STEEL
DESIGN
Joseph E. Bowles
Professor of Ciod Engineering
S k;lrrLTURALSTEEL DEIGN
CONTENTS
9
t
2.
&.@!
'67890
DODO 89876543210
Preface
'Uh.*<
of Congress Cataloging in Publication Data
Bor ' s. 'oseph E
rdral steel des~gn
.>graphy:
p.
udes mdex.
Bulldmg, Iron and steel. 2. Steel,
Str~
3. Structures, Theory of. I. T ~ t l e
'TA6!' "478
624'.1821
7918155
1SBW r 370067651
!
11
12
13
14
15
16
17
18
19
110
11 1
112
113
114
115
Types of Structures
Design Procedures
Steel as a Structural Matenai
Steel Products
Steel Strength
Temperature Effects o n Steel
Structural Design Codes
Building Loads
Highway a n d Railroad Bridge Loads
Impact Loads
Earthquake Loads
Fatigue
Steel Structures
Accuracy of Computations and Electronic Calculators
Structural Engineering Computations in SI
Methods of A n a l y s ~ s
Beam Analysis
Determinate Structures
Truss Analysis
h a d Frame Analys~s
Bndge Analysls
27
28
29
2 I0
2 1 1
48
49
4 10
4 1 1
412
4 13
414
General Considerations
Design of Beams by the Elastic Method
Design of Continuous Beams
Web Buckling and Crippling
Shear Criteria
Strong versus W e a k  h i s Bending
Deflections
Biaxial Bending and Bending on Unsymmetrical
Sections
Shear Center of Open Sections
Design of Laterally Unsupported Beams
Beams with Nonparallel Flanges
Design of Bridge Stringers and Floor Beams
Composite Beams
Beam Design Using Load Resistance Factor Design
(LRFD)
BeamColumn Design
Introduction
General Considerations of Axial Load with Bending
Effective Lengths of Columns in Building Frames Developing the BeamColumn Design Formulas
Determination of the Interaction Reduction
Coefficient C,,,
AASHTO and AREA BeamColumn Design Formulas
BeamColumn Design Using Interaction Equations
Stepped Columns and Columns with Intermediate Axial
Load
Control of Sidesway
>
BeamColumn Design Using LRFD
Welded Connections
General Conslderat~ons
Weldmg Electrodes
Types of Joints and Welds
Lamella Teanng
Onentation of Welds
Welded Connectlons
Eccentrically Loaded Welded Connections
Welded Column Base Plates
Welded End Plate Connections
Welded Comer Connectlons
Fillet Weld Design Using LRFD
Appendix
A1
A2
A3
Index
The primary purpose of this textbook is to provide the basic material for th
course in structural steel design. The text contains elements of both buil
and bridge design for use in the structural engineering sequence of c i d
gineering programs. If the instructor wishes to emphasize building frames,
text is also suitable for an introduction to structural steel design in arc
programs.
Approximately equal emphasis is given to fps and SI units. In the dis
material both systems of units are used; the examples and homework pro
are in either fps or SI. This format was arrived at throu$ discussions wi
nurnber of interested faculty members and people in industry. The conse
was that the text discussion should continue to use both systems of units beca
transition to metric is not occurring as rapidly in the construction indust
other areas'of engineering. Dual usage seems necessary to provide both
and instructor with a feeling for what is a reasonable member size (num
deflection, or other design parameter in both systems of units.
Practical SI instruction requires use of design data, a n d since none
readily available, I have assembled a set of computergenerated rolled
section data tables as a supplement to the text. These tables are in ge
agreement with the AISC and ASTM A6 specifications. This bound set of
also includes edited material from the AISC, AASHTO, and AREA
tions. It is intended that the textbook, together with the supplemental Stnict
Steel Design Data (SSDD) manual, will provide adequate material for a
design course without the need for any other reference material. T b e
material should be sufficient to enable students to design routine (an
notsoroutine) structural members in either fps or SI units and by any
the three steel design specifications which are most likely to control the deslgn
PXEFA
'
fl PttZFACE
$~$he>,seris simply given the loads for each design problem. Admttedly, the more
reall, IC design problenls require more physical and mental effort on the part of
"/
\
,'r'lcs;udent and more grading effort on the part of the instructor. This extra
,
~".ff~:t can be offset somewhat by assigning fewer total problems, but including
t~i;;s In which loads are glven, to bu~ldconfidence, and some with design
prc i !i31:is, to bulld des~gnsk~ll.
7 iie following text sequence might be appropriate in the semester system:
i'l
L1
AC KNQWLEDGMENTS
Several persons and organizations have provlded considerable encouragement
arid assistance in produclng this textbook. First, I should llke to express my
,incere appreciation to Dr. Peter Z. Bulkeley, Dean of Englneenng and Technol
1 TYPES OF STRUCTURES
e structural engineer wlll be concerned wlth the design of a v a n
d the follo\wng:
structures including, but not necessarily I ~ m ~ t eto,
GENERALDESIGN COP.SIDER%TI
4,
%)
~ of e[astlcrp,
d
E~ The typlcal
l
~rdnge for
~ all steels (relatively l n d e ~ e
n as 29 000 ksl or 200 000 MPa
E
2(1 + P )
where p = Poisson's ratlo taken as 0.3 for steel. Using ,u = 0.3 9ves G =
11 000 ksl or 77 000 MPa.
as
3. coefficlenf of expnnsron, cu T h e coefficient of expansion may be
G =
//
a = 11
AL
25 x
a(T, 
per " C
T,)
L
C = ; ( F  32)
y j o9m Q
w
0 0
3$2&Q
m
0
k?
8
4
3
g 3
;aa
. 9 9
ggmz
5)
a
rd
"0
o
+
..
*9' 0
.O
$ 2
k;"
.
2
*,
.
2
m m
+ ? Y r :
w  0 0
"
"
%
3
Q)
88
$898
I
m
m
o m  o
Q W W t 
S: , 8 2 3
N
N N m m
%Q%Z
.z
ba
<.ga
.V1
L
Z
m
m
.L
8%3$
Note that lb mass and lb weight or force have been used interc
ably in the fps system because the accelerationproducing force
of gravity. This cannot be done in the SI system, since the ne
derived unit that defines the force necessary to accelerate a
'1 m/s2. The acceleration due to gravity is approximately 9
xample: Given: A rolled structural shape weighs 300
Required: mass/m and weight/m.
Solution: mass/m = kg/m = 1.488164(300) = 446.
wei&t/m = kN/m = 0.0145939(3
k 444
B
E
V1
$i nh
M
I
aJ
Z;
E"
g3
g 7;
E 6 s
8 5
1 4 . 1 W Shapes
X,e most commonly used structural shape is the wideflange or W shape. This is
a doubly symrnetncal (symmetrical about both the x and y axes) shape consisting i;f two rectangularshaped flanges connected by a rectangular web plate. The
flange faces are essentially parallel with the inner flange distance for most of the
graiips, with a constant dimension.? There is some variation due to roll wear
and other factors, but the distance is held constant within ASTM tolerances.
h e shape is produced as illustrated in Fig. 11.
Khe dssignation: W16 X 40 means a nominal overall section depth of 16 in with
a weight of 40 Ib/ft.
, The des~~narron:
W410 X 59.5 is the same W 16 as above with a nominal depth in
1
rr~m(based on the approximate average depths of all the W16 sections and
rounded to the nearest 5 mm) and with a mass of 59.5 kg/m.
&or to 1978, at least ope W section in a group designation was "exactly" the
t:)$kpnal depth given (i.c., one W16 was 16.00 in deep; one W18 was 18.00 in
"';,$@p).
,
Now the closest W16 is the W16 x 40, with a designated depth of 16.01
$0: There can be substantial deviations between the nominal and actual depth
. .(<.g., the W21 ranges from 20.66 to 22.06 in). For the W14 the SI equivalent is
W360, but the actual range is 349 to 570 mm (in this case the "average" was too
. fr*
'from the nominal value and W360 was somewhat arbitrarily used).
'
It should be noted that the rolled product will contract on cooling and at a
vat~ablerate depending on the thickness at any point on the cross section. The
rolls used to produce the shapes will undergo wear, and coupled with the
enoinlous forces involved in the rolling process, only shapes of nominal dimension (varying from theoretical or design values) can be produced. American
Soclety for Testing and Materials (ASTM) specification A6 in Part 4 gives
a:lowabie rolling tolerances, including amount of flange and web warping and
deviation of web depth permitted for the section to be satisfactory. Generally,
the maxipum permissible variation in depth as measured in the plane of the web
is 5 in or 3 mm. Note, however, that the permissible difference in depth
between two rolled beams with a theoretical depth of 16.01 can produce extreme
depth:; of 15.885 to 16.135 in or a difference of 4 in or 6 mm. These variations
should be' kepr rn mind, particularly when converting to SI dimensions for
detailing, clearances, and mating of parts.
14.2 S Shapes
These are doubly symmetric shapes produced in accordance with dimensions
adopted in 1896 and were formerly called 1 beams and American Standard
%
t The several sections with a constant nominal depth. Where a group consists of a large number
of S~CIICJI~S,
a second inner flange distance may be used.
r
 
