You are on page 1of 20

Journal of Constructional Steel Research 164 (2020) 105765

Contents lists available at ScienceDirect

Journal of Constructional Steel Research


journal homepage: www.elsevier.com/locate/ijcard

Assessment of high-strength concrete encased steel composite


columns subject to axial compression
Binglin Lai a, J.Y. Richard Liew a, *, Akshay Venkateshwaran a, Shan Li a,
Mingxiang Xiong a, b
a
Department of Civil and Environmental Engineering, National University of Singapore, E1A-07-03, 1 Engineering Drive 2, 117576, Singapore
b
School of Civil Engineering, Guangzhou University, Guangzhou 510006, China

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

Article history: This paper investigates the axial load capacity of concrete encased steel composite stub columns with
Received 20 May 2019 high strength concrete and steel materials. A total of 14 column specimens with varying material
Received in revised form strengths and different steel section shapes were tested under concentric compression load. The test
3 August 2019
results revealed that the design methods in EN 1994-1-1 and JGJ 138-2016 overestimate the axial load
Accepted 13 September 2019
Available online 13 November 2019
capacity of high strength concrete encased steel columns. Adding a small percentage of steel fiber in the
high strength concrete was found to improve the compression resistance of the composite section. A new
test database consisting of 51 partially encased composite sections and 82 fully encased composite
Keywords:
Axial capacity
sections was established, covering a wide range of section geometric and material grades. Parametric
Concrete confinement study was then carried out to assess current design methods in predicting the axial capacity of such
Concrete encased steel columns composite columns with respect to material strengths, steel contribution ratio, reinforcement ratio,
High strength concrete section slenderness ratio, confined concrete area ratio, and concrete confinement efficiency. The effec-
High strength steel tiveness of concrete confinement in partially encased composite column was evaluated and a simplified
Spalling method was proposed to compute the enhanced concrete strength based on regression analysis. For fully
concrete encased composite columns, a concrete strength reduction factor was proposed to be used with
EN 1994-1-1 to predict the compression resistance. Design recommendation was made considering the
early cover spalling of high strength concrete and the material compatibility between steel and concrete
in composite column design.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction column connection details [1]. FEC columns outperform Concrete


Filled Tubular (CFT) members in terms of fire and corrosion resis-
Concrete Encased Steel (CES) composite columns has been of tance as the steel section is completely covered by concrete. It is
interest for many researchers due to its excellent structural per- experimentally validated by Zhu et al. [2] and Wang et al. [3] that
formance under both static and seismic loading conditions. The steel section can greatly improve the shear resistance of the com-
application of such columns can be found in basement construction posite columns. For PEC column, concrete is filled in between the
of high-rise buildings and metro railways stations in which such steel flanges. As the steel flange is not externally restrained by the
columns are used to resist the dynamic load due to train movement. concrete, additional check is needed on the plate slenderness ratio
As shown in Fig. 1, CES columns can be divided into two categories: to avoid local buckling.
Fully Encased Composite (FEC) column and Partially Encased With the advancement of material technologies, high strength
Composite (PEC) column. FEC column generally has the similar and ultra-high strength concrete have become commercially
geometric configurations to conventional Reinforced Concrete (RC) available, which motivates the utilization of high strength materials
column except for the presence of steel section, which can be H- in composite columns [4e7]. High strength concrete has been
shaped, cruciform-shaped and even consists of several steel sec- proven to work well with high strength steel in CFT members based
tions depending on specific structural requirement or the beam on previous experimental research [8e10]. As such, a new design
guide [11] was proposed to extend the current EN 1994-1-1 method
to C90/105 concrete and S550 steel for CFT members. However, CES
* Corresponding author. columns behaves in a different manner compared with CFT
E-mail address: ceeljy@nus.edu.sg (J.Y.R. Liew).

https://doi.org/10.1016/j.jcsr.2019.105765
0143-974X/© 2019 Elsevier Ltd. All rights reserved.
2 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Fig. 1. Concrete encased steel composite columns with different types of steel sections.

counterpart. Without continuous and sufficient confinement effect, Some experimental work has been conducted to study the
CES columns always fail by concrete crushing. Therefore, due to the structural behaviour of CES columns, including the axial perfor-
concerns on the inherent brittleness of high strength concrete and mance of stub columns [16e27], slender columns under concentric
the scarcity of relevant experimental work, current design codes, or eccentric compression [28e32], biaxial bending [33,34],
including EN 1994-1-1 [12], AISC 360-16 [13] and JGJ 138-2016 [14], combining constant axial force and transverse loading [35,36], as
all placed a restriction on the use of high strength material. Based well as cyclic loading [37e39]. Most of these tests focused on
on previous studies carried out by Lai et al. [15], which employs normal concrete. Apart from experimental works, some numerical
C100 concrete and S355 steel, high strength CES columns have been simulations were also carried out by Lai et al. [40], Ellobody et al.
successfully applied in several buildings in Singapore. As shown in [41], El-Tawil & Deierlein [42] and Liu et al. [43], which uses
Fig. 2, CES columns are adopted in the basement construction, advanced technique to perform cross-section analysis and member
resisting the high compression force from the superstructure above stability analysis considering material and geometric non-
the ground level. Sufficient concrete cover must be provided to the linearities. Among the above-mentioned research, the axial per-
steel section to achieve minimum 3 h fire rate for such basement formance has attracted much attention, and some analytical
columns. models were proposed to assess the concrete confinement effect

Fig. 2. Concrete Encased Steel (CES) composite columns used in top down construction.
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 3

Fig. 3. Concrete encased steel composite columns: (a). Different concrete confinement regions; (b) Spalling of concrete cover.

[44e46] based on the modification of confinement model devel- showed that the squash load predicted by EC4 method is higher
oped by Mander et al. [47] for RC columns. Most of these models than the test results when the high strength concrete
divide concrete regions into unconfined, partially confined and (fc ¼ 109 MPa) was used.
highly concrete regions. As illustrated in Fig. 3(a), concrete cover Early cover spalling was observed by Eid et al. [48] in high
outside stirrups is regarded as unconfined concrete, and the con- strength concrete columns. As illustrated in Fig. 3(b), the spalling of
crete region within steel flanges is categorized as highly confined concrete cover prevents the cross-section reaching the full plastic
region, in which the confining pressure is provided by the lateral resistance and results in a lower-than-expected load resistance.
reinforcement and the steel section. The remaining zone is partially Same phenomenon was also observed by Foster et al. [49], and it is
confinement region, which is bounded by stirrups and locates believed that cover spalling is an inevitable consequence of the
outside steel section. Considering the non-uniform distribution of placement of steel reinforcements but become noticeable when
confining pressure, effective confined area is incorporated by high strength concrete is used. For normal strength concrete, the
assuming the arch action between two adjacent longitudinal bars effect of cover spalling may be disguised by the strength
and the tips of steel flanges, with the initial tangential angle of 45 . enhancement due to concrete confinement effect [50]. As an
Zhao et al. [45] and Chen & Wu [46] proposed the stress-strain effective technique to minimize the brittleness of high strength
relations of concrete based on different confining conditions, and concrete, steel fibers has been used in CFT members [8,51] and RC
the biaxial stress state of structural steel section is also considered members [50,52]. It was discovered that the addition of steel fiber
to generate the constitutive law of steel in the axial direction. did not significantly improve the compressive performance of high
Despite the existence of confinement effect validated by Chen & strength CFT columns [8,51], but it helped to address the vulnera-
Lin [44] in normal strength CES stub columns, recent experimental bility to premature cover spalling for high strength RC columns
study conducted by Zhu et al. [16] revealed that steel section greatly [50,52].
affects the lateral expansion of concrete and hence the confinement So far, no design guideline allows for the use of high strength
model derived from RC columns may not be applicable for high materials in CES columns, and there is no unified approach to
strength CES stub columns. Tests done by Huang et al. [22] also predict the axial load capacity for a wide range of material

Fig. 4. Concrete encased steel composite sections used in the experimental program.
4 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Table 1
Geometric and material properties of test specimens.

