Accepted Manuscript

Self-Centering Capacity of a Structural Frame Composed of

Steel-Jacketed Concrete Columns and Steel Beams

Hiroyuki Nakahara, Hao Yin

PII: S2352-0124(17)30038-3
DOI: doi: 10.1016/j.istruc.2017.06.004
Reference: ISTRUC 202
To appear in: Structures
Received date: 16 March 2016
Revised date: 12 June 2017
Accepted date: 12 June 2017

Please cite this article as: Hiroyuki Nakahara, Hao Yin , Self-Centering Capacity of
a Structural Frame Composed of Steel-Jacketed Concrete Columns and Steel Beams,
Structures (2017), doi: 10.1016/j.istruc.2017.06.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

Self-Centering Capacity of a Structural Frame Composed of Steel-Jacketed

Concrete Columns and Steel Beams
Hiroyuki NAKAHARA1 and Hao YIN2
1.Graduate School of Engineering, Nagasaki University, Nagasaki, Japan
E-mail: nakaharahiroyuki@nagasaki-u.ac.jp
2.Department of Engineering, A Plus Design Consultancy Co., Ltd., Sichuan, China

PT
E-mail: yinhao7176@163.com

Abstract

RI
A structural frame composed of steel-jacketed concrete columns and steel beams with damage control

SC
capacity has been studied for the structural performances through the experiments. Four specimens,
which were 1/3 scaling model of cruciform subassemblies, were made and subjected to cyclic force in
order to be investigated on its self-centering capacities mainly. The test results showed that this
NU
structure remained lower damage after the cyclic loading test than the normal reinforced concrete
structure. The load carrying capacities of the specimens were estimated through full plastic moment of
the section of the column by taking into consideration of the confining effect of concrete jacketed by
steel tube.
MA

1. Introduction

In 1995, Hyogo-ken Nanbu Earthquake attacked Kobe City in Japan and killed more than six
D

thousand people. Although the earthquake broke down some buildings completely, most of
reinforced concrete buildings remained with slight damages. The remained buildings were,
TE

however, removed because of its residual deformations and apparent cracks. To decrease the
number of removed buildings which still possessed enough seismic performances, many
researchers focused on the damage control for the buildings after earthquakes.
P
CE

The damage control capacity includes the reducing the residual deformation and the
preventing the noticeable cracks, which bring users anxieties to collapse the building. Based
on the purpose of making the damage controlled building, the authors proposed structural
AC

frame which was composed of steel beams and concrete columns jacked by steel tubes
referred to the experimental work by Yoshimura [5], who showed that the jacketed concrete
column possessed efficient ductility. The column included diaphragms to connect steel beams.
The cruciform specimen was made by combining the pre-casted column and steel beams each
other by bolts.

There are many papers which focused on the self-centering behaviors of many kinds of
structural types. These studies are summarized and overviewed by Pessiki [1] which reviews
55 references about the studies of self-centering behaviors mainly conducted by researchers
in Lehigh University. The post-tensioned concrete wall was investigated by Seo et.al. [3].
The post-tensioned steel connections by Ricles et.al. [2]. Our test specimens include these
typical features: 1) steel-jacketed concrete columns which fail in flexure were used, 2) new
diaphragm was proposed and used, 3) there is no longitudinal bar but tensile tendons which
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

penetrate to the columns. The tests of the fifteen specimens were conducted under axial force
and cyclic force corresponding to the situation under seismic excitation. In this paper, four
specimens are chosen to be described its structural properties.

The load versus deformation relations obtained by the test showed significant self-centering
capacities and stable behaviors without abrupt strength deterioration. The flexural failure of
the column occurred in the all of the specimens. The self-centering behaviors derived from
the elastic behavior of unbounded steel bar built in the columns and axial force sustained by
the column. The load carrying capacities were also investigated by comparing to the full
plastic moment of the section of the column by taking the confining effect of concrete
jacketed by steel tube into consideration.

PT
2. Specimen and Test Setup

RI
Fig.1 shows the test specimens which are 1/3 scaling model of cruciform subassemblies.
They were composed of concrete column confined by steel tube and H-shape steel beams.
The specimen includes the diaphragms made by split-Ts and steel plates as shown in Fig.1 (b).

