- Email: [email protected]
Experimental behaviour of reinforced concrete-filled steel tubes under eccentric tension
Journal of Constructional Steel Research 136 (2017) 91–100
Contents lists available at ScienceDirect
Journal of Constructional Steel Research journal homepage: www.elsevier.com/locate/jcsr
Experimental behaviour of reinforced concrete-filled steel tubes under eccentric tension
MARK
Ju Chena, Jun Wanga, Wei Lib,⁎ a b
Department of Civil Engineering, Zhejiang University, Hangzhou 310058, China Department of Civil Engineering, Tsinghua University, Beijing 100084, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Concrete-filled steel tubes Eccentric tension Reinforcing bar Angle Design method
This paper presents research on the reinforced concrete-filled steel tubular (CFST) members subjected to eccentric tension, where the embedded components are the reinforcing bars or steel angles. A total of 8 full-scale specimens were designed and tested with main parameters of specimen types and load eccentricities. In particular, there was no direct connection between the outer tube and embedded components. Therefore the tensile load was transferred from the outer steel tube to the reinforcing bars or angles through the filled concrete. The failure mode, the load versus deformation relationships and the strain responses were recorded and analysed. The simplified design equations were also proposed for the elastic tensile stiffness and tension versus moment interaction relationships of reinforced CFSTs under eccentric tension.
1. Introduction It is well known that the concrete-filled steel tube (CFST) has excellent structural performance owing to the composite action between the concrete and steel tube. Various researchers have conducted investigations of the CFST behaviour under compression, bending, tension and combined loadings [1–8]. It was found that the core concrete was confined by the steel tube and the buckling mode of steel tube was altered by the concrete. Therefore several codes of practice have been published and the CFST has been widely used all over the world [9–12]. Sometimes CFST members may be subjected to combined tension and bending, such as the ones in latticed electricity transmission towers, as shown in Fig. 1. The previous research found that the CFSTs had higher tensile strength than the bare steel tubular counterparts. Pan and Zhong [13] conducted experimental and theoretical investigations, which found that the tensile strength of CFST was 1.1 times higher than that of the hollow steel tube, attributing to the hoop stress developed in the outer tube. Han et al. [14] studied the influence of interface conditions on CFST tensile members. The results showed that the perfectly bonded and debonded interfaces had limited effect on the tensile strength of CFST using normal concrete. Li et al. [15–17] conducted a series of investigations on the CFST and the concrete-filled double skin steel tubes (CFDST) under tension. It was found that the filled concrete changed the stress status of steel tubes and enhanced the capacities of members under concentric or eccentric tension. For the
⁎
Corresponding author. E-mail address: [email protected] (W. Li).
http://dx.doi.org/10.1016/j.jcsr.2017.05.004 Received 24 February 2017; Received in revised form 7 May 2017; Accepted 9 May 2017 0143-974X/ © 2017 Elsevier Ltd. All rights reserved.
eccentrically loaded CFSTs, the in-filled concrete worked well with the outer steel tube, and a large deformation capacity was observed for all members, for the final end rotation exceeded 0.1 rad [17]. In order to enhance the ultimate strength of CFSTs, the most effective way is to increase the cross-sectional area of steel. However, the steel tube with large thickness may increase difficulties in manufacturing and installation. When comparing with embedding a steel tube inside, embedding the reinforcing bars or steel angles could be an easier solution for the tensile CFSTs as the convenience of pouring the concrete. Chen et al. [18] has conducted experimental research on the CFSTs with embedded reinforcing bars and angles under concentric tension. It was found that the embedded components could effectively increase the ultimate strength of concentrically loaded tensile member owing to the increase of cross-sectional area of steel. It is also noted that the strength of embedded components could be fully developed through certain bonding length. However, there is still a lack of studies for the behaviour of CFSTs with reinforcing bars or angles under eccentric tension. The contribution of these reinforcing components may be underestimated, which may result in an over-conservative design of CFST tensile members. This paper is a companion one with the CFSTs with reinforcing bars or angles under concentric tension [18]. The experimental investigation on 8 full-scale specimens is conducted. The test parameters include the types of the reinforcing components and the load eccentricities applied. The fail modes, load and deformation relationships and strain responses are recorded and analysed. The load carrying capacities of the CFSTs
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Nomenclature Ac As Asa Asr D d E Ec Es Esa Esr (EA)′T (EA)′T-test (EA)′T-cal fcu fck
ft fu fy fysa fysr L l M Mu Mur T t Tu Tur
cross-sectional area of concrete cross-sectional area of steel tube cross-sectional area of angle cross-sectional area of reinforcing bar outside diameter of steel tube diameter of reinforcing bar elastic modulus of steel elastic modulus of concrete elastic modulus of steel tube elastic modulus of angle elastic modulus of reinforcing bar elastic tensile stiffness under eccentric tension elastic tensile stiffness under eccentric tension obtained from the experimental load-axial elongation curve elastic tensile stiffness under eccentric tension predicted using the proposed equation compressive strength of concrete of 150 × 150 × 150 mm cube characteristic compressive strength of concrete, fck = 0.67fcu
Tu-Eq3 Tu-Eq4 Tu-Test ta α δ ξ
predict the stiffness and tension versus moment relationships of CFSTs with embedded components under eccentric tension.
with reinforcing bars or angles subjected to eccentric tension are also assessed. The main objectives of this research are as follows. 1) to provide a new series of test data for CFSTs with angles and reinforcing bars under eccentric tension; 2) to evaluate the contribution of embedded angles and reinforcing bars for the overall behaviour of eccentrically loaded CFSTs; and 3) to propose simplified design equations which could reasonably
2. Experimental investigation 2.1. General description A total of 8 full-scale specimens were designed for the eccentric tension test. The main parameters are the type of the cross section, the tensile load eccentricity and the connection pattern. Two kinds of embedded components, i.e., the angle with connecting plate and the reinforcing bars were used in the tests. These embedded components were originally designed as construction details and the strength contribution was not considered tentatively. The load eccentricity ratios (λ) of 0.1 and 0.2 were designated, where λ is obtained by dividing the load eccentricity (e) with the radius of steel tube (D/2) for specimens with circular cross section. Two specimens were designed to evaluate the effectiveness of the flange connection for column segments. The nominal diameters of the longitudinal reinforcing bars and the stirrup were 16 mm and 8 mm, respectively. The distance between two stirrup layers was 200 mm. The cross-sectional profile of the steel angle was L-56 × 5 mm, and battens of 156 × 50 × 4 mm were used to connect four steel angles together. The nominal length of each specimen was 4000 mm, while the nominal outer diameter of the steel tube was approximately 400 mm. The schematic view of CFST specimens with reinforcing bars and angles are shown in Fig. 2(a) and (b), respectively. The labels of the specimens are listed in Table 1, where the characters ‘AG’, ‘E’, ‘FC’ and ‘RB’ represent the angle, eccentricity, flange connection and reinforcing bars, respectively. Two load eccentricities, i.e., 20 mm and 40 mm were used in the test, and the corresponding labels are E20 and E40, respectively. The last character ‘A’ or ‘B’ represents the test specimen with the same parameters. Table 1 also lists the measured cross-sectional dimensions and the specimen length for each specimen.
Nt Wind load
tensile strength of concrete ultimate tensile strength of steel yield stress of steel yield stress of angle yield stress of reinforcing bar length of test specimen length of angle moment flexural strength of CFST flexural strength of CFST with reinforcing bars or angles tension thickness of steel tube tensile strength of eccentrically loaded CFST tensile strength of eccentrically loaded CFST with reinforcing bars or angles tensile strength predicted using Eq. (3) tensile strength predicted using Eq. (4) tensile strength of test specimen thickness of angle steel ratio (=As/Ac) percentage elongation after fracture confinement factor
Concrete Steel tube Embedded element
Ground
2.2. Specimen end and connection Two 30-mm-thick steel end-plates were welded to both ends to ensure full contact between the specimen and the bearings. There are
Fig. 1. Member in composite transmission tower under eccentric tension.
92
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
156
8@200 Concrete
End ribs
200
Steel tube
Connecting plate 156×50×4@200
Steel tube
Concrete
End ribs
240 400
Stiffener
Angles 4*L56×5
Batten 156×50×4@200
50
Longitudinal bars 8* 16
Stirrup 8@200
156 400
Stiffener
End plate
End plate
(a) CFST with embedded reinforcing bars
(b) CFST with embedded angles
Steel tube
30 204
Stiffener
Nuts
First layer of flange
End plate
Second layer of flange
204
Vertical rib Horizontal rib
60 Steel tube
M36 High-strength bolt
First layer of flange (c) Vertical and horizontal ribs at the ends
Steel tube
(d) Flange connection between tubes in the middle Fig. 2. Configuration of test specimens (unit: mm).
