Professional Documents
Culture Documents
Abstract: In the design of steel structures, composite effects of stud shear connectors are generally measured using ordinary push-out tests.
Furthermore, based on those results, the evaluation formulas of the ultimate shear strength are given in design guidelines. However, a concrete
slab is subjected to reversed stress during an earthquake, whereas existing tests consider only compressive stresses on concrete. The mechani-
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
cal behavior in existing structures thereby might be different from that under compressive force alone. This research proposes a component
model of a composite beam modeling the stress in actual buildings. Furthermore, cyclic loading tests were conducted on 14 specimens with
different specifications of the stud shear connector, concrete, and rebar. The results showed that the ultimate shear strength is considerably
lower than that under compressive stress. Consequently, this paper presents equations to assess structural performance precisely considering
various influential factors of composite structures. DOI: 10.1061/(ASCE)ST.1943-541X.0002294. © 2019 American Society of Civil
Engineers.
Author keywords: Composite beam; Concrete slab; Cyclic loading test; Headed stud; Ultimate shear strength.
Stud Rebar
which can result in a difference of the stress transfer between the
concrete slab and the stud shear connector in fully reversed cyclic
50
130
130
240
O
110
200
Tension
Compression
130
130
50
Compression Z Concrete
slab Z
150
Tension
Y X X Y
Fig. 1. Stress history of concrete slab: (a) positive bending; and Fig. 2. Component model in composite structure (unit: millimeters):
(b) negative bending. (a) side view; and (b) front view.
(a)
: Displacement gauge
Stud relative displacement, d
Headed stud Ref. point: H-section steel (upper)
Mes. location: H-section steel (lower)
Slip displacement
Slip displacement
of slab, ssl
of stud, ssc
Ref. point: Footing beam
Ref. point: H-section steel (upper) Mes. location: Concrete slab
Mes. location: Concrete slab
(b)
Fig. 3. Loading frame (unit: millimeters): (a) specimen setup; and (b) installation of displacement gauges.
The pitch of the studs and the slab thickness were determined expected that the tensile stress was transferred to the rebar, as re-
in accordance with the recommendations of JSSC (1996). The ported by Lin et al. (2013). Hence, the specification of rebar and its
surface of a jig contacting the concrete was coated with lubricat- allocation were assumed to be influential on the tension capacity.
ing oil before casting of the concrete to eliminate adhesion and The specifications of headed stud shear connectors and rebar were
friction. selected considering the practical construction and their availability
Fig. 3 depicts the loading frame. The specimen was placed on the market. Additionally, the loading protocol was included
widthwise on the footing beam. External force was applied using as an experimental parameter to clarify the influence of the fully
a horizontal jack with a capacity of 1,000 kN. Warping of the reversed stress on the concrete. In this experimental series, No. 2
concrete slab was constrained by steel plates and H-section steel was the reference specimen. The rebar arrangements in No. 8,
tied by the steel bars. No. 9, and No. 10 are shown in Fig. 4. The strain gauge attachments
A list of specimens is presented as Table 1. The 14 specimens are shown in Fig. 5. In the reference specimen (No. 2), two addi-
had 7 experimental parameters: (1) stud shank diameter (ϕsc ¼ 16, tional strain gauges were placed on the top and bottom of the stud
19, and 22 mm), (2) stud height (hsc ¼ 80, 100, and 130 mm), at x ¼ 60 mm.
(3) slab width (B ¼ 300, 400, and 500 mm), (4) concrete strength
(Fc ¼ 29.1, 38.7, and 64.8 N=mm2 ), (5) pitch of rebar (b ¼ 200,
400, and 200=400 mm), (6) diameter of rebar (ϕrb ¼ 6, 10, and Loading Protocols
13 mm), and (7) loading protocol (fully reversed cyclic or posi- The loading amplitude was controlled by the relative displace-
tively cyclic). The designation system is presented with Table 1. ment between the studs located at z ¼ 120 and −120 mm in
Rib heights of reinforcing bars were 0.6, 0.8, and 1.0 mm in 6-, Fig. 2 (hereinafter designated as stud relative displacement, d).
10-, and 13-mm-diameter rebar, respectively. Additionally, the The protocol gradually increased the loading as reversed cyclic
transverse bars were identical to the longitudinal bars in all spec- loading or positive cyclic loading. The increment was 0.2 mm
imens. Regarding the selection of the experimental parameters, it is up to d ¼ 1.0 mm and 0.5 mm over d ¼ 1.0 mm of stud relative
widely reported that the stud diameter, stud height, slab width, and displacement. Compressive and tensile stresses were applied, re-
compressive strength of concrete substantially affect the ultimate spectively, on the positive- and negative-side loadings. The speci-
shear strength of stud shear connectors in compressive loading men was pulled out after the final loading cycle (d ¼ −8.0 mm)
(AISC 2016; CEN 2004a; Tagawa et al. 1995). Furthermore, it was was completed.
