Engineering Structures 172 (2018) 967–980

Contents lists available at ScienceDirect

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

Predicting pull-out behaviour of 4D/5D hooked end fibres embedded in T

normal-high strength concrete
⁎
Sadoon Abdallaha,b, Mizi Fana,b, , David W.A. Reesb
a
College of Material Engineering, Fujian Agriculture and Forestry University, Fuzhou 350002, PR China
b
College of Engineering, Design and Physical Sciences, Brunel University London, Uxbridge, UB8 3PH London, United Kingdom

A R T I C LE I N FO A B S T R A C T

Keywords: Previous research (Abdallah et al., 2016) conducted by the authors had proposed an analytical model to predict
Pull-out behaviour the pull-out behaviour of 4DH/5DH fibres embedded in ultra-high performance concrete. The present paper is an
Concrete strength extension to the prior study and focuses on the pull-out load-slip behaviour of these fibres embedded in normal-
An elastic-plastic response high strength concrete combinations. Based on experimental investigations, an elastic-plastic moment expression
Plastic hinge
has been proposed to represent the partially plastic hinge formed during pull-out of each circular steel fibre. This
Hooked end fibres
prediction takes into account the variation of concrete strength, geometrical and tensile properties of the fibres.
Partial deformation in straightening of the hook as well as fibre rupture was also included in the analysis. The
model developed is validated against experimental pull-out results of various hooked end fibres, for which
reasonable to very good predictions are found.

1. Introduction
Steel fibre reinforced concrete (SFRC) has become increasingly used
in modern structural engineering applications [2]. It is well established
that short steel fibres, when added to cementitious materials, improve
their tensile response, ductility, energy absorption and crack resistance
[3–5]. Another main advantage of fibre reinforcement is to enable the
concrete to persist in carrying load after cracking, the so called post-
cracking behaviour [6,7]. The effectiveness of fibre in transferring
stress from the matrix is dependent on the bond mechanisms that apply
to the fibre/matrix interface [8–10].
The bond refers to the mechanism through which the applied tensile
force is transmitted between the steel fibres and the surrounding ce-
mentitious matrix [11]. A part of this force is resisted by the cementi-
tious matrix, whilst the remainder is resisted by the fibres. The tensile
strength of SFRC can be quite variable, depending mainly on the fibre/
matrix bond strength. For the case where the fibres have a weak bond to
the matrix, pull-out at low loads is likely to occur and thus the fibre
does not contribute much to resisting the propagation of cracks [12].
Conversely, if the bond is too strong, the fibre may rupture before it can
contribute fully to the post-crack strength. Therefore, an understanding
of bond mechanism is a key factor to controlling the mechanical
properties of SFRC [13]. Currently, pull-out tests upon a single fibre are
used to investigate the bond characteristics at the interface [14].
Steel fibres for the use in concrete reinforcement can be categorized
into deformed and undeformed types. The latter (i.e. straight) fibre is
rarely used in practice and almost all fibres commercially available
today are mechanically deformed in their original geometry [15,16].
These deformations apply either to the fibre ends, such as hooks, pad-
dles, and buttons or to the fibre length, such as indented, crimped and
polygonal twisted fibres [17]. Of these, the hooked end fibre of single
bend has been the most widely used compared with other types [18].
Recently, hooked end steel fibres of improved shape were introduced,
namely, 4D (double bend) and 5D (triple bend) hooked ends. These
fibres are designed to increase the anchorage capacity of a concrete
structure to bear complex loading including tension, compression and
shear [19].
The higher performance need is matched to a multi-bend hook in a
refined matrix. The two-bend hook and traditional matrix material are
suitable for many applications where the requirement for a certain load
of safety is ensured. Should that level be sacrificed, say, in favour of a
minimum weight design, the hooked end fibre can be expected to
bridge tensile cracking in the matrix until the fibre force attains its pull-
out limit. The latter would be chosen for the hook geometry and matrix
quality from the various combinations available. The greatest pull-out
force deems that a bend pull attains 100% plasticity across the fibre
section when both the steel and the concrete are at their strongest [1].
Lesser pull-out forces spread less plasticity in normal-medium strength
combinations where slip is facilitated by a crushing failure at the fibre/
matrix interface. The pull-out force theory has adopted two alternative

⁎
Corresponding author at: College of Engineering, Design and Physical Sciences, Brunel University London, Uxbridge, UB8 3PH London, United Kingdom.
E-mail address: mizi.fan@brunel.ac.uk (M. Fan).