SC
The beam flanges were connected to the diaphragms with high strength bolts. The diaphragm
of type C was composed of two split-Ts made by cutting off from H-shape steel members and
9mm thick steel plate which connected them each other by welding. The test matrix is shown
NU
as Table 1. The main parameters were 1) type of assembly diaphragm, 2) strength of steel bar,
3) axial force ratio, and 4) with or without the duct in the concrete column.
MA

Table 1: Test matrix

Assembly Yield strength Tensile strength

Specimen Steel bar (mm) Duct
diaphragm (MPa) (MPa)

D

EHN19D Type. E PC13 1243 1288

CLS16N SS4009 325 438
TE

CLS19N Type. C SS40010

327 452
CLN14N SS40013
P
CE
AC
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

(8)
(1) (9)
(10)
(2)
(3)

9 165 9

type.C

(11)

(4) (5) (6)

PT
(10)

4.5 178 4.5

(7)

RI
type.E

(1)Steel plate 2202206 (2) Steel tube -2002006 (3) Round steel bar

SC
(4) Beam H-3001506.59 (5) Sliding stopper (6) High strength bolt

(7) Connecter H-200200812 (8) Split-T made of H-3001506.59

NU
(9) Rubber (10) Teflon sheet (11) Split-T made of plates
MA

(a) Cross-sectional view (b) 3D view of assembly diaphragm

Figure 1: Test specimen (unit:mm)
D

The manufacturing processes of the specimens are shown in Fig.2. The whole assemble
process had no welding. Only by piling up the steel tubes and assembly diaphragms
TE

alternatively, we were able to complete smoothly. The clearances between the steel tube and
the assembly diaphragm kept that the steel tube did not sustain the axial stress. The round
steel bars in Fig.1 (3) penetrated the column longitudinally. The steel bars only in EHN19D
P

were set in the ducts in the specimen. The other steel bars touched concrete directory. All
CE

bars were round and greased in which the bond stress was negligible. The tensile stress of the
bars penetrated through the column acted to realize the self-centering behaviors.
AC

Figure 2: Construction process of test

STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

Table 2: Material properties of steel

Yield Tensile Young's
strength strength modulus Yield Elongation
(Unit : mm) Standard
ratio (%)
MPa (MPa GPa
-2002006 STKR400 396 484 187 0.818 28.8
Flange 304 448 196 0.679 40.8
H-3001506.59
Web 356 484 202 0.737 38.4
Steel plate
- 247 407 192 0.609 38.8
Diaphragm t=6

PT
type.E Steel plate SS400
- 332 420 195 0.791 36.1
t=4.5
Split T
Diaphragm

RI
H-300150 Web 361 456 205 0.757 36.9
type.C
6.59

SC
Table 3: Properties of concrete
Strength of cylinder test Water/Cement ratio Fine aggregate ratio Slump Air
Specimen
(MPa) (%) (%) (cm) (%)
NU
EHN19D 39.2 42 51.7 21.5 5.9
CLS16N 49.9 42 49.1 22.5 4.6
MA

CLS19N
44.0 45 47.8 22.2 4.5
CLN14N
D

Table 4: Shearing force of beam corresponding to each failure mode

TE

Flexure of column Slip of bolt Tension of split T Flexure of beam

Specimen
(kN) (kN) (kN) (kN)
P

EHN19D 44 76
CE

CLS16N 48
114 162
CLS19N 117
43
CLN14N
AC

Table 2 and Table 3 list the material properties of steel and concrete, respectively. These
structural frames were designed to be failed at the column ends by bending moment. In Table
4 the calculated strength of each failure mode were compared each other. The calculated
values were shown as the shearing force of the beam. The flexural capacity of column was
calculated when the axial force ratio assumed to be equal to 0.2. Table 4 showed that the all
specimens attained the flexure capacities of the column first.
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

3. Loading apparatus and measurement

Loading apparatus for the test is shown in Fig.3. The top and bottom of the column set on pin
supports. Axial force was introduced by 1000 kN jack over the column. Also axial force was
added by pre-tension steel bars in the column. Table 5 shows the axial force N and axial force
ratio N/N0 at the beginning of loading. Notation N0 was axial strength calculated by
multiplying the sectional area and the strength of the concrete cylinder. Two 100 kN
hydraulic joined to the both ends of beams by pin supports were used to apply the inversely-
symmetric deformations of the specimens alternatively.
1000kN loadcell 1000kN jack

