International Journal of Steel Structures

June 2011, Vol 11, No 2, 133-144

DOI 10.1007/s13296-011-2003-5

www.springer.com/journal/13296

Parameters Affecting Response and Design of Steel Moment

Frame Reduced Beam Section Connections: An Overview
D. S. Sophianopoulos1,* and A. E. Deri2
1
Ph.D. P.E., Assistant Professor, Department of Civil Engineering, University of Thessaly, Pedion Areos, 38334, Volos, Greece
2
MSc, P.E., Ph.D. Candidate, Department of Civil Engineering, University of Thessaly, Pedion Areos, 38334, Volos, Greece

Abstract

The most crucial parameters that affect the response as well as the design of Steel Moment Frame (SMF) Reduced Beam
Section (RBS) connections are discussed. After a brief historical background concerning the development of RBS connections,
according to the reviewed literature these highly interdependent parameters are: (a) connection strength, (b) RBS profiles/sizing
and location, (c) stiffness of beam, moment frame and connection, (d) use of deep columns and associated instability
phenomena, (e) existence of composite floor slabs/lateral bracing, (f) strength and ductility of the column panel zone-beam
instability, (g) connection type and (h) column axis orientation. Their effect on the design and the seismic performance of RBS
connections is presented and recommendations, for future research required, in order to fully implement and refine the RBS
concept in European Standards and Practices, are given.

Keywords: steel moment frames, RBS connections, ductility, seismic response, design

1. Development of RBS Moment Connections

The concept of the RBS is owed to the European
researcher Plumier (1990), who, in order to design safe
structures in seismic zones, developed the idea of
“prefabricating” dissipative zones outside but close to the Figure 1. Trapezoidal straightcut steel beam-to-column RBS
connections, by creating specific weakened zones-formed connection (Plumier, 1990).
by reducing the actual cross-section of the profile, in
which yielding may occur safely and in a required ductile involved the reduction of the beam section adjacent to the
way. As a schematic example of a steel beam-to-column beam end connection to the column, thus providing a
Moment Frame connection Plumier was the first to predetermined reliable yielding “fuse” in the main
introduce the trapezoidal cut RBS, in which the flanges of member. The development of a plastic hinge at this point
the beam are trimmed in a manner shown in Fig. 1. would prevent overloading and overstraining of the
This concept, not in actual use at the time, which was connection, in full conformity with the concept of Plumier
actually a by-product of limited experiments with small mentioned earlier. Again, Arbed’s reduction in the beam
European shapes, was patented by the late European steel section was associated with a trapezoidal linear cut of
producer S. A. Arbed (‘‘Antiseismic steel structural work,’’ both flanges, identical to the one given in Fig. 1. In parallel,
U.S. Patent No. 148, 642, 1992), as reported by Iwankiw and in order to offer a beam-to-column connection with
(2004). Arbed generously waived any commercial royalty high ductility, research in Taipei resulted in the invention
rights of his patent, and hence his idea was thereafter of a tapered-cut (with or without filled edges) RBS
offered for public use in the USA. It intentionally scheme, which was patented both in Europe and the US
and owned by the National Science Foundation of Taipei
Note.-Discussion open until November 1, 2011. This manuscript for (EP 0699808B1, 1999; US Patent 559040, 1997).
this paper was submitted for review and possible publication on The damages that steel moment frames suffered during
August 7, 2010; approved on May 6, 2011. the Northridge and Kobe earthquakes in 1994 and 1995
© KSSC and Springer 2011 respectively, prompted the design profession and construction
*Corresponding author industry to review, study and revise material, design, and
Tel: +30-24210-74145; Fax: +30-24210-74169, workmanship criteria for these frames. Hence, in order to
E-mail: dimsof@civ.uth.gr fully identify and investigate the poor performance of
134 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