https://doi.org/10.1016/j.engstruct.2018.06.066
Received 15 October 2017; Received in revised form 5 April 2018; Accepted 18 June 2018
Available online 28 June 2018
0141-0296/ © 2018 Elsevier Ltd. All rights reserved.
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Table 1
Mix design of mixtures.
Concrete type Mix proportions (kg/m3) W/B Compressive strength

Cement (type) Fly ash Aggregates Sand Superplasticizer Water (–) (MPa)

8–10 mm 0–mm
NSC 364(32.5R) – 979 812 – 200 0.55 33
MSC 350(52.5 N) 107 660 1073 – 205 0.45 54
HSC 480(52.5 N) 45 850 886 6 210 0.40 72

Fig. 1. Geometrical properties of hooked end steel fibres.

approaches. The first uses the equations of static force and moment
equilibrium [1,20] and the second adopts the principle of virtual work
[21,22]. Both admit both static and dynamic friction in which slip oc-
curs under a critical force that surpasses the fibre’s limiting elastic
force. The maximum pull-out force is attained in overcoming the fric-
tional bond from matrix contact and in straightening the hook to in-
itiate a continuous slip. With a broken bond slip continues under a
reduced pull-out load required only to oppose the dynamic friction.
The present paper examines those fibre-matrix conditions for a pull-
out force where a full straightening of the fibre is not observed. The
elastic-plastic moment expression is assumed to accommodate the
plastic penetrations of 30, 50 and 70% of the fibre sectional area. The
model, when compared to alternative moment expressions, is validated
against experimental results for all three hooked end fibres embedded
in normal-high strength concretes. This study is necessary in order to
predict more reliably the pull-out behaviour of hooked end fibres for
various fibre geometries and concrete strengths.
2. Experimental investigation
2.1. Materials and specimens preparation
Three different concrete grades namely, normal strength concrete
(NSC), medium strength concrete (MSC) and high strength concrete
(HSC) were included in the experimental programme. The NSC mix
design was prepared using ordinary Portland cement (i.e. CEM II 32.5R)
whilst the other two mixes all employed high strength Portland cement
(CEM III 52.5N). Fly ash, washed sea sand as the fine aggregate and
crushed granite having a maximum size of 10 mm as coarse aggregates
were also used for the preparation of the NSC, MSC and HSC mixtures.
A superplasticizer called TamCem 23SSR was used for the preparation
of the HSC mixture. The detailed mix proportions used in this study are
summarized in Table 1.
Three types of commercially available hooked end steel fibres,
namely Dramix 3DH, 4DH and 5DH, were used in the pull-out tests.
These fibres have a same length (60 mm), diameter (0.9 mm) and aspect
ratio (l/d = 65) and only differ in the hook geometry and tensile
strength. The geometrical properties of these fibres are depicted
(stacked) in Fig. 1 and detailed in Table 2. To determine the fibre/
matrix interface characteristics, straight fibres (i.e. 3DS, 4DS, and 5DS)

Table 2
The measured geometric and mechanical properties of hooked-end fibres.
Fibre type σua (MPa) σyb (MPa) lf (mm) df (mm) Hook length (mm) Hook angles (°) Hook height (mm)

L1 L2 L3 L4 θ1 θ2 β H1 H2

3D 65/60 BG 1150 1100 60 0.90 2.12 2.95 – – 45.7 45.5 67.5 1.85 –
4D 65/60 BG 1500 1400 60 0.90 2.98 2.62 3.05 – 30.1 30.8 75.0 4.37 2.20
5D 65/60 BG 2300 2150 60 0.90 2.57 2.38 2.57 2.56 27.9 28.2 76.0 2.96 1.57

a
Ultimate strength.
b
Yield strength.