Specimen B  D (mm) b  h  tw  tf (mm) L (mm) fc (MPa) fys (MPa) fus (MPa) fyr (MPa) Vf (%)

S500-C90-SP90-H 280  280 180  180  10  18 700 95.9 590 657 578 0
S500-C90-SP60-H 280  280 180  180  10  18 700 88.7 590 657 578 0
S500-C90-SP60-C 280  280 80  180  10  18 700 92.9 590 657 578 0
S500-C90SF-SP60-H 280  280 180  180  10  18 700 93.2 590 657 578 0.5
S500-C130-SP90-H 300  300 200  200  10  18 700 136.6 590 657 578 0
S500-C130-SP60-H 300  300 200  200  10  18 700 135.3 590 657 578 0
S500-C130-SP60-C 300  300 100  200  10  18 700 132.8 590 657 578 0
S500-C130SF-SP60-H 300  300 200  200  10  18 700 126.6 590 657 578 0.5
S690-C90-SP60-H 260  260 160  160  10  18 600 94.6 739 863 578 0
S690-C90SF-SP60-H 260  260 160  160  10  18 600 93.6 739 863 578 0.5
S690-C130-SP60-H 260  260 160  160  10  18 600 127.4 739 863 578 0
S690-C130SF-SP60-H 260  260 160  160  10  18 600 128.7 739 863 578 0.5
S690-C130-SP60-C 260  260 70  160  10  18 600 136.4 739 863 578 0
S690-C130SF-SP60-C 260  260 70  160  10  18 600 134.6 739 863 578 0.5

Note: fc, fys, fus and fyr refer to concrete compressive cylinder strength, steel section yield strength, steel section ultimate strength and yield strength of longitudinal rein-
forcement bars, respectively. Vf refers to the fiber content by volume.

strengths. Besides, literature survey of previous research indicates a 2000-ton actuator test frame in NUS Structural Lab. The test spec-
conflict viewpoint over whether concrete confinement effect or imens include two concrete grades of C90 and C130 and two steel
concrete cover spalling shall be incorporated in the prediction, grades of S500 and S690. As shown in Fig. 4, rectilinear stirrups
especially when high strength concrete is used. To address the were adopted and spaced at either 90 mm or 60 mm interval, giv-
research gaps outlined in the above, this paper investigates the ing the volumetric ratio of stirrups rv ranging from 1.33% to 2.35%.
axial load capacity of CES stub columns from the perspective of Several specimens were reinforced with 0.5% micro steel fibers by
design, inclusive of both FEC columns and PEC columns. This paper volume, aiming to overcome the potential cover spalling issue as
starts with a brief summary of experimental program, followed by mentioned before.
the comparison between test results and analytical results pre- The specimen details are shown in Table 1. Different cross-
dicted by EN 1994-1-1 and JGJ 138-2016. Then a statistic study is sectional dimensions are designed to cater for the capacity of test
carried out based on the database including 51 PEC column data machine. For specimens with S500 steel section, the column height
and 82 FEC column data. Finally, a formula is proposed to assess the is set to be 700 mm, while the specimens with S690 steel are
confinement effect of PEC columns and recommendations are put 600 mm high. The actual concrete compressive strength was ob-
forward for the design of FEC columns with high strength concrete tained by testing the 100  200 mm cylinders at the same day of
and high strength steel. column test. As reported in Table 1, the actual compressive strength
of C90 concrete ranges from 88.7 MPa to 95.9 MPa, and that of C130
concrete ranges from 126.6 MPa to 136.6 MPa. The yield strength of
2. Summary of experimental program
steel section is taken as the minimum value of web and flange plate.
For S500 steel section, the yield strength reaches 590 MPa and for
2.1. Specimens and materials
S690 it is 739 MPa, both of which exceed the material strength limit
stipulated in EN 1994-1-1.
A total of 14 CES stub column specimens were tested using the

Fig. 5. Test setup and instrumentation: (a) reaction frame; (b) 2000 tons actuator; (c) Test setup and instrumentation.
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 5

Fig. 6. Typical failure patterns of test specimens.

Each specimen is designated with a string of letters and specimen with S500 steel section and C130 concrete mixed with
numbers, which begins with steel grade, followed by concrete 0.5% steel fiber, the clear spacing of stirrups is 60 mm, and the steel
grade and stirrup spacing, and end with the shape of steel section. section is H-shaped.
For example, the label “S500-C130SF-SP60-H” refers to the The compression machine and test setup is shown in Fig. 5. A

Fig. 7. Load-axial shortening curves of test specimens.


6 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Table 2
Comparison between test result of this experimental program and analytical prediction using EN 1994-1-1 and JGJ 138-2016.

Specimen rv (%) NEC4 (kN) NJGJ (kN) NTest (kN) NTest/NEC4 NTest/NJGJ

S500-C90-SP90-H 1.44 10,943 10,747 9611 0.878 0.894


S500-C90-SP60-H 2.15 10,520 10,299 9926 0.944 0.964
S500-C90-SP60-C 2.15 11,085 10,839 10,156 0.916 0.937
S500-C90SF-SP60-H 2.15 10,784 10,579 11,241 1.042 1.063
S500-C130-SP90-H 1.33 15,130 15,094 12,003 0.793 0.795
S500-C130-SP60-H 1.99 15,042 15,001 12,862 0.855 0.857
S500-C130-SP60-C 1.99 15,606 15,453 13,569 0.869 0.878
S500-C130SF-SP60-H 1.99 14,451 14,375 13,253 0.917 0.922
S690-C90-SP60-H 2.35 10,574 10,277 9327 0.882 0.908
S690-C90SF-SP60-H 2.35 10,526 10,226 10,601 1.007 1.037
S690-C130-SP60-H 2.35 12,233 12,034 10,317 0.843 0.857
S690-C130SF-SP60-H 2.35 12,300 12,105 11,284 0.917 0.932
S690-C130-SP60-C 2.35 12,951 12,745 10,796 0.834 0.847
S690-C130SF-SP60-C 2.35 12,861 12,649 10,947 0.851 0.865
Mena 0.896 0.911
St.dev 0.065 0.070

Table 3
Specimen details of partially encased columns.

Ref. Specimen designation Structural steel section (b  h  tw  tf) Stirrup spacing Longitudinal bars As/Ac Ar/Ac d fc fy,s fy,r Cross-section Type

Type Size Class mm % % MPa MPa MPa

[53] SRC-1-2-1 Ca 170  360  20  20 1 e e 33.02 0.00 0.76 35 280 e A


SRC-1-2-2 C 170  360  20  20 1 e e 33.02 0.00 0.80 27 280 e A
SRC-1-2-3 C 170  360  20  20 1 e e 33.02 0.00 0.75 36.7 280 e A
SRC-2-2 C 166  366  16  25 1 e e 33.59 0.00 0.80 28.2 285 e A
SRC-3-2 C 175  357  25  26 1 e e 34.06 0.00 0.81 27.3 285 e A
SRC-4-2 C 160  370  10  30 1 e e 32.13 0.00 0.77 27.7 235 e A
SRC-5-2 C 180  350  30  10 1 e e 33.14 0.00 0.78 27.7 235 e A
SRC-6-2 C 180  370  30  20 1 e e 42.29 0.00 0.78 36.6 235 e A
SRC-7-2 C 160  350  10  20 1 e e 24.00 0.00 0.68 38.5 280 e A
SRC-8-2 C 140  360  20  25 1 e e 37.33 0.00 0.81 29.7 280 e A
SRC-9-2 C 208  360  20  16 1 e e 29.70 0.00 0.78 28.3 280 e A
SRC-1-3 C 170  360  20  20 1 e e 33.02 0.00 0.80 36.7 380 e A
SRC-2-3 C 166  366  16  25 1 e e 33.59 0.00 0.80 38.1 380 e A
SRC-3-3 C 175  357  25  26 1 e e 34.06 0.00 0.80 38.1 380 e A
SRC-4-3 C 160  370  10  30 1 e e 32.13 0.00 0.79 38.9 355 e A
SRC-5-3 C 180  350  30  10 1 e e 33.14 0.00 0.79 38.9 355 e A
SRC-6-3 C 180  370  30  20 1 e e 42.29 0.00 0.83 36.7 355 e A
SRC-7-3 C 160  350  10  20 1 e e 24.00 0.00 0.76 36.7 380 e A
SRC-8-3 C 140  360  20  25 1 e e 37.33 0.00 0.81 38.9 380 e A
SRC-9-3 C 208  360  20  16 1 e e 29.70 0.00 0.78 38.9 380 e A
[45] SRC-A-1 C 160  330  10  10 3 e e 15.87 0.00 0.63 47.4 437 e A
SRC-A-2 C 160  330  10  10 3 e e 15.87 0.00 0.57 64.7 460 e A
SRC-B-1 C 170  360  20  20 1 e e 33.02 0.00 0.79 47.4 457 e A
SRC-B-2 C 170  360  20  20 1 e e 33.02 0.00 0.75 64.7 508 e A
SRC-B-3 C 170  360  20  20 1 e e 33.02 0.00 0.81 45.7 515 e A
SRC-C-2 C 110  330  10  10 1 e e 15.00 0.00 0.62 47.4 437 e A
[54] PEC-1 C 60  150  3  6 1 100 e 14.88 0.00 0.64 25 257 e B
PEC-2 C 60  150  3  6 1 100 4d14 14.88 4.06 0.53 25 257 297 B
PEC-3 C 60  150  3  6 1 100 4d14 14.88 4.06 0.53 25 257 297 B
[27] A-04 C 112  450  10  14 1 e e 19.84 0.00 0.77 28.44 416 e C
A-06 C 168  450  10  14 1 e e 16.87 0.00 0.75 28.14 416 e C
A-08 C 224  450  10  14 3 e e 16.00 0.00 0.74 28.14 416 e C
A-10 C 280  450  10  14 3 e e 16.06 0.00 0.74 28.14 416 e C
[55,56] H3 H 400  400  8  8 4 120 e 6.29 0.00 0.33 60 394 e D
H4 H 400  400  8  8 4 200 e 6.29 0.00 0.33 58.9 394 e D
H5 H 400  400  8  8 4 400 e 6.29 0.00 0.32 61.7 394 e D
[57] C2 H 450  450  9.7  9.7 4 225 e 6.81 0.00 0.48 32.7 370 e D
C3 H 450  450  9.7  9.7 4 337.5 e 6.81 0.00 0.48 32.4 370 e D
C4 H 450  450  9.7  9.7 4 450 e 6.81 0.00 0.48 31.9 370 e D
C5 H 450  450  9.7  9.7 4 225 e 6.81 0.00 0.46 34.3 370 e D
C6 H 450  450  6.35  6.35 4 337.5 e 4.38 0.00 0.37 32.7 374 e D
C7 H 450  450  6.35  6.35 4 300 e 6.68 0.00 0.48 31.9 374 e D
[58] C8 H 600  600  12.88  12.88 4 600 e 6.78 0.00 0.46 34.2 360 e D
C9 H 600  600  12.91  12.91 4 600 e 6.79 0.00 0.46 34.2 360 e D
C10 H 600  600  12.81  12.81 4 300 e 6.74 0.00 0.45 34.2 360 e D
C11 H 600  600  9.71  9.71 4 600 e 5.04 0.00 0.37 34.2 345 e D
C12 H 600  600  12.86  12.86 4 300 e 6.77 0.00 0.46 34.2 360 e D
[59] P1 H 125  131  3.18  3.18 4 95 4d8 7.95 1.34 0.32 55 323 569 E
P1R H 125  131  3.18  3.18 4 95 4d8 7.95 1.34 0.32 55 323 569 E
P2 H 125  131  3.18  3.18 4 63 12d4.2 7.93 1.12 0.32 55 323 600 F
P2R H 125  131  3.18  3.18 4 63 12d4.2 7.93 1.12 0.32 55 323 600 F
a
Where “C”, “H”, “I” denote steel section with Cruciform shape, H-shape and I-shape, respectively.
Table 4
Specimen details of fully encased columns.

