INFLUENCE OF FIBER SHAPE ON THE STRENGTH OF STEEL

FIBER REINFORCED CONCRETE

Remigijus Šalna1, Gediminas Marčiukaitis2
Vilnius Gediminas Technical University, Saulėtekio ave. 11, LT-10223 Vilnius, Lithuania.
E-mail: 1remigijus.salna@vgtu.lt, 2gelz@vgtu.lt

Abstract. The paper presents analysis and comparison of theoretical models, evaluating the shape of steel fiber for
anchoring properties in concrete. The analysis shows that all models evaluate this effect differently. Either way all
discussed modes leads to different calculation results of strength of steel fiber reinforced concrete. According to this
analysis, the test results of steel fiber pull-out tests is made and discussed in paper. The theoretical-empirical coeffi-
cient of evaluation of fiber shape, based on analysis of mentioned models and the test results, is suggested. This coef-
ficient is verified with tests results of steel fiber reinforced concrete elements, subjected to bending and compression.

Keywords: steel fiber, SFRC, steel fiber reinforced concrete, strength.

1. Introduction et al. 1989; Zhang and Li 2002; Kanda and Li 1999;

It is well-known, that straight steel fiber, without
any modification for increasing anchorage properties
(waved, crimped, bend-up ends, etc.), is not useful in
economic as well as in strength case. However, not all
modified fibers are optimal when applied in common
performance of steel fiber reinforced concrete (SFRC).
Factors, such as yield strength of fiber f y , fiber shape,
strength of concrete f c have an influence on it. As dif-
ferent experimental research shows, in some cases the
anchorage of steel fiber is enough and it fails, when nor-
mal stresses exceed the yield strength of fiber; other wise
the fiber can be pulled out of concrete matrix by deform-
ing its anchorage shape, without exceeding the yield
strength. Besides, most of SFRC strength calculation
methods require evaluating by tests the bond stresses of
steel fiber with concrete matrix.
An experimental research on different shape steel fi-
bers pull-out tests was carried out. The purpose of the
tests was to evaluate the influence of fiber shape on the
strength of SFRC and to suggest a theoretic-empirical
model.
2. Analysis of SFRC strength calculation methods and
derivation of theoretic-empirical coefficient,
evaluating the shape of steel fiber
As it shows the analysis of SFRC strength calcula-
tion methods (Šalna 2008) of different authors (Li and
Wang 2002; Nelson et al. 2002; Li 2002; Li 1992; Wang
763
Kanda et al. 2000; Leung and Li 1991; Mallaj et al. 1995;
Stang et al. 1995; Волков and Газин 2007), three groups
of models can be distinguished:
• the calculation of strength of SFRC is based on
additive law;
• the calculation of strength of SFRC is based on
principles of failure mechanics;
• the calculation of strength of SFRC is based
only on empirical equations.
The principle of additive law (Maлмейстер et al.
1980), is mainly applied in analyzed models; however,
the correction coefficients in analyzed models is esti-
mated differently. For example, some authors replace
chaotic displacement of steel fibers to properly oriented
(Филлинс and Харис 1980; Портной and Салибеков
1979; Рабинович 2004; Marčiukaitis 1998), others em-
ploy stochastic theory or empirical coefficients (Leung
and Li 1999; Li 1992; Stang et al. 1995; Kanda and Li
2000; Kanda et al. 2000; Zhang and Li 2002). Unfortu-
nately, final values of those coefficients are significantly
different: coefficient, evaluating the angle of fiber varies
from π / 2 to (π / 2) 2 ; the coefficient, evaluating chaotic
displacement of fibers, varies from π / 2 to 1/2
(Рабинович 2004). The anchorage properties and shape
of fiber is also evaluated differently. These factors cause
that the final calculated strength of SFRC is different.
(Šalna and Marčiukaitis 2007).
According to the classical suggestions of theory of
design of composite materials, the strength σ sfrc and
modulus of elasticity Esfrc of SFRC can be expressed in
the following way:
σ sfrc = σ f V f ψ f + σc (1 − V f ) , (1)
Esfrc = E f V f ψ f + Ec (1 − V f ) , (2)
where ψ f is the coefficient of correction, depending on
properties of deformation of fibers and concrete, fibers