50
50
130
130
130
200
200
200
240
240
240
110 110 110
200
200
200
130
130
130
50
50
50
0 150 200 150 0 150 200 150 150 200 150
500 500 500
Fig. 4. Arrangement of rebar (unit: millimeters): (a) No. 8; (b) No. 9; and (c) No. 10.
Strain gauge
ε ui − ε bi
εi = (i=1, 2, 3) Strain gauge
2
92.5
ε ε u1
130
ε u3 u2
ε b3 ε b2 ε b1
9 7 . 5 5 0 50 97.5
280
Strain gauge
120
500
500
15
120
57.5 15
95
92.5
95
10 120
130
47.5
x Side Front
Back Front
(a) (b) (c)
Fig. 5. Attachment of strain gauges (unit: millimeters): (a) stud and jig; (b) rebar; and (c) concrete slab.
stress and the ultimate stress of the H-section steel were, respec-
tively, 291 and 427 N=mm2 (Table 6). In this experiment, the axial Results of Cyclic Loading Tests of Component
strain of H-section steel did not exceed the yield axial strain in the Model
loading test.
Cyclic Behavior of Component Model of Composite
Beam
Table 2. Mix proportions
Unit material content (kg=m3 )
This section investigates the cyclic behavior and ultimate shear
strength of the component model of the composite beam based on
w∶c s∶a Water Cement Sand Gravel Admixture the results of cyclic loading tests. Figs. 6(a–n) portray the hysteresis
51.0 47.1 179 351 814 950 4.21 loops of the respective specimens. In larger stud relative displace-
41.5 45.5 170 410 775 960 4.71 ment, the initial gradient of load-displacement relation becomes
35.9 45.8 172 479 755 902 4.79 considerably small (e.g. d = –8.0 mm to 0 mm), while it gradually
increases after the stud shear connector contacts with the concrete
surface (e.g. d = 0 mm to 8.0 mm). Therefore, the specimens draw
Table 3. Material properties of concrete
the slip hysteresis loop under the fully reversed syclic loading.
Overall, the specimens had larger ultimate shear strengths in
w∶c Compressive strength Tensile strength Modulus of elasticity positive-side loading than in negative-side loading. The ultimate
(%) (N=mm2 ) (N=mm2 ) (N=mm2 ) strength in the negative side remained 27% (No. 11) to 49%
51.0 29.1 3.6 20,111 (No. 3) of that in the positive side because cracks originated in
41.5 38.7 4.2 24,051 the slab under tensile stress and the concrete lost strength during
35.9 64.8 5.7 33,877 negative-side loading. The largest ultimate shear strengths in both
the positive and negative sides were obtained in No. 13 [Fig. 6(m)],
in which the slab was high-strength concrete (Fc ¼ 65 N=mm2 ).
Table 4. Material properties of headed stud Furthermore, it should be noted that the shear strength under tensile
stress deteriorated rapidly, especially in No. 2, No. 6, No.7, No. 9,
Diameter Length Yield stress Ultimate stress Elongation No. 10, and No. 11, in contrast to that under positive-side loading.
(mm) (mm) (N=mm2 ) (N=mm2 ) (%)
This immediate degradation implies the rapid loss of the composite
16 130 411 473 33 effect between the steel beam and the concrete slab under negative
19 130 391 486 25 bending, even though the current design provisions do not differ-
22 80 398 461 25 entiate the capacity of stud shear connectors in terms of the stress
22 100 351 446 25 condition in each bending deformation (AISC 2016; CEN 2004a;
22 130 384 464 27
AIJ 2010a). Therefore, the mechanical performance under tension
loading needs to be scrutinized to secure the composite effect in the
structural design.
Table 5. Material properties of rebar Figs. 7(a–c) respectively portray the fracture processes of
No. 2 at d ¼ 0, −4.0, and −8.0 mm. Cracks occurred at the slab
Diameter Yield stress Ultimate stress Elongation
(mm) (N=mm2 ) (N=mm2 ) (%)
center and embedded positions of the studs. The crack width ex-
panded gradually with an increase in the stud relative displace-
6 360 506 31 ment (Fig. 7).
