materials

Article
Seismic Behavior of Steel Column Base with
Slip-Friction Connections
Chengyu Li, Qi Liu and Gongwen Li *
School of Urban Construction, Wuhan University of Science and Technology, Wuhan 430065, China;
lichengyu@wust.edu.cn (C.L.); lq12345890@163.com (Q.L.)
* Correspondence: ligongwen@wust.edu.cn




Received: 9 August 2020; Accepted: 28 August 2020; Published: 9 September 2020


Abstract: Traditional rigid column base connections are damaged to different degrees after
an earthquake and the damage is generally nonrecoverable. Thus, the cost of repairing or dismantling
is quite high. A new type of slip-friction column base connection is proposed in this paper, which
aims to replace the yielding energy dissipation of the traditional column base connection by the
sliding friction energy dissipation between the arc endplates, thus achieving the design objective of
damage-free energy dissipation. Finite element simulation was conducted to study the hysteretic
performance of the proposed connections considering different axial compression ratios. The research
indicates that both kinds of the proposed connections show good energy dissipation behavior and the
increase of axial compression force can increase the energy dissipation ability. It also shows that the two
kinds of connections can achieve the objective of damage-free energy dissipation. For the proposed
connection, future research is still needed such as corresponding tests in the laboratory, the effect of
the connection on the whole structure, and the re-centering systems for the connections.

Keywords: column base; slip-friction connection; hysteretic behavior; damage-free energy dissipation

1. Introduction and Background

Steel structure has been widely used in high-frequency seismic regions because of its excellent
performances in previous earthquakes. However, in the 1994 Northridge Earthquake [1] in the
United States and the 1995 Hanshin Earthquake [2] in Japan, traditional rigid column base connections
have been damaged to varying degrees, and the cost of repairing or dismantling after an earthquake is
very high. Yang et al. [3], Nakashima et al. [4], and Borzouie [5] conducted a series of cyclic performance
tests of traditional exposed column base connections. The test results show that the yielding of column
base connections will produce residual deformation at the top of the column, and the axial shortening
due to the yielding of column base can cause undesirable effects in the building during earthquake
excitation [6]. Moreover, it is difficult for the building structure to recover to the initial vertical state
when yielding occurs. Therefore, it is very difficult to repair or remove the traditional column base
connections after the first floor is damaged. Akira et al. [7] pointed out that whether the damaged parts
are repaired or dismantled after an earthquake, it will cause great waste of resources and social labors.
Thus, to solve this problem, the concept of a low damage connection is proposed to dissipating
seismic energy by setting auxiliary energy-consuming components to keep the main structure from
damage. One of the most commonly used low damage connections is friction connection and it
has been studied by many scholars. The sliding hinge joint (SHJ) beam-column friction connection
which is known as asymmetric friction connection (AFC) was proposed by Clifton [8] for the first
time (Figure 1 [9]) and it has been further developed by many scholars [9–16]. This kind of AFC
beam-column connection generally uses friction damper to replace the traditional rigid structure
connections. Through pushover analysis, this kind of connection can control the structural damage,

Materials 2020, 13, 3986; doi:10.3390/ma13183986 www.mdpi.com/journal/materials

Materials 2020, 13, x FOR PEER REVIEW 2 of 23

Materials
Materials 2020,
13,13, x FOR PEER REVIEW 2 of
2223
beam-column 2020,connection
3986 generally uses friction damper to replace the traditional rigid 2structure of

connections. Through pushover analysis, this kind of connection can control the structural damage,
beam-column connection generally uses friction damper to replace the traditional rigid structure
while the
while residual
the residual
connections. lateral displacement
lateral
Through displacement
pushover still
analysis, exists
still this
exists after
after
kind alarge
largeearthquake.
of aconnection earthquake. ForFor
can control column
column
the basebase
structural friction
damage,friction
connections
connections
while the (Figure
(Figure22lateral
residual [17]),Borzouie
[17]), Borzouie et al.et
displacement [17] al. [17] after
conducted
still exists conducted
experimental experimental
studies onFor
a large earthquake. thestudies on the
cyclic performance
column base cyclic
friction
performance
columnofbase
ofconnections columnstrongbase
(Figure 2and strong
weak
[17]), andaligned
axis
Borzouie weak
et al. axis
[17] aligned
asymmetric
conducted asymmetric
friction friction
connection,
experimental andconnection,
studies iton
was thefound and it
cyclic
was that
found the proposed
that the
performance details
proposed
of column for
basebase connections
details
strong and forweak with
base friction
alignedconnections
connections
axis provide
with friction
asymmetric repeatable
connection,cycles
frictionconnections provide
and it
without
was found
repeatable strength
cyclesthat degradation
the proposed
without or any
strengthdetailsrequirements
degradation for repair
for baseorconnections or replacement
any requirements following
with friction an
connections
for repair earthquake.
provide
or replacement
following an earthquake. Freddi et al. [18] and Dimopoulos et al. [19] also proposed a new steel
Freddi et
repeatableal. [18]
cycles and Dimopoulos
without et
strength al. [19] also
degradation proposed
or any a new type
requirements of a sway
for damage-free
repair or replacement
type of a
sway damage-free steel column base connection and the results of finite element simulationM-θ
column
followingbasean connection
earthquake. andFreddi
the results
et al. of
[18] finite
and element
Dimopoulos simulation
et al. verified
[19] also the accuracy
proposed a of the
new type of a
verified
curve.
swayThis kind of connection
damage-free steel column canbase
effectively
connection avoid yielding
and of theofcolumn
the results base, thus,
finite element achieving
simulation the
verified
the accuracy of the M-θ curve. This kind of connection can effectively avoid yielding of the column
objective of damage-free
the accuracy of the M-θenergycurve.dissipation.
This kind ofBorzouieconnection et al.
can[20] also conducted
effectively cyclic tests
avoid yielding on Grade
of the column
base, thus, achieving the objective of damage-free energy dissipation. Borzouie et al. [20] also
10.9 bolts
base, in AFC.
thus, The test
achieving results
the showed
objective that Grade 10.9
of damage-free bolts had
energy less ductility
dissipation. compared
Borzouie et al.to[20]
Grade also
conducted cyclic
8.8conducted
bolts, andcyclictests
Gradetests on Grade
10.9 on
bolts 10.9 bolts
fractured
Grade
in
for large
10.9 bolts
AFC. The
sliding
in AFC.
test results
Thedisplacements.
showed
Thus, that
test results showed
that
Grade Grade
8.8 bolts
Grade
10.9
10.9are
bolts
used
bolts had
had
less ductility
inless compared
thisductility to Grade 8.8 bolts, and Grade 10.9
paper. compared to Grade 8.8 bolts, and Grade 10.9 bolts fractured for large sliding bolts fractured for large sliding
displacements.
displacements. Thus, Grade
Thus, 8.8 8.8
Grade bolts areare
bolts usedused ininthis
thispaper.
paper.

Figure 1. Beam-column asymmetric friction connection (AFC) [9].