moment connections, as designed prior to the aforementioned 3. RBS profiles/sizing and location
seismic events, mainly two novel design strategies were
introduced. The one is associated with the strengthening Three main RBS configurations have been the subject
of the connection and the other with weakening the beam of a significant number of experimental and theoretical
framing. Both strategies force the plastic hinge away works (Iwankiw and Mohamadi, 2004; Jin and El Tawil,
from the face of the column, and problems related to 2005). The straight-cut, the tapered-cut, referred to also as
potential fragility of groove welds under triaxial stress teardrop (Carter and Iwankiw, 1998) and a slight
conditions are avoided. modification (Chen and Tu, 2004), and the radius-cut
The first approach provides a beam-to-column connection (circular, radiused), while two more rather marginal types
stronger than the beam itself, through reinforcing a short have also been reported. All these are depicted in Fig. 2,
portion of the beam near the column by using cover while more geometrical details and nominal sizing values
plates, ribs, haunches or side plates. The second approach concerning the three main types mentioned above are
however, referred as Reduced Beam Section (RBS), is given in Fig. 3 (from Carter and Iwankiw, 1998; Iwankiw
based on the selective removal of beam material adjacent and Carter, 1996).
to the connection, either from the web or from the flanges. Earlier experiments on straight-cut RBS beam-to-column
The reduction in cross sectional area will reduce the connections, conducted within the ARBED-EU project
moment capacity at a discrete location of the beam, where (Plumier, 1997), using HEB 300 columns and HEA 260
yielding will be concentrated, and thus the connection is beams, have shown that the shape of the reduction in the
protected against early fracture. The use of RBS by either beam section and the technique for realizing it should be
utilizing corrugated web I-girders (Mirgharderi et al., defined on the basis of the following: (a) the length of the
2010) or providing voids in the web (Hedayat and plastic hinge is of the order of the beam height, (b) the
Celikag, 2009) is still in its early stages and will not be technique must take into account two type of stress
considered herein. On the other hand, the well known concentrations, local ones resulting from flame cutting
“dogbone” connection, in which both beam flanges are and global ones, resulting from the shape of the reduction
trimmed, has gained a considerable attention, and a great section and (c) local stress concentrations are of higher
number of experimental and theoretical investigations order than the global ones. Moreover, in a quite informative
have been carried out and reported in the literature. In issue of Steel Tips (Moore et al., 1999), a very comprehensive
contrast to reinforced moment connections, the RBS evaluation of test results for straight cut, tapered cut and
avoids elevated force demand on the supporting column radius cut RBS moment frame connections, which had
(Iwankiw and Zoruba, 2002), increased weld-metal volume been realized up to that time, was summarized and
and joint restraint as well as supplementary material reported.
requirements. In what follows, the most important For the tapered cut (Chen et al., 1996), although the
parameters affecting the seismic response and the design intension is to allow the section modulus of the beam to
of this type of RBS connections will be discussed. match the seismic moment gradient in the reduced region
-thus promoting more uniform yielding, stress concentrations
2. Connection Strength at the re-entrant corners of the flange cut may lead to
fracture there or in the minimum beam section (FEMA
In order to ensure acceptable seismic performance, an 355D, 2000). This phenomenon has been reported for
adequate combination of connection stiffness, strength both straight and tapered cut RBS laboratory tests after
and ductility should be provided (Jin and El Tawil, 2005), significant plastic rotations. This is not the case however
a fact that is also true for any connection type. Several for the radius cut RBS, which appears to minimize stress
basic strength principles that may be used in the design of concentrations, hence reducing the danger of fractures
steel moment frames with RBS connections to achieve a within the reduced section (Engelhardt et al., 1996, 1998),
desirable balance between seismic ductility and strength the inelastic deformations distribute over the length of the
demand have been reported (Iwankiw, 1997; Carter and reduced section-providing therefore the greatest ductility
Iwankiw, 1998). Depending on the location and sizing of (Carter and Iwankiw, 1998) and the fabrication of such a
the RBS (discussed below), on the beam span and its cut is relatively simple. For all these reasons, the radius
structural shape, and on various other parameters (such as cut RBS has been prequalified in FEMA, in ANSI/AISC,
strength degradation due to local instability, strain hardening, as well as in Canadian Standards (FEMA 350, 2000;
floor slab, shear effects, plastic deformation, inelastic ANSI/AISC 358-05, 2005; CISC/ICCA, 2008) and depicted
drift-also addressed in later sections herein) new design in Fig. 4, accompanied by recommended sizing limits. It
criteria and enhanced equations, regarding the strength of should be noted however that the radius cut RBS
an RBS connection have been proposed (Iwankiw and connection is prequalified only for shallow columns.
Mohamadi, 2004; Han et al., 2009). Further details, in The radius cut RBS has only been incorporated in
conjunction with the contents of the next paragraphs, can Eurocode 8 (Design of structures for earthquake resistance)
be found in the works cited above. and more specifically in the “EN 1998-3: 2005, Assessment
 Parameters Affecting Response and Design of Steel Moment Frame Reduced Beam Section Connections: An Overview 135

Figure 2. Types of RBS configurations.

Figure 3. Schematic of widely accepted reduced beam section profile options (Carter and Iwankiw, 1998).

and retrofitting of buildings”, for improving the local schematic and the recommendations for sizing in design
ductility of steel beams and as a steel beam-to-column are shown in Fig. 5, from where one may readily perceive
connection upgrading alternative, in order to shift the the similarities between the US, Canadian and European
plastic hinge away from the face column. The corresponding dimensioning.
136 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

Figure 4. Prequalified circular (radius) cut RBS connection for seismic applications and recommended sizing limits.

4. Stiffness of Beam, Moment Frame and Iwankiw, 1998). Early post-Northridge researches reported
Connection that the decrease is of the order of 5% (Chen et al., 1996;
Iwankiw, 1997), which can be considered quite marginal
Evidently, trimming of the beam flanges in the RBS for structural design purposes. Since then, a significant
area, results in decrease of stiffness for the beam, the number of theses and technical papers were published
connection and the moment frame. The amount of this aiming to evaluate the stiffness matrix of RBS frame
decrease depends on numerous parameters, such as the elements, the knowledge of which is fundamental to the
geometry and sizing of the cut, the dimensional and quantification of elastic and inelastic performance of steel
inertia properties of the frame elements and the connection frames with RBS connections. An excellent state-of-the-
as well as on other non-structural items (Carter and art review containing the mathematical problem formulation
 Parameters Affecting Response and Design of Steel Moment Frame Reduced Beam Section Connections: An Overview 137