10 372 509 28 Figs. 8(a and b) respectively present the distribution of bending
13 350 493 26
strain of the stud in positive-side and negative-side loadings. The
horizontal axis was the position along the stud shear connector. The
bending strain was calculated by dividing the remainder of strains
Table 6. Material properties of H-section steel at the upper and lower sides of the stud by 2 [Fig. 5(a)]. Further-
Parameter Value
more, the position of the rebar and the yield bending strain εy;sc are
also depicted in Fig. 8. The bending strain of the stud reversed
Yield stress (N=mm2 ) 291 at around x ¼ 35 mm [Fig. 8(a)]. This double curvature originated
Ultimate stress (N=mm2 ) 427 from the constraint of the horizontal and rotational movement at the
Elongation (%) 44
head of the stud. In addition, the bending strain near the welded part
0 0 0 0
Q [kN] Ultimate strength Q [kN] Ultimate strength Q [kN] Ultimate strength Q [kN] Ultimate strength
600 600 600 600
0 0 0 0
Q [kN] Ultimate strength Q [kN] Ultimate strength Q [kN] Ultimate strength Q [kN] Ultimate strength
600 600 600 600
0 0 0 0
400 400
200 200
0 0
-200 -200
-10 -5 0 5 10 -10 -5 0 5 10
d [mm] d [mm]
(m) (n)
Fig. 6. Hysteresis curves: (a) No. 1; (b) No. 2; (c) No. 3; (d) No. 4; (e) No. 5; (f) No. 6; (g) No. 7; (h) No. 8; (i) No. 9; (j) No. 10; (k) No. 11; (l) No. 12;
(m) No. 13; and (n) No. 14.
ε [%] ε [%]
0.2 Rebar 0.2 Rebar
εy εy
0 0
εy εy
-0.2 -0.2
0 20 40 60 80 100 0 20 40 60 80 100
(Welded) x [mm] (Welded) x [mm]
(a) (b)
crack crack
tension
softening
(a) (b) (c)
Fig. 9. Influence of reversed cyclic loading: (a) first positive side loading; (b) following negative-side loading; and (c) following positive-side
loading.
(x ¼ 15 mm) exceeded the yield bending strain at d ¼ 1.0 mm. 371 kN, whereas that of No. 1 was 436 kN in the ultimate state.
The bending strain of the stud in the negative-side loading was This deterioration originated by the loading protocol was inferred
much smaller than that in the positive-side loading [Fig. 8(b)]. as starting from the cracking of the concrete occurring in the
This magnitude relation was the same as the hysteresis loop in negative side. It decreased the normal force in the following
Fig. 6(b). positive-side loading. The mechanism of degradation is presented
The loading protocol influence on the cyclic behavior and ulti- in Figs. 9(a–c). In the first positive-side loading, the stud shear
mate shear strength can be refined based on the results of No. 1 connector received normal force N s;f from the concrete near the
[Fig. 6(a)] and No. 2 [Fig. 6(b)]. The component model, which welded part and the headed part [Fig. 9(a)], although the opposite
was subjected to the positive cyclic loading, showed no pinching direction of normal force, N s;b , supported the stud at the headed
hysteresis [Fig. 6(a)]. The ultimate shear strength in the positive part. In the following negative-side loading, the concrete began
side in No. 2, which was under fully reversed cyclic loading, was to crack and gradually lost its tensile strength [Fig. 9(b)]. Therefore,
characteristic slip capacity does not exceed 6.0 mm. Moreover, the slab. In addition, the height of the stud in No. 2 was the same as
slip performance of the component model in this research was the embedded position of the rebar in the back side of the slab.
lower than that of the ordinary push-out specimens, which usually Therefore, the longer stud (No. 2) smoothly transferred the stress
reach ultimate shear strength at 30% of the stud diameter (JSCE to the rebar, in contrast to No. 5 and No. 6. The ductility thereby
2014). Civjan and Singh (2003) reported that the structural perfor- differed with the same stud shank diameter. However, the ultimate
mance of stud shear connectors decreases considerably under cyclic strengths in the negative side were still almost identical, irrespective
loading. The component models in this research also give the deg- of the stud height. In addition, the stud relative displacement at
radation of slip capacity, as Civjan and Singh (2003) demonstrated the ultimate shear strength remained at −3.9 to −4.4 mm in No. 2,
in earlier experiments. The hysteresis curves of No. 1 and No. 2 No. 5, and No. 6.
