You are on page 1of 12

Engineering Structures 126 (2016) 405416

Contents lists available at ScienceDirect

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

Experimental evaluation of flexural behavior of composite beams with


cast-in-place concrete slabs on precast prestressed concrete decks
Hetao Hou a, Xiang Liu a, Bing Qu a,b,, Tianxiang Ma a, Haining Liu a, Mingyuan Feng a, Bo Zhang c
a
School of Civil Engineering, Shandong University, Jinan, Shandong Province 250061, China
b
Dept. of Civil and Environmental Engineering, California Polytechnic State University, San Luis Obispo, CA 93407, USA
c
Institute of Standard Construction Quota of Shandong Province, Jinan, Shandong Province 250001, China

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

Article history: This paper focuses on a new type of steel-concrete composite beams consisting of cast-in-place concrete
Received 10 September 2015 slabs on Precast Prestressed Concrete Decks (PPCDs). To evaluate flexural performance of such composite
Revised 14 July 2016 beams, this research team conducted an experimental investigation on twelve specimens. Test results
Accepted 31 July 2016
show that the composite beams with PPCDs can exhibit desirable and stable flexural performance under
monotonic loading. Based upon the test results, this paper also evaluates the influence of some key design
parameters (e.g., concrete slab thickness, amount of shear studs and longitudinal reinforcement ratio) on
Keywords:
flexural performance of the composite beams.
Testing
Precast
2016 Elsevier Ltd. All rights reserved.
Flexure
Composite
Beam

1. Introduction but will reduce headroom and more importantly result in extra
costs.
Steel-concrete composite beams combine the advantages of This research team explored a practical solution to the above
structural steel such as high strength, ductility and ease of erection issue, in which the Precast Prestressed Concrete Decks (PPCDs)
and those of reinforced concrete such as high rigidity and low cost. were used as an alternative to the steel corrugated decks of con-
A conventional steel-concrete composite beam consists of a steel ventional steel-concrete composite beams. The PPCDs, which will
beam with I-shaped cross-section supporting a concrete slab cast be described in detail in a following section, are a new type of con-
on a steel corrugated deck. Through application of appropriate con- struction units recently developed for modular construction [8].
nectors (e.g., headed shear studs) between the steel beam and the While keeping the other favorable features of conventional com-
concrete slab, their relative slip can be restrained, enabling shear posite beams, the composite beams with PPCDs offer additional
force transfer between them and achievement of the composite aesthetic values (such as flat surface and similar or even reduced
beam action [1]. Recent investigations, both analytical and experi- floor depth) and can be more affordable in some cases.
mental, have shown that this type of structural components can The objective of this research was to evaluate flexural perfor-
exhibit high stiffness and strength, and behave in a ductile manner, mance of the composite beams with PPCDs. Through testing of a
making them a viable and cost-effective alternative to traditional series of twelve specimens, this research team investigated dam-
structural steel or reinforced concrete beams [27]. Nevertheless, age progression, failure mechanism and other aspects of the com-
there are remaining issues limiting the widespread acceptance of posite beams under monotonic loading. The test results obtained
such composite beams. For example, the wrinkled surface of steel from this investigation form a basis for a better understanding of
corrugated deck as part of the steel-concrete composite beam the fundamental behavior of such composite beams and help pro-
may be unfavorable to building owners and architects in some mote their applications in future building constructions. The fol-
cases. Adoption of suspended ceilings could alleviate this problem, lowing sections describe in detail the proposed composite beams,
specimen design and construction, material properties, test setup,
loading program, observations, test results, and influences of key
construction parameters on flexural performance of the composite
Corresponding author at: Dept. of Civil and Environmental Engineering, beams.
California Polytechnic State University, San Luis Obispo, CA 93407, USA.
E-mail address: bqu@calpoly.edu (B. Qu).

http://dx.doi.org/10.1016/j.engstruct.2016.07.065
0141-0296/ 2016 Elsevier Ltd. All rights reserved.
406 H. Hou et al. / Engineering Structures 126 (2016) 405416