Table 1 Information of test specimens. Specimen label
Specimen type
L (mm)
D (mm)
t (mm)
l (mm)
ta (mm)
d (mm)
e (mm)
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FC-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FC-E20
CFST without embedded component CFST with reinforcing bars CFST with reinforcing bars CFST with reinforcing bars & with flange connection CFST with angles CFST with angles CFST with angles CFST with angles & with flange connection
3999 3997 4003 3998 3998 3998 3998 3999
400.1 401.0 399.8 400.3 400.4 400.4 400.4 400.0
6.01 6.01 6.01 6.01 6.01 6.01 6.01 6.01
– – – – 56 56 56 56
– – – – 6 6 6 6
– 16 16 16 – – – –
20 20 40 20 20 20 40 20
Note: L is the length of the test specimens; D is the diameter of the steel tube; T is the thickness of the steel tube; l is the width of the angle leg; t is the thickness of the angle; d is the diameter of the reinforcing bar.
several 4-mm-thick vertical and horizontal ribs inside the steel tube at both ends. Fig. 2(c) shows the dimensions of ribs. These inner ribs were used to enhance the load transfer between the steel tube and concrete. The CFST column segments are often connected by flange connections in real structures. For the specimens with flange connection, the connection located at the centre of the specimen. For the specimen with the flange connection, two layers of flanges were designed at the end of each column segment to ensure the load transferring. The thicknesses of flanges were 24 mm and 16 mm for the first and second layers, respectively. Sixteen M36 grade 8.8 high-strength bolts were used in the connection. The details of the flange connection are shown in Fig. 2(d).
Table 2 Material properties of steel. Type
E (GPa)
fy (MPa)
fu (MPa)
δ (%)
Angle Steel tube (4 mm) Steel tube (6 mm) Reinforcing bar 16 mm
205.2 220.0 215.0 205.0
376 558 458 418
555 639 560 573
69.0 65.0 70.7 38.5
93
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Eccentric Connection
Central line of tension
Load eccentricity
Central line of specimen e Eccentric Connection
(a) Schematic view
(b) Photo Fig. 3. Test setup.
2.3. Fabrication and material properties
2.5. Loading and measurement
The steel tubes were fabricated by cold-forming a flat virgin plate into a round shape, and then welding two edges together by a straight electric welding. The tube ends were machined to get the specified length. The reinforcing bars or angles were placed inside the steel tubes before filling the concrete, and no connection was made between the steel tubes and the reinforcing bars or angles. The self-compacting concrete was filled into all specimens without any vibration to ensure the compactness of the core concrete. A very small settlement of concrete occurred on top of the column due to the shrinkage of concrete. The high-strength concrete was used to fill this longitudinal settlement before welding the top end plate. The strength of the steel tube may be higher than that of virgin plate due to the cold forming process. The material properties of steel tubes were obtained by the coupon test, where the coupons were extracted from the middle of the tube and far away from the welding. The material properties of the reinforcing bars were obtained by the direct tensile test. Table 2 lists the measured average values of the Young's modulus (Es), the yield stress (fy), the tensile strength (fu), and the percentage elongation after fracture (δ). The concrete properties were measured after 28 days of curing of concrete. The average cubic strength (fcu), tensile strength (ft), and elastic modulus (Ec) of concrete were 61.6 MPa, 4.06 MPa, and 36,260 MPa, respectively. The tests were conducted soon after the 28 days, therefore the material properties measured at 28 days can be regarded as those at the test days.
The displacement control method was used during the loading at a constant speed of 1.0 mm/min for all specimens. An initial load of approximately 10% of the design strength of the specimen was applied on each specimen before the test to eliminate any possible slippage at the specimen ends. The ultimate strength of the specimen was defined as the strength when the maximum longitudinal strain of the midheight reached 5000 με according to Han et al. [14]. Therefore the test was stopped when the strain gauge reading at the mid-height reached 5000 με. The applied load and readings of transducers were recorded at regular intervals during the tests. The tensile load was measured by the load sensor attached to the hydraulic actuator. A total of 40 3 mm × 2 mm strain gauges, including longitudinal and circumferential ones, were mounted on the outer surface of steel tube around the mid-height. The working range of the strain gauges was ± 20,000 με. For the specimen with flange connections, the strains of the flange bolts were also measured. The strains of the embedded reinforcing bars and angles were not recorded in the test, for the placement of strain gauges might disturb the interaction between steel and concrete. Four vertical LVDTs were placed at both ends to obtain the elongation and rotation of the specimen, and 3 horizontal LVDTs were used to record the overall bending of the upper segment. Fig. 4 shows the placement of the LVDTs. 3. Test results and analysis 3.1. Failure modes
2.4. Test setup
The test was conducted in a smooth and controlled way. According to LVDTs and strain gauges, the test specimens have overall bending and elongation deformation when the ultimate strength was reached. The lateral displacements under different loading levels are shown in Fig. 5, where Tu-test is the measured tensile strength. The failure mode of the inner concrete was examined after the test. The outer steel tube was removed to show the crack patterns of the core concrete, as shown in Fig. 6. The widths of cracks were measured by the microscope and were marked in the same figure. The width of the crack and distance between cracks can reflect the damage degree of core
Fig. 3 shows the schematic view and the photo of the test setup. A tensile loading machine with a capacity of 8,000 kN was designed for the test. The tension was applied by a hydraulic actuator. The specimen was fixed at the bottom and the eccentric tension was applied on the top through the spherical bearing on the top, where the specimen could rotate around the bearing. The specimen was connected to the bearing and the bottom base by high-strength bolts.
94
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
e
S1
1-1 I
Displ. transducer
Tension e
Strain gauges
S3 S5 1-3 S2 S4 1-2
A1-5
A1
A2
1-4
A3-5
A3
200
Strain gauges
Displ. transducer
200
Tension
I
A4
Fixed end
(b)Strain gauges along the longitudinal direction
(a)Displacement transducer
Side A
Central line of tension
1-3 e
1-4
1-2
Welding
Side B
1-1
Central line of specimen
(c) Hoop strain gauges on cross section (I-I) Fig. 4. Instrumentation and measurement.
3.2. Tension versus axial elongation relationships
concrete under different parameters and evaluate the effectiveness of the connection and embedded components. It is noted that these values are obtained after the unloading, which is less than the maximum width during the test due to the closure of cracks. For the specimen with 20 mm eccentricity (CFST-AG-E20A), the average distance between two cracks on the ‘Side A’ was 98 mm, while that on the ‘Side B’ was almost the same. The ‘Side A’ and ‘Side B’ represent the sides close to and far away from the central line of tension, respectively, as shown in Fig. 4(c). The average widths of the cracks on the side A and the side B were 0.36 and 0.28 mm, respectively. For the specimen with 40 mm eccentricity (CFST-AGE40), the average distance between two cracks on the side A was 86 mm, while that on the side B was 105 mm. The average widths of the cracks on the side A and side B were 0.35 and 0.31 mm, respectively. The influence of the eccentricity on the crack width is not significant within the scope of the test parameter. For the specimens with flange connection (CFST-AG-FC-E20), a large crack was observed at the midheight, due to the discontinuity of the steel tube.
Fig. 7 depicts the tension versus axial elongation (T-Δ) relationships of the eccentrically loaded CFST specimens, and it also shows the corresponding T-Δ relationships of the concentrically loaded CFST specimens from the companion paper. It can be seen that the typical T-Δ relationship for the reinforced CFST under eccentric tension consists of three stages, i.e., the elastic stage, the elastic-plastic stage and the plastic stage. For the elastic stage, the tangent line of T-Δ relationship is used to calculate the elastic tensile stiffness (EA)′T, where the original point and the point corresponding to the proportion limit of steel (0.8fy) are connected together (Han et al. [14]). The values obtained from the test (EA)′T-test are listed in Table 3. The initial stiffness of eccentrically loaded specimen decreased with the increase of the load eccentricity. For instance, the initial stiffness of CFST-RB-E20 (with 20 mm eccentricity) was 23% higher than that of CFST-RB-E40 (with 40 mm eccentricity). The initial stiffness also increased when the reinforcing bars and angles are embedded. The initial stiffness of CFST-RB-E20 95
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
4.5
gradually because of the hardening effect. As the maximum longitudinal strain at the mid-height reached 5000 με, the specimen reached its ultimate tensile strength and was unloaded after then. Table 4 presents the measured tensile strengths (Tu-test) of all specimens. It can be found that the embedded reinforcing bars and angles significantly enhanced the tensile strength of the specimen. The tensile strengths of specimens CFST-RB-E20 and CFST-AG-E20 were 17.6% and 24.2% higher than that of CFST-E20, respectively. The flange connection has minor influence on the tensile strength, for the ultimate strengths of specimens with or without flange connection are almost the same. The tensile strength decreased with the increase of the load eccentricity. The tensile strength of CFST-AG-E40 decreases 1.4% than that of CFSTAG-E20 series. For all specimens, the stiffness of the unloading branch was similar to that of the initial loading branch.