were almost identical in the small loading amplitude under com- The slab width influence can be understood by Figs. 6(b, g,
pressive stress, although the slab in No. 2 was subjected to tensile and h). The ultimate shear strength in the positive side was consid-
stress in the negative-side loading, which indicates that the relation erably lower in No. 7 [Fig. 6(g)], with 300-mm slab width, whereas
between the shear force and stud slip displacement in the positive the discrepancy of the ultimate shear strength in the positive side
side does not differ even under different loading protocols up to did not appear when the slab width was greater than 400 mm
d ¼ 1.5 mm (6.8% of the stud diameter). This proves that the in- [Figs. 6(b and h)]. However, the stud relative displacement at the
fluence of concrete damage on the shear force in the positive side is ultimate shear strength increased slightly with slab width. A similar
not prominent at small loading amplitudes, although it is indispen- trend in the ordinary push-out test was reported in earlier research
sable in the ultimate state. The preceding discussion proves the ne- (Tagawa et al. 1995). Tagawa et al. (1995) concluded that the slab
cessity of incorporating the interrelation of the mechanical behavior width influence on the mechanical properties of stud shear connec-
in the positive and negative sides in the evaluation process. tors almost vanishes if the slab has a width greater than or equal to
In terms of the stud shank diameter, the ultimate shear 400 mm. The ultimate shear strength in the negative side exhibits
strength in the positive side enlarges with larger stud diameter a moderate positive relation with the slab width. This enhancement
[Figs. 6(b–d)]. The following reasons are inferred: (1) the increase is assumed to derive from the larger cross-sectional area of the con-
of the stud stiffness transfers greater stress to the slab; and (2) the crete slab.
wider aspect area carries larger normal force to the slab. The ulti- In terms of the pitch and number of rebar, Figs. 6(h, i, and j)
mate shear forces in the positive side were 225 kN in No. 3 (ϕsc ¼ depict different performance in each specimen. The largest ultimate
16 mm), 293 kN in No. 4 (ϕsc ¼ 19 mm), and 371 kN in No. 2 shear strength in the negative side loading was obtained in No. 10
(ϕsc ¼ 22 mm). The ultimate shear strength was positively propor- [Fig. 6(j)], with 16 rebar in total. However, the enlargement of
tional to the stud shank diameter on the positive side. Moreover, the No. 10 was only 7% in the positive side and 6% in the negative
stud relative displacement at the ultimate shear strength increased side compared with No. 8, whereas the total cross-sectional area of
with smaller stud diameter. Presumably, the concrete damage did rebar doubled, which suggests that the outer rebars do not contrib-
not become severe because the stud absorbed the deformation at ute to the ultimate shear strength as much as their expected yield
the same loading amplitude. However, the ultimate shear forces in axial strength. Figs. 10(a–c) depict the stress distribution of rebar
the negative-side loading showed no specific trend, unlike those embedded in the concrete slab under negative-side loading in
0.18 kN 0.37 kN 0.34 kN 0.01 kN 5.93 kN 12.6 kN 5.66 kN 4.57 kN 7.50 kN 15.1 kN 6.36 kN 7.30 kN
(0.7%) (1.4%) (1.3%) (0.1%) (22.4%) (47.6%) (21.3%) (17.2%) (28.3%) (60.0%) (24.0%) (27.5%)
5.3 kN
(20%)
13.3 kN
(50%)
26.5 kN
0.37 kN 0.81 kN 0.88 kN 0.33 kN 13.5 kN 26.5 kN 26.5 kN 12.5 kN 21.0 kN 26.5 kN 26.5 kN 16.5 kN (100%)
(1.4%) (3.1%) (3.3%) (1.2%) (51.1%) (100%) (100%) (47.0%) (79.4%) (100%) (100%) (62.2%)
(a) (b) (c)
Fig. 10. Stress distribution under negative-side loading (No. 10): (a) d ¼ −1.0 mm; (b) d ¼ −2.5 mm (ultimate shear strength); and
(c) d ¼ −4.0 mm.