5. Use of Deep Columns and Associated

Instability Phenomena
The growing popularity of the RBS moment connection
combined with the fact that design engineers often use
deep columns in steel Special Moment Frames to control
drift, necessitated further experimental and analytical
works, since, as mentioned earlier, the RBS concept was
validated and prequalified for columns with shallow wide
flange sections only. The first experimental as well as
analytical investigation concerning the RBS moment
connections with deep columns was performed in the
University of California, San Diego (Uang and Chi, 2000;
Figure 5. RBS and sizing recommendation of Eurocode Chi and Uang, 2002); three full scales tests under a
8-Part 3. standard SAC loading history were reported.
Yielding was firstly observed in the panel zone, afterwards
required to determine the stiffness matrix of a 2-D RBS web local buckling in the RBS region occurred, followed
frame element has been recently presented (Bartley and by lateral-torsional buckling and flange local buckling,
Chambers, 2008). In this work one may also find proposed while no weld fracture was observed. These results
solutions to the foregoing problem, accompanied by the showed that the deep column is prone to twisting in the
most significant mathematical expressions, as well as absence of lateral bracing near the RBS region or of
approaches for accommodating the stiffness of the composite floor slab. Chi and Uang (2002) concluded that
element when structural analysis software does not accept the twisting was caused by two factors. First, since RBS
user-defined stiffness matrices. beams buckle more laterally, they introduce torsion and
More specifically, the most important contributions related out-of-plane bending in the columns and second, the
to the above formulation are those of Grubbs (1997), torsional property of deep section has a tendency to
Chambers et al. (2003), Mortensen et al. (2008), Iwankiw produce higher warping stresses to the column, schematically
and Mohammadi (2004), Dumonteil (2006), and Kim and represented in Fig. 8. Moreover, their subsequent analytical
Engelhardt (2007). Basically, the stiffness matrix [k] of a study revealed that the high ratio of the deep sections is
two-dimensional frame element relates the forces and mainly responsible for the large warping stress, and a
displacements depicted in Fig. 6, that is {p}=[k]{δ}. The verification procedure was presented for the deep column
above studies used various methods, based on this well RBS connection.
known relationship, to formulate the equation for the In Chapter 3 of the June 2002 Steel Tips Issue (Shen et
stiffness matrix of a 2-D RBS frame element shown in al., 2002), the results of an analytical investigation
Fig. 7, as for instance conjugate beam, virtual work and concerning the cyclic behavior of beam-to-column RBS
Castigliano’s theorem. The reader may refer to the above connections with deep columns was presented. More
manuscripts for details and to the review article mentioned specifically, the specimen DC-2 of the previous study
earlier for comparative conclusions. (without doubler plates, thus more realistically representing

Figure 6. Loads and deformations of a 2-D frame element (Bartley and Chambers, 2008).

Figure 7. Nonprismatic beam element with reduced beam sections (Chambers et al., 2003; Kim and Engelhardt 2007).
138 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

of RBS moment connections to deep columns, by Ricles

and collaborators (Ricles et al., 2004; Zhang and Ricles,
2006a,b). The test matrix of the experimental program
consisted of six full-scale interior RBS connections to
deep wide flange section columns; five of the specimens
had a composite floor slab and the remaining one no floor
slab and a supplement lateral brace at the end of the RBS.
The test setup, a typical connection detail with floor slab
and the corresponding detail of beam lateral bracing can
be found in the aforementioned papers. The results of the
whole study revealed the beneficial role of the presence
of a composite floor slab, since it provides restraint to the
beam, thus reducing the magnitude of the lateral
movement of both top and bottom beam flanges in the
RBS as well as the strength degradation due to beam
instability in the RBS. Evidently, it was shown that the
design recommendations proposed earlier by Chi and
Uang (2002) overestimated the column torsional warping
normal stress (since their work did not include floor slab
Figure 8. Column torsion and weak-axis bending due to and supplemental lateral braces), a finding being in
lateral-torsional buckling of beam (Chi and Uang, 2002). conformity with the postulations risen by Shen et al.
(2002), outlined in the previous paragraph.
These experimental results were verified analytically
current design practice) was modeled in ABAQUS and via 3-D FEM simulations of RBS connections in perimeter
analyzed. Additionally, after successfully simulating the special moment resisting frames with deep columns
cyclic behavior of this tested specimen, the ABAQUS (Zhang and Ricles 2006b). Parametric studies under
model ABQ-DEEP prototype was used to model the inelastic monotonic and cyclic loading were performed
connection assembly with various column sizes. The including parameters identified as: (a) beam-to-column
parametric studies conducted had a three-fold objective: connection type, (b) column sections, (c) composite floor
(a) to find out whether significant characteristics in an slap, (d) panel zone strength and (e) beam web slenderness.
RBS connection with deep columns are considered on not It was also shown that an RBS connection to deep
in current design practice, (b) to establish the effects of columns possesses less potential for ductile fracture in the
floor slab and transverse beams in bracing the connection, connection region than a welded reinforced flange
in order to prevent lateral movement of hinge areas and connection and that this potential is increased when the
(c) to investigate the connection assembly performance in column section properties are reduced, leading to larger
the presence of lateral bracing, by introducing actual strain and stress.
lateral supports from transverse beams and the concrete The contents of this section, of the preceding ones and
with metal deck floors, thus representing a common of the ones to follow clearly indicate the coupling effect
situation existing in almost all steel framed buildings. The of the parameters discussed herein, which is the main aim
conclusions drawn from the above analytical study were of this overview. In what follows, repeated references on
as follows: (a) no considerable reasons were found to similar effects will not be presented, but the reader should
suggest not using deep column sections in any moment address the cited papers for more detailed information.
frame, including special moment ones, (b) the studied
connections with deep columns satisfy the FEMA 6. Existence of Composite Floor Slabs-lateral
requirements, (c) the column twisting observed in the Bracing
experiments of Chi and Uang (discussed earlier in this
section) would not have occurred if the specimens had The role of this specific parameter in conjunction with
realistically represented actual buildings, i.e. if the the use of deep columns has already been discussed in the
specimens had transverse beams connected to the panel previous section. In addition to the above and in
zone and floor slabs, (d) using deep columns is a potential chronological order, preliminary results on research of
for saving in the cost of material and construction and (e) floor slab effects with RBS connections (Tremblay et al.,
the specimens without floor bracing tested by Chi and 1997) indicate that the floor slab should be detailed in a
Uang are unrealistic and non-representative of actual manner minimizing its interaction with the RBS. This can
structures. be accomplished via the omission of shear studs near and
Further experimental and analytical work was carried at the RBS and by providing a physical separation between
out in Lehigh University, concerning the seismic behavior the slab and the column (as for instance using a nominal
 Parameters Affecting Response and Design of Steel Moment Frame Reduced Beam Section Connections: An Overview 139