2. Descriptions of composite beams with PPCDs

As shown in Fig. 1, a PPCD is essentially a special precast con-


crete panel on which T-shaped ribs are regularly placed. The panel
and the ribs are reinforced by prestressed steel wires parallel with
the ribs. In addition, holes are regularly provided in the ribs to
allow passage of steel rebars perpendicular to the ribs.
Fig. 2 illustrates a composite beam with PPCDs. As shown, the
PPCD segments are placed on top of a steel beam and support
the cast-in-place concrete slab. The PPCDs and cast-in-place con-
crete are integrated as a monolithic slab through the dowelling
and interlocking actions of the PPCD ribs and steel rebars penetrat-
ing the ribs. Shear studs are attached along the top flange of the
steel beam and extended into the cast-in-place portion of the con- Fig. 2. Sketch of a composite beam with cast-in-place concrete slab on PPCDs.
crete slab to transfer the shear force between the concrete slab and
steel beam, hence enabling development of the composite beam
action. Similar to the steel corrugated decks in conventional com-
shown in Fig. 4, mortar was used to fill the joint between adjacent
posite beams, the PPCDs, if sized properly, can function as the
PPCDs. Steel rebars with diameters of 8 mm, 10 mm and 12 mm
working platform and formwork for the cast-in-place concrete,
were used to reinforce the cast-in-place concrete of the specimens.
which accelerate the construction process and eliminate the exces-
Arrangement of the steel rebars in each specimen is presented in
sive cost associated with formwork erection and removal.
detail in Fig. 5. As shown, steel rebars were provided at both the
top and the bottom of the cast-in-place concrete slab. The steel
3. Specimen design and construction rebars at the bottom of the cast-in-place concrete slab were to
enhance the integrity of the slab. Therefore, only the steel rebars
It was perceived that the composite beams with PPCDs can at the top of the slab were considered in calculation of longitudinal
replicate the viable flexural behavior of the conventional compos- reinforcement ratio, q. Specific reinforcement ratio, PPCD rib orien-
ite beams with steel corrugated decks. However, very limited tation, number of PPCDs used, and slab thickness of each specimen
research, particularly experimental work, had been completed to are provided in Table 1.
confirm if the composite beams with PPCDs behave as expected. Shear studs were evenly distributed along the top flange of the
To this end, this research team tested twelve specimens designated steel beam in each specimen. Diameter and length of each shear
as CB1CB12. Specimens CB1CB11 were composite beams con- stud were 19 mm and 100 mm, respectively. Note that shear studs
sisting of PPCDs while Specimen CB12 was a composite beam with are the critical component controlling whether or not the compos-
a full-depth cast-in-place concrete slab included here for compar- ite beam action can develop in the specimens [9]; however, there
ison purpose. All the specimens employed H were no analytical models readily available for calculating the
200  150  100  6  8 as the steel beams. Here, the designation required number of shear studs in the composite beams with
of welded beam of I-cross-section reflects the depth, width of top PPCDs. To this end, the formulae recommended in the Chinese Code
flange, width of bottom flange, web thickness and flange thick- for Design of Steel Structures (GB50017) [10] for the composite
nesses of a steel member (unit in mm). Width of the concrete slab beams with full-depth cast-in-place concrete slab were used to
was kept constant at 800 mm in all specimens; however, thickness estimate the shear strength of each stud (which was determined
of the concrete slab, hf, was varied from 130 mm to 150 mm. All to be 79.5 kN). The required number and interval of shear studs,
the specimens had the same length of 3000 mm (measured from nr and s, respectively, were determined based upon the shear
beam ends) and span of 2700 mm. strength of an individual shear stud and the shear force to be trans-
The PPCDs in Specimens CB1CB10 were arranged so that the ferred along the interface between steel beam and concrete slab
PPCD ribs were perpendicular to the beam longitudinal direction. (taken as tensile yield strength of the steel beam in this research).
Specimen CB11 had the PPCD ribs parallel with the beam longitu- The actual number of shear studs provided in each specimen, na,
dinal direction. The same type of PPCDs was used in Specimens was determined based upon nr. Table 1 lists the detailed design
CB1CB11. Fig. 3 shows the detailed geometries and reinforcement information about shear studs. As shown, in all specimens except
distributions of a typical PPCD segment. As shown, steel wires with Specimens CB7 and CB8, na was selected to be the same as nr. Spec-
the nominal diameter of 4.8 mm and prestressed to 30% of their imens CB7 and CB8 respectively had na values higher and lower
yield strength (approximately prestressed to 471 MPa) were than nr and these two specimens were included for comparison
spaced at 100 mm to reinforce the PPCD along the direction paral- purpose. It is recognized that the shear studs need to have both
lel with the ribs. Moreover, two wires were used to reinforce each sufficient strength and stiffness to eliminate the slip deformation
rib along its longitudinal direction. Rib interval was kept at between the slab and steel beam for forming the fully composite
500 mm. Rectangular openings (110 mm  40 mm) were spaced bending moment resisting action. However, the factor, nr, is only
at 200 mm along each rib to allow passage of the steel rebars. As related to strength. Therefore, the specimens with na larger than

Fig. 1. A typical PPCD (before removal of extended rebars and bottom concrete panel at the PPCD ends).
H. Hou et al. / Engineering Structures 126 (2016) 405416 407

(a) cross-section (b) side view (extended rebars and bottom panel
at the ends may be cut out if needed)

Fig. 3. Detailed geometries and rebar distributions of a typical PPCD.

or equal to nr are not necessarily to exhibit the fully composite


behavior. Nevertheless, the two factors, nr and na, were introduced
here to compare the performance difference made by different
numbers of shear studs.
Note in the specimens with PPCDs (i.e., Specimens CB1CB11)
that a gap existed between the two adjacent PPCDs on different
sides of the steel beam. The width of the overlap zone between
the PPCD and the top flange of the steel beam, a, was varied from
0 mm to 60 mm in this investigation (see Fig. 5). As shown, a smal-
ler value of a corresponds to a wider gap between the adjacent
Fig. 4. Joint between adjacent PPCDs.
PPCDs. Table 1 lists specific values of a of the specimens. Note that

(a) Side view

(b) Plan view of CB1 to CB10

(c) Plan view of CB11

(d) Section A-A of CB1 to CB10 (e) Section A-A of CB11 (f) Section A-A of CB12

Fig. 5. Detailed geometries and rebar distributions of tested specimens.


408 H. Hou et al. / Engineering Structures 126 (2016) 405416

Table 1
Summary of specimen details.

