You are on page 1of 11

Engineering Structures 49 (2013) 135–145

Contents lists available at SciVerse ScienceDirect

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

Flexural and shear capacity of composite steel truss and concrete beams
with inferior precast concrete base
Leopoldo Tesser ⇑, Roberto Scotta
Department of Civil Environmental Architectural Engineering, University of Padua, Via Marzolo 9, Padua, Italy

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

Article history: The paper presents a set of 24 lab tests on 12 composite steel truss and concrete beams with inferior pre-
Received 14 February 2012 cast concrete base completed by second phase concrete cast. The experiments investigate the shear and
Revised 18 August 2012 flexural strengths of the beams with different depth, width and transverse reinforcement inclination. The
Accepted 7 November 2012
beams failure modes are studied focusing on the crack patterns and on the interaction between concrete
Available online 25 December 2012
cast at different times. The obtained results are compared with theoretical evaluations of typical resisting
mechanisms of steel-concrete composite and reinforced concrete structures using updated European and
Keywords:
American design Standards. The main features of the composite steel truss and concrete beams are quan-
Composite steel truss and concrete beams
Hybrid truss beams
titatively and qualitatively discussed.
Flexural strength Ó 2012 Elsevier Ltd. All rights reserved.
Shear strength

1. Introduction generally assigned to the web bars immersed in concrete acting


as studs, without using other specific connectors between the
Nowadays in Italy ‘‘Composite Steel Truss and Concrete beams’’ two materials (e.g. headed stud shear connectors or similar that
(hereinafter CSTC beams), also called hybrid truss beams, have are required in standard composite beams).
been increasingly used since they allow high construction speed The original embedded steel truss is made by welding steel ele-
and minimum site labour and they provide economic convenience. ments that are plate, straight and curved bars. The typical assem-
These beams are composed by prefabricated steel trusses embed- bly is the plate as the bottom chord, two or more straight bars as
ded in cast-in-place concrete. They bear their own weight and top chord and one or more sequences of bent bars to form the diag-
the weight of the slabs without any provisional support during onal truss members (Fig. 1a). The web bars usually converge to the
floor construction and then they collaborate with the hardened top chord giving the truss a typical triangle section and providing
concrete topping. The completion concrete does not have any addi- adequate stiffness against torsion and out of plane buckling. The
tional longitudinal or transverse reinforcement except optional bottom chord is conceived to support the slab. Successively a dif-
longitudinal bar pieces to recover the continuity of multi-span ferent kind of bottom chord was developed by replacing the steel
beams. The inventor of these beams was Salvatore Leone, who pat- plate with a precast reinforced concrete base (Fig. 1b). The diagonal
ented them in 1967 and submitted his own production rules and bars are welded to some lower straight bars embedded in the pre-
assessment methods to the Italian Superior Council of Public cast concrete base. Dealing with these beams, the early academic
Works [1]. theoretical and experimental investigations focused on the buck-
The CSTC typology stands in between the reinforced concrete ling of bare trusses in the first phase [2], on the strength of the con-
and the composite steel concrete ones since it has some features nection between steel base and concrete topping [3] and on the
of both of them. Since the truss is designed to carry all the loads flexural strength [4,5].
applied before the concrete hardening, two phases can be distin- A similar construction technique called ‘‘Slim Floor’’ appeared in
guished in the life of CSTC beams. In the first phase the structure Scandinavia in the eighties with later employments and develop-
made only by the truss is considered while the latter phase con- ments in Europe and in the rest of the world [6,7]. The beams
cerns the structure composed also by the completion concrete cast. can be assessed using the Standards for composite structures and
The connection between the truss and the concrete topping is praise several applications. Likewise the CSTC beams, the Slim
Floor girder is characterized by a steel beam embedded in cast-
in-place concrete and by a lower wide flange plate supporting
⇑ Corresponding author. Address: University of Padua, Department of Civil
the slab in the first phase. A Slim Floor girder differs from a CSTC
Environmental and Architectural Engineering, Via Marzolo 9, 35121 Padua, Italy.
Tel.: +39 049 8275619; fax: +39 049 8275612. beam because of the solid web of the steel sections (wide flange
E-mail address: leopoldo.tesser@dicea.unipd.it (L. Tesser). or rectangular hollow) and of the shear connectors. In fact the first

0141-0296/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.engstruct.2012.11.004
136 L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145

shear was developed specifically for triangle section truss with


chord bars and precast concrete base.
The paper presents a set of 24 lab tests on twelve CSTC beams
made by precast concrete base and by the second phase concrete
topping. The geometry and the static schemes are designed to
investigate both the flexural and shear failure modes by two dis-
tinct tests on each beam. The obtained results are evaluated and
compared with resisting mechanism evaluations driven from cur-
rent design codes on steel-concrete composite and reinforced con-
crete structures such as Eurocode 2 [19], Eurocode 4 [20], Model
Code 2010 [21], ACI 318-08 [22].

2. Beams subjected to test

Twelve CSTC beams have been subjected to experimental tests.