968
S. Abdallah et al.
Fig. 2. Pull-out test setup.
were also tested. The straight fibre tests were prepared by cutting off
the hooked ends of the 3DH, 4DH and 5DH fibres.
The pull-out tests on single steel fibres were performed using cubic
matrix specimen with a side dimension of 100 mm. In each test spe-
cimen, three single steel fibres were carefully placed through a hole in
the side of each mould. The embedded length was one half of the
overall fibre length (i.e. 30 mm). For compressive strength tests, five
cubes (100 × 100 × 100 mm) were also prepared. During preparation,
the dry components were firstly mixed for approximately 1 min before
water and the superplasticizer were added. They were then mixed for
9 min, a period which experience has shown is appropriate to result in a
homogenous mixture. After casting and vibration, the specimens were
covered with a thin polyethylene film to avoid retaining the escaping
moisture and left for 24 h at room temperature. The specimens were
then removed from moulds and cured for a further 28 days in the
conditioning chamber, where the temperature was held at 20 ± 2 °C
and the relative humidity at 96 ± 4%. All specimens were tested at an
age of 30 ± 2 days and the average properties of five specimens were
taken from both for the compressive strength and pull-out tests.
2.2. Test setup
The pull-out tests were performed using a specially designed grip
system, as illustrated in Fig. 2, which was attached to an Instron 5584
universal testing machine. The grips were designed such that the forces
applied to the fibre reflected those experienced by fibres bridging a
crack. The body of the gripping system was machined in a lathe using
mild steel and had a tapered end to allow the insertion of four M4 grub
screws (Fig. 2). These were then tightened around the steel fibre to an
969
Engineering Structures 172 (2018) 967–980
equal torque for an even distribution of gripping pressure to minimise
the deformation of the fibre ends and avoid breakage at the tip. Two
linear variable differential transformer (LVDT) transducers were used to
measure the distance travelled by the steel fibre relative to the concrete
face during testing (i.e. the pull-out distance). They were held in place
using aluminium sleeves on either side of the main grip body (Fig. 2).
The LVDT’s had ball bearings at the tips to allow for accurate readings
on the face of the samples. The sample was secured to the Instron base
using clamps with riser blocks and M16 studs. The specimen was po-
sitioned on a round brass disc to remove any discrepancies in the
sample base and to allow for distortion. In each pull-out test, where a
displacement rate of 10 µm/s was adopted.
2.3. Experimental results
To determine the interfacial characteristics between fibre and ma-
trix, pull-out test on straight fibres were carried out. The average pull-
out load-slip curves of straight fibre lengths of (3DS, 4DS and 5DS)
geometry pulled from the NSC, MSC and HSC matrix are shown in
Fig. 3. The curve on the left shows slip up to 30 mm and the right up to
1 mm. The latter shows that pull-out behaviour of each straight fibre is
characterized generally by a rapid increase of load up to peak, followed
by a load drop indicating full fibre/matrix debonding. Afterwards, the
pull-out process occurs under frictional resistance where the pull-out
load gradually decreases with increasing slip. It can also be observed
that all straight fibres of different strengths embedded in the same
matrix showed approximately similar peak load values, as expected.
From a common bond (3D, 4DS and 5DS), however, the comparison
between the three concretes show that the peak load increases
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Fig. 3. Average pull-out-slip response of straight steel fibres taken from 3DS, 4DS and 5DS bends. (a) Total pull-out curve and (b) detail up to a slip of 1 mm.

significantly as the compressive strength of the matrix increases. In matrix showed higher deformation and straightening of the hook than
comparison with the NSC and MSC, the percent increase in HSC’ peak those of NSC and MSC. However, the full deformation and straightening
loads is approximately 96% and 42%, respectively. of all hooked end fibres pulled-out from HSC did not occur in contrast
The pull-out load-slip curves of hooked end steel fibres (3DH, 4DH to UHPC [1].
and 5DH) are plotted in Fig. 4a–c where each curve presented is the
average curve of five specimens. As shown in Fig. 4a–c, the pull-out
3. Analytical pull-out model for 4DH/5DH fibres
response increases significantly as matrix strength increases for all
hooked end fibres. In each comparison between the three hooked end
In the prior study by Abdallah et al. [1], an analytical model of the
fibres the highest anchorage effect is provided by the hook of 5DH fibre
pull-out behaviour of 4DH and 5DH hooked end fibres; based on the
which has increased pull-out strength considerably. In each case, it can
frictional pulley model was proposed. The pull-out process of each
be seen that the pull-out strength of 5DH fibres is greater than that of
hooked end fibre was divided according to the fibre geometry (i.e.
the 3DH and 4DH fibres for all concretes tested.
number of bends) into four, five and six stages for 3DH, 4DH and 5DH
To understand further the effect of matrix strength on pull-out be-
fibres, respectively as shown in Figs. 6–8. From static force and moment
haviour of hooked end fibres, deformation and straightening processes
equilibrium, the pull-out force P in each pull-out stage can be de-
of the 3DH, 4DH and 5DH fibres embedded in different matrixes were
termined as given by Eqs. (1)–(4). The mathematical derivation has
examined under an optical microscope (Fig. 5). Scrutiny of the mor-
been explained in detail in [1].
phology of deformation shows that the deformation and straightening
of the hook increases significantly with the matrix strength for all FPH
ΔP′ =
hooked end fibres. All hooked end fibres pulled-out from the HSC 1−μ × cosβ (1)

