Seccion reducida

STRUCTURALSTEELEDUCATIONALCOUNCIL
TECHNICALINFORMATION&PRODUCTSERVICE
AUGUST 1999
Design of Reduced Beam Section
(RBS) Moment Frame Connections
by
Kevin S. Moore, James O. Malley, Michael D. Engelhardt

DESIGN OF REDUCED BEAM SECTION (RBS) MOMENT FRAME CONNECTIONS
ABOUT THE AUTHORS
KEVIN S. MOORE is a Design Engineer with Degenkolb Engineers in San Francisco, Califor-
nia. He earned his M.S. degree at The University of Texas at Austin working under the direc-
tion of Dr. J. A. Yura and Dr. M. D. Engelhardt. While conducting research, Kevin assisted
Dr. Engelhardt with material testing for the '~UTTests," some of the first moment connection
tests following the 1994 Northridge earthquake. He was the lead engineer for a 5-stolry SMF
building utilizing RBS connections constructed in San Francisco and is a registered Profes-
sional Engineer in California.
JAMES 0. MALLEY is a Senior Principal at Degenkolb Engineers in San Francisco, Califor-
nia. He is the Project Director for Topical Investigations of the SAC Joint Venture Partnership.
The SAC Joint Venture was created to develop guideline documents for the design, evaluation,
and repair of steel moment frame buildings in response to the damage caused by the North-
ridge earthquake. Jim has been involved with many steel design and peer review projects,
including the 5-story SMF building listed above. He is a member of the AISC Committee on
Specifications and Chair of the Seismic Subcommittee and has authored numerous papers on
steel design and construction throughout his career. He is also a registered Structural Engi-
neer in California.
MICHAEL D. ENGELHARDT is an associate professor of Civil Engineering at The University
of Texas at Austin. Mike teaches courses on structural steel design at The University of Texas
and conducts research on seismic resistant steel framing. His previous work includes major
contributions to the development and validation of eccentrically braced frames (EBFs). Mike
has been an active participant in moment connection research since the 1994 Northridge
earthquake and has worked extensively on RBS related research. Mike is a member of AISC
Task Committee Number 113 on Seismic Design and is a registered Professional Engineer in
California.

CONTENTS
I.
.
3.
4.
o
6.
.
INTRODUCTION ....................................................................................................... 1
1.1 DESCRIPTION OF SMF ................................................................................. 1
1.2 BACKGROUND OF RBS ............................................................................... 2
HISTORY OF THE DEVELOPMENT OF RBS SMF CONNECTIONS ........................ 3
2.1 INITIAL RESEARCH ....................................................................................... 3
SUMMARY OF TEST RESULTS .............................................................................. 4
3.1 OVERVIEW OF TEST RESULTS FOR RADIUS CUT RBS SPECIMENS ........... 4
RBS DESIGN PROCEDURE FOR SMFS .................................................................. 6
4.1 RBS DESIGN ................................................................................................. 6
4.2 RBS SIZING .................................................................................................. 7
4.3 STEP-BY-STEP PROCEDURE ...................................................................... 10
4.4 ADDITIONAL DESIGN CONSIDERATIONS ................................................... 15
RBS DESIGN EXAMPLE ....................................................................................... 18
PROCEDURES FOR ACCEPTANCE OF DESIGN BY BUILDING AUTHORITIES ...21
6. I COMMUNICATION ....................................................................................... 21
6.2 METHODOLOGY ......................................................................................... 22
6.3 CONSTRUCTION DOCUMENTS ................................................................... 22
FABRICATION AND INSPECTION ISSUES ........................................................... 22
7.1 CUTTING AND GRINDING ........................................................................... 22
7.2 WELDING .................................................................................................... 23
REFERENCES ....................................................................................................... 25
APPENDIX A ......................................................................................................... Ai
LIST OF FIGURES
1.1
1.2
2.1
2.2
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
5.1
5.2
5.3
PRE-NORTHRIDGE MOMENT CONNECTION DETAIL ................................................. 1
RADIUS CUT RBS MOMENT CONNECTION ............................................................... 2
TAPERED CUT RBS MOMENT CONNECTION ............................................................ 3
EXAMPLE OF LABORATORY BEHAVIOR OF RADIUS CUT RBS TEST SPECIMEN ..... 4
(A) DETAIL OF TEST SPECIMEN ........................................................................... 4
(B) RESPONSE OF TEST SPECIMEN ..................................................................... 4
MOMENT DIAGRAM AND BEAM GEOMETRY FOR RBS ............................................. 7
GEOMETRY OF RADIUS CUT RBS ............................................................................. 8
TYPICAL MOMENT FRAME BEAM WITH RBS CONNECTIONS ................................... 8
BEAM AT MINIMUM SECTION OF RBS .................................................................... 10
FREE BODY DIAGRAM BETWEEN CENTERS OF RBS ............................................ 11
FREE BODY DIAGRAM BETWEEN CENTER OF RBS
AND FACE OF COLUMN FLANGE ............................................................................ 12
FREE BODY DIAGRAM FOR CALCULATION OF COLUMN MOMENTS ...................... 13
COMPARISON OF TEST RESULTS FOR COVER PLATED AND RBS CONNECTIONS 17
RBS DIMENSIONS ................................................................................................... 18
PORTION OF EXAMPLE BEAM BETWEEN RBS CENTERS ....................................... 19
CONNECTION DETAIL FOR DESIGN EXAMPLE ....................................................... 21

I. Introduction
When subjected to a major earthquake, build-
ings designed to meet the design require-
ments of typical building codes, such as the
UniI'orm Building ~ Code (1997), are expected
to have damage to both structural and non-
structural elements. The structural design for
large seismic events must therefore explicitly
consider the effects of response beyond the
elastic range. The "Special Moment Frame"
(SMF) steel building system is designed such
that the connections between the frame
beams and columns absorb substantial
energy and provide major contributions to the
displacement ductility demand.
1.1 Description of SMF
Recent studies by Lee (1997) and others have
demonstrated that this assumption is far dif-
ferent from the actual behavior.
l
~ ~--~ C.P.~70T-4
I I : I . 7/8" A325-XBOLTS 1
A SMF lateral force resisting system is often
preferred by building owners and architects
because this type of system provides large
unobstructed spaces throughout the build-
ing plan. This "open" layout offers the most
flexibility for programming the spaces as well
as architectural appointments. For these rea-
sons, steel buildings with SMF systems are
quite common in major commercial and
institutional structures. Furthermore, the
SMF system is considered by many to be one
of the most ductile steel building systems
available to the engineer. For this reason,
SMF systems have been widely used in areas
of high seismicity.
SMFs are typically comprised of connec-
tions between wide flange beams and
columns where beam flanges are welded to
column flanges utilizing complete joint pene-
tration welds. Figure 1.1 shows a typical
unreinforced design detail for a beam-to-col-
umn connection used in SMF systems prior
to the 1994 Northridge earthquake. Common
practice prior to the Northridge earthquake
was to either bolt or weld the web to the col-
umn shear plate, and to weld the beam
flanges to the column flange using a com-
plete joint penetration groove weld. Histori-
cally, designers have assumed that beam
shear is transferred to the column by the
beam web connection and the moment is
transferred through the beam flanges.
Figure 1. I Pre-Northridge
Moment Connection Detail
In the design of SMF connections, the
engineer must set objectives for both load
and deformation capacities. Usually, the load
capacity requirement is based on the plastic
moment of the beam. The connection must
be strong enough to develop the strength of
the beam, thus reducing the risk of brittle
failure in the connection. Inelastic deforma-
tion capacity is required to assure ductility in
predetermined locations when subjected to
large deformation demands.
After some of the problems observed in
SMF connections after the Northridge earth-
quake, a common philosophy has been to
design the connection to remain nominally
elastic at the column face, and force the
inelastic deformation of the frame to occur in
a portion of the beam, away from the con-
nection. This philosophy is executed by using
a "capacity design" approach. The plastic
moment and associated shear of the beam is
based on probable strengths of materials.
These maximums then become the design
loads for the connection. The connection of
the beam to the column flange is then
designed using nominal material properties.
Most post-Northridge connection designs
locate the plastic hinge (where inelastic

DESIGN OF REDUCED BEAM SECTION (RBS} MOMENT FRAME CONNECTIONS
deformations are concentrated in the SMF
beam) away from the column flange through
reinforcing a short portion of the beam near
the column. By increasing the strength of the
beam in this region, a plastic hinge will tend
to form just adjacent to the reinforced por-
tion of the beam. The inherent difficulty with
utilizing a reinforced beam-column connec-
tion is the increased material and labor costs
associated with this connection and the SMF
system as well as requiring welds that are
difficult and costly to make and inspect.
1.2 Background of RBS
Another type of connection developed to force
the inelastic deformation away from the
beam-column interface is referred to as a
"Reduced Beam Section" connection (RBS) or
"dogbone". This connection relies on the
selective removal of beam flange material
adjacent to the beam-to-column connection,
typically from both top and bottom flanges,
to reduce the cross sectional area of the
beam. This reduction in cross sectional area
will reduce the moment capacity at a discrete
location in the beam. Various shapes of
cutouts are possible, including constant cut,
tapered cut, radius cut and others. Figure
1.2 illustrates a radius cut RBS connection.
The Luxembourg-based steel manufac-
turing company, ARBED, held a 1992 US
patent on the reduced beam section (RBS).
' A
~L~--..
Figure 1.2
Radius Cut RBS Moment Connection
Following the Northridge earthquake, they
waived all patent and claim rights associated
with the RBS for the benefit of the profession.
This gracious gesture allowed further devel-
opment of the concept for use in post-North-
ridge SMF buildings.
The shape, size and location of the RBS
all have an effect on the connection demands
and performance. Various shapes have been
tested and used in new construction during
the past several years. Test programs have
been performed to investigate straight cut
(Plumier, 1997), taper cut (Chen, et.al. 1996)
and radius cut (Engelhardt 1997; Tremblay,
et.al. 1997; Popov, et.al. 1998) reduced beam
sections.
The RBS forces yielding and hinge forma-
tion to occur within the reduced section of
the beam and limits the moment that can be
developed at the face of the column. By
reducing demands on the beam flange groove
welds and the surrounding base metal
regions, the RBS reduces the possibility of
fractures occurring in this vulnerable region.
Although the RBS essentially weakens the
beam, its impact on the overall lateral
strength and stiffness of a steel moment
frame is generally quite small.
The inelastic deformation focused in an
RBS connection remains in the reduced
beam section, which can be designed and
located such that minimal protective meas-
ures need to be taken at the connection of
beam to column. The smaller moment gener-
ated at the face of the column for an RBS
connection, in addition to reducing stress
levels on the welds, also offers some advan-
tages in satisfying strong column-weak beam
requirements and in minimizing column
doubler plate requirements.
Fabrication and erection of the RBS con-
nection avoids the addition of strengthening
plates and special weldments that are
required of many post-Northridge moment
connections. Consequently, the RBS connec-
tion is very competitive from a cost perspec-
tive. Because of the competitive cost and
established performance based on extensive
testing and analysis, the RBS connection
appears to be a cost effective, consistently
performing connection for use in the seismic
design of SMF building structures.
2

