Practical Advanced Analysis For Semi-Rigid Space Frames: Seung-Eock Kim, Se-Hyu Choi

International Journal of Solids and Structures 38 (2001) 9111±9131
www.elsevier.com/locate/ijsolstr
Practical advanced analysis for semi-rigid space frames

Seung-Eock Kim a,*, Se-Hyu Choi b
a
Department of Civil and Environmental Engineering, Sejong University, 98 Koonja-dong, Kwangjin-ku, Seoul 143-747, South Korea
b
Construction Technology Research Institute, Sejong University, 98 Koonja-dong, Kwangjin-ku, Seoul 143-747, South Korea
Received 15 November 2000; in revised form 15 May 2001
Abstract
A practical advanced analysis of semi-rigid space frame is developed. Herein, the nonlinear behavior of beam-
to-column connections is discussed, and practical modeling of these connections is introduced. The proposed analysis
can predict accurately the combined nonlinear eects of connection, geometry, and material on the behavior and
strength of semi-rigid frames. Kishi±Chen power model is used to describe the nonlinear behavior of semi-rigid con-
nections. Stability functions are used to capture second-order eects associated with P-d and P-D eects. The column
research council tangent modulus and a parabolic function for gradual yielding are used to represent material non-
linearity. The load±displacements predicted by the proposed analysis compare well with those available experiments. A
case study has been performed for a four story semi-rigid frame. Ó 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Advanced analysis; Semi-rigid connection; Geometric nonlinearity; Material nonlinearity; Steel frame
1. Introduction
Conventional analysis of steel frame structures is usually carried out under the assumption that the
beam-to-column connections are either fully rigid or ideally pinned. However, most connections used in
current practice are semi-rigid type whose behavior lies between these two extreme cases. In the AISC-
LRFD Speci®cation (AISC, 1993), two types of constructions are designated: type fully restrained con-
struction; and type partially restrained construction. The LRFD Speci®cation permits the evaluation of the
¯exibility of connections by rational means when the ¯exibility of connections is accounted for in the
analysis and design of frames.
The semi-rigid connections in¯uence the moment distribution in beams and columns as well as the drift
(P-D eect) of the frame. One way to account for all these eects in semi-rigid frame design is through the
use of a direct second-order inelastic frame analysis known as `àdvanced analysis''. Advanced analysis
indicates a method that can suciently capture the limit state strength and stability of a structural system
and its individual members so that separate member capacity checks are not required. Since the power of
*
Corresponding author. Tel.: +82-2-3408-3391; fax: +82-2-3408-3332.
E-mail address: sekim@sejong.ac.kr (S.-E. Kim).
0020-7683/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 0 - 7 6 8 3 ( 0 1 ) 0 0 1 4 1 - X
9112 S.-E. Kim, S.-H. Choi / International Journal of Solids and Structures 38 (2001) 9111±9131
personal computers and engineering workstations is rapidly increasing, it is feasible to employ advanced
analysis techniques directly in engineering design oce.
During the past 20 years, research eorts have been devoted to the development and validation of several
advanced analysis methods. The advanced analysis methods may be classi®ed into two categories: (1)
plastic-zone method; and (2) plastic hinge method. Whereas the plastic-zone solution is known as the
`èxact solution'', but cannot be used for practical design purposes (Chen and Toma, 1994). This is because
the method is too intensive in computation and costly due to its complexity.
Advanced analyses for two-dimensional steel frame with rigid and semi-rigid connection were developed
by Lui and Chen (1986), Al-Mashary and Chen (1991), Kishi and Chen (1986, 1990), Liew (1992), Kim and
Chen (1996a,b), and Barsan and Chiorean (1999). Second-order inelastic analyses for the space steel frames
with rigid connections were developed by Orbison (1982), Prakash and Powell (1993), and Liew and Tang
(1998). Orbison's method is an elastic±plastic hinge analysis without considering shear deformations. The
material nonlinearity is considered by the tangent modulus Et and the geometric nonlinearity is by a
geometric stiness matrix. Orbison's method, however, underestimates the yielding strength up to 7% in
stocky members subjected to axial force only. DRAIN-3DX developed by Prakash and Powell (1993) is a
modi®ed version of plastic hinge methods. The material nonlinearity is considered by the stress±strain
relationship of the ®bers in a section. The geometric nonlinearity caused by axial force is considered by the
use of the geometric stiness matrix, but the nonlinearity caused by the interaction between the axial force
and the bending moment is not considered. This method overestimates the strength and stiness of the
member subjected to signi®cant axial force. Liew and Tang's method is a re®ned plastic hinge analysis. The
eect of residual stresses is taken into account in conventional beam-column ®nite element modeling.
Nonlinear material behavior is taken into account by calibration of inelastic parameters describing the yield
and bounding surfaces. Liew and Tang's method, however, underestimates the yielding strength up to 7% in
stocky member subjected to axial force only. The analysis for the space steel frames with semi-rigid con-
nections was developed by Shakourzadeh et al. (1999). The work proposed by Shakourzadeh et al. was on
modeling of connections in the analysis of thin-walled space frames.
We shall develop a practical advanced analysis using plastic-hinge concept for semi-rigid space frames.
This paper combines nonlinear behavior of framed members and that of semi-rigid connection. The
analysis is equivalent to the plastic-zone analysis in its accuracy but is much simpler than the plastic-zone
analysis. Lateral torsional buckling of members is assumed to be prevented by adequate lateral braces. Also
a compact W-section is assumed so that the section can develop full plastic moment capacity without local
buckling.
2. Behavior of semi-rigid frames
The important attributes that aect the behavior of semi-rigid steel frame structures may be grouped
into three categories: connection, geometric, and material nonlinearities. The connection nonlinearity in-
dicates the nonlinear moment±rotation relationship of semi-rigid connections. The geometric nonlinearity
includes second-order eects associated with the P-d and P-D eects and geometric imperfections. The
material nonlinearity includes gradual yielding associated with the in¯uence of residual stresses on ¯exure
behavior.
2.1. Nonlinear behavior of connections
The forces transmitted through beam-column connections consist of axial force, shearing force, bending
moment, and torsion. The eect of axial force, shearing force, and torsion is negligible since their defor-
mations are small compared with the rotational deformation of connections. The deformation behavior of
S.-E. Kim, S.-H. Choi / International Journal of Solids and Structures 38 (2001) 9111±9131 9113
Fig. 1. Schematic moment±rotation curves of various semi-rigid connections.
a connection may be customarily described by moment±rotation relationship, and its typical behavior is
nonlinear. The schematic moment±rotation curves of commonly used semi-rigid connections are shown in
Fig. 1. It may be observed that a relatively ¯exible connection has a smaller ultimate moment capacity and
a larger rotation, and vice versa. Herein, Kishi±Chen power model shall be adopted to describe the mo-
ment±rotation relationship of semi-rigid connections (Kishi and Chen, 1990).
If the direction of incremental moment applied to a connection is reversed, the connection will unload
with the initial slope of the moment±rotation curve. This loading and unloading behaviors of connections
can be adequately accounted for by the use of tangent stiness and initial stiness, respectively (Chen and
Lui, 1991). Herein, these stinesses shall be obtained by simply dierentiating Kishi±Chen power model
equation.
2.2. Geometric nonlinearity
The bending moments in a beam-column consist of two types: primary bending moment; and secondary
bending moment. Primary bending moments are caused by applied end moments and/or transverse loads
on members. Secondary bending moments are from axial compressive force acting through the lateral
displacements of a member. The secondary bending moments include the P-d and P-D moments. Herein,
stability functions are used for each member to capture these second-order eects in a direct manner.
2.3. Material nonlinearity
Residual stresses result in a gradual axial stiness degradation. The ®bers that have the highest com-
pressive residual stress will yield ®rst under compressive force, followed by the ®bers with a lower value of
compressive residual stress. Due to this spread of yielding or plasticity, the axial and bending stinesses of a
column segment are degraded gradually along the length of a member. This stiness degradation eect will
be accounted for later by the tangent modulus concept (Liew, 1992).
When a wide ¯ange section is subjected to pure bending, the moment-curvature relationship of a section
has a smooth transition from elastic to fully plastic. This is because the section yields gradually from ex-
treme ®bers which have higher stresses than interior ®bers. The gradual yielding eect leads to the concept
of a hardening plastic hinge which may be represented simply by a parabolic stiness reduction function of
a plastic hinge (Liew, 1992). This will be described later.
3. Practical connection modeling
The connection behavior is represented by its moment±rotation relationship. Extensive experimental