thickness of compressible material such as polystyrene parameter affecting the response of steel beam-to-column
insulation during the pouring of the slab). Moreover, the moment frame connections, regardless of their configuration
effect of floor slabs on the ductility performance of (Bertero et al., 1973; Popov et al., 1975, Engelhardt and
straight-cut RBS connections was experimentally investigated Husain, 1993). For seismic applications, the ductility
by Chen (1998), who conducted a series of large size tests developed in the beam or panel zone prior to connection
of-beam-to-column subassemblies. From his study it was failure is considered as a critically important measure of
deduced that: (a) in the absence of trimming of beam performance.
flanges the composite action of floor slabs induces strain After the Northridge and Kobe events and the gradual
concentration at the bottom flange of the beam causing increase of interest for RBS connections, the effect of the
fracture at that area, (b) the presence of RBS relaxes the panel zone (PZ) in conjunction with other parameters was
sensitivity of beam bottom flange welds, leading to a extensively re-examined. One important goal was to
much more reliable energy dissipation and provides a establish the optimal distribution of material inelastic
higher plastic rotation angle, (c) floor slabs contribute deformation in an RBS connection, since both the RBS
another 20% of extra strength and (d) in the design of the and the panel zone contribute in energy dissipation.
RBS the provided strength should be 10% less than the Initially, in an effort for simulating in a simple manner the
required one. Similar results were reported on tests inelastic response of steel frames with RBS (Shen et al.,
concerning tapered-cut RBS connection with floor slabs 2000; Kitjasateanphun et al., 2001), the panel zone was
(Chen and Chao, 2001). modeled as a rotational spring with a bilinear shear
In order to establish cyclic stability criteria for Steel distortion relationship and incorporated into DRAIN-2DX
Moment Frame RBS connections, a statistical study was software. The cyclic response of RBS connections using
performed (Uang and Fan, 2001), based on test results of two simple models was compared to the response of fine
55 full-scale specimens. Except developing relations 3-D Abaqus models, and a very good agreement was
between the response quantities (i.e. plastic rotation found. It was concluded that the panel zone has a
capacity and strength degradation rate) and the slenderness profound influence on inelastic deformation of steel
ratios of both web and flange local buckling modes, the frames. The proposed PZ model performed adequately
investigation on the effect of concrete slab revealed that and correlated well with the actual performance observed
its presence increases both the strength and the rotation during testing, but special attention should be given to
capacity of RBS beams under positive bending. However, weak panel zones with high demand to capacity ratio.
under negative bending, it was found the slab enhances In the experimental work discussed in the previous
slightly only the plastic rotation capacity, implying that it section (Jones et al., 2002), specimens having strong
does not provide any bracing to the beam bottom flange. panel zones, balanced panel zones and weak panel zones
Further experiments on radius cut RBS moment relative to beam strength were tested. The results showed
connections were carried out at the same period (Jones et that weak panel zones allow for the most stable hysteretic
al., 2002), aiming mainly to investigate the effects of response at large drift levels, since the corresponding
panel zone strength, beam web-to-column flange connection specimens performed in an excellent way without strength
(discussed herein in later sections) as well as of degradation and exhibited very good energy dissipation.
composite behavior with a concrete slab. As far as the Yielding occurred within the panel zones, leaving the
latter parameter is concerned, the presence of a composite RBS segments of the beams essentially elastic, but the
floor-contrary to previously reported findings-did not connections failed due to low-cycle induced fast fracture.
seem to lead to early fracture or other detrimental effects. The effect of panel zone strength and the web connection
As reported in the aforementioned paper’s Summary and type, the latter being the parameter discussed in the next
Conclusions “… to the contrary, the presence of the slab section, were the key variables of an experimental study
was beneficial to specimen performance by enhancing conducted on eight radius cut RBS steel moment
beam stability and delaying strength degradation.…it connections (Lee et al., 2005). This work was to a large
appears that no special treatment is needed for the slab, extent motivated by the fact that, although a significant
such as leaving a gap between the slab and the face of the amount of RBS test data was available at that time, no
column”. No shear studs were placed within the RBS specific recommendations regarding the desirable range
region to avoid possible initiation of beam flange fractures of panel zone strength had been proposed. Aiming to
at a stud weld. come up with a balanced PZ strength criterion on the
basis of their test results and the results of previous
7. Strength and Ductility of the Column experiments, the authors designed the panel zones of the
Panel Zone-beam Instability specimens adopting the following expression for the PZ
design shear strength:
The column panel zone strength and ductility has long 2
been recognized by the scientific community as well as 3bcftcf
Vp = γFycdctp 1 + -------------
- (1)
by the structural engineering practitioners as a crucial db dc tp
140 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

Figure 9. Typical geometry of radius cut RBS (a) and seismic moment profile for RBS design (b), (Lee et al., 2005).