They are characterized by an inferior precast concrete base sur-
rounding the bottom chord of the steel truss. The second phase
cast is completed 30 days after the cast of the base. The beams
are tested in their second phase, that is when the completion con-
crete cast is fully hardened.
On the basis of the theoretical evaluations that are discussed in
the following Section 5, each beam was designed to be submitted
to two distinct tests aiming to investigate firstly their flexural fail-
ure and secondly their shear failure. Hence some geometrical prop-
erties, such as beam depth, beam width, web bars inclination and
longitudinal and transversal reinforcement ratios, were varied to
investigate their influence on the collapse mechanisms.
The geometrical properties of the beams are listed in Table 1
Fig. 1. Classical hybrid truss beams: steel truss with steel base (a) and precast
with the following meanings of the symbols: L is the beam length,
concrete base (b).
b is the beam section width, h is the beam section depth, s is the
distance between two subsequent web bent bars, a is the angle be-
typology has shear studs welded to the top flange or raised pattern tween bent-up web bar and longitudinal axis of member, b is the
rolled into the top surface of the beam that can provide composite angle between the direction of shear force and a web bar in the
action. plane of the beam cross section. The bar nominal diameters are
Another similar composite girder, called FRC-encased steel joist all expressed in millimetres. In addition each concrete base is
composite beam, was proposed in United States in the late nineties 0.10 m deep and is completed with six longitudinal bars £8 and
for application in seismic resistant frames [8]. The key features are with rectangular stirrups £8 spaced 0.15 m. A typical tested beam
the plane steel joist made of angle and plate profiles embedded in section and profile are depicted in Fig. 2 where the introduced
concrete and fibre reinforcement for maintaining the integrity of symbols are illustrated.
the section without stirrups. This plane joist is not required to bear The intervals of geometrical parameters variation were selected
the slab before the concrete cast on-site. Analytical studies on the among their most typical values in practical applications and more
flexural strength and on the shear strength without web reinforce- specifically: the shear span to depth ratio (from 5.6 to 11.3 for the
ments were carried out [9]. The contribution of the chords angles flexural tests and from 2.2 to 4.6 for the shear ones); the width to
to the shear strength by their own strength and by their interaction depth ratio (from 0.5 to 1.7); the longitudinal lower flexural rein-
with the concrete section was not experimentally verified. forcement ratio (from 1.3% to 4.1%); the web shear reinforcement
More recent years saw an increasing number of theoretical, ratio (from 0.10% to 0.32%).
numerical and experimental investigations on CSTC beams in Italy The prescribed material properties are normal strength con-
due to: the need of improving the understanding of their resistant crete of class C50/60 according to EN 206-1 [23] (characteristic
mechanisms, the assessment of safety and serviceability require- cylindrical compressive strength fck,cyl = 50 MPa) with a maximum
ments of new construction Codes, the intention of widening the aggregate size of 15 mm and hot rolled structural steel S355JR
application field, the development of customized trusses. In partic- according to EN 10025-1 [24] (characteristic yield strength
ular a technical solution for recovering the beams continuity across fyk = 355 MPa).
columns was tested [10]. The behaviour of bare steel trusses was
the focus of experimental tests [11–13] as well as of theoretical
and numerical investigations [14,15]. The trusses with concrete 3. Test methodology
base were also tested in the first phase by means of flexural tests
[13,16]. The flexural and shear capacities of hybrid truss beams 3.1. Test technique
after the second phase concrete cast were verified experimentally
and numerically [11–13,16,17] and some assessment methods The experiments were accompanied by the characterization of
were proposed [13,15,16]. Even some novel assemblies for beam- the concrete and steel materials by means of standard tests. In par-
column joints were introduced aiming to obtain seismic resistant ticular it has been verified that the welding points do not affect the
frames [13,16,18]. Nevertheless the number of available experi- mechanical properties of the truss steel as it will be detailed in
mental data, especially those of beam shear failure, are still lim- Section 4.1.
ited; the influence of the interface between distinct concrete Each beam was submitted to two different tests varying the sta-
casts has not been verified yet; no capacity models in flexure and tic scheme configuration.
L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145 137

Table 1
Geometrical properties of the composite beams.

ID Beam geometry Steel truss


L (m) b (m) h (m) Top chord Bottom chord Web s (m) a (°) b (°)
B1 6.00 0.25 0.24 3£28 4£28 2£14 0.40 40 18
B2 6.00 0.30 0.24 3£30 4£30 2£14 0.40 40 25
B3 6.00 0.40 0.24 3£28 4£28 2£14 0.40 40 39
B4 6.00 0.25 0.34 3£30 4£30 2£14 0.40 53 11
B5 6.00 0.30 0.34 3£30 4£30 2£14 0.40 53 16
B6 6.00 0.40 0.34 3£28 4£28 2£14 0.40 53 26
B7 6.00 0.25 0.44 3£26 4£28 2£14 0.40 62 8
B8 6.00 0.30 0.44 3£30 4£30 2£14 0.40 62 12
B9 6.00 0.40 0.44 3£30 4£30 2£14 0.40 62 19
B10 6.44 0.25 0.54 3£30 4£30 2£16 0.46 65 6
B11 6.44 0.30 0.54 3£30 4£30 2£14 0.46 65 9
B12 6.44 0.40 0.54 3£30 4£30 2£14 0.46 65 15

(a)

(b)

Fig. 2. Typical tested CSTC beam section (a) and profile. (b) Measures are expressed in centimetres.

All the tests were carried out under displacement control with
three increasing amplitude loading and unloading cycles up to fail-
ure. The measure instruments captured local and global displace-
ments allowing the characterization of the beams behaviour.