works on connections have been performed, and a large body of moment±rotation data has been collected
(Goverdhan, 1983; Nethercot, 1985; Kishi and Chen, 1986; Chen and Kishi, 1989). Using these abundant
database, researchers have developed several connection models including: linear; polynomial; B-spline;
power; and exponential models. Herein, the three-parameter power model proposed by Kishi and Chen
(1990) is adopted. The Kishi±Chen power model contains three parameters: initial connection stiness Rki ,
ultimate connection moment capacity Mu , and shape parameter n. The power model may be written as (Fig.
2):
h
m 1=n
for h > 0; m > 0 1
1 hn
where m M=Mu , h hr =h0 , h0 reference plastic rotation, Mu =Rki , Mu ultimate moment capacity of the
connection, Rki initial connection stiness, and n shape parameter. When the connection is loaded, the
connection tangent stiness (Rkt ) at an arbitrary rotation hr can be derived by simply dierentiating Eq. (1)
as:
dM Mu
Rkt 2
djhr j h0 1 hn 11=n
Fig. 2. Moment±rotation behavior of three-parameter model.

When the connection is unloaded, the tangent stiness is equal to the initial stiness as:
dM Mu
Rkt Rki 3
djhr j h0
It is observed that a small value of the power index n makes a smooth transition curve from the initial
stiness Rkt to the ultimate moment Mu . On the contrary, a large value of the index n makes the transition
more abruptly. In the extreme case, when n is in®nity, the curve becomes a bilinear line consisting of the
initial stiness Rki and the ultimate moment capacity Mu .
An important task for practical use of the power model is to determine the three parameters for a given
connection con®guration. Herein, the practical procedures for determining the three parameters are pre-
sented for the following four types of connections with angels: single/double web-angle connections; and
top and seat angle with/without double web-angle connections.
The values of Rki and Mu can be determined by a simple mechanical procedure with an assumed failure
mechanism (Kishi and Chen, 1990). For single/double web-angle connections shown in Fig. 3, the initial
connection stiness and the ultimate moment capacity are given by:
ta3 a coshab
Rki G 4
3 ab coshab sinhab
2Vpu V0 2
Mu da 5
6
where G shear moduli, ta thickness of web-angle, a 4:2962 when Poisson's ratio is 0.3, b g1 =da ,
da height of web-angle, g1 distance from the ®xed support line to free edge line as shown Fig. 3(a),
Vpu minimum value of Vpy , V0 maximum value of Vpy , and Vpy plastic shear force per unit length.
For the top and seat angle connections shown in Fig. 4, the initial connection stiness and the ultimate
moment capacity are given by:
3EI d12
Rki 6
1 0:78tt2 =g12 g13
Mu M0s Mp Vp d2 7
where EI bending stiness of angle's leg adjacent to column face, g1 gt D=2 tt =2, gt gage
distance from top angle's heel to center of fastener holes in leg adjacent to column face, D db for rivet
fastener, D W for bolt fastener, db the diameter of the fastener, W the diameter of the nut, tt
thickness of top angle, d1 distance between centers of horizontal legs of top and bottom angles
( d tt =2 ts =2), ts thickness of the bottom angle, d total depth of the beam section, M0s plastic
moment capacity at point C, Mp plastic moment capacity at point H2 of top angle, Vp shear force,
d2 d ts =2 k, and k distance from the top angle's heel to the toe of the ®llet.
For top and seat angle connections with double web-angles shown in Fig. 5, the initial connection
stiness and the ultimate moment capacity are given by:
M 3EIt d12 6EIa d32