No. 11, and No. 12 [Figs. 6(b, k, and l)]. The largest ultimate shear strength in the positive-side loading relates to the compressive
strength in the positive side loading in No. 2 became slightly larger strength of concrete, whereas that in the negative loading is affected
than that in No. 12, whereas No. 12 had the largest ultimate shear by the bond strength between the concrete and reinforcing bars. It is
strength in the negative side loading. It might be inferred that widely recognized that the strengths possesses a positive relation-
the specimen with higher-gauge rebar has greater shear strength. ship with each other (ACI 2003; CEN 2004b; AIJ 2010b). Hence, it
However, the opposite result was demonstrated between No. 2 was expected that the ultimate shear strengths in the positive side
and No. 12. The reversal of the ultimate shear strength in positive and negative side would demonstrate a positive proportional rela-
side between No. 2 and No. 12 is explainable through the concrete tionship. The ultimate shear strength in No. 13 therefore located in
damage in negative-side loading. No. 12, in which the rebar were the upper right in Fig. 11. Furthermore, the stud shear connectors
13 mm in diameter, resisted greater shear force in the negative side, with smaller diameters (No. 3 and No. 4) placed comparatively on
even with a small loading amplitude [Fig. 6(l)]. However, a wide the upper right. As described in the comparison of the stud diam-
crack suddenly appeared to release the larger fracture energy at eter, the compressive stress on the concrete is not localized due to
d ¼ −1.3 mm. Consequently, comparatively little stress in the pos- larger flexibility of slender stud shear connectors in the positive
itive side was carried to the concrete slab even in the positive-side loading. In addition, the ultimate shear strength in the negative side
loading. The ultimate strength in the positive side thereby degraded loading did not differ drastically depending on the diameter of stud
even with larger cross-sectional area of rebar. In No. 11, in which shear connectors. The ultimate shear strength per cross-sectional
the rebar diameter was only 6 mm, the ultimate shear force was area therefore increased with smaller stud diameter. Additionally,
markedly lower on the negative side. ACI (2003) reported that the in Fig. 11, the specimens with a greater number of reinforcing bars
majority of stress transfer between the concrete and the reinforcing (No. 10) or a larger diameter of rebar (No. 12) located relatively to
bars is demonstrated by the bearing of ribs. Furthermore, the other the upper right, whereas that with a coarse arrangement of longi-
contributors (adhesion and friction) gradually diminish with an tudinal bars (No. 11) placed in the lower left. This indicates that the
increase in the slip at the interface. The rib height of reinforcing deterioration mechanism portrayed in Fig. 9 governed the ultimate
bars with 6-mm diameter was the smallest in the experimental shear strength of the component model in the composite beam. In
series, resulting in the rapid degradation of bond strength between summary, the positive proportional relation was demonstrated in
the rebar and concrete. Therefore, the ultimate shear strength and this experimental series, as illustrated in Fig. 11. Furthermore, the
ductility of No. 11 was much less than that of the other specimens. ultimate shear strengths in No. 2, No. 7, No.8, No. 9, No. 10,
The concrete strength was the most crucially important factor No. 11, and No. 12 were identical with the conventional evaluations
affecting shear strength in this experimental series. The ultimate in the AISC specification (AISC 2016), Eurocode 4 (CEN 2004a),
shear force of No. 13, with a concrete strength of 65 N=mm2 ,
was 491 kN in the positive side and −235 kN in the negative side,
whereas those in No. 2 were, respectively, only 371 kN (76%) and
No. 2 3 4 5 6 7 8 9 10 11 12 13 14
−120 kN (51%) [Figs. 6(b and m)]. Enhancement of the ultimate
Symbol
shear strength issued from the high compressive strength of con-
crete is widely described in the literature (e.g., Ollgaard et al. 1971;
qmax /Asa (Negative)
Li and Cederwall 1996; Luo et al. 2016). Regarding the enlarge- 160
ment of the ultimate shear strength in negative-side loading, the
concrete material gradually loses tensile strength with crack expan- 140
sion during cyclic loading. Consequently, the increase of the tensile 120
strength is not a direct factor creating larger ultimate shear strength
in tension loading. Rather, it is assumed to be achieved because 100
of larger bond force between rebar and concrete, which enables the
stress to be transferred between the stud and rebar during negative- 80
side loading.
60
The effectiveness of rebar is presented clearly in Fig. 6(n). The
specimen without reinforcement experienced overall fracture at 150 200 250 300 350
d ¼ −0.5 mm and immediately lost shear strength, which implies qmax /Asa (Positive)
the necessity of the rebar to secure the composite effect between the
Fig. 11. Comparison of ultimate shear strength of positive and
steel beam and concrete slab, particularly under negative bending.
negative sides.
Additionally, the maximum contribution of concrete on the shear
ε rb /ε rb,y [-] Concrete crack ε rb/ε rb,y [-] Concrete crack ε rb/ε rb,y [-] Concrete crack ε rb/ε rb,y [-] Concrete crack
1.5 1.5 1.5 1.5
1 1 1 1
0 0 0 0
Fig. 13. Strain of rebar: (a) front (left side, z ¼ 147.5 mm); (b) front (left side, z ¼ 50 mm); (c) back (left side, z ¼ 147.5 mm); and (d) back
(left side, z ¼ 50 mm).
