Professional Documents
Culture Documents
Abstract: This paper investigated the confinement mechanism of steel-confined concrete in steel-reinforced concrete (SRC) columns. Six
axially loaded SRC-core columns were tested and the corresponding structural performance data were collected and assessed. Considering the
confinement mechanism of both steel-confined and hoop-confined concrete, the concrete in the SRC columns was divided into four regions:
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
highly steel-confined concrete (HSCC), partially steel-confined concrete (PSCC), partially confined concrete (PCC), and unconfined concrete
(UCC). Based on Mander’s model, a theoretical stress–strain model for each concrete region was proposed and implemented in an analytical
model to predict the mechanical behaviors of axially loaded SRC columns from the results of the tests described in this paper and of previous
tests; satisfactory agreement was observed. DOI: 10.1061/(ASCE)ST.1943-541X.0002289. © 2019 American Society of Civil Engineers.
Author keywords: Steel-confined concrete; Confinement mechanism; Region partition; Stress–strain model.
Fig. 1. Regions for the concrete in SRC columns: (a) Type I, UCC and confined concrete; and (b) Type II, UCC, PCC, and HCC.
and cross-shaped steel profiles has gradually increased. Therefore, basis of the confinement mechanism of both steel-confined and
the confinement mechanism of steel with arbitrary sections of hoop-confined concrete. Finally, these stress–strain relationships
confined concrete requires further study. Zhao and Wen (2018) ini- according to the proposed methodology were implemented in the
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
tiated a research program focused on the confinement mechanism analytical model proposed by Chen and Lin (2006) and Chen
of steel-confined concrete and developed a stress–strain model and Wu (2017), and the experimental and predicted axial load–
for steel-confined concrete by modifying the key parameters of displacement relationships of the axially loaded SRC short columns
Mander’s model; this modified model yielded satisfactory results were compared to validate the theoretical stress–strain model.
compared with experimental results (Zhao et al. 2014). However,
the material strength of the specimens was normal. Considering the
extensive use of high-strength materials, the confinement mecha- Experimental Program
nism in high-strength steel–concrete composite columns is worth
studying.
Specimen Design
Clearly, if the confinement mechanism of both the steel-confined
and hoop-confined concrete can be understood and properly consid- Six high-strength SRC-core short columns were tested under axial
ered, a theoretical stress–strain model for concrete in SRC columns compression. Because this experiment investigated the confine-
can be proposed. The confinement mechanism of hoop-confined ment mechanism of steel-confined concrete, the influence of hoop-
concrete was intensively studied by Kent and Park (1971), Sheikh confined concrete was eliminated and the specimens were designed
and Uzumeri (1982), Scott (1982), Park et al. (1982), Mander et al. to consist of only two components: a structural steel section and
(1988), and Cusson and Paultre (1995). In contrast, limited analyti- concrete. For the specimens in Zhao et al. (2014), the grade of
cal studies (El-Tawil and Deierlein 1996; Chen and Wu 2017) have the concrete ranged from C35 to C50 and the grade of the structural
been conducted on the confinement mechanism of steel-confined steel section ranged from Q235 to Q345. To reflect the high-
concrete. Therefore, the current stress–strain model for the concrete strength materials in this study, the grade of concrete ranged from
in SRC columns, especially for HCC, may not be appropriate. The C60 to C70 and the grade of the structural steel section ranged
stress–strain model for the concrete in SRC columns also requires from Q390 to Q460. Considering the influential factors of the con-
further study. finement on the concrete, i.e. the area ratio of steel to concrete, the
In this paper, an experimental study was carried out on six confined dimensions of the steel flange, and the strength combina-
axially loaded high-strength SRC-core columns to investigate the tion of the steel and concrete, the configurations of the specimens
confinement mechanism of steel-confined concrete. Then, consid- were designed as shown in Fig. 2, where lf is the clear width of the
ering the confinement mechanism of both steel-confined and hoop- steel flange and also the confined dimension of the steel flange, lc is
confined concrete, the concrete in the SRC columns was divided the dimension of the steel flange to its opposite steel web, tw is the
into four regions according to the confinement level, and a theo- thickness of the steel web, tf is the thickness of the steel flange, and
retical stress–strain model for each concrete region was proposed bf is the width of the steel flange. The height of the specimens was
by considering the model of Mander et al. (1988). This model has 1,200 mm. The material properties of the structural steel shapes,
wide application and reliability according to the theoretical including the elastic modulus (Es ), the yield strength of the steel
lc
lf
Steel confined
concrete
tf
tw
bf
Shaped
steel
web and flange (fy;w and f y;f , respectively), the ultimate tensile attached to six horizontal aluminum bars, were used to measure the
strength of steel web and flange (f u;w and fu;f , respectively), horizontal strains of the concrete. The 11 strain rosettes, S18–S28,
the percentage elongation of structural steel (δ 10 ), and the cylinder were used to trace the strain distribution of the steel flange; S38–
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
compressive strength of concrete (f c0 ), are listed in Table 1. The S41 were used to measure the vertical and lateral strains of the con-
aluminum bars were cast within the concrete inside the specimens crete surface; and Gauges S29–S37 and S42–S51 were designed to
to measure the strain of internal concrete. The diameter of the alu- not only obtain the vertical strains of the structural steel section but
minum was 4 mm. The strength corresponding to a residual strain also ensure that the application of the axial load was concentric.
of 0.002 was 242 MPa and the ultimate strength was 320 MPa.
Each specimen was labeled in the form SRC-i-j, where i is the
type of structural steel and j represents the grade of the material Experimental Results
strengths. For instance, the label SRC-A-1 refers to the specimen
with a type-A encased structural steel section for which the grade of General Behavior
the concrete is C60 and the grade of the steel is Q390.