.
hd
'V shapes
S \hapes
C shape,
L shape
:;gc
Anierrian Stdndard
bean1 (Ibeatri)
Chdnnel
to 330 MPa and refer to Figs 130 and 136. Similarly, A44
point of 345 MPa, will have a yield strength on the order of
guaranteed values converge.
UReitdllgular
"_,;id
Lshdpe
7 slldpe
1 1 ~ q u dleg
l dtlgle
8 Square
0Rounds
Ihrt
Plate
t:,.j
L Shapes
i i . e , ~shapes are either equal or unequal leg angles. All angles have parallel
14.'; T Shapes
St:
LC
S"
WI.:1
~ r a tees
l are structural members obtatned by splitting W (for WT), S (for
curves for
. ...
G E N E R U DESIGN CONSIDERATIO
"C x 100
Steel is not a flammable material; however, the strength is heavily temperaturedependent, as illustrated in Fig. 14. Both the yield and tensile strength at
1000F is about 60 to 70 percent of that at room (about 70F) temperature. The
drop in strength 1s rather marked at higher temperatures, as shown on the figure,
where the strength at 1600F is only about 15 percent of that at room temperature.
Steel frames enclos~ngrnater~alsthat are flammable will require fire protection to control the ternperature of the metal for a sufficient time for the
occupants to seek safely or for the fire to either consume the flammables or be
extir~guishedbefore the building collapses. In many cases the building does not
t See footnote a of Table 11 In J. E. Bowles, Structura/ Steel Desrgn Data Manual, McGrawHd,
New York, 1980
16.2 IAVTemperatureEffects
Brittle fracture is a failure often associated with low temperatures. Essentially,
brittle fracture is failure that takes place without material yielding. The
stressstrain curves of Fig. 13 indicate that m the usual failure of a tensile
specimen, considerable elongation takes place. As a matter of fact, a minimum
percent elongation is specified for steel in the ASTM standard tensile test.
Inlplicit in steel design is the resultant deformation (yieldixig) of the material
Fireproofing materials
Cinder concrete
Gypsum board
plaster, cclnetit and sand
Expanded shale concrete
Vertnicul~te
Perlite
Unit w e ~ g h t
Pcf
l 10
3040
100
I
I
Usually use 35* nim o f flreprotection for 2h fire rating;
obtain specific thickness values from either tests o r from
the producers o f gypsuln, perlite, etc.
k~/rn~
17.3
4.76.3
15.7
,.,
rigure 15 Methads of producing fireproofing of structural steel members. ( a ) Sprayed fiber. (6)
Lath and plaster. (c) Lightweight concrete (formed). ( d ) Gypsum boarduse boards to build
'hick;us. ( e ) Corner detail of sprayed on fireproofing. Thickness is built in several sprayings. (fl
Bear. dzwl of sprayed fireproofing.
16
occurring is of little aid in settling the resulting damage claims that are sure
follow. Brittle fracture can be controlled in several ways:
1. Detail niembers and their connections to minimize stress concentrations.
2. Specify the fabrication and assembly sequence to minimize residual ten
stresses.
3. Use steels that are especially alloyed for lowtemperature environments
the owner/client may require a more stringent design than the building co
c,piteria. Only in rare cases can the designer get a variance from the local
governing body to deviate in a less conservative manner from the code. Vari
18 STRUCTURAL ST
fps: R
SI:
0.0008 x area
R = 0.0086 x area
(when area
(area
>
150 ft2)
include:
Ceiling materials, including duct work for environmental co
supplies.
xterior walls supported by the frame, including windows, doors, and ba
Interior walls Qat are permanently placed.
'iechanical eqmpment (heating, air conditioning, ventilati
' (such as elevators, including cage, cables, motors).
r .reproofing.
Beams, girders, and columns, including the footings making
frame.
From this list it is evident that any part of the building w
nstalled contributes to the total dead load. Dead loads can
= 12 psf
From Table IV4 of SSDD, the live load = 5.00 kPa. The r
r for a grder based on a contnbutory area as shown is
El1, 18 X 22 (area ABCD):
R = 0.0008(18 X 22) = 0.32
< 0.60
0.0086[(12
4.33 L
4.33(5.00)
d loads have been extensively studied in recent years, particularly for larger
rise structures. Generally, for tall structures ~wndtunnelstudies should bz
or smaller regularshapsd
30 m, the wind pressure
ory to use. The Nation
15
0.75
SOLUTION
Note that a public room is not the same as
,whece the loading is pnmarily seating in fixed or movable
definition we may use a liveload reduction factor.
First, estimate the dead load using Table IV3 of S
weight of concrete = 23.5 kN/m3.
Weight of concrete: 0.125 x 23.5
The sever31 wind values are shown in Fig. El36. The data display is
convenient for computer programming for frame stresses using the com~utergr,ogramdiscussed in Chap. 2.
///
lilllllllllllllllII1111111111'
Figure 18 Snow and other roof live and dead loads.
In addition to the types of pressure or area loads noted, building codes may
stipulate checking for a concentrated load of some magnitude which may be
placed anywhere on the floor or roof. Where roofs are used as recreational areas
or sun decks, the live loads must be adjusted to values based on occupancy in
addition to considering snow and/or wind.
Po~idingis a special roof load that may require investigation. Ponding is a
condition where water collects on a flat roof which has deflected locally
(possibly due to an overload, poor construction, foundation settlement, or
plugged roof drain), causing a concentration of water which in turn increases the
load and deflection, causing a further concentration of water. Noting that a
water depth of 1 in results in a liveload pressure of 5.2 psf, loads are readily
eveloped which can locally fail roof members. Through progressive failure, the
roof may collapse. Ponding design is considered in some detail by Marho
July 1966 AISC Engineering Journal.
'
'
),
,,, ,*,.
,. . .
31 kips IJJ kN
24 k ~ p s108 k N
I6 kips 7 3 kS
Bridge span

33.8 to 145.6 ft
M,.,
= :[(0.9~
+ 4.206)(0.5L + 2.33) 
1l 2 L ] It  kips
26 ?a :27.5 ft
2.53 to 38.86 m
Over 127.5 ft
Over 38.86 m
a W = 40, 30, and 20 kips or equivalent in kN (and is the basic truck load, not the
total).
al axle lodd
nd of train:
. !
,. 3
..' 1
 ..
23 712.0~
35 118.0b
48 800.0~
17 990.0
27 154.0
38 246.0
479.6
282.0
128.1
522.0
626.4
729.3
306.8
367.3
426.4
197.9
1225.3
.U1 values shown are for one rail (onehalf track load). Axle loads shown in diagram. Obtain
for other E loads by proportion. ,,
.4t center of span; other moment values are usually close to center of span, so one may obtain
!utal moment as the sum of w ~ ' / 8 for dead load + live load value shown in the table.
. . :s
,;
I.
f ::cd
total uniform loading, however, must not be less than the following:
SI, kPa
Loaded span
fps. psf
SI, kPa
AASHTO
AREA
100
90
75
75
100
2X100+1X50
2~100+1X50+1X25
As specified by designer
4
XI.xe than 4
Other bridge loadings that must be considered include impact, wind, and
longitudinal forces. Impact and longitudinal forces allow for dynamic effects
from rolling equipment going across as well as for starts and stops made on the
bridge. Impact will be considered in the next section. The wind force is
>elfexplanatory and in the case,of a loaded railroad bridge, the wind against the
train may be'a substantial load.
AREA: 0.15
Other loadings that may require consideration include differential temperatures between top and bottom flanges or chords, ice and snow loads, possibIe
overloads, and for continuous bridges, support (pier) settlements.
.
<..*,,.,
6. ..4
 ..'i
.  . llr,
....,.., 
G E N E W . DESIGN CONSIDERATIONS
~ L O I U I Y
Example 14 Given a highway truss bridge with HS 20 loading. The truss
anels are 7.5 m, and the distance between reactions is 37.5 m. T h e distance
etween trusses (width) is 14.1 m. What is the impact factor, I!?
OLUTION The impact factor will vary for the floor beams, stringers, and
russ, depending on their lengths. For the stringers the impact factor is
l5
= 0.330 > 0.30
therefore. use 1/ = 0.30
//
impact load as
Item
Elevator loads
Macbnery and other moving loads
1.00
> 0.25
f PS
I,
=.
2
< 0.30
I,+ 125 
I,
l5
i0.30
L+38
Example 16 What is the impact for the floor beams of the AREA truss of
Example 15? Floor beams are transverse members connecting the two
trusses at panel points.
SOLUTION
S = 27.6 ft, L = 17 f t
< 80.
3(17)*
+ 40  = 43.1 percent
1600
HQUAKE LOADS
general trends. One is to attempt to model tfie
asses and springs and use a digital computer to
s assumed earthquake accelerations The
the earthquake accelerations based on earthcitation based on building geometry, and apply
GENERAL DESIGN
roof
Elevation
Figure 112 Earthquake zone map for the Umted States. (After Unrform Bu~ldngCode, 19
...
Alloaf....
Wood sheathing
= 3 psf
= 7 psf
= I1 psf
,,.,.'.b..*
"
..
.
.,.
G E N E W DESIGN CONSID
9
Any floor:
+ 30 x 2)(8)(20)/(40 x 30)
,I
.4
4
Floor: (144) (concrete)
12
Ceiling (estimated)
Bar joists and metal pan
Exterior wall at 10ft height
(2 x 40 + 2 x 90)(15)(10)/(40
x 90)
Figure El76
number of stories
0.1(10) = 1.0 s
(F F
lop
)=Wn A,
2 W,hn
(218.6  13.7)
F,, = 0.001351(337
F,
0.001351(337
Summing the horizontal floor loads and including the top value of 13.7
218.59 versus 218.6 kips as a check. These lateral floor loads
further prorated to the several bays in the EW direction for the fr
analysis load condition(s), which includes earthquake forces.
12 FATIGUE
failures which have been attributed t
ue. Fatigue failure is a material fracture caused by a sufficiently Iarg
ulsating stresses, or stress reversals. Th
ture of the material at a location where
oscopic in size) exists. A crack form
nding on stress level, rapidly or slowly (sometimes so slowly that the
re) progresses to failure of the part
Most metals tested under repeated or cyclic loadings display stress r
s qualitatively illustrated in Fig. Irly, these curves were commonly displayed as stress level versus cycIes.
t, the stress range is used as the parameter of interest. The stress range
to
41
!@I
T or R e p 60 (415)
36 (250)
24 (165)
24 (165)
T o r Rev
32 (221)
19 (131)
13 (90)
lob(
T or Rev
2 1 (145)
12.5 (86)
8 (55)
T o r Rev
45 (310)
27.5 (190)
18 (124)
16
TorRev
27(186)
16(110)
10(69)
7(48
T or Rev
45 (310)
27.5 (1%)
18 (124)
16 (110
T or Rev
45 (3 10)
27.5 (190)
18 (124)
16 (110
T or Rev
21 (145)
12.5 (86)
8 (55)
Figure 113 Qualitative plot of stress range F,, versus number of cycles to failure.
. .",., ... ',"....,. 
3.
can be defined as
The AISC, AASHTO, and AREA specifications are very nearly .identical
(where rnembcr failure is not catastrophic) in specifying the stress range and
number of stress cycles. These specifications are based on a large number of
fatigue tests performed (see Fisher, "Fatigue Strength of Steel Members with
Welded Details," AISC Engineering Journal, No. 4, 1977). Typical stress range
values which may be used for all three specifications are given in Table 14. The
reader should consult the Structural Welding Code, Sec. 914, which is the ori
of the material used in the three specifications or the appropriate des
specification for a more complete presentation of fatigue cases and F,.
The largest value in each cycle category in Table 14 is generally applicable
to buildings. Lesser values than shown are necessary for reduced sections,
certain types of joints, type of joining material, and for certain members in
industrial buildings. The AISC (Appendix B) manual, AWS Structural Welding
Code, the AASHTO specifications (Sec. 17.2) or AREA should be consulted for
those situations where fatigue must be considered. Note that fatigue is not
usually considered with wind or earthquake loadings on buildings. Fatigue is
usually not considered for routine building design, since 10 load cycles/day over
_ . . _.
. .  " .,.....
N = 10 x 365 x 2 0 = 73 OM
,,
This is seldom enough cycles of whatever thk;,itress.rang6 to require a reduction
___________
._
_.