Specimen Rib orientationa Amount of PPCDs hf (mm) a (mm) s (mm) na/nr Longitudinal rebars q (%)
number diameter (mm)
CB1 Perpendicular 6 130 40 130 1.0 12 8 0.29
CB2 Perpendicular 6 140 40 130 1.0 12 8 0.27
CB3 Perpendicular 6 150 40 130 1.0 12 8 0.25
CB4 Perpendicular 6 130 0 130 1.0 12 8 0.29
CB5 Perpendicular 6 130 20 130 1.0 12 8 0.29
CB6 Perpendicular 6 130 60 130 1.0 12 8 0.29
CB7 Perpendicular 6 130 40 100 1.30 12 8 0.29
CB8 Perpendicular 6 130 40 150 0.87 12 8 0.29
CB9 Perpendicular 6 130 40 130 1.0 12 10 0.46
CB10 Perpendicular 6 130 40 130 1.0 12 12 0.65
CB11 Parallel 2 130 40 130 1.0 12 8 0.29
CB12 b b 130 b 130 1.0 12 8 0.29
a
Relative to beam longitudinal direction.
b
Not applicable.

Specimens CB1, CB4, CB5 and CB6, which were included to address concrete modulus of elasticity experimentally obtained from the
the influence of overlap width, were tested at the early stage of the prisms (150 mm  150 mm  300 mm) according to the Standard
investigation. Overlap width of the other specimens was deter- for Test Method of Mechanical Properties on Ordinary Concrete [13].
0
mined based upon test data of these specimens. Table 2 presents the values of f c:cube , fc,cube, fc,prism and Ec for the
Construction of each specimen with PPCDs followed the prac- cast-in-place concrete and the concrete used in the precast panels,
tice of conventional steel-concrete composite beams with steel respectively.
corrugated decks. First, the shear studs were welded to the steel The steel rebars with nominal diameters of 8 mm, 10 mm and
beam; then the PPCDs were installed as the formwork; next, the 12 mm were HRB400. The steel wires with the nominal diameter
joints between adjacent PPCDs were filled with M20 mortar (see of 4.8 mm were Grade 1570. Flanges and webs of the steel beams
Fig. 4) and the rebars were placed according to design (see both used Q345 steel; however, their strength values were slightly
Fig. 5); last, the concrete slab was cast and cured. different. Table 3 presents the yield strength, fy, ultimate strength,
fu, and modulus of elasticity, Es of each type of steel.

4. Material properties
5. Test setup loading scheme and instrumentation
Mix of the M20 mortar was designed according to the Specifica-
tion for Mix Proportion Design of Masonry Mortar (JGJ/T98-2010) Four-point bending tests were performed for all the specimens.
[11]. However, actual strength of the M20 mortar was not experi- Fig. 6 schematically shows the test setup. As shown, each specimen
mentally measured. According to the design, the masses of Grade was simply supported at the ends and two identical point loads
42.5 cement, lime and sand (with the water content of 3% and were applied at its one-third points. At each loading step, Zones
the dry density of 1450 kg/m3) in each cubic meter of the M20 1 and 3 of the beam (see Fig. 6) were subjected to constant shear
mortar are 292.5 kg, 57.5 kg and 1500 kg, respectively. Nominal demands while Zone 2 was subjected to pure bending.
yield and ultimate strengths of the steel in shear studs are Prior to testing, the ultimate flexural resistance, Mua, of each
240 MPa and 400 MPa, respectively. However, these properties specimen was calculated through a stress diagram assuming that
were not experimentally verified. Grade 52.5 cement, natural sand the entire steel beam section yields due to tension and the stress
with the fineness modulus of 2.88 (fine aggregate), crushed stones over the compression zone of the slab section is uniform [14]. Load
with the nominal maximum size of 20 mm (coarse aggregate), was monotonically increased through force control during each
water, and fly ash were used in concrete. The masses of cement, test. The load increment, which is one tenth of the load causing
fine aggregate, coarse aggregate, water and fly ash in each cubic Mua in the specimen, was added at each loading step. The target
meter of the cast-in-place concrete are 312.4 kg, 686.7 kg, load of each step lasted ten minutes and fifteen minutes before
1196.1 kg, 159.7 kg and 55.2 kg, respectively, while the masses of and after occurrence of the first visible crack in each specimen.
the corresponding mixtures in the precast concrete are 402.9 kg, The loading protocol was continued beyond the load associated
575.7 kg, 1205 kg, 155.2 kg and 71.2 kg, respectively. with Mua and concluded when one of the following scenarios
Properties of concrete, steel rebars, steel wires, and steel beam occurred: (1) significant visible sliding was observed between steel
flange and beam web were determined from tests of material beam and concrete slab; (2) concrete severely crushed; and (3)
samples. For concrete, tests were conducted on both cubes specimen stability requirements were no longer seemed to meet.
(100 mm  100 mm  100 mm) and prisms (150 mm  150 mm  Each specimen was instrumented to record the following
300 mm). Given that concrete strength associated with the response quantities of interest: strains in concrete slab and steel
150 mm  150 mm  150 mm cubes is more commonly used. beam, deflections at the midpoint and two loading points of each
The concrete strength obtained from the 100 mm  100 mm  composite beam; and the slip between steel beam and concrete
100 mm cubes was converted to that associated with the slab at the ends of each specimen.
150 mm  150 mm  150 mm cubes based upon the prior recom-
mendations (i.e., modifying the concrete strength associated with
Table 2
100 mm  100 mm  100 mm cubes by the coefficient of 0.95) Properties of concrete (unit: MPa).
[12]. The compressive strengths associated with the 100 mm 
0
Category f c:cube fc,cube fc,prism Ec
100 mm  100 mm cubes, 150 mm  150 mm  150 mm cubes,
0 3.51  104
and 150 mm  150 mm  300 mm prisms are denoted as, f c:cube , PPCD 56 53.2 34.05
Cast-in-place concrete 40.6 38.6 25.82 3.23  104
fc,cube and fc,prism, respectively. In addition, Ec represents the
H. Hou et al. / Engineering Structures 126 (2016) 405416 409