970
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

embedded hooked fibre designs. The force P1 is required initially for

complete fibre de-bonding to allow the force to pick up and overcome
the anchorage resistance from the hook. Therefore it seemed natural to
list those key forces P2 and P3 sequentially in the manner given. The
final force P4, P5 and given in Figs. 6, 7 and 8 respectively is that re-
quired to maintain the rate of slip of a straightened fibre where only
kinetic friction resists. An exponential force decay applies with in-
creasing slip this being a common characteristic for all fibres irre-
spective of their pre-deformed geometry whether initially straight or
hooked [23,16].Where ΔP′, ΔP″, ΔP‴, ΔP⁗ represent the pull-out force
due to plastic deformation contribution of one, two, three and four
plastic hinges (PH), respectively. By adding pull-out force at onset of
complete debonding P1 to Eqs. (1)–(4) the pull-out force P as a function
of fibre slip Δ in all main stages can be obtained as follows:
For 3DH fibre:

P2 = P1 + ΔP″ (Eq.(2)) ⎫
P=⎧
⎩ P3 = P1 + ΔP′ (Eq.(1)) ⎬
⎨ ⎭ (5)
For 4DH fibre:

⎧ P2 = P1 + ΔP‴ (Eq.(3)) ⎫
P= P3 = P1 + ΔP″ (Eq.(2))
⎨ ⎬
⎩ P4 = P1 + ΔP′ (Eq.(1)) ⎭ (6)
For 5DH fibre:
⁗
⎧ P2 = P1 + ΔP (Eq.(4)) ⎫
⎪ P3 = P1 + ΔP‴ (Eq.(3)) ⎪
P=
⎨ P4 = P1 + ΔP″ (Eq.(2)) ⎬
⎪ ⎪
⎩ P5 = P1 + ΔP′ (Eq.(1)) ⎭ (7)
The P1 value can be predicted using Eq. (8) of straight fibre developed
by Naaman et al. [23].
P1 = ψτfd (Δ) × (l−Δ) (8)
where ψ is the fibre perimeter, τfd (Δ) is the frictional shear stress
function for a slip Δ, and (l-Δ) is the length of fibre remaining embedded
for any slip Δ.

4. Analytical formulation of elastic-plastic responses

Fig. 9a shows the fibre’s circular section of radius rf with centroidal

axes x and y . A limiting linear elastic stress distribution in Fig. 9b
provides an initial condition for plastic penetration to occur. Penetra-
Fig. 4. Average pull-out-slip response of hooked-end steel fibres. (a) 3DH, (b) tions to depth h shown in Fig. 9c assume an elastic-perfectly plastic
4DH and (c) 5DH fibres.
material model in which the yield stress σy remains constant.
The fully elastic bending moment ME applies to Fig. 9b when the
yield stress σy applies to the section’s top and bottom points upon the
μ × cosβ fibre’s y -axis.
2FPH ⎡1 + 1 − μ × cosβ ⎤
ΔP″ = ⎣ ⎦ Ef σy
ME
1−μ × cosβ (2) = =
I ρE rf

ΔP‴ =
FPH ⎡3 +
⎣
( 2μ ∗ cosβ
1 − μ ∗ cosβ ) ⎡⎣2 (1 + μ ∗ cosβ
1 − μ ∗ cosβ ) + 1⎤⎦ ⎤⎦ which is re-arranged in two alternative forms
Ef I σy I
(1−μ∗cosβ ) (3) ME = =
ρE rf

FPH ⎡4 +
⎢
⎣
( 2μ ∗ cosβ
1 − μ ∗ cosβ ) ⎡3 + 2μ∗cosβ ⎡2 1 +
⎣ ⎣ ( μ ∗ cosβ
1 − μ ∗ cosβ ) + 1⎤⎦ where ρE is the fully elastic curvature. Substituting I =
πr f4
4
for the fibre’s
neutral x -axis:
(
+2 1+
μ ∗ cosβ
1 − μ ∗ cosβ ) + 1⎤⎦ ⎤⎥⎦ πEf r f4 πσy r f3
ΔP⁗ = ME = =
(1−μ∗cosβ ) (4) 4ρE 4 (9)
Each force increase (ΔP) adds to the common datum force P1 to An elastic-plastic bending moment Mep applies to Fig. 9c where
provide a succession of peak values that match the sequence of pull out plastic zones have penetrated inwards to depth h as shown. This mo-
forces required to free each bend from its matrix within the three ment is the sum of two components [24]:

971
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Fig. 5. Deformation and straightening of hook after pull-out test.