Fig. 14. Slip behavior: (a) headed stud; and (b) concrete slab.
pffiffiffiffiffiffiffiffiffiffiffi
0.8Fu πϕ2sc =4 0.29αϕ2sc Fc Ec
qmax ¼ Min ; ð2Þ
γν γν
Comparison with Prevailing Design Codes 8
>
< 0.2 hsc þ 1 ð3 ≤ hsc =ϕsc ≤ 4Þ
Comparison with Evaluation Formula in AISC α¼ ϕsc ð3Þ
>
:
AISC determines the ultimate shear strength of one headed stud 1.0 ðhsc =ϕsc > 4Þ
anchor, qmax , embedded in a solid concrete slab by (AISC 2016)
where γ ν = partial factor. Eurocode 4 (CEN 2004a) recommends
pffiffiffiffiffiffiffiffiffiffiffi
qmax ¼ 0.5Asc Fc Ec ≤ Rg Rp Asc Fu ð1Þ 1.25 for the value of γ ν .
Actually, the evaluation formula in Eurocode 4 is much more
where Fu = specified minimum tensile strength of steel-headed stud conservative than that in AISC (2016). The ultimate shear strengths
anchor. In the component model in this research, Rg and Rp are 1.0 determined by concrete crush and stud failure were, respectively,
and 0.75 because the headed studs are welded directly to the steel 58% and 85% of Eq. (1). Figs. 16(a and b) respectively compare
shape (AISC 2016). the experimentally obtained results with Eurocode 4’s equation in
Figs. 15(a and b) compare the experimentally obtained results the positive side and negative side. Fig. 16(a) shows good agreement
with the evaluation equation in the AISC specification (AISC except for No. 14, which was not reinforced by rebar. In addition,
2016). The horizontal axis is the square root of the product of the experimentally obtained results were assessed conservatively,
Fc and Ec . The vertical axis is the quotient of the shear strength except that of No. 7, which had a 300-mm slab width. Therefore,
per stud and the cross-sectional area of the stud, qmax =Asc . In ad- it can be inferred that the ultimate shear strengths in the positive
dition, Shimada et al. (2016) constructed a database of previous side are roughly evaluated by Eqs. (2) and (3). However, the differ-
ordinary push-out tests. The results of Shimada et al. (2016) are ence of the ultimate shear strength from the specification of rebar
also presented in Figs. 15(a and b) to clarify the difference from diameter and its allocation, which are the influential factors on
conventional specimens. the tension capacity, is not demonstrated by the evaluation formula.
The ultimate shear strengths in the negative side did not exceed those
of the formula [Fig. 16(b)]. The experimental values remained
at 24% at minimum (No. 14) and 61% at maximum (No. 3), which
No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Symbol
Push-out test AISC’s equation
Shimada (2016) φsc=16 mm φsc=19 mm φsc=22 mm
hsc=130 mm hsc=130 mm hsc=80 mm hsc=100 mm hsc=130 mm No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Symbol Symbol
300 300
400 400
200 200
100 100
100 100
0 0 0 0
0 300 600 900 1200 1500 0 300 600 900 1200 1500 0 100 200 300 0 100 200 300
Fc Ec Fc Ec qmax/Asa (Eurocode-4) qmax/Asa (Eurocode-4)
(a) (b) (a) (b)
Fig. 15. Comparison of experimental results and AISC’s equation: Fig. 16. Comparison of experimental results and Eurocode 4’s
(a) positive; and (b) negative. equation: (a) positive; and (b) negative.
1 0.8
This section derives the evaluation formulas of the ultimate shear Eq. (3) Eq. (8)
strength in the positive and negative sides to assess the performance 0.9 0.4
accurately. 0
0.8
Eurocode 4 (CEN 2004a) considers the ratio of stud height, hsc , 2 3 4 5 6 7 8 9 10 0 0.001 0.002 0.003 0.004
to shank diameter, ϕsc , as a coefficient, α. However, the value of α hsc /φ sc 1/φsc2
becomes a constant, 1, when hsc =ϕsc > 4, which means that the (a) (b)
effect of the aspect ratio of a headed stud is neglected. This presump-
tion designates that the ultimate shear strength divided by the cross- Fig. 17. Coefficients αþ and α− arranged by hsc =ϕsc and 1=ϕ2sc :
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
sectional area is assessed as equal in the same concrete property (a) positive; and (b) negative.
when stud shear connectors become narrow. However, Figs. 15
and 16 demonstrate that the influence of hsc =ϕsc should be refined
based on the experimentally obtained results. Based on the preced- By contrast, the coefficient in the negative side, α− , is positively
ing discussion, this research elaborates the coefficient to define the proportional to the inverse of ϕ2sc . This clear relation is given by the
effect of aspect ratio in the positive side, αþ , referring to the exper- consistency of the ultimate shear strength irrespective the specifi-
imentally obtained results. Coefficient αþ is back-calculable using cation of a stud under tensile stress, as described in this paper.