3.2. Test setup

The experimental tests were performed in the Construction


Material Experimentation Laboratory at the Department ICEA of
the University of Padua. The test setup is presented in Fig. 3. The
load was applied by an hydraulic jack setup on a steel contrast frame
firmly anchored to the lab floor. The jack was controlled by an
hydraulic control unit that imposed the prescribed displacements.
A load cell was placed between the jack and a distributor beam to
measure the applied force more precisely. The distributor beam is
simply supported and loaded in its mid-span. Thus the specimen
is loaded by two equal forces whose positions are reported in Ta-
ble 2. The steel beam transferred the load to the specimen by two
low rails. The size of the low rails in the longitudinal direction of Fig. 3. Drawing of the test set-up.
the specimen was 0.12 m. Three neoprene thick layers were inserted
between the rail base and the specimen in order to avoid horizontal
constraint effects. The specimen was then supported by other two for shear tests. For each experiment Table 2 specifies the values of
steel low rails on strengthened steel pedestals. The total weight of the following dimensions: clear span S; minimum distance be-
the superstructure applying the load was about 4.1 kN. tween support axis and load axis A; distance between load axes B.
The specimen static scheme is simply supported beam with two During the tests, local and global displacements were measured.
point loads as illustrated in Fig. 4a for flexural tests and in Fig. 4b For the local measures, strain transducers with a measurement
138 L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145

Table 2 standard compression test on six concrete cylinder specimens


Effective static schemes. and standard tensile test on six steel coupons (3£14 mm and
ID Flexural tests Shear tests 3£30 mm). The steel coupons were cut from undamaged portions
S (m) A (m) B (m) S (m) A (m) B (m) of the tested beams during their demolition. Three of them were
taken from top chords and the other three from truss webs. The
B1 5.78 2.58 0.62 4.00 0.90 0.40
B2 5.78 2.58 0.62 4.00 0.90 0.40
pieces from top chords were extracted in such a way they included
B3 5.78 2.58 0.62 4.00 0.90 0.40 a welding spot that was connecting them with the web bars.
B4 5.78 2.53 0.72 4.00 0.90 0.40 The samples test results are summarized in Tables 3 and 4
B5 5.78 2.58 0.62 4.00 0.90 0.40 where fc0 is the concrete cylindrical compressive strength, fy is the
B6 5.78 2.65 0.48 4.00 0.90 0.40
steel yield strength, fu is the steel tensile strength, At is the ultimate
B7 5.98 2.75 0.48 4.00 0.90 0.40
B8 5.42 2.47 0.48 4.00 0.90 0.40 elongation over five equivalent diameters. The prescribed class for
B9 5.84 2.38 1.02 4.00 0.90 0.40 concrete was the same for the first phase and for the second phase.
B10 6.44 2.98 0.48 5.50 1.00 0.40 The concrete casts were carried out at the prefabrication plant
B11 6.44 2.98 0.48 5.50 1.10 0.40
using the same mix design and the same control procedures. Thus
B12 6.44 2.98 0.48 5.50 1.10 0.40
it can be assumed that the concrete properties of the bases are the
same of those of the toppings. The flexural and the shear tests were
base of 100 mm were used. For the elongations of the steel truss carried out several months after the concrete casts. Hence the dif-
embedded in the concrete cast, some pieces of steel bar were ference in terms of ageing and resistance of the concrete base and
welded perpendicularly to the truss and surrounded by rubber be- the concrete topping were negligible.
fore concrete casting in such a way they could emerge laterally. The results for the steel material are quite homogeneous, the
After concrete hardening, the rubber was removed and the trans- coefficient of variation for the yield strength being lower than
ducer’s knifes were connected with the bars stumps (Fig. 5). The 2.5%. The yielding and necking region were never located in the
accuracy of the transducers is 1/10,000 mm, that is a precision of welding spots. Thus it can be concluded that the welding did not
1 l-strain over the measurement base with a nominal range of affect the mechanical properties of the steel bars.
±2.5 mm that is ±0.025 strain over the given base. For the beam
mid-span deflection and the support adjustment, inductive stan- 4.2. Flexural test results
dard displacement transducers were used. Their accuracy is 0.5%
and 2.5% and their nominal maximum range is 100 mm or The crack patterns after failure can be seen in the drawings of
10 mm respectively. Fig. 4 includes the position of the instruments the beam profiles in Fig. 6 aiming to provide immediate qualitative
for a typical CSTC beam. The electric signals from all the measure- information on the specimens failures. The crack patterns have
ment instruments converged in a data acquisition system con- been realized by photogrammetric technique from the pictures of
trolled by a computer for instantaneous processing. the collapsed specimens and they report the cracks wider than
0.5 mm. The cracks generally appear in the central portions of
4. Test results the beams developing about vertically from the bottom of the con-
crete base to the concrete topping.
4.1. Test on material samples The specimens performances with respect to the strength and
the deflection capacity can be investigated by means of their applied
The material properties were verified by tests on concrete and load-mid span displacement curves. Fig. 7 presents the load-
steel samples. In particular the following tests were carried out: displacement curves of the beams grouped for similar depth. In

(a)

(b)

Fig. 4. Specimens static schemes for flexural (a) and shear (b) capacity tests.
L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145 139

strength. On the contrary in Fig. 7d, the curves of the three deepest
specimens display that the beam width has negligible effects on
stiffness and strength. The beams ductility was not always exploited
since the tests were stopped before the complete collapse.
The flexural strengths of the composite beams are reported in
Table 5 where Pm is the maximum load applied by the jack and
measured by the load cell; Mu,exp is the maximum bending moment
calculated accounting for the specimen self-weight and the weight
of the superstructure. Furthermore elc,max and euc,min are respec-
tively the maximum average strain recorded by the strain trans-
ducer placed on the lower chord and the minimum average
strain recorded by the strain transducers set on the upper chord.
These values can be compared to the value of the yield strain of
0.18% that is the ratio between the average yield strength of the
steel samples 367 MPa and the steel Young modulus assumed
equal to 200 GPa.
In addition, Table 5 collects the flexural failure modes deduced
from visual inspection and strain transducers measurements with
the following meaning of the symbols: Cc indicates compressive
concrete crushing, Sc stands for compressive steel yielding and St
for tensile steel yielding.
All the specimens in the first tests, characterized by shear span
to depth ratio higher than 5.6, failed in flexure as expected. Within
the range of variation of section sizes and longitudinal reinforce-
ment ratios, all the beams showed lower chord yielding. Moreover
the beams number 3 and 6 suffered concrete crushing in the upper
part of the section between the two points of load application.