where Fy is the yield strength of the column, db the beam forced all inelasticity action in the PZ, with the beam
depth, dc the column depth, tp the thickness of the panel remaining elastic, and eventually caused fracture of the
zone, bcf the column flange width and tcf the column beam flange. Therefore, it was recommended that a
flange thickness. Coefficient γ was taken equal to 0.75× balanced panel zone design should be used, meeting the
0.6 for the strong PZ and equal to 0.60 for the medium following criterion:
PZ specimens. Note that Eq. (1) with γ=1 has been
VRBS, p
adopted in the 2002 AISC Seismic Provisions, while the 0.70 ≤ -------------
- ≤ 0.90 (4)
strength reduction factor 0.75 was implemented in the Vp
1997 AISC Seismic Provisions. The above equation may The effect of column PZ ductility on the local and
also be written in the form: lateral instability of radius cut RBS beams in steel moment
connections has also been investigated analytically via a
1-F d t ≈ 0.6F d t
Vp = Vy(1 + CFC) , Vy = ----- yc c p yc c p , FEM based parametric study (Moslehi Tabar and Deylami,
3 2005). Similarly to the previous experimental investigation,
2 the authors in order to determine the strength of each
3bcftcf
CFC = -------------
- (2) panel zone, compared the design shear demand Vr with
db dc tp
the ultimate shear strength of the column PZ Vy. Equation
(1) with γ=1 was used for Vy, while Vr was computed by
As explained by the authors, CFC corresponds to the
the following expression:
column flange contribution to the postyield strength, and
it was used as a measure for the panel zone strength, 1 Lb – 0.5dc 1
Vr = 0.85Mp --------------- – --------------------- ---- (5)
while the measure used for the beam strength was VRBS,p, 0.95db Lb H
i.e. the panel zone shear force corresponding to the
in which Mp is the plastic moment capacity of the beam
development of the actual plastic moment of the RBS,
section. European profiles for the column and beam (IPB
given by:
and IPE respectively) were used and hence, the whole
⎛ 1---[ L + d ]⎞ design combined US guidelines regarding RBS dimensions
M ⎜2 b c ⎟ d
VRBS, p = ⎛⎝--------------
RBS, p⎞
-⎠ × ⎜ ----------------------⎟ × ⎛1 – -----b-⎞
and panel zone strength with European beam and column
(3)
db ⎜ ---- Lb ⎟ ⎝ Hc⎠ profiles as well as beam-to-column connection (welded),
⎝ 2- – e ⎠ a fact appearing for the first time in the reviewed literature.
Moreover, the properties assumed for the steel material
where, MRBS,p refers to the plastic moment at the RBS were fy =250 MPa and fu =370 MPa, that do not correspond
(based on the measured yield strength), and Hc is the to any European Steel Grade. Nevertheless, the FE modeling
column height, with the remaining geometrical properties and consequent parametric study under cyclic loading
shown in Fig. 9. produced interesting results, which to a large extent agree
The results of the tests revealed that in the presence of with the findings of previous relevant studies. The most
a strong PZ all energy dissipation is concentrated in the important conclusion drawn was that partial shear yielding
RBS region, causing a significant amount of buckling, within the panel zone can improve the hysteretic
and that lateral-torsional buckling of beams lead to column response, by avoiding premature instability of beams. It
twisting associated with severe out-of-plane deformations, was found that (a) weak PZs produce stable hysteretic
thus preventing the development of sufficient ductility, as response while their presence may lead to premature
also reported in other studies (Chi and Uang, 2002; Jones fracture of the connection before attaining the required
et al., 2002). Very weak panel zone designs, although plastic rotation of the beam, (b) for strong PZs the
having the ability of dissipating a large amount of energy, flexural moment capacity of the beam is reduced due to
 Parameters Affecting Response and Design of Steel Moment Frame Reduced Beam Section Connections: An Overview 141

lateral and local buckling and (c) for slender beams the bolted web connection, due to premature brittle fracture
hysteretic response is dominated by lateral-torsional of the beam flange at the weld access hole, was
buckling rather by web local buckling. contributed to the increased demand in the beam flanges
due to bolt slippage, by measuring the relevant strain
8. Connection Type data. As a sequel, the practice of providing web bolts
uniformly along the beam depth was brought into
In the ANSI/AISC 358-05 as well as in the CISC 2008 question via a combination of tests and FE analytical
publications, three types of connections are prequalified evaluations (Lee and Kim, 2007). The authors, on the
for steel SMF and IMF for seismic applications, namely basis of the principal stress distribution (main load path)
the Reduced Beam Section (RBS), the Bolted Unstiffened near the column face of radius cut RBS connections
(extended) end plate [BU(E)EP] and the Bolted Stiffened proposed an alternative beam flange bolted connection
(extended) end plate [BS(E)EP]. On the other hand, configuration, consisting of two separate shear tabs. The
although design and detailing rules for steel moment excellent agreement of analytical and experimental
resisting frames have been incorporated in Eurocode 8- results of this study indicates that the proposed design
Part1 (EN 1998-1: 2005), as well as design rules for overcomes the drawbacks of typical beam bolted RBS
beam-to-column connections (in dissipative zones), no steel moment connections. Marginally, a recent work
specific type of connection is recommended for seismic concerning design equations for moment strength of the
applications. Furthermore, as already discussed in Section latter should also quoted (Han et al., 2009).
3, the RBS concept in European Standards is adopted
only in EC8-Part 3, without any direct linking to the 9. Column Axis Orientation
contents and guidelines of the relevant Eurocode for Steel
Structures, i.e. EN 1993-1-8: 2005, Design of joints, All the previously reviewed literature deals with steel
which are valid only for static loading and prescribe three moment frame RBS connections to the strong column
main types of beam-to-column connections, i.e. welded, axis. The only pertinent study concerning cyclic response
bolted with extended or flush end plate and bolted with and design recommendations of weak-axis RBS connections
angle flange cleats. Although a large amount of theoretical was conducted by Gilton and Uang (2002). Accounting
and experimental studies are available for the modeling for the difference between the stress profile along beam
of these types of connections and the static and cyclic flange in strong and weak axis connections, shown in Fig.
behavior of corresponding framed structures (Chan and 10, the tests performed included two different specimens
Chui, 2000 and other works not cited herein), none of (designed, fabricated and constructed with simulated field
these has been investigated in the presence of RBS, conditions), which were loaded quasi-statically using
except in a very recent work, referred to in Section 10. standard SAC protocol.
According to the above remarks, it is of no surprise that FE analytical predictions were thereafter correlated to
the parameters dealt with herein are associated almost with the experimental results, and significant issues were
exclusively with the response of steel moment frame RBS addressed, namely the effect of RBS on stress flow, the
connections that are utilized only in northern America. necessity of far-side continuity plate in the connection
After major concerns that were expressed about the welded configuration, the determination of continuity plate stickout
flange-bolted web pre-Northridge moment connection for length, the effect of its trimming and the beam shear stress
severe seismic applications (Engelhardt and Husain, 1993), profile near grove weld. Moreover, a design procedure was
the RBS concept for enhanced ductility and energy recommended and the following conclusions were drawn:
dissipation demands lead to a significant amount of (a) prevention of brittle weld fracture was achieved in the
works with the effect of the connection type as one of the presence of RBS, which may reduce the strain concentration
major concerns. After re-establishing that sound welding near the groove weld by a factor of about three, (b) most
quality should be kept all times (Chen and Chao, 2001), of yielding and inelastic rotation took place in the vicinity
experiments conducted on radius cut RBS moment of the RBS, with no contribution of column and panel
connections (Jones et al., 2001) examined among other zone, (c) a continuity plate is not necessary on the far side
objectives the merits of a bolted web connection versus a of the one-sided connection, (d) the beam flange tensile
fully welded web connection. Although no dramatic
differences in performance between these two connection
types were perceived (as also stated earlier in Moore et
al., 1999), it appeared that a welded web connection
reduces the vulnerability of the connection to the occurrence
of a near weld fracture, a finding giving an advantage of
this connection type over the bolted one. Similar results
were drawn from a later experimental investigation (Lee Figure 10. Stress profile across beam flange: (a) strong-
et al., 2005). The poor performance of specimens with axis and (b) weak-axis connection (Gilton and Uang, 2002).
142 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