Mep = Me + Mp (10) Ap 1 2 π
= r f (180−2θe ) −rf cosθe (rf −h)
2 2 180
Here the elastic moment Me and the fibre’s curvature ρe apply to the
inner elastic region, respectively: θ h
Ap = r f2 ⎛1− e ⎞ π −2r f2 ⎛⎜1− ⎞⎟ cosθe
σy Ie σy × π (rf −h) 4 πσy Ef × y Ef (rf −h) ⎝ 90 ⎠ ⎝ rf ⎠
Me = = = (rf −h)3; ρe = =
rf −h 4(rf −h) 4 σy σy Hence

=
Ef rf ( ) h
1− r
f
Ap θ 2 h
= R = ⎛1− e ⎞− ⎛⎜1− ⎞⎟ cosθe
σy (11) Af ⎝ 90 ⎠ π ⎝ rf ⎠

The plastic moment contribution applies to the plastic regions within in which
which an elemental strip (Fig. 9a) shows: h h
rf −rf sinθe = h, ∴ = 1−sinθe ; ∴ 1− = sinθe
rf rf
δMp = 2σy δAy = 2σy [2(r f2−y 2 )1/2δy] ×y
Therefore,
Integrating over each plastic zone depth rf ⩽ y ⩽ rf −h :
θ 2
rf R = ⎛1− e ⎞− sinθe cosθe
Mp = 4σy ∫(r −h) y (r f2−y2 )1/2dy
f
⎝ 90 ⎠ π
giving an equation between R and θe
4
Mp = σy [h (2rf −h)]3/2 θ 1
3 (12) ∴ R = ⎛1− e ⎞− sin2θe
⎝ 90 ⎠ π
Substituting Eqs. (11) and (12) into Eq. (10)
Preset values for R = 0.3, 0.5 and 0.7 require a trial solution to each
πσy 4σy θe as shown in Table B.1 (see Appendix B). The results obtained by this
Mep = (rf −h)3 + [h (2rf −h)]3/2
4 3 (13) formula will be compared with those obtained using the following
formula proposed by Alwan et al. [20] Eq. (15) and Xu et al. [25] Eq.
Dividing Eqs. (13) and (9) gives the non-dimensional moment ratio
(16), respectively.
3 3/2
Mep h 16 ⎡ h ⎛ h ⎞ ⎤ πr f2
= ⎜⎛1− ⎟⎞ + ⎜2− ⎟ ⎡ df ⎤
ME ⎝ r f ⎠ 3π ⎢ rf ⎝ rf ⎠ ⎥ (14)
MP = ⎢f y ×
2
× ⎥
3⎦
⎣ ⎦ ⎣ (15)
16
Eq. (14) confirms the shape factor for a fully plastic circular
3π 1 ⎛ θe sin(4θe ) ⎞ cos3 (θe ) ⎤
section when
h
= 1 and that Mep = ME when
h
= 0 .The plastic zone Mep = 4σy r f3 ⎡ − +
rf rf ⎢ sinθe ⎝ 8
⎣ 32 ⎠ 3 ⎥
⎦ (16)
Ap
areas in Fig. 10 are in the ratio with the fibre section area Af as: It is agreed that the degree of plasticity required to unbend a fibre
2

Ap can be estimated numerically from the original hook angle and the
= R; where Af = πr f2 residual curvature following pull-out more precisely than the plastic
Af
penetration depths assumed here. However, this paper follows the spirit
in which the geometry of Fig. 10 shows: of the elastic-perfectly plastic material model and the availability of