Table 3 and cast-in-place concrete slab. Such cracks progressively


Properties of steel (unit: MPa). extended from the ends of Zone 2 (i.e., around the loading points)
Category fy fu Es to the middle portion of Zone 2. At the ultimate loading, the
Grade 1570 1547 1768 1.91  105 debonding failure along the interface between the PPCDs and
HRB 400 390 623 1.87  105 cast-in-place concrete slab in Zone 2 propagated into Zones 1
Q345 (beam flange) 360 503 2.00  105 and 3, severely compromising the specimen resistance in these
Q345 (beam web) 380 521 2.00  105 two zones. The specimen eventually developed diagonal shear
cracks in Zones 1 and 3 as shown in Fig. 8.
In comparison with the specimens consisting of PPCDs (i.e.,
Uniaxial Strain Gauges (SGs), Dial Gauges (DGs) and Linearly Specimens CB1 to CB11), Specimen CB12 behaved in a slightly dif-
Variable Displacement Transformers (LVDTs) were used to record ferent manner. With the increase of compressive loading, the
the strain, deflection and slip quantities, respectively. The instru- minor cracks initially occurred at the bottom face of the concrete
mented locations of SGs, DGs and LVDTs are illustrated in Fig. 7. slab in Zone 2. At the compressive loading of 75% of the ultimate
Note that the two LVDTs attached at each end of the specimen loading, the cracks at the bottom face of the concrete slab became
were used to obtain the slip deformation between concrete slab wider and the beam deflection increased significantly. Further
and steel beam. Ideally, the two LVDTs should be attached to the increasing the compressive loading eventually led to a flexure-
top of the steel beam and the bottom of the concrete slab, respec- dominated failure, which was characterized by concrete crushing
tively. However, the upper LVDT was arranged around the mid- and steel beam yielding over Zone 2. Fig. 9 shows the damage in
depth of the slab for the ease of installation, resulting in underes- Specimen CB12 at the end of the test. Table 4 summarizes the crit-
timates of the slip deformation between the concrete slab and steel ical bending moments in the specimens and the corresponding
beam. Moreover, intervals of the strain gauges through the con- beam deflections, in which Mcr, My and Mu represent the beam
crete slab thicknesses of Specimens CB2 and CB3 (i.e., those with bending moments associated with first visible crack in concrete
thicker slabs) are 50 mm and 55 mm, respectively. Totally, thirty- slab, first yielding in steel beam and ultimate limit state of the
six uniaxial SGs, three DGs and four LVDTs were installed in each specimen; and dy and du represent the beam deflections associated
specimen. with My and Mu, respectively. Also included in the table are the
ratios of du to dy. Note that the ratio captures the deformation
6. Test results capacity of a composite beam loaded beyond its yielding limit. As
shown, the specimen with a full-depth cast-in-place concrete slab
Tests of all the specimens were performed as planned. This sec- (i.e., Specimen CB12) has a higher ratio of du to dy than the speci-
tion describes the behaviors and failure modes observed in the mens with PPCDs. The ratio of du to dy varies from 3.72 to 6.04 with
specimens; followed by interpretation of the recorded test data. an average of 4.94 among Specimens CB1 to CB11. It is recognized
that the deflection limit considering serviceability of a beam is
1/250 of the beam span according to GB50017 [10]. As shown in
6.1. General behavior and failure mode
Table 4, all beams except Specimen CB2 have dy lower than the
limit associated with serviceability. Specimen CB2 exhibited a
Although Specimens CB1CB11 had different test parameters,
slightly higher deflection (about 6% larger) than the limit associ-
all these specimens consisting of PPCDs exhibited similar behavior.
ated with serviceability. Furthermore, Table 4 lists the resultant
Taking Specimen CB1 as example, this section describes the note-
applied load associated with My, which is denoted as Py, together
worthy observations during the experimental investigation. At
with the ratios of Py to dy. Note that the ratio of Py to dy captures
the early stage of testing, initial cracks concentrated at the joints
the stiffness of each specimen up to yielding.
between the adjacent PPCDs (see Fig. 4) over Zone 2 (see Fig. 6).
When the compressive loading reached about 60% of the ultimate
loading, bottom flange of the steel beam yielded and the cracks in 6.2. Moment-deflection backbone curves
the PPCD joints penetrated the mortar infills (see Fig. 4). Mean-
while, visible cracks developed along the interface between the Fig. 10 compares the bending moment vs. beam deflection
PPCDs and cast-in-place concrete slab over Zone 2. Further increas- curves of the specimens. Note that the bending moment and beam
ing the compressive loading beyond about 80% of the ultimate deflection were recorded at the mid-span point at each loading
loading caused a remarkable growth of the crack between PPCD step. For a better comparison, the bending moment and beam

Fig. 6. Test setup.


