engineering structures - universiti teknologi malaysia · tensile capacity of grouted splice...

Engineering Structures 111 (2016) 285–296

Contents lists available at ScienceDirect

Engineering Structures

journal homepage: www.elsevier .com/ locate /engstruct

Tensile capacity of grouted splice sleeves

http://dx.doi.org/10.1016/j.engstruct.2015.12.0230141-0296/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +60 128072616.E-mail addresses: [email protected] (J.H. Ling), [email protected]

(A.B. Abd. Rahman), [email protected] (I.S. Ibrahim), [email protected](Z. Abdul Hamid).

Jen Hua Ling a,⇑, Ahmad Baharuddin Abd. Rahman b, Izni Syahrizal Ibrahim b, Zuhairi Abdul Hamid c

a School of Engineering and Technology, University College of Technology Sarawak, 96000 Sibu, Sarawak, Malaysiab Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor Darul Ta’zim, MalaysiacConstruction Research Institute of Malaysia, Jalan Chan Sow Lin, 55200 Kuala Lumpur, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history:Received 2 June 2015Revised 15 December 2015Accepted 17 December 2015

Keywords:Grouted spliceSleeveConnectionConfinement

This study tests grouted splices connected by two types of sleeves, namely Welded Bar Sleeve (WBS) andTapered Head Sleeve (THS). These sleeves are made from non-proprietary pipe sections, where (a) WBS isfabricated by welding the deformed bars to the inner wall of the pipe, and (b) THS is made tapered withsmaller openings at both ends. To study the behavior, the splice specimens were tested under incremen-tal tensile load at various bar embedded lengths and sleeve diameters. The degree of confinement gen-erated in the sleeve is found to increase with decreasing sleeve diameter. This improves the bondstrength in sleeve, which subsequently increases the tensile capacity of the splice. THS gives a 30% highertensile capacity compared with WBS. With the active confinement, the required bar embedded length ofthe splice can be reduced to 8 times the bar diameter. An analytical model is formulated on the basis ofthe confinement stress as expressed in a function of sleeve dimensions. The model is used to predict thetensile capacities of the splices at a variation range of ±10% of the experimental results. This verifies thecorrelations among the sleeve dimensions, the confinement stress and bond strength of the groutedsplice.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction The idea of using non-proprietary pipes to splice steel bars was

Steel bars are usually adjacently lapped in reinforced concretestructures. This method requires long lapping lengths, and it isnot practical for connecting precast concrete elements. Due tothe simplicity of connecting and short bar embedded length,mechanical and welded splices are commonly used for precast con-crete structure [1–4].

There are many mechanical splices [5–7]. Some of these splicesmay not be easily acquired in certain countries as it is not common.This results in a longer duration and additional cost to acquirethem. Additionally, some of the splices require high precision toensure the alignment of the spliced bars due to tight tolerances.

Using non-proprietary pipes to splice steel bars could be a solu-tion in cases of unavailability or when it is too costly to acquirethose proprietary splices. With knowledge of the design, steel barscould be easily spliced by using pipe sections, of course, with slightmodifications. Also, the designer could easily choose suitable sizesof pipe to fit the tolerance requirements.

proposed by Einea et al. [8] in 1995. Kim [9] adopted the idea andused pipe sections as the beam-column connection. Since then,researchers have developed grouted splices using various non-proprietary materials, such as mild steel pipes [10–14], corrugatedaluminum sleeves [15], spirals [16], square hollow sections [17],and glass fiber reinforced polymers [18–21].

These non-proprietary splices are called grouted splice which isa type of mechanical splice consisting of a sleeve, some grout andtwo spliced bars. The steel bars are spliced and bonded by grout inthe sleeve. The grout mechanically interlocks with the ribs on thebars to prevent the bars from slipping out of the sleeve [11,22–24]. The sleeve confines the grout to increase the bond strength,and subsequently, reduces the required anchorage length of thebars in it [8,13,25,26].

This study intends to develop grouted splices by using pipe sec-tions. While developing grouted splices, the following isconsidered:

� For the splices which rely on gripping mechanism to preventthe slippage of the spliced bars, gripping-slip usually occurs[27]. The bar slips suddenly under load and it is not recoverable.To avoid this, steel bars should be bonded by using grout.

� When steel bars are adjacently spliced, the load eccentricitycauses the connector to inevitably self-align and rotate [28].

Nomenclature

SymbolAc,b contact surface area of spliced bar, mm2

Ac,sl contact surface area of sleeve with grout, mm2

Esl modulus of elasticity of sleeve, N/mm2

Fn confinement force acting on grout, kNPb bond strength of grouted splice, kNPb,ths bond strength of THS, kNPb,wbs bond strength of WBS, kNPu tensile capacity of grouted splice specimen, kNPu,avg average tensile capacity of grouted splice specimen, kNPu,exp experimental tensile capacity of grouted splice speci-

men, kNPu,pre predicted tensile capacity of grouted splice specimen,

kNRr reliability ratioTb tensile capacity of spliced bar, kNTt,sl transverse tensile force of sleeve, kNdb diameter of bar embedded in the sleeve, mm

dsi inner diameter of sleeve, mmdse outer diameter of sleeve, mmdwb diameter of bar welded to WBS, mmft,sl transverse tensile stress of sleeve, N/mm2

fu,g ultimate compressive stress of grout, N/mm2

fy nominal yield stress of spliced bar, N/mm2

lb bar embedded length in sleeve, mmlsl length of sleeve, mms1 standard deviation of tensile capacity of grouted splice

specimens2 standard deviation of displacement of bartsl thickness of sleeve, mma tapered angle of THS, �du displacement of bar at failure, mmet,sl transverse tensile strain of sleeve, mm/mmub bond stress acting on spliced bar, N/mm2

un,b confinement stress acting on spliced bar, N/mm2

un normal confinement stress in sleeve, N/mm2

(a) Welded Bar Sleeve (WBS)

lsl

lb

db

dwb

ts

dsi

Spliced bar

Sleeve Welded bar

Grout

(b) Tapered Head Sleeve (THS)

lb

ls1

ls2

dsidse

db

tsl

Spliced bar

Grout

Sleeve

Fig. 1. Schematic design the grouted splice specimens.

286 J.H. Ling et al. / Engineering Structures 111 (2016) 285–296

This deformation generates undesired stress at the rotatingpoint of the spliced bars, and therefore, hinges are formed atthe spliced bars. For this, steel bars should be aligned end-to-end without eccentricity.

� Some splices require the spliced bars to be threaded for the pur-poses of installation. However, threading causes damage to thebar. It degrades the bond and the tensile capacity of the splicedbar [11,23,29]. Thus, it should be avoided.