Two idealized axial load–displacement relationships are presented
in Fig. 4(a) as Curve A and Curve B, and the test results of axial
Test Setup and Loading System load–displacement relationships are shown in Fig. 4(b). Curve A
Fig. 3(a) presents the experimental setup and loading system. The and Curve B are the same before Point V. Point Y refers to the yield
test used an FCS-30000 kN (Research Laboratory of Earthquake load, N y , which is calculated as N y ¼ f y · As þ fc0 · Ac , where fy
Engineering, Tongji University, Shanghai, China) electrohydraulic is the yield strength of the structural steel section, As is the cross-
servo coordinate loading system with a load capacity of 30,000 kN sectional area of the structural steel, fc0 is the cylindrical compres-
and 400 mm maximum displacement in the vertical direction. The sive strength of the UCC, and Ac is the cross-sectional area of the
two end plates of the specimens were fixed to the testing machine. concrete. Point P is the peak point and indicates that the whole sec-
The upper endplate was restrained against all degrees of freedom, tion of the specimen is in a plastic state and the concrete confine-
and the lower endplate, the loading end, could move only in the ment reaches its maximum value. Along Curve B, the load starts
vertical direction during the loading process. to decrease at a constant rate after Point P because the load increase
The displacement-controlled method was adopted as the loading from the structural steel strengthening is less than the load re-
method in this test. The loading rate was changed from 0.6 to duction due to the degradation of the concrete. In contrast, along
3 mm=min when a decrease in the measured axial load was ob- Curve A, the axial load will not decrease until it reaches the bal-
served during the loading process. The application of the load was anced load N v , which indicates that the effect on the capacity of the
stopped when the vertical displacement of the specimens exceeded specimen achieves a balance between the degradation of the con-
60 mm, which was regarded to indicate specimen failure. crete and strengthening of the steel; then the load will increase be-
cause the effect of the steel strengthening will be superior to that of
the concrete degradation after Point V.
Measurement Arrangement
Specimen SRC-A-1 is used as an example to illustrate the
For each specimen, the displacement of the specimen and the representative phenomenon during each stage of the curve.
strains of the structural steel shapes and concrete were continuously 1. Elastic stage (OY): Both the structural steel shapes and concrete
measured and recorded during the loading process. The arrange- exhibited linear behavior in the early stage of the loading pro-
ment of the displacement transducers is presented in Fig. 3(b). cess, and the load–displacement curve was linear prior to the
Transducers D1–D4 were placed on four sides along the height of yield load N y . No cracks were observed on the concrete surface
the specimen to measure the overall vertical displacement, which [Fig. 5(a)].
was also used to verify the concentricity of the applied axial load 2. Elastoplastic stage (YP): The load–displacement curve became
during the early loading process. Transducers D5–D16 were used nonlinear after the load reached Point Y because of the non-
to measure the horizontal displacement of the steel flange at the linear characteristics of the materials. The concrete confinement
midheight of the specimen to investigate the out-of-plane bending provided by the structural steel shapes achieved its maximum
deformation of the steel flange due to the horizontal expansion of value when the load reached Point P. Several vertical cracks
the concrete under compression. were observed near the midheight of the specimen [Fig. 5(b)]
The arrangement of the strain gauges is presented in Fig. 3(c). and formed gradually from one side of the specimen to the other
In Section 1-1, Gauges S1–S8 were used to obtain the strains of the sides due to the initial imperfection.
steel flange; Gauges S9–S12 were used to measure the strains of 3. Strength degradation stage (after Point P): In this stage, because
the steel web; and Gauges S13–S17, which were attached to five the increase of the load due to the steel strengthening was always
vertical aluminum bars, were used to measure the vertical strains of less than the reduction of the load due to concrete degradation
the concrete, because the inside of the aluminum bars deformed after Point P, the load continuously decreased at a relatively
with the concrete. In Section 2-2, Gauges S52–S61, which were slow rate until the specimen failed. The cracks in the concrete
600
D3 D1
D11 D7
600
D2
D10 D9 D8
(a) (b)
S23S24S25
2 2 2 2 2 2 2 2
S26S27S28
S W N E
(c)
Fig. 3. (a) Test setup; (b) arrangement of displacement transducers; and (c) arrangement of strain gauges.
developed rapidly, and several large pieces of concrete started to flange and the strain of the steel and concrete were measured to
spall, leading to separation between the steel flange and concrete investigate the confinement mechanism of steel-confined concrete.
[Fig. 5(c)]. Fig. 6(a) shows the relationship between N∶N y and Δd , where
N∶N y is the ratio of the applied load N to the yield load N y , and Δd
is the average difference in the horizontal expansion between the
Confinement Mechanism of Steel-Confined Concrete
average measured value of the two sideward displacement trans-
The specimens expanded horizontally under the axial compression ducers and that of the middle displacement transducer. There is lit-
during the loading process. The lateral displacement of the steel tle difference in the horizontal expansion prior to the yield load N y
12000 12000
N (kN)
N (kN)
N (kN)
12000
N (kN)
N (kN)
12000 12000 12000
Peak point
8000 8000 8000 Balanced point
Fig. 4. (a) Typical axial load–displacement relationship; and (b) test results.
[Fig. 4(a)]; however, once the yield load N y is reached, the hori- agreement with the typical Poisson’s ratios. Then, jεch =εcv j begins
zontal expansion difference Δd starts to increase, which indicates to increase rapidly when the axial load reaches 0.75 N y ; however,
that the steel flanges deform out-of-plane and provide effective because the structural steel section enters a plastic state later than
confinement to the concrete. Then, when the load reaches the peak does the concrete, jεsh =εsv j does not change substantially until the
load N p , the concrete confinement peaks and begins to decrease load reaches the yield load N y . Therefore, the increase in jεsh =εsv j
because the steel flanges yield and cannot provide enough con- after 0.75N y indicates that the concrete confinement increases
finement to the concrete as bending deformation rapidly increases, gradually. After the load reaches N p , jεsh =εsv j starts to increase
indicating that the compressive strength of the concrete starts to rapidly, indicating that the concrete confinement provided by the
decrease due to the degeneration of the concrete confinement. structural steel section starts to decrease due to the excessive plastic
In addition to the out-of-plane bending deformation of the steel deformation of the steel flanges.
flange, the strains of steel flange can also reflect the confinement As described previously, the confinement mechanism of steel-
mechanism of steel-confined concrete. Because Poisson’s ratio of confined concrete in high-strength SRC-core columns, which is
the materials cannot be obtained directly from the test, the absolute similar with that in normal-strength SRC-core columns (Zhao et al.