410 H. Hou et al. / Engineering Structures 126 (2016) 405416

Fig. 7. Instrumentation details.

stiffness degradation and the ultimate deflection, they overall exhi-


bit the similar bilinear type of moment-deflection relationship. In
addition, no specimens exhibited strength degradation during the
testing. Moreover, it is found that the bending moment resistance
associated with the first yielding of steel beam remains somewhat
constant and on the order of 60% of the corresponding ultimate
bending moment resistance in all specimens except CB11. Speci-
men CB11, which had PPCD ribs parallel with the beam longitudi-
nal direction, exhibited a lower bending moment resistance
associated with the first yielding of steel beam (i.e., 48% of its ulti-
mate bending moment resistance as shown in Fig. 10).

6.3. Discussion of representative strain data

The strain histories were recorded at the instrumented loca-


Fig. 8. Diagonal crack in Zone 3 of Specimen CB1.
tions for each specimen (see Fig. 7). Focusing on the strain data
from Specimens CB1, CB7, CB8 and CB12, this section discusses
progressive developments of strain profiles of these specimens.
Note that data from some strain gauges that were functioned
improperly at higher loading levels are removed from the result
illustrations and discussions.
Fig. 11(a) and (b) presents the strain profiles under different
loading levels recorded at the mid-span locations of Specimens
CB1 and CB12, respectively. Fig. 11(c) and (d) shows the progres-
sively developed strain profiles across the concrete slabs of the
two specimens recorded at one load application point. Note that
Specimens CB1 and CB12 have similar design parameters (see
Table 1) except that Specimen CB1 adopted PPCDs while Specimen
CB12 had a full-depth cast-in-place concrete slab. As shown, the
strain profiles of both specimens exhibit approximately linear dis-
tributions up to fairly high loading levels. Moreover, as indicated
by the level with zero strains, the neutral axis moves upward in
both specimens as the bending moment demand increases. Based
Fig. 9. Slab damage in Zone 2 of Specimen CB12.
upon linear interpolation of the data shown in Fig. 11(a) and (b),
the neutral axes locate at 67.6 mm and 68.5 mm away from the
deflection of each specimen are normalized by its ultimate bending interface between concrete slabs and steel beams at the close-to-
moment resistance and span, respectively. Also included in the fig- ultimate-state moment ratios 0.92 and 0.88 for Specimens CB1
ures are the limits associated with first yielding in steel beam and and CB12, respectively. Also at these two moment ratios, strain
first visible crack in concrete slab. As shown, these curves remain at the top of the concrete slab reaches 2.439  103 and
primarily linear at the lower loading levels but become nonlinear 2.736  103 in Specimens CB1 and CB12, respectively. The
with the increase of loading. Although the specimens differ in compressive strain in Specimen CB12 achieved a higher level
H. Hou et al. / Engineering Structures 126 (2016) 405416 411

Table 4
Summary of critical bending moments and beam deflections.

Specimen Mcr (kN m) My (kN m) Mu (kN m) My/Mu Py (kN) dy (mm) Py/dy (kN/mm) du (mm) du/dy
CB1 112.5 135.5 247.1 0.55 301.1 7.58 39.72 41.83 5.52
CB2 121.5 142.3 251.1 0.57 316.2 11.48 27.55 52.39 4.56
CB3 139.5 152.1 258.8 0.59 338.0 8.79 38.45 38.26 4.35
CB4 135.0 147.3 254.7 0.58 327.3 8.74 37.45 50.41 5.77
CB5 121.5 136.3 237.6 0.57 302.9 7.75 39.08 45.17 5.83
CB6 135.0 142.6 242.1 0.59 316.9 10.63 29.81 39.58 3.72
CB7 135.0 153.6 262.8 0.58 341.3 9.32 36.62 56.30 6.04
CB8 103.5 128.3 216.0 0.59 285.1 9.56 29.82 36.17 3.78
CB9 112.5 139.7 261.0 0.54 310.4 7.37 42.12 39.49 5.36
CB10 135.0 151.2 252.0 0.60 336.0 7.48 44.92 36.92 4.94
CB11 135.0 123.5 255.6 0.48 274.4 10.65 25.77 48.05 4.51
CB12 112.5 146.3 257.0 0.57 325.1 7.71 42.17 76.56 9.93

Fig. 10. Moment vs. deflection curves of tested specimens.

partially due to the better integrity of the full-depth cast-in-place close-to-ultimate load levels, Specimen CB7 retains the approxi-
concrete slab in Specimen CB12 in comparison with Specimen mately linear strain distribution, suggesting that plane section
CB1. The higher neutral axis location and higher slab compressive remained primarily plane in the specimen. However, the strain
strain in Specimen CB12 are consistent with the fact that Specimen profiles of Specimen CB8 cease to follow the linear distribution at
CB12 achieved a higher flexural resistance in comparison with the corresponding load levels. This is because the shear stud inter-
Specimen CB1 (see Table 4). val in Specimen CB8 is larger (see Table 1), resulting higher slip
Fig. 12(a) and (b) presents the strain profiles under different deformations between the concrete slab and steel girder at higher
loading levels recorded at the mid-span locations of Specimens load levels (which will be discussed in detail in the following sec-
CB7 and CB8, respectively. Fig. 12(c) and (d) shows the progres- tion). Although the neutral axis location is higher in Specimen CB8
sively developed strain profiles across the concrete slabs of the compared with CB7 at the comparable close-to-ultimate states, the
two specimens recorded at one load application point. Note that less successful development of the composite action due to the slip
Specimens CB7 and CB8 had identical design parameters (see between the concrete slab and steel girder led to the lower flexural
Table 1) except that Specimen CB7 had a smaller shear stud strength in Specimen CB8 (see Table 4).
interval compared with Specimen CB8. As shown, at relatively The yield strain of beam flange is also included in Figs. 11 and
low load levels, both specimens have linear strain profiles. At the 12. As shown, fibers at the top portion of the steel beam are not
412 H. Hou et al. / Engineering Structures 126 (2016) 405416