Grouted splices are usually tested under incremental tensileload [8,14,17,29]. Based on ACI-318 [30] and AC-133 [31], the ten-sile capacity should be at least 125% of the nominal yield strengthof the spliced bars. This sets the evaluation criteria to determinethe feasibility of the splices.

This paper fills the gap of knowledge in the study of groutedsplices. Some researchers performed regression analysis to predictthe response of the splice under tensile load [12]. The effect of theconfinement stress is not incorporated in the model, although it isgenerally agreed to increase the bond strength [20,25,32–34].Some studies successfully correlate the bond stress with the con-finement stress [8,26]. However, it is still a distance away fromobtaining the tensile capacity of a grouted splice, as the capacitymay not always be governed by the bond strength.

This paper presents the experimental results of two newgrouted splices. The performance of the splice under tensile loadis evaluated with respect to two parameters, the bar embeddedlength and the sleeve diameter. An analytical model is formulatedto predict the tensile capacity on the basis of (a) the approach usedto measure the confinement stress in the sleeve, which is proposedby Einea et al. [8], and (b) the model correlating the bond stresswith the confinement stress, as demonstrated by Untrauer andHenry [26].

Studies show that the degree of confinement generated in thesleeve is somehow governed by the dimensions and the configura-tions of the sleeve [8,11,29]. Hence, the equations for predictingthe tensile capacity are expressed as a function of the sleevedimensions. This is done by correlating the stress parameter withthe dimension parameters in the equations. To justify this, theequations are then verified by the experimental results.

As an overview of this paper, the test specimens and the exper-imental program are described in Section 2. The experimentalresults are presented and discussed in Section 3. Section 4 presentsthe formulation of an analytical model to predict the tensile capac-ity of the grouted splice. The model is then evaluated for reliability

in Section 5. Finally, in Section 6, the conclusions of the study arepresented.

2. Test methodology

2.1. Test specimens

This study tests two grouted splices, namely Welded Bar Sleeve(WBS) and Tapered Head Sleeve (THS) (Fig. 1). The sleeves aremade from mild steel pipes with the inner diameters, dsi, of50 mm, 65 mm and 75 mm (Table 1). Steel bars, with the nominalyield strength of 500 N/mm2 and the nominal diameter of 16 mm,are spliced at the embedded lengths, lb, of 75 mm, 125 mm and175 mm (Table 1).

Welded Bar Sleeves (WBS) are fabricated by welding four steelbars (nominal yield strength of 500 N/mm2 and bar diameter of10 mm) to the inner surface of the pipes (nominal yield strengthof 250 N/mm2). The ribs on these welded bars interlock with thegrout to prevent the grout from slipping out of the sleeve.

Tapered Head Sleeves (THS) are made tapered with small open-ings at both ends of the pipe (dse = 35 mm). The space volume

Table 1The dimensions of Welded Bar Sleeve (WBS) and Tapered Head Sleeve (THS).

Specimens Units Spliced bar Sleeve

db(mm)

lb(mm)

dsi(mm)

dse(mm)

lsl(mm)

tsl(mm)

dwb

(mm)

WBS-1 3 16 75 50 – 150 4.5 10WBS-2 3 16 75 65 – 150 4.5 10WBS-3 3 16 75 75 – 150 4.5 10WBS-4 3 16 125 50 – 250 4.5 10WBS-5 3 16 125 65 – 250 4.5 10WBS-6 3 16 125 75 – 250 4.5 10WBS-7 3 16 175 50 – 350 4.5 10WBS-8 3 16 175 65 – 350 4.5 10WBS-9 3 16 175 75 – 350 4.5 10

THS-1 3 16 75 50 35 360 4.5 –THS-2 3 16 75 65 35 360 4.5 –THS-3 3 16 75 75 35 360 4.5 –THS-4 3 16 125 50 35 360 4.5 –THS-5 3 16 125 65 35 360 4.5 –THS-6 3 16 125 75 35 360 4.5 –THS-7 3 16 175 50 35 360 4.5 –THS-8 3 16 175 65 35 360 4.5 –THS-9 3 16 175 75 35 360 4.5 –

Data logger

Computer SG 2

Actuator

Pulling force

Reaction forces

SG 3

SG 1

Fig. 2. Instrumental setup for tensile load test.

J.H. Ling et al. / Engineering Structures 111 (2016) 285–296 287

decreases as the grout slips toward the ends of the sleeve as thetensile load increases. This generates excessive confinement stressto improve the bond in the sleeve.

Non-shrink grout (Sika Grout-215), with the nominal strength of70 N/mm2 at day 28, is used as the bonding material in the sleeve.It is mixed into a pour-able state, according to the proportions rec-ommended by the manufacturer (4 l of water: 25 kg of grout). Thegrout is poured into the sleeve prior to insertion of the steel bar.

Table 2 presents the mass, cost and tolerance of the test speci-mens. The cost of a splice sleeve ranges from RM3.79 to RM15.31and the tolerance ranges from 14 mm to 39 mm.

2.2. Test setup

Three strain gauges (SG) are installed on each specimen (Fig. 2);

i. SG1 is installed on the spliced bar at about one bar diameterfrom the surface of the grout. It is used to measure the elon-gation of the bar under the tensile load [35].

Table 2The mass, cost and tolerance of the test specimens.

Specimens Volume Mass Co

Sleeve (mm3) Grout (mm3) Sleeve (kg/unit) Grout (kg/unit) Sle

WBS-1 102,524 217,241 0.805 0.478 3WBS-2 118,428 420,463 0.93 0.925 3WBS-3 129,031 585,397 1.013 1.288 4WBS-4 170,873 362,069 1.341 0.797 5WBS-5 197,380 700,772 1.549 1.542 6WBS-6 215,052 975,661 1.688 2.146 6WBS-7 239,222 506,896 1.878 1.115 7WBS-8 276,333 981,080 2.169 2.158 8WBS-9 301,073 1,365,925 2.363 3.005 9THS-1 224,699 136,738 1.764 0.301 7THS-2 347,798 362,540 2.73 0.798 10THS-3 441,645 545,799 3.467 1.201 13THS-4 224,699 116,632 1.764 0.257 7THS-5 347,798 342,434 2.73 0.753 10THS-6 441,645 525,693 3.467 1.157 13THS-7 224,699 96,525 1.764 0.212 7THS-8 347,798 322,327 2.73 0.709 10THS-9 441,645 505,587 3.467 1.112 13

Notes: The mass is computed from the following values:(a) Densities of steel and grout are 7850 kg/m3 and 2200 kg/m3, respectively.(b) Costs of steel and grout are approximately RM4.00 and RM1.20, respectively.

ii. SG2 is installed transversely on the sleeve at the mid lengthof the bar embedded length. It is used to measure the defor-mations of the sleeve due to splitting expansion of the groutin the sleeve [8,10].

iii. SG3 is installed longitudinally at the mid length of thesleeve. It is used to measure the longitudinal elongation ofthe sleeve under the tensile load.