value of the strain ratio of the horizontal to the vertical strains was 2014), can be described as follows: both the structural steel shapes
obtained to reflect the change in Poisson’s ratio. Fig. 6(b) presents and concrete are in an elastic stage before the axial load achieves
the relationship between N∶N y and jεch =εcv j, where εch and εcv are 0.75N y ; because the difference in the lateral expansion deformation
the average horizontal and vertical strains of the concrete, respec- between the steel flange and concrete is limited, the confinement
tively. Fig. 6(c) shows the relationship between N∶N y and jεsh =εsv j, effect of the structural steel shapes on the concrete is limited before
where εsh and εsv are the average horizontal and vertical strains of 0.75N y ; after 0.75N y, jεch =εcv j starts to increase due to the plastic
the structural steel section. In general, jεsh =εsv j is close to 0.3 and behavior of the concrete, but jεsh =εsv j remains approximately un-
jεch =εcv j is close to 0.2 in the elastic stage. The absolute values of changed. Therefore, the concrete confinement begins to increase
the strain ratio for both the steel and the concrete show good gradually and reaches its maximum value when the load reaches
Cracks
Separation Spalling
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
Fig. 5. Failure process of SRC-A-1: (a) elastic stage; (b) elastoplastic stage; and (c) strength degradation stage.
1.4 1.4
1.2 1.2
Np /Ny
1 1
Ny /Ny Ny /Ny
0.8 0.8
N/Ny
N/Ny
0.75Ny /Ny
0.6 0.6
SRC-A-1 SRC-A-1
0.4 SRC-A-2 0.4 SRC-A-2
SRC-B-1 SRC-B-1
0.2 SRC-B-2 0.2 SRC-B-2
SRC-B-3 SRC-B-3
SRC-C-1 SRC-C-1
0 0
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(a) d (mm)
(b) /
1.4
1.2
1
Ny/Ny
0.8
N/Ny
0.75Ny/Ny
0.6
SRC-A-1
0.4 SRC-A-2
SRC-B-1
0.2 SRC-B-2
SRC-B-3
SRC-C-1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(c) /
Fig. 6. Data analysis of confinement mechanism—ratio of applied load to yield load for (a) horizontal displacement difference of steel flange curves;
(b) strain ratio curves for concrete; and (c) strain ratio curves for steel flange.
the peak load N p . Then the steel flanges yield and deform rapidly, steel shapes. Therefore, considering the distribution of the confine-
resulting in the reduction of the concrete confinement and the ment effect in different directions, the steel-confined concrete can
degeneration of the compressive strength of the concrete. be divided into three regions according to the confinement mech-
However, the concrete confinement cannot be provided uni- anism [Fig. 3(c)], i.e., highly steel-confined concrete (HSCC),
formly in all directions due to the open-type section of the structural Region I; partially steel-confined concrete (PSCC), Region II; and
A x ia l lo a d kN
A x ia l lo a d kN
A x ia l lo a d kN
8000 14000 12000
6000 10500 9000
4000 7000 6000 Zone
Zone Zone
2000 Zone 3500 Zone 3000 Zone
Zone Zone Zone
0 0 0
0 400 800 1200 1600 2000 2400 2800 0 400 800 1200 1600 2000 2400 2800 0 400 800 1200 1600 2000 2400 2800
Strain Strain Strain
A x ial lo ad kN
A x ia l lo a d kN
A x ia l lo a d kN
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
Fig. 7. Axial load ratio–horizontal strain curves for each zone of steel-confined concrete.
UCC, Region III. Region I can be considered to sustain the lateral and concrete. When the peak load was reached, the load shared by
confining stress in any directions; Region II can be considered to the steel section N s was calculated as the measured yield strength
sustain the lateral steel confining stress in only one direction; and times the area of the steel section, and thus the load shared by the
Region III can be considered to sustain no lateral steel confining concrete N c was obtained by subtracting N s from the peak load N p .
stress in any directions. The enhanced strength of concrete fci was obtained by dividing N c
Based on the regional partition of the steel-confined concrete by the area of concrete. The strength incremental factor kci was
mentioned previously, Fig. 7 illustrates the relationships between defined as the ratio f ci ∶f c0 to reflect the strength enhancement of
the axial load and horizontal strain of the different steel-confined concrete. Accordingly, the displacement ratio Δp ∶Δy , where Δy
concrete regions. The horizontal strain is the average measured and Δp are the displacements corresponding to the yield load and
value of the strain gauges attached to the horizontal aluminum bars peak load, revealed the improvement of peak strain.
in the same region. The concrete horizontal strain increases from Fig. 8(a) presents the test results of Specimens SRC-A-1, SRC-
the HSCC to PSCC to UCC under the same axial load; therefore, B-1, SRC-A-2, and SRC-B-2, which illustrate the effect of the area
different regions of the steel-confined concrete experience different ratio As ∶Ac on the strength incremental factor and displacement ra-
levels of confinement effect. Therefore, the experimental investiga- tio Δp ∶Δy . Obviously, kci increases with increasing area ratio
tion in this section provides a theoretical basis for the regional par- As ∶Ac , and the strength enhancement of concrete is more significant
tition of the steel-confined concrete proposed in this paper. for the strength combination of the grade of steel Q390 and the
grade of concrete C60. However, the area ratio has a limited effect
on Δp ∶Δy ; for the strength combination of grade of steel Q390 and
Influential Factors on Confinement of Concrete
grade of concrete C60, Δp ∶Δy decreased by 1% when the area ratio
The test results are listed in Table 2. The ratio of the peak load to the As ∶Ac increased from 0.16 to 0.33, and a 1% increase was observed
yield load varied from 1.04 to 1.15. The strength and deformation for the strength combination of grade of steel Q460 and grade of
capacity of concrete were enhanced by the confinement provided concrete C70. Therefore, the area ratio As ∶Ac has a more positive
by the structural steel section, but the enhancement of strength effect on kci than does Δp ∶Δy . The test results of Specimens SRC-
could not be directly measured from the experiment. The axial A-1 and SRC-C-1 are discussed to investigate the effect of the con-
compressive load of the specimens was shared by the steel section fined dimension of steel flange lf on kci and Δp ∶Δy [Fig. 8(b)].
p/ y
p/ y
Kci
Kci
1.2 1.2 1.2 1.2
1 1 1 1
1.8 1.8
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
1.6 1.6
1.4 1.4
p/ y
Kci
1.2 1.2
1 1
0.8 0.8
Kci p/ y
0.6 0.6
SRC-B-1 SRC-B-2 SRC-B-3
(c) (Q390-C60) (Q460-C70) (Q460-C60)
Fig. 8. Analyses of influential factors: (a) effect of the area ratio of steel to concrete; (b) effect of the confined dimension of the steel flange; and
(c) effect of the strength combination of steel and concrete.