PT
RI
760

Pin

SC
Pin
760

Outer frame
100kN loadcell 100kN loadcell
Pin
NU
100kN jack 100kN jack

Pin Pin
MA

1000 1000

Figure 3: Loading apparatus

D

The aluminum frame for measuring deformations was set on pin support of end of column as
shown in Fig.4. The horizontal and vertical displacements were measured by four transducers.
TE

The drift angle Rb of the beam was obtained by the difference of transducer and divided
by the distance between and . The tensile force of each steel bar was measured by four
load-cells which were set at the end of each steel bar. The load-cell was doughnut-shaped
P

steel which both the diameter and height were 36mm. The steel bar was set through the
CE

15mm hole of the load-cell. The tensile force of steel bar was measured as the compressive
force by the load-cell.
AC

Table 5: Axial force

Total axial force (kN)
Test N0 (kN) Axial force ratio
Applied axial force Tension of steel bar
EHN19D 142 124 1385 0.19
CLS16N 253 24 1764 0.16
CLS19N 264 34 0.19
1555
CLN14N 148 72 0.14
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

2.5

2
1.5

Rb(1/100rad.)
1
0.5
0
-0.5
-1
-1.5
-2
-2.5
0 5 10 15
Transducer
Cycle

PT
Figure 4: Measurement apparatus Figure 5: Deformation histories

RI
The deformation histories of the test are shown in Fig.5 in which vertical axis shows drift
angle Rb of the beam of the specimen and horizontal axis shows number of loading cycles.
The peak drift angle stepwise increased from 0.25/100 rad. to 2.5/100rad.. At each drift, two

SC
cyclic deformations were applied statically.

4. Strength calculation method

NU
This structure was designed to be failed at the column by bending moment. The bending
strength of column Mn was calculated by the sum of bending moments of steel bars Mste and
concrete Mcon. Mste was calculated by the distance between the steel bars multiplying the
MA

difference of the tensile forces of steel bars. And Mcon was obtained by full plastic sate of the
section of the concrete column.
M n M ste M con (1)
D

Nm Nm
M con Dc
ccB b c
TE

2 (2)
Owing to the small measured value of Mste, the strength of column was only calculated by
Mcon in the paper. The axial force Nm on the specimens was equal to the applied vertical load
P

plus the tensile forces of the steel bars. The effective cross section of EHN19D, which had
CE

ducts in the column, was calculated by the concrete section minus the area of ducts. The
strength of confined concrete by the steel tube was obtained by the equation.
ccB Kc cB (3)
AC

Where, cB was the average strength of concrete by standard cylinder test, K c was the
enhanced coefficient of concrete strength by the effect of square steel tube and calculated by
the equation proposed by Y.Sun and K.Sakino [4].

sy 4t 2 D t
Kc 1.0 11.5
cB D 2t 3 (4)
D and t represented the depth and the wall-thickness of steel tube, respectively. sy was
equal to yield stress of steel tube.
 STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

5. The Relations between Loading Force and Deformation and Strength Evaluation
The relations between bending moment of the column Mn and drift angle Rb are shown in
Fig.6. The bending moment of the column was converted from the measured shearing forces
of the steel beams. The mark in the Fig.7 represents the calculated strength by Eq.(2) at
the peak of every cycles. When the bending moment was vanished, the residual deformations
were close to 0 rad. for all of the specimens. These results show that the efficient self-
centering capacity of the proposed beam-column frame exhibits by the loading tests.
Relation between the tensile force T of the steel bars and drift Rb angle are shown in Fig.7.
The vertical axis shows the average tension of the steel bars. In the figure, the broken and
solid lines showed the yield strength Ty and the tensile strength Tc of the steel bar,

PT
respectively. EHN19D showed that the steel bars did not yield through the loading test. With
the increasing of the drift angle, the axial force of column increased as reacting force of the
tensile force of the steel bars. As shown in Fig.6, the bending moment of the column was also