The specimens are tested with an incremental tensile load of0.5 kN/s. The displacements are measured by using the built-inLVDT of the hydraulic actuator. It is assumed that (a) the bond–slipbetween the bar and the grout, (b) the gripping slip between thegrip and the bar, (c) the bond–slip between the bar and the sleeveand (d) the tensile elongation of the sleeve are negligible [11].Thus, the displacement measured is being considered as a resultof the elongation of the spliced bars.

3. Experimental results

The grouted splice specimens are tested at about 28 days aftergrouting. The grout strengths are 76.75 N/mm2 and 68.68 N/mm2

for WBS and THS, respectively.

st Tolerance (mm)

eve (RM/unit) Grout (RM/unit) Total cost (RM/specimen)

.22 0.57 3.79 14

.72 1.11 4.83 29

.05 1.55 5.60 39

.36 0.96 6.32 14

.2 1.85 8.05 29

.75 2.58 9.33 39

.51 1.34 8.85 14

.68 2.59 11.27 29

.45 3.61 13.06 39

.06 0.36 7.42 19

.92 0.96 11.88 19

.87 1.44 15.31 19

.06 0.31 7.37 19

.92 0.9 11.82 19

.87 1.39 15.26 19

.06 0.25 7.31 19

.92 0.85 11.77 19

.87 1.33 15.2 19

Table 3Test results of grouted splice specimens (averages of three specimens).

Specimen Tensile capacity, Pu,avg (kN) Standard deviationof tensile capacity, s1

Displacementat failure, du (mm)

Standard deviation ofdisplacement, s2 (mm)

Failure mode

Welded Bar Sleeve (WBS) WBS-1 86.1 3.03 3.5 0.65 Bar bond–slipWBS-2 79.0 3.75 2.3 0.56 Bar bond–slipWBS-3 75.1 1.58 2.9 0.45 Bar bond–slipWBS-4 127.9 6.98 33.8 4.46 Bar fractureWBS-5 128.0 6.78 29.4 10.34 Bar bond–slipWBS-6 122.4 6.46 29.2 16.70 Bar bond–slipWBS-7 133.5 2.23 26.6 0.17 Bar fractureWBS-8 129.1 7.09 30.6 0.99 Bar fractureWBS-9 133.4 3.72 31.4 0.51 Bar fracture

Tapered Head Sleeve (THS) THS-1 112.2 5.59 3.6 1.31 Bar bond–slipTHS-2 102.1 4.94 3.7 0.40 Bar bond–slipTHS-3 96.1 6.13 3.4 0.38 Bar bond–slipTHS-4 137.0 3.27 30.8 1.96 Bar fractureTHS-5 135.4 2.34 30.3 0.59 Bar fractureTHS-6 134.6 0.29 30.2 0.94 Bar fractureTHS-7 137.7 3.40 25.9 1.13 Bar fractureTHS-8 133.2 0.53 29.2 2.13 Bar fractureTHS-9 135.5 0.71 26.5 1.11 Bar fracture


Table 3 presents the average values of the three specimens withthe same configurations. These values are used (a) for evaluatingthe performance of grouted splices and (b) for justifying the relia-bility of the formulated model.

From Table 3, the following is found:

i. THS generally performs better than WBS in terms of tensilecapacity.

ii. The tensile capacity increases as the bar embedded lengthincreases.

iii. The tensile capacity increases as the sleeve diameterdecreases.

The grouted splices fail in two modes, bar fracture and bond–slip failures (Fig. 3), and give two different load–displacementresponses (Fig. 4). The spliced bar generally fractures after yielding,or slips before or while yielding.

The specimens show a close-to-elastic response at the initialstage (Fig. 4). Although it is insignificant, the stiffness progressivelydegrades due to the development of the internal micro-cracks(Fig. 5) [11]. As the steel bar yields, significant elongation of thebar takes place and plastic response is observed. The steel bareventually fractures as the tensile capacity is achieved.

Based on the failure modes:

(a) Bar bond-slip failure (WBS-5)

The grout keys that interlocked with the bar ribs failed, and thus, the bar slipped

out of the sleeve.

Fig. 3. Typical failure of the g

i. The tensile capacity of the grouted splices is governed by (a)the tensile capacity of the splice bars, and (b) the bondcapacity between the bar and the grout, which is weaker.

ii. In comparison with the bond–slip failure, the specimenswhich fail by bar fracture generally (a) give a higher tensilecapacity and (b) offer a higher degree of ductility.

Figs. 6 and 7 show the effects of the sleeve diameter and the barembedded length. The grouted splices are considered adequatewhen the tensile capacities exceed 125.7 kN, which is equivalentto 1.25fy (as required by ACI-318 and AC-133).

Fig. 6 shows that:

i. The bar embedded length of 125 mm (about 8 times the bardiameter) is adequate for both WBS and THS. It is about 8times the bar diameter.

ii. The tensile capacity increases as the sleeve diameter of WBSand THS decreases. The response is linearly proportional.

iii. The effect of the changing diameter of THS is more signifi-cant than WBS, as observed from the gradient of the curves.

iv. THS gives about 30% higher tensile capacity than WBS withthe same sleeve diameter when the bar embedded length isinsufficient.

(b) Bar fracture failure (THS-6)

The spliced bars yielded,

elongated, and underwent the

necking process before rupture.

routed splice specimens.

0

20

40

60

80

100

120

140

Load

(kN

)

Displacement (mm)

(a) Bar bond slip failure (specimen WBS-1)

The spliced bar slipped before achieving its yield point

0

20

40

60

80

100

120

140

0 5 10 15 20 25 30 0 5 10 15 20 25 30

Load

(kN

)

Displacement (mm)

(b) Bar fracture failure (specimen THS-8)

Specimens with bar fracture failure offer a higher tensile

capacity The spliced bars yield and

elongate significantly before fracture

Fig. 4. Typical load–displacement response in accordance to failure modes.

Micro-cracks

Pulling force

Sleeve

Spliced bar

Grout

Fig. 5. Propagation of internal cracks in grout (modified from Yankelevsky [36]).