DESIGN OFFREDUCED BEAM SECTION (RBS) MOMENT FRAME CONNECTIONS
. History of the Development
of RBS SMF Connections
A number of significant events led to the cur-
rent environment surrounding SMF design
and construction methodologies. Concerns
over material properties, connection geome-
try, design parameters and weld quality are
just a few issues which became a concern
after brittle failures were observed in SMF
moment connections after the 1994 North-
ridge earthquake.
SMF structures were still being designed
and requested by owners for all the reasons
described earlier. The pre-Northridge con-
nection detail had become a driving eco-
nomic factor for the viability of the SMF sys-
tem. To redesign moment connections in a
SMF system utilizing expensive connection
reinforcement techniques made this building
system less competitive.
2.1 Initial Research
A significant amount of research and testing
on RBS moment connections has already
been completed, and additional work is
underway. Appendix A provides a listing of
tests on RBS connections. The list includes
key features of each test, including member
sizes and strengths, connection details, RBS
size and shape, and the plastic rotation
achieved by each test assemblage. As indi-
cated by the data-in Appendix A, successful
tests have been conducted on constant cut,
tapered cut and radius cut RBS specimens.
The tapered cut, shown in Figure 2.1, is
intended to allow the section modulus of the
beam to match the seismic moment gradient
in the reduced region, thereby promoting
more uniform yielding within the reduced
section. This is intended to create a reliable,
uniform hinging location. However, stress
concentrations at the re-entrant corners of
the flange cut may lead to fracture at these
locations. After significant plastic rotation,
both the constant cut and tapered cut RBS
connections, have experienced fractures
within the RBS in some laboratory tests.
These fractures have occurred at changes in
section within the RBS, for example at the
minimum section of the tapered RBS. These
changes of cross-section introduce stress
concentrations that can lead to fracture
within the highly stressed reduced section of
the beam.
I~ ~= ~
Figure 2. I
Tapered Cut RBS Moment Connection
The radius cut RBS appears to minimize
stress concentrations, thereby reducing the
chances of a fracture occurring within the
reduced section (Engelhardt, et.al. 1996).
Furthermore, test results indicate that
inelastic deformations distribute over tl~e
length of the reduced section. The radius cut
is also relatively simple to fabricate.
Figure 2.2 shows an example of a labora-
tory test of a radius cut RBS specimen. The
connection detail is shown in Figure 2.2(a)
and the moment versus plastic rotation
response is shown in Figure 2.2(b). As is typ-
ical of most radius cut RBS tests, this speci-
men showed excellent performance.
As shown in Figure 2.2(a),.it is important
to note that most RBS test specimens, in
addition to incorporating the RBS, also incor-
porated significant improvements in welding
and in other detailing features as compared
to the pre-Northridge connection. All speci-
mens were constructed using welding elec-
trodes that exhibit improved notch tough-
ness as compared to the E70T-4 electrode
commonly used prior to the Northridge
earthquake.
The majority of specimens also incorpo-
rated improved practices with respect to
backing bars and weld tabs. In most cases,
bottom flange backing bars were removed,
backgouged and sealed with a fillet weld, and
top flange backing bars were seal welded to
the column. Weld run-off tabs were removed
in most cases. In addition to welding related
improvements, most specimens also incorpo-
3

~ ~ / B.U, bar to remain
---.--~"~J / ~ Remove weld tabs
• ~"~'>~....... i~'i"" Note:
i ~ ~ 45° ~ All field welds: E71T-8
r ~ ")(S~if~ed CVN=20.-~
~ ~ -~W~,lg4
~ ~ltS: 1" A3~ 25 9" C-C ~
' ~Holes: 1-1/16" DIA. J
• E ~8" x 6" x 2'-6" /
~Z ................. ~
....
~ ~ I k cleaned and ins~cted
~ Reave B.U. bar
IN k Remove ~ld tabs
~8
~oo
" 3'-4" Radius
~ ~ /Grind Smith
5/1~ ~/ ~ / Grind Parallel to Beam Flange
/
~ ~ ~ 2.31"
~ ,~ ~
9" 27"
(a) Detail of ~est Specimen
d CVN = 20 ft-lbs at -20 deg F)
40000 . '
$1:~¢.~B4 I
i.,0000
i -20000
•.30000 I
~0000
.0.0~ .0.114
~ Moment ~ndRotafJonComputed
v,lth Rs~pe¢~to Faca o~Col,~nn
I I I
,-0.03 -0.02 .0.01 0 0.01 0.02 0.03 0.04 0.05
Total Plastic Rotation (radian)
(b) Response of Test Specimen
Figure 2.2 Example of Laboratory Behavior of Radius Cut RBS Test Specimen
rated additional detailing improvements.
Consequently, although the beam flange
cutouts are the most distinguishing feature
of the RBS connection, the success of this
connection in laboratory tests is also likely
related to the many other welding and detail-
ing improvements implemented in the test
specimens, i.e. the use of weld metal with
improved notch toughness, improved prac-
tices with respect to backing bars and weld
tabs, use of continuity plates, etc.
3. Summary of Test Results
The table in Appendix A provides a listing of
RBS test data. While this list may not be
exhaustive or contain every test performed
on RBS beam-column subassemblies or
ancillary testing to support performance, the
list does provide the reader with a substan-
tial amount of documented performance con-
ditions for this connection. The table also
includes RBS tests completed under the SAC
Phase 2 research program as of mid-1999.
These test results have not been formally
published, but are included based on avail-
able test reports.
The AISC Seismic Provisions for Structural
Steel Buildings (1997) require qualification
testing for SMF connection designs. The test
results reported in Appendix A may be useful
in satisfying th~se qualification test require-
ments. Appendix S of the Seismic Provisions
for Structural Steel Buildings provides guide-
lines on extrapolating test results beyond the
tested member sizes.
Appendix A includes listings for 43 RBS
tests. This number does not include tests by
Plumier (1997), or shaking table tests by
Chen, Yeh and Chu (1996). Additional tests
have also been conducted on specimens in
which the RBS was provided in the bottom
flange only for use as a retrofit measure for
existing moment frame connections. These
RBS retrofit tests are not reported in Appen-
dix A. Information on the tests is available in
the AISC Steel Design Guide Series Twelve
(Gross, et.al. 1999).
3.1 Overview of Test Results for
Radius Cut RBS Specimens
This section provides an overview of the test
data listed in Appendix A for radius cut RBS
test specimens. There are 27 radius cut RBS
tests listed in the table. Examination of this
data indicates that these connections devel-
oped plastic rotations ranging from 0.029 rad
to beyond 0.05 rad. These results suggest
that the radius cut RBS connection can
develop large plastic rotations on a consis-
tent basis. Also notable is the fact that a
4

large number of radius cut RBS connections
have been tested under a variety of condi-
tions by a number of different investigators,
and there has not been a single test with
poor performance. This suggests the connec-
tion is quite robust and reliable.
The data in Appendix A demonstrates the
possible ultimate failure modes for the radius
cut RBS connection. In many tests, specimen
strength gradually deteriorated due to local
and lateral torsional buckling, and testing
was terminated due to limitations of the test-
ing equipment or test setup. However, a
number of connections have been loaded well
past the occurrence of local flange buckling
within the RBS, and ultimately failed by low
cycle fatigue fracture of the RBS. Only one of
the 27 radius cut RBS specimens experi-
enced a fracture at the beam-to-column con-
nection. This specimen, designated "DBBW-
C - Beam 2" in Appendix A, fractured in the
beam bottom flange base metal adjacent to
the groove weld, with the fracture initiating
at the weld access hole. However, even this
connection developed 0.038 rad. of plastic
rotation prior to fracture.
Most of the radius cut RBS specimens
have been tested pseudo statically, using a
loading protocol in which applied displace-
ments are progressively increased. However,
one specimen ("S-l") was tested monotoni-
cally to failure. Two specimens ("LS-2" and
"LS-3") were tested using a loading protocol
intended to represent near source ground
motions that contain a large pulse. Finally,
two specimens ("S-4" and "SC-2") were tested
dynamically. The radius cut RBS specimens
have performed well under all of these load-
ing conditions.
A wide range of beam sizes have been
tested with the radius cut RBS. The smallest
beam listed in Appendix A is a W530x82
(Canadian designation) which is roughly
equivalent to a W2 lx50. The heaviest beam
tested is a W36x300. All columns for radius
cut RBS tests have been W14 sections. Most
of the columns have been sized to provide for
a very strong panel zone, although a small
number of tests have included moderate
panel zone yielding. No tests have been con-
ducted on specimens with very weak panel
zones. However, such tests will be completed
during 1999.
Of the 27 radius cut RBS specimens
listed in Appendix A, there are no reported
cases of weld fracture. Beam flange groove
welds for all radius cut RBS specimens have
been made by the self shielded flux cored arc
welding process (SS-FCAW) using electrodes
with a minimum specified CVN toughness of
20 ft.-lbs, at-20 ° F. Three different electrode
designations have been used in these tests:
E71T-8, E70TG-K2, and E70T-6. For one of
the radius cut RBS specimens, details of the
backing bars were not reported. However, for
the remaining 26 specimens in which back-
ing bar details were reported, the bottom
flange backing was removed and the top
flange backing was left in place. For the
majority of these specimens, the top flange
backing was seal welded to the face of the
column, although these seal welds were not
provided in four specimens (WG-1 to WG-4).
Note that only one of the 27 radius cut RBS
specimens used cover plates at the beam-to-
column connection as a supplement to the
RBS.. The remaining 26 specimens used no
supplemental reinforcing measures (cover
plates, ribs, etc.) at the connection.
Dimensions of the RBS cuts for the 27
radius cut specimens vary over a fairly small
range. The distance from the face of the col-
umn to the start of the RBS cut (designated
as L 1 in Appendix A) ranged from 50 to 75%
of the beam flange width. The lengths of the
cuts (designated as LRBS in Appendix A)
have varied from 74 to 82% of the beam
depth. The amount of flange width removed
at the minimum section of the RBS (desig-
nated as FR in Appendix A) has varied from
38 to 55%.
Two types of web connection details have
been used for radius cut RBS test specimens:
a welded and a bolted detail. In the welded
detail, the beam web is welded directly to the
column flange using a complete joint pene-
tration groove weld. For the bolted detail,
fully tensioned high strength bolts are used.
Approximately half the specimens have used
the bolted detail, and half the welded detail.
The data indicates no significant difference in
performance for radius cut specimens.
5

Beam lateral bracing details have also
varied among the radius cut RBS specimens.
Of the 27 specimens, seven are reported to
have provided a brace at the RBS. For the
remaining 20 specimens, the lateral brace
was typically further away from the RBS
placed near the point of load application.
Finally, of the 27 radius cut specimens
listed in Appendix A, six were tested with a
composite concrete floor slab. For Specimens
"SC-1" and "SC-2," a one-inch gap was inten-
tionally left between the face of the column
and the slab, in an attempt to minimize com-
posite action. For Specimens "DBBW-C
Beams 1 & 2" and "DBWW-C Beams 1 & 2,"
no such gap was provided. No detrimental
effects of the slab were observed in any of
these tests. In some tests, the investigators
noted that the slab enhanced overall energy
dissipation by delaying beam instability. Note
that for all composite specimens, no shear
studs were placed in the region of the RBS or
between the face of the column and the start
of the RBS.
As described above, a rather wide range of
conditions has been investigated in RBS test-
ing completed to-date. Testing of RBS con-
nections is continuing under the SAC pro-
gram and for specific building construction
projects. The reader is encouraged to remain
abreast of this data, as it becomes available.
Even though many variables have already
been investigated in RBS testing, there are a
number of conditions that have received less
attention. These conditions, when they arise
in design, should be approached with cau-
tion since data is lacking in these areas. In
such cases, additional testing may be war-
ranted. For example, no radius cut RBS con-
nections to the weak axis of a wide flange col-
umn have been tested, although data for
some other RBS connections to the column
weak axis are available (see Specimens
"COH-3" and "COH-4" in Appendix A). No
specimens with deep columns have yet been
considered. Further, no tests on specimens
with very weak panel zones have been con-
ducted. Future research is underway to
address these and other issues.
4. RBS Design Procedure for
SMFs
The following sections contain recommenda-
tions for the design of new radius cut RBS
moment connections. Based on the suc-
cesses outlined above, and the preference of
engineers designing new SMF structures, the
design methodology presented herein focuses
on the radius cut RBS shape. Globally
important design parameters such as panel
zone participation, beam shear and overall
frame drift are addressed as part of the rec-
ommended procedure. Many important
aspects of moment connection design are
applicable and must be considered when
designing SMF RBS connections. The RBS
design methodology should be performed in
conjunction with available test results as
part of the justification of the design proce-
dure.
The initial part of the SMF/RBS design is
to determine the configuration of the moment
frames, the typical bay sizes, plan dimen-
sions and frame locations. Many of these
requirements are determined by others,
(architects, owners, developers), but the
engineer should influence these decisions
based on sound design practices. One exam-
ple would be to consider the bay size if a
SMF/RBS system is to be utilized. Because
of the high moment gradient ratio associated
with short bays, more beam flange removal
in RBS connections will be required for short
bay frames than long bay frames. In addi-
tion, beam sizes may be affected. With proper
guidance, the engineer can supply informa-
tion that will help the architect develop a
rational, efficient building design. Upon
determination of the basic structural param-
eters, the engineer can begin the member
and connection design process.
4.1 RBS Design
The engineer will begin the design of the
structure by determining the force level and
drift limits to be incorporated as part of the
design. These parameters are typically set by
a model building code such as the Uniform
Building Code (1997) or, in the future, the
6