Height, H (m)
4 3.5
Sine curves
0.3 Tu-test
3
0.6 Tu-test 0.9 Tu-test
2.5 2 0
4
2
6
8
10
12
Lateral displacement, (mm) 3.3. Tension versus strain relationships
(a) CFST-E20
Fig. 8(a) shows the development of tension versus longitudinal strain (T-εL) relationships for the specimen CFST-RB-E20, where the position of strain gauges are denoted in Fig. 4. It can be found that the T-εL relationship also consists of elastic, elastic-plastic and plastic stages. All specimens had positive longitudinal strain during the test, which indicates the cross section was under tension during the eccentric loading. The longitudinal strain of the side close to the tensile load was higher than that of the opposite side. Fig. 8(b) shows the longitudinal strain distribution in the hoop direction under different loading levels, i.e., 0.3Tu-test, 0.6Tu-test, and 0.9Tu-test, where Tu-test is the measured tensile strength. It can be seen that the strain distribution was almost linear along the cross section and was coincident with the plane crosssection assumption. Fig. 9 shows the strain development along the hoop direction, where negative strain values were detected. This indicates the reduction of the diameter of the steel tube. The previous research on concentrically loaded CFST tensile members showed that the core concrete significantly reduced the hoop strain of the steel tube [18]. For the eccentrically loaded specimen, the hoop strain on the side close to the central line of tension is greater than that on the other. For the specimen CFST-RB-E20, the hoop strain on the tensile side was approximately 2 times of that on the compressive side. It is attributing to the combination of load eccentricity and the interaction between the outer steel tube and the inner concrete. Moreover, the hoop strain of the CFST with embedded components is smaller than that of the CFST counterpart when the same load value is reached. For instance, the hoop strain at point 1–3 of CFST-RB-E20 and CFST-AG-E20 is 100 με and 124 με when the load reached 3500kN, respectively, while that of CFST-E20 is 300 με. It is due to the fact that the embedded components resist part of the load and the strain on the outer tube is lower for the reinforced CFST.
4.5 0. T u-test 3 0. Tu-test 6 0. Tu-test 9
Height, H (m)
4
Sine curves
3.5 3 2.5 2 0
3
6
9
12
15
Lateral displacement, (mm)
(b) CFST-RB-E20 4.5 0.3 Tu-test
Height, H (m)
4
0.6 Tu-test 0.9 Tu-test
3.5 3
Sine curves
2.5 2 0
3
6
9
12
15 3.4. Load transfer mechanism
Lateral displacement, (mm)
Fig. 10 shows the schematic view of the load transfer mechanism in different components. The tensile load has been applied to the specimen through the endplate and high-strength bolts at both ends. The load was then transferred to the steel tube through the welding between the endplate and the steel tube. The tensile load was partially transferred to the concrete through the inner horizontal and vertical ribs of the steel tube at both ends, as shown in Fig. 10(a). The embedded reinforcing bars and angles have no direct connection to the outer steel tube or the endplate. However, they were anchored in the end concrete through the macro locking effect and the bond stress, as shown in Fig. 10(b). Finally the tensile load is transferred to the whole cross section.
(c) CFST-RB-FC-E20 Fig. 5. Lateral displacement of eccentrically loaded specimen.
(with embedded reinforcing bars) was 16% higher than that of the CFST counterpart (CFST-E20). For the specimen with flange connection, the initial stiffness was lower than its CFST counterparts. For example, the stiffness of CFST-RB-FC-E20 was 7.1% lower than that of CFST-RB-E20, and the stiffness of CFST-AG-FC-E20 was 0.6% lower than the average one of CFST-AG-E20 series. It may be due to the blot elongation of the flange connection. For the elastic-plastic stage, the stiffness decreases with the increase of the load. The plastic deformation developed and the load increased 96
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
(a) CFST-AG-E20A (‘Side A’)
(b) CFST-AG-E20A (‘Side B’)
(c) CFST-AG-E40 (‘Side A’)
(d) CFST-AG-E40 (‘Side B’) Fig. 6. Concrete cracks of eccentrically loaded CFST specimens (units: mm).
where λ(=2e/D) is the load eccentricity ratio of the CFST specimen, e is the load eccentricity, D is the diameter of the circular cross section; (EA)T is the elastic tensile stiffness for the concentric tensile specimen calculated by Eqs. (1a) and (1b). The value of (EA)T ‐ test' is less than (EA)T ‐ test when the eccentric load is applied. The predicted elastic tensile stiffnesses (EA)T ‐ cal' using Eq. (2) are compared with the measured ones (EA)T ‐ test' in Table 3. The average values and COV of (EA)T ‐ cal'/(EA)T ‐ test' are 0.980 and 0.071 for all specimens, respectively. In general the proposed equation is found to have reasonable accuracy.
4. Simplified design method 4.1. Elastic tensile stiffness There is lack of equation to calculate the elastic tensile stiffness of an eccentrically loaded CFST specimen in current codes of practice [9–11]. For the concentrically loaded CFST specimen with embedded components, the elastic tensile stiffness (EA)T can be calculated as follows according to Chen et al. [18]
(EA)T = Es As + 0.1(Ec Ac + Esr Asr )
(1a)
(EA)T = Es As + 0.1(Ec Ac + Esa Asa )
(1b)
4.2. Tension versus moment interaction curve
where Es and As are the elastic modulus and the cross-sectional area of steel tube, respectively; Ec and Ac are the elastic modulus and crosssectional area of concrete, respectively; Esr and Esa are the elastic modulus of reinforcing bars and the reinforcing angles, respectively; Asr and Asa are the cross-sectional area of reinforcing bars and angles, respectively. The test results showed that the reinforcing bars or angles were able to resist the tensile load with concrete together, and the whole cross section was under tension for each eccentrically loaded specimen. An empirical equation for the calculation of the elastic tensile stiffness of the CFST specimen under eccentric tension (EA)T′ is presented as follows:
(EA)′T = (1 −
λ )2 (EA)
T
For the tension-moment interaction (T-M) curve of CFST member without embedded components, previous research has found that the tensile strength and flexural strength followed a liner relation. The current code of practice proposed a linear tension versus moment interaction curve for the CFST member, where the tensile strength and the flexural strength of the eccentrically loaded member are normalized by those of the concentrically loaded one as follows [12]:
T M + ≤1 Tu Mu
(3)
where T and M are the tension and moment, respectively; Tu and Mu are the ultimate tensile and flexural strength of CFST, respectively. Tu and Mu can be predicted according to the current code of practice.
(2) 97
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al. 5000
Table 4 Comparison of measured and predicted tensile strengths.
4500 4000
Specimens
Tension, T (kN)
3500
Tu-test (kN)
Tu-Eq3 (kN)
Tu-Eq3/Tu-
Tu-Eq4 (kN)
Tu-Eq4/Tu-
test
3000 2500
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FCE20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FCE20
2000 1500
CFST-E20 CFST-RB-E20 CFST-RB-FC-E20 CFST-RB-E40
1000 500 0 0
10
20
30
Axial elongation,
40
50
(mm)
test
3529.0 4151.2 – 4134.7
3165.6 3708.0 3243.3 3708.0
0.897 0.893 – 0.897
3384.6 3982.0 3665.2 3982.0
0.959 0.959 – 0.963
4460.4 4308.9 4321.2 4397.3
3817.4 3817.4 3330.8 3817.4
0.856 0.886 0.771 0.868
4105.6 4105.6 3771.4 4105.6
0.920 0.953 0.873 0.934
Mean COV
0.867 0.052
(a) RB series
0.937 0.035
5000 4500
5000
4000
4500
3000
4000
2500
3500
Tension, T (kN)
Tension, T (kN)
3500
2000 1500
CFST-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-FC-E20 CFST-AG-E40
1000 500 0 0
5
10
15
20
25
Axial elongation,
30
35
3000 2000 1500 1000
40
500
(mm)
0
(b) AG series 5000
A1-5 S1 S2 S3 S4 S5 A3-5
2500
0
1000
CFST-AG series
4500
2000
4000
5000
6000
L( )
(a)
4000 3500
CFST-RB series
450
3000 2500
400
CFST series
2000
Distance, d (mm)
Tension, T (kN)
3000
Longitudinal strain,
1500 1000
Eccentric Concentric
500 0 0
10
20
30
Axial elongation,
40
50
Td
T
350
0
300 250
0.3Tu-test 0.6Tu-test 0.9Tu-test
200 150 100
(mm)
50
(c) Comparison of concentrically and eccentrically loaded specimens
0
Fig. 7. Load-axial elongation (N-Δ) relationships.
0
1000
2000
Longitudinal strain,
Table 3 Comparison of measured and predicted initial stiffness.