Rki 8
hr g1 g12 0:78tt2 g3 g32 0:78ta2
Mu M0s Mpt Vpt d2 2Vpa d4 9

Fig. 3. Web-angle connection. (a) Moderate thick plate modeling and (b) mechanism at ultimate condition.
where EIt , EIa bending stiness of legs adjacent to column face of top angle and web-angle, g3
gc W =2 ta =2, W diameter of nut, ta thickness of top angle, Mpt ultimate moment capacity of top
angle, Vpt shearing
force acting on plastic hinges, Vpa resulting plastic shear force, d4 2Vpu
V0a da =3 Vpu V0a ll ts =2, Vpu shearing force at upper edge of web-angle, V0a shearing force at
lower edge of web-angle, and ll distance from bottom edge of web-angle to compression ¯ange of beam.
As for the shape parameter n, the equations developed by Kishi et al. (1991) are implemented here. Using
a statistical technique for n values, empirical equations of n are determined as a linear function of log10 h0
shown in Table 1.
Fig. 4. Top and seat angle connection. (a) De¯ected con®guration at elastic condition and (b) mechanism at ultimate condition.
4. Practical advanced analysis
4.1. Geometric second-order eects
Stability functions are used to capture the second-order eects since they can account for the eect of the
axial force on the bending stiness reduction of a member. The bene®t of using stability functions is that it
enables only one or two elements to predict àccurately the second-order eect of each framed member
(Kim and Chen, 1996a,b).
Fig. 5. Top and seat angle with web-angle connection. (a) De¯ected con®guration at elastic condition and (b) applied forces in ultimate
state of connection.
The force±displacement equation using stability functions may be written for three-dimensional beam-
column element as
8 9 2 EA 0 0 0 0 0
38 9
> P > L > d >
>
> > 6 > >
> MyA >
> >
> 6 0
EI
S1 Ly
EI
S2 Ly 0 0 0 7>
7>>
>
hyA >
>
>
< >
= 6 EI EI 7<> >
=
MyB 60 S2 Ly S1 Ly 0 0 0 7 hyB
6 7 10
>
> MzA >
> 60 0 0 S3 EILz S4 EILz 0 7> hzA >
>
> >
> 6 7>> hzB >
> >
>
> > 40 5> >
: MzB >
> ; 0 0 S4 EILz S3 EILz 0 >
: >
;
T 0 0 0 0 0 GJ /
L
Table 1
Empirical equations for shape parameter n (Kishi and Chen, 1990)
Connection type N
Single web-angle connection 0:520 log10 h0 2:291 for log10 h0 > 3:073
0:695 for log10 h0 < 3:073
Double web-angle connection 1:322 log10 h0 3:952 for log10 h0 > 2:582
0:573 for log10 h0 < 2:582
Top and seat angle connection 2:003 log10 h0 6:070 for log10 h0 > 2:880
0:302 for log10 h0 < 2:880
Top and seat angle connection with double web-angle 1:398 log10 h0 4:631 for log10 h0 > 2:721
0:827 for log10 h0 < 2:721
where P , MyA , MyB , MzA , MzB , and, T are axial force, end moments with respect to y and z axes and torsion
respectively. d, hyA , hyB , hzA , hzB , and, / are the axial displacement, the joint rotations, and the angle of twist.
S1 , S2 , S3 , and S4 are the stability functions with respect to y and z axes, respectively.
The stability functions given by Eq. (10) may be written as
8 p p p
>
> p qy sinp qy p2 q cosp qy
>
< 2 2 cosp q p q sinpp
p p if P < 0
y y qy
S1 p p p 11a
>
> p2 qy coshp qy p qy sinhp qy
>
: p p p if P > 0
2 2 coshp qy p qy sinhp qy
8 p p
>
> p2 qy p qy sinp qy
>
< p p p if P < 0
2 2 cosp qy p qy sinp qy
S2 p p 11b
>
> p qy sinhp qy p2 qy
>
: p p p if P > 0
2 2 coshp qy p qy sinp qy
8 p p p
> p qz sinp qz p2 q cosp qz
>
< p
p
p if P < 0
2 2 cosp qz p qz sinp qz
S3 p
p
p
11c
>
> p2 q coshp qz p qz sinhp qz
: p p p if P > 0
2 2 coshp qz p qz sinhp qz
8 p p
> p2 qz p qz sinp qz
>
< p p p if P < 0
2 2 cosp qz p qz sinp qz
S4 p p 11d
>
> p qz sinhp qz p2 qz
: p p p if P > 0
2 2 coshp qz p qz sinp qz
where qy P =p2 EIy =L2 , qz P =p2 EIz =L2 , and P is positive in tension.
4.2. Column research council tangent modulus model associated with residual stresses
The CRC tangent modulus concept is used to account for gradual yielding (due to residual stresses)
along the length of axially loaded members between plastic hinges. The elastic modulus E (instead of
moment of inertia I) is reduced to account for the reduction of the elastic portion of the cross-section since
the reduction of the elastic modulus is easier to implement than a new moment of inertia for every dierent
section. From Chen and Lui (1991), the CRC Et is written as
Et 1:0E for P 6 0:5Py 12a

P P
Et 4 E 1 for P > 0:5Py 12b
Py Py
4.3. Parabolic function for gradual yielding due to ¯exure
The tangent modulus model is suitable for the member subjected to axial force, but not adequate for
cases of both axial force and bending moment. A gradual stiness degradation model for a plastic hinge is
required to represent the partial plasti®cation eects associated with bending. We shall introduce the
parabolic function to represent the transition from elastic to zero stiness associated with a developing
hinge. When the parabolic function for a gradual yielding is active at both ends of an element, the slope-
de¯ection equation may be expressed as
8 9 2 Et A 38 9
>
> P >
> L
0 0 0 0 0 > > d > >
>
> > 6 > >
>
> MyA >
>
> 6 0 kiiy kijy 0 0 07 7
>
>
> hyA >>
>
< = 6 7< =
MyB 0 k k 0 0 0 h
66
ijy jjy 7
7
yB
13
>
> MzA >
> 6 0 0 0 kiiz kijz 0 7> > hzA >>
>
> > > >
>
> M > > 4 0 0 0 kijz kjjz 0 5> > hzB >
> >
: zB >; GJ :
>
;
T 0 0 0 0 0 L
/
where

S22 Et Iy
kiiy gA S1 1 gB 14a
S1 L
Et I y
kijy gA gB S2 14b
L

S22 Et I y
kjjy gB S1 1 gA 14c
S1 L

S42 Et I z
kiiz gA S3 1 gB 14d
S3 L
Et Iz
kijz gA gB S4 14e
L

S42 Et Iz
kjjz gB S3 1 gA 14f
S3 L
The terms gA and gB is a scalar parameter that allows for gradual inelastic stiness reduction of the
element associated with plasti®cation at end A and B. This term is equal to 1.0 when the element is elastic,
and zero when a plastic hinge is formed. The parameter g is assumed to vary according to the parabolic
function:
g 1:0 for a 6 0:5 15a
g 4a1 a for a > 0:5 15b