972
S. Abdallah et al.
Fig. 6. Schematic diagram of 3DH fibre pull-out process.
closed solutions provided therein. In fact, a more recent extension given
by the authors shows how the residual stress following an elastic spring-
back involved in the hook’s formation become further necessary con-
siderations for subsequent hook straightening arising from pull out. The
full analysis leading to the penetration depths assumed is not given here
only that they have been shown to correlate with the hinge moment
necessary for fibre unbending in each composite grade. Thus, the
spread of plasticity within each fibre’s section area at 30, 50 and 70%
were judged to be reasonable from the quality of the three mixes and
the degree of straightening observed. The analysis presented shows, in a
meaningful way, how an elastic–plastic hinge moment, less than that
required for complete collapse facilitates the release of the anchorage
provided by the embedded hook.
973
Engineering Structures 172 (2018) 967–980
5. Model validation
Experimentation ascertains the applicability of the proposed ana-
lytical model to predict the pull-out behaviour of various hooked end
steel fibres embedded in different concrete strengths. In addition, the
predictions from the proposed elastic-plastic moment expression were
also compared with those results obtained using formulas developed by
Alwan et al. [20], Xu et al. [25] and Abdallah et al. [1]. Material
properties of the fibres (i.e. fibre geometry and tensile strength) were
used as input parameters to Eq. (13) (Table 2). The P1, P2…,P5 forces
predicted at each stage of pull-out for all hooked end fibres are sum-
marized in Tables 3. Here, the continuous pull-out curves shown were
fitted numerically to P1, P2…,P5 using a fifth degree polynomial function
[1].
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Fig. 7. Schematic diagram of 4DH fibre pull-out process.

Figs. 11–13 show the comparison between the predicted and ex-
perimental pull-out curves of all fibres embedded in NSC, MSC and
HSC. It can be seen that each polynomial has captured the main forces
of pull-out predicting reasonably well the continuous pull-out-slip re-
sponse of all hooked end fibres. The deviations between the proposed
formula and Alwan et al. [20] and Xu et al. [25] shown brackets in
Tables 3, do not exceed 12%. Of these predictions those from the
elastic-plastic moment expression proposed are more consistent in
providing good agreement with each experimental condition in-
vestigated. The proposed elastic-plastic model was also compared with

974
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Fig. 8. Schematic diagram of 5DH fibre pull-out process.

975
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Fig. 9. Stress distribution of the steel fibre circular section.

Fig. 10. The elastic and plastic zone area of the steel fibre circular section.

Table 3
The predicted pull-out forces for all hooked end fibres (coefficient of variation CV% between parentheses).
R Fibre type P1 P2 P3 P4 P5

0.3 3DH Proposed 55 351 174 – –

Alwan 55(0.0) 409(7.6) 198(6.4) – –
Xu 55(0.0) 429(10.0) 208(8.9) – –
4DH Proposed 55 526 327 173 –
Alwan 55(0.0) 623(8.4) 382(7.7) 198(6.7) –
Xu 55(0.0) 663(11.52) 406(10.7) 208(9.1) –
5DH Proposed 55 843 573 357 188
Alwan 55(0.0) 1005(8.7) 679(8.4) 419(7.9) 215(6.6)
Xu 55(0.0) 1072(11.9) 724(11.6) 445(10.9) 227(9.3)

0.5 3DH Proposed 76 437 222 – –

Alwan 76(0.0) 430(0.8) 219(0.6) – –
Xu 76(0.0) 492(5.9) 244(4.7) – –
4DH Proposed 76 656 411 222 –
Alwan 76(0.0) 644(0.9) 403(0.9) 219(0.6) –
Xu 76(0.0) 743(6.2) 461(5.7) 244(4.7) –
5DH Proposed 76 1047 714 448 240
Alwan 76(0.0) 1026(1.0) 700(0.9) 440(0.9) 236(0.8)
Xu 76(0.0) 1195(6.6) 811(6.3) 505(5.9) 265(4.9)

0.7 3DH Proposed 108 525 277 – –

Alwan 108(0.0) 462(6.3) 251(4.9) – –
Xu 108(0.0) 548(2.1) 286(1.5) – –
4DH Proposed 108 779 496 277 –
Alwan 108(0.0) 676(7.0) 435(6.5) 251(4.9) –
Xu 108(0.0) 815(2.2) 5172.0) 286(1.5) –
5DH Proposed 108 1227 843 537 297
Alwan 108(0.0) 1058(7.3) 732(7.1) 472(6.4) 268(5.1)
Xu 108(0.0) 1288(2.4) 883(2.3) 560(2.1) 307(1.6)