4.3. Shear test results

The crack patterns developed during the shear tests are pre-
sented in the drawing of the beam profiles in Fig. 6. In all beams
Fig. 5. Pictures of the beam 2 during flexural test (a) and of the beam 9 during shear at least one diagonal crack appeared in the upper concrete section
test (b).
in the portion between the support and the closest loading axis.
This crack progressively widened and developed at one tip toward
Table 3 the upper section edge and at the other tip toward the interface be-
Concrete cylindrical standard tests. tween concrete casts. Afterwards in this last direction the cracks
advanced horizontally at the interface toward the closest support,
Coupon ID Origin fc0 (MPa)
expect for beam B8. At the peak strength two phenomena have
C1 Beam 4 53.2
been observed jointly or separately. The first one: the crack at
C2 Beam 5 59.9
C3 Beam 5 63.9 the interface enlarged and headed towards the support causing,
C4 Beam 7 54.8 in most cases, a detachment of the concrete base next to the end
C5 Beam 10 60.2 of the lower chord. The second phenomenon: the cracks at the
C6 Beam 11 62.5 interface continued diagonally in the concrete base, in most cases
Average 59.1
they were shifted with respect to the concrete topping. In any case
Standard deviation 4.2
Coefficient of variation (%) 7.2 the cracks of the shear tests did not intersect those of the flexural
tests.
Fig. 8 shows the applied load-deflection curves depth in which
the vertical displacement is measured in the axis of the two point
Table 4
Steel bar standard tests. loads. In Fig. 8a the responses of the specimens B1, B2 and B3, hav-
ing the same amount of web reinforcement, display the influence
Coupon ID Origin Diam. (mm) fy (MPa) fu (MPa) At (%)
of the section width on the shear strength. In fact for the shallow-
S1 Beam 7 14 393 544 13.8 est beams the strength increases as the beam width increases,
S2 Beam 10 14 405 546 14.1
other parameters being constant. On the contrary the curves of
S3 Beam 11 14 394 541 10.8
Average 397 544 12.9
the beams B7, B8 and B9 in Fig. 8c and those of the beams B11
Standard deviation 6.7 2.5 1.8 and B12 of Fig. 8d suggest that section width has moderate influ-
Coefficient of variation (%) 1.7 0.5 14.1 ence in the shear strength of deeper beams. The essential role
S4 Beam 7 30 377 523 17.5 played by the amount of web reinforcement can be noted when
S5 Beam 10 30 366 510 22.1
comparing the curves of the beam B10 with those of the beams
S6 Beam 11 30 359 507 22.3
Average 367 513 20.6 B11 and B12 in Fig. 8d.
Standard deviation 9.1 8.5 2.7 The shear strengths of the specimens are listed in Table 6 where
Coefficient of variation (%) 2.5 1.7 13.2 Pm is the maximum load applied by the jack and measured by the
load cell; Vu,exp is the maximum shear force calculated subtracting
the specimen self-weight and the weight of the superstructure.
Fig. 7a, the responses of the shallow specimens B1 and B3 show the Table 6 also reports the shear failure mode deduced from
significant influence of the beam width on the flexural stiffness and visual inspection and strain transducers measurements with the
140 L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145

Fig. 6. Crack patterns for the flexural (left column) and shear (right column) tests: specimen from B1 to B8 (a) and specimen from B9 to B12 (b).

following meaning of the symbols: Sw yielding of the tensile web 5. Theoretical evaluations
steel bars, D detachment between concrete base and upper portion
from the diagonal cracks to the closest support. The purpose of the present section is to verify the applicability
All the specimens in the second tests, characterized by shear and the precision of well-established mechanics to estimate the
span to depth ratio lower than 4.6, failed in shear as expected. CSTC beams strength.
None of the tested beams showed compression struts failure. The Using the hypotheses and the design formulas provided from
beams with lower web reinforcement ratio, that accounts for the Standards for reinforced concrete and composite structures
web bars inclination and section sizes, suffered relative slip be- [19–22], the flexural and shear capacities are theoretically evalu-
tween base and topping concrete after the web bars yielding. ated and compared with the experimental results.
L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145 141

Fig. 6. (continued)