3000 L(
4000
)
(b)
Specimens
(EA)′T-test (kN)
(EA)′T-cal (kN)
(EA)′T-cal/(EA)′T-test
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FC-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FC-E20
1.373 × 106 1.601 × 106 1.303 × 106 1.488 × 106 1.638 × 106 1.684 × 106 1.407 × 106 1.650 × 106
1.540 × 106 1.567 × 106 1.238 × 106 1.567 × 106 1.576 × 106 1.576 × 106 1.245 × 106 1.576 × 106 Mean COV
1.122 0.979 0.950 1.053 0.962 0.936 0.885 0.955 0.980 0.075
Fig. 8. Load-longitudinal strain relationships.
from the AIJ guide [9], AISC specification [10], Eurocode 4 [11] and DBJ specification [12]. It was found that the AIJ guide and AISC specification gave the most conservative prediction, and the results predicted by the Eurocode 4 and DBJ specification were similar and were closer to the test results [7]. Therefore in this study, the equation in DBJ specification [12] was adopted tentatively. For the CFST with reinforcing bars or angles, the T-M relationship may be different from that of the CFST counterpart because of the embedded components. According to the experimental results, an equation is proposed to describe the T-M relationship of CFST with reinforcing bars or angles as follows:
Previously Han et al. [7] proposed the calculation method for the flexural strength of the CFST, which was adopted in the DBJ specification [12]. Comparisons were also made between the calculation results
98
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
5000
contributions of CFST and embedded components together as follows:
4500
Tur = (1.1–0.4α ) fy As + fysr Asr
(5a)
Tur = (1.1–0.4α ) fy As + fysa Asa
(5b)
Tension, T (kN)
4000 3500 3000
where fy is the yield strength of the steel tube; As is the cross-sectional area of the steel tube; α is the steel ratio defined by α = As / Ac; Ac is the cross-sectional area of the concrete; fysr and fysa are the yield stress of the reinforcing bars and the reinforcing angles, respectively. In Eq. (4), Mur is the ultimate flexural strength of CFST with reinforcing bars and angles, which can be regarded as the summation of contributions from two parts tentatively, i.e., the concrete and steel tube part (the CFST part), as well as the embedded components part. The expressions of the flexural strength of reinforced CFST are as follows:
1-1 1-2 1-3 1-4
2500 2000 1500 1000 500 0 -500
-400
-300
-200
Hoop strain,
-100
H(
0
Mur = Msc,u + Mp
)
Tension, T (kN)
(a) CFST-RB-E20 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 -500
Msc,u = γm Wsc fscy
-300
Hoop strain,
-200 H(
-100
0
)
(b) CFST-AG-E20 5000 4500
1-1 1-2 1-3 1-4
Tension, T (kN)
4000 3500 3000
5. Conclusions
2500 2000
The behaviour of the reinforced CFSTs with reinforcing bars or angles under eccentric tension was investigated experimentally in this study. The failure modes, elastic tensile stiffness and the tensionmoment relationships of test specimens were analysed. The following conclusions could be drawn within the limited scope of this study:
1500 1000 500 0 -500
-400
-300
Hoop strain,
-200 H(
-100
0
1) All test specimens showed ductility behaviour during the test. Although the reinforcing bars or angles were not connected to the steel tube directly, the embedded components worked effectively with the core concrete attributing to the inner ribs and bonding effect. 2) The embedded reinforcing bars or angle significantly enhanced the strength of the eccentrically loaded tensile member. The tensile strengths of specimens CFST-RB-E20 and CFST-AG-E20 were 17.6% and 24.2% higher than those of CFST counterparts without embedded components, respectively. The tensile strength decreased with the increase of load eccentricity. Meanwhile, the elastic tensile stiffness of eccentrically loaded CFST specimen with reinforcing bars or angles was lower than that of the concentrically loaded counterpart.
)
(c) CFST-E20 Fig. 9. Load-hoop strain relationships.
⎛ T ⎞2 M ≤1 ⎜ ⎟ + ⎝ Tur ⎠ Mur
(7)
In Eq. (7), γm is the flexural strength index [7], representing the influence of the confinement degree to the strength, γm = 1.1 + 0.48 ⋅ ln (ξ + 0.1), where the confinement factor ξ is determined by αfy/fck, α is the steel ratio for CFST cross-section, α = As / Ac; fy is the yield stress of outer steel tube, fck is the characteristic strength of concrete cube. In Eq. (7), Wsc = πD3/32, D is the diameter of the outer tube. fscy is the ‘nominal yield strength’ of the composite section [7], fscy = (1.14 + 1.02ξ)fck, fck is the characteristic compressive strength of concrete. The predicted tensile strengths for eccentrically loaded specimens by Eq. (3) (Tu-Eq3) and Eq. (4) (Tu-Eq4) are both presented in Table 4, where the measured results (Tu-test) are also presented for comparison. The average value and the COV of the Tu-Eq4/Tu-test ratio are 0.937 and 0.032, respectively, while the average value and the COV of the Tu-Eq3/ Tu-test ratio are 0.867 and 0.048, respectively. Fig. 11(b) depicts the measured test results (denoted in points) and the predicted tension versus moment relationships. It can be seen that most data points are above the lines predicted by Eqs. (3) and (4) expect the data point CFST-E20. It is due to the fact that the specimen CFST-E20 has no embedded component inside. In general reasonable accuracy has been achieved for the predicted results.
1-1 1-2 1-3 1-4
-400
(6)
where, Msc,u is the ultimate flexural strength of the CFST part, Mp is the ultimate flexural strength of the cross section of embedded components. Msc,u can be expressed as follows [12]:
(4)
The curves of Eqs. (3) and (4) are presented in Fig. 11(a). In Eq. (4), Tur is the ultimate tensile strength of CFST with reinforcing bars and angles. Chen et al. [18] has found that it can be predicted by adding the 99
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Outer tube
Tension
Outer tube
Tension
Tension
Ribs
Tension
Embedded bars Concrete or angles
Embedded bars or angles
Concrete
(a)
Tension
(b) Fig. 10. Load transfer mechanism in the specimen.
1.2
In general good agreement was achieved between the predictions and test results.
1
Acknowledgements Eq.(4)
T/Tu
0.8
The research work described in this paper was supported by a project from the Science and Technology Department of Zhejiang Province (2015C33005).
0.6 0.4 Eq.(3)
References
0.2 [1] L.H. Han, W. Li, R. Bjorhovde, Developments and advanced applications of concrete-filled steel tubular (CFST) structures: members, J. Constr. Steel Res. 100 (9) (2014) 211–228. [2] B. Uy, Local and postlocal buckling of fabricated steel and composite cross sections, J. Struct. Eng. ASCE 127 (6) (2001) 666–677. [3] T. Perea, R.T. Leon, J.F. Hajjar, M.D. Denavit, Full-scale tests of slender concretefilled tubes: axial behavior, J. Struct. Eng. 139 (7) (2013) 1249–1262 (2013). [4] C.W. Roeder, D.E. Lehman, E. Bishop, Strength and stiffness of circular concretefilled tubes, J. Struct. Eng. ASCE 136 (12) (2010) 1545–1553. [5] J. Moon, C.W. Roeder, D.E. Lehman, H. Lee, Analytical modeling of bending of circular concrete-filled steel tubes, Eng. Struct. (2012) 349–361. [6] J. Nie, Y. Wang, J. Fan, Experimental study on seismic behavior of concrete filled steel tube columns under pure torsion and compression–torsion cyclic load, J. Constr. Steel Res. 79 (12) (2013) 115–126. [7] L.H. Han, Further study on the flexural behavior of concrete-filled steel tubes, J. Constr. Steel Res. 62 (6) (2006) 554–565. [8] H. Lu, L.H. Han, X.L. Zhao, Analytical behavior of circular concrete-filled thinwalled steel tubes subjected to bending, Thin-Walled Struct. 47 (3) (2009) 346–358. [9] AIJ, Recommendations for Design and Construction of Concrete Filled Steel Tubular Structures, Architectural Institute of Japan (AIJ), Tokyo, Japan, 2008. [10] ANSI/AISC 360-10, Specification for Structural Steel Buildings, American Institute of Steel Construction (AISC), Chicago, USA, 2010. [11] EN 1994-1-1, Design of Composite Steel and Concrete Structures — Part 1-1: General Rules and Rules for Buildings. Eurocode 4, European Standard, CEN, Brussels, 2004 (2004). [12] DBJ/T13-51-2010, Technical Specifications for Concrete-Filled Steel Tubular Structures, The Housing and Urban-Rural Development Department of Fujian Province, Fuzhou, China, 2010 (in Chinese). [13] Y.G. Pan, S.T. Zhong, Constitutive relations of concrete filled steel tube under tension, Ind. Constr. 20 (4) (1990) 33–37 (in Chinese). [14] L.H. Han, S.H. He, F.Y. Liao, Performance and calculations of concrete filled steel tubes (CFST) under axial tension, J. Constr. Steel Res. 67 (11) (2011) 1699–1709. [15] W. Li, L.H. Han, T.M. Chan, Numerical investigation on the performance of concrete-filled double-skin steel tubular members under tension, Thin-Walled Struct. 79 (6) (2014) 108–118. [16] W. Li, L.H. Han, T.M. Chan, Tensile behaviour of concrete-filled double-skin steel tubular members, J. Constr. Steel Res. 99 (8) (2014) 35–46. [17] W. Li, L.H. Han, T.M. Chan, Performance of concrete-filled steel tubes subjected to eccentric tension, J. Struct. Eng. ASCE 141 (12) (2015) (04015049). [18] J. Chen, J. Wang, W.L. Jin, Concrete-filled steel tubes with reinforcing bars or angles under axial tension, J. Constr. Steel Res. (2017) (Accepted).
0 0
0.2
0.4
0.6
0.8
1
1.2
M/Mu (a) 5000 4800
Eq.(3)
CFST-AG-E20 series
4600
Eq.(4)
CFST-AG-FC-E20
T (kN)
4400 CFST-AG-E40
4200
CFST-RB-E2
4000
CFST-RB-FC-E2
3800 3600
CFST-AG series
CFST-RB series
3400 CFST-E20
3200 3000 0
50
100
150
200
250
300
M (kN m)
(b) Fig. 11. Tension-moment relationship of the reinforced CFST.
3) Design equations were proposed to predict the stiffness and the tension-moment interaction relationship of reinforced CFST under eccentric tension, where the flexural strength of the CFST with reinforcing bars or angles can be tentatively estimated by adding the flexural strength of CFST and the embedded components together.