RI
increasing after Rb=1.5/100 rad. gradually due to the enhancement of the axial force of the
concrete section. The specimens in which used SS400 steel bars were yielding after the Rb

SC
attained by 1.5/100 rad.. After yielding of the steel bars, the strengths of specimens were not
increasing as shown in Fig.6.
50 50
100
NU
Exp. Exp. 50
Exp. Exp.
Cal. Cal.
Ty=165(kN) Ty
25 25 40 Tc
75
Mn(kNm)

Mn(kNm)

30
T(kN)

T(kN)
0 0 50
MA

20

-25 -25 25
10

-50 -50 0 0
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
Rb(1/100rad.) Rb(1/100rad.) Rb(1/100rad.) Rb(1/100rad.)
D
TE

(EHN19D) (CLS16N) (EHN19D) (CLS16N)

50 50 50 50
Exp. Exp.
Cal.
Exp. Exp.
Cal. Ty Ty
P

25 25 40 Tc 40 Tc
Mn(kNm)

Mn(kNm)
CE

30 30
T(kN)
T(kN)

0 0

20 20
-25 -25

10 10
-50 -50
AC

-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
0 0
Rb(1/100rad) Rb(1/100rad.) -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
Rb(1/100rad.) Rb(1/100rad.)

(CLS19N) (CLN14N) (CLS19N) (CLN14N)

Figure 6: Relations between Mn and Rb Figure 7: Relations between T and Rb
Comparisons of the bending moments of the tests and calculations are summarized in Table.6.
Calculated bending moments were obtained as the strength Mn by the equation (1), and the
axial force Nm was equal to the sum of applied axial force and the tensile force of steel bars at
Rb =1.5/100 rad.. The axial force ratio of all the tests were greater than the ratio in Table 5,
because the tensile force of the steel bars had increased at Rb =1.5/100 rad. than that at Rb =0
rad.. The calculated bending moment did not include the Mste,. The calculated strength
underestimated experimental strength from 5% to 11%.
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

Table 6: Comparisons between test and calculation

Exp. strength Cal. strength axial force ratio
Test Exp./Cal.
(kNm) (kNm) at Rb=1.5/100 rad.
EHN19D 29.0 26.1 0.26 1.11
CLS16N 27.5 26.0 0.18 1.06
CLS19N 30.4 28.9 0.24 1.05
CLN14N 25.3 23.6 0.19 1.07

6. Maximum Bending Moment and Self-centering Capacity

The relations between axial force Nm and bending moment Mn of the column are showed in

PT
Fig.8. Solid curves are obtained by the equation (2), and dotted curves and broken curves are
the 0.9 and 1.1 of the solid ones. The mark of represents the experimental strength at the
peaks of every loading cycles, and the mark of represents experimental strength at Rb

RI
=1.5/100 rad.. By the increasing of the axial force corresponding to the increasing of the drift
angle, the bending moment of the concrete section increased consequently.

SC
500 500
Cal. Cal.
1.5% 1.5%
400 400
NU
Nm(kN)
Nm(kN)

300 300

200 200

100
MA

100

0 0
0 10 20 30 40 50 0 10 20 30 40 50
Mn(kNm) Mn(kNm)

500
(EHN19D) (CLS16N)
500
D

Cal. Cal.
1.5% 1.5%
400 400
TE
Nm(kN)

Nm(kN)

300 300

200 200
P

100 100
CE

0 0
0 10 20 30 40 50
0 10 20 30 40 50
Mn(kNm) Mn(kNm)
(CLS19N) (CLN14N)
AC

Figure 8: Relations between Nm and Mn

In order to investigate the residual deformation, the relations of residual drift angle and
maximum drift angle at each cycle are shown in Fig.9. The vertical axis showed the residual
drift angle Rr at the lateral load was 0 kN. The horizontal axis shows the maximum drift angle
Rb0 at each cycle. Each specimen has the regression curve which is drawn in solid or broken
lines. The residual drift angle of EHN19D is larger than those of the others. The residual drift
angle of CLS16N, CLD19N and CLN14N are restrained under 5% of the maximum drift
angle of each cycle.
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

0.5 0.5
EHN19D CLS19N
CLS16N CLN14N
0.4 0.4

Rr(1/100rad.)