International Building Code. Once the force
level is determined based on site condtions,
structural system, seismicity of the region
and target drift limits, the engineer can begin
the design of the seismic system using the
AISC Seismic Provisions for Structural Steel
Buildings {1997).
Based on the required design parameters,
the engineer will determine the beam and
column sizes required to meet drift limits,
etc. It is important that the engineer remem-
ber that the frame is less stiff due to the RBS
design, than a "typical" non-RBS SMF.
After proper beam-column sizes have
been determined for the frame, the RBS
design procedure should be followed to
develop the proper flange reduction to pro-
duce the desired performance. Many of the
design steps and recommendations parallel
information provided in reports referenced at
the end of this document.
The strength of the beam at the minimum
section of the RBS must satisfy code require-
ments under all applicable load combina-
tions including gravity, wind, and other loads
appropriate for the structure under consider-
ation. Beam sizes in typical SMFs are nor-
mally governed by code specified drift limits.
Consequently, even with a reduction in beam
moment due to the addition of the RBS, the
strength of the modified frame will often be
satisfactory for all load combinations. In
some cases, a minor increase in beam size
may be needed.
The addition of RBS cutouts will reduce
the stiffness of a steel moment frame. This
reduction in stiffness, although generally
quite small, may affect the ability of the
frame to satisfy code specified drift limits. A
recent study by Grubbs (1997) evaluated the
reduction in elastic lateral stiffness of steel
moment frames due to the addition of radius
cut RBS connections. This study showed
that over a wide range of frame heights and
configurations, the average reduction in stiff-
ness for a 50 percent flange reduction was on
the order of 6 to 7 percent. For a 40 percent
flange reduction, the reduction in elastic
frame stiffness was on the order of 4 to 5 per-
cent. If this reduction in stiffness is a con-
cern, drift can be computed in the usual
manner using a model that does not explic-
itly account for the RBS, and then increased
by the amounts noted above to account for
the RBS connections. Alternatively, a refined
structural model, including the reduced stiff-
ness at each connection due to the RBS, can
be developed to check the stiffness of the
frame.
4.2 RBS Sizing
The location and size of the RBS will dictate
the level of stress at the beam flange-column
flange connection. The RBS seismic moment
diagram is presented in Figure 4.1 and indi-
cates the Nominal Capacity, the Probable
Demand, and the Nominal Demand for the
RBS beam. Note that M'p RBS is the maxi-
mum moment expected at l~he face of the col-
umn flange when the RBS has yielded and
strain hardened under combined earthquake
and gravity loads. M' p RBS is directly influ-
enced by the Probable iJemand, and the loca-
tion of the RBS. M' P,RBS is later referred to
as Mf in this document.
r--~ r...... ,~;~,~-~, .............................. i
, I ,
~ ~,~as i
~--~,-,,~o~
Moment Diegrem
L~
~am ¢,¢~y
Figure 4. I Moment Diagram and
Beam Geometry for RBS
The overall goal in sizing the RBS cut is to
limit the maximum beam moment that can
develop at the face of the column to values in
the range of about 85 to 100 percent of the
beam's actual plastic moment. This
approach, in effect, limits the average maxi-
mum stress at the beam flange groove welds
to values on the order of the actual yield
stress of the beam. Experiments have shown
that connections detailed in accordance with
7

the recommendations provided below are
capable of safely resisting this level of
moment. As a point of comparison, tests on
pre-Northridge moment connections without
RBS cutouts often show maximum moments
at the face of the column of about 125 per-
cent of M~ or greater (Popov, Stephen 1972;
Tsai, PopoPv 1988; Engelhardt, Husain 1993).
Consequently, the addition of the RBS
cutouts in the beam results in a substantial
reduction in moment at the face of the col-
umn.
Much of the design procedure presented
below follows recommendations of the
Interim Guidelines: Evaluation, Repair, Modi-
fication and Design of Welded Steel Moment
Frame Structures (FEMA 267) (1995) and the
Interim Guidelines Advisory No. 1, Supple-
ment to FEMA 267 (FEMA 267A) (1997), with
several exceptions. Most significant of these
exceptions is that FEMA 267A places a limit
on the maximum stress permitted at the face
of the column equal to ninety percent of the
minimum specified yield stress of the col-
umn. For the case of an A992 (A572 Gr. 50)
column, this results in a limit of 45 ksi. This
limit was established to address concerns
regarding the potential for through-thickness
failures in column flanges. The design proce-
dure limits the maximum stress at the face of
the column to a value on the order of the
actual yield stress of the beam. This excep-
tion to the requirements of FEMA 267A has
been adopted for several reasons. First, spec-
imens designed according to the procedures
described herein have performed well in lab-
oratory tests. Second, satisfying the 45 ksi
stress limit, would result in large flange
cutouts in many cases, or would require sup-
plemental flange reinforcement such as cover
plates or ribs. Further, recently completed
research conducted under the SAC Phase 2
program suggests that the potential for
through-thickness failures is considerably
less than previously thought, and that the
current limit of 45 ksi can most likely be
increased without posing an increase in risk
of fracture initiation.
The design procedure assumes that a
radius cut RBS is provided in both the top
and bottom flanges at the moment connec-
tion at each end of a moment frame beam.
The procedure also assumes the minimum
specified yield stress of the beam is 50 ksi or
less (Gr. 50 beams), and that the minimum
specified yield stress of the column is 50 ksi
or greater (Gr. 50 or Gr. 65 columns).
Figure 4.2 shows the geometry of a radius
cut RBS, and Figure 4.3 shows the entire
moment frame beam. The key dimensions
I~ ~1 ~
a
4c~+ d
R = radius of cut
8c
C
~1
--1
b
Figure 4.2
Geometry of Radius Cut RBS
that must be chosen by the designer are a,
the distance from the face of the column to
the start of the RBS cut, b, the length of the
RBS cut, and c, the depth of the RBS cut at
its minimum section. The radius of the cut R
can be related to dimensions b and c based
on the geometry of a circular arc, using the
equation in Fig. 4.2. The amount of flange
material that is removed at the minimum
section of the RBS is sometimes referred to
the percent flange removal which is com-
puted as (2c/bf.) x 100, where bfis the unre-
duced flange v~idth of the beam~
In past research tests, the dimensions a
and b have generally been chosen based on
the judgment of the researchers. In general,
these dimensions should be kept as small as
• w = uniform beam gravity load ~ II II
RBS RBS
__ ~ ~.~_.£1l.~r.! ~ ~ 1 I } I I t ~ ~ 1 t I } ~ l ~ l ~ ~.!?..t.~.!.|~[~]
' &4i i~ ,- ,n -~ ,n - ~
•
,, lla +~ " L' = distancebe~een ~nters of RBS ~ts ~a+ ~ ~
I ~
L : distance between column ¢entedines
Figure 4.3
Typical Moment Frame Beam with
RBS Connections
8

possible in order to minimize the increase of
moment between the plastic hinge located in
the RBS and the face of t_he column.
The dimension a should be large enough,
however, to permit stress in the reduced sec-
tion of the beam to spread uniformly across
the flange width at the face of the column.
Similarly, the dimension b should be large
enough to avoid excessive inelastic strains
within the RBS. Based on an evaluation of
successful past tests, the following sugges-
tions are made for selecting these dimen-
sions:
(o.s to o.Ts) bf tl)
b ~ (65 to 0.85)d (2)
where by and d are the beam flange width
and delSth. Examination of RBS test data
indicates that successful connection per-
formance has been obtained for a wide range
of values for a and b. Consequently, a great
deal of precision in choosing these values
does not appear justified and Equations 1
and 2 should be considered an approximate
guide.
The remaining dimension that must be
chosen when sizing the RBS is c, the depth of
the cut. The value of c will control the maxi-
mum moment developed within the RBS, and
therefore will control the maximum moment
generated at the face of the column. As noted
above, the final dimensions should be chosen
so that the maximum moment at the face of
the column is in the range of about 85 to 100
percent of the beam's actual plastic moment.
At present, it is suggested to avoid utilizing
flange reductions greater than about 50 per-
cent. Thus, the value of c should be chosen
to be less than or equal to 0.25bf.
The basic approach taken in "this proce-
dure is to choose preliminary values for a, b,
and c, then compute the moment at the face
of the column, and check this moment
against the limit noted above. Some iteration
in the RBS dimensions may be needed to
arrive upon a satisfactory design. Further
design checks are completed upon satisfac-
tory sizing of the RBS.
The beam size will typically be chosen for
drift requirements, followed by some amount
of flange reduction. The designer must exam-
ine the effect of all applied loads at the RBS
location. It is possible that beam size may
need to be adjusted, and different RBS sizing
and location must be determined, to meet all
design criteria.
This RBS sizing determination is also
applicable when retrofitting existing SMF
structures. Access is limited or impossible at
the upper flange of the beam, due to the
presence of a floor slab, so RBS modifications
typically occur at the bottom flange of the
moment beam only. If access is available to
the top flange of the beam, it is recommended
to apply the RBS design methodology to both
flanges. There has been a great deal of effort
and research spent on the use of RBS modi-
fications to existing SMFs. The AISC Design
Guide Series Twelve (1999) that summarizes
this work, contains a significant amount of
information regarding retrofit of SMFs utiliz-
ing RBS connection modifications. It is rec-
ommended that designers using an RBS
approach to retrofit an existing SMF refer to
the AISC document prior to utilizing the
design methodology contained herein.
Upon selection of the beam-column com-
bination to be utilized in the SMF design and
the location, shape and size of the RBS, fur-
ther connection design checks are required
to ensure the design will perform in a ductile
manner.
The first check should be the "Strong Col-
umn-Weak Beam" confirmation. This check
is intended to limit inelastic deformations of
columns outside of their panel zone regions.
It is generally recognized that column yield-
ing is an undesirable mode because of the
possible effect on the column, and in turn,
the global stability of the structural frame.
The AISC Seismic Design Provisions (1997)
outline an acceptable design level for the
beam/column relationship. As a minimum,
this AISC proviso should be met.
RBS connection design must also address
the panel zone. The panel zone is subjected
to large shear forces as the beams reach their
full capacity. Based on FEMA 267A (1997),
the panel zone must be strong enough to
develop at least 80% of the shears associated
with Mfl The panel zone requirements can be
met in one of two ways. One way is to provide
a column with a thick enough web to resist
the required shear in accordance with the
9

design requirements. The other way to sup-
ply sufficient panel zone shear resistance is
to add doubler plates to the selected section.
Doubler plates should consist of the required
additional thickness of steel, added to one or
both sides of the column web. Fabricators
indicate that the use of a heavier column sec-
tion, instead of doubler plates and other
labor intensive reinforcing details, may result
in a more economical structural frame.
The final design check to be performed on
the selected beam-column combination is the
beam shear. The maximum beam shear is
developed in the section of the beam between
the RBS and the column flange face, where
gravity shear and seismic shear coincide. At
this location, shear capacity of the beam sec-
tion needs to be checked to ensure that the
beam will have adequate shear capacity after
the plastic hinge in the beam develops due to
applied lateral loads.
The following step-by-step presentation
outlines the RBS design procedure relating to
the removal of the beam flange and the
checks required to ensure proper behavior
and correlation with test and research
results.
4.3 Step-by-step Procedure
STEP 2 Compute the plastic section modu-
lus at the minimum section of the
RBS.
Figure 4.4 shows a cross-section of the
beam at the minimum section of the RBS.
b~
"~'~"""''~P~ions cutfromflange
d/2 ~ ~ tw
PlasticNeutralAxis
d/2
/./.~Portions cutfromflange
/
_ __ ~ ,~,'~t
~ ~.~
c c
Figure 4.4
Beam at Minimum Section of RBS
Based on the dimensions shown in this fig-
ure, ZRB S can be computed as follows:
STEP 1 Choose trial values for RBS dimen-
sions a, b, and c.
The trial values for a and b should be
chosen within the limits of Equations 1 and
2. To establish a trial value of c, a flange
reduction of about 40 percent is suggested
for the initial design iteration. Thus, choose c
~ 0.20 bf As noted earlier, values for c in
excess of approximately 0.25bf are not rec-
ommended.
a (O.Sto 0.75) bf
b ~ (0. 65 to O.85) d
10
Z~s = Z b - 2 c t.f (d - t.f ) (3)
Where:
ZRB S = plastic section modulus at min-
imum section of RBS
(1)
= plastic section modulus for full
beam cross-section
(i.e. without flange cutouts)
other variables as shown in Figure 4.4.