volume, orientation and anchorage properties, mostly
observed in test results; σ f ,σc are normal stresses of
fibers and concrete; E f ,Ec are modulus of elasticity of
steel fibers and concrete; V f is the volume of steel fibers.
Equation (1) shows that strength of SFRC mainly
depends on stresses of steel fiber σ f . Obviously, the
magnitude of stresses σ f depends on fiber anchorage
properties. If anchorage is proper, the steel fiber should
reach the yield strength f y . In the classic theory of rein-
forced concrete, the anchorage length l f , an depends on
bond stresses τ f of fiber and concrete matrix. The bond
stresses and can be defined as
τ f d f l f , an π = f y A f . (3)
Unfortunately, the behavior of crimped or hooked
steel fiber is different: The tangential stresses in hook acts
together with bond stresses.Thus, the bond stresses τ f
can be expressed in the following way:
τ f = τ1 + τ2 , (4)
where τ f is total bond stresses; τ1 – bond stresses in
straight part of the fiber; τ2 – bond stresses in hook.
The bond stresses in hook τ2 can be expressed by
bond stresses in the straight part of fiber τ2 = τ1kan . Then,
equation (4) can be rewritten:
τ f = τ1 + τ2 = τ1 + τ1kan = τ1 (1 + kan ) , (5)
where kan is the coefficient of effectiveness of anchoring
of steel fiber. This coefficient depends on the type of steel
fiber, shape of hook, failure character during deforma-
tions.
Applying the medium value of normal stresses σ f
of steel fiber, expressed by tangential stresses τ f (Mar-
čiukaitis 1998) and equation (5), the new equation for
calculating normal stresses, evaluating geometrical pa-
rameters of steel fiber, can be derived:
lf lf
σf = τf = τ1 (1 + kan ) . (6)
df df
The normal stress (Eq. 6) is employed by defining
strength of SFRC. In this case, it is necessary to replace
chaotic displacement of steel fibers to a properly oriented
764
one. As the analysis and comparison of different SFRC
strength design models show (Šalna and Marčiukaitis
2007), the most accurate results can be obtained by ap-
plying multiplication of two coefficients, suggested in
Рабинович (2004):
λ op λ p = 0, 41 . (7)
Eventually, from equations (1, 2, 6, 7), the strength
of SFRC can be expressed in following way:
σsfrc = 0, 41σ f V f + σc (1 − V f ) =
lf (8)
= 0, 41(1 + kan ) τ1 Vf + σc (1 − V f ) .
df
According to adequacy of equations (1, 2, 8), the co-
efficient of correction ψ f can be expressed by ratio be-
tween normal stress and yield strength of steel fiber. Fi-
nally, equations (1) and (2) can be expressed as follows:
⎛ σf ⎞
σ sfrc = 0, 41⎜ ⎟ σ fuV f + σc (1 − V f ) , (9)
⎜σ ⎟
⎝ fu ⎠
⎛ σf ⎞
Esfrc = 0, 41⎜ ⎟ E f V f + Ec (1 − V f ) . (10)
⎜σ ⎟
⎝ fu ⎠
The equation (8) and (9) show, that the main factor,
influencing the strength of SFRC, is normal stresses σ f
of steel fiber. The magnitude of σ f can be defined from
equation (6), when bond stresses from pull- out tests are
known.
3. The pull-out tests of different shape of steel fibers
For evaluating different bond stresses and different
failure characters of steel fibers in concrete matrix, the
test research consisted of five different types of fibers
(Fig 1): the most popular of them with hooked ends –
MPZ 50 and MPZ 60 ( d f = 1 mm, f y = 1100 MPa), two
types of waved fibers – MPS 50 ( d f = 1 mm) and MPD
50 ( b × h = 2 ,5 × 0 ,8 ) ; and the last type – fibers, cut from
steel plate MGP 32 ( b × h = 1, 2 × 0 ,9 ) was carried out.
The half of fiber, fixed in special stands to projected
position, is casted in concrete and left for hardening un-
der normal conditions for 28 days. The strength of con-
crete determined f c ,cube = 37 ,94 MPa. The specimens
were fixed in clamps of hydraulic testing machine
(Fig 2). The loading speed was 0,025 kN per second. The
average values of pull-out force, estimated by tests of
each series are presented in Table 1.
Fig 1. Different types of steel fibers tested
 τ2
a) N/2 N/2 b)
τ1

l2 l1
4 lan

'
1 τ2

3 2 h
' τ1
τ2

4 l2 l1
lan

N/2 N/2
τ1
f cu
Fig 2. The pull-out test of steel fiber. a –the testing τ1
scheme of steel fiber pull out test: 1– steel fiber; 2 –
concrete cube 100×100×100 mm; 3 – anchoring bolt;
4 – clamps of hydraulic jack; b – the failure character l2 l1
of pulled-out steel fiber lan