Ref Specimen designation Cross-section (mm) Structural steel section Main bars Transverse bars d As/Ac fc fy,s fy,r fy,h Steel fiber Cross-section type

Type Size Configuration Type % MPa MPa MPa MPa

[16] C-I-M40 200  200 I H68  100  4.5  7.6 12d10 D6.5@40 Ma 0.10 3.76 93 254 427 335 0 A
C-I-M60 200  200 I H68  100  4.5  7.6 12d10 D6.5@60 M 0.10 3.76 93 254 427 335 0 A
C-I-R40 200  200 I H68  100  4.5  7.6 12d10 D6.5@40 R 0.10 3.76 93 254 427 335 0 B
C-I-R60 200  200 I H68  100  4.5  7.6 12d10 D6.5@60 R 0.10 3.76 93 254 427 335 0 B
C-þ-M40 200  200 C 2H68  100  4.5  7.6 12d10 D6.5@40 M 0.18 7.90 94 254 427 335 0 C
C-þ-M60 200  200 C 2H68  100  4.5  7.6 12d10 D6.5@60 M 0.18 7.90 94 254 427 335 0 C
C-þ-M80 200  200 C 2H68  100  4.5  7.6 12d10 D6.5@80 M 0.18 7.90 94 254 427 335 0 C
C-þ-R40 200  200 C 2H68  100  4.5  7.6 12d10 D6.5@40 R 0.18 7.90 94 254 427 335 0 D
C-þ-R60 200  200 C 2H68  100  4.5  7.6 12d10 D6.5@60 R 0.18 7.90 94 254 427 335 0 D
[17] SRC1 280  280 H H150  150  7  10 12d16 D8@140 R 0.31 5.58 29.5 296 350 453 0 E
SRC2 280  280 H H150  150  7  10 12d16 D8@75 R 0.32 5.58 28.1 296 350 453 0 E
SRC3 280  280 H H150  150  7  10 12d16 D8@35 R 0.31 5.58 29.8 296 350 453 0 E
SRC4 280  280 C 2H175  90  5  8 12d16 D8@140 R 0.37 6.42 29.8 345 350 453 0 F

B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765


SRC5 280  280 C 2H175  90  5  8 12d16 D8@75 R 0.37 6.42 29.8 345 350 453 0 F
SRC6 280  280 C 2H175  90  5  8 12d16 D8@35 R 0.38 6.42 29.5 345 350 453 0 F
SRC7 280  280 I H150  75  5  7 12d16 D8@140 R 0.17 2.41 28.1 303 350 453 0 B
SRC8 280  280 I H150  75  5  7 12d16 D8@75 R 0.18 2.41 26.4 303 350 453 0 B
SRC9 280  280 I H150  75  5  7 12d16 D8@140 R 0.17 2.41 28.1 303 350 453 0 B
SRC10 280  280 I H150  75  5  7 12d16 D8@75 R 0.17 2.41 29.8 303 350 453 0 B
[18] CSRC1 400  400 H H150  150  7  10 12d16 D8@100 M 0.21 2.54 26.88 301.7 356.7 340 0 G
CSRC2 400  400 H H150  150  7  10 12d16 D8@100 M 0.18 2.54 35.2 301.7 356.7 340 0 G
CSRC3 400  400 H H150  150  7  10 12d16 D8@200 M 0.18 2.54 35.2 301.7 356.7 340 0 G
CSRC4 400  400 H H150  150  7  10 12d16 D8@100 M 0.25 3.28 26.88 296.7 356.7 340 0 G
CSRC5 400  400 H H150  150  7  10 12d16 D8@100 M 0.21 3.28 35.2 296.7 356.7 340 0 G
[19] src1 280  280 C 2H175  90  5  8 4d16 D8@140 R 0.40 6.28 23.9 274 453 606 0 F
src2 280  280 C 2H175  90  5  8 4d16 D8@100 R 0.41 6.28 23.5 274 453 606 0 F
src3 280  280 C 2H175  90  5  8 12d16 D8@100 R 0.34 6.42 21.8 274 453 606 0 F
src4 280  280 C 2H175  90  5  8 12d16 D8@100 R 0.32 6.42 25.3 274 453 606 0 F
src5 280  280 C 2H160  50  3.2  4.5 4d16 D8@190 R 0.20 2.45 26 271 453 606 0 F
src6 280  280 C 2H160  50  3.2  4.5 4d16 D8@140 R 0.20 2.45 26.3 271 453 606 0 F
src7 280  280 C 2H160  50  3.2  4.5 12d16 D8@140 R 0.16 2.50 25 271 453 606 0 F
src8 280  280 C 2H160  50  3.2  4.5 4d16 D8@100 R 0.20 2.45 26.6 271 453 606 0 F
src9 280  280 C 2H160  50  3.2  4.5 12d16 D8@100 R 0.16 2.50 24.6 271 453 606 0 F
src10 280  280 C 2H160  50  3.2  4.5 12d16 D8@100 R 0.16 2.50 24.3 271 453 606 0 F
[20] C4 200  200 I S.I$B No.10 4d12 D8@200 R 0.24 2.75 18.48 240 400 320 0 H
C9 200  200 I S.I$B No.10 4d12 D8@100 R 0.24 2.75 18.48 240 400 320 0 H
[21] GJ-1 300  300 H H200  200  8  12 8d12 D8@100 R 0.23 7.49 104.2 368.7 503.6 486 0 I
GJ-2 300  300 H H200  200  8  12 8d12 D8@100 R 0.23 7.49 102.6 368.7 503.6 486 0 I
GJ-3 300  300 H H200  200  8  12 8d12 D8@75 R 0.21 7.49 115.2 368.7 503.6 486 0 I
GJ-4 300  300 H H200  200  8  12 8d12 D8@50 R 0.23 7.49 105.3 368.7 503.6 486 0 I
GJ-5 300  300 H H200  200  8  12 8d12 D8@50 R 0.20 7.49 126.9 368.7 503.6 486 0.3%ma I
GJ-6 300  300 H H200  200  8  12 8d12 D8@50 R 0.19 7.49 130.3 368.7 503.6 486 0.4%m I
GJ-7 300  300 H H200  200  8  12 8d12 D8@50 R 0.19 7.49 135.7 368.7 503.6 486 0.5%m I
[22] C50e0 240  240 H H154.4  161.8  8  11.5 8d16 D6@150 R 0.38 9.08 51.1 363 356.8 339.4 0 I
C100e0 240  240 H H154.4  161.8  8  11.5 8d16 D6@150 R 0.24 9.08 109.3 363 356.8 339.4 0 I
C100F 240  240 H H154.4  161.8  8  11.5 e e e 0.23 8.81 123.8 363 0 0 1%m J
Authors C10 250  250 H H154.4  161.8  8  11.5 e e e 0.26 8.06 101.3 381 0 0 0.5%h þ 0.5%m J
C20 250  250 H H154.4  161.8  8  11.5 e e e 0.24 8.06 111.6 381 0 0 1.0%h þ 0.5%m J
C20A 250  250 H H154.4  161.8  8  11.5 e e e 0.25 8.06 107.4 381 0 0 1.0%h J
C30 250  250 H H154.4  161.8  8  11.5 e e e 0.23 8.06 120.4 381 0 0 1.5%h þ 0.5%m J
C50-S120-SF0 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@120 R 0.33 7.16 52.3 375 550 510 0 I
C120-S120-SF0 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@120 R 0.19 7.16 125.6 375 550 510 0 I
C120-S90-SF0 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@90 R 0.19 7.16 125.6 375 550 510 0 I
C120-S60-SF0 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@60 R 0.19 7.16 125.6 375 550 510 0 I
(continued on next page)