Figure 1. Beam-column asymmetric friction connection (AFC) [9].
Figure 1. Beam-column asymmetric friction connection (AFC) [9].

Figure 2. Column base friction connection (adapted from [17]).

Figure 2. Column base friction connection (adapted from [17]).
On the basis of previous studies, a new kind of steel column base with slip-friction arc endplates is
proposed
On(astheshown
basisinofFigure 3). For
previous this kind
studies, a of connection,
new kind of ifsteel
the moment
columnunderbase the horizontal
with load is
slip-friction arc
less than theisslip
endplates moment,
proposed
Figure the
2.(as connection
shown
Column does3).
in Figure
base not
friction rotate
For thisand
connection kindit of
behaves
(adapted like[17]).
connection,
from theif ordinary
the moment column
underbase
the
connection.
horizontalHowever,
load is lesswhen
thanthe moment
the exceedsthe
slip moment, theconnection
slip moment, thenot
does connection
rotate and rotates. Underlike
it behaves large
the
earthquakes,
ordinary the
column connection
base dissipates
connection. seismic
However, energy
when through
the the
moment rotational
exceeds
On the basis of previous studies, a new kind of steel column base with slip-friction arc friction
the between
slip the
moment, arc
the
connection
endplates rotates.(as
is proposed Under
shown largeinearthquakes,
Figure 3). For the this
connection
kind ofdissipates
connection, seismic energy
if the momentthrough
underthe the
horizontal load is less than the slip moment, the connection does not rotate and it behaves like the
ordinary column base connection. However, when the moment exceeds the slip moment, the
connection rotates. Under large earthquakes, the connection dissipates seismic energy through the
are welded together, the upper arc endplate and the lower arc endplate are connected by
high-strength bolts, and the pre-pressure between the upper and lower arc endplates is produced by
applying a pretension force to high-strength bolts. The upper arc endplate is provided with oblong
bolt holes so that the connection can rotate along the oblong holes, and the lower arc endplate is
provided with13,ordinary
Materials 2020, 3986 circular bolt holes. Under horizontal external force, when the moment 3 of 22
reaches the slip moment, the upper steel column and the upper arc endplate will rotate, and the
relative slip between the upper and lower arc endplates will cause friction. The maximum rotation
endplates and the friction between the arc endplates and bolts. The sliding friction energy dissipation
angle of the upper steel column and the upper arc endplate can be controlled by the size of the
is used to replace the energy dissipation caused by material yielding of the traditional connection.
oblong bolt hole opened on the upper arc endplate. The connection can be divided into a concave
Compared to the AFC introduced above (Figures 1 and 2), the axial compression force transferred from
connection (Figure 3a) and a convex connection (Figure 3b) according to the orientation of arc
the column may increase the friction force between the arc endplates. Thus, the proposed connection
endplates.
may have better energy dissipation behavior.

(a)

(b)
Figure
Figure Steel column
3. 3.Steel column base
basewith slip-friction
with arc endplates:
slip-friction (a) concave
arc endplates: (a) connection; (b) convex connection.
concave connection; (b) convex
connection.
As is shown in Figure 3, the connection is composed of steel column, upper and lower arc
endplates, high-strength bolts, and base plate. Among them, the steel column and the arc endplates
are welded together, the upper arc endplate and the lower arc endplate are connected by high-strength
bolts, and the pre-pressure between the upper and lower arc endplates is produced by applying a
pretension force to high-strength bolts. The upper arc endplate is provided with oblong bolt holes
so that the connection can rotate along the oblong holes, and the lower arc endplate is provided
with ordinary circular bolt holes. Under horizontal external force, when the moment reaches the slip
moment, the upper steel column and the upper arc endplate will rotate, and the relative slip between
the upper and lower arc endplates will cause friction. The maximum rotation angle of the upper steel
column and the upper arc endplate can be controlled by the size of the oblong bolt hole opened on the
upper arc endplate. The connection can be divided into a concave connection (Figure 3a) and a convex
connection (Figure 3b) according to the orientation of arc endplates.
Materials 2020, 13, x FOR PEER REVIEW 4 of 23

Materials 2020, 13, 3986

2. Objectives 4 of 22
Materials 2020, 13, x FOR PEER REVIEW 4 of 23

The main purpose of this paper is to study the seismic behavior of a steel column with the
2.2.Objectives
Objectives
proposed slip-friction connections to check the ability of damage-free energy dissipation. As is
introducedTheThe main purpose
main finite
above, purpose ofofthisthis
element paper
paper is to is
analysis to study
study
(FEA)the is athecommonly
seismic seismic
behaviorbehavior
ofused of a steel
a steelmethod
column to column
with withseismic
the proposed
study the the
slip-friction
proposed connections
slip-friction to check
connections the ability
to of
check damage-free
the ability energy
of dissipation.
damage-free
behavior of steel column base connections. Thus, cyclic FEA simulation of the proposed connection As is
energy introduced above,
dissipation. As is
finite element
introduced analysis
above, (FEA)
finite is a commonly
element analysis used(FEA)method to study theused
is a commonly seismic behavior
method of steelthe
to study column
seismic
will be conducted to study the energy dissipation and seismic behavior. Furthermore, the seismic
base connections.
behavior Thus, cyclic
of steel column baseFEA simulationThus,
connections. of thecyclic
proposed
FEAconnection
simulationwill be conducted
of the proposed to study
connection
behavior
theof concave
energy and convex
dissipation and seismic connections will be compared
behavior. Furthermore, in this
the seismic paper.
behavior of concave and convex
will be conducted to study the energy dissipation and seismic behavior. Furthermore, the seismic
connections
behavior will be compared
of concave in this
and convex paper. will be compared in this paper.
connections
3. Finite Element Analysis
3. Finite Element Analysis
3. Finite Element Analysis
3.1. Size of Specimen
3.1. Size of Specimen
3.1. Size of Specimen
The section of theofspecimen
The section the specimen usedusedininFEA
FEA is is H 300mm
H 300 mm× ×300 300mmmm × 10 × 10mmmm × 15×mm, 15 mm,
which which
meansmeans
the section The section
is H-shaped
the section of the
is H-shaped specimen
(Figure
(Figure4)4)and used in FEA
the
and the height
height is (H)
H(H)
300 mm
is
is 300 300× 300
mm, themm
mm, the
width × 10 ofmm
width ×of 15the
the flange mm, which
(B)flange
is means
(B)
300 mm, is 300
the
mm, the section
thethickness
thicknessis of
H-shaped
ofthe
the web
web (t(Figure
w )(tis 4)10
w)10ismm, and
mm,
and the height
and
the the(H)
thickness is the
thickness
of 300flange
mm, the
of the
(t width
f ) isflange
15 mm. of fthe
(tThe flange
) istotal
15 mm. (B)The
length is
of 300
total
mm,
the the thickness
specimen is of
1700 the
mm. webAs (t
is ) is
shown
w 10 mm,
in and
Figure the
5, the thickness
inner of
radius
length of the specimen is 1700 mm. As is shown in Figure 5, the inner radius (R) of the upper arc the
(R) flange
of the (t f) is 15
upper arc mm. The
endplates total
lengthis of
is 0.7 H,
endplates the
0.7the specimen
H,width of theisarc
the width 1700
of mm.
endplates
the As(W)is shown
arc endplates (W)inisFigure
is 500 mm, 5, the
the thickness
500 mm, theinner
of theradius
thickness (R)the
arc endplates
of of arc
the upper
is 20 mm, arc
endplates is
endplates
the lengthisof 0.7oblong
H, thebolt
widthholes of the arc endplates
(Figure 5a) in the (W) upper is 500 mm, the thickness
arc endplate of the arc
is 42 mm (Figure 6),endplates
and the is
20 mm, the length of oblong bolt holes (Figure 5a) in the upper arc endplate is 42 mm (Figure 6), and
diameter
20 mm, theoflength
circularofbolt holesbolt
oblong (Figure
holes5b) in the 5a)
(Figure lower arc endplate
in the upper arcisendplate
22 mm. is 42 mm (Figure 6), and
the diameter of circular bolt holes (Figure 5b) in the lower arc endplate is 22 mm.
the diameter of circular bolt holes (Figure 5b) in the lower arc endplate is 22 mm.