100
Contents lists available at ScienceDirect
Journal of Constructional Steel Research journal homepage: www.elsevier.com/locate/jcsr
Experimental behaviour of reinforced concrete-filled steel tubes under eccentric tension
MARK
Ju Chena, Jun Wanga, Wei Lib,⁎ a b
Department of Civil Engineering, Zhejiang University, Hangzhou 310058, China Department of Civil Engineering, Tsinghua University, Beijing 100084, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Concrete-filled steel tubes Eccentric tension Reinforcing bar Angle Design method
This paper presents research on the reinforced concrete-filled steel tubular (CFST) members subjected to eccentric tension, where the embedded components are the reinforcing bars or steel angles. A total of 8 full-scale specimens were designed and tested with main parameters of specimen types and load eccentricities. In particular, there was no direct connection between the outer tube and embedded components. Therefore the tensile load was transferred from the outer steel tube to the reinforcing bars or angles through the filled concrete. The failure mode, the load versus deformation relationships and the strain responses were recorded and analysed. The simplified design equations were also proposed for the elastic tensile stiffness and tension versus moment interaction relationships of reinforced CFSTs under eccentric tension.
1. Introduction It is well known that the concrete-filled steel tube (CFST) has excellent structural performance owing to the composite action between the concrete and steel tube. Various researchers have conducted investigations of the CFST behaviour under compression, bending, tension and combined loadings [1–8]. It was found that the core concrete was confined by the steel tube and the buckling mode of steel tube was altered by the concrete. Therefore several codes of practice have been published and the CFST has been widely used all over the world [9–12]. Sometimes CFST members may be subjected to combined tension and bending, such as the ones in latticed electricity transmission towers, as shown in Fig. 1. The previous research found that the CFSTs had higher tensile strength than the bare steel tubular counterparts. Pan and Zhong [13] conducted experimental and theoretical investigations, which found that the tensile strength of CFST was 1.1 times higher than that of the hollow steel tube, attributing to the hoop stress developed in the outer tube. Han et al. [14] studied the influence of interface conditions on CFST tensile members. The results showed that the perfectly bonded and debonded interfaces had limited effect on the tensile strength of CFST using normal concrete. Li et al. [15–17] conducted a series of investigations on the CFST and the concrete-filled double skin steel tubes (CFDST) under tension. It was found that the filled concrete changed the stress status of steel tubes and enhanced the capacities of members under concentric or eccentric tension. For the
⁎
Corresponding author. E-mail address: [email protected] (W. Li).
http://dx.doi.org/10.1016/j.jcsr.2017.05.004 Received 24 February 2017; Received in revised form 7 May 2017; Accepted 9 May 2017 0143-974X/ © 2017 Elsevier Ltd. All rights reserved.
eccentrically loaded CFSTs, the in-filled concrete worked well with the outer steel tube, and a large deformation capacity was observed for all members, for the final end rotation exceeded 0.1 rad [17]. In order to enhance the ultimate strength of CFSTs, the most effective way is to increase the cross-sectional area of steel. However, the steel tube with large thickness may increase difficulties in manufacturing and installation. When comparing with embedding a steel tube inside, embedding the reinforcing bars or steel angles could be an easier solution for the tensile CFSTs as the convenience of pouring the concrete. Chen et al. [18] has conducted experimental research on the CFSTs with embedded reinforcing bars and angles under concentric tension. It was found that the embedded components could effectively increase the ultimate strength of concentrically loaded tensile member owing to the increase of cross-sectional area of steel. It is also noted that the strength of embedded components could be fully developed through certain bonding length. However, there is still a lack of studies for the behaviour of CFSTs with reinforcing bars or angles under eccentric tension. The contribution of these reinforcing components may be underestimated, which may result in an over-conservative design of CFST tensile members. This paper is a companion one with the CFSTs with reinforcing bars or angles under concentric tension [18]. The experimental investigation on 8 full-scale specimens is conducted. The test parameters include the types of the reinforcing components and the load eccentricities applied. The fail modes, load and deformation relationships and strain responses are recorded and analysed. The load carrying capacities of the CFSTs
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Nomenclature Ac As Asa Asr D d E Ec Es Esa Esr (EA)′T (EA)′T-test (EA)′T-cal fcu fck
ft fu fy fysa fysr L l M Mu Mur T t Tu Tur
cross-sectional area of concrete cross-sectional area of steel tube cross-sectional area of angle cross-sectional area of reinforcing bar outside diameter of steel tube diameter of reinforcing bar elastic modulus of steel elastic modulus of concrete elastic modulus of steel tube elastic modulus of angle elastic modulus of reinforcing bar elastic tensile stiffness under eccentric tension elastic tensile stiffness under eccentric tension obtained from the experimental load-axial elongation curve elastic tensile stiffness under eccentric tension predicted using the proposed equation compressive strength of concrete of 150 × 150 × 150 mm cube characteristic compressive strength of concrete, fck = 0.67fcu
Tu-Eq3 Tu-Eq4 Tu-Test ta α δ ξ
predict the stiffness and tension versus moment relationships of CFSTs with embedded components under eccentric tension.
with reinforcing bars or angles subjected to eccentric tension are also assessed. The main objectives of this research are as follows. 1) to provide a new series of test data for CFSTs with angles and reinforcing bars under eccentric tension; 2) to evaluate the contribution of embedded angles and reinforcing bars for the overall behaviour of eccentrically loaded CFSTs; and 3) to propose simplified design equations which could reasonably
2. Experimental investigation 2.1. General description A total of 8 full-scale specimens were designed for the eccentric tension test. The main parameters are the type of the cross section, the tensile load eccentricity and the connection pattern. Two kinds of embedded components, i.e., the angle with connecting plate and the reinforcing bars were used in the tests. These embedded components were originally designed as construction details and the strength contribution was not considered tentatively. The load eccentricity ratios (λ) of 0.1 and 0.2 were designated, where λ is obtained by dividing the load eccentricity (e) with the radius of steel tube (D/2) for specimens with circular cross section. Two specimens were designed to evaluate the effectiveness of the flange connection for column segments. The nominal diameters of the longitudinal reinforcing bars and the stirrup were 16 mm and 8 mm, respectively. The distance between two stirrup layers was 200 mm. The cross-sectional profile of the steel angle was L-56 × 5 mm, and battens of 156 × 50 × 4 mm were used to connect four steel angles together. The nominal length of each specimen was 4000 mm, while the nominal outer diameter of the steel tube was approximately 400 mm. The schematic view of CFST specimens with reinforcing bars and angles are shown in Fig. 2(a) and (b), respectively. The labels of the specimens are listed in Table 1, where the characters ‘AG’, ‘E’, ‘FC’ and ‘RB’ represent the angle, eccentricity, flange connection and reinforcing bars, respectively. Two load eccentricities, i.e., 20 mm and 40 mm were used in the test, and the corresponding labels are E20 and E40, respectively. The last character ‘A’ or ‘B’ represents the test specimen with the same parameters. Table 1 also lists the measured cross-sectional dimensions and the specimen length for each specimen.
Nt Wind load
tensile strength of concrete ultimate tensile strength of steel yield stress of steel yield stress of angle yield stress of reinforcing bar length of test specimen length of angle moment flexural strength of CFST flexural strength of CFST with reinforcing bars or angles tension thickness of steel tube tensile strength of eccentrically loaded CFST tensile strength of eccentrically loaded CFST with reinforcing bars or angles tensile strength predicted using Eq. (3) tensile strength predicted using Eq. (4) tensile strength of test specimen thickness of angle steel ratio (=As/Ac) percentage elongation after fracture confinement factor
Concrete Steel tube Embedded element
Ground
2.2. Specimen end and connection Two 30-mm-thick steel end-plates were welded to both ends to ensure full contact between the specimen and the bearings. There are
Fig. 1. Member in composite transmission tower under eccentric tension.
92
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
156
8@200 Concrete
End ribs
200
Steel tube
Connecting plate 156×50×4@200
Steel tube
Concrete
End ribs
240 400
Stiffener
Angles 4*L56×5
Batten 156×50×4@200
50
Longitudinal bars 8* 16
Stirrup 8@200
156 400
Stiffener
End plate
End plate
(a) CFST with embedded reinforcing bars
(b) CFST with embedded angles
Steel tube
30 204
Stiffener
Nuts
First layer of flange
End plate
Second layer of flange
204
Vertical rib Horizontal rib
60 Steel tube
M36 High-strength bolt
First layer of flange (c) Vertical and horizontal ribs at the ends
Steel tube
(d) Flange connection between tubes in the middle Fig. 2. Configuration of test specimens (unit: mm).