7
Table 4 (continued )

8
Ref Specimen designation Cross-section (mm) Structural steel section Main bars Transverse bars d As/Ac fc fy,s fy,r fy,h Steel fiber Cross-section type

Type Size Configuration Type % MPa MPa MPa MPa

C120-S120-SF5 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@120 R 0.18 7.16 126.7 375 550 510 0.5%m I
C120-S60-SF5 240  240 H H152.9  157.6  6.5  9.4 8d13 D10@60 R 0.18 7.16 126.7 375 550 510 0.5%m I
C120-S60-SF0-C 240  240 C 2H70  140  10  12 8d13 D10@60 R 0.26 10.95 125.6 380 550 510 0 K
[23] DH-TI-75 600  600 C 2H175  350  6  9 12d29 D13@75 R 0.25 3.00 34.6 437 439 472 0 F
DH-TI-90 600  600 C 2H175  350  6  9 12d29 D13@90 R 0.25 3.00 34.6 437 439 472 0 F
DH-SP-60 600  600 C 2H175  350  6  9 16d25 D13@60 SP 0.25 3.00 34.6 437 451 472 0 L
DH-SP-75 600  600 C 2H175  350  6  9 16d25 D13@75 SP 0.25 3.00 34.6 437 451 472 0 L
DH-SP-95 600  600 C 2H175  350  6  9 16d25 D13@95 SP 0.25 3.00 34.6 437 451 472 0 L
DH-SP-115 600  600 C 2H175  350  6  9 16d25 D13@115 SP 0.25 3.00 34.6 437 451 472 0 L
[24] HSRC-SP1 D411 C 2H120  206  12  12 20d10 D10@50 SP 0.54 8.24 37 517 426 426 0 M
HSRC-SP2 D411 C 2H120  206  12  12 20d10 D10@50 SP 0.54 8.24 37 517 426 426 0 M
[25] SRC4 600  600 C 2H100  220  6  9 12d25 D13@125 SP þ M 0.16 1.72 32.3 386 469 463 0 N
SRC5 600  600 C 2H175  350  6  9 12d25 D13@150 SP þ M 0.24 2.98 32.3 386 469 463 0 N
SRC6 600  600 C 2H175  350  6  9 12d25 D13@125 SP þ M 0.24 2.98 32.3 386 469 463 0 N

B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765


[26] SCN4A 100  100 H H20  20  5  5 4d8 D6@50 R 0.22 2.62 27 350 470 415 0 H
SCN4B 100  100 H H25  25  5  5 4d8 D6@50 R 0.27 3.43 27 350 470 415 0 H
SCH6A 150  150 H H30  30  5  5 4d8 D6@75 R 0.14 1.83 41 350 470 415 0 H
SCH6B 150  150 H H45  45  5  5 4d8 D6@75 R 0.20 2.88 41 350 470 415 0 H
[27] SRC-1A-HP3-28 560  560 C 2H168  450  10  14 4d25 D10@110 R 0.48 6.04 28.14 416 436 337 0 O
SRC-1A-HP3-42 560  560 C 2H168  450  10  14 4d25 D10@110 R 0.48 6.04 28.14 416 436 433 0 O
SRC-1A-HP4-28 560  560 C 2H168  450  10  14 4d25 D13@110 R 0.48 6.04 28.14 416 436 302 0 O
SRC-1A-HP4-42 560  560 C 2H168  450  10  14 4d25 D13@110 R 0.48 6.04 28.14 416 436 426 0 O
SRC-3A-HP4-42 560  560 C 2H168  450  10  14 12d25 D13@110 M 0.44 6.12 28.14 416 436 426 0 P
SRC-1B-HP3-28 670  670 C 2H168  450  10  14 4d25 D10@110 R 0.40 4.14 28.14 416 436 337 0 O
SRC-1B-HP3-42 670  670 C 2H168  450  10  14 4d25 D10@110 R 0.40 4.14 28.14 416 436 433 0 O
SRC-1B-HP4-28 670  670 C 2H168  450  10  14 4d25 D13@110 R 0.40 4.14 28.14 416 436 302 0 O
SRC-1B-HP4-42 670  670 C 2H168  450  10  14 4d25 D13@110 R 0.40 4.14 28.14 416 436 426 0 O
SRC-3B-HP4-42 670  670 C 2H168  450  10  14 12d25 D13@110 M 0.37 4.17 28.14 416 436 426 0 P
a
Where “M”, “R”, “SP” denote multiple stirrup, rectilinear stirrup, and spiral stirrup, respectively. “h” and “m” denotes hook-end fiber and micro straight fiber, respectively.
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 9

Fig. 8. Geometric configurations of partially encased steel composite columns reported in literature [27,45,53e59].

mega-shape short steel column attached to the actuator was used is compared with analytical predictions using EN 1994-1-1 and JGJ
to transfer the load to specimens. Concentric compression was 138-2016. Both design codes adopt similar conception to predict
ensured by adjusting the specimen location until the readings of the axial load capacity by summing the compression resistance of
the 4 LVDTs roughly coincide with each other. Displacement- each material component. The difference lies in the treatment of
controlled mode was employed in the loading process and pre- reduction factor to account for unfavourable conditions under pure
loading was applied up to 20% of predicted maximum load. Actual compression scenario. The formulae in the two design codes are
loading was then applied with a rate of 0.2e0.3 mm/min, and the given in Eqs. (1) and (2).
test terminated when the stirrup failed and the concrete crushed
severely. EN 1994  1  1 NEC4 ¼ 0:85Ac fc þ As fys þ Ar fyr (1)

 
2.2. Test results JGJ 138  2016 NJGJ ¼ 0:9 Ac fc þ As fys þ Ar fyr (2)

The typical failure modes are shown in Fig. 6. Specimens with where Ac, As and Ar refer to area of concrete, steel section and
high strength plain concrete C90 or C130 exhibit similar failure longitudinal reinforcement bars, respectively.
mode, which is characterised by premature spalling of concrete The comparison between the test and predicted results is given
cove, followed by the local bucking of longitudinal reinforcement in Table 2. EN 1994-1-1 and JGJ 138-2016 give very similar pre-
bars, outward bending or even fracture of lateral reinforcement diction, although the value of NTest/NEC4 is slightly lower than NTest/
bars, and the crushing of the concrete enclosed by stirrups. For the NJGJ. For all specimens with C130 concrete, the predicted capacity
fiber reinforced composite columns, spalling of concrete cover was are all higher than the test results, and adding 0.5% steel fiber still
not apparently observed. Several longitudinal, horizontal and di- generates overestimation. This indicates that the 0.5% steel fiber
agonal cracks initiated and propagated on the concrete surface and may not be sufficient to overcome the brittleness of C130 concrete.
the steel fibers were pulled out from the cracks. The observed Comparing the specimens with C90 and C130 concrete, the former
phenomenon was similar to those reported by Huang et al. [22] on generally has a higher ratio of test to predicted result, and the same
their physical tests on the fiber reinforced CES column without trend is also observed for test specimens with S500 and S690 steel,
reinforcement bars. which reveals the material compatibility issues between high
The load-axial shortening behaviour is plotted in Fig. 7. It can be strength concrete and high strength steel. This issue will be further
seen that most specimens behave in a ductile manner with a very elaborated in the statistical analysis.
smooth transition from the ascending branch to the post-peak
behaviour. For some specimens with cruciform steel section, the
residual strength even reaches comparable to the peak load. The 3. Statistical analysis
comparison between the specimens with concrete C90 and C130
indicates that the former exhibit a more ductile post-peak behav- 3.1. Database of test results
iour. Addition of steel fibers enhances both the ultimate load and
residual strength, but this effect is more pronounced for specimens This section assesses the accuracy of the current design method
with C90 concrete. Since this study mainly aims to evaluate the in predicting the compression resistance of CES columns, and per-
axial load capacity from experimental and statistical analysis, the forms parametric study to identify key parameters that affect the
detailed information of strain development in steel and concrete is accuracy of the predictions. A database is established based on
not discussed here. available test data from the literature. A total of 51 test data on PEC
The maximum load obtained from the load-displacement curve columns and 82 test data on FEC columns were summarized in
10 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Fig. 9. Geometric configurations of fully encased steel composite columns reported in literature [16e27].