Table 1. Test rezults

Fig 3. Models, estimating shape of steel fibers: a – es-
Type d f, n, h, l1, l2, lan, N, timation of bond stresses τ2 in hook (wave) of fiber
mm vnt. mm mm mm mm kN in part of l2 ; b – estimation of bond stresses τ2 by
MPZ60 1,0 0,5 3,5 24,5 5,5 30 0,87
shape of fiber ( h , l2 ); c – estimation of hook by con-
MPZ50 1,0 0,5 3,5 20,5 5,5 26 0,86
crete bearing strength f cu .
MSP50 1,0 3,8 2,2 0 8 30 1,25
MPD50 2,5x0,8 4,0 1,8 0 6,5 26 1,65
The similarity of all models is that the bond stres-
MPG32 1,2x0,9 0,5 1,0 12,2 2,8 15 0,13 ses τ1 , acting in the straight part of steel fiber l1 , are as-
sumed equal to ultimate bond strength fbd . The only
4. Analysis of different models, estimating the shape of difference of these models is estimation of bond stresses
steel fiber and derivation of empirical coefficient τ 2 in the hooked part l2 . For example, the bond stresses
τ 2 according to first model (Fig 3, a) are calculated from
The classic theory of reinforced concrete expresses
a pull-out force N, and expressed in a modified length of
the bond stresses of steel bar and concrete matrix by the
hook of fiber l2 :
following way:
τ = N / ( πdlan ) , (11) N − τ1πd f l1
τ2 = . (12)
πd f l2
where N is pull-out force of steel bar from concrete
matrix, πdlan – bond area of steel bar and concrete The equation (12) demonstrates that bond stresses
matrix. τ 2 in hooked (waved) part l2 is always greater if com-
The ultimate bond strength of steel bar and concrete pared to τ1 . Besides, the distance l2 is respectively
matrix τ = fbd is usually calculated from the presented smaller. In this case, the ratio of bond stresses τ2 / τ1 can
equation. Furthermore, the optimal pull-out force should clearly express the effectiveness of hook (wave).
be equal to the ultimate force of steel fi- The second model (Fig 3, b) estimates the bond
ber: fb lan π = f y A f . Therefore, these equations have a stresses τ '2 , acting in a fictitious area, considered by
clear physical sense only for straight fibers. When fibers geometrical shape of hook:
are hooked or waved, the calculation of bond stresses is
complicated. The Fig 3 presents three, widely used mo- N − τ1πd f l1
τ '2 = . (13)
dels of steel fiber anchoring in concrete matrix 2 ( h + d f ) l2
(Курбатов and Попов 1985; Bentur et al. 1985;
Рабинович 2004).
765
 The benefits of this model is that the geometrical pa- Table 3. The average values of obtained bond stresses τ2 and
rameters of steel fiber ( h , l2 ) can be evaluated by the anchoring coefficient kan
calculation of τ2 .
Coefficient of
The influence of hook is estimated per concrete Bond Anchoring
Type of parameter
stresses coefficient
bearing strength f cu , which resists to the shape of hook steel fiber variation
τ2 , MPa kan
according to the third model (Fig 3, c): kan
N − τ1πd f l1, at MPZ 60 9,13 2,20 0,05
f cu = , (14) MPZ 50 10,75 2,59 0,02
d f hn
MPS 50 15,60 3,76 0,13
where n is the number of hooks (waves).
MPD 50 9,71 2,34 0,22
However, our tests results show (Fig 1, b), that all
steel fibers were deformed after pull-out tests (all waves MPG 32 10,71 2,58 0,09
and hooks were straighten). It means that concrete bear-
ing stresses f cu were less comparing to bond stresses τ2 , 5. Verification of coefficient suggested by tests results
and the fictitious area is destroyed, thus, the second and
The suggested equation (19) was verified by tests re-
third model is less accurate.
sults (Šalna, 2008). The tests program was aimed at find-
As the results of other authors (Ashour and Has-
ing out the influence of different fiber volume on the
sanain 1992; Li and Ward 1992) show, the bond stresses
compression and tension strengths of SFRC. In the tests
τ1 for the straight part of steel fiber are constant and can
reported, only one type of fiber MPZ50 Art. No. 1010
be assumed to be 4,15 MPa. Applying the mentioned (Metalproducts Ltd) was applied. The nominal yield
value and employing the first model, the bond stresses in strength of steel fiber was f yk = 1100 MPa. The speci-
hook τ2 is calculated (Table 3). With bond stresses in
mens were reinforced with three different volume frac-
hook τ2 , the medium normal stresses σ f , med are equal to: tions: 1, 1,5 and 2 % (78,5 kg/m3, 117,8 kg/m3, and
157 kg/m3, respectively).
( τ1 + τ2 ) l f The compression strength of SFRC was obtained by
σinc , mid = , (15)
df testing cubes (150×150×150 mm, each series consists of
4 specimens), and tension strength – by 3 point bending
and the strength of all SFRC according to the additive prisms (100×100×400 mm). The tests results and com-
law: parison of suggested equation (19) are presented in Ta-
ble 4.
σSFRC = σ f ,mid V f + σc (1 − V f ) . (16)
Table 4. Comparison of calculated and observed tests results
As it has been mentioned, the bond stresses τ2 can (Šalna, 2008)
be expressed by bond stresses τ1 : Vf , f c ,obs , ft ,obs , f c ,cal , ft ,cal , f c ,obs ft ,obs