0
5.1. Flexural capacity n ¼ fy ðkAs  k A0s Þ=ðbr fc0 beff dÞ ð2Þ

Referring to the final composite section made by the composite


truss and the completing concrete cast, the positive bending ulti- where fc0 is the concrete cylindrical compressive strength that can be
mate strength is here considered. adequately reduced to account for long lasting loads, beff is the
The design bending strength is determined by the flexural the- effective compressive concrete width, d is the effective depth,
ory according to world standards on reinforced concrete and com- n = x/d is the normalized neutral axis depth, As and A0s are the tensile
0
posite structures (e.g. Eurocode 2 [19], Eurocode 4 [20], Model and compressive steel areas respectively, k = rs/fy and k ¼ r0s =fy are
Code 2010 [21], ACI 318-08 [22]). The efficacy of the connection the tensile and compressive normalized steel stresses respectively,
between concrete base and topping is assessed in the following ka = a/d is the normalized depth of the concrete compression centre
Section 5.2. and br is a coefficient for the integration of the stress distribution
The plastic resistance moment of the composite cross section is over the section. If the stress block simplification is assumed then
calculated assuming the following hypotheses: full interaction be- ka = 0.4 and br = 0.8. If the tensile steel and the compressive steel
tween steel truss and concrete; the tensile strength of concrete is are both yielded then k = k0 = 1, otherwise the steel effective stresses
neglected; the effective area of concrete in compression resists a depend on the effective depth and the solution of Eq. (2) is
stress up to its cylinder compressive strength fc0 with a parabola– nonlinear.
rectangle stress distribution shape over the whole depth between The theoretical strengths of the CSTC beams are presented in
the neutral axis and the most compressed fibre of concrete (alter- the fifth column of Table 5. Fig. 9 plots the ultimate bending mo-
natively a stress block assumption can be adopted as suggested by ments obtained from the experiments vs. the theoretical ones.
Eurocode 2 [19] without significantly loosing in precision when the The relative error of the theoretical evaluation with respect to
tensile steel yields). the experimental results is shown in the sixth column of Table 5
Thus the ultimate positive bending moment can be determined and it is negative if conservative. The relative error mean is
by the following formula: 11.6% and its standard deviation is about 5.8% demonstrating
that the experimental results are estimated very well by the theo-
M u;d ¼ fy As dð1  ka nÞ ð1Þ retical evaluations. The negative error mean is likely due to the
steel hardening not taken into account by the conservative theoret-
and by imposing the equilibrium of the section normal solicitation ical evaluations. The standard deviation is comparable with those
leading to of the materials strength properties.
142 L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145

250
(a) 180 (b)
160

140 200

Applied load [kN]

Applied load [kN]


120
150
100

80
100
60
B1 B4
40 50
B2 B5
20
B3 B6
0 0
0 100 200 300 0 50 100
Displacement [mm] Displacement [mm]

350
(c) 350 (d)
300 300

Applied load [kN]


Applied load [kN]

250 250

200 200

150 150

100 B7 100 B10


B8 B11
50 50
B9 B12
0 0
0 20 40 60 80 0 20 40 60
Displacement [mm] Displacement [mm]

Fig. 7. Applied load vs. mid-span displacement curves for flexure tests for 240 mm (a), 340 mm (b), 440 mm (c) and 540 mm (d) beam depth.

Table 5 transverse reinforcement are not fulfilled. The nominal strength


Flexural tests’ results. against vertical shearing is generally estimated as the sum of a con-
ID Pm Mu,exp elc,max euc,min Flexural Mu,d Error crete and a steel contributions as follows:
(kN) (kN m) (%) (%) failure mode (kN m) (%)
V u;d ¼ V c;d þ V s;d ð3Þ
B1 111 155 0.85 0.34 St–Sc 140 9.7
B2 150 206 0.82 0.55 St–Sc 162 21.4 where Vc,d and Vs,d are the nominal shear strengths provided by con-
B3 143 201 1.94 0.41 St–Sc–Cc 158 21.4 crete and by shear reinforcement respectively and that can be ex-
B4 211 281 2.46 0.29 St–Sc 263 6.4
pressed (in Newton) as follows if applied to the CSTC beams:
B5 208 284 1.06 0.29 St–Sc 262 7.7
B6 193 276 3.01 0.42 St–Sc–Cc 234 15.2 V c;d;EC2 ¼ 0 ð4Þ
B7 276 399 2.88 0.61 St–Sc 329 17.5
B8 313 404 1.78 0.31 St–Sc 364 9.9 qffiffiffiffi
B9 303 385 0.42 0.11 St 369 4.2 V c;d;MC ¼ kv bz fc0 ð5Þ
B10 332 519 0.39 0.22 St–Sc 465 10.4
B11 326 514 0.42 0.23 St–Sc 479 6.8 qffiffiffiffi
B12 328 524 1.16 0.23 St–Sc 476 9.2
V c;d;ACI ¼ 0:165bd fc0 ð6Þ
Average 11.6
Standard deviation 5.8
V s;d;EC2 ¼ V s;d;MC ¼ Av fyv ðcot h þ cot aÞ sin a cos bz=s ð7Þ

5.2. Shear capacity


V s;d;ACI ¼ Av fyv ð1 þ cot aÞ sin a cos bd=s ð8Þ
For the hybrid beams, the shear force is carried by the interaction fc0
in which is the specified compressive concrete strength in MPa (in
between steel truss and concrete. The main effect of the two differ- the tests the compressive strength of the base is the same of the
ent phases of concrete casting is the presence of the interface be- upper portion), kv is a coefficient depending on the compressive
tween base and topping. Such interface lies horizontally along the concrete strength ranging from 0.145 to 0.178 for the tested beams,
beam and is intentionally not roughened. Thus the shear strength z is the effective shear depth that is the inner lever arm of the com-
should be assessed as the minimum of the two potential shear fail- posed section in mm, Av is the area in mm2 of the a couple of rein-
ure mechanisms: the first one is the Mörsch mechanism, typical of forcing bars subjected to tension spacing s in mm, fyv is the specified
concrete beams with web reinforcement, and the second one is yield strength of transverse reinforcement in MPa and h denotes the
the shear transfer at the interface between the two concrete casts. inclination of the compression struts respect to the beam axis, while
The first mechanism is assessed with the formulas by Eurocode the other symbols are already defined. The cited Codes [19,21,22]
2 (EC2) [19], Model Code 2010 (MC) [21] and ACI 318-08 (ACI) [22] contain also some limitations on the maximum shear force (such
even if their prescriptions on the minimum spacing between as the conditions on compressive concrete struts) that are not
L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145 143

300 450
(a) (b)
400
250
350

Applied load [kN]


Applied load [kN]
200 300

250
150
200

100 150
B1 B4
100
B2 B5
50
50
B3 B6
0 0
0 10 20 30 40 0 10 20 30 40
Displacement [mm] Displacement [mm]

600 500
(c) (d)
450
500
400

350

Applied load [kN]


Applied load [kN]

400
300

300 250

200
200
B7 150 B10

B8 100 B11
100
50 B12
B9
0 0
0 10 20 30 0 10 20 30
Displacement [mm] Displacement [mm]

Fig. 8. Applied load vs. mid-span displacement curves for shear tests for 240 mm (a), 340 mm (b), 440 mm (c) and 540 mm (d) beam depth.