where a is a force-state parameter that measures the magnitude of axial force and bending moment at the
element end. The term (a) may be expressed by AISC-LRFD and Orbison, respectively:
Fig. 6. Full plasti®cation surface of AISC-LRFD.
AISC-LRFD
Based AISC-LRFD bilinear interaction equation (Kanchanalai, 1977), the cross-section plastic strength
of the beam-column member may be expressed (Fig. 6)
P 8 My 8 Mz P 2 My 2 Mz
a for P 16a
Py 9 Myp 9 Mzp Py 9 Myp 9 Mzp
P My Mz P 2 My 2 Mz
a for < 16b
2Py Myp Mzp Py 9 Myp 9 Mzp
Orbison
Orbison's full plasti®cation surface (Fig. 7) (Orbison, 1982) of cross-section is given by
a 1:15p2 m2z m4y 3:67p2 m2z 3:0p6 m2y 4:65m4z m2y 17
where, p P =Py , mz Mz =Mpz (strong axis), my My =Mpy (weak axis).
4.4. Shear deformation
The stiness coecients should be modi®ed to account for the eect of the additional ¯exural shear
deformation in a beam-column element. The ¯exural ¯exibility matrix can be obtained by inversing the
¯exural stiness matrix as
Fig. 7. Full plasti®cation surface of Orbison.

2 3
kjj kij
hMA kii kjj kij2 kii kjj kij2
5 MA
4 18
hMB kij kii MB
kii kjj kij2 kii kjj kij2
where kii , kij , and kjj are the elements of stiness matrix in a planar beam-column. hMA and hMB are the slope
of the neutral axis due to bending moment. The ¯exibility matrix corresponding to ¯exural shear defor-
mation may be written as
" 1 1
#
hSA GAs L GAs L MA
1 1 19
hSB GAs L GAs L
MB
where GAs and L are shear rigidity and length of the beam-column, respectively. Total rotation at the A and
B is obtained by combining Eqs. (18) and (19) as

hA hMA hSA
20
hB hMB hSB
The force±displacement equation including ¯exural shear deformation is obtained by inversing the ¯exi-
bility matrix as
2 3
kii kjj kij2 kii As GL kii kjj kij2 kij As GL
6 7
MA 6 k kjj 2kij As GL kii kjj 2kij As GL 7 hA
6 ii 7 21
MB 4 kii kjj kij2 kij As GL kii kjj kij2 kij As GL 5 hB
kii kjj 2kij As GL kii kjj 2kij As GL
The force±displacement equation may be written for three-dimensional beam-column element as
8 9 2 Et A 3
>
> P > > L
0 0 0 0 0 8 > d >
9
>
> > > >
>
> MyA >>
>
6 0 Ciiy Cijy
6 0 0 07 >
7>
>
>
hyA >
>
> >
<M = 6 0 C > >
yB 6 ijy Cjjy 0 0 0 7 hyB =
7<
6 7 22
>
> MzA >> 6 0 0 0 Ciiz Cijz 0 7> hzA >
>
> >
> 6 7>
>
>
>
>
>
>
> MzB >> 4 0 0 0 Cijz Cjjz 0 5>> h >
>
: >
; : zB > ;
T 0 0 0 0 0 GJ /
L
in which
2
kiiy kjjy kijy kiiy Asz GL
Ciiy 23a
kiiy kjjy 2kijy Asz GL
2
kiiy kjjy kijy kijy Asz GL
Cijy 23b
2
kiiy kjjy kijy kjjy Asz GL
Cjjy 23c
2
kiiz kjjz kijz kiiz Asy GL
Ciiz 23d
kiiz kjjz 2kijz Asy GL
2
kiiz kjjz kijz kijz Asy GL
Cijz 23e
Fig. 8. Beam-column element with semi-rigid connections.
2
kiiz kjjz kijz kjjz Asy GL
Cjjz 23f
where Asy and Asz are the eective shear ¯exure shear areas with respect to y and z axes, respectively.
4.5. Eect of semi-rigid connection
The connection may be modeled as a rotational spring in the moment±rotation relationship represented
by Eq. (22). Fig. 8 shows a beam-column element with semi-rigid connections at both ends. If the eect of
connection ¯exibility is incorporated into the member stiness, the incremental element force±displacement
relationship of Eq. (22) is modi®ed as
8 9 2 Et A 38 9
>
> P >> L
0 0 0 0 0 > > d >>
>
> > 6 0 C C > >
>
> MyA >
>
> 6 iiy ijy 0 0 07 7
>
>
> hyA >
>
>
< = 6 0 C C 0 0 07 < =
MyB 6 ijy jjy 7 hyB
6 24
>
> MzA >
> 6 0 0 0 Ciiz Cijz 0 7