(a) strain profiles across the entire beam section (b) strain profiles across the entire beam section
(at midpoint of CB1) (at midpoint of CB12)

(c) strain profiles across the concrete slab (d) strain profiles across the concrete slab
(at one loading point of CB1) (at one loading point of CB12)

Fig. 11. Comparison of strain profiles: Specimens CB1 vs. CB12.

yielded at the close-to-ultimate states. Table 5 compares the neu- result trend are observed among all specimens. All the curves tend
tral axis location, Zur (defined as the distance from the neutral axis to exhibit increased slopes under a higher mid-span bending
to the interface between steel girder and concrete slab) of each moment, suggesting that the slip increases to a higher degree
specimen under the close-to-ultimate moment, Mur. Moreover, as when a composite beam is loaded into inelastic range. At the end
shown in Figs. 11(a) and (b) and 12(a) and (b), the tensile strain of the test, the beam end slip was found to be less than 0.1% of
recorded at the bottom of the concrete slab is higher than that of beam span (i.e., 2.7 mm) in all specimens.
the top of the steel beam at a given loading level, suggesting that
slip deformations occurred at the interface of concrete slab and
7. Discussion of influences of key parameters
steel beam. Table 6 further reports the strain at the top of concrete
slab, ect, and that at the bottom face of bottom flange of steel beam,
As listed in Table 1, some design parameters were intentionally
esb, in Zone 2 of each specimen under the critical bending moments
varied in the specimens. As such, the test data allow a further dis-
including Mcr, My and Mur.
cussion on the influences of these parameters on flexural perfor-
mance of the specimens.
6.4. Relative slip at beam end

As a consequence of the shear force transferred by the shear 7.1. Influence of slab thickness
studs, slip between the concrete slab and steel beam was observed
at ends of each specimen. Fig. 13 shows the beam end slip mea- Specimens CB1, CB2 and CB3 all had PPCDs. As listed in Table 1,
sured at different levels of mid-span bending moment. As shown, they had the same amount of reinforcement and same geometries
for a given mid-span bending moment, the slip at beam end varies except that their concrete slab thicknesses were in increasing order
from specimen to specimen. However, no significant differences in (130 mm, 140 mm and 150 mm, respectively). As listed in Table 4,
H. Hou et al. / Engineering Structures 126 (2016) 405416 413

(a) strain profiles across the entire beam section (b) strain profiles across the entire beam section
(at midpoint of CB7) (at midpoint of CB8)

(c) strain profiles across the concrete slab (d) strain profiles across the concrete slab
(at one loading point of CB7) (at one loading point of CB8)

Fig. 12. Comparison of strain profiles: Specimens CB7 vs. CB8.

the critical bending moment resistances, Mcr, My and Mu all


Table 5 increase slightly when the slab thickness increases, indicating that
Neutral axis location measured at the close-to-ultimate state. flexural resistance of a composite beam with PPCDs may benefit
Specimen Mur (kN m) Zur (mm) from the use of a thicker slab. To better understand the effect of
slab thickness on stiffness degradation, Fig. 14 shows the bending
CB1 225.0 67.63
CB2 225.0 75.71
moment-deflection curves of these three specimens. As shown, at
CB3 235.4 82.59 the early stage of the loading, all the curves have a very similar
CB4 225.0 68.23 slope, suggesting that slab thickness has a negligible impact on
CB5 225.0 65.40 specimen initial stiffness. However, when the loading is increased
CB6 225.0 67.47
beyond the limit associated with beam yielding and up to a mod-
CB7 236.3 63.79
CB8 202.5 72.62 erate level, Specimen CB2 which had the intermediate slab thick-
CB9 225.0 68.30 ness of 140 mm exhibits the most severe stiffness degradation. A
CB10 225.0 69.24 further examination on Specimen CB2 revealed that construction
CB11 225.0 67.14
defects existed at the joints between adjacent PPCDs, resulting in
CB12 225.0 68.49
414 H. Hou et al. / Engineering Structures 126 (2016) 405416

Table 6
Strain data at the pure-bending portion of the specimen under critical bending
moments (unit: 103).

Specimen Strain at Mcr Strain at My Strain at Mur


ect esb ect esb ect esb
CB1 0.667 1.301 0.872 1.800 2.439 4.271
CB2 0.718 1.377 0.919 1.800 2.683 4.96
CB3 0.753 1.556 0.947 1.800 2.711 4.611
CB4 0.845 1.639 0.928 1.800 2.883 a
CB5 0.918 1.436 1.068 1.800 3.01 6.085
CB6 0.764 1.553 0.870 1.800 2.326 4.573
CB7 0.732 1.543 0.825 1.800 2.562 5.234
CB8 0.712 1.382 0.856 1.800 2.464 4.726
CB9 0.930 1.491 1.092 1.800 2.412 4.439
CB10 1.017 1.522 1.137 1.800 2.535 4.586
CB11 0.834 1.917 0.725 1.800 2.536 a
CB12 0.754 1.402 1.110 1.800 2.736 4.537
a
Strain gauge failed prematurely.

Fig. 14. Moment deflection curves of specimens CB1, CB2 and CB3.