.._I
SOLUTION
f,,=16(12)=28ksi>24ksi
N.G.
The section is inadequate for this number of cycles; increase the section so
that f,, < 24 ksi. Note that the section is adequate for "allowable" static
stresses.
T ~ L I ~ S soli~rnn
on
h type, where traffic passes between the trusses. The decktype trus
preferred if clearance beneath the truss is not a factor, because pr
Many truss bridges combine both types (see Fig. 119) o
of a truss for the longer spansand girders for the short
s a common practice. This latter scheme is illustrated I
with bridge trusses are shown in Fig. 120 (see also Fi
n bridge design is to use girder structures, which req
sses. In all cases as much welding is used as
lions either welded or fabricated using highstrength
ACCURACY
leng:;. uf the diagonal members.
113.2 Bridges
the 10in slide rule was the principal computational tool in the structur
er's office, the computations rarely exceeded three significant dipits. This
satisfactory, for the reasons presented earlier in this
y and as implied in the example computations.
Presently, the electronic calculator and/or the digital computer are almost
nivetsally used for structural computations because of both the greater corn
lexity of structural configurations and the greater speed of performing calcul
ally setting the decimal. Now it is almost mandato
ethical and economical reasons) to provide several iterations on a
he design. This step almost always requires use of
These calculating devices can give a rather large number of digits to any
blem is how to treat this increased computing capacity.
ot any better than the input, but with a large number of
arently significant digits it looks very impressive. In nearly all design offices,
design computations are checked by a second person as a design precaution,
ner carries a large number of string calculations on an
ronic calculator, the results will differ from those .obtained where the
ker truncates intermediate steps, then reenters these values and continues
computations. Where the discrepancies are not large, the question arises of
ether the problem has been "checked" or whether one (or both) of the
persons has made a design omission.
For these reasons it is suggested that regardless of the initial input data
accuracy, computations should be camed to as many decimal places as is
aps) to obtain good checking convergence. The extra
alculation effort is minimal. Any intermediate steps should be written to the
me precision as they are used in subsequent calculations (e.g., do not write
06.1 and then use 106.153 in the following computations).
50
STRUCTURAL S T B ~ LDESIGN
To convert
to
kilogram (kg)
:mund (lb)
pound
';ips (1000 Ib)
ib/ft
: ips/ft
!b/ft
psi
psi
ksi
kilonewton (kN)
kg
kN
kN
kN/m
kN/m
kg/m
MPa
kPa
MPa
kPa
kPa
kPa
kN . m
kN m
.ASI.
psf
isf
Lip . ft (moment)
kip . in
What are the wind forces at points 1 through 9 of the roof shown in Fig. Pi7?
Answer: ~ ( 5 =
)  18.07 kips.
Multiply by
0.009806650
0.4535924
,
0.004448222
4.448222
0.014593727
14.593727
1.488 16404
0.006894757
6.8947577
6.8947577
6894.7577
0.04788026
47.88026
1.35584
0.1 129862
PROBLEMS
11 What is E, for a steel with F, = 40 ksi?
12 What is E, for a steel with F, = 365 MPa?
Answer: 0.00182.
1 rep roofing
Figure PI3
.
.,,.
c,
.,.
,.,,., ".,..,..
1.1What is the increase in weight of a W410 X 46.1 rolled section with 55 mm of vermiculite
fi:.eproofing as in Fig. P131 (See the comment in Prob. 13 before starting this problem.)
i5 What is the R factor for the interior column at the top floor and three floors down of Ex
1I?
16 What is the R factor for beam B1 and column B2 of Example 12?
Answer: M = 7865 kN . m; I
40.1 percent.
12 What are the story shears in the NS direction of the building of Example 17?
'
. of
ben
Ib )
Figure 11.1 Two of world's tallest buildings using structural steel frameworks. (
(4424 Sears Tower building in Chicago, Illinoiscently
the
tallest buil
337111John Hancock Center building, also in Chicago, Illinois. View from the sears
ANA
I
.*
The method of analysis depends oh. the complexlty of the structure and
whether it is rigid (indeterminate) or simply framed. An analysis may consider
the structure as either two or threedimensional. Simple framed structures are
generally determinate; that is, the three equations of statics (2FA,2 F,, Cbf =
are sufficient to obtain the internal member forces. In any case, with simpEe
ming the ends of the members are assumed to have no moment resistance
ransfer to adjacent members. The internal member forces of determinate
tructures are readily obtained by hand calculations and with considerabIe
efficiency using pocket calculators.
Rigid,framed structures are generally indeterminate since the member ends
transfer shear forces and moments to the adjacent members. Indeterminate
ructures require d e f o m o n compatibility to supplement the equations of
atics to determine3e internal member forces. DigitaEomputers are used to
obtain solutions for all but the simplest indeterminate structures. Continuous
beams and certain simple rigid frame structures have solutions that can be
ELEMENTSOF FXAME,TRUSS, M
.D
'
,q = 4 + 6 + 3 = I3
'
'

~ I ,
21 portal
Mc
v.=I
:'
= 3P L= 
2 2
PL
4
PL
2
+,C ,
 wx2
EIy" = M = moment = + c 1 x
2
+ c2
 wx3
c1x2
EZy' = slope = + 6
2
+ C2x + C,
TERMINATE STRUCTURES
'.
successively:
EIy"' = V = shear =  wx
Loadbrar~rigwail ,
 wx4
EIy = deflection = 24
Tile general equation for the elastic curve for a beam produ~cesfour
integration. Constants C2 and C4 = 0 for simple beams, and the w
for beams loaded with concentrated loads.
The general beam equation is sometimes useful in that the appro
i n f ~ i n a t i o nmay be approximately obtained by replacing a ser
ce~tratedloads with the equivalent uniform load value.
,@,earnde.fl,c.ctionsare often limited for both buildings and bridges. A va
of L/360 to as low as L/1000 can be fo
occupancy and/or interior finish. Where' deflection
approximate deflections should be computed before a de
The reader should note that highly refined deflectio
possible because of the uncertainties in loading. It should also b
deflection is heavily depende,nCton the moment of inertia_an
constant for steel. Therefore, where a rigid deflection criterion i
use of highstrength steel ma'y not be economical, since the sectlo
fix?? by the moment of inertia rather than by the bending stress.
ical sense almost all beamcolumn connections can, and do, transmi
2j

erea&Iy _
optimized for least
  weigh~,sincehes_geometry.
Indeterminate
trusses are not SO readily
.
60 snivcl.rfRA~STEEL DESIGN
V Z ) The
~ . nodal equilibrium in matrix notation is
P=AF.,
',
'
x = ( A S A ~ ) P'
p4
 x4
as
e product of A x SA
Truss member
PI + F I cos cx = 0
ZFv = O
66
,," ..,.,.....,.
ELEMENTS OF FWLIE,
p6
VP = 6
1VlII = 2
NPPI = 7
12(0.707 11)
6
()'
0.707 11
12
RL=4(72)
6
 48.0
,= 50.91
kN  m
70 STRUCTURAL
STEEL DESIGN
P4 =
.",,.,,
4,
7 mPP1)
(does
not matter)
(ASA~)'{P}
night include:
NLC
b a d combination

~ l e a d+ live
Dead + live in alternate bays
Dead + live snow
Dead live f snow
Dead + live + snow one side
Dead live wind
..
1
2
3
+
+
the end of the coding sequence is checked and both satisfy statics,
problem has been correctly solved for that irlpltt data.
Problem checks can be performed in several ways, as illustrated in
+ wind
of computational effortoften
a story are also often convenient, as the column axial forces are direct1
Advantage can often be taken of symmetry of structure and lo
two office building examples that follow (without wind). Sometimes t
matrix can be used to advantage, as illustrated in the column check of th
building.
Truss checking is similar to that for rigid frames. Always take
truss geometry by checking members in which the internal force is
zero. If the computer output does not give zero and the load
there is something surely wrong with the input data, such as mispunching an NP
or a mismatch of H and V (either length or signs). Perform any additional
statics checks near each end of the truss at joints where a minimum of bars
connect. If these joints do not satisfy statics, the more
either and the problem needs to be reprogrammed.
7'6"
1
Z
2 1'1"
7
3
(2)
(5)
(3)
16)
(9)
110)
(12)
(13)
10 ( 1 6 )
I I
(19)
(20)
14123)
I j
(75)
(26)
117)
124)
,.s.v
4
(;
17 45.47.47
.\'P = 4 b
.YPPI = 47
18 47
r.2.i
Elevation o f typical interior bent
5 @ 18'6"
 
Plan view
~
SOLUTION
The general computer coding is partially shown .in Fig. E23a
and also on Fig. E23d (computer output sheet). This coding gives the
following computer program control data information:
NP = 46
NBAND = 11
The next step is to develop the beam and column loadings for this
typical interior bent (refer to Fig. E23b, which displays loads after comcputations).
Roof:
Dead load: estimate 5 in of concrete on a metal deck supported by steel
bar joists.
NM
Dead load
0.080(18.5)
26(members)
O.lO(l8.5)
= 1.85 kips/ft
.02(6)18.5 = ? . 2 ? k
Concrete: (5/12)144
Metal deck and joists (estimated)
5 psf
= 5 psf
=70 psf
31.5 percent
"
or
Note that the L = 80 psf value is from Table IV4. Use a 30 perce
reduction.
Live load = 8OC1.00 0.30) = 56 psf
In a 7.5ft span, R = 7.5(18.5) = 138.8 ft2 < 150 (no reduction).
The equivalent beam loads for the roof are:
..
. ... ..,..,.... . .
Other floors:
The exterior brick veneer wall will contribute column loads for' t
terior wall finish, fixtures, etc., of 5 psf gives a total ='45 psf.. .."
Column load
L'
~ & l e c any
t beam weight e tributing shear to the roof line of exteri
columns, since the values are tw small to be either reliable or to affect t
design.
Now make some preliminary member size estimates with the followin
practical considerations:
r
rO
T rr 0C0 0 0
0.0. .. 0 e
0 0CC 0
*r
a ~0O 0
oo
 O
~ .O
O*
o C c ~  ~ O
M,
I
u
"I
;f
the moment is M =
lue is somewhere
   in
A
.a
0.925
900000000*000O_t_O
rr*
* D *
00
0 0 O O O 0 C c 0 O C C 0 0 0 3 0 0 0 0 0 0 9 0 0 0
l 0 0 0 ~ 0 C 3 0 0 1 0 0 0 0 0 0 0 0 5 0 0 0 0 0 0
0
0 0
C 0
0 0
0 0
C 0
C 0
C 0
0 0
0 0
00
00
00
00
00
00
00
00
C0
00
00
00
00
00
00
00
0
ldddd222dddd222;ddd22fdd
0000000000000000000000000
Q O C O C o o ~ O 9 O O O O I O e O O O O * o  ~
;d;;;;;;d;;;;;;dJ;;:;n.dddO
* ~ * * . * * n D r * * *
i
= 142.7'ft , kips
NN
* r O r * * ~ * ~ w ~ ~ ~
NN
*.+
L
6
.
1 I.OU
in
1, = 3 I 6.u in'
This computation
utions + the uniform beam loads from roof to the first flo
'c01umn~
Z
C D D o C o O O O O o O 0 E O O O O o c O o o o o O
For columns:
Estimate a length coefficient (Chap. 7) K = 1.2 +KL = 12 x
14.4 ft. Estimate F, = 16 ksi (som thing less than 22 ksi).
Papp,,
........................
12!0(12)
= 60 in3
24
n =
O O o C O O O O O O O O O o O O O O O O O O o o
dl,d;dd;ddddld;,hdb:dD'O'd::
*.*<*.wm~=*
*rwnrnccn.oQ*
000000000000000000000000
000000000000000000000000
00000000000000000000O0O0
dddddddddddddddddddddddd
000000000000000000000000
O O O O o O O O o O O O O O O O o O O O O O O O
000000000000000000000000
dddddddddddddddddddddddd
0 0 0 0 ~ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ~ 0 0
0000000000000000000000~0
000000000000000000000000
.=.
O O O o 0 O O O O O o O O C C O O o o o O o o o
u.
 ,a
................._...._._
   *=*
 