Fig. 10. Two types of steel fibers used in fibe-reinforced CES columns.
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 11

Fig. 11. Comparison between test results and analytical results predicted by EN 1994-1-1 and JGJ 139-2016.

reinforcement.
As for the FEC columns, a total of 16 cross-section types were
reported from previous research. Some are reinforced with recti-
linear stirrups [16,17,19e23,26,27], some have multiple stirrups
with cross ties [16,18,27]. The specimens tested by Liang et al. [23]
and Shih et al. [25] were heavily reinforced with more than one
types of stirrups as depicted in Fig. 9 (L) & (N), resulting in signif-
icant strength enhancement due to confinement effect. Huang et al.
[22] investigated the behaviour of fiber reinforced CES columns
without reinforcement bars, and Chen et al. [24] studied the axial
response of circular CES columns with spiral stirrups.
Some unpublished tests previously conducted by the authors
are also summarized in Table 4, which consists of 11 test data on
CES columns with and without steel fiber. Two types of steel fibers
are used, and the close up view is given in Fig. 10. The hook-end
fiber is 60 mm long with 0.9 mm diameter, and the micro steel fi-
Fig. 12. Evaluation of axial capacity of partially encased composite columns based on ber is 13 mm in length with very small diameter of 0.16 mm. The
section classification of structural steel. experimental program reported in this paper adopted the micro
*
Where “C þ 1” denotes Class 1 cruciform section, “C þ 3” denotes Class 3 cruciform steel fibers as shown in Fig. 10(b).
section, “H þ 4” denotes Class 4H-section.
As plotted in Fig. 11, the comparisons between the test result and
analytical result indicate that both EN 1994-1-1 and JGJ 138-2016
predict the axial load capacity of CES columns with same level of
accuracy, although the former is a bit closer to test result for both
PEC and FEC columns. The following study will only focus on EN
1994-1-1 for the ease of analysis. Since PEC and FEC columns
behave in quite different manner and the key design parameter also
differs, the statistical analysis will be undertaken separately so as to
provide more reliable and specific conclusions.

3.2. Evaluation on partially encased composite (PEC) columns

3.2.1. Axial load capacity


As PEC column is filled with concrete within steel flanges, only
Fig. 13. Width and thickness of steel flange and steel web. the web component is fully restrained by concrete encasement,
whereas the flanges are susceptible to outward local buckling un-
less a certain requirement of width to thickness ratio is fulfilled. As
Tables 3 and 4, respectively. It shall be noted that only the double listed in Table 3, according to EN 1993-1-1 [60], all the PEC columns
symmetric CES stub columns subjected to pure compression are with H-section are categorized into Class 4 section, while cruciform
included in the database. The cross-section configurations of sections have either Class 1 or Class 3 steel section. Fig. 12 dem-
collected specimens varied widely as sketched in Fig. 8 and Fig. 9. It onstrates that most specimens with Class 4 steel section cannot
can be seen that some PEC columns are internally reinforced with reach the EC4 predicted axial capacity, indicating that the load
longitudinal reinforcement bar and stirrups [54e59], while some carrying capacity of PEC columns may be restricted by the elastic
are just infilled with plain concrete [27,45,53] without any local buckling of slender steel section.
12 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Fig. 14. Influence of different parameters on the accuracy of EN 1994-1-1 prediction for partially encased columns.

3.2.2. Parametric study sffiffiffiffiffiffiffiffiffi


As the load-carrying capacity of PEC columns is affected by 2bf þ tw 235
several design parameters, this section provides some design  44 (3)
tf fy
guidance by evaluating NTest/NEC4 with the variation of material
strength, steel area ratio As/Ac, steel contribution ratio d, as well as where the definition of bf, tf, hw and tw are illustrated in Fig. 13. Since
the width to thickness ratio of steel flange bf/tf and steel web hw/tw. most specimens in Table 3 have steel yield strength about 400 MPa,
According to EN 1994-1-1, steel contribution ratio is calculated as: and the thickness of web and flange are almost identical, the width
d ¼ Asfys/NEC4. As for the width to thickness ratio of steel section, EN to thickness ratio of flange (bf/tf) for all specimens can be approx-
1994-1-1 stipulated the following limitations as Eq. (3): imately calculated as 16.36 for the ease of further comparative
study.
Since current test data only covers the concrete compressive
strength ranging from 25 MPa to 65 MPa and steel yield strength
within 235 MPa to 515 MPa, no apparent tendency is observed in
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 13

Fig. 15. comparison between the test result and analytical predictions using the effective width of Class 4 section: (a) Effective width method in EN 1993-1-5; (b) Comparison.

ensuring the width to thickness ratio is within the Class 3 section


range.
As EC4 does not provide any explicit solution for PEC columns
with slender steel section (Class 4), the concept of effective cross-
section is adopted here following the same procedure specified in
EN 1993-1-5 [61]. As illustrated in Fig. 15(a), the effective area of
steel flange is determined by incorporating a reduction factor r. The
full equations are available in EN 1993-1-5, Clause 4.4 [61].
Adopting the effective width method, conservative estimation can
be attained except for a few outliers as plotted in Fig. 15(b). The
local buckling of steel section occurs at 55% of maximum load as
reported by Pereira et al. [59], thus leading to the unexpected low
capacity, which may attribute to the initial imperfections existed in
the steel sections. Nevertheless, the remaining test data reported in
[55,57,58] indicates the conservative prediction, revealing that the
effective width method shall be used to predict the axial load ca-
pacity of PEC columns with slender steel section.

3.2.3. Evaluation of concrete confinement effect


As seen from Figs. 12 and 14, some test results are much higher
than EC4 prediction by 30%, which may result from confinement
effect provided by steel sections. As sketched in Fig. 16, the effective
confined area of PEC columns is considered to be the zone bounded
by parabolas connecting two adjacent flange tips [44e46,53]. It is
believed that both the steel flange and web contribute to the
confining pressure acting on concrete. Steel flange can be viewed as
Fig. 16. Effective confined region and confining mechanism of partially encased a cantilever that deflects outward freely due to lateral deformation
composite columns. of concrete. The intersection of flange and web functions as a rigid
component thus a bending moment will be generated to resist the
expansion of concrete. Steel web is under biaxial stress state as a
Fig. 14(a, b) regarding the effect of material strength variation on result of axial compression and lateral tension caused by the con-
the accuracy of EC4’s predictions. From Fig. 14(c, d), with the in- crete expansion. Therefore, the width to thickness ratio of both
crease of steel area ratio As/Ac from 0.05 to 0.4, the ratio of test to steel flange and web is believed to exert certain effect on the
EC4’s prediction also display an increasing trend, indicating that concrete confinement efficiency. In additions, steel area ratio, steel
steel section may plays a major role on the load carrying capacity of contribution ratio can also affect confinement effect as demon-
PEC columns. On the other hand, higher steel area ratio always strated in Fig. 14.
generates higher level of concrete confinement effect, which results Assuming steel section reaches the yield strength at maximum
in concrete strength enhancement and leads to higher axial load load, the confined compressive strength fcc can be back-calculated
capacity accordingly. This similar trend is observed for the steel by substituting NTest into Eq. (1). The strength enhancement ratio
contribution ratio. The higher the steel contribution ratio, the fcc/fc is plotted as functions of steel area ratio (As/Ac), steel contri-
higher the capacity is attained. Different from the effect of steel area bution ratio (d), and the width to thickness ratio of steel flange (bf/
ratio, width to thickness ratio of steel flange and web exhibit a tf) and web (hw/tw). It is observed from Fig. 17 that using Class 4
negative effect. As indicated from Fig. 14(e), violating the EC4 steel section is unfavourable as fcc/fc is always <1.0, in which case
requirement calculated from Eq. (3) always results in NTest/NEC4 PEC columns may fail by steel section elastic buckling instead of
<1.0, suggesting that local buckling of steel section must be concrete crushing. Excluding the outliers highlighted in Fig. 17(a),
considered for PEC column, otherwise it should be avoided by the trend line is plotted to give the best fit of each data distribution.
14 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Fig. 17. Relations of confinement effect with different parameters.

In order to improve confinement effect, both the steel area ratio and !0:25  0:27
steel contribution ratio shall be increased, while the width to fcc bf hw
thickness ratio of steel section must be restricted. ¼ 1:24e1:95As =Ac þ 0:04e1:3d  0:03  1:45
fc tf tw
Based on the regression analysis of collected data, a formula is
proposed to evaluate the confinement effect for PEC columns as (4)
below Eq. (4): As shown in Fig. 18, good agreement is reached between test
results and analytical predictions using the proposed model. It
should be noted that this model is derived from test date with
concrete compressive strength within 25 MPa to 65 MPa and steel
yield strength within 235 MPa to 515 MPa, beyond this limitation
the proposed formulae shall be re-evaluated.