Table 6 it provides the best correlation between the experimental results of


Shear tests’ results. the shear strength respect to the theoretical evaluations. Probably
ID Pm Vu,exp Shear failure Vu,d,ACI Error Vu,d,MC Error Vu,d,EC2 Error
the angle of 26.5° offers the best estimation because of the follow-
(kN) (kN) mode (kN) (%) (kN) (%) (kN) (%) ing peculiarities of the CSTC beams: interface between different
B1 149 113 Sw–D 138 22 140 24 109 3
concrete casts, smooth web reinforcement, large diameter of web
B2 192 145 Sw–D 146 1 150 4 107 26 reinforcement, large web bar spacing. Taking into account the
B3 263 197 Sw 160 19 153 22 99 50 experimental evidence and the National Annex to EN 1992-1-1 of
B4 293 219 Sw–D 214 2 223 2 189 14 1993 [26] as well as other important investigations [27–29], it
B5 382 284 Sw–D 230 19 233 18 189 33
might be suggested to assume a minimum value of about 26.5°
B6 280 211 Sw–D 259 23 242 15 179 15
B7 492 364 Sw–D 283 22 284 22 267 27 for the inclination of concrete struts in the shear design of CSTC
B8 496 368 Sw 306 17 309 16 260 29 beams.
B9 501 373 Sw 350 6 336 10 254 32 The shear transfer mechanism at the interface between con-
B10 457 370 Sw–D 380 3 388 5 370 0 crete casts is assured by the friction and the crossing inclined
B11 376 301 Sw–D 357 19 353 17 286 5
B12 381 309 Sw 416 35 380 23 284 8
web bars, as it will be discussed in the following. The shear–fric-
Average 1 0 20 tion strength can be evaluated according to EC2 [19], MC [21]
Standard deviation 19 17 15 and ACI [22] with the following formulas:

V i;d;EC2 ¼ V i;d;MC ¼ Av f fyv f ðl sin a þ cos aÞ cos b þ cfct Acf ð9Þ

reported here in sake of brevity since they are all respected by the
V i;d;ACI ¼ Av f fyv f ðl sin a þ cos aÞ cos b ð10Þ
tested beams. The Model Code 2010 [21] evaluates the angle h of
the compression struts according to the Modified Compression Field where Avf is the area of shear friction reinforcement subjected to
Theory of Vecchio and Collins [25]. In the case of Eurocode 2 [19], tension in mm2, fyvf is their yield strength in MPa, l is the friction
this angle is prescribed to be within the range 21.8° 6 h 6 45.0° coefficient assumed equal to 0.6 for the upper portion placed
(2.5° P cot h P 1.0°) or a stricter range decided by the National An- against the hardened base not roughened and for normal-weight
nexes. The minimum value of the angle h provides the highest the- concrete, c is the cohesion coefficient equal to 0.35 for concrete sur-
oretical strength of the tensile–shear failure mechanism. The angle face without curing, fct is the tensile concrete strength and Acf is the
h = 26.5° of compression struts has been selected in this work since area of concrete section resisting shear transfer in mm2.
144 L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145

600 400
B9
B12 B8 B10
B7
B10 350
500
B11 B12
300
B8 B11
B5
400
B7 250
Mu,exp [kNm]

Vu,exp [kN]
B9 B4
B5 B6
300 200
B3
B4
B6
B2 150
200 B2
B3
B1
100 ACI
B1
100 MC
50
EC2
0 0
0 200 400 600 0 200 400 600
Mu,d [kNm] Vu,d [kN]

Fig. 9. Ultimate bending moments obtained from the tests vs. theoretical ones. Fig. 10. Ultimate shear forces obtained from the tests vs. theoretical ones.