7>> hzA >
>
>
> > > >
>
> M > > 4 0 0 0 Cijz Cjjz 0 5>
>
> hzB >
>
: zB >; GJ : / ;
>
T 0 0 0 0 0 L
where
2
Cijy

Ciiy Cjjy
Ciiy RktYB RktYB

Ciiy 25a
RY
2
Cijy

C C
Cjjy RiiyktYAjjy RktYA

Cjjy 25b
RY
Cijy
Cijy 25c
RY
2
Cijz

C C
Ciiz RiizktZBjjz RktZB

Ciiz 25d
RZ
2
Cijz

Ciiz Cjjz
Cjjz RktZA RktZA

Cjjz 25e
RZ
Cijz
Cijz 25f
RZ
2
Ciiy Cjjy Cijy
RY 1 1 25g
RktYA RktYB RktYA RktYB
2
Ciiz Cjjz Cijz
RZ 1 1 25h
RktZA RktZB RktZA RktZB
in which RktYA tangent stiness of connections A in the Y -direction, RktYB tangent stiness of connec-
tions B in the Y -direction, RktZA tangent stiness of connections A in the Z-direction, and RktZB tangent
stiness of connections B in the Z-direction.
5. Veri®cation study
In the open literature, no available benchmark problems of semi-rigid space frames are available for
veri®cation study. One way to verify the proposed analysis is to make separate veri®cations for the eects of
semi-rigid connections of a planar frame and for the nonlinear behavior of the column and the space rigid
frame.
5.1. Eect of semi-rigid connections of a planar frame
Stelmack (1982) studied the experimental response of two ¯exibly connected steel frames. A two-story,
one-bay frame among his study is selected as a benchmark frame in the present study. The benchmark
frame was fabricated from the same A36 W5 16 sections, and all column bases are pinned supports in
Fig. 9. The connections used in the frame were bolted top and seat angles connections of L4 4 1=2
made of A36 with bolt fasteners of A325 3/4-in. D, and its experimental moment±rotation relationship is
shown in Fig. 10. Gravity loading of 10.7 kN (2.4 kips) was ®rst applied at third points of the beam of the
®rst ¯oor, and then a lateral load was applied as the second loading sequence. The lateral load±displace-
ment relationship was provided by the experimental work.
Fig. 9. Con®guration and load condition of Stelmack's two-story semi-rigid frame.