Eq. (4) by solving Eq. (2) for α. Here, qþ max denotes the ultimate Therefore, assessing α− by the stud aspect ratio is inadequate be-
shear strength per stud shear connector on the positive side cause the stud height is not a determinant of the performance in
negative-side loading. Additionally, constructing the evaluation for-
γ ν qþ mula of α− employing 1=ϕ2sc enables a reduction of the stud shank
αþ ¼ p ffiffiffiffiffiffiffiffiffiffiffi
max
ð4Þ
0.29ϕ2sc Fc Ec diameter term from Eq. (2). This research thereby proposes the pre-
diction model of α− as Eq. (8) through single regression analysis.
The ultimate shear strength in the negative side does not relate The intercept is intentionally fixed as 0 to make the regression equa-
directly to the compressive strength of concrete. In addition, the tion pass the origin. However, it has sufficient accuracy
concrete loses its tensile strength in the ultimate state. Therefore,
550
substituting the compressive strength in Eq. (4) with the tensile α− ¼ ð8Þ
strength is inadequate for the evaluation. Instead, this research used ϕ2sc
the bond strength, Fbd , between the concrete and rebar. The bond
strength of deformed steel bars is determined using Eq. (5) (AIJ Prediction of the ultimate shear strength in the positive side,
2010b). Furthermore, the coefficient in the negative side α− is qþ 0 −0
max (Eva.), and the negative side, qmax (Eva.), might be imple-
back-calculable by Eq. (6). Here, q− max denotes the shear strength
mented by substituting Eqs. (7) and (8) in Eq. (2). The ultimate
per stud shear connector on the negative side shear strengths in the positive and negative sides are interrelated
under the reversed cyclic loading (Fig. 11). For that reason, it is
1 1 necessary to consider cracks in concrete, which weaken the stress
Fbd ¼ Min Fc ; 1.35 þ Fc ð5Þ transfer between a stud and concrete. Accordingly, the evaluation
10 25
formula of the ultimate strength in the negative side is constructed
γ ν q− first because it is the critical mechanism in this experimental series.
α− ¼ p ffiffiffiffiffiffiffiffiffiffiffiffiffi
max
ð6Þ It is readily apparent that the ultimate shear strength in the negative
0.29ϕ2sc Fbd Ec side deeply relates to the rebar allocation and its diameter. Fig. 10
shows that the axial force of rebar is not uniform in the cross section
Fig. 17 presents the obtained coefficients αþ and α− arranged
of a concrete slab. The rebar near the edge of slab has roughly 50%
by the stud aspect ratio hsc =ϕsc in Fig. 17(a) and the inverse of ϕ2sc
of the yield axial force in No. 10. Consequently, this study makes
in Fig. 17(b). The displayed specimens in Fig. 17 are No. 2, No. 3,
the following assumptions in the evaluation: (1) steel bars arranged
No. 4, No. 5, and No. 6, which possessed the same concrete proper-
with 200-mm pitch contribute 100% of their cross-sectional area;
ties but different specifications of studs. In Fig. 17(a), the coeffi-
and (2) steel bars arranged with 400-mm pitch contribute 50% of
cients exceeded Eq. (3) in hsc =ϕsc > 4 and the value of the
their cross-sectional area. The effective cross-sectional area of rebar
coefficient increased up to 1.25 in hsc =ϕsc ¼ 8.1, which reveals
is calibrated based on these presumptions. Fig. 18(a) depicts the
that the guidance of Eurocode 4 (CEN 2004a) secures conservative
correction factor, β − ¼ q− −0
max ðExp:Þ=qmax (Eva.), arranged by the
assessments in spite of its inaccuracy. Fig. 17(a) clearly shows that
yield axial strength of the effective cross-sectional area of rebar,
the coefficient in the positive side, αþ , is positively proportional to
fy;ef . Here, q−max (Exp.) is the experimental value in the negative-
hsc =ϕsc . Consequently, the following formula can be derived by
side loading. For simplicity, β − obtained in No. 2, No. 9, No. 10,
single regression analysis. This study inherits the existing function
No. 11, No. 12, and No. 14 are displayed in Fig. 18(a). Fig. 18(a)
of Eurocode 4 (CEN 2004a) in hsc =ϕsc ≤ 4
shows that the correction factor is generally positively proportional
8 to the effective yield strength of rebar, fy;ef (unit: kN). Therefore,
>
> hsc
>
< 0.2 þ 1 ð3 ≤ hsc =ϕsc ≤ 4Þ the correction factor is induced through the single regression
þ
ϕsc
α ¼ ð7Þ analysis as
>
>
> 0.055 hsc þ 14.2
: ðhsc =ϕsc > 4Þ
ϕsc β − ¼ 0.002f y;ef þ 0.78 ð9Þ
fy,eq [kN] Fig. 19(b) compares the evaluated and experimentally obtained
(a) (b) results. The accuracy of prediction is improved compared with
Fig. 16(a), although Eurocode 4 (CEN 2004a) tends to make
Fig. 18. Evaluation of the ultimate shear strength in the negative side: conservative assessments. However, the proposed formulas of
(a) reduction factor; and (b) comparison. the ultimate shear strength adequately include the influential factor
considering the interrelation between the positive- and negative-
side loadings, which represents the genuine mechanical behavior
of the component model of the composite beam.