STEP 3 Establish the expected yield stress
of the beam.
The expected yield stress for the beam
can be determined from Section 6.2 of the
Buildings (1997). According to these provi-
sions:
Fye = Ry Fy (4)
where:
Fye = expected yield stress
= minimum specified yield stress
= ratio of expected to minimum
specified yield stress
= 1.5 for A36 steel
The factor of 1.15 in Equation 5 accounts
for strain hardening, and is based on strain
hardening vaiaes measured in RBS tests.
STEP 5 Compute the shear force at the
center of the RBS cuts at each end
of the beam.
The shear at the center of the RBS can be
computed from a free body diagram of the
moment frame beam taken between RBS
centers. Such a free body diagram is illus-
trated in Figure 4.5 for the case of a uni-
formly distributed gravity load w.
f
R~BS RBS
I w = uniform beam gravity ~oad •
l!.~.,~ ~ ~ t ~ I t t t t I t t I t ~ ~ I t I I I t t ~ t.!..!,{
. . . . . .
RBSRBS! i RBS RBS
i L' = distance between centers of RBS ' -I
Figure 4.5
Free Body Diagram Between
Centers of RBS
= 1.1 for A572 Gr. 50 and A992
steel
The value of Fve recognizes that the
actual yield strengtl~of structural steel can
significantly exceed the minimum specified
value.
Summing moments about each end of
this free body diagram results in the follow-
ing:
2MRBs wL'
V~S - L ' + -~- (6a)
STEP 4 Compute the maximum moment
expected at the center of the RBS.
MRBS = 1.15 ZRBS Fye (5)
2 MRBs wL'
V~O~S - L' 2 (6b)
where: where:
MRBS =
ZRBS =
maximum moment expected at
the center of the RBS
plastic section modulus at min-
imum section of the RBS
expected yield stress of beam
VRBS V' BS = shear force at the center
of the RBS at each end
of beam
L' = distance between centers of
RBS
W = uniformly distributed gravity
load on beam
11

For gravity load conditions other than a
uniform load, the appropriate adjustment
can easily be made to the free body diagram
and to Equations 6a and 6b.
Equations 6a and 6b assume that plastic
hinges will form at the RBS at each end of the
beam. If the gravity load on the beam is very
large, the plastic hinge at one end of the
beam may move toward the interior portion
of the beam span. If this is the case, the free
body diagram in Figure 4.5 should be modi-
fied to extend between the actual plastic
hinge locations. To check if Equations 6a and
6b are valid, draw the moment diagram for
the segment of the beam shown in Figure
4.5, i.e., for the segment of the beam between
the centers of the RBS cuts. If the maximum
moment occurs at the ends of the spans,
then Equations 6a and 6b are valid. If the
maximum moment occurs within the span,
and exceeds Mp.e of the beam (see Equation
8), then the modification described above will
be needed.
expected at the face of the column.
M f = Mp,Bs + VRBs a +
where:
(7)
= maximum moment expected at
the face of the column
allother variables as previous~
defined
Equation 7 neglects the gravity load on
the portion of the beam between the center of
the RBS and the face of the column. This
simplifies the equation and introduces little
error. If desired, the gravity load on this
small portion of the beam can be included in
the free body diagram and in Equation 7.
STEP 7 Compute the plastic moment of the
beam based on the expected yield
stress.
Mpe = Zb Fy e (8)
The moment at the face of the column can
be computed from a free body diagram of the
segment of the beam between the center of
the RBS and the face of the column flange.
Such a free body diagram is illustrated in
Figure 4.6.
RBS
- - Mf ....."~". VRBs MRBs
~ ,
I- b
---N a +.-Z-
Figure 4.6
Free Body Diagram Between Center of
RBS and Face of Column Flange
Summing moments about the left end of
this free body diagram results in the follow-
ing:
where:
Mpe = plastic moment of beam based
on expected yield stress.
STEP 8 Check that Mfis in the range of 85
to 100 percent of Mpe.
M.f ~0.85 to 1.0 (9)
m pe
If Equation 9 is not satisfied, modify the
values of c and/or a and b as needed, and
repeat Steps 2 through 8. Note that this
check on moment at the face of the column is
simplified for design purposes, based on
more detailed analyses and past test results.
The actual force transfer mechanism and
state of stress and strain at this location is
quite complex due to the constraint gener-
ated by the connection to the column flange.
For more detailed information on the issue,
the reader is referred to (Lee, et.al. 1997).
12

STEP 9 Strong Column-Weak Beam Check
To check strong column-weak beam Z Mc
requirements, the procedure presented in
FEMA 267A (1997) will be used, with minor Where:
modifications. The equation to be used to
check this requirement (from Equation Vc =
7.5.2.5-1 of FEMA 267A (1997)) is as follows:
= Mct+ Me b (14)
shear force in the columns
above and below the connection
~ Z¢(F~c- J~) > 1.0 (10) Mct
ZMc
= column moment
above connection
immediately
where: Mcb = column moment immediately
below connection
plastic section modulus of the
column section above and
below the connection
ht distance from top of beam to
point of inflection in the col-
umn above the connection
YMc = minimum specified yield stress
of the column
= axial stress in the column
above and below the connection
~VMc sum of the column moments at
the top and bottom of the panel
zone corresponding to the
development of MRB S at the
center of the RBS in the
attached beams
Figure 4.7 shows a free body diagram that
can be used to estimate column moments
when checking Equation 10. This free body
cuts the beams at the RBS centers and cuts
the columns at assumed points of inflection
(often taken as mid-height of the adjacent
stories for design purposes).
Based on Figure 4.7, £'Mc can be esti-
mated from the following equations:
, ,(de _,~
Z M R~s + (VR~s + V~s)~- + a +
2J
V~ : (11)
h t + d b + h b
Mct = Vcht (12)
Mcb = Vchb (13)
d c = depth of column
hb distance from bottom of beam
to point of inflection in the col-
umn below the connection
d b = depth of beam
All other variables as previously defined.
Mct
~ -,,~-.-.~V~i C
i Mcb
I
l
I I
a+(b/2) dc a+(b/2)
Figure 4.7
~ MRBS
V
RBS
Free Body Diagram for
Calculation of Column Moments
ht
db
hb
13

The approach presented in FEMA 267A
(1997) accounts for the difference in column
shear forces above and below the connection,
whereas the simplified approach above
assumes the same shear force is present in
the columns above and below the connec-
tion. Although the approach in FEMA 267A
(1997) may be somewhat more accurate, the
computation of Vc presented in Equation 11
above is simpler to implement, and is still
reasonably accurate for initial design pur-
poses considering the numerous uncertain-
ties involved in the strong column-weak
beam design philosophy. The reader is
referred to Section 7.5.2.5 of FEMA 267A
(1997) to implement a more accurate calcu-
lation for Vc to be used in the final design
check.
STEP 10 Check Panel Zone
To check the column panel zone, the pro-
cedure used in Section 6.6.6.3.7 of FEMA
267A (1997) will be used. This section
requires that the panel zone have sufficient
strength to develop the shear force developed
by 0.8 £'M/: Based on this approach, the
panel zone'shear force can be computed as
follows:
M? = maximum moment expected at
opposite column face
The value of My computed according to
Equation 7 combines the, seismic moment
due to (2XMRBs)/L' with the moment due to
gravity load. On the side of the column oppo-
site to that where My is developed, the
moment at the face of" the column will be
somewhat smaller since the gravity load
moment will oppose the seismic moment.
This somewhat smaller moment is calculated
using Equation 17.
The strength of the panel zone can be cal-
culated as follows:
3bct~
V = 0.55Fycdct 1 + dbdc--~~ (18)
where:
V = panel zone shear strength
M'f = M~S + V~S a + (15)
•Mf= Mf+ M~r (16)
o.8Z
Vpz - 0.8Vc (17)
0.95 db
Where:
bc = width of column flange
tcf = thickness of column flange
= total thickness of panel zone
including doubler plates
STEP 11 Check Beam Shear
Vpz panel zone shear force corre-
sponding to the development of
80 percent of the maximum
expected column face moments
maximum moment expected at
the face of the column, calcu-
lated according to Equation 7
The final design check should be made to
ensure that the beam has adequate capacity
for shear asssociated with lateral and gravity
loads. This check combines the beam shear
associated with the plastic moment within
the RBS using Equation 6a, combined with
the portion of gravity load adding shear to
the beam within the section between the RBS
14

center and the column flange. This can be
calculated using Equation 19:
VRBs q
(/-/,)
W - -
2 (19)
2
4.4 Additional Design Considera-
tions
In addition to establishing the dimensions of
the RBS cut, there are a number of addi-
tional design and detailing features that may
significantly affect connection performance
and economy of this system. These items are
discussed below.
The procedure presented above for sizing
the RBS cut permits a range of acceptable
values for the dimensions a, b and c. Fabri-
cation can likely be simplified by standardiz-
ing these dimensions over a large number of
beams on a project. Making small changes
on the RBS dimensions from beam to beam
is not likely to improve connection perform-
ance and may unnecessarily increase fabri-
cation costs. The designer may wish to con-
sult with a fabricator before finalizing the
RBS dimensions to identify ways of reducing
fabrication costs. For example, if the fabrica-
tor is making RBS cuts using a torch
mounted on a guide with a fixed radius, the
economy of the connection may be improved
by maintaining a constant radius of cut R
over a large number of connections.
The RBS cut is normally made by thermal
cutting in the fabrication shop. The cut
should be made to avoid nicks, gouges, and
other discontinuities. After the cut is made,
the surface should be ground, to aid in
reducing the potential for fractures occurring
in the RBS at high plastic rotations and low
cycle fatigue. The grinding should be done to
avoid producing grind marks perpendicular
to the beam flange, since they are perpendi-
cular to the direction of principal stress.
These marks can act as stress risers. Varia-
tions on grinding methods may be possible to
reduce fabrication effort.
Another consideration for design of RBS
moment connections is welding. Research
conducted since the Northridge earthquake
has demonstrated the importance of weld
metal toughness in the groove welds of seis-
mic resistant moment connections (Kauf-
mann, et.al. 1996; Tide 1998 I. The AISC Seis-
mic Provisions (1997) recommends the use of
a filler metal with a minimum specified ten-
sile strength of 70 ksi, (assuming a 50 ksi
base material specified yield) and a minimum
specified CVN value of 20 ft.-lb, at -20 ° F.
Previous research tests on RBS connections
have generally employed the self-shielded
flux cored arc welding process (FCAW), using
E70TG-K2, E71T-8 or E70T-6 electrodes. All
of these electrodes provide a minimum spec-
ified CVN of 20 ft.-lb, at -20 ° F. A number of
other FCAW electrodes are available that pro-
vide this minimum CVN value. In addition,
successful tests on other types of connec-
tions have employed the shielded metal arc
welding {SMAW) process using an E7018
electrode. The final choice of welding process
and electrode is best left to the fabricator.
Other factors, such as the mixing of different
filler metals in the same weld joint may result
in lower CVN values for the combination,
than for one of the filler metals alone. A
paper written on this subject, "The Effects of
Intermixed Weld Metal on Mechanical Prop-
erties" (Johnson, Quintana 1998), may be
useful to the engineer when considering the
inter-mixing of weld filler metals.
At the beam flange complete joint pene-
tration welds, it is recommended that the
weld run-off tabs be removed at both the top
and bottom flanges, and that the edges of the
groove welds be ground smooth. The pre-
ferred final profile of the weld tab ground
surface is radiused, to further reduce the
possibility of fracture at these locations. This
will minimize any potential notches intro-
duced by the presence of the weld tabs, or by
discontinuities contained in the weld metal
within the run-off regions. In addition, it is
recommended that the bottom flange steel
backing be removed and a reinforcing fillet be
placed at the base of the weld after the joint
is backgouged to sound metal. This require-
ment is intended both to eliminate the notch
effect produced by the steel backing, and to
permit better inspection and ultrasonic test-
ing of the weld. At the top flange groove weld,
15