Fig. 10. Comparison of moment±rotation behavior by experiment and three-parameter power model for veri®cation study.
Herein, the three parameters of the power model are determined by a curve-®tting and Eqs. (4)±(9)
combined with Table 1. The curves by the experiment and by the curve-®tting result in a good agreement as
shown in Fig. 10. The parameters by Kishi±Chen equations and by the experiment show a dierence to
some degree as shown in Table 2 and Fig. 10. In spite of this dierence, Kishi±Chen equations are pre-
ferably in practical design since experimental moment±rotation curves are not available in design stages, in
general. In the analysis, the gravity load is ®rst applied, followed by the lateral load. The lateral dis-
placements by the proposed methods and by the experiment compare well in Fig. 11. As a result, the
proposed analysis is adequate in predicting the behavior and strength of semi-rigid connections.
5.2. Column with three-dimensional degree of freedom
A simply supported column with three-dimensional degree of freedom is shown in Fig. 12. W8 31
column of A36 steel is used for the analysis. The column strength calculated by the proposed analysis, Euler
solution, and DRAIN-3DX based on the slenderness parameter kc are compared in Fig. 13.
The strength of the proposed analysis compares well with Euler's theoretical solution. The maximum
error from the proposed analysis is 1.31% for the practical range of columns (kc 6 2:0). However, DRAIN-
3DX produces the maximum error of 21.16%. The large error value is a result of not considering the in-
teraction of the axial force and bending moments when considering geometric nonlinear eect.
Table 2
Comparison of the three parameters of power model for veri®cation
Parameters Curve-®tting Kishi±Chen
Initial stiness 4,520 kN-m/rad (40,000 kip-in/rad) 3,374 kN-m/rad (29,855 kip-in/rad)
Ultimate moment 24.9 kN-m/rad (220 kip-in/rad) 20.9 kN-m/rad (185 kip-in/rad)
Shape parameter 0.91 1.65
Fig. 11. Comparison of lateral displacements by experiment and proposed methods for veri®cation study.
Fig. 12. Column.
5.3. Nonlinear behavior of space rigid frame
Fig. 14 shows Orbison's six-story space frame (Orbison, 1982). The yield strength of all members is 250
MPa (36 ksi) and Young's modulus is 206,850 MPa (30,000 ksi). Uniform ¯oor pressure of 4.8 kN/m2 (100
psf) is converted into equivalent concentrated loads on the top of the columns. Wind loads are simulated by
point loads of 26.7 kN (6 kips) in the Y -direction at every beam-column joints.
The load±displacement results calculated by the proposed analysis compare well with those of Liew and
Tang's (considering shear deformations) and Orbison's (ignoring shear deformations) results (Tables 3 and
4, and Fig. 15). The ratios of load carrying capacities (calculated from the proposed analysis) over the
applied loads are 2.057 and 2.066. These values are nearly equivalent to 2.062 and 2.059 calculated by Liew
and Tang and Orbison, respectively.
Fig. 13. Comparison of column strength.
Fig. 14. Space frame of six-story.
Table 3
Result of analysis considering shear deformation
Method Proposed Liew's
Plastic strength surface LRFD Orbison Orbison
Ultimate load factor 1.990 2.057 2.062
Displacement at A in Y -direction (mm) 208 219 250
Table 4
Result of analysis ignoring shear deformation
Method Proposed Orbison's
Plastic strength surface LRFD Orbison Orbison
Ultimate load factor 1.997 2.066 2.059
Displacement at A in Y -direction (mm) 199 208 247
Fig. 15. Comparison of load±displacement of six-story space frame.
Fig. 16. Three-dimensional semi-rigid frame. (a) Con®guration and (b) load case.
Table 5
Parameter for connection
Initial stiness Ultimate moment Shape parameter
63,148 kN-m/rad (557,858 k-in/rad) 107 kN-m/rad (950 kip-in/rad) 0.524
Fig. 17. Load±displacement relationship.
6. Case study
Fig. 16 shows a four-story semi-rigid frame. Each story was 3.65 m (12 ft) high and 4.5 m (15 ft) wide.
The frame was subjected to concentrated gravity and lateral loads. The members were W14 82 for the
column and W16 40 the beam. The yield strength of all member was 248 MPa (36 ksi) and Young's
modulus was 206,850 MPa (30,000 ksi). The beam connections were top- and seat-angles of L6
4 9=16 7. All fasteners were A325 3/5-in. Diameter bolts. Three parameters computed are given in
Table 5. The load±displacement curves of semi-rigid, rigid, and hinged connection are compared in Fig. 17.
The rigid, semi-rigid, and hinged frame collapse when the applied load ratio reached 1.87, 1.45, and 0.81,
respectively. Thus, semi-rigid connection is a very crucial element to be considered in advanced analysis.
7. Conclusions
In this paper, a practical advanced analysis of three-dimensional semi-rigid frames has been developed