From Fig. 7,

i. The tensile capacities of both WBS and THS increase as thebar embedded length increases.

ii. With the same bar embedded length, THS offers a highertensile capacity than WBS.

iii. The effect of the bar embedded length in THS is more signif-icant than WBS, as observed from the gradient of the curves.

iv. THS requires a shorter bar embedded length to achieve 1.25fycompared with WBS. The interception between the curveswith the line 1.25fy shows that the required bar embeddedlength for THS and WBS are at least 100 mm and 120 mm,respectively (about 6–8 times the bar diameter).

60

70

80

90

100

110

120

130

140

40 45 50 55 60 65 70 75 80

Tens

ile c

apac

ity (k

N)

Sleeve diameter (mm)

1.25fy

lb = 75 mm

lb = 175 mm

lb = 125 mm


THS gives a higher tensile capacity than WBS under the

same sleeve diameter.

Fig. 6. Tensile capacity ve

THS generates bond strength more efficiently than WBSbecause (as discussed in the previous study [12]):

i. The inclined wall of THS generates a uniform compressivefield around the spliced bars, while WBS is unable to gener-ate the compressive field.

ii. The entire sleeve of THS contributes to generating confine-ment stress in the sleeve, while WBS only relies on the ribson the welded bars.

iii. The degree of confinement generated in the sleeve increasesas the grout stretches and slips toward the ends of thesleeve.

60

70

80

90

100

110

120

130

140

Tens

ile c

apac

ity (k

N)

Sleeve diameter (mm)40 45 50 55 60 65 70 75 80

1.25fy

lb = 75 mm

lb = 125 mm

lb = 175 mm


rsus sleeve diameter.

60

70

80

90

100

110

120

130

140

Tens

ile c

apac

ity (k

N)

Bar embedded length (mm)

60

70

80

90

100

110

120

130

140

40 60 80 100 120 140 160 18040 60 80 100 120 140 160 180

Tens

ile c

apac

ity (k

N)

Bar embedded length (mm)


1.25fy

dsi = 75 mm dsi = 65 mm dsi = 50 mm

dsi = 75 mm dsi = 65 mm dsi = 50 mm

1.25fy


With the same bar embedded length, THS gives a higher tensile capacity than WBS

Fig. 7. Tensile capacity versus bar embedded length.


iv. THS generates active confinement as the grout slips towardthe narrower ends of the sleeve. WBS generates passive con-finement, and responds only upon the propagation of split-ting cracks.

Figs. 8–10 present the typical strain responses of the splicedbars and sleeve with respect to the load and strain, representingthe categories of (a) WBS bond–slip failure, (b) WBS bar fracturefailure, (c) THS bond–slip failure, and (d) THS bar fracture failure.

The spliced bars behave similarly regardless of the type ofgrouted splice. The bars deform elastically prior to yielding atabout 500 N/mm2. Then, plastic response takes place where thesteel bar elongated significantly and eventually fractured (Fig. 8).

The transverse strain of the sleeve indicates the confinementcondition in the sleeve. A linear compressive strain develops atthe early stage (Fig. 9(c)). This is due to longitudinal stretching ofthe grouted splice under load. Then, compressive strain stopsdeveloping and transforms into tensile strain (Fig. 9(d)). Underhigh tensile load, splitting cracks develop around the spliced bars[37]. The sleeve controls the radial expansion of the grout due tosplitting cracks and thus transverse tensile strain develops. At thatmoment, the grout in the sleeve is under confinement. This con-finement state ensures the grout keys interlock with the bar ribseffectively, and prevents degradation of the bond performancedue to splitting cracks [11,23].

Under tensile load, the grout in THS breaks and slides towardthe ends of the sleeve. Thus, a sudden increment of the longitudi-nal strain of the sleeve occurs (Fig. 10(c) and (d)) [38]. The longitu-dinal strain in WBS remains linear throughout the test as the groutdoes not break and the sleeve remains elastic.

4. Analytical model

The analytical model is formulated to predict the tensile capac-ities of the grouted splices. Fig. 11 illustrates the relationshipbetween the transverse tensile force, Tt,sl, and the confining stress,un, in the sleeve [8]. This model is built to suit both WBS and THS,neglecting the effects of the welded bars on WBS to the confine-ment stress in the sleeve.

The transverse tensile stress of the sleeve, ft,sl, is a function ofthe strain, et,sl, and the modulus of elasticity, Esl.

f t;sl ¼ et;slEsl ð1ÞIt is also governed by the transverse cross sectional area of the

sleeve,

f t;sl ¼Tt;sl

tsllbð2Þ

where tsl is the thickness of the sleeve, and lb is the bar embeddedlength.

Thus,

Tt;sl ¼ et;sllbtslEsl ð3ÞReferring to Fig. 11(a), the principle of static equilibrium pro-

poses that

2Tt;sl ¼ undsilb ð4ÞTherefore, the normal confinement stress, un, is expressed as

un ¼ 2et;sltslEsl

dsið5Þ

The confinement force acting on the grout, Fn, is the multiplica-tion of the confinement stress, un, with the inner surface area of thesleeve, Ac,sl.

Fn ¼ unAc;sl ð6Þ

where Ac;sl ¼ pdsilb for WBS; and ð7Þ

Ac;sl ¼ plb2 cosa

d2si � dsidse þ d2

se

dsi � dse

!for THS ð8Þ

The confinement stress acting on the bar, un,b, is the confine-ment force, Fn, per unit contact surface area of the embedded bars.

un;b ¼ dsiun

dbfor WBS; and ð9Þ

un;b ¼ un

2db cosad2si � dsidse þ d2

se

dsi � dse

!for THS ð10Þ

Untrauer and Henry [26] claimed that the bond strengthbetween the steel and concrete increases in proportion to thesquare root of normal pressure and concrete strength. The correla-tion between the bond stress, ub, and the confinement stress, un,b, isas expressed in Eq. (11).

ub ¼ ðAþ Bffiffiffiffiffiffiffiffiun;b

p Þffiffiffiffiffiffiffif u;g

qð11Þ

where A and B are constant values obtained from the experimentalresults (Fig. 12), which are 0.223 and 1.578, respectively. fu,g = ulti-mate compressive strength of grout.