Fig. 10. Regional partition of the concrete in SRC columns with different steel profiles.
' / f'
Confined strength ratio f cc c
Larger confining stress ratio f l'2 / f c'
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
the steel flange (El-Tawil and Deierlein 1996; Zhao and Wen 2018). Confinement Effectiveness Coefficient of HSCC and PSCC
The bending moment applied at the base of the flange can be equiv- Regarding the confinement effectiveness coefficient of the concrete
alent to the resistance developed by the steel flange (Fig. 13), and confined by H-shaped steel, Mirza and Skrabek (1992) assumed
the first yielding of the steel flange is considered as the limiting that the effective confined concrete area is the total concrete area
condition minus the area of the two parabolas with heights equal to one-
quarter of their length [Fig. 15(a)]. Using a similar procedure to that
f l;f · l2f t2f · f y;f
¼ ð15Þ proposed by Mander et al. (1988) to define the confinement effec-
2 6 tiveness, the confinement effectiveness coefficient K e0 is
The lateral confining stress is effective confined concrete area
K e0 ¼ ð17Þ
total confined concrete area
t2f · fy;f
fl;f ¼ ð16Þ
3 · l2f It is necessary to consider the effect on the distribution of the
lateral confining stress due to the out-of-plane bending deforma-
where fy;f = yield strength of the steel flange. tion of the steel flange. Fig. 15(b) shows that the arching action is
f l' f
lf
f l' w
Y
tf
X lf X
H-shaped steel section Cross-shaped steel section
fl f
A lf B A lf B
PSCC UCC
lb
lb
G C G C
Z1
f l' w f l' f
lf
lf
lf
lf
F F
E lf lb D E lf lb D
Y HSCC PSCC
Y
X f l' w
X
(a) (b)
Fig. 14. Distribution of lateral confining pressure: (a) actual distribution; and (b) simplified distribution.
tf
tf
lf Ineffective lf
t
45 45
confined 45
Effective region Effective
confined confined
h-2t f
region region
h-2t
h-2t
h
h
h
45
(h-2t)/4
tf
t
tf
(h-2tf )/ 4 h (h-2tf )/4
(a) (b) (c)
Fig. 15. Region for effective confined concrete: (a) H-shaped steel–confined concrete; (b) steel tube–confined concrete; and (c) cross-shaped steel–
confined concrete.
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
assumed to occur in the form of a second-degree parabola with and OF are connected with the PSCC. The HSCC sustains the lat-
an initial tangent slope of 45° at each side (Huang et al. 2008). eral confining stress all around. More specifically, Sides GE and EF
Considering both the arching action and out-of-plane bending de- sustain the confinement provided by the steel web and Sides GO
formation of the steel flange, the effective confined region of the and OF sustain the confinement essentially provided by the steel
H-shaped steel-confined concrete is defined as shown in Fig. 15(c) flange. Therefore, in accordance with the region partition of con-
(Zhao and Wen 2018). crete, the thick line is equivalent to the steel component to present
In Fig. 16(a), the shaded part in the upper left is the HSCC. the function of the lateral confining stress. Because the steel web is
Sides GE and EF are connected with the steel web, but Sides GO located at the symmetric axis of the cross section, the steel web
f l' f
Ineffective confined region
Effective
tf
G O confined
region
45
fl'w fl' f
Y E F
l f /2 lf
tw
lf tf
fl' w tw
X
Simplified stress state Effective confined region
(a)
f l' f
Ineffective confined region
lf / 2
Effective
tf
A B
confined
region
lb
Y G O
t lf lb / 2
X fl 'w
Simplified stress state Effective confined region
(b)
Fig. 16. Simplified stress state and effective confined region: (a) HSCC; and (b) PSCC.
Verification of Proposed Stress–Strain Model for Stress–Strain Model for Structural Steel Sections
Concrete With the expansion of the inner concrete under compression, the
steel flange provides confinement for the concrete that undergoes
To verify the theoretical stress–strain model for the concrete in SRC out-of-plane bending deformation. In this case, the steel web is in a
columns proposed in this paper, it was implemented into the ana- tensile state in the horizontal direction, weakening the compres-
lytical model proposed by Chen and Lin (2006) and Chen and Wu sive strength of the steel web. Chen and Wu (2017) considered the
(2017) to predict the axial load–displacement relationships and biaxial stress state of steel web and adopted the modified Giuffrè-
ultimate load from previous experiments on axially loaded SRC Menegotto-Pinto model to propose a stress–strain model for the
short columns. In previous studies adopting this analytical model, steel web. The horizontal tensile stress of steel web fyt;w , corre-
the cylinder compressive strength of concrete ranged from 23.5 to sponding to the limiting condition adopted in Eq. (15), can be
94.0 MPa and the yield strength of structural steel ranged from 235 calculated according to force equilibrium
0 0
(a) (b)
(c)
Fig. 17. Theoretical stress–strain relationships in the analytical model: (a) concrete; (b) structural steel section; and (c) longitudinal bars.
0
2 · lf · fl;f Comparison of Analytical Results and Experimental
fyt;w ¼ ð20Þ Results
tw
Using the preceding analytical model, the predicted results of the
The axial compressive stress of the steel web fyv;w can be ob- axial compressive behavior of SRC short columns were obtained
tained according to the biaxial stress ellipse theory (Chen and Wu and compared with the experimental results of axially loaded SRC
2017) short columns to validate the stress–strain model proposed in this
f2yv;w þ f2yt;w − fyv;w · fyt;w ¼ f 2y ð21Þ study.