Figure 4. Size of H-shaped section.

Figure 4.4.Size
Figure Sizeof
ofH-shaped section.
H-shaped section.

(a) (b)
Figure 5. Holes in arc endplates: (a) upper arc endplate; (b) lower arc endplate.
(a)5. Holes in arc endplates: (a) upper arc endplate; (b) lower(b)
Figure arc endplate.

Figure 5. Holes in arc endplates: (a) upper arc endplate; (b) lower arc endplate.
 Materials 2020, 13, 3986 5 of 22
Materials 2020, 13,
Materials x FOR
2020, PEER
13, x FOR REVIEW
PEER REVIEW 5 of 23 5 of 23

Figure 6. Size of oblong bolt holes (unit: mm).

Figure 6. Size of oblong bolt holes (unit: mm).
Figure 6. Size of oblong bolt holes (unit: mm).
3.2. Material Model
3.2. Material Model
3.2. MaterialThe steel column and arc endplates are made of Q345B steel, which means the yield strength of
Model
The steel column and arc endplates are made of Q345B steel, which means the yield strength of
the steel is 345 Mpa. The bolts are 8.8 grade M20 high-strength bolts, which means the tensile
the steel is 345 Mpa. The bolts are 8.8 grade M20 high-strength bolts, which means the tensile strength
The steel iscolumn
strength 800 MPa, andthearc endplates
yield strength are
ratiomade
is 0.8,of Q345B
and steel, which
the nominal diameter means
of thethe
boltyield strength of
is 20 mm.
is 800 MPa, the yield strength ratio is 0.8, and the nominal diameter of the bolt is 20 mm. Finite element
the steel is element
Finite 345 Mpa. The involves
analysis bolts are the 8.8 grade
plastic M20 high-strength
development bolts,
of materials and the which
influence
analysis involves the plastic development of materials and the influence of material nonlinearity and
means the tensile
of material
nonlinearity
strength and geometric
is 800 nonlinearity.
MPa, the yield nonlinearity. The
strength ratioparameter corresponding
is 0.8, and parameter
the nominal settings
diameter in the
of12.0, ANSYS
the Swanson
bolt is 20 mm.
geometric The corresponding settings in the ANSYS (Version
(Version
element12.0,
Finite Analysis Swanson
analysis Analysis
involves theSystems
plasticInc., Houston, TX, of
development USA) [21] analysis
materials and are
theshown in Table
influence of 1.
material
Systems Inc., Houston, TX, USA) [21] analysis are shown in Table 1. The stress–strain
The stress–strain relationship of the steel material is shown in Figure 7. The von Mises yield criterion
nonlinearity andofgeometric
relationship nonlinearity.
the steel material is shownThe corresponding
in Figure parameter
7. The von Mises settingsis in
yield criterion the ANSYS
selected,
is selected, the Poisson's ratio is set as 0.3, and the strengthening model MKIN (multilinear kinematic
the 12.0,
(Version Poisson’s ratio is Analysis
Swanson set as 0.3, and the strengthening
Systems Inc., Houston,modelTX,MKIN
USA) (multilinear kinematic
[21] analysis hardening)
are shown in Table 1.
hardening) is adopted in the analysis model.
is adopted in the analysis model.
The stress–strain relationship of the steel material is shown in Figure 7. The von Mises yield criterion
is selected, the Poisson's ratio is setTable
as 0.3,
Table and the
1. Material
1. Material strengthening
properties
properties model MKIN (multilinear kinematic
of the steel.
of the steel.
hardening) is adopted in the analysis
Elastic Modulusmodel. Yield Strength Tensile Strength Poisson’s Ratio
SteelSteel
TypeType Elastic Modulus Yield Strength Tensile Strength Poisson’s Ratio
E/GPaE/GPa fy/MPa
fy /MPa fufu/MPa
/MPa υ υ
High-strength Table 1. Material properties of the steel.
High-strength bolt 206 206 640
640 800
800 0.3 0.3
bolt
Other steel 206 345 560 0.3Poisson’s
Elastic Modulus Yield Strength Tensile Strength Ratio
Steel Other
Type steel 206 345 560 0.3
E/GPa fy/MPa fu/MPa υ
High-strength
206 640 800 0.3
bolt
Other steel 206 345 560 0.3

Figure 7. Stress–strain
Figure 7. Stress–strain relationship
relationship of
of the
the steel
steel material.
material.
3.3. Finite Element Model
3.3. Finite Element Model
The FEA model before meshing is shown in Figure 8. All the components here, such as high-strength
The FEA model before meshing is shown in Figure 8. All the components here, such as
bolts and arc endplates, are built by solid elements Solid185. The contact elements CONTA174 and
high-strength bolts and arc endplates, are built by solid elements Solid185. The contact elements
TARGE170 are used to simulate the contact between the two arc endplates and between arc endplates
CONTA174 and TARGE170 are used to simulate the contact between the two arc endplates and
and bolts. The pre-tension element PRETS179 is used on of
thethe
topsteel
areamaterial.
of the high-strength bolts to
between arc endplatesFigure 7. Stress–strain
and bolts. relationship
The pre-tension element PRETS179 is used on the top area of the
apply the pre-tension force. The element meshing of the arc endplates and bolts is shown in Figure 9.
high-strength bolts to apply the pre-tension force. The element meshing of the arc endplates and
3.3. Finite
boltsElement
is shownModel
in Figure 9.