0

100

200

300

400

500

600

700

Stre

ss (N

/mm

2 )

strain x10-6 (mm/mm)

THS 8A

THS 8B

THS 8C

0

100

200

300

400

500

600

700

Stre

ss (N

/mm

2 )


THS 2A

THS 2B

THS 2C

(d) Specimen THS-8 (c) Specimen THS-2

Spliced bar

SG1

0

100

200

300

400

500

600

700

Stre

ss (N

/mm

2 )

Strain x10-6 mm/mm

WBS 7A

WBS 7B

WBS 7C

0

100

200

300

400

500

600

700

0 2000 4000 6000 8000 100000 2000 4000 6000 8000 10000

0 2000 4000 6000 8000 100000 2000 4000 6000 8000 10000

Stre

ss (N

/mm

2 )

Strain x10-6 mm/mm

WBS 2AWBS 2BWBS 2C

Spliced bar

SG1

Linear response

(a) Specimen WBS-2 (b) Specimen WBS-7

Linear response with the same gradient

Fig. 8. Stress–strain responses of the spliced bars (SG1).

0

20

40

60

80

100

120

140

-100 0 100 200 300 400

Load

(kN

)


THS 2A

THS 2B

THS 2C0

20

40

60

80

100

120

140

-100 0 100 200 300 400

Load

(kN

)


THS 8A

THS 8B

THS 8C

0

20

40

60

80

100

120

140

-100 0 100 200 300 400

Load

(kN

)

Strain x10-6 mm/mm

WBS 2A

WBS 2B

WBS 2C

0

20

40

60

80

100

120

140

-100 0 100 200 300 400

Load

(kN

)

Strain x10-6 mm/mm

WBS 7AWBS 7BWBS 7C


Spliced bar

SG2

Spliced bar

SG2

(c) Specimen THS-2 (d) Specimen THS-8

Linear transverse tensile strain

The development of compressive strain

stopped and transformed into tensile strain

Fig. 9. Transverse strain responses of sleeve (SG2).


0

20

40

60

80

100

120

140

0 200 400 600 800

Load

(kN

)


THS 8A

THS 8B

THS 8C

0

20

40

60

80

100

120

140

0 200 400 600 800

Load

(kN

)

Strain x10-6 mm/mm

WBS 2A

WBS 2B

WBS 2C

0

20

40

60

80

100

120

140

0 200 400 600 800

Load

(kN

)

Strain x10-6 mm/mm

WBS 7A

WBS 7B

WBS 7C

0

20

40

60

80

100

120

140

0 100 200 300 400 500 600

Load

(kN

)


THS 2A

THS 2B

THS 2C


(c) Specimen THS-2 (d) Specimen THS-8

Spliced bar

SG3

Spliced bar

SG3

Linear strain response

Linear strain response

Sudden increment of strain

Sudden increment of strain

Fig. 10. Longitudinal strain responses of sleeve (SG3).

un

un

un

Tt,sl Tt,slTt,slTt,sl un

(a) Grouted splice (b) Grout (c) Sleeve

Fig. 11. Free body diagram of stress in a grouted splice [8].


The bond strength of the grouted splice, Pb, can be predicted bymultiplying the bond stress, ub, with the contact area of theembedding bar.

Pb ¼ pdblbub ð12Þ

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0 2.0 4.0 6.0 8.0

un,b

u b/f u,g

578.1223.0 ,,

+= bngu

b ufu

Fig. 12. Relationship of ub/pfu,g versus

pun,b.

Eq. (13) is obtained by substituting Eq. (11) into Eq. (12).

Pb ¼ pdblbffiffiffiffiffiffiffif u;g

qð0:223 ffiffiffiffiffiffiffiffi

un;bp þ 1:578Þ ð13Þ

The dimension of the sleeve somehow influences the confine-ment stress in the sleeve, as (a) indicated by the experimentalresults and (b) reflected through the equation derivation process.For this, the confinement stress, un,b, is substituted by the dimen-sion parameters of the grouted splice, as expressed in Eqs. (14)and (15).

Pb;wbs ¼ plbffiffiffiffiffiffiffiffiffiffiffiffidbf u;g

q0:223

ffiffiffiffiffiffiffiffiffiffidsiun

pþ 1:578

ffiffiffiffiffidb

p� �ð14Þ

Pb;ths ¼plb

ffiffiffiffiffiffiffiffiffiffiffiffidbf u;g

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffifficosaðdsi � dseÞ

p 0:158ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiunðd2

si � dsidse þ d2seÞ

q�

þ1:578ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffidb cosaðdsi � dseÞ

q �ð15Þ

Substituting Eq. (5) into Eqs. (14) and (15), yields

Table4

Theco

mpu

tation

ofthepred

ictedou

tcom

esin

compa

riso

nwiththeex

perimen

talresu

lts.

Specim

enCom

putation

ofbo

ndan

dtensile

capa

cities

ofgrou

tedsp

lice

connection

Com

pariso

nof

thepred

ictedan

dex

perimen

talresu

lts

e t,sl(le)

T t,sl(kN)

[Eq.

(3)]

u n(N

/mm

2)

[Eq.

(5)]

u n,b(N

/mm

2)

[Eq.

(9)]

u b(N

/mm

2)

[Eq.

(11)]

P b(kN)

[Eq.

(16)]

T b(kN)

[Eq.

(21)]

P u,pre(kN)a

P u,exp

(kN)(Tab

le3)

Rr [Eq.

(22)]

Pred

icted

Failure

Mod

ebActual

Failure

Mod

e(Tab

le3)

Rem

arks

c

WBS

-111

37.8

4.18

13.06

4.17

78.8

135.9

78.8

86.1

1.09

Bar

bond–

slip

Bar

bond–

slip

pW

BS-2

140

9.7

3.98

16.17

3.97

81.8

135.9

81.8

79.0

0.97

Bar

bond–

slip

Bar

bond–

slip

pW

BS-3

976.7

2.39

11.20

2.39

76.8

135.9

76.8

75.1

0.98

Bar

bond–

slip

Bar

bond–

slip

pW

BS-4

617.1

2.27

7.09

2.29

119.6

135.9

119.6

127.9

1.07

Bar

bond–

slip

Bar

fracture

XW

BS-5

139

16.1

3.96

16.09

3.97

136.2

135.9

135.9

128.0

0.94

Bar

fracture

Bar

bond–

slip

XW

BS-6

123

14.2

3.03

14.20

3.03

133.2

135.9

133.2

122.4

0.92

Bar

bond–

slip

Bar

bond–

slip

pW

BS-7

30.5

0.11

0.34

0.11

131.7

135.9

131.7

133.5

1.01

Bar

bond–

slip

Bar

fracture

XW

BS-8

111.8

0.32

1.30

0.31

141.3

135.9

135.9

129.1

0.95

Bar

fracture

Bar

fracture

pW

BS-9

223.6

0.55

2.58

0.54

149.2

135.9

135.9

133.4

0.98

Bar

fracture

Bar

fracture

p

[Eq.

(10)]

[Eq.