976
S. Abdallah et al.
Fig. 11. Comparison of predicted and experimental pull-out load-slip curves for
3DH fibre. (a) NSC, (b) MSC and (c) HSC.
the developed plastic moment model by Abdallah et al. [1]. It can be
seen from Figs. 11–13 that the plastic moment model significantly
overestimates the mechanical anchorage contribution provided by the
hook. The elastic-plastic material model provides the most convenient
estimate when plastic penetration occurs in the section symmetrically
without work hardening. The area consumed by fully yielded material
in the two segments leaves a central area to which the elastic con-
tribution to the applied moment applies. The paper shows how to
simplify the estimate realistically while retaining satisfactory accuracy
compared to the exact second moment of area for the elastic core given
here and by others. Justification for this lies in the inherent error the
idealised material model invokes but safe in the knowledge of its lower
977
Engineering Structures 172 (2018) 967–980
Fig. 12. Comparison of predicted and experimental pull-out load-slip curves for
4DH fibre. (a) NSC, (b) MSC and (c) HSC.
bound character. Thus any attempt to admit hardening would raise the
moment carrying capacity, both Mep and MP, but the ratio Mep/MP used
here would not be altered significantly.
6. Conclusions
In this study, an elastic-plastic response model has been developed
to predict the pull-out behaviour of three hooked end fibres embedded
in three low-high strength concretes in variations combinations. Based
on experimental investigations, the complete deformation and
straightening of the hook of each fibre pulled-out from NSC, MSC and
HSC was not observed. The amount of deformation and straightening of
the hook was found to be directly related to matrix strength. That is, the
S. Abdallah et al.
Fig. 13. Comparison of predicted and experimental pull-out load-slip curves for 5DH fibre. (a) NSC, (b) MSC and (c) HSC.
assumption that an elastic-plastic moment expression accommodates
plastic bend-ratios of 30, 50 and 70% of the fibre sectional area, cor-
responded to the fibres embedded in NSC, MSC and HSC, respectively.
The model considered the variation of the concrete strength, geome-
trical and tensile properties of the fibres. In addition, the model was
also able to take into account a 100% bend ratio in which the condition
for the fibre rupture may occur.
Comparison between alternative moment expressions and experi-
mental pull-out results has validated the present proposal. Moreover,
the prediction from the proposed expression described the main pull-
978
Engineering Structures 172 (2018) 967–980
out forces most consistently. The verifications presented here also re-
inforce the validity and applicability of the proposed model for pre-
dicting the pull-out-slip response of various hooked end fibres em-
bedded in various concrete strengths presented in pervious paper [1].
Acknowledgements
The first author gratefully acknowledges the financial support of the
Ministry of Higher Education and Scientific Research of Iraqi
Government for this Ph.D. project.
S. Abdallah et al. Engineering Structures 172 (2018) 967–980

Appendix A. Nomenclature

P pull-out force
Δ relative slip of the fibre
ΔP pull-out force due to plastic hinges contribution
PH plastic hinge
FPH rotational friction component
I second moment area
ME bending moment
Me elastic moment
Mp plastic moment
δA incremental strip area
δy incremental strip depth
µ friction coefficient
τfd frictional shear stress
df fibre diameter
rf fibre radius
σy fibre yield strength
σu fibre ultimate strength
Ef modulus of elasticity of fibre
θe angular penetration measure
Af total area of the steel fibre circular section
Ap plastic penetration area
ρE fully elastic curvature
ρe curvature of inner elastic zone

Appendix B

See Table B.1.

Table B.1
Trial solutions for R = 0.3, 0.5 and 0.7.

θe 1−
θe 1
sin2θe R
90 π

60 0.333 0.2757 0.0543

45 0.500 0.3184 0.1816
44 0.511 0.3181 0.1919
42 0.533 0.3165 0.2168
40 0.553 0.3135 0.2365
36 0.600 0.3027 0.2973
35.8 0.602 0.3020 0.3000
25 0.722 0.2438 0.4782
24.2 0.731 0.238 0.4930
24.1 0.732 0.2373 0.4947
24.08 0.732 0.2371 0.4953
24.02 0.733 0.2367 0.4964
24.01 0.733 0.2365 0.4996
15 0.833 0.1592 0.6740
14.1 0.843 0.1504 0.6930
14 0.844 0.1494 0.6950
13.9 0.845 0.1488 0.6970
13.8 0.846 0.1475 0.6992

References concrete subjected to dynamic loading – a review. Eng Struct 2016;124:405–17.