ox
,8
0 0 0     C 3 0     C 0 0     ~ 0 0
Sx =
~ * r 0 0 0 0 =  C o O O C O  ~ 0 0 O O O O e o
* C r 0 0 ' 0 0 . ~ . 0 0 0 0 *  * 0 0 0 0 * C * 0
8
(142.7) == 95.1 ft . kips
12
Ma, = 118.9, say 120 ft . kips
M,
0 0 0 0 0 0 0 0 0 0 0 0 0 C 0 0 0 0 0 C 0 0 0 0
0
0 0
0 0
0 0
0 0 0 C
0 0 C
0 0 0 c
0 0 0 0 0 0
~ 0 0
C C
0 0 0 0 ~
0
+ 1.48)(21'3)2
.
L
.
c
.
.
00
..
;l;dddd:;:dddd.;:dddd2;*
dddddddddddddddddddddddd
000000000000000000000000
000000000000000000000000
000000000000000000000000
dddddddddddddddddddddddd
00000000000000000000000000
00000000000000000000000000
'
ddd;;;;ddd;;;;ddd;;;:ddddd
'"I.
' Z
4
W
.
'4
\ l I T F I I I tN"IE
U
h/D
YO
YP
ND
N>
ND
NO
NP
YO
yo
NO
Yo
LIP
qo
YP
YD
q P
NP
LIP
ND
1
7
10
=
=
=
=
=
=
=
YP =
"
I =
VP =
NO =
No =
NO
VP =
YP =
YO =
LIP =
ND =
No =
YO
.,D
YO
NP
NP
=
=
=
=
NO
yo =
NO =
NP =
NP =
YD T
NO =
=
No =
NO
OANJ h I O T H =
13
AND U NX =
NX =
NX =
NX =
WX =
NX =
VX =
NX
NX =
NX =
NX =
NX =
MX =
NX
NX=
YX =
NX =
NX =
NX =
NX =
NX =
YX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
NX =
FIX =
NX =
NX =
NX =
NX =
NX =
NX =
YX =
YX =
NX =
NX =
=
=
=
690
11
I 2
114
15
16
17
18
19
20
21
22
23
24
25
26
27
2R
20
30
31
32
33
34
35
36
37
38
30
40
41
42
4'
44
45
46
1
2
4
5
6
7
8
0
10
11
12
13
14
19
16
17
18
l a
20
71
22
23
24
25
76
77
ZP
29
10
31
32
31
34
3<
36
77
38
39
40
41
42
43
44
4T
46
0.00394
O.OO4Rl
0.09541
0.00169
0.00033
0.19343
0.00169
0.00033
0.19343
0.00394
0.00481
0.09941
0.00272
0.00081
0.08302
0.00121
0.00006
0.17721
0.00121
0.00006
0.17721
0.00272
0.00082
0.OR302
0,00275
0.00061
0.05121
0.00125
0.00007
0.14018
0.00125
0.00007
0.14018
0.00275
0.00061
0.05121
0.00353
0.00136
0.00023
0.08232
0.00136
0.00023
0.Oq232
0.00353
0.00068
0.00068
XAL F3RCE. K
3
4
5
6
(I
10
11
12
13
14
15
16
17
6.07
22.77
38.21
38.21
82.42
34.19
20.14
20.1+
38.11
3R.70
22.49
2249
1.01
0.*6
1.01
58.49
81.26
81.21
58.64
0.72
0.54
0.12
64.52
49.19
87.27
30.45
18.14
19.14
30.45
63.13
50.34
81.36
8721
49.19
64.52
30.1.3
11.44
11.94
30.33
8136
50.34
63.13
6431
IF1
A = 9.13 in2
110 in"
I,
48:
A = 14.10 in2
From computer o u t p u t for values shown which have been reduced 25 percent I'or wind
5o.that o u t p u t is directly comparable t o D + L o u t p u t :
7.
..l
1,
... ...,...*.,_
101.55
Actual F =  = ]35,4k
0.75
..
R3= 12.09~
V1= 8 965
C h e ~ k l n gcomputer o u t p u t for column (member 25) uslng computer o u t p u t and the statics o f node 1 4
Use W8
40:
A = 11.70 in2
Ix = 146 in4
I, = 184.0 inJ
ample 24 We will design the small office building of Example 2units. Note that the dimensions are slightly different than when
oft conversion. The same general design considerations/parameters
pply as in that example except that the brick veneer wall will be take
00 mm (approximately 2 x thickness of Example 23). Figure E24a
lays the general building layout and SI dimensions.
LUTION The computer program will solve either SI or fps probIems. I
cessary to include a "UNITS" card with fps or SI identification for use in
the FORMAT statements to identify the units of the output.
Since'the same building design criteria are being used, we may pi%
directly to coding the joints (refer back to Fig. E23d since frame is s ~ i l x )
eveloping the frame loads.
oof loads:
Dead load: estimate 130 mm of concrete on a metal deck
steel bar joists.
= 3.055
kPa
=0.263 kPa
= 0.263
kPa
3.581 kPa
Live load: a liveload reduction can be used in the 6.5 x 5.9 m area
Basic live load = 4.0 kPa (Table IV4).
21.1 k N / m
16.5 k N / m
23.6 kN/m
Dead load
Elevat~ono f t y p ~ c a lInterlor bent
59
,!
1
59
59
1
L
1
5.9
,
1
59
3.581(5.9)
Plan vrew
Figure E24a
58100'0
11590'0
',5100'0ZLZOO'O
010b9'1OhlCh',
. .
01200'0fbO1L.kEUOUb'f
EVIOU0
0kEII.EbV8bb.b
~0100'0U0015'19b9L0.4
bLkOO.0
ObULP'ZO99Z5'8
8?Z00'0siSOE'ZEZEIS.8
E~IOO'O
bLLL0.bIEIZS'Y
blIOO'0SZBfb'ZVEZSS'8
Zli00'0
l'd
01100'0
LLEOL'I91LOO'O
OLlOO'OOCZOO'O
IlbO9'1ZfU20.0ULZOO'Ob
b
9OZOO.OSETOO'O
ElZIZ'EZIb00'0
5E100'0bObO9'1LZOEO'O
bLZOO.0
ULEUS'ZILPZO'O
18200'01Ulbl'tbuZOO'O
LE100'0
bZ1bl295000'0EiIOO'OOEEYS'ZIECZO00UZ00'0
12
09
bk
dC
~r
VZ
= XN
= XN
XN
= Xh
 Xh
XN
Xh
xh
XN
Xh
= Xh
XN
Xh
52
?C
VL
><
ZE
rL
ZL
le
Or
bZ
VZ
LZ
iZ
22
1Z
UZ
01
11
LI
VI
51
?I
LI
Ol.bCb511
sb'51Z00'0
OI'bZbS110u'Z~ObLl
UL'LIZ
KN
= XN
= XN
=
=
=
=
OI'bZb$ll
56'5120 0 '0
Ol'bZb'YI1OU'ZLOI,CI
Ok'iIZ
Of
br
Hk
LC
9 LC
6ZZEl'O
ZZfOO'O
Z
1
XN
XN
XN
XN
XN
= XN
OU'ZlObLI5b.GIZ00'0
01bzb511
11'91Z00'0
01'501911
1 1  1 000
01SO19I1
11'01Z00'0
01'G01011
so3
NIS
?I 01
(11
1:
lfr
;I
6 B L
P \ P
PSI;.
19L.b
XI
S6SXOItM
1
f ; I
131113W
Z
= XN
= XN
..
.
..
OZ'OI
OE'EBEZEI
000
0'r@EZ~1
MYW
~ H L

 o h
=
=
=
dN
oh
dh
AllJIlSVl3 OOY
Od3ZYON ON
92
= S138M3Y
dh
OOb
3J
ON
U
L
9
5
b
2
I
= cN
= dh
oh
ON
dN
ON
=
=
=
=
=
=
dN
= dk
S3181N3
tIl3jllS
030151 3 3 l ~ l iL N
31N U N i l Oi dlh G V
9C
 dN
dN
dh
= dk
'XIdllWd
P ~ H
L ~ +N d ~E ~ NZ ~ NI ~ .N3 8 h z r
21
I1
01
NW
ONV
000'58
OOZ'LZ.
OOZ'LZ
0000
000'0
O.OOOOOL
11 31I1JM
E
= llONJ3 O V J l 30 ON
62
HC
= XN
OOZ.LZ
OOZ'LZ
000'0
0000
OOZ'LZ
002LZ
000'58
000'58
000'0
000'0
OOZ'LZ
OOZ'LZ
000.5u
OOO'b8
000.0
000'0
OOZ'LZ
OOZ'LZ
0000
000'0
0000
000'0
009t
OO*..
CC
Zi
II C
0k
=

= XN
= XN
= XN
= XN
= H l O l H ONVB
1~33
dN
ah
dk
ON
dN
5k
21
= XN
S N V i O V M MO Yvl ' X I d l V W  X
2 ~ 3 3
=
=
00'0
Ob'ZlOhCI56*$1Z00'0
O I ' ~ L ~ ~ I
1I'VIZ00'0
01~b01011
XI%
. . .
OCZSZ'OI
OESOO'O
= dk
= dk
01
00'0
OZ'ZZIOZ'ZZI000
00'0
O L ~ E B E 2 E I O C  E B E Z E I 00'uLI00.8~100'0
00'0
OO'uL9211
00'8L9ZII
00'8L100'8LI000
00'0
OO'LfL9ZII 00'8L9ZII~Aluo)
= dN
= dh
= dN
3H1
89
= RI
2081.85
0.75
Now consider C F,,
53.64
O.K.
0:
17.60
14.92
J1
LJ
47.18
Member forces
. .
.'
m
~ c * ~m c~c * ~c o ,~ r c ~o c u
o m ~c o ~
m o ~ o ~* D ~O
dd:dd'::;;;'dd;';:;;;;f;;r:
0 0 f f " m N * O * C N " ~
"IOtCN"
m m m R  C I O n O m O m N n o m C ~ m O * C C O O o
m F C E m N N m ~ O C * m n * 0 ~ 0 0 3 e 0 3 0 0 0
d;:;;;;;d'd';;;d:;:d;;d;;d
~IDP~NC*QONV
I
I , , ,
m n n o m *
. .. . .
*l,"...a~
n m r p t
""ti 'i"
'(Mem.
22)
'(24)
= 10.2
, ,,,=
CFh
15.3(2) = 40.8 k N
O.K.
ELEMENTS OF F R h X E ,
+ purlins, estimated at
= 5 psf
=20 psi
20.6 Ib/ft
A = 6.06 in2
3 x
i:
.A = 3.38 in2
(15
+ 27)(11)/2
+ 19.2)(10)/2
120
=2
92
w=
120
= 0.09 kip/ft
Length of top
= 144 ft
= 45 ft
= 80ft
W(72)
I
The weights of the trusses have been rounded up to account for connections.
The truss will be analvzed for two load conditions.
+ dead on
= 6.84 kips
=
3.42 kips
Main truss:
Interior nodes: (0.10 0.50)(12)
= 7.20 kips
.,...
Exterior nodes: 7.20/2
= 3.60 kips
Snow load at 25 psf of horizontal (span) projection:
4812
= 7.50 kips
=
3.75 kips
5 74
12
1 1 31'
60
sln 0 = 0 19612
LOS 0 = 0 98058
12
L==1224'
cos 0
0 = tdn
=
7.50'
8 = tan'
'1
Py
18
72
 = 14 04'
3 87'
12
 = 12 37'
Los e
= 0 025(25)(12 3 7 ) = 7 73'
r':, =
Figure E t S c
ELEMENTS OF FRA!!IE,
~ 1 1 0 1 9c ~
rhotTlo*
? N O I T I O N NO
YQ
STIFFIII
THF P  Y A T R I X ,
ENTRIES
AND
INK
954
IL
FORCE. K
&XI
D E S I G N EN0 VOlENlS
CO99EClE0 GO. F E N bNC Y l Y O
INEL9 EN0 F I R S T I . FTX

8.15

..
..
YO r
qp =
'$0 =
YP =
YO =
YP
YO =
YD =
YP =
NO =
N O =
NO z
NP =
10
11
12
13
14
15
16
17
1R
14
20
21
22
23
24
25
=
=
26
=
YD
NP
NO
ND
No
NP
NO
ND
N'P
YP
NO
NP
NP
YO
NP
NP
NO
ND
Yo
YD
YP
NP
l o
rlD
;
.
27
28
24
t . PO'
=
=
31
32
=
=
=
79
34
35
96
37
=
=
5
38
30
;
.
40
41
47
41
44
45
;:
$2
.NO
yb
NP
ND
NP
ND
NP
yo
YO
NP
NO
=
r
49
50
51
=
=
=
=
52
53
=
=
=
56
54
55
57
5R
Figure E25e
..