(17)]

THS-1

176

12.2

6.49

26.73

6.49

81.3

135.9

81.3

112.2

1.38

Bar

bond–

slip

Bar

bond–

slip

pTH

S-2

324

22.4

9.20

30.53

9.2

92.6

135.9

92.6

102.1

1.10

Bar

bond–

slip

Bar

bond–

slip

pTH

S-3

181

12.5

4.46

14.82

4.45

81.8

135.9

81.8

96.1

1.17

Bar

bond–

slip

Bar

bond–

slip

pTH

S-4

258

29.8

9.54

39.29

9.56

146.8

135.9

135.9

137.0

1.01

Bar

fracture

Bar

fracture

pTH

S-5

426

49.1

12.09

40.11

12.09

164.9

135.9

135.9

135.4

1.00

Bar

fracture

Bar

fracture

pTH

S-6

375

43.3

9.24

30.70

9.25

159.9

135.9

135.9

134.6

0.99

Bar

fracture

Bar

fracture

pTH

S-7

50.9

0.20

0.82

0.18

128.7

135.9

128.7

137.6

1.07

Bar

bond–

slip

Bar

fracture

XTH

S-8

233.8

0.67

2.22

0.68

143.0

135.9

135.9

133.2

0.98

Bar

fracture

Bar

fracture

pTH

S-9

117

19.0

2.89

9.60

2.9

176.3

135.9

135.9

135.5

1.00

Bar

fracture

Bar

fracture

p

aW

henP b

6Tb,P

u¼

P b.O

therwise,

P u¼

Tb.

bW

henP b

6Tb;thegrou

tedsp

lice

isex

pected

tofailby

barbo

nd–

slip.O

therwise,

itis

conside

redto

failby

barfracture.

c‘‘p

”indicatesthat

thepred

ictedan

dtheex

perimen

talfailure

mod

esarethesame,

‘‘X”indicatesotherwise.


Pb;wbs ¼ plbffiffiffiffiffiffiffiffiffiffiffiffidbf u;g

q0:315k1 þ 1:578

ffiffiffiffiffidb

p� �ð16Þ

Pb;ths ¼plb

ffiffiffiffiffiffiffiffiffiffiffiffidbf u;g

qk3

0:223k1k2 þ 1:578k3ffiffiffiffiffidb

p� �ð17Þ

where k1 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiet;sltslEsl

qð18Þ

k2 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðd2

si � dsidse þ d2seÞ

qð19Þ

k3 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffidsi cosaðdsi � dseÞ

qð20Þ

The tensile capacity is expressed in Eq. (21)

Tb ¼ k4pf yd2b

4

!ð21Þ

where k4 is a factor that correlates the ultimate capacity of the steelbars with its nominal yield strength, fy. For the steel bars withfy = 500 N/mm2, k4 is 1.35, as obtained in the test results [38].

The tensile capacity of a grouted splice is governed by either (a)the bond strength of the grout, Pb, or (b) the tensile strength of thespliced bar, Tb, whichever is weaker. Thus, when Pb 6 Tb, Pu ¼ Pb.Otherwise, Pu ¼ Tb.

5. Results and verifications

To determine the reliability of the model, the parameters usedin the experimental test are substituted in the equations (Table 4).The predicted tensile capacities of the specimens, Pu,pre, are com-pared with the experimental results, Pu,pre, in the form of a ratioRr (Eq. (22)).

Rr ¼ Pu;exp

Pu;preð22Þ

In this study, the model is formulated based on the followingassumptions:

a. Confinement stress induced by the sleeve is equally dis-tributed throughout the length of the sleeve.

b. Confinement stress can be quantified from the transversetensile strain of the sleeve.

c. Bond stress is equally distributed throughout the embeddedlength of the bar.

d. The total confinement stress caused by the sleeve is equiva-lent to the total confining stress acting on the spliced bars.

These assumptions are required at this exploratory stage whennew grouted splices are developed. The purpose is to simplify thederivation process when encountering the complicated mathemat-ical manipulations. For this reason, researchers are also makingsimilar assumptions [8,39].

The predicted outcomes, Pu,pre, in Table 4 match reasonably wellwith the experimental results, Pu,exp. The reliability ratios, Rr, of thespecimens generally range from 0.90 to 1.10.

However, specimens THS-1 and THS-3 are out of the range. Theexperimental results, Pu,exp, are 38% and 17% higher than the pre-dicted results, Pu,pre, for THS-1 and THS-3 respectively. Unrealisti-cally high bond stress tends to develop along a very short barembedded length [40], which is 75 mm in this case. This sets thelimitation of the model, where the actual performance of thegrouted splice could be underestimated when a very short barembedded length is provided.

The formulation process of the model reveals that the bondstrength generated in the sleeve can be computed by incorporating(a) the dimensions of the sleeve (tsl, dsi, dse, and a), (b) the diameter

0.00

0.50

1.00

1.50

2.00

2.50

WBS

1

WBS

2

WBS

3

WBS

4

WBS

5

WBS

6

WBS

7

WBS

8

WBS

9

THS1

THS2

THS3

THS4

THS5

THS6

THS7

THS8

THS9

Prediction Untrauer & Henry, 1965Einea et. al., 1995 Xu et. al., 2012, P1Xu et. al., 2012, P2

Rel

iabi

lity

ratio

, Rr

Specimens

Fig. 13. Comparison of the reliability ratio for various bond-confinement models from literature.

Table 5Factors for the bond-confinement models by different researchers.

Models Bond-confinementmodel (Eq. (11))

Description of test specimensb

Aa Ba Test specimen Sleeve diameter Bar size/yieldstrength

Embeddedlength

Bondingmaterials/compressivestrength

Confinement

Prediction in this study 0.223 1.578 Grouted splicespecimen

50 mm, 65 mmand 75 mm

16 mm/500 N/mm2 4.69db,7.81db,10.94db

Grout/76.75 N/mm2,68.68 N/mm2

Passive, varies.

Untrauer and Henry(1965)

0.0332 1.553 Concrete pull-out blocks

– #6/92 ksi 15db Concrete/36.1–69.2 ksi Active, constant

Einea et al. (1995) 0.0374 1.485 Grouted splicespecimens

2 in, 3 in and1½ in

#6/60 ksi 8db–13db Grout/6.5 ksi, 8.0 ksi,10.0 ksi

Passive pressure,varies

Xu et al. (2012)[41] �0.0032 2.0191 Concrete pull-out blocks

– 16 mm/335 N/mm2 5db Concrete/42.5 N/mm2 Active, constant,Perpendicular tothe bar ribs (P1)

0.0277 1.9505 Active, constant,Parallel to thebar ribs (P2)

a The factors A and B have been converted into SI units.b Unit conversion: 1 in. = 25.4 mm, #6 bar = 19 mm diameter. 1 ksi = 6.89 N/mm2.


and the embedded length of the spliced bars (db and lb), and (c) thecharacteristic and the response of the materials (fu,g, Esl and et,sl).