The formula (10) of ACI 318-08 [22] is based on the shear–fric- error about 20%) but provide lower relative error standard
tion model that provides a conservative prediction of shear transfer deviation (15%) and, most of all, consistently safe estimations
strength. Some modified relationships give closer estimates of of the shear strength. In the opinion of the authors the following
shear-transfer strength (Mattock [30]). For instance, when the considerations can be adduced. Firstly the spacing of the bent
shear–friction reinforcement is inclined to the shear plane, such web bars in the composite beams is higher than the validity lim-
that the shear force produces tension in that reinforcement, the its of the standards’ formulas. Secondly the web bars are smooth
nominal shear strength Vi,d,ACI can be modified as follows: and have reduced adherence to concrete. Thirdly the web bar
2
diameters are generally higher respect to those used in ordinary
V i;d;ACI ¼ Av f fyv f ðlsina þ cos aÞ cos b þ K 1 Acf sin a ð11Þ reinforced concrete beams. All these aspects produce wider
where K1 = 2.758 MPa is a shear transfer coefficient for normal- cracks and consequently decrease the efficacy of the tensile–
weight concrete. shear concrete contribution. The estimations of the Eurocode 2
If the interface can be considered smooth, the Model Code 2010 [19], that neglects the concrete contribution, fit very well the
[21] evaluates the shear contribution rising from the dowel action experimental results right where the other Codes seem to fail.
exerted by the diagonal bars subjected to compression as In fact the Eurocode 2 [19] offers an excellent approximation
qffiffiffiffiffiffiffiffiffiffiffi of five tests out of twelve with relative errors between 15%
V dowel;d;MC ¼ aF Av fc fyv f fc0 ð12Þ and 0% and therefore the selection of the compressive concrete
strut inclination finds confirmation in a significant range of most
aF being a coefficient depending on the surface roughness suggested typical beam depths. Moreover those five tests actually showed
equal to 1.4 and Avfc being the area of diagonal bars subjected to wider shear cracks justifying the basic hypothesis inherent in
compression in mm2. the derivation of the formula.
The theoretical beams shear strengths according to the Codes
are presented in Table 6. In all cases the Mörsch mechanism pro-
vides lower strength than that of the shear transfer at the inter- 6. Conclusions
face and it determines the overall shear capacity. This is in
agreement with the appearance of sliding between concrete base The following conclusions can be drawn.
and topping after a significant widening of the diagonal cracks,
which occurs after the tensile diagonal bars yielding in the 1. The flexural tests evidenced the following results: all the
experiments. Table 6 also reports the relative errors of the theo- beams suffered yielding of the steel bottom chord; the
retical evaluation respect to the experimental results. Fig. 10 yielding of the steel top chord happened in most of the
plots the ultimate shear strengths calculated from the tests vs. beams; only two beams underwent crushing of the com-
the theoretical ones. It shall be mentioned that the experimental pressed concrete; the cracks mainly appeared in the central
shear capacity of reinforced concrete beams generally shows portions of the beams developing about vertically from the
higher scatter than the flexural one since it involves some mech- bottom of the concrete base to the concrete topping; no sig-
anisms, such as the shear friction and dowel action, that are lar- nificant sliding between the concrete base and the concrete
gely dependent on several physical parameters difficult to topping was detected.
master. Comparing the theoretical evaluations with the experi- 2. In the shear tests all the CSTC beams were affected by
mental results, the following considerations can be pointed out. inclined cracks in the portion between the load application
For what deals with Model Code 2010 [21] and ACI 318-08 point closer to the support and the support itself. While all
[22], despite a satisfactory estimation on the average (mean rel- the cracks occurred in the upper part of the section, only
ative error included from 0% to 1%), the relative error standard few of them continued in the concrete base. After the wid-
deviation of 17–19% is significantly higher than the one from ening of those cracks, visible detachments or sliding
the flexural test. Furthermore some experimental results fell between the concrete base and topping appeared.
noticeably below the theoretical evaluations. The estimations of 3. The theoretical flexural capacity was estimated by the the-
the Eurocode 2 [19] are notably conservative (mean relative ory commonly adopted by most world Standards on rein-
L. Tesser, R. Scotta / Engineering Structures 49 (2013) 135–145 145