.
.

0.00
*
.
.
.



.

...


rjrsFs
NO
rrlu
FO~CE.I
chapters to design (or redesign the members). Note that the maxim
lateral displacement is now on the order of 4.7 in at the base of the sh
truss and only 4.0 in at the top of the main truss, indicating some additio
bending in the interior columns. Note that this displacement is occunin
the wind NLC, as one would reasonably expect.
I,
I,
2151.9
76 X
low6m
Two interior columns (members 24, 25, 27, 28, 67, 68. 69, 70): W690
= 0.72
kPa
= 0.24
kPa
= 0.96 kPa
IS
'1@ .. .KUCTURAL
STEEL DESIGN
~u. J I K U L ~UKAL
SIEEL DESIGN
KN &NO
YP
ND ;
N
=
VD =
VP =
LiO
3
4
5
6
7
NP
qu
NO
NO
vu =
rrr =
=
<D
8
9
10
11
12
17
UO
'40
\ Y
16
NP
17
19
Nu
NP
Y O :
'40
,do
NO
L(P
..
2
=
YP =
Ye =
v;
NU
*iP
40
14
15
1R
70
71
77
73
24
7r,
26
27
%.
2R
2s
30
31
U I
=
=

YO
34
35
NP
YP
=
UP =
ND =
UD
NO
S'
N %,
.1R
=
=
=
'40
VP
=
=
No =
LID =
YO
NO
37
3,
36
37
38
39
40
41
42
43
44
45
16
47
Vo
"R
yo
4Q
'40
50
No
U'
NP
V?
hi"
=
=
:Z
53
54
55
YO
56
YO
YP
=
NP =
NO
YO
','o i
LIP =
No =
YP =
Yo =
ND =
NV ;
NO =
hi0 ;
NO
.,D
=
=
VP
VO
YO =
=
Urn
ND
ND
=
LID =
YO
ii,.,..
57
58
59
60
61
62
61
64
55
66
67
be
69
70
71
'2
73
74
75
Ih
77
78
79
80
KMR
!X,
MU 01 R A D I A N S
CONOI'IOH
YO
AXIAL
FORCE, K N
gles with long legs back to back for both top and bottom chords, find the an
PROBLEMS
The following problems are of two 'types. The first four problems are for the student to obtain
familiarity with using the computer program given in the Appendix to solve structural problems
(other computer programs, such as STRESS or STRUDL, may also be used, but the program in the
Appendix is likely to be considerably faster to run). Use E = 200 000 MPa or 29 000 ksi. Obtain the
solution in either fps or SI units, as assigned by the instructor.
One or more of the last three problems should be used for the design projects to be carried
along with other later chapter problems.
21 Given the following beam. Obtain a W section that limits the deflection at point A to 1 in or 25
mm. Note that all NPE(2) can be made NPPI.
Answer: MA = 171.6 ft kips or 236.2 kN . m.
w=
4 kipsift or 60 k N / m
sok
2.4
200 k N
Wind
Figwe PZ1
22 Given the following academic beam/frame. Find a W section that limits the'deflection at point
A to 0.5 in or 12.5 rnm.
110
BRlDGE DESX
30 rt 3.0
29 t 0.5
13.75 r 0.5
7.5 r 0.5
20.0 t 2.0
3.25
6.0 C 1.5
f
g
h
9.1 r 1.0
8.8 r 0.20
4.20 r 0.15
2.3 r 0.2
6.1 r 0.6
1
2.0 + 0.5
for the analysls program for the wheel loads at selected pos~tlonsalong the span. It is sugg
use wheel d~stance~ncrementsof 5 f t or 1.8 m.
'
1. The knee brace is plnned to the column, but the column is contmuous to the truss.
2. The lateral crane lmpact load wdl be applied to the column at the attachment of the runway
guder to the column, as shown m the figure.
3. An optional vehcal member is used to reduce the K L / r of the bottom chord. If you use h s
member, it should be pinned to the bottom chord, but the bottom chord is continuous across this
connecuon (thus wll have bendng).
4. Sketch where you would place w n d bracmg, but do not design.
27 Given the hghway bndge truss shown in Fig. P27, make a computer analysis for the s
truck loadmg ass~gnedby the mstructor. Use the followmg dimensions and mtial data:
Dimensions/section,
fps:
25 r 3 ft
3.
Members
Bottom chord (1,4, . . ,21,25)
Top chord (6, 10, 15, 19, 23)
Vertxxils (3, 7, . . . , 20, 24)
W8 x 48
2 C12 x 30
W8 X 40
SI:
7.5
+Im
NTRODUCTION
.
,.
l i e STRUCW
STEEL BESIGN
h we obtaiq
60P3 = 60%
e, =
S o ~ u n oWe
~ must apply a factored load of sufficient magnitude t
0.8 P3(48)
 P3(60)
0.64(29 000)
1.0(29 000)
velop f,
Fy in the three bars. At this time, the bar forces are simply
P I = A, Fy
P2 = A2F,
P3 = A3F,
P I + P2 + P, = P,,,,,,,
P,,= 36(0.64) + 36(0.75)
or simply P,
+ 1.25P3 + P, = 30 kips
P, = 30/3.05
= 9.836 kips
///
12.295
l 'ow let us reconsider Example 31 using "plastic analysis" in the following
exnrple.
n hat are the bar forces when all three bars have yielded?
Within material homogeneity and rolling tolerances, as well as practical measuring limitations.
120
proportions are such that the section can become fullv ~ l a s t i cbefore the nnce
strain hardening (i.e., .depth/web thickness and flange width/flange' thicknes
not too large).
moment
in t
W e ' will now investigate
 .    rnnrent
*vy.
 in detail the Plastic
following several paragraphs. Referring to Fig. 31, the moment at initial yield i
A
A..
.LLW
My = S,F,
where S, is the section modulus, I / c . The moment of inertia I and the distanc
from the neutral axis to the extreme fiber c are as in any mechanicsofmaterial.
~ ~C F C P Fshnwn
textbook. The plastic moment M...Y , bv ins~ectionof
  the
 S..".. nn t*. ~ e
cross section in Fig. 31 with a fully plastic section, M = Mp and noting that the
neutral axis at this point divides the area in two narts
avo3
  with
 rliqtnnre /17 t* nv u.vu
centroid from neutral axis, is
..
 
The value of A7 is called the plastic modulus, Z , so that we may rewrite the
moment as
3 ,
Mp