250
Mid-span bending moment (kN.m)

200

150

100

Specimen CB1 ( a = 40 mm)


Specimen CB4 ( a = 0 mm)
50 Specimen CB5 ( a = 20 mm)
Specimen CB6 ( a = 60 mm)
Fig. 13. Slip at beam end under different levels of mid-span bending moment.

a further reduced post-cracking flexural stiffness in the beam. 0


Focusing on the data from Specimens CB1 and CB3, it is found that 0.0 0.4 0.8 1.2 1.6 2.0
Specimen CB3 (which had a thicker concrete slab) exhibits a less Mid-span deflection normalized
significant degradation in flexural stiffness.
by beam span (%)

Fig. 15. Moment deflection curves of Specimens CB1, CB4, CB5 and CB6.
7.2. Influence of overlap width

slightly higher flexural resistance. As such, the overlap width of


As listed in Table 1, Specimens CB1, CB4, CB5 and CB6 had the
40 mm was kept in all the other specimens with PPCDs (see
same parameters except that the width of the overlap zone
Table 1). Note that the overlap width on the order of 40 mm was
between the PPCD and the top flange of the steel beam was varied
also recommended for the composite beams consisting hollow core
among these specimens. Fig. 15 compares the mid-span bending
slabs in recent research [16].
moment vs. mid-span deflection curves of these specimens. Among
these specimens, Specimen CB6 had the largest overlap width of
60 mm. As shown, this specimen exhibits the lowest initial stiff- 7.3. Influence of shear stud interval
ness. This is because a larger overlap width corresponds to a smal-
ler volume of cast-in-place concrete surrounding the shear studs As shown in Table 1, the only difference among Specimens CB1,
(see Fig. 5d), resulting in a higher level of flexibility of the shear CB7 and CB8 is that they had different shear stud intervals. Based
studs. Such an observation is consistent with the prior research on the original design, Specimens CB7 and CB8 had shear studs
on shear connectors in solid precast concrete planks [15]. Speci- more than and less than required, respectively; and Specimen
men CB4 represents an extreme case in which the PPCDs are not CB1 had the exact required amount of shear studs. Technically, a
overlapped on the steel beam. As shown in Fig. 15, this specimen fully composite beam would not exhibit increased ultimate flexu-
has the highest flexural resistance. However, it is recognized that ral resistance if interval of the shear studs reduces. The fact that
Specimen CB4 may be less practical since temporary construction the flexural strength of Specimen CB7 is higher than that of Spec-
aid is necessary when the overlap zone does not exist. Between imen CB1 reveals that the actual strength of the shear studs
Specimens CB1 and CB5, there are no remarkable differences in adopted in this investigation are lower than expected, leading to
their structural responses except that Specimen CB1 achieved a formation of the partially composite action in Specimen CB1 (and
H. Hou et al. / Engineering Structures 126 (2016) 405416 415

Fig. 16. Moment deflection curves of Specimens CB1, CB9 and CB10. Fig. 18. Moment deflection curves of Specimens CB1 and CB12.

7.5. Influence of rib orientation

Specimens CB1 and CB11 are identical except that their PPCD
ribs were oriented along two different orthogonal directions.
Fig. 17 shows the mid-span bending moment vs. mid-span deflec-
tion curves of these two specimens. As shown, although these two
specimens eventually achieved similar ultimate flexural resis-
tances, the flexural rigidity of Specimen CB11 is lower than that
of Specimen CB1. This is primarily due to the fact that the ribs
when oriented along the beam longitudinal direction do not pro-
vide sufficient doweling action for the cast-in-place concrete slab,
which could lead to sliding deformation of the cast-in-place con-
crete slab relative to the underneath PPCDs.

7.6. Influence of PPCDs

Specimens CB1 and CB12 allow a direct performance compar-


ison between the composite beams with a cast-in-place concrete
slab on PPCDs and a full-depth cast-in-place concrete slab.
Fig. 18 shows the mid-span bending moment vs. mid-span deflec-
tion curves of these two specimens. As shown, Specimens CB1 and
Fig. 17. Moment deflection curves of Specimens CB1 and CB11. CB12 exhibit the same stiffness up to the mid-span deflection asso-
ciated with the first yielding of steel beam. Beyond that point,
Specimen CB12 exhibits a much higher degree of stiffness degrada-
the other specimens with the same shear stud design). Neverthe-
tion in comparison with Specimen CB1. This is partially due to the
less, as shown in Table 4, a larger amount of shear studs tends to
fact that the compressive strength and modulus of elasticity of the
lead to higher values of Mcr, My and Mu. As shown in Fig. 13, for
concrete in PPCDs are higher than those of cast-in-place concrete
a given mid-span bending moment up to 225 kN m, the specimen
(see Table 2). At the limit state, Specimen CB12 has a higher
with a smaller shear stud interval consistently exhibits a smaller
mid-span deflection than Specimen CB1; however, their ultimate
slip deformation at its end, suggesting the beneficial contribution
flexural resistances are comparable. Therefore, strengthwise, per-
of shear studs in ensuring development of the composite beam
formance of Specimens CB1 and CB12 are similar.
action.