The specimens were classified into two groups. One group con-
For steel flange, the Giuffrè-Menegotto-Pinto model (Giuffrè tained SRC-core columns that were comprised of only structural
and Pinto 1970) is also applicable. Drawn on the definition of Chen steel section and concrete, and the other group contained ordinary
and Wu (2017), the parameters in the Giuffrè-Menegotto-Pinto SRC columns. A total of 101 specimens consisting of 34 SRC-core
model are defined as follows. The constant representing the radius columns (Chen et al. 2010; Zhao et al. 2014; and the experiments
of the transitional area between the elastic range and hardening conducted in the present paper) and 67 SRC columns (Chen and
branch R ¼ 10, and the ratio β of the hardening stiffness Esh to Yeh 1996; Tsai et al. 1996; Weng et al. 2006; Chen et al. 2010; Shih
the elastic modulus Es of steel is et al. 2013; Liang et al. 2014; Chen et al. 2015; Liu et al. 2015; Zhu
Esh ðfu − fy Þ=ðεu − εy Þ et al. 2014; Xiao et al. 2017) are investigated in this section. Apart
β¼ ¼ ð22Þ from the specimens studied in this paper, the geometry and material
Es Es
properties of several representative specimens are summarized
where f u = ultimate strength of the structural steel; and εy and εu = in Table 3 and the complete data are listed in Table S1, including
strains corresponding to the yield and ultimate strength of the struc- the cross-sectional dimensions, the measured yield strength and ul-
tural steel, respectively. The elastic modulus of the steel Es is timate tensile strength of the steel flange and steel web, the yield
206 GPa. The proposed stress–strain models for the steel section strength and configuration of the longitudinal bars and hoops, and
are illustrated in Fig. 17(b). In these models, εy and εyv;w are the cylinder compressive strength of the UCC. The cylinder com-
the strains corresponding to fy and fyv;w , and Esh;f and Esh;w pressive strength of the UCC ranged from 20.6 to 94.0 MPa and the
are the hardening stiffness of the steel flange and steel web. yield strength of the structural steel ranged from 235 to 517 MPa,
which is suitable for the range of parameters for which this analyti-
Stress–Strain Model for Longitudinal Bars cal model has been validated. More details are given in the corre-
The stress–strain model for longitudinal bars proposed by Chen and sponding references.
Lin (2006) is adopted in this analytical model [Fig. 17(c)]. It is The load ratios of the predicted axial bearing capacity to the
assumed that the longitudinal bar will undergo yielding during a yield load N p;a ∶N y are presented in Table 2. The load ratio N p;a ∶N y
yield plateau under compression. The compressive stress of longi- ranged from 1.06 to 1.12 and the average error between N p;a ∶N y
tudinal bars starts to decrease when the axial compressive strain is and N p;t ∶N y was 2.67%, which indicates that the analytical model
εco . Considering the bucking of the bars after the concrete cover can accurately reflect the enhancement of the bearing capacity of
undergoes crushing, the compressive stress of the bar will decrease the specimens in this paper due to the enhancement of the concrete
to 20% of the yield strength and then maintain a constant value strength. The parameters obtained according to the analytical model
(Chen and Lin 2006). are listed in Table 4. For some specimens, the stresses in the hoops
© ASCE
designation (mm) Type Size bars Configuration Type f y;f fu;f fy;w f u;w f y;l fy;h f c0 Reference
SRC1 280 × 280 H H150 × 150 × 7 × 10 12d16 D6@140 R 296 350 296 350 350 453 29.5 Chen and Yeh (1996)
SRC4 280 × 280 C 2 H175 × 90 × 5 × 8 12d16 D6@140 R 329 446 329 446 350 453 29.8
src1 280 × 280 C 2 H175 × 90 × 5 × 8 4d16 D6@140 R 318 444 378 466 453 606 23.9 Tsai et al. (1996)
src2 280 × 280 C 2 H175 × 90 × 5 × 8 4d16 D6@100 R 318 444 378 466 453 606 23.5
H6-S17 200 × 200 H H120 × 60 × 7 × 10 4d13 D6@170 R 316 448 266 351 388 538 20.6 Weng et al. (2006)
C4-S9 200 × 200 C 2H120 × 40 × 4 × 7 4d13 D6@90 R 266 351 296 372 388 538 20.6
A-04 — C 2H450 × 112 × 10 × 14 — — — 416 547 418 528 — — 27.0 Chen et al. (2010)
1A-HP4-28 560 × 560 C 2H450 × 168 × 10 × 14 4d25 D13@110 R 416 547 418 528 436 302 27.8
Src5 600 × 600 C 2H350 × 175 × 6 × 9 12d25 D13@150 M 386 529 421 571 469 463 32.3 Shih et al. (2013)
Src6 600 × 600 C 2H350 × 175 × 6 × 9 12d25 D13@125 M 386 529 421 571 469 463 32.3
SRC-1-2(1) — C 2H360 × 170 × 20 × 20 — — — 280 515 280 515 — — 35.0 Zhao et al. (2014)
SRC-1-2(2) — C 2H360 × 170 × 20 × 20 — — — 280 515 280 515 — — 27.0
DH-SP-75 600 × 600 C 2H350 × 175 × 6 × 9 16d25 D13@75 SP 454 574 437 546 451 472 34.6 Liang et al. (2014)
DH-SP-95 600 × 600 C 2H350 × 175 × 6 × 9 16d25 D13@95 SP 454 574 437 546 451 472 34.6
C-I-M40 200 × 200 H H100 × 68 × 4.5 × 7.6 12d10 D6.5@40 M 254 368 254 368 427 335 93.0 Zhu et al. (2014)
C-+-M60 200 × 200 C 2H100 × 68 × 4.5 × 7.6 12d10 D6.5@60 M 254 368 254 368 427 335 94.0
HSRC-SP1 D411 C 2H206 × 120 × 12 × 12 20d10 D10@50 SP 516 635 516 635 426 426 36.5 Chen et al. (2015)
HSRC-SP2 D411 C 2H206 × 120 × 12 × 12 20d10 D10@50 SP 516 635 516 635 426 426 36.5
CSRC1 400 × 400 H H150 × 150 × 7 × 10 12d16 D8@100 M 301 440 301 440 356 340 26.5 Liu et al. (2015)
CSRC3 400 × 400 H H175 × 175 × 7.5 × 11 12d16 D8@200 M 297 442 301 440 356 340 34.8
E00-1 450 × 450 H H120 × 106 × 12 × 20 32d8 D3.3@80 M 408 603 523 603 438 597 48.3 Xiao et al. (2017)
E00-2 450 × 450 H H120 × 106 × 12 × 20 32d8 D3.3@80 M 398 603 411 603 438 597 44.7
Note: H = H-shaped steel section; C = cross-shaped steel section; R = rectangular hoops; M = multiple hoops; SP = spiral hoops; d = diameter of longitudinal bars; and D = diameter of hoops.