The predicted bond strength is then compared with the pre-dicted tensile capacity of the spliced bars, which is taken as 1.35times the specific yield strength of the spliced bars. This value isconcluded from the statistical analysis of the experimental results[38]. The expected failure mode of the grouted splice is determinedby this value. The bar is expected to fracture outside of the sleeve ifthe bond strength exceeds the tensile capacity of the bar. Other-wise, the grouted splice would endure bond–slip failure.

However, due to the minor inconsistency of the quality of thesteel bars, the bar does not always fail at the load of 1.35 timesthe specified yield strength. The predicted failure mode merelyshows a likelihood of a grouted splice to undergo the bar fractureor the bar bond–slip failures.

For this, four out of eighteen failure modes are incorrectly pre-dicted. Due to (a) inconsistent quality of the material in the realityand (b) inevitable assumptions that tend to idealize conditionswhen deriving equations, it is still considered acceptable.

Both the experimental results (Table 3) and the predicted out-comes (Table 4) prove that the sleeve configuration influences

the degree of confinement stress. THS generates a higher confine-ment stress compared with WBS. Thus, different designs of thesleeve would require different equations for predicting the tensilecapacities. This model might be only suitable for WBS and THSwith 16 mm bar diameter, but it gives an essential basis for adesigner to predict the capacity of grouted splices made from pipesections.

Although it is an analytical derivation, parts of the results areobtained through a statistical approach. The factors that correlatethe confinement stress, un,b, and the bond stress, ub, are obtainedfrom linear regression analysis based on the model proposed byUntrauer and Henry [26] (Eq. (11)). To be rigorous, the currentamount of data might not be sufficient to give a good regressionmodel. As a result, the accuracy in predicting the bond stressreduces when the confinement stress, un,b, increases (Fig. 12).

Nevertheless, compared with the other models from the litera-ture, such as by Untrauer and Henry (1965), Einea et al. (1995), andXu et al. (2012), it is giving the closest prediction of the experimen-tal results thus far. When the constants A and B for Eq. (11) are0.223 and 1.568, the reliability ratios are closer to 1.0 comparedwith the others (Fig. 13).


The models given by Untrauer and Henry (1965), Einea et al.(1995), and Xu et al. (2012) are not suitable for this case to predictthe tensile capacities of WBS and THS due to the differences in theaspects of the test specimens and the materials used, as high-lighted in Table 5.

6. Conclusion

This study proposes an economical solution to produce groutedsplice using pipe sections. It is easy to produce, requiring basictools and simple fabrication techniques, such as cutting, hammer-ing and welding.

The study consists of two main parts, the experimental studyand the analytical derivation. The experimental results show that

i. THS (tapered shape) offers 30% higher tensile capacity com-pared with WBS (pipe shape). It is more effective in generat-ing the active confinement in the sleeve.

ii. The tensile capacity increases as the bar embedded lengthincreases. The required bar embedded length of the spliceis about 6–8 times the bar diameter.

iii. The tensile capacity increases as the sleeve diameterdecreases. For this, the sleeve diameter should be as smallas possible, provided sufficient tolerance of about 25 mmis provided [42].

iv. The tensile capacity of a grouted splice is governed by (a) thetensile capacity of the spliced bars, and (b) the bond capacitybetween the bar and the grout, which is weaker.

v. In comparison with the bond–slip failure, the specimens thatfail by bar fracture generally (a) give a higher tensile capac-ity and (b) offer a higher degree of ductility.

For the analytical derivation, equations are derived on the basisof the confinement stress, which are then expressed in the func-tions of sleeve dimensions. These equations manage to predictthe tensile capacity of Welded Bar Sleeve (WBS) and Tapered HeadSleeve (THS) at the variation of ±10%.

The derivation steps reveal that the bond strength generated inthe sleeve can be computed by incorporating (a) the dimensions ofthe sleeve, (b) the diameter and the embedded length of the splicedbars, and (c) the characteristic and the response of the materials.

The proposed model requires experimental data to determinethe constants in Eq. (11). The constants A and B of 0.223 and1.568 give the closest prediction of the tensile capacities of WBSand THS thus far.

At this exploratory stage, the verification of the model is basedon the reliability of the derived equations, rather than the accuracyof the equations. For the time being, it is rather heavily dependenton the experimental results. The results would be more accurate asthe amount of the sample increases. To further verify the equa-tions, additional models other than experimental results mightbe required, such as numerical modeling.

The experiments and analysis presented are based on staticloads. The elongation and the slippage of the spliced bars are notsegregated while measuring the displacement. At this moment, itis uncertain whether the splices are applicable for cyclic and fati-gue loads, unless adequately tested.

The analytical model might only be valid for THS and WBS, andprobably, some other sleeves with identical shapes and configura-tions. To apply the model beyond the range of the parameterstested in this study, further investigation is required.

At this note, the authors intend to further explore the connec-tions in terms of response under fatigue loads and with varioussizes of bar.

Acknowledgments

The authors thank the financial support of Construction Indus-try Research Institute of Malaysia (CREAM) and ConstructionIndustry Development Board (CIDB) under Research Grant Vot73713.

References

[1] Soudli KA, Rizkalla SH, LeBlanc B. Horizontal connections for precast concreteshear wall subjected to cyclic deformations Part 1: mild steel connections. PCI J1995;40(4):78–96.

[2] Ameli MJ, Parks JE, Brown DN, Pantelides CP. Seismic evaluation of groutedsplice sleeve connections for reinforced precast concrete column-to-cap beamjoints in accelerated bridge construction. PCI J 2015;60(2):80–103.

[3] Haber ZB, Saiidi MS, Sanders DH. Seismic performance of precast columns withmechanically spliced column-footing connections. ACI Struct J 2014;111(3):639–50.

[4] Belleri A, Riva P. Seismic performance and retrofit of precast concrete groutedsleeve connections. PCI J 2012;57(1):97–109.

[5] ACI Committee 439. Mechanical connections of reinforcing bars. FarmingtonHills: American Concrete Institute; 1999.

[6] Lancelot HB. Mechanical splices of reinforcing bars: proprietary couplers fortension and compression splices; ready and able when ordinary lap splicesaren’t suitable. Concr Constr 1985.