= Lr,,
Example 33 What is the section modulus S,, plastic modulus Z , and th
shape factor f for the rectangular shape shown in Fig. E33?
'
..
* a
The plastic modulus and shape factor for a W shape can be computed
manner similar to the rectangular shape of Example 33. Here convenient use
made of the tables for T shapes, as illi~stratedin the following ex,mple.
Example 34 Compute the plastic section modulus and shape factor for
W610 x 241.1 rolled shape.
SOLUTION
The value of A/2 is readily obtained from the WT table CVT30
X 120.5), since this T shape is made from splitting a W610 shape. The
value in the table also locates the center of the area of the T but is with
respect to,the flange.
From Table V18 of SSDD, obtain
I I ,
I'
I
c
0
0
xr
I1
iY
From Table V3, the depth of a W610 x 241.1 is 635 mm. The total
area = 30.77 X lo' m2.
&,
rnrn = 8.498 m
Figure E33
ll
=I  =bh3= .  bh2
c
12(h/2)
6
m3
2
m3 and
The value given in' Table V3 for 2, = 7.659 x
discrepancy is due to the extra digits used by the computer in computing
directly as opposed to rounding for Table V18 and the use of 0.497
0.498 above.
From Table V3, the section modulus of a W6 10 x 241.1 is
S,
6.78 x
m3
200(0.4)~
= 5.333
m3
ELASTIC,
BEHAVIOR OF STRUCTURAL
Z bli2/6
/=I50
Z=
h( b/12  b,/iZ 1
/ ' = L.SI1
( bh2  
bill,?
(bli'
b,/i:
Z = b/iZ/3
f = 2.00
esign method. The current elastic design procedures as found in the several
esign specifications is based on the linear stressstrain response tb the elastic
mit, but there is implicit recognition of the steel behavior beyond the elastic
limit. Elastic design as commonly used places the limiting steel stress as the yield
that used the limiting stress of F, in the elastic design
safety factor F = 1. A,.safety factor of 1 is unacceptble, as it allows for no future changes in structural use/occupancy, or for
material properties (flaws, under dimensions of sections, and minor mztallurgj
z = 4r3/3
/=I70
J
= + ( r i  r j )
ri
32 =ri + r j
= from tables
f = Z / S = 1.10 t o 1.18
Average = 1.1 4
Figure 33 Plastic section modulus and shape factor for selected cross sections.
= f(SF,)
Now if we take A S / S
AR/R
F, we obtain
Steel design may be based on the yield strength (termed plastic, or limit states,
design) or on an elastic design. In limit states design the analysis proceeds based
on assumed plastic behaviorthe member continues to strain from E, to e,, (see
Fig. 13b) with no increase in load. Elastic design allows for this unique
behavior of steel but limits the working stresses to the elastic region of the
This value of F is taken as the basic value of F for use in the elastic design
ethod for structural steel for structures other than bridges for highways and
railroads. Railroad and highway bridges are generally subjected to a more
overloading, so the un
hostile environment and a greater possibility
factor A S = AR is taken as 0.29, which gives F = 1/0.55 = 1.82.
The value of F = 1/0.6 ,is modified to 1/0.66 when the crosssection
geometry is such that a plastic hinge can fully develop at the most high
stressed point. Rolled shapes whose section geometry is such that the plast
hinge can fully develop so that the basic value of F = 1/0.66 can be used are
termed compact shapes. The geometry criteria for these shapes will be considered
in Chap. 4.
For A36 steel the basic allowable stresses using the previously defined
safety factor becomes
Fa
EFLECIIONS
uld a plastic hinge develop at a point along a beam or column, a very
ection would result. This deflection would, however, have no meaning
ould result in a structure collapse. No structure is
OF STRUCTURAL
r v
137.5 MPa
We note that the optional rounding of 21.6 ksi to 22 ksi (as should be done using
desk calculations) can create a slight computational discrepancy if a digital
computer is used, unless a rounding procedure for this grade of steel is set in the
computer program. The author suggests that the rounding to 22 ksi be done,
since it is allowed and A36 steel is the most common grade used. It is not
recommended (at this time) to round 137.5 MPa to 140 MPa, since some
rounding up has already taken place to obtain 250 MPa from 36 ksi.
2hlP a
MP
F,S
l l l i l l l l l l l l l l l l l l l l
fMY
F,S
This value of F is used in plastic (or limit states) design as a load factor
which the working or design loads are multiplied to obtain the "ultimate'
beam.
My = S,F,. Since y = Mp 
When f
=,
Example 35 Derive an expression for hfp for the fixedend beam s
and select a W shape with adequate Z for P,,, = 120 kN.
P, = 170 k N
SOLUTION
Three hinges are necessary to produce a collapse mechanism
Note that the beam is indeterminate to second degree (no horizontal load
so Z F, = 0 has no significance). From symmetry the three hinges necess
to form the mechanism must be as shown in Fin.
 E35. The effect of
fixedend moments is to reduce the simple beam moment diagram as shown
byi,the dashed lines. For hinges to form it is necessary that the moment
val'ue be Mp, and it is evident that Mp will form first at the fixedend
locations, since the elastic moment is largest at those points. Further
increases in moment increases the elastic moment into the plastic range. It is
also evident that the only other possible location for M, is under the
concentrated load, since the moment at this point will be the next location
where the elastic moment is large enough that increases in Pw to P, will
orce the moment into the plastic range. When this hinge forms, the
. !ructure cbllapses (theoretically) and no further increase in load is possible.
With this consideration, we have (again referring to Fig. E35)
m3 furnished
O.K.
Pu = Pw X load factor
= 120(1.7) = 204 kN
,,q
Zx = 0.6566 x
m3
The beam must cany its own weight, so for selfweight the simple beam
moment is M = w ~ ~ /For
8 .plastic analysis use the same concept as for the
concentrated load, which gives
W360 X 38.7
Example 36 Given the propped cantilever beam shown in Fig. U6, it is
required to obtain a general expression for itl, and design the beam if
ww = 5 kN/m and F, = 250 MPa. Also derive a general expression for the
location of M, in the span.
ELASTIC,
2 R,x
0
Taking moments of beam segment AC about A ( Z M , = 0) gives
M~ +
Mp
+ R,L;=Ow L2
2
Substituting Eq. ( a ) for Rc in Eq. (c), then substituting Eq. (c) into Eq. (b)
for Mp, we obtain
x 2 + 2 x L  L* = 0
Solve this by completing the square, we obtain
Figure E37
Right span:
x = 1.414L  L = 0.414L
Now a general expression for Mp can be obtained from Eqs. ( a ) and (b):
Mp
0.08579~~~~
Using Eq. (34) with the given beam length and loading, the value of Mp is
M, = 0.08579(45
1 . 7 ) ( 6 )=
~ 236.26 kN . m
z,=236'26  0.9451 x
250
The maximum value of M, from either span is used for design (since beam
runs across both spans using a constant section).
usethis
beam weight is
AZ,
0.5 1
45
(0.945 1 )
0.0107
=
0.9558
!< 1.0861 x
m3
O.K.
bracing requirements).
From Table 112, select a W24 x 68 w ~ t hZ,= 176.4 in3. Check tde beam
weight effect as approx~mately (if a borderline case is found, one may be
justified in the additional work for an exact analys~s):
AM'
w, L~
AM; + 2
2 = 8
Example 37 Given the twospan continuous beam shown in Fig. E37,
select an economical W section using plastic design and A36 steel.
I
AZ,
= 170.0
Total Zx(reqd)
2.17(12)
36
= 0.72 in3
O.K.
occupancy; other values are also used (e.g., 1.5 for maximum sn
case
+R = $(FdD
+R 2 l . l ( l . 1 0
+ 1 . 4 1 + 1.6PVm,,)
+R
l.l(l.10
l.l(l.10
+ 1.4L)
+R
>
1.5Sm,,)
The general objective with LRFD is to assess each item that influences
design of a structure rather than "lumping" several effects together, as,
example, simply adding the dead and live loads to obtain the composite lo
Larger factors are used with those items that carry more uncertainty, such
snow and wind loads (liveload factors of 1.5 and 1.6 versus the deadload fa
+ FLL)
factors
Stress condition
Suggesteda
Canada
AASHTO
~ & s i o nmembers
Yielding (4)
Fracture (F,)
0.88
0.74
0.90
0.90
1.O
sending
Rolled sections and plate girders
0.86
0.90
1.o
0.90
1.o
1.0
Columnsb
TJ 2 0.16
0.16 < 7
1.0
TJ > 1.0
:!hear
Webs of beams and girders
0.86
0.90
1.o
Cvnnections
BoltsC
'Welds
0.70 1.00
0.80
0.90
0.90
0.86
0.900.25~~
0.65
" See Journal of Structural Division, ASCE ST9, September 1978 (contains eight
Fapers on LRFD).
7 = ( K L / ~ ~ ) V F (~K/ =E length factor as given in Chap. 6).
See Sec. 810.
Metal Structures (New York: McGrawHill Book Company). or Johnston, Guide ro Srabiliry Deri2n
Criteria for Metal Structures, 3rd ed. (New York: John Wiley 8i Sons. Inc.).
,
7
136
where
ELASTIC, PLASTIC,
F,,=
+,
rC
:p
9
u7
h'[
i h I
i =O;Y
k c = I 1 5 ,re111
FIdnze ,dnle >I.
it
,h.111=
k; = 0 63 l l ~ n g e \
k: = 4 5 for web
J,
$,
Figure 37 Compress~oncharactens~csfor rolled shapes shown. Note that 111s generafly nec
to mvesbgate the cntical b / t ratlo, which may be as shown for a W shapc w t h a cover plate
welded or bolted.
Fcr = X E ~ ; ( ; )
13
Fcr
=
0.1875
c2
4  Fcr
Since Fc, is on both sides of Eq. (39), we must solve for Fcr/X by trial. Once the
value is obtained, this can be used in Eq. (38a) for the ratio of b / t , which is
~sually,the item of interest. From Eq. (38a) the b / t ratio is
Lt
'9
For kc
= 11 3
=pC
65
say 
vz
Fcr / A
3.615, X = 1, and Fc, = 0.754, the limiting b / t ratio for A36 steel is
",/@=Em
t
0.75 x 36
= 62.3
I t is often useful in using Eq. (39) to set up a table of A vs. F, with values from
FCr = 0.754 to F,. For A36 steel, typical values are as follows:
/
i;,,ksl
/F
27.,
28.0;
2ii 0
33.0
36
E,, ksl
1.00
0.922
0.741
0.407
0.0
29 M)O
26 738
21 489
11 803
0
say 11
k,n2XE
12(1  p 2 ) ( b / l ) 2
138 s d u m
STEEL DESIGN
139
I is usual to assume that the four plate edges are simply supported in shear and
t e sha'ar stress F,, = </\/5. This value must be combined with the safety
factor of 1/0.6 so that the design shear stress becomes
. . ."
C
r7
C
crs
l y
 'y
Fs =
SFX*
1.67xfl
2.89
Mo~t~practlcal
steel design problems consider either buckling in compression 6r b'uckling in shear. w h e r e both stresses act simultaneously, the reader
should consult books such as those by Bleich, Johnston, and Timoshenko and
Goodier cited in an earlier footnote.
a
where all terms have been previously identified except be, which is shown in Fig.
38. For a very long, thin, simply supported plate, it appears that the theoretical
value of k: is 3.615 for this equation. The use of postbuckling strength of plates
is not often directly evaluated. It is used more often in a more indirect manner;
for example, AISC allows use indirectly via Appendix C3, which states: "When
the widththickness ratio of a uniformly compressed stiffened element exceeds
the applicable limit given in Sec. 19.2.2, a reduced effective width, b,, shall be
used in computing the . . . ."
PROBLEMS
3.1 What is the allowable b,/2t, for any W section using a steel with F;. = 50 ksi?
Answer: 9.2.
3.2 What is the allowable b,/2!, for any W section with F, = 250 MPa?
33 What is the plastic moment capacity (kN . m) of a W 9 2 0 x 200.9 section for Fy
Answer: 2865.2 kN . m.
34 Verify 2, in Table V  3 for a W 6 1 0 x 241.1 rolled shape.
345 MPa?
35 What is Z, for the geometrical shape shown in Fig. P35 if it is used for a beam?
*
bending for the span and loading shown in Fig. P36. Use
4 kipa't't
37 Select the lightest W section to sorisfy bending for the span and loading shown in Fig. P37. Use
Fy = 250 MPa and plastic design.
Answer: W 6 1 0 X 101.2.
340 k;\'
50 kN1m
+7 in
8 IT]1
FIgwp P3.7
38 Select the lightest W section for bending for the beam shown in Fig. P38. Use F, = 345 MPa
and plastic design.
>