by combining nonlinear behavior of framed members and that of semi-rigid connection. The conclusions of
this study are as follows:
1. The proposed methods can predict accurately the combined eects of connection, geometric, and mate-
rial nonlinearities for semi-rigid frames.
2. The practical procedures for determining connection parameters are provided for a given connection
con®guration.
3. Stability functions enable only one or two element per member to capture second-order eects that make
the proposed analysis practical.
4. The CRC tangent modulus and a parabolic function consisting of member forces for gradual yielding
predict inelastic behavior reasonably well.
5. The strengths predicted by these methods are compared well those available experiments.
6. In the case study, the rigid, semi-rigid, and hinged frame collapse when the applied load ratio reached
1.87, 1.45, and 0.81, respectively. Thus, semi-rigid connection is a very crucial element to be considered
in advanced analysis.
7. The proposed analysis can be used in lieu of the costly plastic zone analysis.
Acknowledgements
This work presented in this paper was supported by funds of National Research Laboratory Program
(2000-N-NL-01-C-162) from Ministry of Science & Technology in Korea. Authors wish to appreciate the
®nancial support.
References
Al-Mashary, F., Chen, W.F., 1991. Simpli®ed second-order inelastic analysis for steel frames. Struct. Engng. 69, 395±399.
American Institute of Steel Construction, 1993. Manual of steel construction, load and resistance factor design, second ed., vols. 1 and
2, Chicago, IL.
Barsan, G.M., Chiorean, C.G., 1999. Computer program for large de¯ection elasto-plastic analysis of semi-rigid steel frameworks.
Comput. Struct. 72, 699±711.
Chen, W.F., Kishi, N., 1989. Semi-rigid steel beam-to-column connections: data base and modeling. ASCE J. Struct. Engng. 115 (1),
105±119.
Chen, W.F., Lui, E.M., 1991. Stability Design of Steel Frames, CRC Press, Boca Raton, FL.
Chen, W.F., Toma, S., 1994. Advanced Analysis of Steel Frames, CRC Press, Boca Raton, FL.
Goverdhan, A.V., 1983. A collection of experimental moment±rotation Curves and evaluation of prediction equations for semi-rigid
connections. Mater's Thesis, Vanderbilt University, Nashville, TN, 490.
Kanchanalai, T., 1977. The design and behavior of beam-columns in unbraced steel frames, AISI Project no. 189, report no. 2, Civil
Engineering/structures Research Lab., University of Texas, Austin, TX, 300.
Kim, S.E., Chen, W.F., 1996a. Practical advanced analysis for braced steel frame design. ASCE J. Struct. Engng. 122 (11), 1266±1274.
Kim, S.E., Chen, W.F., 1996b. Practical advanced analysis for unbraced steel frame design. ASCE J. Struct. Engng. 122 (11), 1259±
1265.
Kishi, N., Chen, W.F., 1986. Data base of steel beam-to-column connections, Structural Engineering Report no. CE-STR-86-26,
School of Civil Engineering, Purdue University, West Lafayette, IN, 653.
Kish, N., Chen, W.F., 1990. Moment-rotation relations of semi-rigid connections with angles. ASCE J. Struct. Engng. 116 (7), 1813±
1834.
Kishi, N., Chen, W.F., Goto, Y., Matsuoka K.G., 1991. Applicability of three-parameter power model to structural analysis of ¯exibly
jointed frames, Proc. Mechanics Computing in 1990's and Beyond, Columbus, OH, 233±237.
Liew, J.Y.R., 1992. Advanced analysis for frame design, Ph.D. Thesis, School of Civil Engineering, Purdue University, West Lafayette,
IN, 392.
Liew, J.Y.R., Tang, L.K., 1998. Nonlinear re®ned plastic hinge analysis of space frame structures, Research Report no. CE027/98,
Department of Civil Engineering, National University of Singapore, Singapore.
Lui, E.M., Chen, W.F., 1986. Analysis and behavior of ¯exibly jointed frames. Engng. Struct. 8, 107±118.
Nethercot, D.A., 1985. Steel beam-to-column connections ± a review of test data and its applicability to the evaluation of joint
behavior in the performance of steel frames, CIRIA Project Record, RP 338.
Orbison, J.G., 1982. Nonlinear static analysis of three-dimensional steel frames, Report no. 82-6, Department of Structural
Engineering, Cornell University, Ithaca, New York.
Prakash, V., Powell, G.H., 1993. DRAIN-3DX: Base program user guide, version 1.10, A Computer Program Distributed by NISEE/
Computer Applications, Department of Civil Engineering, University of California, Berkeley.
Shakourzadeh, H., Guo, Y.Q., Bato, J.L., 1999. Modeling of connections in the analyses of thin-walled space frames. Comput. Struct.
71, 423±433.
Stelmack, T.W., 1982. Analytical and experimental response of ¯exibly-connected steel frames, M.S. Thesis, Department of Civil
Environmental and Architectural Engineering, University of Colorado, 134.