The FEA model before meshing is shown in Figure 8. All the components here, such as
high-strength bolts and arc endplates, are built by solid elements Solid185. The contact elements
CONTA174 and TARGE170 are used to simulate the contact between the two arc endplates and
between arc endplates and bolts. The pre-tension element PRETS179 is used on the top area of the
high-strength bolts to apply the pre-tension force. The element meshing of the arc endplates and
Materials
Materials 2020,
2020, 13,
13, x3986
FOR PEER REVIEW 66 of
of 23
22
Materials 2020, 13, x FOR PEER REVIEW 6 of 23

(a) (b)
(a) (b)

The
8.8.The8.
FigureFigure finite
finite element
element analysis
analysis (FEA) (FEA)
model model
of of connections:
connections: (a) concave connection;
The finite element analysis (FEA) model of connections:(a)
(a)concave
concaveconnection;
connection; (b) convex
(b) convex
(b) convex
connection. connection.
connection.

(a) (b)
(a) (b)

(c)
(c)
Figure 9. Element meshing of arc endplates and bolts: (a) upper arc endplate; (b) lower arc endplate;
Figure(c)9.high-strength
Element bolts. of
Element meshing
meshing of arc
arc endplates
endplates and
and bolts:
bolts: (a)
(a) upper
upper arc
arc endplate;
endplate; (b)
(b) lower
lower arc
arc endplate;
endplate;
(c) high-strength bolts.
3.4. Constraints and Loading
3.4. Constraints
3.4. Constraints and and Loading
Loading
The friction model used here is the “extended Lagrange” friction formula, and the friction
coefficient
The friction is model
set as 0.4.usedBinding
here constraints
is are usedLagrange”
to simulatefriction
the welding effect, such theasfriction
upper
The friction model used here is the
the “extended
“extended Lagrange” friction formula,
formula, and
and the friction
steel
coefficient iscolumn and upper arc endplate, lower steel column and lower arc endplate, as well as screw
coefficient is set
setas as0.4.
0.4.Binding
Bindingconstraints
constraints areareused
usedto to
simulate
simulate thethewelding
welding effect, suchsuch
effect, as upper steel
as upper
column and nut.
and upper arc endplate, lower steel column and lower arc endplate, as well as screw and nut.
steel column and upper arc endplate, lower steel column and lower arc endplate, as well as screw
The The fixed
fixed constraint
constraint UX =UXUY = UY
= UZ= UZ= = RX
RX = = RY
RY = =RZRZ==00 is is applied
applied to to the
the bottom
bottom of
ofthe
the base
base plate.
plate.
and nut.
In order
In order to to apply
apply the vertical
the vertical load and horizontal loadonon thetoptopofofthetheupper
upper steel
steel column,
column, aa rigid
The
coverfixed
plateconstraint
is set as the UX =load
UY =and
loading UZhorizontal
plate =onRXthe= top
load
RY of
= RZ 0 the
the =column
is applied
in the to the bottom
model. The lengthof the
and base rigid
plate.
width of
cover
In orderplate is set
to applyplate as the
the are loading
vertical plate
loadasand on the top of the column in the model. The length and width of
the loading the same thehorizontal
upper steelload column,on the
andtop the of the upper
thickness is 10steel
mm.column,
The materiala rigid
of
the loading
cover plate plate
is set arethe
as the same asplate the upper thesteel column, and the thickness isThe
10 mm. The material of
the cover plate is setloading
as ideal linearon elastic,top
andofthe
theelastic
column in the
modulus ismodel.
10 3 times
3 times
length
the elasticand widthof
modulus of
the cover
the loading plate
plate is set as ideal linear elastic, and the elastic modulus is 10
are the same as the upper steel column, and the thickness is 10 mm. The material of the elastic modulus of
the steel column.
the steel column.
the cover plate
After is theset as ideal of
modeling linear elastic, and
the specimen, thethe elastic
loads are modulus
applied inissteps. 103 times
Firstly,thethe
elastic modulus
pretension forceof
After
the steel the
column.
of 125 modeling of the specimen, the loads are applied in steps.
kN is applied to the high-strength bolts. Then, an axial vertical load (shown in Table 2) is Firstly, the pretension force
of 125 kN the
After
applied is to
applied
the toptoofthe
modeling ofthe high-strength
the specimen,
upper bolts.
the
steel column, loads Then,
and are an
theappliedaxialinaxial
applied vertical
steps.load load
Firstly, (shown in Table
the pretension
is converted into 2) is
force
a surface
applied to the top of the upper steel column, and the applied axial load
of 125 kN is applied to the high-strength bolts. Then, an axial vertical load (shown in Table 2) is is converted into a surface load,
which isto
applied applied
the toptoofthe theloading
upper plate on the top
steel column, of the
and the column.
applied axial In Tableload2,isn converted
is the axialinto compression
a surface
Materials 2020, 13, 3986 7 of 22
Materials 2020, 13, x FOR PEER REVIEW 7 of 23
Materials 2020, 13, x FOR PEER REVIEW 7 of 23

load,which
load, whichisisapplied
appliedtotothe
theloading
loadingplateplateononthe
thetop
top ofthethecolumn.
column.InInTable
Table 2,nnisisthe
theaxial
axial
ratio. The cyclic horizontal displacement is applied to the oftop of the column and the2,loading history
compression
compression ratio. The
ratio. 10, cyclic horizontal
cyclic ∆horizontal
Thewhere displacement
displacement is applied to the top of the column and the
is shown in Figure y is the displacement of is applied
the top of to
thethe top ofcorresponding
column the column and to the
loading history
loading angle
history is shown in Figure 10, where Δy is the displacement of the top of the column
rotation of is shown
0.006 in Figure 10, where Δy is the displacement of the top of the column
radian.
corresponding to the rotation
corresponding to the rotation angle angleofof0.006
0.006radian.
radian.
Table 2. Value of the axial vertical load.
Table2.2.Value
Table Valueofofthe
theaxial
axialvertical
verticalload.
load.
n 0.1 0.2 0.3
n 0.1
0.1 0.2
0.2 0.3
0.3
Verticaln load (kN) 409 817 1227
Verticalload
Vertical load(kN)
(kN) 409
409 817
817 1227
1227

Figure10.
Figure 10.Loading
Loadinghistory
historyofofhorizontal
horizontaldisplacement.
displacement.

4.4.Results and
Resultsand Discussion
andDiscussion
Discussion
4. Results
4.1. Rotation
4.1.Rotation Center
RotationCenter
Center
4.1.
The
Therotation
rotationvector
vectordiagram
diagramof the
theconcave
concaveconnection
connectionis shown
shownin Figure
Figure11.11.The
Theideal
idealrotation
center The
is rotation
the centervector
of diagram
the arc ofofthe
endplate, concave
while connection
it can be isisshown
observed ininFigure
that during 11. The
the ideal
rotation
rotation
rotation
ofofthe
center is the
center is thethe center
center of the
of the arc endplate,
arc rotation
endplate, while
while it
it cancanbebe observed
observed that
that during
during the rotation
the rotation of the
the
connection,
connection, theposition
position ofofthe
the rotation center
center does
does not
notcoincide
coincide with
with thetheideal
idealrotation
rotation center
centerand
and itit
connection,
drifts the position of the rotation center does not coincide with the ideal rotation center and it
driftswith
drifts with
with
thethe
the
rotation
rotation
rotation
of the
of
steelsteel
ofthe
the column.
steel column.
It can
column. It
also
Itcan be
canalso observed
also beobserved
be
that the
observed rotation
that
that
center moves
therotation
the rotation closer
centermoves
center moves
and closer
closer andto the ideal
closer rotation center when the vertical axial load is increased.
closer and closer totothe
the idealrotation
ideal rotation center
center when
when thevertical
the vertical axial
axial loadisisincreased.
load increased.