Table 1 Information of test specimens. Specimen label
Specimen type
L (mm)
D (mm)
t (mm)
l (mm)
ta (mm)
d (mm)
e (mm)
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FC-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FC-E20
CFST without embedded component CFST with reinforcing bars CFST with reinforcing bars CFST with reinforcing bars & with flange connection CFST with angles CFST with angles CFST with angles CFST with angles & with flange connection
3999 3997 4003 3998 3998 3998 3998 3999
400.1 401.0 399.8 400.3 400.4 400.4 400.4 400.0
6.01 6.01 6.01 6.01 6.01 6.01 6.01 6.01
– – – – 56 56 56 56
– – – – 6 6 6 6
– 16 16 16 – – – –
20 20 40 20 20 20 40 20
Note: L is the length of the test specimens; D is the diameter of the steel tube; T is the thickness of the steel tube; l is the width of the angle leg; t is the thickness of the angle; d is the diameter of the reinforcing bar.
several 4-mm-thick vertical and horizontal ribs inside the steel tube at both ends. Fig. 2(c) shows the dimensions of ribs. These inner ribs were used to enhance the load transfer between the steel tube and concrete. The CFST column segments are often connected by flange connections in real structures. For the specimens with flange connection, the connection located at the centre of the specimen. For the specimen with the flange connection, two layers of flanges were designed at the end of each column segment to ensure the load transferring. The thicknesses of flanges were 24 mm and 16 mm for the first and second layers, respectively. Sixteen M36 grade 8.8 high-strength bolts were used in the connection. The details of the flange connection are shown in Fig. 2(d).
Table 2 Material properties of steel. Type
E (GPa)
fy (MPa)
fu (MPa)
δ (%)
Angle Steel tube (4 mm) Steel tube (6 mm) Reinforcing bar 16 mm
205.2 220.0 215.0 205.0
376 558 458 418
555 639 560 573
69.0 65.0 70.7 38.5
93
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Eccentric Connection
Central line of tension
Load eccentricity
Central line of specimen e Eccentric Connection
(a) Schematic view
(b) Photo Fig. 3. Test setup.
2.3. Fabrication and material properties
2.5. Loading and measurement
The steel tubes were fabricated by cold-forming a flat virgin plate into a round shape, and then welding two edges together by a straight electric welding. The tube ends were machined to get the specified length. The reinforcing bars or angles were placed inside the steel tubes before filling the concrete, and no connection was made between the steel tubes and the reinforcing bars or angles. The self-compacting concrete was filled into all specimens without any vibration to ensure the compactness of the core concrete. A very small settlement of concrete occurred on top of the column due to the shrinkage of concrete. The high-strength concrete was used to fill this longitudinal settlement before welding the top end plate. The strength of the steel tube may be higher than that of virgin plate due to the cold forming process. The material properties of steel tubes were obtained by the coupon test, where the coupons were extracted from the middle of the tube and far away from the welding. The material properties of the reinforcing bars were obtained by the direct tensile test. Table 2 lists the measured average values of the Young's modulus (Es), the yield stress (fy), the tensile strength (fu), and the percentage elongation after fracture (δ). The concrete properties were measured after 28 days of curing of concrete. The average cubic strength (fcu), tensile strength (ft), and elastic modulus (Ec) of concrete were 61.6 MPa, 4.06 MPa, and 36,260 MPa, respectively. The tests were conducted soon after the 28 days, therefore the material properties measured at 28 days can be regarded as those at the test days.
The displacement control method was used during the loading at a constant speed of 1.0 mm/min for all specimens. An initial load of approximately 10% of the design strength of the specimen was applied on each specimen before the test to eliminate any possible slippage at the specimen ends. The ultimate strength of the specimen was defined as the strength when the maximum longitudinal strain of the midheight reached 5000 με according to Han et al. [14]. Therefore the test was stopped when the strain gauge reading at the mid-height reached 5000 με. The applied load and readings of transducers were recorded at regular intervals during the tests. The tensile load was measured by the load sensor attached to the hydraulic actuator. A total of 40 3 mm × 2 mm strain gauges, including longitudinal and circumferential ones, were mounted on the outer surface of steel tube around the mid-height. The working range of the strain gauges was ± 20,000 με. For the specimen with flange connections, the strains of the flange bolts were also measured. The strains of the embedded reinforcing bars and angles were not recorded in the test, for the placement of strain gauges might disturb the interaction between steel and concrete. Four vertical LVDTs were placed at both ends to obtain the elongation and rotation of the specimen, and 3 horizontal LVDTs were used to record the overall bending of the upper segment. Fig. 4 shows the placement of the LVDTs. 3. Test results and analysis 3.1. Failure modes
2.4. Test setup
The test was conducted in a smooth and controlled way. According to LVDTs and strain gauges, the test specimens have overall bending and elongation deformation when the ultimate strength was reached. The lateral displacements under different loading levels are shown in Fig. 5, where Tu-test is the measured tensile strength. The failure mode of the inner concrete was examined after the test. The outer steel tube was removed to show the crack patterns of the core concrete, as shown in Fig. 6. The widths of cracks were measured by the microscope and were marked in the same figure. The width of the crack and distance between cracks can reflect the damage degree of core
Fig. 3 shows the schematic view and the photo of the test setup. A tensile loading machine with a capacity of 8,000 kN was designed for the test. The tension was applied by a hydraulic actuator. The specimen was fixed at the bottom and the eccentric tension was applied on the top through the spherical bearing on the top, where the specimen could rotate around the bearing. The specimen was connected to the bearing and the bottom base by high-strength bolts.
94
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
e
S1
1-1 I
Displ. transducer
Tension e
Strain gauges
S3 S5 1-3 S2 S4 1-2
A1-5
A1
A2
1-4
A3-5
A3
200
Strain gauges
Displ. transducer
200
Tension
I
A4
Fixed end
(b)Strain gauges along the longitudinal direction
(a)Displacement transducer
Side A
Central line of tension
1-3 e
1-4
1-2
Welding
Side B
1-1
Central line of specimen
(c) Hoop strain gauges on cross section (I-I) Fig. 4. Instrumentation and measurement.
3.2. Tension versus axial elongation relationships
concrete under different parameters and evaluate the effectiveness of the connection and embedded components. It is noted that these values are obtained after the unloading, which is less than the maximum width during the test due to the closure of cracks. For the specimen with 20 mm eccentricity (CFST-AG-E20A), the average distance between two cracks on the ‘Side A’ was 98 mm, while that on the ‘Side B’ was almost the same. The ‘Side A’ and ‘Side B’ represent the sides close to and far away from the central line of tension, respectively, as shown in Fig. 4(c). The average widths of the cracks on the side A and the side B were 0.36 and 0.28 mm, respectively. For the specimen with 40 mm eccentricity (CFST-AGE40), the average distance between two cracks on the side A was 86 mm, while that on the side B was 105 mm. The average widths of the cracks on the side A and side B were 0.35 and 0.31 mm, respectively. The influence of the eccentricity on the crack width is not significant within the scope of the test parameter. For the specimens with flange connection (CFST-AG-FC-E20), a large crack was observed at the midheight, due to the discontinuity of the steel tube.
Fig. 7 depicts the tension versus axial elongation (T-Δ) relationships of the eccentrically loaded CFST specimens, and it also shows the corresponding T-Δ relationships of the concentrically loaded CFST specimens from the companion paper. It can be seen that the typical T-Δ relationship for the reinforced CFST under eccentric tension consists of three stages, i.e., the elastic stage, the elastic-plastic stage and the plastic stage. For the elastic stage, the tangent line of T-Δ relationship is used to calculate the elastic tensile stiffness (EA)′T, where the original point and the point corresponding to the proportion limit of steel (0.8fy) are connected together (Han et al. [14]). The values obtained from the test (EA)′T-test are listed in Table 3. The initial stiffness of eccentrically loaded specimen decreased with the increase of the load eccentricity. For instance, the initial stiffness of CFST-RB-E20 (with 20 mm eccentricity) was 23% higher than that of CFST-RB-E40 (with 40 mm eccentricity). The initial stiffness also increased when the reinforcing bars and angles are embedded. The initial stiffness of CFST-RB-E20 95
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
4.5
gradually because of the hardening effect. As the maximum longitudinal strain at the mid-height reached 5000 με, the specimen reached its ultimate tensile strength and was unloaded after then. Table 4 presents the measured tensile strengths (Tu-test) of all specimens. It can be found that the embedded reinforcing bars and angles significantly enhanced the tensile strength of the specimen. The tensile strengths of specimens CFST-RB-E20 and CFST-AG-E20 were 17.6% and 24.2% higher than that of CFST-E20, respectively. The flange connection has minor influence on the tensile strength, for the ultimate strengths of specimens with or without flange connection are almost the same. The tensile strength decreased with the increase of the load eccentricity. The tensile strength of CFST-AG-E40 decreases 1.4% than that of CFSTAG-E20 series. For all specimens, the stiffness of the unloading branch was similar to that of the initial loading branch.