.iiii
#!fI
(n
(hi
 ..
"...,.
I...
U l L L b
ULOlUIY

inoment together with the loads can be used to backcompute the critical shear.
on mechanics of materials:
For bending:
For shear:
The section is checked to ke sure that it is adequate to carry its own weight,
and finally the working load deflestions are checked. Sometimes deflections
2. The hghtest section in the group is at the top and the heavlest is at the
is
"4.
.
/DESIGN OF B
Lb = 0 < LC
rind use Fb = 0.665 = f x 36 = 24 ksi.
Rearranging Eq. (4l), the required section modulus with strodg axis
bending is
We must find a section with a somewhat greater value of S so that the beam.
,
weight can be carried. From Table 111, select
furnished
<
S,,,, = 138
furnished
If we use this beam, the actual fb= 24(140.45)/140 = 24.077 ksi, or this
beam will be overstressed 77 psi based on the given load conditions. We
may:
1. Use engineering judgment that this small overstress is acceptable, or
2. Use the next lightest section, which in this case is a
W24
With Sx so much larger and beam weight the same as in the earlier
section, it is not necessary to again check the beam weight.
The reader should observe that from a deflection stand~ointthe W24 is
the better selection.
///.
Example 42 What size of beam should be used in Example 41 if the beam
depth is limited to 16 in?
Sreqd is approximately 145 in3, since the weight will have to be
SOLUTION
larger than 68 lb/ft of the most economical section selected in the example
Always use the maximum depth possible when the beam is laterally sup
:;ported. Therefore, by inspection of the W16's in Table 13, select ,
SOLUTION
First, draw shear and moment diagrams.
Second, obtain M,,, = 436 kN . m. Assuming full lateral support, w
obtain
Fb = 0.67 X 250 = 167 MPa
The required section modulus is
Try
W16 X 89:
W 14 x 90:
W12 x 106: S
145.0 in3 d
12.89 in
> 16
N.G.
(next try)
Note that 167 MPa x lo3 = 167 x l d kPa, which is consistent with 436
kN . m. Check Table VII for Sx somewhat larger. Find:
154
Col.
r
S,,,
= S,
+ AS,
= 2.611
Col
>,
:!.,.
..
cur
.,,.,
..
O.K.
m3
major difference
Continuous beams are designed similar to simple beams.
when
using
AISC
specifications
is
that
if the section is compact and is not a
"
,
. ...
cantilever beam, the section may be designed on the basis of using either:
1. 0.9 X largest negative moment in span, or
2. Positive moment based on maximum positive moment from moment diagram
0.1 x average of the negative span moments,
+ +
+ M for design is
+ M = 1i1.3 + 57.86 +2 140.6 (0.1) = 121.2 < 126.5kN  rn
Since the largest moment is  126.5 kN . m, this is used to select the sectionNote that the compression flange is part on the top and part on the bottom,
with inflection points as shown on the moment diagram of Fig. E44.
Tentatively assume that LC will be larger than the values for the bottom
flange; if so, no lateral bracing will be required and we can use Fb = 0.665,
SOLUTION
From the computer, output obtain beam end moment values of
r,,
,The
156 W U C l V R A L
STEEL DESIGN
midspan brace to the bottom flange for the center span, to produce
& = 2'3
  1.13 rn < 1.78
2
&Check the beam for selfweight (weight = 0.45 kN/m). By proportion
( s i ~ c ethe beam loading is uniform)
t
S,,,, = S
DESIGN OF B W L I S FOR B M N G
"I
distance or column base plate width in the following way. The needed reactio
distance is obtained by considering an area in web compression defined by the
reaction length + an additional distance using a 1 : 1 (45") slope through the k
distance of the section. The section property tables tabulate k for the several
rolled sections. The k distance is measured from the outer flange face to the top
of the fillet transitioning',the webtoflange interface. At this location the
resulting web area in compression is nearly (if not exactly) a minimum. At a
reaction the area in web compression is
A, = (A' + k ) t ,
The allowable stress at this location is taken by AISC (see SSDD Sec. 110.10.1
to be
F,
At a reaction with j
0.75Fy
R I A , , we obtain
At a concentrated load in the span, the distance k can develop on both sides
of the load as illustrated in Fig. 46. For thls condition, we obta~n
where
Figure 45 Web failures to avoid in design. (a) Web crippling. (6) Web buckling.
Figure 46 Bearing length for concentrated loads on beams accord~ngto AISC spcnf~cahons.
5 SHEAR CRITERZA
The shear stress distribution across any section subjected to bending c
computed using the equation presented earlier:
When large uniform loads are carried through the flange to the we
be necessary to check the compression stress& and limit the valuedto
&
VQ
5 0.75Fy
f, = It
Example 45 What is the allowable reaction for a W16 x 40 using the basic
value of N = 3; in, with A36 steel? What column load can be transmitted
using a W 8 X 31?
A plot of shear stress using this equation is illustrated in Fig. 47. We note that
the average shear stress based on
f" = dlw
=
=
1.03 in
0.305 in
R = (N
in = N.
differs somewhat from the maximum value shown in Fig. 47 (in this case about
23 percent) but is considerably easier to compute. AISC allows use of Eq. (4.T)
for either rolled or fabricated (plate girders) sections.
The USHTO and AREA specifications simply allow computation of f,
based on the area of the "gross" section. This can be interpreted as in the AISC
specifications (i.e., f, = V/dt,).
The allowable shear stress F, for rolled sections is computed as:
+ 2k)t(0.75F,)
= [8.W + 2(1.03)](0.305)(0.75 X 36) = 82.8 kips
P = (N
F, = 0.405
~ x a m ~ 46
l e What is the allowable reactidn for a W460 x 74.4 .section
using the basic value of N = 89 mm and F, = 345 MPa? What column load
can be transmitted using a W200 x 46.1?
F,
(AISC)
(AASHTO)
0.335
F, = 0.35Fy
(AREA)
k = 27.8 mrn
tw = 9.0 mm
50
\
X: 9 ~
C114h0 X 1 0 5 . 7
///
100 \I PJ
107 4
Y rzc
\!I'd
1.7
71.6
47 Theoretical and average shear stress distribution on section for conditions shown
160
k~p\/ft
.  ,L
Figure E47
SOLUTION
Data for a W24 x 94:
dz24.31in
tw = 0.515 in
k=1.53ln
S,
222.0 in3
where B= 0.005 for fps units
= 0.0019 for SI units
wL
f,= 0.44
2twd
W L ~
f 8Sx = 0.665
From Eq. (a),
0.4(36)(2)(24.31)(0.515)
L
Substltutlng this w into Eq. (b), we obtain
=
360.6
L
/
=
47 DEFLECTIONS
360'6L  0.66(36)
L
8%
24(8) (222)
360.6 x 12
= 2(24.3 1)(0.515)
(N
.
14.4 ksi
Deflect~onestimates under working load are often required to ensure that floo
O.K.
36.61(9.85)2(12) = 24.0 ka
8(222.0)
36.61(9.85)
fv
9.85 ft
N =  R
0.K.
= WL/~=
+ k)twF, = R
t w Fa
180.3
 1.53 = 11.44 in
0.515(27)
L STEEL DESION
For L
and
=.
12.6 mm (approximately
in)
//
.,
r! .
Equation (49) is commonly used for spandrel beams and roof purli
where there are both vertical and horizontal force components. When the load is
applied to the top flange of the member (common for purlins on sloping roofs)
and is separated into components perpend~cularand pafallel to the X and Y
axes, the force parallel to the Y axis does not pass through the origin of
This is a very complex stress state that involves biaxial bending and torsion.
approximate solution is obtained as
I6
1.
The preceding sections have considered bending about either the X or the Y axi
of W, M, or S shapes. In all these cases the moment produced by the load
zpplied perpendicular to the X or Y axis with a line of action through the o
of axes. This produced the type of bending stress that can be computed usi
Eq. (41). The W, M, and S sections are symmetrical with respect to
,lane containing the X and Y axes and thus produce principal axes.
xres are such that
that is, using onehalf the section modulus of the section for the tangential
bending component and the (+) sign used to obtain the stress.
An alternative form of Eq. (49) is widely used for biaxial bending and
similar form for combining bending and axial stresses. To obtain this form,
setf, = Fb and then divide both sides of the equation by the allowable stress
(noting that this is done even if Fbx# Fb), delete the (), and reorder to ob
1%
FbXSX
+ =
MY
1.0
F b s,
Example 410 Deslgn a roof purlin using a channel section for the si
sheds and for the most cntical loadlng cond~tionof Example 25. Space the
purlins 6 ft horizontally as shown in Fig. E410 and use sag rods at the
midbay spacing, giving an unsupported length of 12.5 ft for bendin
moments about the Y axis of the purlin.
SOLUTION
Additional data from Example 25:
Dead load
Live load
= 20.0
snow load
25 cos 14.04
25 cos B
re
= 24.25
'Total
.
.?
Vertical load/ft
psi
= 44.25
psi
psf
= 6.1 8(0.04425)
= 0.273
kip/ft
>?
.!&
15
i.
{
values, we obtain
*hk
tJ
fb =  2046x  1220y
b t point B, x =  0.0419, y
fb
MPa (compression)
0.0673.
= 167.8 MPa (
= 
+ = tension)
The shear center locates the point with respect to a cross section to apply
flexural load so that no twisting (or torsion) occurs when shear stresses due t
bending act on the plane through the point. Thus. if the loading passes through
the shear center, the section may be analyzed for simple bending and shear
using Eqs. (49) and (49a). I f the beam loading does not pass through the she
center, a torsion moment is developed that produces torsional s
stresses of
Ve' t
where Ve'= shear and shear eccentricity with respect to the shear center
t = thickness of element where shear stress is desired
J = torsional constant of section, for a thin rectangle
bt
J= b/t 4 10 (section webs and some flanges)
3
At point C, x
0.1 10, y
fb =
 0.0673.
 142.9 MPa ( 
compression)
If we use fb = M c / I :
At point A :
bt
J =   0.21t4
b/r
4 (stubby flanges as for channeis)
3
The computation of the shear center is complicated for all but the sim
shapes. Fortunately, most sections have the shear center at a convenient Iocation
(see Fig. 410), for example:
rr
Using these values for a C10 x 30 with d = 10.00 in. b, = 3.033 in, t f =
0.436 in, and t, = 0.673 in, we obtain E, = 0.705 in, as in AISC and Table 16
This deflection is computed at 14.2 mm (over
/$
At midspan:
L
x=
f,
Shear enter l o ~ a t ~ o n
E,
v = o
82
20.7
At x ,= L / 4
3.13 ft:
M,
j.7 in kips
Qe reader should note that there are several equations used in various
dp,;g~ tables which give Eo values for channels slightly different from those of
d ,,!14), depending on the assumptions made for shear flow in the flanges.
The computation for shear center of more complicated shapes is beyond the
s c c ~ ~ofe this text, and the reader is referred to any text on advanced mechanics
of ?,~terials.
1' I.: bending stress for conditions where torsion is developed by the applied
lo22 * ,t passing through the shear center is approximately
"
_.
,,
= 0.3
Note that if the channel beam were designed without considering torsio
and shear center location, the apparent bending stress at rnidspan indicates
that the section is considerably overdesigned. When considering the sh
center effects, the section is not adequate unless it can be assumed that
analysis is too approximate and that the end connections are such as to
restrain rotation so that the larger bending stresses do not develop. The
author would assume that the section is considerably underdesigned and use
a different section, as there is too much design risk involved for the small
amount of savings obtainable from using a lighter section.
The torsion shear stress can be evaluated as
Ve' t,
L=,
B r ~ i kv e n e e r
I+
Figure E412
I 2 45'
= 21,
Eo = 0.705 in
b, = 3.033 in
0.927
+ 0.152 = 1 08 1n4(vs
1 22 In SSDD)
///
= shear modulus =
= I 1 150ksifory=0.3