04019009-14
© ASCE
(MPa) PSCC (MPa) PCC (MPa)
Specimen
designation f l;f (MPa) f r;h (MPa) fl;h (MPa) 0
K e;hs 0
K e;ps 0
K e;h f l1 f l2 khs fl1 fl2 kps f l1 f l2 kp Errora (%) Reference
SRC-A-1 2.59 — — 0.804 0.562 — 2.08 2.08 1.28 0 1.46 1.10 — — — 0.34 This paper
SRC-A-2 2.73 — — 0.804 0.562 — 2.19 2.19 1.23 0 1.53 1.09 — — — 4.20
SRC-B-1 10.83 — — 0.804 0.562 — 8.71 8.71 1.91 0 6.09 1.20 — — — −2.33
SRC-B-2 12.04 — — 0.804 0.562 — 9.68 9.68 1.83 0 6.77 1.18 — — — 0.50
SRC-B-3 12.21 — — 0.804 0.562 — 9.82 9.82 2.07 0 6.86 1.22 — — — −0.24
SRC-C-1 5.83 — — 0.804 0.282 — 4.69 4.69 1.61 0 1.79 1.11 — — — 2.25
SRC1 1.93 435 0.75 — 0.508 0.379 — — — 0.28 1.27 1.16 0.28 0.28 1.07 2.04 Chen and Yeh (1996)
SRC4 3.89 434 0.75 0.804 0.599 0.361 3.39 3.39 1.63 0.27 2.60 1.21 0.27 0.27 1.06 5.06
src1 3.76 448 0.77 0.804 0.599 0.229 3.28 3.28 1.74 0.18 2.43 1.22 0.18 0.18 1.05 1.84 Tsai et al. (1996)
src2 3.76 464 1.12 0.804 0.599 0.229 3.50 3.50 1.82 0.26 2.51 1.24 0.26 0.26 1.07 4.97
H6-S17 15.00 446 0.90 — 0.176 0.329 — — — 0.30 2.94 1.29 0.30 0.30 1.10 0.65 Weng et al. (2006)
C4-S9 13.41 481 1.83 0.804 0.331 0.329 6.18 6.18 2.29 0.60 5.04 1.43 0.60 0.60 1.19 2.40
A-04 10.45 — — 0.804 0.156 — 8.10 8.10 2.29 0.00 1.63 1.13 — — — 1.32 Chen et al. (2010)
1A-HP4-28 4.35 302 1.53 0.804 0.400 0.512 4.29 4.29 1.80 0.78 2.53 1.31 0.78 0.78 1.18 0.66
Src5 1.46 455 3.71 0.804 0.577 0.360 1.75 1.75 1.33 0.58 1.42 1.20 0.58 0.58 1.12 6.51 Shih et al. (2013)
Src6 1.46 463 2.47 0.804 0.577 0.381 1.92 1.92 1.36 0.75 1.59 1.24 0.75 0.75 1.15 8.39
SRC-1-2(1) 6.64 — — 0.804 0.562 — 5.34 5.34 1.79 0 3.73 1.18 — — — −3.40 Zhao et al. (2014)
SRC-1-2(2) 6.64 — — 0.804 0.562 — 5.34 5.34 1.96 0 3.73 1.20 — — — −4.49
DH-SP-75 1.72 472 3.15 0.804 0.577 0.513 3.00 3.00 1.50 1.62 2.61 1.38 1.62 1.62 1.29 1.75 Liang et al. (2014)
DH-SP-95 1.72 472 2.49 0.804 0.577 0.513 2.66 2.66 1.45 1.28 2.27 1.32 1.28 1.28 1.24 4.25
C-I-M40 4.85 335 3.71 — 0.184 0.510 — — — 1.89 2.78 1.19 1.89 1.89 1.13 1.17 Zhu et al. (2014)
C-+-M60 4.85 335 2.47 0.804 0.483 0.437 4.98 4.98 1.33 1.08 3.42 1.16 1.08 1.08 1.08 −3.17
HSRC-SP1 8.49 426 3.71 0.804 0.618 0.238 7.71 7.71 2.01 0.88 6.13 1.35 0.88 0.88 1.16 6.25 Chen et al. (2015)
04019009-15
HSRC-SP2 8.49 426 3.71 0.804 0.618 0.238 7.71 7.71 2.01 0.88 6.13 1.35 0.88 0.88 1.16 5.28
CSRC1 1.96 340 1.51 — 0.335 0.507 — — — 0.76 1.42 1.27 0.76 0.76 1.19 1.81 Liu et al. (2015)
CSRC3 1.71 340 0.75 — 0.335 0.349 — — — 0.26 0.84 1.12 0.26 0.26 1.05 −1.80
E00-1 24.63 421 0.32 — 0.372 0.762 — — — 0.24 9.41 1.27 0.24 0.24 1.03 3.31 Xiao et al. (2017)