τ2 = τ1kan . % MPa MPa MPa MPa f c ,cal f t ,cal

(17)
It is obvious, that the anchoring coefficient kan is 0 50,88 5,71 – – – –
equal to a ratio of hooked fiber pull-out force N and pull 1,0 53,78 6,78 52,81 5,73 1,02 1,18
out force of steel fiber without hook ( N ref = τ1πd f lan ): 1,5 54,03 8,21 53,77 6,93 1,00 1,18
2,0 56,37 8,95 54,73 8,13 1,03 1,10
N
kan = . (18) 0 37,94 4,41 – – – –
N ref 1,0 40,10 5,1 40 4,86 1,00 1,05
The calculated in such a way values of coefficient of 1,5 40,29 5,75 41,02 6,07 0,98 0,95
anchoring kan are presented in Table 3. 0 41,56 5,13 – – – –
Combining equations 7, 16–18, the strength of 2,0 46,04 7,3 45,6 7,49 1,01 0,97
SFRC can be finally expressed by the following equation:
As the comparison of observed by tests and calcu-
τ1 (1 + kat ) l f
σsfrc = 0, 41 Vf + σc (1 − V f ) . (19) lated results shows (Table 4), the compression strength of
df SFRC corresponds well to tests results
( f test / f calc = 0 ,98 ÷ 1, 03 ). The tension strength of SFRC
The benefit of suggested equation (19) is a possibili-
correspond to the tests results not so accurate
by f test / f calc = 0 ,95 ÷ 1,18 . The differences can be ex-
ty to evaluate all mechanical properties of materials, as
well as the type and anchorage properties of steel fiber.
plained by big plastic strains of SFRC, appearing in the
specimen, subjected to bending.