Height, H (m)
4 3.5
Sine curves
0.3 Tu-test
3
0.6 Tu-test 0.9 Tu-test
2.5 2 0
4
2
6
8
10
12
Lateral displacement, (mm) 3.3. Tension versus strain relationships
(a) CFST-E20
Fig. 8(a) shows the development of tension versus longitudinal strain (T-εL) relationships for the specimen CFST-RB-E20, where the position of strain gauges are denoted in Fig. 4. It can be found that the T-εL relationship also consists of elastic, elastic-plastic and plastic stages. All specimens had positive longitudinal strain during the test, which indicates the cross section was under tension during the eccentric loading. The longitudinal strain of the side close to the tensile load was higher than that of the opposite side. Fig. 8(b) shows the longitudinal strain distribution in the hoop direction under different loading levels, i.e., 0.3Tu-test, 0.6Tu-test, and 0.9Tu-test, where Tu-test is the measured tensile strength. It can be seen that the strain distribution was almost linear along the cross section and was coincident with the plane crosssection assumption. Fig. 9 shows the strain development along the hoop direction, where negative strain values were detected. This indicates the reduction of the diameter of the steel tube. The previous research on concentrically loaded CFST tensile members showed that the core concrete significantly reduced the hoop strain of the steel tube [18]. For the eccentrically loaded specimen, the hoop strain on the side close to the central line of tension is greater than that on the other. For the specimen CFST-RB-E20, the hoop strain on the tensile side was approximately 2 times of that on the compressive side. It is attributing to the combination of load eccentricity and the interaction between the outer steel tube and the inner concrete. Moreover, the hoop strain of the CFST with embedded components is smaller than that of the CFST counterpart when the same load value is reached. For instance, the hoop strain at point 1–3 of CFST-RB-E20 and CFST-AG-E20 is 100 με and 124 με when the load reached 3500kN, respectively, while that of CFST-E20 is 300 με. It is due to the fact that the embedded components resist part of the load and the strain on the outer tube is lower for the reinforced CFST.
4.5 0. T u-test 3 0. Tu-test 6 0. Tu-test 9
Height, H (m)
4
Sine curves
3.5 3 2.5 2 0
3
6
9
12
15
Lateral displacement, (mm)
(b) CFST-RB-E20 4.5 0.3 Tu-test
Height, H (m)
4
0.6 Tu-test 0.9 Tu-test
3.5 3
Sine curves
2.5 2 0
3
6
9
12
15 3.4. Load transfer mechanism
Lateral displacement, (mm)
Fig. 10 shows the schematic view of the load transfer mechanism in different components. The tensile load has been applied to the specimen through the endplate and high-strength bolts at both ends. The load was then transferred to the steel tube through the welding between the endplate and the steel tube. The tensile load was partially transferred to the concrete through the inner horizontal and vertical ribs of the steel tube at both ends, as shown in Fig. 10(a). The embedded reinforcing bars and angles have no direct connection to the outer steel tube or the endplate. However, they were anchored in the end concrete through the macro locking effect and the bond stress, as shown in Fig. 10(b). Finally the tensile load is transferred to the whole cross section.
(c) CFST-RB-FC-E20 Fig. 5. Lateral displacement of eccentrically loaded specimen.
(with embedded reinforcing bars) was 16% higher than that of the CFST counterpart (CFST-E20). For the specimen with flange connection, the initial stiffness was lower than its CFST counterparts. For example, the stiffness of CFST-RB-FC-E20 was 7.1% lower than that of CFST-RB-E20, and the stiffness of CFST-AG-FC-E20 was 0.6% lower than the average one of CFST-AG-E20 series. It may be due to the blot elongation of the flange connection. For the elastic-plastic stage, the stiffness decreases with the increase of the load. The plastic deformation developed and the load increased 96
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
(a) CFST-AG-E20A (‘Side A’)
(b) CFST-AG-E20A (‘Side B’)
(c) CFST-AG-E40 (‘Side A’)
(d) CFST-AG-E40 (‘Side B’) Fig. 6. Concrete cracks of eccentrically loaded CFST specimens (units: mm).
where λ(=2e/D) is the load eccentricity ratio of the CFST specimen, e is the load eccentricity, D is the diameter of the circular cross section; (EA)T is the elastic tensile stiffness for the concentric tensile specimen calculated by Eqs. (1a) and (1b). The value of (EA)T ‐ test' is less than (EA)T ‐ test when the eccentric load is applied. The predicted elastic tensile stiffnesses (EA)T ‐ cal' using Eq. (2) are compared with the measured ones (EA)T ‐ test' in Table 3. The average values and COV of (EA)T ‐ cal'/(EA)T ‐ test' are 0.980 and 0.071 for all specimens, respectively. In general the proposed equation is found to have reasonable accuracy.
4. Simplified design method 4.1. Elastic tensile stiffness There is lack of equation to calculate the elastic tensile stiffness of an eccentrically loaded CFST specimen in current codes of practice [9–11]. For the concentrically loaded CFST specimen with embedded components, the elastic tensile stiffness (EA)T can be calculated as follows according to Chen et al. [18]
(EA)T = Es As + 0.1(Ec Ac + Esr Asr )
(1a)
(EA)T = Es As + 0.1(Ec Ac + Esa Asa )
(1b)
4.2. Tension versus moment interaction curve
where Es and As are the elastic modulus and the cross-sectional area of steel tube, respectively; Ec and Ac are the elastic modulus and crosssectional area of concrete, respectively; Esr and Esa are the elastic modulus of reinforcing bars and the reinforcing angles, respectively; Asr and Asa are the cross-sectional area of reinforcing bars and angles, respectively. The test results showed that the reinforcing bars or angles were able to resist the tensile load with concrete together, and the whole cross section was under tension for each eccentrically loaded specimen. An empirical equation for the calculation of the elastic tensile stiffness of the CFST specimen under eccentric tension (EA)T′ is presented as follows:
(EA)′T = (1 −
λ )2 (EA)
T
For the tension-moment interaction (T-M) curve of CFST member without embedded components, previous research has found that the tensile strength and flexural strength followed a liner relation. The current code of practice proposed a linear tension versus moment interaction curve for the CFST member, where the tensile strength and the flexural strength of the eccentrically loaded member are normalized by those of the concentrically loaded one as follows [12]:
T M + ≤1 Tu Mu
(3)
where T and M are the tension and moment, respectively; Tu and Mu are the ultimate tensile and flexural strength of CFST, respectively. Tu and Mu can be predicted according to the current code of practice.
(2) 97
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al. 5000
Table 4 Comparison of measured and predicted tensile strengths.
4500 4000
Specimens
Tension, T (kN)
3500
Tu-test (kN)
Tu-Eq3 (kN)
Tu-Eq3/Tu-
Tu-Eq4 (kN)
Tu-Eq4/Tu-
test
3000 2500
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FCE20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FCE20
2000 1500
CFST-E20 CFST-RB-E20 CFST-RB-FC-E20 CFST-RB-E40
1000 500 0 0
10
20
30
Axial elongation,
40
50
(mm)
test
3529.0 4151.2 – 4134.7
3165.6 3708.0 3243.3 3708.0
0.897 0.893 – 0.897
3384.6 3982.0 3665.2 3982.0
0.959 0.959 – 0.963
4460.4 4308.9 4321.2 4397.3
3817.4 3817.4 3330.8 3817.4
0.856 0.886 0.771 0.868
4105.6 4105.6 3771.4 4105.6
0.920 0.953 0.873 0.934
Mean COV
0.867 0.052
(a) RB series
0.937 0.035
5000 4500
5000
4000
4500
3000
4000
2500
3500
Tension, T (kN)
Tension, T (kN)
3500
2000 1500
CFST-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-FC-E20 CFST-AG-E40
1000 500 0 0
5
10
15
20
25
Axial elongation,
30
35
3000 2000 1500 1000
40
500
(mm)
0
(b) AG series 5000
A1-5 S1 S2 S3 S4 S5 A3-5
2500
0
1000
CFST-AG series
4500
2000
4000
5000
6000
L( )
(a)
4000 3500
CFST-RB series
450
3000 2500
400
CFST series
2000
Distance, d (mm)
Tension, T (kN)
3000
Longitudinal strain,
1500 1000
Eccentric Concentric
500 0 0
10
20
30
Axial elongation,
40
50
Td
T
350
0
300 250
0.3Tu-test 0.6Tu-test 0.9Tu-test
200 150 100
(mm)
50
(c) Comparison of concentrically and eccentrically loaded specimens
0
Fig. 7. Load-axial elongation (N-Δ) relationships.
0
1000
2000
Longitudinal strain,
Table 3 Comparison of measured and predicted initial stiffness.
3000 L(
4000
)
(b)
Specimens
(EA)′T-test (kN)
(EA)′T-cal (kN)
(EA)′T-cal/(EA)′T-test
CFST-E20 CFST-RB-E20 CFST-RB-E40 CFST-RB-FC-E20 CFST-AG-E20A CFST-AG-E20B CFST-AG-E40 CFST-AG-FC-E20
1.373 × 106 1.601 × 106 1.303 × 106 1.488 × 106 1.638 × 106 1.684 × 106 1.407 × 106 1.650 × 106
1.540 × 106 1.567 × 106 1.238 × 106 1.567 × 106 1.576 × 106 1.576 × 106 1.245 × 106 1.576 × 106 Mean COV
1.122 0.979 0.950 1.053 0.962 0.936 0.885 0.955 0.980 0.075
Fig. 8. Load-longitudinal strain relationships.