(a)
(a) (b)
(b)

(c)
(c) (d)
(d)
Figure 11. Rotation vector diagram of concave connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.
Figure11.
Figure 11.Rotation
Rotationvector
vectordiagram
diagramofofconcave
concaveconnections:
connections:(a)
(a)n n= =0;0;(b)
(b)n n= =0.1;
0.1;(c)(c)n n= =0.2;
0.2;(d)
(d)n n= =0.3.
0.3.
Materials 2020, 13, 3986 8 of 22

Materials 2020, 13, x FOR PEER REVIEW 8 of 23

Figure 12 shows the rotation vector diagram of the convex connection. It can be observed that
Figure
there is 12 shows
no drifting the
of the rotation
rotation vector
center diagram
during of the convex
the rotation connection.
of the convex It can and
connection, be observed that
the position
ofthere is no drifting
the rotation of consistent
center is the rotation center
with during
the ideal the rotation
rotation of all
center for thethe
convex connection,
specimens, which and
meansthe
position of the rotation center is consistent with the ideal rotation
that the vertical load has no effect on the position of the rotation center. center for all the specimens, which
means that the vertical load has no effect on the position of the rotation center.

(a) (b)

(c) (d)

Figure 12. Rotation vector diagram of convex connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.
Figure 12. Rotation vector diagram of convex connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.
4.2. M-θ Curve and Energy Dissipation Behavior
4.2. M-θ Curve and Energy Dissipation Behavior
The M-θ curve for the concave connection is shown in Figure 13, where M is the moment at the
rotation The M-θ and
center curve
θ for therotation
is the concaveangle
connection is shown inItFigure
of the connection. can be13, where M
observed is the
that moment
during at the
the initial
rotation
stage center
of the and θthe
loading, is the
M-θrotation
curve angle of the
basically connection.
linearly It canand
increases, be observed that during
the connection the initial
is in elasticity,
while when the slip moment is reached, the connection starts to rotate, and a slipping segmentelasticity,
stage of the loading, the M-θ curve basically linearly increases, and the connection is in appears.
while when the slip moment is reached, the connection starts to rotate,
It is shown in Figure 13 that the M-θ curve for each circle is a parallelogram, which means that and a slipping segment
it has
a good energy dissipation behavior. It can also be observed that after the second circle, the peakmeans
appears. It is shown in Figure 13 that the M-θ curve for each circle is a parallelogram, which load
that it has a good energy dissipation behavior. It can also be observed that after
corresponding to the positive radian of each circle decreases step by step, and the degree decreases as the second circle, the
peak
the load corresponding
vertical load increases. to the positive
Moreover, radian
as the of each
vertical load circle decreases
increases, the peakstepload
by step,
of theand the degree
second circle
decreases as the vertical load increases. Moreover,
increases gradually and the wave-like jitter decreases. as the vertical load increases, the peak load of the
second circle increases gradually and the wave-like jitter decreases.
Materials 2020, 13, 3986 9 of 22
Materials 2020, 13, x FOR PEER REVIEW 9 of 23

(a) (b)

(c) (d)

Figure 13. M-θ curves of concave connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.
Figure 13. M-θ curves of concave connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.

The M-θ curve of the convex connection is shown in Figure 14. It can also be observed that
The M-θ curve of the convex connection is shown in Figure 14. It can also be observed that the
the shape of the hysteresis loop is close to a parallelogram, which reflects the large slip deformation
shape of the hysteresis loop is close to a parallelogram, which reflects the large slip deformation
characteristics of the convex connection. As the vertical load increases, the curve shows a downward
characteristics of the convex connection. As the vertical load increases, the curve shows a downward
trend similar to the concave connection, while the curve of the convex connection is smoother and the
trend similar to the concave connection, while the curve of the convex connection is smoother and
curves after the second circle nearly overlap, especially for the specimen under a high vertical load.
the curves after the second circle nearly overlap, especially for the specimen under a high vertical
It can be seen from the M-θ curves that the hysteresis loop under the cyclic load is plump, and the
load. It can be seen from the M-θ curves that the hysteresis loop under the cyclic load is plump, and
large hysteresis loop area indicates that the convex connection has good energy dissipation ability.
the large hysteresis loop area indicates that the convex connection has good energy dissipation
The hysteresis loop of the convex connection also shows a “small wave” jittering, which may be
ability.
related to the slight deformation of the upper arc endplate. As the vertical load increases, the M-θ
The hysteresis loop of the convex connection also shows a “small wave” jittering, which may be
curves become smoother.
related to the slight deformation of the upper arc endplate. As the vertical load increases, the M-θ
Figure 15 shows the outline of typical M-θ curves of column base AFC in Borzouie’s research [17].
curves become smoother.
Comparing the curves in Figure 15 with those in Figures 13 and 14, it can be concluded that the
Figure 15 shows the outline of typical M-θ curves of column base AFC in Borzouie’s research
hysteresis loops of the proposed connections are plumper, especially when n = 0.2, which means a
[17]. Comparing the curves in Figure 15 with those in Figures 13 and 14, it can be concluded that the
better energy dissipation ability. This is because the axial compression force transferred from the
hysteresis loops of the proposed connections are plumper, especially when n = 0.2, which means a
column may increase the friction force between the arc endplates for the proposed connections.
better energy dissipation ability. This is because the axial compression force transferred from the
column may increase the friction force between the arc endplates for the proposed connections.
It can also be observed in Figures 13 and 14 that when θ reaches 1/50, which is the limited
inter-story displacement angle under large earthquakes in the Chinese code [22], the bearing
capacity and energy dissipation ability of the column remains steady, which means that the
proposed connections have good seismic behavior under large earthquakes, and the limitation of a
Materials 2020, 13, x FOR PEER REVIEW 10 of 23
Materials 2020, 13, x FOR PEER REVIEW 10 of 23
displacement of structure with the proposed connections could be designed according to the existing
Materials 2020, 13, 3986 10 of 22
displacement
code [22]. of structure with the proposed connections could be designed according to the existing
code [22].