766
Conclusions
1. The different models of estimation of fiber shape on
the strength of SFRC are analyzed. The analysis
shows, that the authors estimate this effect differ-
ently.
2. Due to the analysis of models discussed, the pull-out
test of four widely used steel fibers is made. The
tests show that all specimens fail by straightening
the hooks and waves of fibers. It means that some
models, evaluating the shape of steel fiber by bear-
ing strength of concrete or bond stresses, acting in a
fictitious area can be not so accurate.
3. From the analysis of different models and pull-out
tests, the empirical anchoring coefficient, estimating
mechanical properties of materials of SFRC was
suggested.
4. The comparison of tests results with the coefficient
suggested shows a good agreement (ratio of ob-
served and calculated values for compression of
SFRC – 0,98 ÷ 1, 03 , for tension – 0 ,95 ÷ 1,18 ).
References
Ashour, S. A.; Hassanain, G. S.; Wafa, F. F. 1992. Shear Be-
havour of High Strength Fiber Reinforced Concrete
beams, ACI Structural Journal 89(2): 176–184.
Bentur, A.; Mindesst, S.; Diamond, S. 1985. Pull out process in
steel fiber reinforced cement, International Journal of
Cement Compozites 8(1): 29–37.
Kanda, T.; Lin, Z.; Li, V. C. 2000. Modeling of Tensile Stress-
Strain Relation of Pseudo Strain-Hardening Cementitious
Composites, ASCE Journal of Materials in Civil Engi-
neering 12(2): 147–156. doi:10.1061/(ASCE)0899-
1561(2000)12:2(147)
Kanda, T.; Li, V. C. 1999. Effect of Apparent Fiber Strength
and Fiber-Matrix Interface on Crack Bridging in Cement
Composites, ASCE Journal of Engineering Mechanics
125(3): 290–299. doi:10.1061/(ASCE)0733-
9399(1999)125:3(290)
Li, V. C.; Wang, S. 2002. Flexural Behavior of GFRP Rein-
forced Engineered Cementitious Composites Beams, ACI
Materials Journal 99(1): 11–21.
Li, V. C. 1992. A Simplified Micromechanical Model of Com-
pressive Strength of Fiber Reinforced Cementitious Com-
posites, Journal of Cement and Concrete Composites
14(2): 131–141.
doi:10.1016/0958-9465(92)90006-H
Li, V. C.; Ward, R.; Hamza, A. M. 1992. Steel and synthetic
fibers as shear reinforcement, ACI Materials Journal
89(5): 499–508.
Leung, C. K. Y., Li, V. C. 1991. New Strength Based Model for
the Debonding of Discontinuous Fibers in an Elastic Ma-
trix, Journal of Materials Science 26: 5996–6010.
doi:10.1007/BF01113875
767
Maalej, M.; Li, V. C., Hashida, T. 1995. Effect of Fiber Rupture
on Tensile Properties of Short Fiber Composites, ASCE
Journal of Engineering Mechanics 121(8): 903–913.
doi:10.1061/(ASCE)0733-9399(1995)121:8(903)
Marčiukaitis, G. 1998. Statybinių kompozitų kūrimo ir progno-
zavimo principai. [Design principles of composite materi-
als]. Vilnius: Technika. 134 p.
Nelson, P. K.; Li, V. C.; Kamada, T. 2002. Fracture Toughness
of Micro-Fiber Reinforced Cement Composites, ASCE
Journal Materials in Civil Engineering 14(5): 384–391.
doi:10.1061/(ASCE)0899-1561(2002)14:5(384)
Stang, H.; Li, V. C.; Krenchel, H. 1995. Design and Structural
Applications of Stress-Crack Width Relations in Fiber Re-
inforced Concrete, Journal of Materi-
als and Structures 28: 210–219. doi:10.1007/BF02473251
Wang, Y.; Backer, S.; Li, V. C. 1989. A Statistical Tensile
Model of Fiber Reinforced Cementitious Compos-
ites, Journal of Composites 20(3): 265–274.
doi:10.1016/0010-4361(89)90342-X
Zhang, J.; Li, V. C. 2002. Effect of Inclination Angle on Fiber
Rupture Load in Fiber Reinforced Cementitious Compos-
ites, Composite Science and Technology 62(6): 775–781.
Šalna, R. 2008. Dispersinio armavimo įtaka gelžbetoninių be-
sijų perdangos plokščių praspaudimui [Influence of steel
fibers on punching shear strength of beamless reinforced
concrete slabs]. Vilnius: Technika. 146 p.
Šalna, R.; Marčiukaitis, G. 2007. The influence of shear span
ratio on load capacity of fibre reinforced concrete ele-
ments with varous steel fibre volumes, Journal of Civil
Engineering and Management 13(3):
Волков, В. И.; Газин, Э. М., Бабекин, В. В. 2007.
Инженерные методы проектирования фибробетонных
конструкций [Design methods of SFRC structures].
Бетон и железобетон 4(547): 20–22.
Курбатов, Л. Г. 1985. Проектирование и изготовление
сталефибро-бетонных конструкций: Обзорная
информация. [Design and production of SFRC structures.
Basic informatikon]. ЦНТИ Госражданстроя. Москва.
55 c.
Малмейстер, А. К.; Тамуж, В. П.; Тетерс, Г. А. 1980.
Сопротивление полимерных и композитных ма-
териалов. [Resistance of polimeric and composite mate-
rials]. 3-е изд. Рига: Зинатне. 471 c.
Портной, К. И.; Салибеков, С. Е.; Светлов, И. Л.; Чибарев,
В. М. 1979. Структура и свойства композиционных
материалов. [The structure and properties of composite
materials]. Мocква: Машиностроение. 255 с.
Рабинович, Ф. Н. 2004. Композиты на основе дисперсно
армированных бетонов. Вопросы теории и
проектирования, технология, конструкции: Моногра-
фия. [Composite structures, based on fiber reinforcement.
The questions of theory, design, technology and struc-
tures]. Москва: АСВ. 506 с.
Филинс, Д.; Харрис Б. 1980. Прочность, вязкость
разрушения и усталостная выносливость полимерных
композиционных материалов. [Strength, failure and re-
sistance of fatigue of polimeric composite materials].
Мocква: Химия, 471 с.