forced concrete and steel concrete composite structures. push-out tests fit for the determination of load-slip laws of REP composite
truss beams). Science of Constructions Institute Acts, University of Trieste;
The relative errors between theoretical evaluations and
1980.
experimental results showed a modest and conservative [4] Giordano G, Spadea G. Stato Ultimo in flessione di travi in cemento armato con
mean value and a low standard deviation comparable with doppia armatura tipo REP: ricerca sperimentale (Flexural ultimate state of
those of the materials strength properties. reinforced concrete beams with double reinforcement REP type: experimental
research). Report no. 65, Structural Department of University of Calabria; 1983.
4. The excellent agreement between the flexural tests results [5] Giordano G, Ombres L, Spadea G. Modellazione teorica e controllo
and the theoretical evaluations proves that the theory of sperimentale del comportamento a rottura di travi inflesse di tipo REP
plastic resistance moment for composite cross sections (Theoretical modelling and experimental verification of the collapse behaviour
of REP type bended beams). L’Industria Italiana del Cemento; 1987. p. 617.
[19,20] can correctly esteem the flexural strength of hybrid [6] Mullett DL. Composite floor systems. UK: The Steel Construction Institute,
truss beams. Hence the basic hypotheses of the assessment Blackwell Science; 1998.
method, particularly the preservation of plane sections and [7] Naccarato PA. ‘‘Low floor-to-floor heights’’, Modern steel
construction. American Institute of Steel Construction; 2000.
the absence of sliding between concrete and steel, are valid. [8] Khuntia M, Goel SC. Experimental study of FRC-encased steel joist composite
5. The crack patterns evidenced in the shear experimental beams. J Struct Eng ASCE 1999;125(5):495–501.
tests and the measurement of the web truss elongation, [9] Khuntia M, Goel SC. Analytical study of FRC-encased steel joist composite
beams. J Struct Eng, ASCE 1999;125(5):503–9.
confirm that the Mörsch’s theory [27] can be effectively [10] Di Marco R. Sperimentazione su travi continue REP con traliccio tipo TR
applied to evaluate the shear strength of the CSTC beams. (Experimental tests on hyperstatic REP beams with TR truss type). Internal
6. The theoretical shear capacity of CSTC beams was evalu- Report University IUAV of Venice; 2004.
[11] Scotta R, Tesser L. Preliminary experiences and basic concepts on the
ated by ACI 318-08, Model Code 2010 and Eurocode 2 even
structural performance of hybrid trussed beams. In: 4th Spec. conference on
if their conditions of the transverse reinforcement maxi- the conceptual approach to structural design, Venice; 2007.
mum spacing are not fulfilled by the CSTC beams. The esti- [12] Scotta R, Tesser L. Sperimentazione su travi tralicciate miste REPÒ – NOR
mations of the first two Codes, which account for the shear (Experimental tests on composite truss beams REPÒ – NOR). VII Italian
Workshop on Composite Structures, Aesse Stampa, Benevento; 2009. p. 69–76.
concrete contribution, showed a good agreement in the [13] Tesser L. Composite steel truss and concrete beams and beam-column joints
average but a high standard deviation and some overesti- for seismic resistant frames: modelling, numerical analysis and experimental
mations respect to the experimental results. On the con- verifications. Ph.D. thesis, University of Padua, Italy; 2009.
[14] Vincenzi L, Mazzotti C, Savoia M. Stabilità in fase I del traliccio metallico delle
trary, the Eurocode 2 approach, which neglects the shear travi reticolari miste (stability in the first phase of the steel truss of composite
concrete contribution in the tensile–shear failure mecha- steel truss and concrete beams). XVII CTE Congress, Rome 2008;2:741–50.
nism, demonstrated a mean relative error of the theoretical [15] Quaranta G, Petrone F, Marano GC, Trentadue F, Monti G. Structural design of
composite concrete–steel beams with spatial truss reinforcement elements.
evaluation respect to the tests results of about 20% on the Asian J Civ Eng 2011;12(2):155–78.
conservative side when an angle of compression concrete [16] Amadio C, Macorini L, Sorgon S, Suraci G. A novel hybrid system with RC-
struts of 26.5° was adopted. It also provided a slightly lower encased steel joists. EJECE 2011;15(10):1433–63.
[17] Borri A, Grazini A. Analisi del comportamento di travi tralicciate in c.a. per il
relative error standard deviation and consistently safe miglioramento della risposta sismica degli edifici (analysis of the reinforced
estimations. concrete truss beam behaviour for the improvement of the building seismic
7. On the basis of the experimental tests and of the theoretical capacity). XII ANIDIS convention, Pisa; 2007.
[18] Scotta R, Tesser L. Comportamento di nodi trave-pilastro sismoresistenti in
evaluations, it is the authors opinion that the evidence of
struttura mista di tipo tralicciato soggetti ad azioni cicliche (‘‘Experimental
wider shear cracks respect to those typical on reinforced behavior of beam-column joints of steel truss and concrete composite
concrete beams is caused by: high spacing of bent web bars structure subjected to cyclic loads’’). Progettazione sismica (Seismic design),
in the composite beams; smoothness and low adherence of vol. 3; 2011. p. 47–62, ISSN: 1973–7432, ISBN: 978-88-6198-066-2.
[19] Eurocode 2. Design of concrete structures Part 1–1: General rules and rules for
the web bars themselves; high web bars diameters respect buildings. Comité Européen de normalisation, Brussels, Belgium; 2003.
to those used in ordinary reinforced concrete beams. The [20] Eurocode 4. Design of composite steel and concrete structures Part 1–1:
formation of wider cracks can reduce the efficacy of the General rules and rules for buildings. Comité Européen de normalisation,
Brussels, Belgium; 2004.
tensile–shear concrete contribution. [21] fib. Fib Bullettin 56: Model Code 2010. First complete draft, vol. 2. Fédération
Internationale du Béton (fib), Lausanne, Switzerland; 2010.
[22] ACI 318. Building code requirements for structural concrete (ACI 318-08) and
commentary. American Concrete Institute; 2008.
Acknowledgement [23] EN 206-1. Concrete – Part 1: Specification, performance, production and
conformity. Comité Européen de Normalisation, Brussels, Belgium; 2000.
The authors would like to warmly thank Tecnostrutture S.r.l. for [24] EN 10025-1. Hot rolled products of structural steels. General technical delivery
conditions. Comité Européen de normalisation, Brussels, Belgium; 2004.
providing the specimens and for assuring commendable technical [25] Vecchio FJ, Collins MP. The modified compression-field theory for reinforced-
support during all the experimental campaign. concrete elements subjected to shear. ACI J 1986;83(2):219–31.
[26] National Annex to EN 1992-1-1 – Eurocode 2, Superior Council of Public
Works, Rome, Italy; 1993.
References [27] Kupfer H. Erweiterung der mörsch’schen fachwerkanalogie mit hilfe des
princips vom minimum der formänderungsarbeit. CEB. Bull Inf 1964;40:
44–57.
[1] Leone S. Procedimenti di calcolo per travi REP (‘‘Method of calculus of REP
[28] Toniolo G, Di Prisco M. Tecnica delle Costruzioni vol. 2A – Cemento Armato –
beams’’). Associazione Produttori travi REP (‘‘REP Beams Producers
Calcolo agli stati limite (Construction Technique vol. 2A – Reinforced Concrete
Association’’), Milan, March 1972.
– Limit state design). 3rd ed., Zanichelli, Bologna, Italy; 2010.
[2] Papia M. Indagine teorica e sperimentale sui fenomeni di instabilità flesso-
[29] Nielsen MP, Hoang LC. Limit analysis and concrete plasticity. 3rd ed. Boca
torsionale nelle travi REP (theoretical and experimental investigation on the
Raton, FL, USA: CRC Press; 2010.
flexural-torsional buckling phenomena of the REP beams). Science of
[30] Mattock AH. Shear transfer in concrete having reinforcement at an angle to the
Constructions Institute Acts, University of Palermo; 1977.
shear plane. In: Shear in Reinforced Concrete, SP-42. Farmington Hills
[3] Puhali R, Smotlack I. Relazione sulle prove di push-out atte a determinare le
(MI): American Concrete Institute; 1974. p. 17–42.
leggi di carico-scorrimento delle travi in sistema composto REP (Report on the

You might also like