(a) (b)
(a) (b)

(c) (d)
(c) (d)
Figure 14. M-θ curves of convex connections: (a) n = 0; (b) n = 0.1; (c) n = 0.2; (d) n = 0.3.
Figure M-θ
14.14.
Figure curves
M-θ of of
curves convex connections:
convex connections: n=
(a)(a) (b)n n= =0.1;
n =0;0;(b) (c)nn==0.2;
0.1;(c) (d)nn==0.3.
0.2;(d) 0.3.
300
200 Borzouie's Research
Borzouie's Research
Convex Connection 300 Convex Connection
200
150 Borzouie's Research
Borzouie's Research 200
Convex Connection Convex Connection
150
100
200
100
100
50
M (kN.m)
M (kN.m)

100
50
0 0
M (kN.m)
M (kN.m)

-500 0
-100
-50
-100
-100
-100
-150 -200

-150
-200 -200
-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 -300
-200 (rad) -0.05 -0.04 -0.02 0.00 0.02 0.04 0.05
-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 -300 (rad)
(rad) -0.05 -0.04 -0.02 0.00 0.02 0.04 0.05
(rad)

(a) (b)
(a) (b)
Figure 15.M-θ
Figure15. M-θcurves
curvesof
ofbase
basebending
bendingabout
aboutthe
thestrong
strongaxis: (a)nn==0;0;(b)
axis:(a) (b)nn== 0.2
0.2 (adapted
(adaptedfrom
from[17]).
[17]).
Figure 15. M-θ curves of base bending about the strong axis: (a) n = 0; (b) n = 0.2 (adapted from [17]).
It can also be observed in Figures 13 and 14 that when θ reaches 1/50, which is the limited inter-story
To study the effect of vertical load on the energy dissipation ability of the connections, the
displacement angle under large earthquakes in the Chinese code [22], the bearing capacity and energy
To study
equivalent the effect
viscous of vertical
damping ratio isload on the energy
calculated dissipation
for convex abilityThe
connections. of the connections,
equivalent the
viscous
dissipation ability of the column remains steady, which means that the proposed connections have
equivalent viscous damping ratio is calculated
damping ratio he is calculated by the following formula:for convex connections. The equivalent viscous
good seismic behavior under large earthquakes, and the limitation of a displacement of structure with
damping ratio he is calculated by the following formula:
the proposed connections could be designed according to the existing code [22].
 Materials 2020, 13, 3986 11 of 22

To study the effect of vertical load on the energy dissipation ability of the connections, the
equivalent
Materials 2020, 13, xviscous damping
FOR PEER REVIEWratio is calculated for convex connections. The equivalent viscous
11 of 23
damping ratio he is calculated by the following formula:

1 1S SABCOA
he h=e  ABCOA , , (1) (1)
2π2S∆BOD
SBOD
where
where SABCOA
SABCOA andandSBODSBOD
are are shown
shown in in Figure16.
Figure 16.The calculated hhe eisislisted
The calculated in in
listed Table 3. It3.isItshown
Table in in
is shown
Table 3 that he increases
Table 3 that he increases as the axial compression ratio increases, which indicates that the
as the axial compression ratio increases, which indicates that the increase of increase of
axial force may increase the energy dissipation ability of the connections.
axial force may increase the energy dissipation ability of the connections.

Figure 16. Definition of equivalent viscous damping ratio.

Figure 16. Definition of equivalent viscous damping ratio.
Table 3. The equivalent viscous damping ratio he .
Table 3. The equivalent viscous damping ratio he.
n he
n 0 0.55 he
0 0.1 0.63 0.55
0.2 0.72
0.1 0.3 0.81 0.63
0.2 0.72
4.3. Deformation of0.3
the Arc Endplates 0.81

It can be observed in the FEA that there is a small opening between the upper arc endplate and
4.3. Deformation of the Arc Endplates
lower arc endplate (Figure 17). It can be observed from Figure 17 that the deformation at the center
along
It can the widthindirection
be observed the FEAofthatthethere
arc plate
is a is the largest.
small openingThe maximum
between deformation
the upper ∆s1max of
arc endplate andthelower
upper arc plate under a different axial compression ratio is listed in Table 4. It is shown
arc endplate (Figure 17). It can be observed from Figure 17 that the deformation at the center along in Table 4 that
∆s1max direction
the width decreases as
ofthe
theaxial compression
arc plate ratio increases
is the largest. and the maximum
The maximum value of
deformation ∆s1max
Δs1max of/W is upper
the 1/647. arc
plateItunder
can be observed from the FEA that the arc endplates are under an elastic state during the whole
a different axial compression ratio is listed in Table 4. It is shown in Table 4 that Δs1max
loading process.
decreases as the axial compression ratio increases and the maximum value of Δs1max/W is 1/647. It can
be observed from the FEA
Table that the of
4. ∆s1max arcupper
endplates are under
arc endplate for bothan elastic
kinds state during the whole loading
of connections.
process.
n ∆s1max ∆s1max /W
0 0.773 1/647
0.1 0.684 1/731
0.2 0.541 1/924
0.3 0.497 1/1006
 Materials 2020, 13, x FOR PEER REVIEW 12 of 23
Materials 2020, 13, 3986 12 of 22

(a)

(b)

Figure 17. Deformation of the arc endplates: (a) concave connections; (b) convex connections.
Figure 17. Deformation of the arc endplates: (a) concave connections; (b) convex connections.
4.4. Working State of Bolts
Table 4. Δs1max of upper arc endplate for both kinds of connections.
The deformation and stress state of high-strength bolts during the loading process are investigated
to check the n working state. The rowΔof bolts are numbered as 1, 2, 3, and 4 Δ
s1max s1max/W
along the direction of the
0 0.773 1/647
horizontal load from the left to the right. Figures 18–21 show the deformation along the horizontal
direction 0.1
(x direction) and the von Mises 0.684stress of each row of bolts when the loading
1/731 step reaches 2∆y ,
−2∆y , 3∆0.2
y , −3∆ y , 6∆ y , −6∆ y , and 0.541
the end of the last load step. 1/924
0.3 0.497 1/1006
It is shown in Figures 18 and 20 that the deformation of the bolts 1 and bolts 2 are nearly symmetric
with bolts 4 and bolts 3, respectively, and the deformation generally decreases as the axial compression
4.4. Working
ratio State
increases. It isofshown
Bolts in Figure 19 that the maximum von Mises stress of the bolts for concave
connections during the loading
The deformation process
and stress is 625.66
state Mpa, which is
of high-strength smaller
bolts thanthe
during the yield stress
loading with 640
process are
MPa,
investigated to check the working state. The row of bolts are numbered as 1, 2, 3, and 4 along for
and this means that all the bolts are under elastic state during the loading process. However, the
convex connections,
direction it is shown
of the horizontal loadin Figure
from the 21
leftthat theright.
to the maximum
Figuresvon Mises
18–21 stress
show theofdeformation
some of the bolts
along
exceeds the yield stress, which means that some of the bolts are under plastic state during
the horizontal direction (x direction) and the von Mises stress of each row of bolts when the loading the loading
process. Thus,2△
step reaches some of the bolts need to be replaced after large earthquakes for convex connections. It
y, −2△y, 3△y, −3△y, 6△y, −6△y, and the end of the last load step.
can also
It isbeshown
observed from Figures
in Figures 18 and 1920and 21 that
that the axial compression
the deformation ratio
of the bolts has little
1 and bolts effect
2 areon the
nearly
stress development of both kinds of connections.
symmetric with bolts 4 and bolts 3, respectively, and the deformation generally decreases as the axial
compression ratio increases. It is shown in Figure 19 that the maximum von Mises stress of the bolts
for concave connections during the loading process is 625.66 Mpa, which is smaller than the yield
stress with 640 MPa, and this means that all the bolts are under elastic state during the loading
process. However, for convex connections, it is shown in Figure 21 that the maximum von Mises
stress of some of the bolts exceeds the yield stress, which means that some of the bolts are under
plastic state during the loading process. Thus, some of the bolts need to be replaced after large
earthquakes for convex connections. It can also be observed from Figures 19 and 21 that the axial
compression ratio has little effect on the stress development of both kinds of connections.
Materials 2020, 13, 3986 13 of 22
Materials 2020, 13, x FOR PEER REVIEW 13 of 23

(a) (b)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(c)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(d)

(e) (f)

Figure 18. Cont.