[1] Abdallah S, Fan M, Rees DWA. Analysis and modelling of mechanical anchorage of
4D/5D hooked end steel fibres. Mater Des 2016;112:539–52.
[2] Yoo D, Yuan T, Yang J, Yoon Y. Feasibility of replacing minimum shear reinforce-
ment with steel fibers for sustainable high-strength concrete beams. Eng Struct
2017;147:207–22.
[3] Amin A, Foster SJ, Kaufmann W. Instantaneous deflection calculation for steel fibre
reinforced concrete one way members. Eng Struct 2017;131:438–45.
[4] Parvez A, Foster SJ. Fatigue of steel-fibre-reinforced concrete prestressed railway
sleepers. Eng Struct 2017;141:241–50.
[5] Soufeiani L, Raman SN, Jumaat MZB, Alengaram UJ, Ghadyani G, Mendis P.
Influences of the volume fraction and shape of steel fibers on fiber-reinforced
[6] Abdallah S, Fan M, Zhou X, Geyt S. Anchorage effects of various steel fibre archi-
tectures for concrete reinforcement. Int J Concr Struct Mater 2016:1–11.
[7] Zhang F, Ding Y, Xu J, Zhang Y, Zhu W, Shi Y. Shear strength prediction for steel
fiber reinforced concrete beams without stirrups. Eng Struct 2016;127:101–16.
[8] Abu-Lebdeh T, Hamoush S, Heard W, Zornig B. Effect of matrix strength on pullout
behavior of steel fiber reinforced very-high strength concrete composites. Constr
Build Mater 2011;25:39–46.
[9] Abbas AA, Syed Mohsin SM, Cotsovos DM. A simplified finite element model for
assessing steel fibre reinforced concrete structural performance. Comput Struct
2016;173:31–49.
[10] Cunha VMCF, Barros JAO, Sena-Cruz JM. A finite element model with discrete
embedded elements for fibre reinforced composites. Comput Struct
2012;94–95:22–33.

979
S. Abdallah et al.
[11] Abdallah S, Fan M, Cashell KA. Bond-slip behaviour of steel fibres in concrete after
exposure to elevated temperatures. Constr Build Mater 2017;140:542–51.
[12] Abdallah S, Fan M, Cashell KA. Pull-out behaviour of straight and hooked-end steel
fibres under elevated temperatures. Cem Concr Res 2017;95:132–40.
[13] Feng J, Sun W, Wang X, Shi X. Mechanical analyses of hooked fiber pullout per-
formance in ultra-high-performance concrete. Constr Build Mater 2014;69:403–10.
[14] Lee Y, Kang S, Kim J. Pullout behavior of inclined steel fiber in an ultra-high
strength cementitious matrix. Constr Build Mater 2010;24:2030–41.
[15] Banthia N, Trottier J. Concrete reinforced with deformed steel fibers, part I: bond-
slip mechanisms. ACI Mater J 1994;91.
[16] Abdallah S, Fan M, Rees DW. Bonding mechanisms and strength of steel fiber-re-
inforced cementitious composites: overview. J Mater Civ Eng 2018;30:04018001.
[17] Georgiadi-Stefanidi K, Mistakidis E, Pantousa D, Zygomalas M. Numerical model-
ling of the pull-out of hooked steel fibres from high-strength cementitious matrix,
supplemented by experimental results. Constr Build Mater 2010;24:2489–506.
[18] Won J, Lee J, Lee S. Flexural behaviour of arch-type steel fibre reinforced ce-
mentitious composites. Compos Struct 2015;134:565–71.
980
Engineering Structures 172 (2018) 967–980
[19] Abdallah S, Fan M, Rees DWA. Effect of elevated temperature on pull-out behaviour
of 4DH/5DH hooked end steel fibres. Compos Struct 2017;165:180–91.
[20] Alwan JM, Naaman AE, Guerrero P. Effect of mechanical clamping on the pull-out
response of hooked steel fibers embedded in cementitious matrices. Concr Sci Eng
1999;1:15–25.
[21] Chanvillard G. Modeling the pullout of wire-drawn steel fibers. Cem Concr Res
1999;29:1027–37.
[22] Soetens T, Van Gysel A, Matthys S, Taerwe L. A semi-analytical model to predict the
pull-out behaviour of inclined hooked-end steel fibres. Constr Build Mater
2013;43:253–65.
[23] Naaman AE, Namur GG, Alwan JM, Najm HS. Fiber pullout and bond slip. I:
Analytical study. J Struct Eng 1991;117:2769–90.
[24] Rees DWA. Chapter 5 - Elastic-perfect plasticity. In: Rees DWA, editor. basic en-
gineering plasticity. Oxford: Butterworth-Heinemann; 2006. p. 127–60.
[25] Xu B, Ju JW, Shi H. Progressive micromechanical modeling for pullout energy of
hooked-end steel fiber in cement-based composites. Int J Damage Mech 2010.
1056789510385260.