Rr(1/100rad.)
0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3
Rb0(1/100rad.) Rb0(1/100rad.)

PT
Figure 9: Relations between Rr and Rb0
EHN19D did not exhibit the apparent self-centering capacity, even though the steel bars
remained elastic. The main reason was that its ducts eliminated the dowel action of the steel

RI
bars. The other specimens did not have the ducts, so that the dowel effect of the steel bars
acted to aid the self-centering capacity.

SC
Fig.10 shows damage situation in the loading cycle of which peak deformation was Rb
=2.5/100 rad.. The left side picture (a) shows the peak deformation and the right side picture
shows the situation when the lateral load was 0. As shown in the left side picture, the
NU
horizontal clack was apparently observed, but the clack closed when the lateral load was 0 as
shown in the right side picture (b). During the loading test, the damage could be only
observed except for at the end of the column. Referring to the strain obtained from the gauges
MA

which pasted on the steel tubes and steel beams, the yield of them were not observed. This
showed that the structural frame had the continuous usage performance after the attacks by
large earthquakes due to its damage control capacity which included the self-centering
capacity and the reduction capacity against clacks. According to the original aim of the study,
D

this structure showed the advantages to keep the value of the building after the excitations by
earthquakes.
P TE
CE
AC

(a) Picture at peak deformation (b) Picture after unloading

Figure 10: Damage situations
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

7. Conclusions
The following conclusions were achieved through this study:
1) The relations between bending moment and story drift angle of all specimens showed the
self-centering behavior apparently. The self-centering capacity was affected by the dowel
action of the steel bars which penetrated the concrete section longitudinally.
2) Calculated strength by the full plastic state of the confined concrete section was slightly
underestimated the experimental strength from 5% to 11%.
3) Through the construction of the test specimens, the easy constructing procedure of this
structure was observed.

PT
4) From the appearance evaluations, the damages were only observed at the ends of and the
other parts had no damages. This structure provided the damage control capacity to

RI
realize the continuous use of the building attacked by large scale earthquakes.

SC
List of notations
bc inner breadth of the steel tube
NU
D depth of the steel tube
Dc inner depth of the steel tube
MA

Kc strength enhance ratio by confining effect

Mcon bending moment of concrete section
Mn bending moment of section
D

Mste bending moment of steel section

TE

Nm axial force of the section

Rb drift angle of beam of the specimen
P

Rbo maximum drift angle in the cyclic steps

CE

Rr residual drift angle of the specimen

T tensile force of the steel bars
AC

t thickness of steel tube

Tc tensile strength of the steel bars
Ty yield strength of the steel bars
cB strength of concrete
ccB strength of concrete confined by steel tube
sy yield stress of the steel tube
STRUCTURES-D-16-00038
ACCEPTED MANUSCRIPT
By Hiroyuki Nakahara and Hao Yin

Acknowledgements
The authors wish to express their thanks to Mr. H. Kubotera, technician of Kyushu University,
and Mr. T. Miyamoto, graduate student of Kyushu University, for their assistance in the
experiment and preparation of this paper.

References
[1] Pessiki, S.: Sustainable seismic design, Procedia Engineering ,171,2017,pp.33 39

[2] Ricles, J.M., Sause, R. Peng, S.W. and Lu L.W.: Experimental evaluation post-tensioned steel connections,
Journal of Structural Engineering, Vol.128, No.7, 2002, pp.850-859

PT
[3] Seo, C-Y. and Sause, R.: Ductility Demands on Self-Centering Systems under Earthquake Loading, ACI
Structural Journal, 102:2, 2005, pp.275-285.

RI
[4] Sun, Y. and Sakino, K.: Ductility Improvement of Reinforced Concrete Columns with High-Strength
Materials, Transactions of the Japan Concrete Institute (JCI Vol.15), 1993, pp.455-462

SC
[5] Yoshimura, K., Tomii, M., Sakino, K. and Tanaka, A.: Experimental Study on R/C Subassemblages to
Prevent a Short Column from Shear Failure by Using a Steel Square Tube, Proceedings of the 9th World
Conference on Earthquake Engineering, 1988, pp.737-742.
NU
MA
D
P TE
CE
AC