3.3. Evaluation on fully encased composite (FEC) columns

3.3.1. Axial load capacity


Different from PEC columns, FEC column is not susceptible to
local buckling due to the fully restraint of concrete encasement. A
general comparison between test and design prediction is given in
Fig. 19. FEC columns with cruciform steel section have excellent
load-carrying capacity compared with I-section and H-section.
Another useful technique to improve the load capacity is to change
the configuration of lateral reinforcement. Fig. 19(b) indicates that
using multiple stirrups with cross-ties or spiral hoops can lead to
much higher resistance.

Fig. 18. Comparison between the fcc/fc from physical test and fcc/fc from proposed 3.3.2. Parametric study
model. As indicated in Fig. 20(a, b), using high strength materials
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 15

Fig. 19. Evaluation of axial capacity of fully encased composite columns based on: (a) Steel section shape; (b) Configuration of stirrups.

always lead to overestimation on the axial load capacity. For FEC longitudinal reinforcement bars [52]. Once the interfacial crack is
columns with concrete compressive strength <50 MPa, the average initiated, core concrete applies certain lateral force to the main bars.
ratio of test to prediction is 1.13, however it drops down to 0.95 Combining with the second-order effect induced by axial force,
when concrete strength increase beyond 50 MPa. Similar trend is longitudinal bars tend to buckle outward, which further exacerbate
also observed with the increase of steel yield strength. The aver- cover spalling.
aged NTest/NEC4 is 1.07 and 0.90 for normal strength steel and high In order to provide insightful guidance from design purpose, and
strength steel beyond 550 MPa. Other parameters, including steel make the design prediction on the safe side, confinement effect is
areas ratio, steel contribution ratio, longitudinal reinforcement neglected in this paper, although it really helps to improve the post-
ratio and confined area ratio (ratio of core concrete area to gross peak ductility and residual strength [16]. Based on the review of
area), are graphically illustrated in Fig. 20(cef). It is interesting to international building codes, the axial load capacity of CES columns
find that increasing steel area ratio results in the decrease of NTest/ can be generalized below in Eq. (5):
NEC4, which is completely different from the findings on PEC col-
umns. Nevertheless, it may not be prudent to conclude the negative 0 0
NPre ¼ acc Ac fc þ As fys þ Ar fyr (5)
correlation between As/Ac and the axial load capacity, since all test
data differs a lot in other parameters, which may disguise the effect where acc is the concrete strength reduction coefficient associated
of As/Ac. As for the other three parameters, no clear tendency is with concrete grade and other factors inclusive of size effect, long
observed due to the scattered data distribution. term effect, and the difference between in-place concrete and
Fig. 20(g) demonstrates that the increase of stirrup volumetric concrete cured in standard cylinder moulds, etc. fy0 refers to the
ratio does not bring about the increase of NTest/NEC4, which may effective strength of high strength steel, which can be determined
attribute to the large variation of concrete strength. Since high following the material compatibility criterion as illustrated in
strength concrete requires higher confining ratio to overcome the Fig. 22. It shall be noted that the concrete axial strain corresponding
inherent brittleness [62], a non-dimensional confinement index to peak stress is capped at 0.0028 according to EN 1992-1-1 [63],
rvfyh/fc proposed by Razvi & Saatcioglu [62] is adopted here for and hence the effective strength will be lower than the yield
more appropriate assessment. As indicated from Fig. 20(h), this strength when high strength steel over 550 MPa is used.
confinement index clearly displays a positive correlation with NTest/ Substituting the NPre in Eq. (5) with test result NTest, acc can be
NEC4, thus it shall be viewed as a critical criterion to judge the back calculated. As plotted in Fig. 23, acc decreases with the in-
design of transverse reinforcement, especially for FEC columns crease of concrete compressive strength. The value 0.85 adopted by
with high strength concrete. This index will be further analysed in EN 1994-1-1 [12] and AISC 360-16 [13] is conservative for concrete
the following subsection. strength below 50 MPa, but apparently it is not feasible for high
strength concrete exceeding 90 MPa. The equations stipulated in
3.3.3. Evaluation of concrete strength reduction coefficient building code CSA-A23.3 [64], NZS 3101 [65], AS 3600 [66], ACI ITG-
Based on current confinement models for CES composite section 4.3R [67] and proposed by Awati & Khadiranaikar [68], Kim et al.
[42,44e46], concrete at different zones exhibits quite different [69] are summarized in Table 5. Although most of these models
constitutive laws. The maximum stress of unconfined and confined place limitations on concrete strength range, they are still used here
concrete does not occur simultaneously, and hence the axial ca- for comparison purpose.
pacity cannot be directly predicted unless the full range load- Among these models, CSA-A23.3 [64] gives the most conserva-
deformation relation is generated, which is beyond the scope of tive estimations, but it seems to be overly conservative for normal
this study and also not practical for design implementation. Addi- strength concrete. Other models can predict the normal strength
tionally, the early cover spalling observed in high strength concrete CES columns with sufficient margin of safety, but they generate
may also offset the benefit brought by concrete confinement effect. overestimations when concrete compressive strength is beyond
As sketched in Fig. 21, cover spalling occurs due to the formation of 90 MPa, so more reduction is needed to give conservative predic-
weak plane between the concrete core and concrete cover. Core tion. To this end, following the form of reduction factor given by
concrete is confined by the stirrups whereas the concrete cover Liew et al. [11], a new formula is proposed in Eqs. (12) and (13) to
expands outward without any restraint. Consequently, the concrete evaluate the axial capacity of CES columns as an extension of EN
cover will delaminate from the specimens and results in a loss of 1994-1-1. Since EN 1994-1-1 limits the maximum concrete grade at
capacity. Another source of cover spalling may be the bending of C50, the new proposed formulae introduces a coefficient h to
16 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Fig. 20. Influence of various parameters on the accuracy of EN 1994-1-1 prediction for fully encased composite columns.
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 17

account for the brittleness of concrete with compressive strength


exceeding 50 MPa.

acc ¼ 0:85h (12)

. . .
 1:0  ðf  50Þ 200 50N mm2 < fc  90N mm2
c
h¼ .
0:8 fc > 90N mm2
(13)
As the research on CES columns with ultrahigh strength con-
crete is still limited, it is not advisable to impose excessive reduc-
tion on concrete compressive strength just in order to make all the
test data in Fig. 23(b) fall above the propose method. Based on the
previous research on high strength and ultrahigh strength concrete
conducted by Liew et al. [11], h shall be kept constant at 0.8 when
concrete strength exceeding 90 MPa.
Fig. 21. Mechanism of spalling of concrete cover. As observed from Fig. 23, for some specimens with concrete
C120/C130, acc falls below the prediction of all models in Table 5
and the proposed method as well, which indicates that CES col-
umn with such high strength concrete may not present the desired
excellent axial load capacity, despite the significant cost input to
produce high strength concrete. Based on the current usage of high
strength concrete in high-rise buildings, the maximum concrete
grade in CES columns shall be conservatively capped at C100, and
the corresponding structural steel grade shall be limited to S550
based on the material compatibility criterion illustrated in Fig. 22.
Beyond this limit the material strength cannot be fully developed,
thus the increase in the cost of production does not bring about
proportional increase in the axial capacity.
The relation between acc and the confinement index rsfyh/fc is
plotted in Fig. 24, which can be written as acc ¼ 0.6(rvfyh/fc)0.184
based on the fitting of data distribution. For fully encased steel
columns with high strength concrete, increasing the volumetric
ratio of transverse reinforcement is helpful to increase the load-
carrying capacity and make the best use of high strength mate-
rials. Reducing the stirrup bar spacing, using spiral hoops, and
changing the arrangement of stirrups from rectilinear configuration
to multiple configuration are all applicable from the statistical
analysis mentioned above. Since the practical range of transverse
reinforcement volumetric ratio is restricted to cater for concreting
Fig. 22. Determination of effective strength of high strength steel. casting issue, another technique is to employ higher grade
confining steel, which is proved to be effective in confining high
strength concrete columns [62].

Fig. 23. Comparison of acc values from tests with predictions from (a) design codes; (b) analytical models [68,69] and proposed method.
18 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

Table 5
Formulae of acc in different design codes and analytical models.