Practical Advanced Analysis For Semi-Rigid Space Frames: Seung-Eock Kim, Se-Hyu Choi

Uploaded by

Copyright:

Available Formats

Practical Advanced Analysis For Semi-Rigid Space Frames: Seung-Eock Kim, Se-Hyu Choi

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Practical Advanced Analysis For Semi-Rigid Space Frames: Seung-Eock Kim, Se-Hyu Choi

Uploaded by

Copyright:

Available Formats

International Journal of Solids and Structures 38 (2001) 9111±9131

Practical advanced analysis for semi-rigid space frames

2. Behavior of semi-rigid frames

2.1. Nonlinear behavior of connections

Fig. 1. Schematic moment±rotation curves of various semi-rigid connections.

2.2. Geometric nonlinearity

2.3. Material nonlinearity

3. Practical connection modeling

The connection behavior is represented by its moment±rotation relationship. Extensive experimental

Fig. 2. Moment±rotation behavior of three-parameter model.

M 3EIt d12 6EIa d32

Mu  M0s  Mpt  Vpt d2  2Vpa d4 9

4. Practical advanced analysis

4.1. Geometric second-order eects

Et  1:0E for P 6 0:5Py 12a

4.3. Parabolic function for gradual yielding due to ¯exure

g  4a 1 a for a > 0:5 15b

Fig. 6. Full plasti®cation surface of AISC-LRFD.

4.4. Shear deformation

Fig. 7. Full plasti®cation surface of Orbison.

Fig. 8. Beam-column element with semi-rigid connections.

4.5. Eect of semi-rigid connection

5.1. Eect of semi-rigid connections of a planar frame

Fig. 9. Con®guration and load condition of Stelmack's two-story semi-rigid frame.

5.2. Column with three-dimensional degree of freedom

Fig. 12. Column.

5.3. Nonlinear behavior of space rigid frame

Fig. 13. Comparison of column strength.

Fig. 14. Space frame of six-story.

Fig. 15. Comparison of load±displacement of six-story space frame.

Fig. 17. Load±displacement relationship.

You might also like

Mu M0s Mpt Vpt d2 2Vpa d4 9

4.1. Geometric second-order eects

Et 1:0E for P 6 0:5Py 12a

g 4a1 a for a > 0:5 15b

4.5. Eect of semi-rigid connection

5.1. Eect of semi-rigid connections of a planar frame