from the AIJ guide [9], AISC specification [10], Eurocode 4 [11] and DBJ specification [12]. It was found that the AIJ guide and AISC specification gave the most conservative prediction, and the results predicted by the Eurocode 4 and DBJ specification were similar and were closer to the test results [7]. Therefore in this study, the equation in DBJ specification [12] was adopted tentatively. For the CFST with reinforcing bars or angles, the T-M relationship may be different from that of the CFST counterpart because of the embedded components. According to the experimental results, an equation is proposed to describe the T-M relationship of CFST with reinforcing bars or angles as follows:
Previously Han et al. [7] proposed the calculation method for the flexural strength of the CFST, which was adopted in the DBJ specification [12]. Comparisons were also made between the calculation results
98
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
5000
contributions of CFST and embedded components together as follows:
4500
Tur = (1.1–0.4α ) fy As + fysr Asr
(5a)
Tur = (1.1–0.4α ) fy As + fysa Asa
(5b)
Tension, T (kN)
4000 3500 3000
where fy is the yield strength of the steel tube; As is the cross-sectional area of the steel tube; α is the steel ratio defined by α = As / Ac; Ac is the cross-sectional area of the concrete; fysr and fysa are the yield stress of the reinforcing bars and the reinforcing angles, respectively. In Eq. (4), Mur is the ultimate flexural strength of CFST with reinforcing bars and angles, which can be regarded as the summation of contributions from two parts tentatively, i.e., the concrete and steel tube part (the CFST part), as well as the embedded components part. The expressions of the flexural strength of reinforced CFST are as follows:
1-1 1-2 1-3 1-4
2500 2000 1500 1000 500 0 -500
-400
-300
-200
Hoop strain,
-100
H(
0
Mur = Msc,u + Mp
)
Tension, T (kN)
(a) CFST-RB-E20 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 -500
Msc,u = γm Wsc fscy
-300
Hoop strain,
-200 H(
-100
0
)
(b) CFST-AG-E20 5000 4500
1-1 1-2 1-3 1-4
Tension, T (kN)
4000 3500 3000
5. Conclusions
2500 2000
The behaviour of the reinforced CFSTs with reinforcing bars or angles under eccentric tension was investigated experimentally in this study. The failure modes, elastic tensile stiffness and the tensionmoment relationships of test specimens were analysed. The following conclusions could be drawn within the limited scope of this study:
1500 1000 500 0 -500
-400
-300
Hoop strain,
-200 H(
-100
0
1) All test specimens showed ductility behaviour during the test. Although the reinforcing bars or angles were not connected to the steel tube directly, the embedded components worked effectively with the core concrete attributing to the inner ribs and bonding effect. 2) The embedded reinforcing bars or angle significantly enhanced the strength of the eccentrically loaded tensile member. The tensile strengths of specimens CFST-RB-E20 and CFST-AG-E20 were 17.6% and 24.2% higher than those of CFST counterparts without embedded components, respectively. The tensile strength decreased with the increase of load eccentricity. Meanwhile, the elastic tensile stiffness of eccentrically loaded CFST specimen with reinforcing bars or angles was lower than that of the concentrically loaded counterpart.
)
(c) CFST-E20 Fig. 9. Load-hoop strain relationships.
⎛ T ⎞2 M ≤1 ⎜ ⎟ + ⎝ Tur ⎠ Mur
(7)
In Eq. (7), γm is the flexural strength index [7], representing the influence of the confinement degree to the strength, γm = 1.1 + 0.48 ⋅ ln (ξ + 0.1), where the confinement factor ξ is determined by αfy/fck, α is the steel ratio for CFST cross-section, α = As / Ac; fy is the yield stress of outer steel tube, fck is the characteristic strength of concrete cube. In Eq. (7), Wsc = πD3/32, D is the diameter of the outer tube. fscy is the ‘nominal yield strength’ of the composite section [7], fscy = (1.14 + 1.02ξ)fck, fck is the characteristic compressive strength of concrete. The predicted tensile strengths for eccentrically loaded specimens by Eq. (3) (Tu-Eq3) and Eq. (4) (Tu-Eq4) are both presented in Table 4, where the measured results (Tu-test) are also presented for comparison. The average value and the COV of the Tu-Eq4/Tu-test ratio are 0.937 and 0.032, respectively, while the average value and the COV of the Tu-Eq3/ Tu-test ratio are 0.867 and 0.048, respectively. Fig. 11(b) depicts the measured test results (denoted in points) and the predicted tension versus moment relationships. It can be seen that most data points are above the lines predicted by Eqs. (3) and (4) expect the data point CFST-E20. It is due to the fact that the specimen CFST-E20 has no embedded component inside. In general reasonable accuracy has been achieved for the predicted results.
1-1 1-2 1-3 1-4
-400
(6)
where, Msc,u is the ultimate flexural strength of the CFST part, Mp is the ultimate flexural strength of the cross section of embedded components. Msc,u can be expressed as follows [12]:
(4)
The curves of Eqs. (3) and (4) are presented in Fig. 11(a). In Eq. (4), Tur is the ultimate tensile strength of CFST with reinforcing bars and angles. Chen et al. [18] has found that it can be predicted by adding the 99
Journal of Constructional Steel Research 136 (2017) 91–100
J. Chen et al.
Outer tube
Tension
Outer tube
Tension
Tension
Ribs
Tension
Embedded bars Concrete or angles
Embedded bars or angles
Concrete
(a)
Tension
(b) Fig. 10. Load transfer mechanism in the specimen.
1.2
In general good agreement was achieved between the predictions and test results.
1
Acknowledgements Eq.(4)
T/Tu
0.8
The research work described in this paper was supported by a project from the Science and Technology Department of Zhejiang Province (2015C33005).
0.6 0.4 Eq.(3)
References
0.2 [1] L.H. Han, W. Li, R. Bjorhovde, Developments and advanced applications of concrete-filled steel tubular (CFST) structures: members, J. Constr. Steel Res. 100 (9) (2014) 211–228. [2] B. Uy, Local and postlocal buckling of fabricated steel and composite cross sections, J. Struct. Eng. ASCE 127 (6) (2001) 666–677. [3] T. Perea, R.T. Leon, J.F. Hajjar, M.D. Denavit, Full-scale tests of slender concretefilled tubes: axial behavior, J. Struct. Eng. 139 (7) (2013) 1249–1262 (2013). [4] C.W. Roeder, D.E. Lehman, E. Bishop, Strength and stiffness of circular concretefilled tubes, J. Struct. Eng. ASCE 136 (12) (2010) 1545–1553. [5] J. Moon, C.W. Roeder, D.E. Lehman, H. Lee, Analytical modeling of bending of circular concrete-filled steel tubes, Eng. Struct. (2012) 349–361. [6] J. Nie, Y. Wang, J. Fan, Experimental study on seismic behavior of concrete filled steel tube columns under pure torsion and compression–torsion cyclic load, J. Constr. Steel Res. 79 (12) (2013) 115–126. [7] L.H. Han, Further study on the flexural behavior of concrete-filled steel tubes, J. Constr. Steel Res. 62 (6) (2006) 554–565. [8] H. Lu, L.H. Han, X.L. Zhao, Analytical behavior of circular concrete-filled thinwalled steel tubes subjected to bending, Thin-Walled Struct. 47 (3) (2009) 346–358. [9] AIJ, Recommendations for Design and Construction of Concrete Filled Steel Tubular Structures, Architectural Institute of Japan (AIJ), Tokyo, Japan, 2008. [10] ANSI/AISC 360-10, Specification for Structural Steel Buildings, American Institute of Steel Construction (AISC), Chicago, USA, 2010. [11] EN 1994-1-1, Design of Composite Steel and Concrete Structures — Part 1-1: General Rules and Rules for Buildings. Eurocode 4, European Standard, CEN, Brussels, 2004 (2004). [12] DBJ/T13-51-2010, Technical Specifications for Concrete-Filled Steel Tubular Structures, The Housing and Urban-Rural Development Department of Fujian Province, Fuzhou, China, 2010 (in Chinese). [13] Y.G. Pan, S.T. Zhong, Constitutive relations of concrete filled steel tube under tension, Ind. Constr. 20 (4) (1990) 33–37 (in Chinese). [14] L.H. Han, S.H. He, F.Y. Liao, Performance and calculations of concrete filled steel tubes (CFST) under axial tension, J. Constr. Steel Res. 67 (11) (2011) 1699–1709. [15] W. Li, L.H. Han, T.M. Chan, Numerical investigation on the performance of concrete-filled double-skin steel tubular members under tension, Thin-Walled Struct. 79 (6) (2014) 108–118. [16] W. Li, L.H. Han, T.M. Chan, Tensile behaviour of concrete-filled double-skin steel tubular members, J. Constr. Steel Res. 99 (8) (2014) 35–46. [17] W. Li, L.H. Han, T.M. Chan, Performance of concrete-filled steel tubes subjected to eccentric tension, J. Struct. Eng. ASCE 141 (12) (2015) (04015049). [18] J. Chen, J. Wang, W.L. Jin, Concrete-filled steel tubes with reinforcing bars or angles under axial tension, J. Constr. Steel Res. (2017) (Accepted).
0 0
0.2
0.4
0.6
0.8
1
1.2
M/Mu (a) 5000 4800
Eq.(3)
CFST-AG-E20 series
4600
Eq.(4)
CFST-AG-FC-E20
T (kN)
4400 CFST-AG-E40
4200
CFST-RB-E2
4000
CFST-RB-FC-E2
3800 3600
CFST-AG series
CFST-RB series
3400 CFST-E20
3200 3000 0
50
100
150
200
250
300
M (kN m)
(b) Fig. 11. Tension-moment relationship of the reinforced CFST.
3) Design equations were proposed to predict the stiffness and the tension-moment interaction relationship of reinforced CFST under eccentric tension, where the flexural strength of the CFST with reinforcing bars or angles can be tentatively estimated by adding the flexural strength of CFST and the embedded components together.
100