Figure 11. Details of test specimens (Pachoumis et al., 2009).

force may be effectively reduced by allowing a continuity yielding occurred at the beam’s bottom flange near the
plate stick out of at least 76 mm from the column flange face of the column; if the welds weren’t that strong, brittle
tips, (e) if the beam flange width is less than 70% of the fracture would have occurred. The strain concentration
width of the continuity plate the corner of the plate should near the face of the column was also confirmed by the FE
be trimmed and finally (f) the majority of the beam shear analysis and the authors believe that the key parameters
is taken by the beam web. for the RBS design, as specified in EC8, should be
readjust, in order to safely apply to the European profiles.
10. Use of RBS in European Steel Moment The beam-to-column connection dealt with in this study,
Resisting Frame Connections: although (except some extreme cases) rigid and due to
Recommendations for Future Refinement symmetry advantageous for seismic applications, is rather
and Implementation marginal for use in everyday engineering practice, mainly
due to quality and economic reasons-extensive welds on
From the contents of the previous Sections based on the site. On the other hand, the extended end plate bolted
existing literature, although the RBS concept is originated connection, quite similar to the group of the US prequalified
from Europe (Plumier and Arbed), it is more than bolted stiffened/unstiffened extended end plate connections
profound that its use has not been implemented in is widely used in Europe for steel moment frame
European Civil Engineering Practice, and that there is connections. This because the extended end plate is shop-
marginal research available and minimal inclusion of welded to the beam (with quality prescribed) and the
design recommendations in the relevant Eurocodes. Only beam bolted to the column on site, thus reassuring easiness
one recent publication (Pachoumis et al., 2009) has been in construction and fully balanced dimensioning, and in
detected by the authors of this overview, concerning rigid case of severe damage under reversal loading, that could
fully welded beam-to-column connections with high by design be concentrated only on the beams, repair by
quality welds at the face of the columns (as prescribed in replacing them would be possible.
EC3-1-8) with RBS under cyclic loading. In the Focusing attention on the combination of extended end
experimental part of this study, the two specimens tested plate SMF connections and RBS, the authors suggest that
(differing only in the geometry of the radius cut RBS), the following actions should be followed (not necessarily
with details presented in Figure 11, were designed-as far in the order given below) in order to fully implement and
RBS dimensioning and location are concerned-according refine the RBS concept in European Standards and
to the guidelines given in Eurocode 8-Part 3.The beam Practices: (a) Perform an optimization study for the best
profiles chosen were very common in steel moment use of HEA or HEB columns and IPE (or HEA) beams
resisting frames of European structural steelwork. in an SMF with RBS extended end plate connections
In parallel, a detailed FE modeling and analysis of the under monotone (static) loading, utilizing the component
above subassemblies was carried out, with results presented method for determining the design moment resistance of
in terms of beam moment-rotation hysteretic responses in the relevant joint, combined with the RBS radius cut
very good agreement with the experimental ones. These dimensioning suggestions given in EC8, (b) after
results are summarized in what follows. Specimen RBS1 establishing optimal combinations of use of the above
exhibited excellent performance under cyclic loading, commonly used European I-profiles, conduct extensive
since the plastic hinge was developed at the RBS area and simulations via FE analysis, in order to establish whether
the objective rotation (0.03 rad) was exceeded and these combinations found are suitable for cyclic loading
extensive yielding took place at the beams’ flanges. This applications, accounting for all the parameters involved
was not the case however for specimen RBS2, whose and discussed earlier in this study, (c) organize a full
performance was very poor. The expected formation of European experimental program in order to verify and
the plastic hinge at the RBS region did not take place, and calibrate the theoretical findings and seek for the best
 Parameters Affecting Response and Design of Steel Moment Frame Reduced Beam Section Connections: An Overview 143