Materials 2020, 13, 3986 14 of 22
Materials 2020, 13, x FOR PEER REVIEW 14 of 23

2△y −2△y 3△y −3△y 6△y −6△y LAST

(g)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(h)

(i) (j)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(k)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(l)

Figure 18. Cont.

Materials 2020, 13, 3986 15 of 22
Materials 2020, 13, x FOR PEER REVIEW 15 of 23

(m) (n)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(o)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(p)

Figure18.
Figure 18. Deformation
Deformation of ofthe
thebolts
boltsforfor
concave
concave connections:
connections: (a) n(a) = 0, 1;bolts
= 0,nbolts (b) n1;= 0.2,
(b) n = 0.2,
bolts 1; (c) n = 1;
bolts
= 0, development
(c)0,ndevelopment of deformation for bolts
of deformation for1;bolts
(d) n 1;
= 0.2,
(d) n = 0.2, development
development of deformation for bolts 1;
of deformation for(e)bolts
n = 1;
= 0,2;bolts
(e)0, nbolts (f) n =2;0.2, n = 0.2,
(f) bolts 2; (g) (g) n = 0, development
n = 0,2;development
bolts of deformation of for bolts 2; (h)for
deformation n = bolts 2; (h) n = 0.2,
0.2, development
development of deformation for bolts 2; (i) n = 0, bolts 3; (j) n = 0.2, bolts 3; (k) n = 0,ofdevelopment
of deformation for bolts 2; (i) n = 0, bolts 3; (j) n = 0.2, bolts 3; (k) n = 0, development deformation of
for bolts 3; (l)
deformation fornbolts
= 0.2,3;development of deformation
(l) n = 0.2, development for bolts 3; (m)
of deformation fornbolts
= 0, bolts
3; (m)4;n(n) n =bolts
= 0, 0.2, bolts
4; (n)4;n (o)
= 0.2,
n = 0,4;development
bolts of deformation
(o) n = 0, development for bolts 4; (p)for
of deformation n =bolts
0.2, development
4; (p) n = 0.2,ofdevelopment
deformation for of bolts 4.
deformation
for bolts 4.
Materials 2020, 13, 3986 16 of 22
Materials 2020, 13, x FOR PEER REVIEW 16 of 23

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 1 stress for bolts 1 stress for bolts 1 stress for bolts 1

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 2 stress for bolts 2 stress for bolts 2 stress for bolts 2

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 3 stress for bolts 3 stress for bolts 3 stress for bolts 3

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 4 stress for bolts 4 stress for bolts 4 stress for bolts 4

Figure 19.The
Figure19. Thevon
vonMises
Mises stress
stress of the bolts
bolts for
for concave
concaveconnections.
connections.
Materials 2020,
Materials 13,x3986
2020, 13, FOR PEER REVIEW 1723of 22
17 of

(a) (b)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(c)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(d)

(e) (f)
Figure 20. Cont.
Materials 2020, 13, 3986 18 of 22
Materials 2020, 13, x FOR PEER REVIEW 18 of 23

2△y −2△y 3△y −3△y 6△y −6△y LAST

(g)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(h)

(i) (j)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(k)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(l)

Figure 20. Cont.

Materials 2020, 13, 3986 19 of 22
Materials 2020, 13, x FOR PEER REVIEW 19 of 23

(m) (n)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(o)

2△y −2△y 3△y −3△y 6△y −6△y LAST

(p)

Figure
Figure 20. 20. Deformation
Deformation of of
thethe bolts
bolts for
for convexconnections:
convex connections: (a) n=
(a) n = 0,
0, bolts (b)nn==0.2,
bolts1;1;(b) 0.2,bolts 1; 1;
bolts n =n = 0,
(c)(c)
0, development
development of deformation
of deformation for bolts
for bolts 1; (d)
1; (d) n =n 0.2,
= 0.2,development
developmentof of deformation
deformation for forbolts
bolts1;1;(e)(e)
n =n = 0,
bolts0,2;
bolts 2; (f) n = 0.2, bolts 2; (g) n = 0, development of deformation for bolts 2; (h) n = 0.2, development
(f) n = 0.2, bolts 2; (g) n = 0, development of deformation for bolts 2; (h) n = 0.2, development of
of deformation for bolts 2; (i) n = 0, bolts 3; (j) n = 0.2, bolts 3; (k) n = 0, development of deformation
deformation for bolts 2; (i) n = 0, bolts 3; (j) n = 0.2, bolts 3; (k) n = 0, development of deformation
for bolts 3; (l) n = 0.2, development of deformation for bolts 3; (m) n = 0, bolts 4; (n) n = 0.2, bolts 4; (o)
for bolts 3; (l) n = 0.2, development of deformation for bolts 3; (m) n = 0, bolts 4; (n) n = 0.2, bolts 4;
n = 0, development of deformation for bolts 4; (p) n = 0.2, development of deformation for bolts 4.
(o) n = 0, development of deformation for bolts 4; (p) n = 0.2, development of deformation for bolts 4.
Materials 2020, 13, 3986 20 of 22
Materials 2020, 13, x FOR PEER REVIEW 20 of 23

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 1 stress for bolts 1 stress for bolts 1 stress for bolts 1

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 2 stress for bolts 2 stress for bolts 2 stress for bolts 2

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 3 stress for bolts 3 stress for bolts 3 stress for bolts 3

n = 0, n = 0.1, n = 0.2, n = 0.3,

stress for bolts 4 stress for bolts 4 stress for bolts 4 stress for bolts 4

Figure
Figure 21.21.The
Thevon
vonMises
Misesstress
stress of the
the bolts
boltsfor
forconvex
convexconnections.
connections.