it is recommended that the steel backing be
seal welded to the face of the column using a
minimum size fillet weld, typically a 5/16" fil-
let. Analysis has indicated that the notch
effect of the steel backing is not as severe at
the top flange, and that welding the steel
backing to the column further reduces the
notch effect. Further, defects are less likely
at the top flange weld since the groove weld
is not interrupted by the beam web, as it is at
the bottom flange.
Many researchers and designers believe
that the weld access hole has an important
effect on connection performance. Although
current research is addressing issues related
to the weld access hole, there appears to be
no consensus as of yet on the optimum size
and shape. Consequently, pending further
research, access hole geometry should con-
form to the requirements shown in Figure 5.2
of AWS D1.1-98 (AWS 1998). There is no
indication that weld access hole size, within
the AWS limits, will adversely affect the per-
formance of RBS moment connections.
Therefore, size and shape of the access hole
should be left to the fabricator to conform to
AWS recommendations.
Another important aspect of well-behaved
moment connections are the continuity
plates between the column flanges. All of the
successful tests on RBS connections for new
construction (Appendix A) have employed
continuity plates. However, no RBS tests to
date have omitted continuity plates, so it is
unclear under what conditions continuity
plates are actually required. Pending the out-
come of further research, it is recommended
that continuity plates be provided for all RBS
connections, with a continuity plate thick-
ness similar to the beam flange thickness.
Welds that attach a continuity plate to the
column flange or web, should be made with
an electrode with a rated CVN of at least 20
ft.-lb, at -20 ° F. Based on experimental
results, removal of backing bars from conti-
nuity plate welds, however, does not appear
to be necessary. When welding the continuity
plates to the column, welding in the "k-area"
of the column should be avoided (AISC
1997}.
All welding should be specified to be in
conformance with the latest edition of AWS
D 1.1. Acceptance criteria for ultrasonic test-
ing of groove welds is recommended to be in
conformance with Table 5.2 of AWS D 1.1-98.
Additional useful information on welding
moment connections can be found in a num-
ber of references listed at the end of this doc-
ument.
Recent tests have shown that RBS con-
nections with bolted web details can meet the
recommended plastic rotation demands of
FEMA 267 (1995). However, it should be
noted that at large rotation demands, the
bolted detail appears to be more susceptible
to fracture initiating near the weld access
hole. This issue is the subject of further SAC
sponsored research. Until more definitive
guidance is provided in the upcoming SAC
Guidelines, the engineer should carefully
consider required connection and SMF per-
formance when choosing a beam web con-
nection.
The majority of the welded web connec-
tion tests have utilized a complete joint pen-
etration (CJP) groove weld between the beam
web and column flange over the full depth of
the web. The shear tab, which is welded to
the column and bolted to the beam web, is
still provided. This shear tab serves several
purposes. First, it acts as backing for the
CJP groove weld. Second, it carries erection
loads and helps maintain the frame in a
plumb position until welding at the connec-
tion is completed. Since the shear tab is pro-
vided for erection purposes only, it is recom-
mended that the design of the shear tab be
left to the fabricator. However, to ensure that
the shear tab does not resist loads in the
event that excessive plastic rotations cause
the web connection to fracture, the designer
could consider indicating that the shear tab
be fabricated with short horizontal slotted
holes.
Traditionally the shear tab would be
welded on both sides. However, when utiliz-
ing a web CJP weld, the "~backside" fillet weld
may pose potential filler metal mixing and fit
up problems. The engineer should work with
the fabricator to generate an acceptable
welding sequence. As an alternative to a CJP
groove weld, the beam web connection can
also be made using a heavy fillet welded
shear tab. The shear tab is typically welded
16

to the column using either fillet welds or a
CJP groove weld. The shear tab, in turn, is
then welded to the beam web with fillet
welds. An example of such a connection can
be found in "Moment Frame Connection
Development and Testing for the City of Hope
National Medical Center" (Zekioglu, et.al.
1997).
If the engineer chooses to use a bolted
web connection, all aspects of the connection
should be designed to resist the full shear
applied to the beam due to gravity and earth-
quake loads. Short slotted holes may be uti-
lized to futher protect the shear tab and
beam web from possz'bie excesive deflections
when the connection in subjected to large
rotations as the system undergoes inelastic
action during an earthquake. It should be
noted that structural steel erectors prefer
standard holes to slotted holes to aid in erec-
tion.
One of the most discussed aspects of RBS
design, and one of the most important, is the
supplemental lateral bracing required for
this system. FEMA 267A (1997) recommends
that a lateral brace be provided near the
RBS. The following discussion presents an
analysis of test results that did not have lat-
eral bracing provided near the RBS.
Virtually all moment connections that
dissipate energy by yielding of the beam are
subject to varying degrees of beam instability
at large levels of inelastic rotation. This is
true both for reinforced connections (cover
plates, ribs, haunches, etc.) and for RBS con-
nections. This instability generally involves a
combination of flange buckling, web buckling
and lateral torsional buckling and typically
results in deterioration of the beam flexural
strength, with increasing inelastic rotations.
In the experience of some researchers, the
degree of instability and associated strength
deterioration for RBS connections tested in
the laboratory have been no more severe, and
perhaps somewhat less severe than for many
types of reinforced connections. This is
demonstrated by the connection test results
shown in Figure 4.8.
This figure shows a plot of beam tip load
versus beam tip displacement for two differ-
ent test specimens. These two specimens
were virtually identical, except for the con-
nection detail. Both specimens were con-
structed with the same member sizes
(W36xlS0 beam and W14x426 column) and
heats of steel, and tested in the same test
setup with identical member lengths, identi-
cal member end support conditions, and
identical lateral bracing. Both specimens
were subjected to the same loading history.
The only difference was that one specimen
was constructed with a cover plated connec-
tion and the other with an RBS connection.
Both specimens were provided with a single
beam lateral support near the point of load
application.
250
200
150
100 .
~ 50.
~ o.
.~-~0.
-100,
-150.
-200,
-250
-6
Cover'Pla~ed Connectlon ~.______,~_
-~--~--,~
RBS Connection ] * ~
'~ ~ -
_ _ _
- - - -
~ '~"'~'~'({~:;e~
•II
.~ -2
~. ~.~ :~-~
:°~*°" ~~'°~~
, ,
Displacement (inches)
Figure 4.8
Comparison of Test Results for
Cover Plated and RBS Connections
As can be seen from Figure 4.8, the peak
strength of the RBS connection is less than
that of the cover-plated connection. This, of
course, is expected and is in fact a potential
advantage of the RBS in that it reduces the
moment generated at the connection and the
moment delivered to the column. After reach-
ing their peak strength, both connections
exhibited some strength deterioration due to
combined flange, web and lateral torsional
buckling in the beam. Note however that the
rate of deterioration is less for the RBS spec-
imen. In fact, at large inelastic deformations,
the RBS exhibits the same strength as the
cover-plated connection. This comparison
demonstrates the observation made above,
i.e., RBS connections exhibit no more
strength deterioration, and perhaps some-
what less deterioration than reinforced con-
nections.
17

The test data summarized in Appendix A
indicates that many RBS connection tests
have been conducted without an additional
lateral brace at the RBS. There is no instance
where an investigator reported unusually
severe or unacceptable strength deterioration
due to the absence of a lateral support near
the RBS. Futher, as discussed above,
strength degradation in the RBS is compara-
ble to that seen in many other connection
types for which no additional lateral bracing
is presesntly required. Consequently, based
on currently available data, an additional lat-
eral brace at the RBS does not appear neces-
sary in order to achieve acceptable perform-
ance. However, the designer should still
adhere to the normal code provisions for
beam lateral support and for beam flange
and web slenderness limits. Lateral bracing
for beams in Special Moment Frames should
be provided at a maximum spacing of 2500
/FY, as required by Section 9.8 of the AISC
is~nic Provisions (1997}.
As described earlier, most moment con-
nections show gradual strength degradation
at large levels of plastic roatation due to com-
bined local and lateral torsional buckling of
the beam. This occurs for the RBS as well as
for most other connection types, as illus-
trated in Figure 4.9. Reducing the lateral
support spacing in the region of the plastic
hinge from that required in Section 9.8 of the
AISC Seismic Provisions may therefore
reduce the rate of strength degradation for
most types of moment connections. Further
definitive recommendations and research
results will be provided in the upcoming SAC
Guidelines.
If a designer should choose to provide a
lateral brace at the RBS, the brace should
not be located within the reduced section of
the beam. Welded or bolted brace attache-
ments in this highly strained region of the
beam may serve as fracture initiation sites.
Consequently, if a lateral brace is provided, it
should be located at or beyond the end of the
RBS that is farthest from the face of the col-
umn. If bracing is to be provided as part of
the design, requirements and recommenda-
tions can be gathered from documents such
as FEMA 267A (1997) and "Fundamentals of
Beam Bracing" (Yura 1993).
5 RBS Design Example
Description of Design Example Project
• Commercial Office Building/Medical
Office Building
• Located in San Francisco, California
• Distance from Nearest Earthquake
Fault: ~ 9 kilometers (San Andreas)
• High Seismicity Zone with Near Fault
Characteristics
Description of Design Example Frame
Perimeter Moment Frames
Frame centerline dimensions:
story height = 13' - 0"
bay width = 22' - 8"
Beam: W24x117
A572 Gr. 50 (A992)
Fyb = 50 ksi
Column: W14x311
A572 Gr. 50 (A992)
Fyc = 50 ksi
Gravity load on beam:
(1.2D + .5L per
Sect. 9.2c of AISC Seismic Provisions):
2 kips/ft (0.17 kips/in)
Gravity loads are due to floor tributary
loads as well as exterior wall loads.
Design typical interior moment
connection of perimeter frame.
I~ V l ~
a
R = radius of cut = 4c~+ b~
8c
_1-- I
b
Figure 5.1 RBS Dimensions
18