J. Struct. Eng.
were less than the yield strength, reflecting that it is effective to bearing capacity, and loading path. Additionally, the axial bearing
adopt the procedure of Chen and Wu (2017) to avoid overestimating capacities were also compared [Fig. 18(b)]. Table 4 shows the
the lateral confining stress from the hoops. The strength incremental results of the axial bearing capacity from the experiment and ana-
factors for the HSCC ranged from 1.23 to 2.07, the strength incre- lytical model, N p;t and N p;a , for the specimens corresponding to
mental factors for the PSCC ranged from 1.09 to 1.48, and the Table 3. Overall, the absolute error of 82% of specimens was less
strength incremental factors for the PCC ranged from 1.05 to 1.36. than 7%, and the absolute average error of all specimens was
Clearly, for most specimens the confinement of the HSCC was 4.21%. The average value of N p;t ∶N p;a was 1.014 and the coeffi-
highest because the HSCC was confined by both the steel section cient of variation was 0.0514, which indicates that the predicted
and hoops in two directions. Therefore, the concrete confinement capacities agreed well with the test results. Fig. 18(c) also compares
provided by the structural steel section cannot be ignored in the the axial load–displacement curves of the test results and results
comparisons of the strength incremental factors for the HSCC, from three analytical models (Chen and Lin 2006; Chen and
PSCC, and PCC. Wu 2017; and the present paper). All three analytical models
Fig. 18(a) compares the experimental and analytical axial load– can accurately predict the mechanical behaviors of SRC columns.
displacement curves for the tests studied in this paper and some Compared with the Chen and Wu (2017) model and the model
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
previously performed representative tests. All results substantially presented in this paper, the use of the degrading model for struc-
begin at the origin, but the starting points of each result are at cer- tural steel applied by Chen and Lin (2006) may overestimate the
tain intervals to clearly present the comparisons. The predicted re- strength degeneration for the postpeak behavior, which can be
sults obtained using the analytical model were in good agreement found in the comparison of Specimen HSRC-SP2. The average
with the experimental results in terms of the initial stiffness, axial ratio of the predicted capacities according to the analytical model
16000 21000
f y, f =437,460,437 MPa fy , f =280,280,280,380 MPa
14000 18000 fy ,w =280,280,280,380 MPa
f y, w =437,460,437 MPa
12000 f c' =47.4,64.7,47.4 MPa fc' =35.0,27.0,36.7,36.7 MPa
SRC -A-1/2 SRC -C-1 15000 SRC-1-1(1)/1(2)/2/3
10000
N (kN)
N (kN)
12000
8000
9000 SRC-1-3
6000
SRC-A-1 SRC -A-2 6000 SRC-1-1(1) SRC-1-1(2) SRC-1-2 Test
4000 SRC -C-1
Test ( Zhao et al. 2014)
2000 3000
Analysis Analysis
0 0
15 (mm) 15 (mm)
7000 9000
fc' =23.5,21.8,24.3 MPa fc' =29.5,29.8,29.8 MPa
6000 f y , f =318,318,371 MPa fy , f =296,329,296 MPa
7500
f y , w =378,378,349 MPa fy ,w =296,329,296 MPa
5000 f y ,l =466,466,447 MPa fy ,l =350,350,350 MPa
6000
src-2 src-3 src-10 SRC1 SRC5 SRC10
N (kN)
N (kN)
4000
4500
3000
3000
2000 src -3 src -10 SRC5
src -2 Test SRC1
( Tsai et al.1996) 1500 SRC10( ChenTest
and Yeh 1996)
1000
Analysis Analysis
0 0
3 (mm) 3 (mm)
40000 20000
f c' =27.8,27.8,27.8 MPa f c' =36.5,36.5 MPa
35000 f y , f =416,416,416 MPa f y, f =516,516 MPa
16000 f y, w =516,516 MPa
30000 f y , w =418,418,418 MPa
f y ,l =436,436,436 MPa f y, l =426,426 MPa
25000 HSRC-SP1/2
1A-HP4-28 3A-HP4-42 12000
N (kN)
N (kN)
1A-HP4-42
20000
15000 8000 HSRC-SP2
HSRC-SP1
10000 3A-HP4-42
Test Test
1A-HP4-28 1A-HP4-42 4000 ( Chen et al. 2015)
( Chen et al. 2010)
5000 Analysis Analysis
0 0
5 (mm) 2 (mm)
(a)
Fig. 18. Comparisons of experiments and analyses: (a) axial load–displacement curves; (b) axial compressive capacities; and (c) curves for Speci-
mens SRC4 and HSRC-SP2.
Np,a (kN)
16000 Liang et al. (2014)
Zhu et al. (2014)
Chen et al. (2015)
12000
Liu et al. (2015)
Xiao et al. (2017)
8000 The authors
Mean value = 1.014
4000 Coefficient of Variation = 0.0514
0
0 4000 8000 12000 16000 20000 24000 28000
Downloaded from ascelibrary.org by Sergio Barreiro on 07/20/19. Copyright ASCE. For personal use only; all rights reserved.
Np,t (kN)
(b)
6000 15000
5000 12000
4000
9000
N (kN)
N (kN)
3000
Test 6000
2000 Chen and Lin (2006) Test
HSRC-SP2 Chen and Lin (2006)
SRC4 Chen and Wu (2017) 3000
1000 Chen and Wu (2017)
The authors The authors
0 0
0 1 2 3 4 5 6 0 5 10 15 20 25 30
(mm) (mm)
(c)
(Chen and Lin 2006) to the test results was 1.013 and the coeffi- SRC columns. Finally, the comparisons of the axial behavior of the
cient of variation was 0.0368; the average ratio of the predicted experimental results and the predicted results proved that the theo-
capacities according to the analytical model (Chen and Wu 2017) retical stress–strain model proposed in this study for the concrete
to the test results was 0.993 and the coefficient of variation was in SRC columns is accurate and effective. The conclusions of this
0.0532. Although the error levels of the predicted capacity of these paper are as follows:
three analytical models were similar, the stress–strain model for 1. Open-type structural steel shapes can provide confinement
concrete proposed in this paper, unlike the stress–strain relation- for concrete by the out-of-plane bending deformation of steel
ships in the other two models, possesses a theoretical basis of the flanges to enhance the compressive strength and ductility of the
confinement mechanism of both steel-confined and hoop-confined concrete;
concrete. Therefore, the analytical model proposed in this paper is 2. Based on the confinement mechanism of steel-confined and
suitable for arbitrary SRC columns regardless of the configuration hoop-confined concrete, the concrete in SRC columns was
of the reinforcement, material strength, and configuration of the divided into four regions, namely HSCC, PSCC, PCC, and
encased steel section. UCC; and
In general, the analytical model established in this section effec- 3. A theoretical stress–strain model for the concrete in SRC col-
tively predicts the axial compressive behavior of SRC columns, and umns was proposed according to the confinement mechanism in
the theoretical stress–strain model proposed in this study for the SRC columns, and comparisons of the experimental and analy-
concrete in SRC columns is accurate and satisfactory. tical results of the axial behavior of SRC columns proved that
this model is accurate and effective.