possible application of RBS connections in seismic action on seismic performance of steel moment
design and, (d) through the results of all the previous connections with reduced beam sections.” Journal of
actions, incorporate into the relevant Eurocodes (EC3 and Constructional Steel Research, 57(4), pp. 417-434.
EC8) design recommendations and guidelines for the use Chen, S. J. and Tu, C. T. (2004). “Experimental study of
of RBS in steel moment frames with European I-beam jumbo size reduced beam section connections using high-
sections, and more specifically in frames using extended strength steel.” Journal of Structural Engineering, ASCE,
130(4), pp. 582-587.
end plate bolted connections.
Chi, B. and Uang, C. M. (2002). “Cyclic response and
design recommendations of reduced beam section
11. Concluding Remarks moment connections with deep columns.” Journal of
Structural Engineering, ASCE, 128(4), pp. 464-473.
In this overview, and within length limitations, a detailed Dumonteil, P. (2006). “In-Plane Properties and Modeling of
analysis and discussion of the parameters affecting the Reduced Beam Sections.” Engineering Journal, AISC,
cyclic response and design of Steel Moment Frame RBS 43(2), pp. 103-109.
connections is presented. Aiming to motivate the European Engelhardt, M. D. and Husain, A. S. (1993). “Cyclic-loading
Scientific as well as everyday practice Structural Engineering performance of welded flange-bolted web connections.”
Community to formally include the RBS concept in Journal of Structural Engineering, ASCE, 119(12), pp.
European Standard Practice, some suggestions are 3537-3550.
introduced. It is believed that the RBS is underestimated Engelhardt, M. D., Winnerberger, T., Zekany, A. I., and
in Europe and that much more attention should be given. Potyraj, T. J. (1996). “The dogbone connection: Part II.”
Modern Steel Construction, AISC, 36(8), pp. 46-55.
Engelhardt, M. D., Winnerberger, T., Zekany, A. I., and
References Potyraj, T. J. (1998). “Experimental investigation of
dogbone moment connections.” Engineering Journal,
ANSI/AISC 358-05 (2005). Prequalified Connections for AISC, 4th Quarter, pp. 128-139.
Special and Intermediate Steel Moment Frames for EN 1993-1-8 (2005). Eurocode 3: Design of Steel
Seismic Applications. American Institute of Steel Structures-Part 1.8, Design of joints. European
Construction, Chicago, Illinois, USA. Committee for Standardizations, Brussels.
Bartley, T. C. and Chambers, J. J. (2008). “State-of-the-art Gilton, C. S. and Uang, C. M. (2002). “Cyclic response and
analysis of frames with reduced beam section design recommendations of weak-axis reduced beam
connections.” Proc. Structures 2008: Crossing Borders, section moment connections.” Journal of Structural
ASCE, Canada, pp. 1-10. Engineering, ASCE, 128(4), pp. 452-463.
Bertero, V. V., Krawinkler, H., and Popov, E. P. (1973). FEMA 350 (2000). Recommended Seismic Design Criteria
“Further studies on seismic behavior of steel beam-to- for New Steel Moment-Frame Buildings. Federal
column subassemblages.” Report UCB/EERC-73-27, Emergency Management Agency, Washington, DC.
Earthquake Engineering Research Center, University of FEMA 355D (2000). State of the Art Report on Connection
California at Berkeley, USA. Performance. Federal Emergency Management Agency,
CISC/ICCA (2008). Moment Connections for Seismic Washington, DC.
Applications. Canadian Institute of Steel Construction, Grubbs, K. V. (1997). The Effect of the Dogebone Connection
Markham, Ontario, Canada. on the Elastic Stiffness of Steel Moment Frames. MSc
Carter, C. J. and Iwankiw, N. R. (1998). “Improved ductility Thesis, University of Texas at Austin.
in seismic steel moment frames with dogbone Han, S. W., Moon, K. H., and Stojadinovic, B. (2009).
connections.” Journal of Constructional Steel Research, “Design equations for moment strength of RBS-B
46(1-3), pp. 253. connections.” Journal of Constructional Steel Research,
Chambers, J. J., Almudhafar, S., and Stenger, F. (2003). 65(5), pp. 1087-1095.
“Effect of reduced beam section frame elements on Hedayat, A. A. and Celikag, M. (2009). “Post-Northridge
stiffness of moment frames.” Journal of Structural connection with modified beam end configuration to
Engineering, ASCE, 129(3), pp. 383-398. enhance strength and ductility.” Journal of Constructional
Chan, S. L. and Chui, P. P. T. (2000). Non-Linear Static and Steel Research, 65(7), pp. 1413-1430.
Cyclic Analysis of Steel Frames with Semi-Rigid Iwankiw, R. N. and Carter, C. J. (1996). “The Dogbone: A
Connections. Elsevier Science Ltd., Kidlington, UK. new idea to chew on.” Modern Steel Construction, AISC,
Chen, S. J., Yeh, C. H., and Chu, J. M. (1996). “Ductile steel 36(4), pp. 18-23.
beam-to-column connections for seismic resistance.” Iwankiw, R. N. (1997). “Ultimate strength considerations for
Journal of Structural Engineering, ASCE, 122(11), pp. seismic design of the reduced beam section (internal
1292-1299. plastic hinge).” Engineering Journal, AISC, 34(1), pp. 3-
Chen, S. J. (1998). “Effects of floor slabs on the seismic 16.
behavior of steel beam-to-column connections with Iwankiw, N. and Mohamadi, J. (2004). “Elastic In-Plane
reduced beam section.” Journal of Constructional Steel Stiffness for a Circular Cut Reduced Beam Section.”
Research, 46(1-3), pp. 218. Engineering Journal, AISC, 41(1), pp. 23-36.
Chen S. J. and Chao, Y. C. (2001). “Effect of composite Iwankiw, N. and Zoruba, S. (2002). “Steel moment frames:
144 D. S. Sophianopoulos and A. E. Deri / International Journal of Steel Structures, 11(2), 133-144, 2011