Design codes Equations

CSA-A23.3 [64] acc ¼ 0.85  0.0015fc  0.67 20 MPa  fc  80 MPa (6)



NZS 3101 [65] 0:85 fc  55 MPa (7)
acc ¼
0:85  0:004ðfc  55Þ  0:75 55 MPa  fc  70 MPa

AS 3600 [66] acc ¼ 1  0:003fc (8)


0:72  acc  0:85
8
ACI ITG-4.3R [67] < 0:85 fc  55 MPa (9)
acc ¼ 0:85  0:0022ðfc  55Þ 55 MPa  fc  124 MPa
:
0:7 fc > 124 MPa

Analytical models

Awati and Khadiranaikar [68] 0:85 fc  70 MPa (10)
acc ¼
0:85  0:002ðfc  70Þ  0:75 70 MPa  fc  120 MPa

Kim et al. [69] 0:85 fc  69 MPa (11)
acc ¼
0:85  0:003ðfc  69Þ  0:75 fc > 69 MPa

the axial capacity of CES stub columns with very high strength
concrete.
3. For partially encased stub columns with concrete compressive
strength ranging from 25 to 65 MPa and steel yield strength
within 235 to 515 MPa, the key parameters affecting the design
prediction are steel area ratio, steel contribution ratio, as well as
the slenderness ratio of steel sections. Class 4 steel section
should be avoided for partially encased composite section un-
less the local buckling of the flange plate is accounted for by
using the effective width concept.
4. Based on the regression analysis, a simplified equation is pro-
posed to assess the confinement effect of partially encased
composite section, which takes into account of steel area ratio,
steel contribution ratio, width to thickness ratio of steel flange
and web, and good agreement with test results is observed.
5. Adopting cruciform steel section and stirrups with cross-ties or
spiral hoops enhances the axial load capacity of fully encased
Fig. 24. Coefficient acc as a function of confinement index rsfyh/fc. composite sections. EC4’s method gives conservative estimation
of the composite sections made of normal strength steel section
and concrete with compressive strength up to 50 MPa.
4. Conclusions 6. A non-dimensional confinement index rvfyh/fc parameter is
proposed to evaluate the axial load capacity of fully encased
Fourteen high strength CES stub columns were tested under steel composite sections, especially when high strength con-
pure compression. The tests results were added into a database crete is used. In order to make the best use of high strength
consisting of additional 51 partially encased composite columns concrete, higher volumetric ratio of transverse reinforcement is
and 82 fully encased composite to perform statistical analysis and required or higher strength stirrup bars may be used.
evaluate their axial load capacity. Existing design methods in EN 7. In view of the brittleness of high strength concrete, the
1994-1-1 and JGJ 138-2016 were assessed through the parametric maximum concrete grade allowed in concrete encased steel
study. Key parameters affecting the predictions were identified composite columns shall be conservatively capped at C100, and
accordingly. Finally, an empirical equation was developed to esti- the corresponding structural steel grade shall be limited to S550
mate the concrete confining effect. Design recommendation for based on the material compatibility criterion. Further research is
concrete encased steel composite columns was then proposed. The needed to address the brittleness and early spalling of high
findings reported in this paper support the following conclusions: strength concrete for its application in concrete encased steel
composite columns.
1. Current design methods in EN 1994-1-1 (EC 4) and JGJ 138-2016
cannot be used, without modification, to predict the axial load Acknowledgement
capacity of concrete encased composite columns with high
strength concrete (C90/C130) and high strength steel (S500/ The authors would like to acknowledge the support provided by
S690). The higher the material strength, the less conservative is Building & Construction Authority in Singapore under project grant
the codes’ prediction. R-302-000-168-490. JFE Steel Corporation and Yongnam Engi-
2. Adopting a closer spacing of stirrup reinforcement with 2% neering & Construction are also greatly acknowledged for supply-
volumetric ratio and adding 0.5% steel fiber enable the concrete ing and preparing the test specimens.
encased steel composite columns with C90 concrete and S500/
S690 steel section to reach or exceed the EC4’s predicted axial
References
capacity. However, for the tested specimens with C130 concrete,
the predicted load is still not conservative compared to the test [1] T. Bogdan, M. Chrzanowski, Mega columns with several reinforced steel
results, and higher dosage of steel fibers is needed to improve profileseexperimental and numerical investigations, in: Proceedings of the
12th International Conference on Advances in Steel-Concrete Composite
B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765 19