Section Properties: From Equation 5:
W24x117:
d b = 24.26 in.
bf = 12.80 in.
fw = 0.85 in.
= 0.55 in.
Zxb = 327 in. 3
W14x311:
d c = 17.12 in.
bcf = 16.23 in.
tcf = 2.26 in.
tcw = 1.41 in.
Zxc = 603 in. 3
STEP 1 Choose trial values for RBS dimen-
sions a, b and c
MRB S = 1.15 ZRBS_Fye
= 1 15x218x55
= 13789 in-kip
STEP 5 Compute the shear force at the
centers of the RBS at each end of
the beam
L'=L-dc-2 a+ =272-17.12-2 7+ =222in.
From Equations 6a and 6b:
2Me~s wL' 2×13789 0.17x222
Vm~s - - - + - ~ =143kips
L' 2 222 2
a -~'(0.5 to 0.75) bf
~6 in. to 10 in.
Try: a = 7 in.
b ~(0.65 to 0.85) d b
~ 16 in. to 21 in.
Try: b = 19 in.
c ~0.2 bf
~2.6 in.
Try: c = 2.75 in.
STEP 2 Compute the plastic section modu-
lus at the minimum section of the
RBS
From Equation 3:
ZRBS = Zxb- 2 ctf(d b-t~
= 327 - 2 x 2.75 x 0.85 x (24.26 - 0.85)
= 218 in.3
STEP 3 Establish the expected yield stress
of the beam
For A572 Gr. 50 steel, Ry = 1.1.
From Equation 4:
V~s _ 2M~s wL'_ 2×13789 0.17×222 =105kips
L' 2 222 2
Figure 5.2 shows the shear force diagram,
the bending moment diagram, and the free
body diagram the for the portion of the beam
between RBS centers. Observe that the max-
imum moment occurs at the ends, i.e., at the
centers of the RBS. If the gravity load were
extremely large, compared to the moment
143 105
V (kip)
M (kip-in)
13789
-13789
Fy e = RyFy b = 1.1x50 = 55ksi
expected at the center of the RBS
~ REDS w= 0.17 kips/in. ~ RIBS
Ii.,.l..i ~ I i ~ I i I I I I ~ t I t i i I I I t I ~ I i i.l..!j
. . . . . . . . t J143 ' "~05k~
,
L' ~ 222 in.
Figure 5.2
Portion of Example Beam
between RBS Centers
19

,
developed due to applied lateral loads, the
curved portion of the moment diagram could
drive the plastic hinge toward the column,
away from the RBS. This example indicates
that the gravity load is not large enough to
form a plastic hinge within the span, away
from the RBS. Consequently, the calcula-
tions above for the moment and shear forces,
at the RBS cuts, are valid.
expected at the face of the column
Ms
From Equation 7:
=Mees + Veas(a+2b-/=13789+ 143(7+~) = 16149in-kip
STEP 7 Compute the plastic moment of
the beam based on the expected
yield stress
From Equation 8:
Mpe = Zxb Fye = 327 x 55 = 17985 in-kip
STEP 8 Check that Mfis in the range of 85
to 100 percent of Mpe
From Equation 9:
ZMc
> 1.0 (Equation 10)
Returning to the example, assuming that
points of inflection in the columns occur at
their mid-heights, and assuming an axial
stress (fa) of 15 ksi in the columns under
combined earthquake and gravity loading,
the following calculations result.
From Equations 11, 12, 13 and 14:
h~+ db+ hb
2x13789+ (143+ 105(17;12 + 7 + ~)
156
= 217kips
Met
Mcb
= Vc ht
= 217 x (156 - 24.26)/2
= 14294 in-kip
14294 in-kip
= 2x14294 = 28588 in -kip
Mf 16149- -
Mpe 17985
- - - 0.90 OK
Thus, the preliminary dimensions are OK.
Use: a = 7in.
b = 19in.
c = 2.75 in.
STEP 9 Strong Column-Weak Beam Check
To check strong column-weak beam
requirements, the procedure presented in
FEMA 267A (1997) will be used, with the
minor modifications noted in Section 4. The
final equation to be used to check this
requirement (from Equation 7.5.2.5-1 of
FEMA 267A) is as follows:
~Zc(Fyc-.f~) 2×603(50-15)
- = 1.5 > 1 OK
~M~ 28588
STEP 10 Check Column Panel Zone
To check the column panel zone, the pro-
cedure discussed in Section 4 will be used.
Based on the example, the column panel
zone shear is computed as follows:
Mf = 16149 in-kip (Equation 7)
From Equations 15, 16 and 17:
27Mf = Mf+ M:f
= 16149 + 15522 = 31674 in-kip
,
Mf=M~Bs+V~Bsa+ =13789+1057+ =15522in-kip
i|1
20

Vez- 0.8z..,~'Mr 0.8Vc Vc- 0.8x31671 0.8x217=926kips
0.95dt) 0.95×24.26
Panel zone strength is computed as fol-
lows:
From Equation 18:
=0.55F~,~d~tIlL+3b~ft~d+d~t1
I 3x16"23x(2"26)~]
= 0.55xSOx17.12x1.41 1+ 24.26xlT.12xl.41J= 946kips
946 > 926 .'.No doubler plates required
STEP 11 Check Beam Shear
From Equation 19:
w (l-l') /272~222/
V~ 4 2 0.17 -
' 2 143 ÷
2
= 145kips
V, = A,,,Fy = (0.55)(24.26)(5 O) = 667 kips > 145 kips
RBS flange reduction is approximately 43
percent. Consequently, it is expected that the
inclusion of tlae RBS the beams will increase
interstory drift by about 5 percent.
S~e~c Abut
,~
~ .B.U.barto remain
I / ~ ~Remove weldtabs
IE 718"x 6" ~,.~,,.T-~-~'~"r-.~/~ IP
{B.S.) ~ I ! [ I / _1 16 ~WeldB.U.barIocoiutnn
•~l~.l I~ .~_5 .° - - N ....
~ l / *~
I.t' i Iw2,.,,7
~i I.II'i Ig;-~'~-------------~,~,:,,d~,,,~,,~,,oo,~d,~, tose~v asbac i~g C~,~ -- ~
IZ ....ooo,0,.to,.]~,~ ,_~ ~ ~ ,~,~.~ : ~ columnand beam byfabdcato~.
II 5/16 cleanedand inspected
.
Configure platecomes to ~ 17 75" Radius
=.o,o0,...... /.of column GrindSmooth
~ ~ J ~ 1 ~ 2.75"7.3"
2.75"
5/'I'~ NI welds: ET0
~lI groovewelds: electrodes must be rat~;Ifor
'° CVNof atteast20It-fosat -20deg.F.
Allweldingshallconformto AWS D1.1
Figure 5.3 Connection Detail for
Design E~mple
6 Procedures for Acceptance
of Design by Building
Authorities
Continuity Plates
Use continuity plates with a thickness
approximately equal to the beam flange
thickness. The beam flange thickness is 0.85
inches. Therefore, use 7/8" thick continuity
plates (0.875"). Connect continuity plates to
column flanges using CJP groove welds, and
the web using double fillet welds. The cor-
ners of continuity plates should be config-
ured to avoid welding into the k-area of the
column.
Beam Web Connection
Connect beam web to column flange
using CJP groove weld over full depth of web
(between weld access holes).
A drawing of a generic final connection
detail is shown in Figure 5.3. The resulting
frame should be checked for all code speci-
fied strength and drift limits. Note that the
The design of SMF building systems require
that the design account for inelastic defor-
mation demands on the connection. The
Buildings (1997), Section 9.2, presents the
requirements for SMF structures. The RBS
connection is an option that can meet
requirements set by building codes and con-
sensus documents. The following comments
are intended to describe actions that can be
followed to help facilitate the permitting
process for a SMF building system.
6.1 Communication
It is recommended that early in the process,
the Structural Engineer of Record communi-
cate with the building official regarding the
proposed use and pertinent aspects of the
RBS moment connection. The engineer may
need to provide background documentation
to the building official if he or she is unfamil-
iar with the design and terminology relating
21

to the design. The use of this document may
aid the building official in understanding the
design intent.
6.2 Methodology
Once the building official understands the
design intent and system behavior, it is
important to clearly state the design method-
ology to be used early so that any misunder-
standings can be avoided. This document
presents a general design methodology, uti-
lizing some simplifying assumptions and
some of the better aspects of many different
design methods. There are other ways to
design an RBS moment connection and SMF
system than that represented in this docu-
ment. If other methods are utilized, the engi-
neer should be sure to clearly indicate the
method used and the important aspects that
show design compliance with the governing
building code.
Any design methodology utilized should
correlate well with other published methods,
test results and research papers. Section 9.2
of the AISC Seismic Provisions require that
the design be based on qualifying cyclic tests.
The table in Appendix A will help to satisfy
this requirement for the RBS connection.
Any significant deviation from established
methodologies or tests should be justified. It
is important to understand that many rec-
ommendations contained in this document
are based on experimental research. Design
equations and RBS sizing values are based
on successful research, both analytically and
experimentally. Therefore, any new design
equations should be comparable to estab-
lished equations.
6.3 Construction Documents
After a design is complete, it is imperative to
convey the information accurately on con-
struction documents. While calculations are
important and describe the final constructed
connection, construction documents provide
direction to the fabricator and erector. The
elements expressed on the drawings will be
more important to the final quality of the
design than any calculation.
The documentation related to the RBS
connection should be clear and concise, yet
provide enough detail for the fabricator to
properly incorporate all the difficult and
important aspects of the connection. The
information should be such that any fabrica-
tor or erector can utilize the information pro-
vided, and construct the final connection in
such a manner that the performance will
directly correlate with the design intent.
Important aspects of the design to be
included in the drawing details are welding
details, RBS shape and location, notes
regarding grinding of the RBS after cutting,
shear tab detail information and beam web to
column flange connection details. It is rec-
ommended to provide a set of notes specific
to the RBS connections, relating to welding
practices and connection construction proce-
dures to help the contractor understand the
connection and the importance it has on the
building system performance. Reference to
applicable portions of AWS D I.1 and other
AWS or AISC documents should be included
in these notes to clearly state a level of
expected quality. This level of information
will also facilitate obtaining the appropriate
level of inspection required for this type of
connection.
7 Fabrication and Inspection
Issues
A number of fabrication and inspection
issues are important to ensure a well-con-
structed RBS connection. As discussed ear-
lier proper fabrication and erection of this
connection is a critical portion of the sys-
tem's performance. If welds are poorly
placed, the stress at which fracture initiates
and propagates is much lower than the
stress a tough weld metal, placed with care,
can resist. Cutting and grinding are critical
aspects of fabrication which must be well
executed to produce a high quality final con-
nection.
7.1 Cutting and Grinding
The cut portion of both the curved RBS sec-
tion, as well as the preparation of the end of
22

•
the beam, needs to be smooth and free of
notches. This smoothness is important for
reasons discussed earlier. Many fabrication
shops have the ability to make virtually
notch free thermal cuts. While this is a ben-
efit to reduce the number of perpendicular
notches, which may present stress risers,
small imperfections exist that may affect con-
nection performance.
Therefore, it is important to clearly iden-
tify what is the adequate amount of material
to remove (by grinding) from the cut surface.
FEMA 267A (1997) discusses a level of
acceptable surface roughness value less than
or equal to 1000 as defined in ANSI/ASME
B46.1. This level is difficult to determine
without a significant amount of equipment
and expertise. Therefore, this document rec-
ommends that the thermal cuts be ground
smooth in the following manner: "It is impor-
tant that the pattern of any cuts made in the
flange be proportioned so as to avoid sharp
cut corners. All comers should be rounded to
minimize notch effects and in addition, cut
edges should be cut or ground to have a sur-
face roughness meeting the requirements of
AWS C4.1-77 class 4, or smoother."
The designer should discuss the intent
with the fabricator and develop criteria for an
acceptable mock-up to be made for reference
during fabrication inspections. The final
grinding that the engineer and fabricator
have agreed upon, should be inspected by
the fabricator's representative as well as the
owner's testing agency, to ensure compliance
with the accepted mock-up.
Many beams used for SMF systems are
large with thick flanges and webs. Shear
punching holes in these thick portions of the
member could lead to localized delamination
or tearing. In situations where hole diame-
ters are smaller than the base material
thickness, the designer may consider that
holes required for fabrication of elements and
portions of the RBS beam be drilled rather
than punched. No research results indicate
that a reduction in connection performance
is attributable to punching holes in RBS
beams.
7.2 Welding
Welding is a very critical part of the proper
fabrication of this connection. A significant
amount of effort has been made to produce a
beam with a reduced section modulus,
designed to yield prior to developing
moments which deliver very high stresses to
beam flange - column flange welds. However,
if the welding required for this connection is
done poorly, the stress at which brittle
behavior may occur is much lower than the
engineer expects. Good welds, using tough
filler metal, will resist higher loads than sur-
rounding base metal. Therefore, it is impera-
tive that the welding for this type of connec-
tion be of high quality, to produce a
connection that will perform as designed.
Any specific issues related to welds, such
as weld profiles, sequence, submittal of
materials or certifications that are consid-
ered important for compliance of the fabrica-
tor's work to meet the design intent, should
be clearly stated in the construction docu-
ments. Items such as preheat should be
addressed in the project specifications and
construction drawings. Typically, AWS will
adequately address most issues, and for new
design will provide the fabricator ample
direction to complete the construction in a
safe and high quality manner.
The engineer should be clear in the proj-
ect specifications and construction drawings
that filler metals shall not be mixed in such
a way as to produce a CVN value below that
specified and required for a single filler
metal. Most fabrication shops presently use
gas shielded FCAW methods for welds to
columns and beams. The erection crews,
especially when welding complete joint pene-
tration groove welds, typically use self
shielded FCAW. Also, there are different filler
metals used for the flat position as well as
other positions. Some combinations of filler
metals in the same joint may produce a com-
bined CVN value, which could present "brit-
fie behavior". The engineer should carefully
review the information provided in "The
Effects of Intermixed Weld Metal on Mechan-
ical Properties" (1998) and the submitted
WPS prior to fabrication to ensure that the
fabricator and erector are not creating a
23