5. Conclusions
5. Conclusions
Refined FEA models are built to investigate the seismic behavior of the proposed two kinds of
slip-friction connections with different axial compression ratios, and the conclusions can be summarized
as follows:
Materials 2020, 13, 3986 21 of 22

(1) Both kinds of connections show good energy dissipation behavior, and the increase of axial
compression force can increase the energy dissipation ability of the connections;
(2) The M-θ curves of the convex connection are smoother and steadier than those of the
concave connection;
(3) The energy dissipation ability of the proposed connections is better than that of the traditional
AFC, especially when the axial compression force is applied to the column;
(4) All the components of the connections except for some of the bolts in the convex connection
are under elastic state during the whole loading process, which means that the two kinds of
connections can achieve the objective of damage-free energy dissipation.

6. Recommendations for Future Research

It is worth noting that all the conclusions here are based on the FEA results while there are no
corresponding tests. The authors have designed the specimens and the tests will be conducted in the
laboratory to validate the FEA results in the future.
Furthermore, the research is mainly focused on the seismic behavior of the proposed column base
connection while further researches should be conducted to study the effect of the connection on the
whole structure, such as the effects of displacements on the bending moment and shear in the rest of
the structure.
In addition, external force is needed to push the column with the proposed connections back to
the initial vertical state, which is not convenient. Therefore, re-centering systems should be developed
for this kind of connection to help the structure back to the initial state after earthquakes.

Author Contributions: Conceptualization, C.L.; methodology, Q.L.; formal analysis, Q.L.; writing—original
draft preparation, Q.L.; writing—review and editing, G.L.; supervision, C.L. and G.L.; funding acquisition, C.L.
All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China, grant number 51878522.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
AFC asymmetric friction connection
FEA finite element analysis
SHJ sliding hinge joint
B width of the flange
E elastic modulus
fy yield strength
fu tensile strength
H height of the H-shaped section
he equivalent viscous damping ratio
M moment at the rotation center
n axial compression ratio
R inner radius of the arc endplates
tf thickness of the flange
tw thickness of the web
W width of the arc endplates
υ Poisson’s ratio
θ rotation angle of the connection
displacement of the top of the column corresponding to a rotation angle of 0.006 radian∆s1max :
∆y
maximum deformation of the upper arc endplate
Materials 2020, 13, 3986 22 of 22

References
1. Miller, D.K. Lessons learned from the northridge earthquake. Eng. Struct. 1998, 20, 249–260. [CrossRef]
2. Mahin, S.A. Lessons from damage to steel buildings during the northridge earthquake. Eng. Struct. 1998,
20, 261–270. [CrossRef]
3. Yang, N.; Zhong, Y.; Meng, Q.; Zhang, H. Hysteretic behaviors of cold-formed steel beam-columns with
hollow rectangular section: Experimental and numerical simulations. Thin Walled Struct. 2014, 80, 217–230.
[CrossRef]
4. Nakashima, M.; Liu, D. Instability and complete failure of steel columns subjected to cyclic loading.
J. Eng. Mech. 2005, 131, 559–567. [CrossRef]
5. Borzouie, J. Low Damage Steel Base Connections. Ph.D. Thesis, University of Canterbury, Christchurch,
New Zealand, 2016.
6. Macrae, G.; Urmson, C.R.; Walpole, W.R.; Moss, P.; Hyde, K.; Clifton, C. Axial shortening of steel columns in
buildings subjected to earthquakes. Bull. N. Z. Soc. Earthq. Eng. 2009, 42, 275–287. [CrossRef]
7. Akira, W.; Mamoru, I.; Keizo, S.; Shigetaka, A.; Hiroki, K. Design of Damage Control of Building Structures;
Qu, Z.; Pei, X.Z., Translators; China Architecture Press: Beijing, China, 2014; pp. 1–60. (In Chinese)
8. Clifton, C.G. Semi-Rigid Joints for Moment Resisting Steel Framed Seismic Resisting Systems. Ph.D. Thesis,
University of Auckland, Auckland, New Zealand, 2005.
9. Macrae, G.; Clifton, G.C.; Mackinven, H.; Mago, N.; Butterworth, J.; Pampanin, S. The sliding hinge joint
moment connection. Bull. New Zealand Soc. Earthq. Eng. 2010, 43, 202–212. [CrossRef]
10. Golondrino, J.C.; MacRae, G.; Chase, J.; Rodgers, G.W.; Clifton, G.C. Velocity effects on the behavior of
asymmetrical friction connections (AFC). In Proceedings of the 8th STESSA Conference, Shanghai, China,
1–3 July 2015.
11. Khoo, H.-H.; Clifton, C.; Butterworth, J.; Macrae, G.; Ferguson, G.; Clifton, G. Influence of steel shim hardness
on the sliding hinge joint performance. J. Constr. Steel Res. 2012, 72, 119–129. [CrossRef]
12. Latour, M.; Piluso, V.; Rizzano, G. Experimental analysis of innovative dissipative bolted double split tee
beam-to-column connections. Steel Constr. 2011, 4, 53–64. [CrossRef]
13. Latour, M.; Piluso, V.; Rizzano, G. Experimental analysis on friction materials for supplemental damping
devices. Constr. Build. Mater. 2014, 65, 159–176. [CrossRef]
14. Latour, M.; D’Aniello, M.; Zimbru, M.; Rizzano, G.; Piluso, V.; Landolfo, R. Removable friction dampers for
low-damage steel beam-to-column joints. Soil Dyn. Earthq. Eng. 2018, 115, 66–81. [CrossRef]
15. Latour, M.; Rizzano, G.; Santiago, A.; da Silva, L.S. Experimental response of a low-yielding, self-centering,
rocking column base joint with friction dampers. Soil Dyn. Earthq. Eng. 2019, 116, 580–592. [CrossRef]
16. Piluso, V.; Montuori, R.; Troisi, M. Innovative structural details in MR-frames for free from damage structures.
Mech. Res. Commun. 2014, 58, 146–156. [CrossRef]
17. Borzouie, J.; Macrae, G.A.; Chase, J.G.; Rodgers, G.W.; Clifton, G.C. Experimental studies on cyclic
performance of column base strong axis aligned asymmetric friction connections. J. Struct. Eng. 2016, 142,
04015078. [CrossRef]
18. Freddi, F.; Dimopoulos, C.A.; Karavasilis, T.L. Rocking damage-free steel column base with friction devices:
Design procedure and numerical evaluation. Earthq. Eng. Struct. Dyn. 2017, 46, 1–19. [CrossRef]
19. Dimopoulos, C.A.; Freddi, F.; Karavasilis, T.L. Rocking Damage-Free Steel Column Base with Friction Devices:
Development of Advanced 3D Finite Element Models in Abaqus; EUROSTEEL: Copenhagen, Denmark, 2017.
20. Borzouie, J.; Macrae, G.A.; Chase, J.G.; Rodgers, G.W. Cyclic performance of asymmetric friction connections
with grade 10.9 bolts. Bridge Struct. Eng. 2015, 45, 53–62.
21. SAS. Version 12.0, ANSYS; Swanson Analysis Systems Inc.: Houston, TX, USA, 2011.
22. GB 50011-2010. Code for Seismic Design of Building; China Architecture & Building Press: Beijing, China, 2010.
(In Chinese)

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).