resolution of recent seismic detailing and material shape moment connections subjected to cyclic loading:
issues.” Journal of Constructional Steel Research, 58(5- Experimental analysis and FEM simulation.” Engineering
8), pp. 495-510. Structures, 21(1), pp. 216-223.
Iwankiw, N. (2004). “Seismic design enhancements and the Plumier, A. (1990). “New Idea for Safe Structures in Seismic
reduced beam section detail for steel moment frames.” Zones.” Proc. IABSE Symposium-Mixed Structures
Practice Periodical on Structural Design and Construction, Including New Materials, Brussels, Belgium, pp. 431-
ASCE, 9(2), pp. 87-92. 436.
Jin, J. and El-Tawil, S. (2005). “Seismic performance of Plumier, A. (1997). “The Dogbone: Back to the Future.”
steel frames with reduced beam section connections.” Engineering Journal, AISC, 2nd Quarter, pp. 61-67.
Journal of Constructional Steel Research, 61(4), pp. 453- Popov, E. P., Bertero, V. V., and Chandramouli, S. (1975).
471. “Hysteretic Behavior of steel beam-to-column
Jones, S. L., Fry, G. T., and Engelhardt, M. D. (2002). subassemblages.” Report UCB/EECR-75-11, Earthquake
“Experimental evaluation of cyclically loaded reduced Engineering Research Center, University of California at
beam section moment connections.” Journal of Structural Berkeley, USA.
Engineering, ASCE, 128(4), pp. 441-451. Ricles, J. M., Zhang, X., Fisher, J. W., and Lu, L. W. (2004).
Kim, K. D. and Engelhardt, M. D. (2007). “Nonprismatic “Seismic performance of deep column-to-beam welded
beam element for beams with RBS connections in steel reduced beam section moment connections.” Proc. 5th
moment frames.” Journal of Structural Engineering, International Workshop Connections in Steel Structures
ASCE, 133(2), pp. 176-184. V: Behaviour, Strength and Design, Amsterdam, The
Kitjasateanphun, T., Shen, J., Srivanich, W., and Hao, H. Netherlands, pp. 211-222.
(2001). “Inelastic analysis of steel frames with reduced Shen, J., Kitjasateanphun, T., and Srivanich, W. (2000).
beam sections.” The Structural Design of Tall Buildings, “Seismic performance of steel moment frames with
10(4), pp. 231-244. reduced beam sections.” Engineering Structures, 22(8),
Lee, C. H., Jeon, S. W., Kim, J. H., and Uang, C. M. (2005). pp. 968-983.
“Effects of panel zone strength and beam web connection Shen, J., Astaneh-Asl, A., and McCallen, D. B. (2002). “Use
method on seismic performance of reduced beam section of deep columns in special steel moment frames.” Steel
steel moment connections.” Journal of Structural Tips, Structural Steel Education Council, American
Engineering, ASCE, 131(12), pp. 1854-1865. Institute for Steel Construction, Inc.
Lee, C. H. and Kim, J. H. (2007). “Seismic design of Tremblay, R., Tchebotarev, N., and Filiatrault, A. (1997).
reduced beam section steel moment connections with “Seismic Performance of RBS Connections for Steel
bolted web attachment.” Journal of Constructional Steel Moment Resisting Frames: Influence of Loading Rate
Research, 63(4), pp. 522-531. and Floor Slab.” Proc. Stessa 97, Kyoto, Japan, pp. 664-
Mirgharderi, S. R., Torabian, S., and Imanpour, A. (2010). 671.
“Seismic performance of the Accordion-Web RBS Uang, C. M. and Chi, B. (2000). “Cyclic response and
Connection.” Journal of Constructional Steel Research, design recommendations of RBS moment connections
66(2), pp. 277-288. with deep columns.” Proc. 4th International Workshop on
Moore, K. S., Malley, J. O., and Engelhardt, M. D. (1999). Connections in Steel Structures (Connections IV),
“Design of Reduced Beam Section (RBS) Moment Frame Raonake, VA, USA, pp. 287-296.
Connections.” Steel TIPS, Structural Steel Educational Uang, C. M. and Fan, C. C. (2001). “Cyclic stability criteria
Council, Moraga, CA. for steel moment connections with reduced beam
Mortensen, B., Chambers, J., and Bartley, T. (2008). sections.” Journal of Structural Engineering, ASCE,
“Reduced Beam Section Spring Constants.” Engineering 127(9), pp. 1021-1027.
Journal, AISC, 45(2), pp. 107-116. Zhang, X. and Ricles, J. M. (2006a). “Experimental
Moslehi Tabar, A. and Deylami, A. (2005). “Instability of evaluation of reduced beam section connections to deep
beams with reduced beam section moment connections columns.” Journal of Structural Engineering, ASCE,
emphasizing the effect of column panel zone ductility.” 132(3), pp. 346-357.
Journal of Constructional Steel Research, 61(11), pp. Zhang, X. and Ricles, J. M. (2006b). “Seismic behavior of
1475-1491. reduced beam section moment connections to deep
Pahoumis, D. T., Galoussis, E. G., Kalfas, C. N., and columns.” Journal of Structural Engineering, ASCE,
Christitsas, A. D. (2009). “Reduced beam section 132(3), pp. 358-367.