potential problem by inappropriately mixing
filler metals.
Parameters should be set for quality con-
trol of shop welding and fabrication. The fab-
ricator must have an acceptable Quality Con-
trol (QC) procedure in place throughout the
fabrication of the project. In addition, Quality
Assurance measures should be taken to help
ensure that the QC procedure is being imple-
mented and followed. Typically QA or Verifi-
cation Inspection is provided by special
inspectors, hired by the owner. It is the
responsibility of the engineer to establish
inspection protocol, request a pre-fabrication
and pre-erection meeting, and impress upon
the fabricator and erector the important
issues surrounding the RBS connection
details and construction. Complete joint pen-
etration groove welds should be inspected by
a Level II qualified NDT inspector as defined
in the AWS D 1.1. Each joint should be ultra-
sonically tested and all welds associated with
the connection should receive continuous
special inspection. Field inspection should be
sensitive to such issues as weld preparation
and fit-up, weld profile and weld pass
sequence, back-up bar removal and grinding
of run-off tabs. The inspectors should
develop an acceptable protocol for inspection
and reports in regards to welding and con-
nection completion.
24

References
"AISC Initiates Research Into k Area Crack-
ing," Modern Steel Construction, Vol. 37,
No. 9, September 1997, pp.23-24.
Grubbs, K.V., "The Effect of the Dogbone
Connection on the Elastic Stiffness of
Steel Moment Frames," M.S. Thesis,
Department of Civil Engineering, the Uni-
versity of Texas at Austin, Austin, Texas,
August 1997.
Blodgett, O., Funderburk, S., and Miller, D.,
"Fabricators' and Erectors' Guide to
Welded Steel Construction," The Lincoln
Electric Company, Cleveland, 1997.
International Conference of Building Officials
(ICBO), The Uniform Building Code
(UBSC), April 1997.
Chen, S.J., Yeh, C.H. and Chu, J.M, "Ductile
Steel Beam-to-Column Connections for
Seismic Resistance," Journal of Structural
Engineering, Vol. 122, No. 11, November
1996, pp. 1292-1299.
Iwankiw, N., "Ultimate Strength Considera-
tions of Seismic Design of the Reduced
Beam Section (Internal Plastic Hinge),"
Engineering Journal, American Institute
of Steel Construction, Inc., Vol. 34, No. 1,
First Quarter 1997.
Engelhardt, M.D. and Husain, A.S., "Cyclic
Loading Performance Of Welded Flange -
Bolted Web Connections," Journal of
Structural Engineering, ASCE, Vol. 119,
No. 12, December 1993.
Engelhardt, M.D., Winneberger, T., Zekany,
A.J. ,and Potyraj, T., ~The Dogbone Con-
nection: Part II." Modern Steel Construc-
tion, August 1996.
Engelhardt, M.D., Winneberger, T., Zekany,
A.J. ,and Potyraj, T., "Experimental Inves-
tigation of Dogbone Moment Connec-
tions," Proceedings: 1997 National Steel
Construction Conference, American Insti-
tute of Steel Construction, Chicago, May
1997.
Johnson, M., Quintana, M., '~The Effects of
Intermixed Weld Metal on Mechanical
Properties, Part III," Proceedings, Interna-
tional Conference on Welded Construc-
tions in Seismic Areas, AWS, October
1998.
Kaufmann, E., Xue, M., Lu, L., and Fisher,
J., "Achieving Ductile Behavior of
Moment Connections," Modern Steel Con-
struction, Vol. 36, No. 1, American Insti-
tute of Steel Construction, January 1996.
Lee, K., Goel, S.C., Stojadinovic, B., "Bound-
ary Effects in Welded Steel Moment Con-
nections," Research Report No. UMCEE
97-20, December 1997.
Engelhardt, M.D. and Sabol, T.A., "Reinforc-
ing of Steel Moment Connections with
Cover Plates: Benefits and Limitations,"
Engineering Structures, Vol. 20, No. 6, pp.
510-520, 1998.
Noel, S. N., "Reduced Beam Section Design
for Seismic Retrofit of Steel Moment
Frame Connections," M.S. Thesis, Divi-
sion of Structural Engineering, University
of California, San Diego, 1997.
Gross, J., Engelhardt, M., Uang, C., Kasai,
K., and Iwankiw, N., "Modification of
Existing Steel Welded Moment Frame
Connections for Seismic Resistance,"
Steel Design Guide Series Twelve, Ameri-
can Institute of Steel Construction, Inc.,
Chicago, 1999.
Plumier, A., "The Dogbone: Back to the
Future," Engineering Journal, American
Institute of Steel Construction, Inc. 2nd
Quarter 1997.
25

Popov, E. and Stephen, R., "Cyclic Loading of
Full Size Steel Connections," Bulletin No.
21, American Iron and Steel Institute,
1972.
SAC Joint Venture, Background Reports on
Metallurgy, Fracture Mechanics, Welding,
Moment Connections and Frame Systems
Behavior, Published by the Federal Emer-
gency Management Agency, Report FEMA
288, 1996.
SAC Joint Venture, Interim Guidelines: Eval-
uation, Repair, Modification and Design of
Welded Steel Moment Frame Structures,
Published by the Federal Emergency
Management Agency, Report FEMA 267,
August 1995.
SAC Joint Venture, Interim Guidelines Advi-
sory No. 1 - Supplement to FEMA 267,
Published by the Federal Emergency
Management Agency, Report FEMA 267A,
March 1997.
Seismic Provisions for Structural Steel Build-
ings, American Institute of Steel Con-
struction, Inc., Chicago, April 15, 1997.
"Structural Welding Code - Steel," AWS D 1.1-
98, American Welding Society, Miami,
1998.
Tide, R., "Stability of Weld Metal Subjected to
Cyclic Static and Seismic Loading," Engi-
neering Structures, Vol. 20, Nos. 4-6,
April-June 1998.
Tsal, K.C. and Popov, E.P., "Steel Beam-Col-
umn Joints In Seismic Moment Resisting
Frames", Report No. UCB/EERC - 88/19,
Earthquake Engineering Research Cen-
ter, University of California at Berkeley,
1988.
Yura, J.A., "Fundamentals of Beam Bracing,"
Proceedings, Structural Stability Research
Council Conference, "Is Your Structure
Suitably Braced?," 1993.
Zekioglu, A., Mozaffarian, H. and Uang, C.,
"Moment Frame Connection Development
and Testing for the City of Hope National
Medical Center," Proceedings; Structures
Congress XV, Portland, April 13-16, 1997,
American Society of Civil Engineers,
1997.
26

APPENDIX A
Summary of Experiments on Reduced Beam Section Moment Connections for New Construction
Ref
[1]
[1]
[1]
[1]
[1]
Spec.
YC-1
YC-2
PC-1
PC-2
PC-3
Beam
Built-up W shape
d=23.6", b~=l1.8",
tf=0.79", tw=0.47"
Lb=73"
A36 steel
Fy_f=40 ksi
Fo.~=66 ksi
Fy.w=40 ksi
Fu.w=65 ksi
Column
Built-up Box:
19.7"xl 9.7"x.79"
Lc = 87"
A572 Gr. 50
Fy=56 ksi
Fu=82 ksi
Flange Welds
SS-FCAW
E70T-7
No weld tabs used
Web
Connection
Bolted:
7-7/8" A325
RBS Details
and Other
Flange
Modifications
Tapered cut
L1=2"
LRBS=I3.8"
FR=20%
Tapered cut
L~=2"
LRBS=17.7"
FR=25%
Tapered cut
L1=4.7"
LRBS=I5.7"
FR=34%
Tapered cut
L1=4.7"
LRSS= 17.7"
FR=42%
Tapered cut
L1=4.7"
LRss=I7.7"
FR=42%
Op
(%)
2.4
2.9
4.1
4.8
3.8
Comments
Fracture of beam
flange initiating at weld
access hole
Fracture of beam
access hole
Fracture of beam
access hole
Fracture of beam
access hole
Fracture of beam
access hole
I m~

Ref
[2]
[2]
[2]
[2]
[3,4]
[3,4]
Spec.
DBT-
1A-99-
176
Beam
W30x99
A572 Gr. 50
L~=138"
Column
W14x176
A572 Gr. 50
Lc=168"
Flange Welds
SS-FCAW
E70TG-K2;
backing bar removed
Web
Connection
Bolted:
7-1" A325
RBS Details
and Other
Flange
Modifications
Tapered cut
L1=7.5"
LRBS=20.25''
DBT-
1B-99-
176
DBT-
2A-150-
257
DBT-
2B-150-
257
ARUP-
1
Fy.w= 61.6 ksi
Fu.w= 82.8 ksi
W30x99
A572 Gr. 50
Lb=138"
Fy.w= 51.5 ksi
Fu.w= 72.1 ksi
W36x150
A572 Gr. 50
Lb=138"
F~.w= 60.2 ksi
Fu.w= 72.3 ksi
W36x150
A572 Gr. 50
Lb=138"
Fy.w= 62.9 ksi
Fu.w= 83.1 ksi
W36x150
A572 Gr. 50
Lb=132"
Fy.w=55.6 ksi
Fu.w=70.7 ksi
W14x176
A572 Gr. 50
Lc=168"
Fy.w=55.5 ksi
Fu.w=71.8 ksi
W14x257
A572 Gr. 50
Lc=168"
Fy.w=59.6 ksi
Fu.w=75.2 ksi
W 14x257
A572 Gr. 50
Lc=168"
Fy.w=64.5 ksi
Fu.w=83.2 ksi
W 14x426
A572 Gr. 50
Lc=136"
at bottom flange
SS-FCAW
E70TG-K2
backing bar left in
Bolted:
9-1" A325
welded
(heavy shear
tab groove
FR=45%
Tapered cut
L1=7.5"
LRBS=20.25"
FR=45%
Tapered cut
L1=9"
LaBs=24"
FR=45%
Tapered cut
L1=9"
LRBS=24''
FR=45%
Tapered cut
L1=9"
LABS=24"
COH-1
Fy.f=55.5 ksi
Fu4=73 ksi
Fy.w=62.5 ksi
Fu-w=77 ksi
W27x178
A572 Gr. 50
Lb=132"
Fy.f=44 ksi
Fu.f=62 ksi
Fy.w=46 ksi
Fu-w=62 ksi
W 14x455
A572 Gr. 50
Lc=136"
Fy.f=55 ksi
Fu4=84 ksi
Fy.w=54 ksi
Fu-w=86 ksi
place w/seal weld at
top flange;
backing bar removed
at bottom flange
welded to
column and
fillet welded
to beam
web)
FR=44%
top & bottom
flanges
reinforced with
vertical ribs
Tapered cut
L~=7"
LABS=20"
FR=38%
top & bottom
flanges
reinforced with
vertical ribs
0p
(%)
2.8
4.0
Comments
no failure; test stopped
due to limitations in
test setup
no failure; test stopped
due to limitations in
test setup
3.5 ' Fracture of beam top
flange near groove
we d
1.7 Fracture of beam top
flange we d;
propagated to divot-
type fracture of
column flange
3.5 Flange fracture at
minimum section of
RBS
3.5
A-2

