastani S, Pantazopoulou S. Direct tension pullout bond test:
concrete: optimization on plastic shrinkage cracking. Proceed- Experimental results. J Struct Eng. 2010;136(6):731–743.
ings of the Second International Conference on Future Con- https://doi.org/10.1061/(ASCE)ST.1943-541X.0000159.
crete. Abu Dhabi, UAE; 2011. 27. Tastani S, Brokalaki E, Pantazopoulou S. State of bond along
9. Zhang P, Li Q. Effect of polypropylene fiber on durability of lap splices. J Struct Eng. 2015;141(10):04015007. https://doi.
concrete composite containing fly ash and silica fume. Comp org/10.1061/(ASCE)ST.1943-541X.0001243.
Part B. 2013;45:1587–1594. https://doi.org/10.1016/j. 28. Azizinamini A, Pavel R, Hatfield E, Ghosh S. Behavior of lap-
compositesb.2012.10.006. spliced reinforcing bars embedded in high-strength concrete.
10. Saje D, Bandelj B, Šušteršič J, Lopatič J, Saje F. Shrinkage of ACI Struct J. 2000;96(5):826–835.
polypropylene fibre reinforced high performance concrete. 29. Koseki K, Breen JE. Exploratory study of shear strength of
J Mater Civil Eng. 2011;23(7):941–995. https://doi.org/10. joints for precast segmental bridges. Centre for Transportation
1061/(ASCE)MT.1943-5533.0000258. Research, The University of Texas at Austin, Research report
11. Choi W, Jang S, Yun H. Bond and cracking behavior of lap-spliced 248-1; 1983.
reinforcing bars embedded in hybrid fiber reinforced strain- 30. ACI Committee 318. Building Code Requirements for Struc-
hardening cementitious composite (SHCC). Comp Part B. 2017; tural Concrete (ACI 318-19) and Commentary (ACI 318R-18).
108:35–44. https://doi.org/10.1016/j.compositesb.2016.09.086. Farmington Hills, MI: American Concrete Institute; 2019.
12. Marchand P, Baby F, Khadour A, et al. Bond behavior of rein- 31. ACI, American Concrete Institute. Cracking of concrete mem-
forcing bars in UHPFRC. Mater Struct. 2016;49:1979–1995. bers in direct tension. ACI 224.2R-92, Farmington Hills, MI;
https://doi.org/10.1617/s11527-015-0628-0. 2004.
13. Metelli G, Marchina E, Plizzari GA. Experimental study on 32. ASTM. Standard test method: compressive strength of Cylindri-
staggered lapped bars in fiber reinforced concrete beams. Comp cal Concrete Specimens, C39/C39M. West Conshohocken, PA:
Struct. 2017;179:655–664. https://doi.org/10.1016/j.compstruct. ASTM; 2010a.
2017.07.069. 33. ASTM. Standard test methods and definitions for mechanical
14. Darwin D, Tholen M, Idun E, Zuo J. Splice strength of high rel- testing of steel products—Annex A9 methods for testing steel
ative rib area reinforcing bars. ACI Struct J. 1996;93(1):95–107. reinforcing bars. A370. West Conshohocken, PA: ASTM;
https://doi.org/10.1016/j.aej.2014.01.009. 2010b.
15. Eligehausen R. Bond in tensile lapped splices of ribbed bars 34. International Federation for Concrete. “Bond of reinforcement
with straight anchorages. Publication 301. Berlin: German in concrete.” Bulletin No. 10, Report by former CEB Task
Institute for Reinforced Concrete, 1979. Group 2.5, Lausanne, Switzerland; 2000.
FAWZY ET AL. 15

35. International Federation for Concrete. “fib model code 2010 –

Final draft.” Bulletin No. 65, Lausanne, Switzerland; 2010. Ahmed Atta, Professor of Concrete
36. Joint ACI-ASCE Committee 408. Bond and Development of Structures, Department of Structural
Straight Reinforcing Bars in Tension (ACI 408R-03). American Engineering, Faculty of Engineering,
Concrete Institute, Farmington Hills, MI; 2003; 49 pp. Tanta University, Tanta, Egypt.
Email: [email protected].
edu.eg
AUTHOR BIOGRAPHIES

Tarek Fawzy, Professor of Concrete

Structures, Department of Structural Hamdy M. Afefy, Professor of Con-
Engineering, Faculty of Engineering, crete Structures, Construction Engi-
Tanta University, Tanta, Egypt. neering and Management
Email: [email protected]. Department, Faculty of Engineering,
edu.eg Pharos University, Alexandria. On
leave from Tanta University, Egypt.
Email: [email protected];
[email protected]
Abdel-Hakim Khalil, Professor of
Concrete Structures, Department of Mohamed Ellithy, Assistant Lec-
Structural Engineering, Faculty of turer, Department of Structural
Engineering, Tanta University, Tanta, Engineering, Faculty of Engineer-
Egypt. ing, Tanta University, Tanta, Egypt.
Email: abdelhakim.khalil@f-eng. Email: mohamed.ellithy@f-eng.
tanta.edu.eg tanta.edu.eg

How to cite this article: Fawzy T, Khalil A-H,

Atta A, Afefy HM, Ellithy M. Experimental and
analytical study on tension lap splices in
nonconventional concrete using different
techniques. Structural Concrete. 2020;1–15. https://
doi.org/10.1002/suco.202000403