[7] Haber ZB, Saiidi MS, Sanders DH. Behavior and simplified modeling ofmechanical reinforcing bar splices. ACI Struct J 2015;112(2):179–88.

[8] Einea A, Yamane T, Tadros MK. Grout-filled pipe splices for precast concreteconstruction. Precast/Prestressed Concr Inst J 1995;40(1):82–93.

[9] Kim YM. A study of pipe splice sleeves for use in precast beam-columnconnections. The University of Texas at Austin; 2000. Master, 116 pg.

[10] Henin E, Morcous G. Non-proprietary bar splice sleeve for precast concreteconstruction. Eng Struct 2015;83:154–62.

[11] Abd. Rahman AB, Ling JH, Ibrahim IS, Abd. Hamid Z. Performance of groutedsleeve connectors subjected to incremental tensile loads. Malaysian Constr ResJ (MCRJ) 2010;6(1):39–55.

[12] Ling JH, Abd. Rahman AB, Ibrahim IS, Abdul Hamid Z. Behaviour of groutedpipe splice under incremental tensile load. Constr Build Mater 2012;33(1):90–8.

[13] Alias A, Sapawi F, Kusbiantoro A, Zubir MA, Abdul Rahman AB. Performance ofgrouted splice sleeve connector under tensile load. J Mech Eng Sci (JMES)2014;7:1094–102.

[14] Alias A, Zubir MA, Shahid KA, Abd. Rahman AB. Structural performance ofgrouted sleeve connectors with and without transverse reinforcement forprecast concrete structure. Proc Eng 2013;53:116–23.

[15] Ling JH, Abd. Rahman AB, Abd. Hamid Z. Failure modes of aluminium sleeveunder direct tensile load. In: 3rd International conference on postgraduateeducation (ICPE-3). Penang: Sarawak Universiti Sains Malaysia (USM); 2008.

[16] Aldin Hosseini SJ, Abd Rahman AB. Analysis of spiral reinforcement in groutedpipe splice connectors. Gradevinar 2013;65(6):537–46.

[17] Ling JH, Abd. Rahman AB, Ibrahim IS. Feasibility study of grouted spliceconnector under tensile load. Constr Build Mater 2014;50(1):530–9.

[18] Sayadi AA, Ahmad Baharuddin AR, Sayadi A, Bahmani M, Shahryari L. Effectiveof elastic and inelastic zone on behavior of glass fiber reinforced polymersplice sleeve. Constr Build Mater 2015;80:38–47.

[19] Koushfar K, Abd. Rahman AB, Ahmad Y, Osman MH. Bond behavior of thereinforcement bar in glass fiber-reinforced polymer connector. Gradevinar2014;66(4):301–10.

[20] Tastani SP. Experimental evaluation of direct tension – pullout bond test. In:International symposium ‘‘bond in concrete – from research to standard”.Budapest; 2002.

[21] Tibbetts AJ, Oliva MG, Bank LC. Durable fiber reinforced polymer bar spliceconnections for precast concrete structures. Composites & Ploycon 2009.Tampa, FL, USA: American Composites Manufacturers Association; 2009.

[22] Cox JV, Herrmann LR. Development of a plasticity bond model for steelreinforcement. Mech Cohes-frict Mater 1998;3(2):155–80.

[23] Ling JH, Abd. Rahman AB, Mirasa AK, Abd. Hamid Z. Performance of CS-Sleeveunder direct tensile load: Part I – failure modes. Malaysian J Civ Eng 2008;20(1):89–106.

[24] Sayadi AA, Abd. Rahman AB, Jumaat MZ, Johnson Alengaram U, Ahmad S. Therelationship between interlocking mechanism and bond strength in elastic andinelastic segment of splice sleeve. Constr Build Mater 2014;55:227–37.

[25] Moosavi M, Jafari A, Khosravi A. Bond of cement grouted reinforcing barsunder constant radial pressure. Cement Concr Compos 2005;27:103–9.

[26] Untrauer RE, Henry RL. Influence of normal pressure on bond strength. ACI J1965;65:577–85.

[27] Coogler KL, Harries KA, Gallick M. Experimental study of offset mechanical lapsplice behavior. ACI Struct J 2008;105:478–87.

[28] Coogler KL. Investigation of the behaviour of offset mechanicalsplices. University of Pittsburgh; 2006.

[29] Jansson PO. Evaluation of grout-filled mechanical splices for precast concreteconstruction. R-1512. Michigan: Michigan Department of TransportationMDOT; 2008.


[30] ACI-318-11. Building code requirements for structural concrete andcommentary. Farmington Hills, MI: American Concrete Institute; 2011.

[31] AC-133. Acceptance criteria for mechanical connector systems for steelreinforcing bars. ICC Evaluation Service, Inc.; 2008.

[32] Alavi-Fard M, Marzouk H. Bond of high-strength concrete under monotonicpull-out loading. Mag Concr Res 2004;56(9):545–57.

[33] Robins PJ, Standish IG. The influence of lateral pressure upon anchorage bond.Mag Concr Res 1984;36:195–202.

[34] Soroushian P, Choi K-B, Park G-H, Aslani F. Bond of deformed bars to concrete:effects of confinement and strength of concrete. ACI Mater J 1991;88(2):227–32.

[35] ASTM A1034/A. Standard test methods for testing mechanical splices for steelreinforcing bars. ASTM International; 2005.

[36] Yankelevsky DZ. Bond Action between concrete and a deformed bar – a newmodel. ACI J 1985;82(2):154–61.

[37] Thompson MK, Jirsa JO, Breen JE, Klingner RE. Anchorage behaviour of headedreinforcement: literature review. Austin: University of Texas; 2002.

[38] Ling JH. Behaviour of grouted splice connections in precast concrete wallsubjected to tensile, shear and flexural loads. Universiti Teknologi Malaysia;2011. PhD 276 pg.

[39] Zhang B, Benmokrane B. Pullout bond properties of fiber-reinforced polymertendons to grout. J Mater Civ Eng 2002;14(5):399–408.

[40] Abrishami HH, Mitchell D. Analysis of bond stress distributions in pulloutspecimens. J Struct Eng 1996;122(3):255–61.

[41] Xu F, Wu Z, Zheng J, Hu Y, Li Q. Experimental study on the bond behavior ofreinforcing bars embedded in concrete subjected to lateral pressure. J MaterCiv Eng 2012.

[42] ACI Committee 117. Standard specifications for tolerances for concreteconstruction and material. Farmingto Hills: American Concrete Institute;1990.

engineering structures - universiti teknologi malaysia · tensile capacity of grouted splice...

Documents