Structures, ASCCS 2018, 2018 Jun 5, pp. 267e274. [31] C.S. Kim, H.G. Park, K.S. Chung, I.R. Choi, Eccentric axial load testing for
[2] M. Zhu, J. Liu, Q. Wang, X. Feng, Experimental research on square steel tubular concrete-encased steel columns using 800 MPa steel and 100 MPa concrete,
columns filled with steel-reinforced self-consolidating high-strength concrete J. Struct. Eng. 138 (8) (2012) 1019e1031.
under axial load, Eng. Struct. 32 (8) (2010) 2278e2286. [32] C.S. Kim, H.G. Park, K.S. Chung, I.R. Choi, Eccentric axial load capacity of high-
[3] Q. Wang, D. Zhao, P. Guan, Experimental study on the strength and ductility of strength steel-concrete composite columns of various sectional shapes,
steel tubular columns filled with steel-reinforced concrete, Eng. Struct. 26 (7) J. Struct. Eng. 140 (4) (2013), 04013091.
(2004) 907e915. [33] P.R. Munoz, C.T. Hsu, Behavior of biaxially loaded concrete-encased composite
[4] M. Khan, B. Uy, Z. Tao, F. Mashiri, Behaviour and design of short high-strength columns, J. Struct. Eng. 123 (9) (1997) 1163e1171.
steel welded box and concrete-filled tube (CFT) sections, Eng. Struct. 147 [34] C. Dundar, S. Tokgoz, A.K. Tanrikulu, T. Baran, Behaviour of reinforced and
(2017) 458e472. concrete-encased composite columns subjected to biaxial bending and axial
[5] L.H. Han, C.C. Hou, X.L. Zhao, K.J. Rasmussen, Behaviour of high-strength load, Build. Environ. 43 (6) (2008) 1109e1120.
concrete filled steel tubes under transverse impact loading, J. Constr. Steel [35] S.A. Mirza, V. Hyttinen, E. Hyttinen, Physical tests and analyses of composite
Res. 92 (2014) 25e39. steel-concrete beam-columns, J. Struct. Eng. 122 (11) (1996) 1317e1326.
[6] N. Saleh, A. Ashour, D. Lam, T. Sheehan, Experimental investigation of bond [36] T. Naka, K. Morita, M. Tachibana, Strength and hysteretic characteristics of
behaviour of two common GFRP bar types in highestrength concrete, Constr. steel-reinforced concrete columns, Trans. AIJ 250 (1997) 47e58 (in Japanese).
Build. Mater. 201 (2019) 610e622. [37] J.M. Ricles, S.D. Paboojian, Seismic performance of steel-encased composite
[7] M. Khan, B. Uy, Z. Tao, F. Mashiri, Concentrically loaded slender square hollow columns, J. Struct. Eng. 120 (8) (1994) 2474e2494.
and composite columns incorporating high strength properties, Eng. Struct. [38] C. Campian, Z. Nagy, M. Pop, Behavior of fully encased steel-concrete com-
131 (2017) 69e89. posite columns subjected to monotonic and cyclic loading, Procedia Eng. 117
[8] M.X. Xiong, D.X. Xiong, J.Y.R. Liew, Axial performance of short concrete filled (2015) 439e451.
steel tubes with high-and ultra-high-strength materials, Eng. Struct. 136 [39] W. Zhu, J. Jia, J. Gao, F. Zhang, Experimental study on steel reinforced high-
(2017) 494e510. strength concrete columns under cyclic lateral force and constant axial load,
[9] M.X. Xiong, D.X. Xiong, J.Y.R. Liew, Behaviour of steel tubular members infilled Eng. Struct. 125 (2016) 191e204.
with ultra high strength concrete, J. Constr. Steel Res. 138 (2017 Nov 1) [40] B. Lai, J.Y. Liew, S. Li, Finite element analysis of concrete-encased steel com-
168e183. posite columns with off-center steel section, in: Proceedings of the 12th In-
[10] M.X. Xiong, D.X. Xiong, J.Y.R. Liew, Flexural performance of concrete filled ternational Conference on Advances in Steel-Concrete Composite Structures,
tubes with high tensile steel and ultra-high strength concrete, J. Constr. Steel ASCCS 2018, 2018 Jun 5, pp. 297e303.
Res. 132 (2017) 191e202. [41] E. Ellobody, B. Young, D. Lam, Eccentrically loaded concrete encased steel
[11] J.Y.R. Liew, M.X. Xiong, D.X. Xiong, Design of concrete filled tubular beam- composite columns, Thin-Walled Struct. 49 (1) (2011) 53e65.
columns with high strength steel and concrete, Structures 8 (2016) 213e226. [42] S. El-Tawil, G.G. Deierlein, Strength and ductility of concrete encased com-
[12] EN 1994-1-1, Eurocode 4: Design of Composite Steel and Concrete Structures- posite columns, J. Struct. Eng. 125 (9) (1999) 1009e1019.
Part 1-1, General Rules and Rules for Buildings, 2004. [43] S.W. Liu, Y.P. Liu, S.L. Chan, Advanced analysis of hybrid steel and concrete
[13] ANSI/AISC 360-16, Specification for Structural Steel Buildings, American frames: part 1: cross-section analysis technique and second-order analysis,
Institute of Steel Construction (AISC), Chicago, USA, 2016. J. Constr. Steel Res. 70 (3) (2012) 26e36.
[14] JGJ 138-2016, Code for Design of Composite Structures, Ministry of Housing [44] C.C. Chen, N.J. Lin, Analytical model for predicting axial capacity and behavior
and Urban-Rural Development of the People’s Republic of China, China, 2016 of concrete encased steel composite stub columns, J. Constr. Steel Res. 62
(In Chinese). (2006) 424e433.
[15] B.L. Lai, J.Y.R. Liew, T.Y. Wang, Buckling behaviour of high strength concrete [45] X. Zhao, F. Wen, T.M. Chan, S. Cao, Theoretical stressestrain model for con-
encased steel composite columns, J. Constr. Steel Res. 154 (2019) 27e42. crete in steel-reinforced concrete columns, J. Struct. Eng. 145 (4) (2019),
[16] W.Q. Zhu, G. Meng, J.Q. Jia, Experimental studies on axial load performance of 04019009.
high-strength concrete short columns, Proc. Inst. Civ. Eng.-Struct. Build. 167 [46] S. Chen, P. Wu, Analytical model for predicting axial compressive behavior of
(9) (2014) 509e519. steel reinforced concrete column, J. Constr. Steel Res. 128 (2017) 649e660.
[17] C.C. Chen, S.C. Yeh, Ultimate strength of concrete encased steel composite [47] J.B. Mander, M.J. Priestley, R. Park, Theoretical stress-strain model for confined
columns, in: Proceedings of the Third National Conference on Structural En- concrete, J. Struct. Eng. 114 (8) (1988) 1804e1826.
gineering, 1996, pp. 2197e2206 (in Chinese). [48] R. Eid, K. Kovler, I. David, W. Khoury, S. Miller, Behavior and design of high-
[18] Y. Liu, Z.X. Guo, P.H. Xu, L.P. Jia, Experimental study on axial compression strength circular reinforced concrete columns subjected to axial compres-
behavior of core steel reinforced concrete columns, J. Build. Struct. 36 (4) sion, Eng. Struct. 173 (2018) 472e480.
(2015) 68e74 (in Chinese). [49] S.J. Foster, J. Liu, S.A. Sheikh, Cover spalling in HSC columns loaded in
[19] K.C. Tsai, Y. Lien, C.C. Chen, Behaviour of axially loaded steel reinforced con- concentric compression, J. Struct. Eng. 124 (12) (1998) 1431e1437.
crete columns, J. Chin. Inst. Civil Hydraul. Eng. 8 (4) (1996) 535e545 (in [50] S.J. Foster, M.M. Attard, Strength and ductility of fiber-reinforced high-
Chinese). strength concrete columns, J. Struct. Eng. 127 (1) (2001) 28e34.
[20] K.Z. Soliman, A.I. Arafa, T.M. Elrakib, Review of design codes of concrete [51] A. Le Hoang, E. Fehling, B.L. Lai, D.K. Thai, N. Van Chau, Experimental study on
encased steel short columns under axial compression, HSRC J. 9 (2013) structural performance of UHPC and UHPFRC columns confined with steel
134e143. tube, Eng. Struct. 187 (2019) 457e477.
[21] Y. Du, Liew J.Y.R. Qi Honghui, Experimental study on behaviours of high [52] S.J. Foster, On behavior of high-strength concrete columns: cover spalling,
strength concrete encased columns, in: 9th International Conference on Steel steel fibers, and ductility, ACI Struct. J. 98 (4) (2001) 583e589.
and Aluminium Structures (ICSAS19) Bradford, UK, 3e5, July, 2019, [53] X. Zhao, H. Qin, Y.Y. Chen, Experimental study on constitutive model of steel
pp. 979e990. confined concrete in SRC columns with cruciform steel section, J. Build. Struct.
[22] Z. Huang, X. Huang, W. Li, Y. Zhou, L. Sui, J.Y.R. Liew, Experimental behaviour 35 (4) (2014) 268e279 (in Chinese).
of very high-strength concrete-encased steel composite column subjected to [54] M. Ebadi Jamkhaneh, M.A. Kafi, Experimental investigation of octagonal
axial compression and end moment, in: Proceedings of the 12th International partially encased composite columns subject to concentric and eccentric
Conference on Advances in Steel-Concrete Composite Structures, ASCCS 2018, loading, J. Rehabil. Civil Eng. 7 (1) (2019) 113e126.
2018 Jun 5, pp. 323e329. [55] B.S. Prickett, R.G. Driver, Behaviour of partially encased composite columns
[23] C.Y. Liang, C.C. Chen, C.C. Weng, S.Y. Yin, J.C. Wang, Axial compressive made with high performance concrete, in: Structural Engineering Report no
behavior of square composite columns confined by multiple spirals, J. Constr. 262. Dept of Civil and Environmental Engineering, University of Alberta, AB,
Steel Res. 103 (2014) 230e240. Canada, 2006.
[24] S.W. Chen, P. Wu, Q. Liu, Z.X. Hou, L.B. Qiu, Studies on axially compressed SRC [56] M. Begum, R.G. Driver, A.E. Elwi, Behaviour of partially encased composite
column using Q460 high strength steel, in: In the 8th International Conference columns with high strength concrete, Eng. Struct. 56 (2013) 1718e1727.
on Behavior of Steel Structure in Seismic Area, 2015, pp. 131e132. [57] R. Tremblay, B. Massicotte, I. Filion, R. Maranda, Experimental study on the
[25] T.H. Shih, C.C. Chen, C.C. Weng, S.Y. Yin, J.C. Wang, Axial strength and ductility behaviour of partially encased composite columns made with light welded H
of square composite columns with two interlocking spirals, J. Constr. Steel steel shapes under compressive axial loads, in: Proc., 1998 SSRC Annual
Res. 90 (2013) 184e192. Technical Meeting, 1998, pp. 195e204.
[26] M.S. Rahman, M. Begum, R. Ahsan, Comparison between experimental and [58] T. Chicoine, R. Tremblay, B. Massicotte, J.M. Ricles, L.W. Lu, Behavior and
numerical studies of fully encased composite columns, World Acad. Sci. Eng. strength of partially encased composite columns with built-up shapes,
Technol. 10 (6) (2016 Apr 1) 762e769. J. Struct. Eng. 128 (3) (2002) 279e288.
[27] C.C. Chen, C.C. Tao, S.C. Yang, W.M. Tsai, Experimental stress-strain behavior [59] M.F. Pereira, S. De Nardin, A.L. El Debs, Structural behavior of partially encased
of the concrete confined by cross-H steel cross-sectional shape, J. Chin. Inst. composite columns under axial loads, Steel Compos. Struct. 20 (6) (2016)
Civil Hydraul. Eng. 22 (4) (2010) 423e434 (in Chinese). 1305e1322.
[28] G.T. Zhao, C.H. Wang, C.Y. Gao, C.X. Wang, Experiment study on the capacity of [60] EN 1993-1-1, Eurocode 3: Design of Steel Structures-Part 1-1. General Rules
SRC long column subjected to eccentric compression, J. Baotou Univ. Iron Steel and Rules for Buildings, 2005.
Technol. 25 (4) (2006) 384e400 (in Chinese). [61] 1993-1-5, Eurocode 3: Design of Steel Structures-Part 1-5. Plated Structural
[29] Q. Yu, Z.D. Lu, Research on the static performance of eccentric steel reinforced Elements, 2006.
concrete column, Build. Struct. 39 (6) (2009) 34e38 (in Chinese). [62] S.R. Razvi, M. Saatcioglu, Strength and deformability of confined high-strength
[30] D.J. Han, K.S. Kim, A study on the strength and hysteretic characteristics of concrete columns, ACI Struct. J. 91 (6) (1994) 678e687.
steel reinforced concrete columns-in the case of columns loaded eccentrically, [63] EN 1992-1-1, Eurocode 2: Design of Concrete Structures-Part 1e1. General
J. Archit. Inst. Korea 11 (4) (1995) 183e190 (in Korean). Rules and Rules for Buildings, 2004.
20 B. Lai et al. / Journal of Constructional Steel Research 164 (2020) 105765

[64] Canadian Standard Association (CSA) A23.3, Design of Concrete Structures, Report on Structural Design and Detailing for High Strength Concrete in
Mississauga, Canada, 2014. Moderate to High Seismic Applications, American Concrete Institute, 2007.
[65] Standard New Zealand (NZS) 3101, Concrete Structure Standard, Part 1- the [68] M. Awati, R.B. Khadiranaikar, Behavior of concentrically loaded high perfor-
Design of Concrete Structures, New Zealand, 2006. mance concrete tied columns, Eng. Struct. 37 (2012) 76e87.
[66] AS (Australian Standard), Concrete structures, in: AS 3600-2009, Australia [69] S. Kim, H.C. Mertol, S. Rizkalla, P. Zia, A. Mirmiran, Behavior of high-strength
Standard, Sydney, Australia, 2009. concrete rectangular columns, in: Seventh International Congress on Ad-
[67] ACI (American Concrete Institute) Innovation Task Group (ITG), 4.3 R-07, vances in Civil Engineering, 2006 Oct 11, pp. 1e10.

You might also like