8. Conclusions
7.4. Influence of longitudinal reinforcement ratio
This research team conducted an experimental investigation on
Specimens CB1, CB9 and CB10 are only different in that their twelve steel-concrete composite beams. The tested specimens
longitudinal reinforcement ratios are in increasing order (0.29%, included eleven composite beams with cast-in-place concrete slabs
0.46% and 0.65%, respectively). Fig. 16 compares the mid-span on PPCDs and one with a full-depth cast-in-place concrete slab.
bending moment and mid-span deflection curves of these speci- Based upon the test results obtained, the following significant con-
mens. As shown, a specimen with a larger longitudinal reinforce- clusions were drawn:
ment ratio tends to be stiffer even after it is loaded beyond the
limit associated with yielding in the steel beam. However, it is  The composite beams consisting of cast-in-place concrete slabs
found that the ultimate flexural resistance does not differ signifi- on PPCDs overall exhibited desirable and stable flexural
cantly among these specimens. performance. At the ultimate state, failure mode of the beams
416 H. Hou et al. / Engineering Structures 126 (2016) 405416

is characterized by the crushing of concrete slab and yielding of However, any opinions, findings, conclusions and recommenda-
steel beam. The beam mid-span bending moment vs. deflection tions presented in this paper are those of the authors and do not
curves exhibit the typical bilinear flexure-dominated trend. Test necessarily reflect the views of the sponsors. Finally yet impor-
results confirm that a composite beam with cast-in-place con- tantly, the authors wish to thank three anonymous reviewers for
crete slab on PPCDs, if designed properly, can have the same ini- their careful evaluations and insightful comments that helped
tial flexural stiffness and ultimate flexural resistance as the one improve the paper.
having the same design parameters but a full-depth cast-in-
place concrete slab. References
 Concrete slab thickness when varying between 130 mm and
150 mm affects flexural stiffness degradation of the composite [1] Johnson RP. Composite structures of steel and concrete. 3rd
ed. Malden: Blackwell Scientific Publications; 2004.
beams with PPCDs but it has a negligible impact on initial flex- [2] Hsu C-TT, Punurai S, Punurai W, Majdi Y. New composite beams having cold-
ural stiffness and ultimate resistance of such beams. formed steel joists and concrete slab. Eng Struct 2014;71:187200.
 A 40 mm wide overlap zone between the PPCD and the top [3] Nie J, Fan J, Cai CS. Experimental study of partially shear-connected composite
beams with profiled sheeting. Eng Struct 2008;30:112.
flange of the steel beam was found to be adequate for composite [4] Ranzi G, Bradford MA, Ansourian P, Filonov A, Rasmussen KJR, Hogan TJ, et al.
beams consisting of PPCDs and it is recommended for practice. Full-scale tests on composite steel-concrete beams with steel trapezoidal
 The PPCDs with the ribs along the longitudinal direction of steel decking. J Constr Steel Res 2009;65:1490506.
[5] Tahmasebinia F, Ranzi G, Zona A. Probabilistic three-dimensional finite
beam, although do not seem to affect beam strength, tend to element study on composite beams with steel trapezoidal decking. J Constr
reduce flexural stiffness of the composite beam and therefore Steel Res 2013;80:394411.
should be used with caution in practice. [6] Nie J, Cai CS. Steel-concrete composite beams considering shear slip effects. J
Struct Eng, ASCE 2003;129:495506.
 Adding longitudinal rebars in the compression zone of a com-
[7] Nie J, Cai CS, Wang T. Stiffness and capacity of steel-concrete composite beams
posite beam with PPCDs tends to increase its initial flexural with profiled sheeting. Eng Struct 2005;27:107485.
stiffness but does not help significantly increase flexural resis- [8] Wu F, Huang H, Chen W. Experimental study and calculating methods on
bending rigidity of precast prestressed concrete ribbed panels for composite
tance of the composite beam.
slabs. J Hun Univ (Nat Sci) 2011;38:17.
[9] Grant JA, Fisher JW, Slutter RG. Composite beams with formed steel deck. AISC
Eng J 1977;12:2443.
Acknowledgements [10] GB50017. Code for design of steel structures. Beijing: Ministry of Housing and
Urban-Rural Development of the Peoples Republic of China; 2003.
[11] Specification for mix proportion design of masonry mortar (JGJ/T98-2010).
This research was supported by the 2014 University and Insti- Beijing: Ministry of Housing and Urban-rural Development of the Peoples
tute Innovation Plan of Jinan, the Fundamental Research Funds of Republic of China; 2010 [in Chinese].
[12] GB50010. Code for design of concrete structures. Beijing: Ministry of Housing
Shandong University China under Award No. 2014JC041, the and Urban-Rural Development of the Peoples Republic of China; 2010.
2014 Research Program of Shandong Province for Developing Inno- [13] Standard for test method of mechanical properties on ordinary concrete (GB/T
vative Wall Materials and Energy-Saving Construction Technology 500812002). Beijing: Ministry of Construction (MOC) & General
Administration for Quality Supervision, Inspection and Quarantine
under Award No. 139, the General Program of National Natural (GAQSIQ); 2003 [in Chinese].
Science Foundation of China under Award No. 51578324 and the [14] AISC. Load and resistance factor design specification for structural steel
Recruitment Program of Global Youth Expert. In addition, the third buildings. Chicago: American Institute of Steel Construction, Inc.; 1999.
[15] Moy SSJ, Tayler C. The effect of precast concrete planks on shear connector
author was supported by the Tom and Lucia Chou Fund. Further-
strength. J Constr Steel Res 1996;36:20113.
more, the authors wish to acknowledge the technical support from [16] Lam D. Capacities of headed stud shear connectors in composite steel beams
the Laboratory of School of Civil Engineering, Shandong University. with precast hollowcore slabs. J Constr Steel Res 2007;63:116074.

You might also like