[3,4]
[3,4]
[3,4]
[3,4]
COH-4
~¢ =~
COH-5
|~
[5,6]
[5,6]
Spec. Beam Column Flange Welds Web
Connection
RBS Details
and Other
Flange
Modifications
COH-2
(~ =¢ ~
COH-3 Wl 4x455
A572 Gr. 50
Lc=136"
Fy.f=55 ksi
Fu.f=84 ksi
Fyow=54 ksi
Fu-w=86 ksi
Beam connected to
column web
W33x152
A572 Gr. 50
Lb=132"
Fy.f=57.6 ksi
Fu.f=78.5 ksi
Fy.w=62 ksi
Fu-w=84.5 ksi
Tapered cut
L1=9"
LRBS=26"
FR=43%
top & bottom
flanges
reinforced with
vertical side
plates
Ref
DB1 Wl 4x426
A572 Gr. 50
Lc=136"
W 14x426
A572 Gr. 50
Lc=136"
Fy.f=50 ksi
Fu4=74.5 ksi
Fy.w=50 ksi
Fu.w=75 ksi
W33x152
A572 Gr. 50
Lb=132"
F~4=62.8 ksi
Fu.f=86 ksi
F~.w=69.1 ksi
Fu.w=93.7 ksi
SS-FCAW
E71T-8
backing bar left in
top flange;
backing bar removed
at bottom flange
W36x160
L~=134"
Fy.f=54.7 ksi
Fu4=75.6 ksi
Fy.w=53.5 ksi
Fu-w=79.2 ksi
welded
(beam web
W36x150
Lb=134"
Fy.f=41.4 ksi
Fu4=58.7 ksi
Fy.w=47.1 ksi
Fu-w=61.8 ksi
DB2
Constant cut
L1=9"
groove
welded to
column)
LRBS=I9.5"
FR=40%
Radius cut
L1=9"
L~Bs=27"
FR=40%
Gp Comments
(O/o)
3.8
3.2
4.0
1.8
2.0 Flange fracture at
RBS
3.0 Testing stopped due"
to limitations of test
setup
A-3

Ref
[5,6]
[5,6]
[5,6]
[7]
Spec.
DB3
DB4
DB5
DB1
Beam
W36x170
L~=134"
Fy.f=58 ksi
Fu.f=73 ksi
Fy,w=58.5 ksi
Fu.w=76.7 ksi
W36x194
Lb=134"
Fy.f=38.5 ksi
Fu4=58.6 ksi
Fy,w=43.6 ksi
Fu.w=59.8 ksi
W30x148
Lb=134"
Fy.f=46.6 ksi
Fu.f=64.5 ksi
Fy.w=48.5 ksi
Fu.w=65.4 ksi
W36x135
A36 Steel
Lb=134.5"
Column
W 14x426
A572 Gr. 50
Lc=136"
W 14x426
A572 Gr. 50
Lc=136"
Fy4=50 ksi
Fu4=74.5 ksi
Fy,w=50 ksi
Fu.w=75 ksi
W 14x257
A572 Gr. 50
Lc=136"
Fy.f=48.7 ksi
Fu.f=69 ksi
Fy.w=49.4 ksi
Fu.w=66.2 ksi
W 14x257
with 1-5/16" thk.
cover plates
(cover plates welded
across flanges of
W14x257 to form
box)
A572 Gr, 50
L~=132"
Flange Welds
SS-FCAW
E71T-8
(details of backing
and weld tabs not
available)
Web
Connection
Not
Available
RBS Details
and Other
Flange
Modifications
Radius cut
L1=9"
LRBS=27''
FR=40%
Radius cut
L1=9"
LRBS=27"
FR=38%
Radius cut
L1=5"
LRas=25"
FR=38%
Radius cut
L1=8"
LRBS=28''
FR=40%
~p
(%)
3.8
3.7
4.0
3.0
Comments
Testing stopped due
setup; significant
column panel zone
yielding
Testing stopped due
setup
A-4

Ref
[8]
[8]
[8]
[8]
[8]
Spec. Beam Column
S-1
S-2A
SC-1
S-3
S-4
W530x82 (Canadian
Designation)
d=20.8", bf=8.2",
tf=0.52", tw=0.37"
wt.=54 Ib/ft.
Lb=142"
CSA G40.41-350W
steel
Fy.f =52.4 ksi
Fo.f=76.6 ksi
Fy.w=57.5 ksi
Fu.w=81 ksi
(~
W 14x120
A572 Gr. 50
Lc=120"
Flange Welds
SS-FCAW
E71T-8
backing bar left in
top flange;
backing bar removed
at bottom flange
Web
Connection
Bolted:
5-1" A325
RBS Details
and Other
Flange
Modifications
Radius cut
L1=4.7"
LRss=l5.7"
FR=55%
0p
(%)
9.0
3.6
3.4
note
(8)
note
(9)
Comments
Specimen loaded
monotonically; testing
stopped due to
limitations of test
setup
Testing stopped due
setup
Composite slab
included (6); testing
stopped due to
limitations of test
setup
statically applied
simulated earthquake
loading (7); testing
stopped due to
reaching end of
loading; no connection
failure
dynamically applied
stopped due to
reaching end of
failure
A-5

Ref
[8]
[11]
[11]
[11]
[11]
[12]
[12]
Spec.
SC-2
LS-1
Beam Column
W30x99
A572 Gr. 50
W14x176
A572 Gr. 50
Flange Welds
SS-FCAW
E70T-6
Web
Connection
welded
(Beam web
RBS Details
and Other
Flange
Modifications
Radius cut
L1 = 7"
LS-2
LS-3
LS-4
DBBW
Beam 1
Lb = 141"
Fy.f= 54.0 ksi
Fu4= 71.9 ksi
Fy.w= 58.0 ksi
Fu.w= 74.8 ksi
W36x150
A572 Gr. 50
Lb = 141"
Lc = 150"
Fy.f= 55.5 ksi
Fu4= 74.0 ksi
Fy.w= 54.0 ksi
Fu.w= 73.1 ksi
(~
W 14x398
A572 Gr. 50
Lc = 146"
backing bar left in
top flange;
backing bar removed
at bottom flange
~
SS-FCAW
E70T-6
backing bar left in
groove
welded to
column)
Bolted:
10 - 1" A490
LaBs = 20"
FR = 50%
Radius cut
L1 = 9"
LaBS = 27"
FR = 50%
DBBW
Beam 2m
Fy.f= 54.3 ksi
Fo.f= 68.8 ksi
Fy.w= 59.4 ksi
Fu.w= 72.0 ksi
Fy = 53.0 ksi
Fu = 73.0 ksi
(based on CMTR)
top flange;
backing bar removed
at bottom flange
.
0p
(%)
Note
(9)
Comments
Composite slab
included (6);
dynamically applied
stopped due to
reaching end of
failure
4.0 No connection failure
+1.0 note (12)
/-5.0
-1.0/ note (12)
+5.0
4.0 No connection failure;
testing stopped due to
limitations of test
setup
4.0 No connection failure;
test stopped due to
limitations of test
setup;
see note (13)
4.0
A-6

Ref
[12]
[12]
[13]
[13]
[13]
[13]
Spec.
DBBW-
C
Beam 1
DBBW-
C
Beam 2
DBWW
Beam 1
DBWW
Beam 2
DBWW
-C
Beam 1
DBWW
-C
Beam 2
Beam Column Flange Welds Web
Connection
W36x150
A572 Gr. 50
Lb= 141"
Fy.f= 54.3 ksi
Fu.f= 68.8 ksi
Fy.w= 59.4 ksi
Fu.w= 72.0 ksi
¢¢
W 14x398
A572 Gr. 50
Lc = 144"
Fv = 53.0 ksi
Fu = 73.0 ksi
(based on CMTR)
SS-FCAW
E70T-6
backing bar left in
top flange;
backing bar removed
at bottom flange
((
welded
(Beam web
groove
welded to
column)
RBS Details
and Other
Flange
Modifications
Op
(%)
5.0
3.8
3.5
Comments
Low cycle fatigue
fracture in RBS;
see note (14)
Fracture of bottom
beam flange adjacent
to groove weld;
fracture initiated at
weld access hole;
see note (14)
No connection failure;
test stopped due to
limitations of test
setup
see note (13)
3.5
5.0 Low cycle fatigue
5.0
fracture in RBS
see note (14)
Low cycle fatigue
fracture in RBS
A-7

Ref Spec.
[14] WG-1
[14] WG-2
[14] WG-3
[14j
Notes:
Beam
W33x201
A572 Gr. 50
Lb = 160.5"
F~.f= 52.0 ksi
Fu-f= 72.8 ksi
Fy.w= 51.5 ksi
Fu-w= 68.0 ksi
W36x300
A572 Gr. 50
Lb = 159"
F~.f= 56.0 ksi
Fu4= 72.9 ksi
Fy.w= 56.7 ksi
Fu.w= 74.5 ksi
WG-4 "
Column
W14x311
A913 Gr. 65
Lc = 152"
Fy.f = 69.0 ksi
Fu4= 88.3 ksi
Fy-w= 68.0 ksi
F..w= 86.5 ksi
5/8" doubler plates
(A572 Gr. 50)
provided on each
side of column web
W14x550
A913 Gr. 65
Lc = 152"
Fy.f= 67.0 ksi
Fu4= 86.8 ksi
Fy.w= 68.1 ksi
Fu.w= 87.6 ksi
Flange Welds
SS-FCAW
E70TG-K2;
backing bar removed
at bottom flange
Web
Connection
Bolted:
13-1" A490
Bolted:
20 - 1" A490
(2 rows of 10
bolts each)
RBS Details
and Other
Flange
Modifications
Radius cut
L1 = 9.3"
LRBS= 25"
FR = 54%
Radius cut
L1 = 10"
Lass = 27"
FR = 51%
~p
(%)
2.9
2.9
3.5
Comments
fracture of RBS at
local buckle in RBS
see note (15)
No connection failure;
test stopped due to
limitations of test
setup
1~
" 4.5 "
1. All specimens are single cantilever type, except DBBW, DBBW-C, DBWW, and DBWW-C
2. All specimens are bare steel, except SC-1, SC-2, DBBW-C and DBWW-C
3. All specimens subject to quasi static cyclic loading, with ATC-24, SAC or similar loading protocol, except S-1, S-3, So4, SC-2, LS-2 and LS-3
4. All specimens provided with continuity plates at beam-to-column connection, except Popov Specimen DB1 (Popov Specimen DB1 was provided with
external flange plates welded to column).
5. Specimens ARUP-1, COH-1 to COH-5, S-1, S-2A, S-3, S-4, SC-1, SC-2 and LS-4 provided with lateral brace near loading point and an additional
lateral brace near RBS; all other specimens provided with lateral brace at loading point only.
6. Composite slab details for Specimens SC-2 and SC-2:118" wide floor slab; 3" ribbed deck (ribs perpendicular to beam) with 2.5" ~oncrete cover;
normal wt. concrete; welded wire mesh reinforcement; 3.4"dia. shear studs spaced at 24" (one stud in every other rib); first stud located at 29" from
face of column; 1" gap left between face of column and slab to minimize composite action.
A-8

Seccion reducida

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Seccion reducida

Similar to Seccion reducida (20)

Recently uploaded

Recently uploaded (20)

Seccion reducida