Conclusions
Acknowledgments
An experimental study of six SRC-core columns was carried out to
investigate the confinement mechanism of steel-confined concrete The authors gratefully acknowledge the financial support of the
in high-strength SRC columns; based on this work, a theoretical State Key Laboratory of the Ministry of Science and Technology
stress–strain model was proposed, considering the confinement of China under Grant No. SLDRCE15-B-01. The third author is
mechanism of both structural steel shapes and hoop-confined con- also grateful for the support from the Chinese National Engineering
crete. Furthermore, this model was implemented in an analytical Research Center for Steel Construction (Hong Kong Branch) at The
model to predict the axial compressive behavior of axially loaded Hong Kong Polytechnic University.
concrete;
0
K e;hs = confinement effectiveness coefficient of highly Supplemental Data
steel-confined concrete;
0
K e;ps = confinement effectiveness coefficient of partially Tables S1 and S2 are available online in the ASCE Library (www
steel-confined concrete; .ascelibrary.org).
k = strength incremental factor;
kc = strength incremental factor of unconfined concrete; References
kci = strength incremental factor of confined concrete from
test results; Chen, C. C., and N. J. Lin. 2006. “Analytical model for predicting axial
khs = strength incremental factor of highly steel-confined capacity and behavior of concrete encased steel composite stub col-
concrete; umns.” J. Constr. Steel Res. 62 (5): 424–433. https://doi.org/10.1016/j
.jcsr.2005.04.021.
kp = strength incremental factor of partially confined Chen, C. C., C. C. Tao, S. C. Yang, and W. M. Tsai. 2010. “Experimental
concrete; stress-strain behavior of the concrete confined by cross-H steel cross-
kps = strength incremental factor of partially steel-confined sectional shape.” [In Chinese.] J. Chin. Inst. Civ. Hydraul. Eng. 22 (4):
concrete; 423–434.
.1061/(ASCE)0733-9445(1995)121:3(468). Sheikh, S., and S. M. Uzumeri. 1982. “Analytical model for concrete
El-Tawil, S., and G. G. Deierlein. 1996. Fiber element analysis of confinement in tied columns.” J. Struct. Div. 108 (12): 2703–2722.
composite beam-column cross-sections. Ithaca, NY: Cornell Univ. Shih, T. H., C. C. Chen, C. C. Weng, Y. L. Yin, and J. C. Wang. 2013.
El-Tawil, S., and G. G. Deierlein. 1999. “Strength and ductility of concrete “Axial strength and ductility of square composite columns with two in-
encased composite columns.” J. Struct. Eng. 125 (9): 1009–1019. terlocking spirals.” J. Constr. Steel Res. 90 (5): 184–192. https://doi.org
https://doi.org/10.1061/(ASCE)0733-9445(1999)125:9(1009). /10.1016/j.jcsr.2013.07.021.
Ellobody, E., and B. Young. 2011. “Numerical simulation of concrete en- Tsai, K. C., Y. Lien, and C. C. Chen. 1996. “Behavior of axially loaded
cased steel composite columns.” J. Constr. Steel Res. 67 (2): 211–222. steel reinforced concrete columns.” [In Chinese.] J. Chin. Inst. Civ.
https://doi.org/10.1016/j.jcsr.2010.08.003. Hydraul. Eng. 8 (4): 535–545.
Giuffrè, A., and P. Pinto. 1970. “Il comportamento del cemento armato per Weng, C. C., H. S. Wang, Y. Li, and C. Y. Liang. 2006. “Analytical and
sollecitazioni cicliche di forte intensità.” Giornale del Genio Civile experimental study on seismic design of hoop reinforcement of SRC
5 (1): 391–408. columns.” [In Chinese.] Struct. Eng. 21 (3): 55–83.
Huang, Y. S., Y. L. Long, and J. Cai. 2008. “Ultimate strength of rectan- Xiao, C., F. Deng, T. Chen, and Z. Zhao. 2017. “Experimental study on
gular concrete-filled steel tubular (CFT) stub columns under axial com- concrete-encased composite columns with separate steel sections.” Steel
pression.” Steel Compos. Struct. 8 (2): 115–128. https://doi.org/10 Compos. Struct. 23 (4): 483–491. https://doi.org/10.12989/scs.2017.23
.12989/scs.2008.8.2.115. .4.483.
Kent, D. C., and R. Park. 1971. “Flexural members with confined con- Zhao, X., H. Qin, and Y. Y. Chen. 2014. “Experimental study on constit-
crete.” J. Struct. Div. 97: 1969–1990. utive model of steel confined concrete in SRC columns with cruciform
Lee, T. K., and A. D. E. Pan. 2001. “Analysis of composite beam-columns steel section.” [In Chinese.] J. Build. Struct. 35 (4): 268–279.
under lateral cyclic loading.” J. Struct. Eng. 127 (2): 186–193. https:// Zhao, X. Z., and F. P. Wen. 2018. “Theoretical study on confinement
doi.org/10.1061/(ASCE)0733-9445(2001)127:2(186). mechanism and stress-strain model for steel confined concrete in SRC
Liang, C. Y., C. C. Chen, C. C. Weng, Y. L. Yin, and J. C. Wang. 2014. columns.” [In Chinese.] Eng. Mech. 35 (5): 36–46.
“Axial compressive behavior of square composite columns confined by Zhu, W. Q., G. Meng, and J. Q. Jia. 2014. “Experimental studies on axial
multiple spirals.” J. Constr. Steel Res. 103 (103): 230–240. https://doi load performance of high-strength concrete short columns.” Struct.
.org/10.1016/j.jcsr.2014.09.006. Build. 167 (9): 509–519. https://doi.org/10.1680/stbu.13.00027.