Consequently, the ultimate shear strength is predicted by In this research, the specimens in the experimental series
Eq. (10) with Eq. (9). In Eq. (10), the term of the stud shank diam- covered hsc =ϕsc ¼ 3.6–8.1 in the stud specification. The applicable
eter has already been reduced through the derivation process. scope range of the proposed evaluation formulas is therefore
Fig. 18(b) shows that the proposed formulas precisely evaluate defined within hsc =ϕsc ¼ 3.6–8.1.
the experimentally obtained results
pffiffiffiffiffiffiffiffiffiffiffiffiffi Conclusion
0.8Fu πϕ2sc =4 159.6β − Fbd Ec
q−
max ¼ Min ; ð10Þ
γν γν This study demonstrated cyclic loading tests of a component
model of a composite beam with several influential factors on
Finally, the ultimate shear strength in the positive side is the mechanical performance. Based on the experimentally obtained
calibrated considering the influence of reversed cyclic loading. results, the cyclic behavior and stress transfer between the headed
As Fig. 11 suggests, damage to the concrete slab strongly affects stud and concrete slab were refined. In addition, the applicability
the structural performance in the positive side, particularly in of prevailing equations to assess the shear strength of headed stud
specimens with low ultimate shear strength in the negative side. connectors was investigated in terms of those under compressive
Based on the consideration presented previously, the correction and tensile stresses. The findings are summarized as follows:
factor in the positive side, β þ ¼ qþ þ0
max ðExp:Þ=qmax (Eva.), was com- 1. The ultimate shear strengths under tensile stress were 27%–49%
pared with the evaluated ultimate shear strength in the negative of those under compressive stress. Those under tensile stress did
side, where qþ þ0
max (Exp.) and qmax (Eva.) respectively represent the not differ drastically depending on the stud specifications.
experimental and evaluated values in the positive side loading 2. The normal force of concrete decreases with crack expansion
during cyclic loading. Therefore, the ultimate shear strength de-
grades when the slab is subjected to fully reversed cyclic loading
compared with that under positive cyclic loading.
No. 2 3 4 5 6 8 9 10 11 12 13
3. The ultimate strength of the component model of a composite
Symbol
beam was below that of the evaluation formula [Eq. (1)] in the
β+ qmax +/Asa (Exp.) AISC specification. Specifically, the value in the negative side
1.2 350 was only 17%–44% of the results predicted by the guideline.
4. Eurocode 4 [Eq. (2)] roughly predicts the ultimate shear strength
Eq. (11)
1.1 of a composite beam subjected to fully reversed cyclic loading.
300 However, the ultimate shear strength under tensile stress was
1
much lower than the assessed value of Eurocode 4.
5. The derived equations considering the interrelation between
Wide cracks 250 compressive and tensile sides can appropriately predict the
0.9
ultimate shear strength of the component model of a composite
beam.
0.8 200 In future research, a finite-element model will be constructed
80 100 120 140 160 180 200 200 250 300 350
−
qmax (Eva.) [kN] qmax+/Asa (Eva.) based on the experimental results in this paper. Furthermore, a com-
prehensive parametric study will be demonstrated to interpolate and
(a) (b) supplement the influential parameters of the experiment. The effec-
tive axial strength of rebar and the mechanism of bearing force be-
Fig. 19. Evaluation of ultimate shear strength in the positive side:
tween stud shear connectors and concrete slab will be presented as
(a) correction factor; and (b) comparison.
a continuous function through the calibrated results.