ll",,. ,
'
hen*
or, alternatively,
where
J=c,t:orsion
constant previously defined and given in the table of
ib
:'section properties in SSDD
C, = warping constant =$ t,bf3(d  t,)2
~h~ AISC specifications can now be developed using the preceding several
ations and some additional simplification and rounding of numbsrs. Since
ut 1946, the AISC specifications have used the following formula based On
g a safety factor of 1.67 and a factor Cb to
a c  r ~ n tfor moment gradient (the latest modification) into Eq. (421), to obtain
Fb =
Fb =
12 OOOC,
Ld/A j
82 700Cb
MIAj
2 0.6.
(SI units)
The minimum L/rT ratio for using Eq. (426) is found by equating F,
.6Fy to obtain (with slight rounding):
Substituting this value and using 0.667 instead of 0.6 for the first term (1.0) and
approximately 0.75 for the second term in Eq. (424), we obtain
F,
(426)
+ [Eq. (1.57)12
2
Eq. (1.56b)
M = 204P
1. Initially assume that Fb = 0.6Fy and make a tentative section selection using
a table such as Table 111 of SSDD, which also gives LCand L,. If the actual
unbraced length Lb I LC or L,, a direct solution can be obtained since the
unbraced length will not be a factor.
2. If the tentative section indicates L, < Lb, the unbraced length.may be a
design factor. Using the tentative section or one somewhat larger, compute Fb
using Eq. (423). If this equation gives a value of Fb that satisfies bending, a
so:ution (but possibly not the best) is obtained.
3. If Eq. (423) does not supply a (satisfactory) solution, the designer must use
either Eq. (426) or (427), depending on the L / r , ratio. Use the largest F,
from either Eq. (423) or from the controlling equation (426) or (427) as
determineckby L/r,.
+ 952.2 in . kips
From Table 13, obtain the section properties for a W36 x 300:
16.655 in
r,
1.680 in
Example 413 Given a girder using a W36 x 300 supporting two columns as
shown in Fig. E413, what is the maximum column load using the AISC
specifications and A36 steel? Assume that the girder is restrained against
rotation only at the ends. The columns may provide some lateral restraint
m
"
rli
(0.03 x 35 + 30)(0.1)
35
16.6 ksi
0.09 kips/ft
M
S
952.2
204P
M due to beam
Fb + M = SFb
f b C  =
85.6 ki
'"he preceding example was easy to check, since the beam size has been
s e l e c ~ ~and
d it is only necessary to determine the allowable bending stress.
most design situations the problem is more of an iterative process, in that wh
th? loads.~w,igiven,we do not know what section wi1l:be
011c;:casion: one may use charts such as those in AISC; which give the allbwa
mom:nt fbr several unbraced lengths or, alternatipely, computergenera
ta&s such as Table 113 or VI3 of SSDD, which giye'the a1
se!ccted shapes for several unbraced lengths Lb. We\should also observe that
Eq. (423) controls the design, the use of A36 steel is the 'most 'economic
so!:~!ion (the reader should verify why this is true).
Example 414 Given the laterally unsupported girder
for a crane runway in an industrial warehouse. Select t
that also limits deflection to L/360. Use A36 steel.
trolley travels on a 90lb railroad rail fastened to the top of the flange
(90 lb = 90 lb/yd = 30 lb/ft).
=
= 237 ft
M,,,,,
16.6(1110)
P = l7 473'8
204
. kips
,.
..
.
.'.
1 x 296:
14 x 90:
W16
S,
131.0
L,
39.9
S, = 143.0
L,,
34
L,
28
100: S,
W18 x 97:
175.0
S, = 188.0
L, = 24.1
.r
21(143)
==250ft,kips>237
12
O.K.
,,
,,,
.., .,..,..,.
= 3ok
Figure E414
S ~ L U T IFind
~ N the maximum moment. Write an equation for M in terms of
and take d M / d x = 0.
[15x + 15(x + 6)][L  ( x f 6 ) ]
M =
L
dM  0 = 60x  270 + 30L
x = 13 ft from left end
dx
M = (30 x 13 + 90)(35  13  6 ) = 219.4 f t . kips
35
Aload
15(19)'(1612
3 EIL
1
= (47.80
3 IL
=
15(13)(16) ( L Z6EIL
131 
162)
+ 43.03)(1728) = 1.5 in
'('2)l.2<l.jin
360 3 360
Use a larger section; a side computation indicates that a W18 x 97 does not
2
3
4 or more
100
90
75
Table 41 Wheel load distribution coefficients for AASHTO bridge design
used as S/coefficient where S = stringer (beam) spacing
1 lane
concrete floor
steel grid deck
steel grid deck
+ lanes
( ) = SI value
5.5 (1.676)
4.0 (1219)
5.0 (I 524)
7.0 (2.134)
4.5 ( 1.372)
6.0 (1.828)
Note: For exterior stringers use statics to o b m n effectwe wheel loads causmg
bending and shear (refer to Fig. 4 15).
Bending moment in transverse floor beams w~thoutstnngers:
With concrete deck
With steel grid < 4 ~n (102 mm)
> 4 in (102 mm)
6.0 (1.828)
4.5 (1.372)
6.0 (1.828)
.P, LN
M =  S + 2 P , f t . ~ p ~ / f t M=9.74 r,
32
<f
Use M ' = 0.8M if three or more stnngers.
>, am*
I
*..*
""
488
DESIGN OF B%W
S ===
F,
137.5
591
 9.01(5.79)2
Y*
8
The deadload shear is
4'44/2(5.79 2
.m
1.15PwD
s
1.15(110/80)(80
x 4.448)(0.76) = 280.6 kN
1.524
with P / 2 placed on each rail. The liveload shear is
=
Pw _ 280'6
=
2
140.5 kN
3(5 79)2
30.5
= 79.5 percent
+
40 0.76
150
The design liveload shear is
140.5 x 1.795 = 252.2 kN
The design liveload moment is
=
ML
280.6(2.177 x 1.795)
2
548 kN
c,,
1'
+ 548 = 591 kN . m
'd
= 28.3
a
"
&
'L
+ 252.2 = 280.5 kN
4.30~10~m'
m3
d
t , = 13.8
> 8.50 mm
758 mm
We note that the weight is 0.08 kN/m larger than assumed, but the sec
modulus is more than adequate for this small difference. Check the sh
f
t, = 19.3 mm
0.555, since th
WL2 + track
Md = 8
FO
"
 V dtw
280.5
0.758(13.8)
O.K.
(The end connections to the plate grder wlll be designed in Example 89
Use a W760 x 160.7/ 1.58 rolled section.
~i
190
a*",:
STRUCTURAL STEEL DESIGN
DESIGN OF B U \ S FOR B
The two basic construction methods for producing composite beams are:
1. Shored construction. The steel beams are put in place and the formwork for
the concrete slab is added. This assembly is then shored (braced or propped)
so that no (or relative small amounts of) deflection can occur, and the
concrete is poured. After the concrete has hardened for'about 7 days (about
70 to 75 percent off,' obtained), the shoring is removed. At this point the
stresses in the composite beam are due to the dead weight of the steel beam
plus a proportionate share of the concrete deck.
2. Unshored construction. The steel beams are placed and formwork (metal
decking may be the necessary formwork as in Fig. 416) supplied for the
concrete deck (refer to Fig. 417). The concrete is poured and at this time the
steel beam carries the dead load of steel, formwork (as used), and the
concrete. After the concrete hardens, any formwork is r
at this stage of the construction that the steel beam h
stressed with the weight of the steel beam plqs a proporti
weight of the concrete deck.
II
.0
(,'I
i =
*pJl#
se srn.~llestvalue ol b
b = LII
b=b,+i
AASHTO ~ = I s + ~ I 2, I
h=b,+ 16i
\\SIITO h = I:[
'"
2000
2500
3000
3500
4000
5000
6000
13.8
17.2
20.7
24.1
27.6
34.5
41.4
2550 @SI)
2850
3100
3400
3600
4000
4400
The actual bending stresses in a composite beam are limited by the materi
Shored
!
b 4
Use
is
fq I = 2
4
= 33.07
8
use 34 connectors
< 81 (maximum)
Step 7. Check the shear with both dead and live loads actingV = 30.96 klps
V
j==
dt,
30.96
= 4.9 ksi << 0 . 4 5
1g(0.355)
O.K.
~
=~ W
= L! 5(0.13X8)(3@)(12~)
= 0.32 in
384EI
384(29 000)(2056.1)
L
<360
b/ = 2033  210 =
+
1823 mm
2
The slab dead load, including future wearing surface based on D = 230
The design moment for the deck slab using strength design is
M,,,
M,
40 rnm haunch
v
Cross frames
. . . ..
.'I..,
77.200
4 (? 5550
of concrete, is
(0.230 + 0.040)(23.6 kN/m3) = 6.37 kPa
Increase 10 percent for roadway debris, snow, etc.:
6.37 X 1.10 = 7.0 kPa
The deadload moment is computed as
use 1 = 0.30
L 38
Since the s / L = 2.033/22.4 = 0.091 is so small, oneway principal slab
reinforcement perpendicular to the traffic flow will be used in the slab
design. For this case and five spans providing slab continuity across the
JY
= 0.40
kN/m
=3.60 kN/m
2~iscellaneous,
including debris, formwork, etc.
Total
(I I!
= 1.00 kN/m
= 18.00 kN/m
15
= 0.25
22.4 38
From Sec. 19, the liveload moment due to total truck load on span of this
length is
b = 0.42 + 1.613 = 2.033 rn
b = 0.42 + 161 = 3.86 m
Use b = 2.033 m; for f,' = 27.6 MPa, n = 8.
The distribution factor for live load to a stringer of a bridge with two or
more lanes is obtained from Table 41 as
Factor
1.676
1.676
Slnce 1.213 wheels = 1.213/2 = 0.6065 axle, the adjusted (for deck1 [slab
action) stringer bending moment is
It is usually sufficient to select a steel beam one or more sizes Iess than
that required for noncomposite action. The section required for noncomposite action is approximately (and flange laterally supported)
s x =  = MT
0.555
Tentatively, try a W920
The maximum shear occurs when one of the rearmost truck wheel's [144kN
(32kip) axle] load is at the beam end:
1129 + 1090.5
0.55(250)
16.14 x 103
m3
.*
364.6/3.58:
Ix=6701.3x106m4
S,=14.67X103m3
t, = 34.3 mm
This value must also be adjusted for plate action and impact:
b, = 419 mrn
A = 46.52 x
m2
<< 2
3
O.K.
$>
'
I*.
i..)
sgF
6'
'
snnL DnsmN
AY = ZM*.,
A = 46.52 b'd'
= 46.52 + 0.254(215) = 101.13 x
Final:
I top '%=!!??!=26.1Mpa
Stop
41.77
26.1
8
f,==
O.K.
m2
The section stress profiles are shown in Fig. E417d. Check 100 percen
overload stresses against 150 percent of allowable stresses:
f, = 2(1091) 
f,,,,,,, =
106.2. MPa
20.54
77.0 + 106.2 = 183.2 MPa
<
1 5(0 55F,)
also O.K.
Figure E417c
= 6701.3
+ 46.52(0.327)~x
Id
Dedd load
Lnc. l o ~ d
Figure E417d
= 6701.3
= 16 121.7
m4
""P
= I31
SbOt
l6 121a7
+ 40 215 = 41.77 x 103
l6 121'7
= 20.54 x 103
458 327
m3
Initial:
f;=f,==
M"
S,
m3
I 129
14.67
 77.0 MPa
5130 kN
A F
v , = x =
2
5815 kN
@6
DESIGN OF B W M S FOR B
1.
.v
95
1.
Ka
R .A
Figure W17f
I'
22.4Ra
4078.8
R,
182 kN
0 . 4 d 2 m
V = 182(0.6065)(1.25) = 138 k N
E=4 7 4 0 e
N, =
5130
= 38.7
132.6
4
use 39 (multiples of 3)
I
0.5022 mm
0 .K~
Using the spacings above, the first 3.44 m requires 1 1 groups of studs,
next 3.43 m requires 9 groups, and the next 4.33 m requires 9 zrou~s.for
total of 29 groups at 3 studs each, for a total of 87 studs vs. th
for ultimate shear considerations.
Step 5. Design the diaphragms.
Lateral load:
Wind at 24 kPa X exposed surface area and, referring to Fig
E417a: H = 4 6 0 + 225 + 270+ 916mm = 1.871 m
Curb load at 500 lb/ft for side opposite
Wind
0.323(%)
= 132.6 kN
. The minimum number of studs required from the end to the middbn
tance of L = 22.4/2 = 11.2 m) is
24(1.871)
Total
Note: there is no wind load from the truck because the deck acts as
diaphragm.
With diaphragms spaced at 5.55 m,
Pdiaph
= 52.2(5.55)
289.7 kN
(interior diaphragm)
 289.7
W530
$rk@yproportion since Q / I = constant,
65.5/0.64:
d = 525 mm
1, =
8.9 rnrn
A = 8.39 x
r, = 32 mm
Y to
> j
> 8.0 mm
O.K.
O.K.
m'
L = 2.033 m
1
Flgum FX18
SOLUTION
Since the girder is symmetrical on the floor plan and
are closely spaced, assume that the loads will give a uniform girder I
bv, =
+(FdD
+ FLL)
Using load factors glven in Sec. 37 (and Table 31) and noting th
no live load reduction for floor area since the load factors are the
statistical terms, we obtain
3

I
w,
l.l(l.815
,w
.
wUL'
4.7 1 (25)2
Mu== 367.97 f t  kips
8
8
(4 = 0.86 from Table 31)
Mu = OFyZ
+ = 0.86
Mu = +F,Z
vu=
'
+( v3 )
dt,"
and Mu is computed using factored dead and live loads. This is illustrated in the
following example.
Example 418 Given the floor system shown in Fig. E418 for an off
and live
80 psf; use LRFD design and A36 steel.
Make a preliminary design selection of a W shape with depth no
factor.
By proportion,
Aw Z
0.075(142.6)
A Z = U = 2.27 in3
4.7 1
Wu
O.K.
+ F~
Use a W 2 1
v3
62 section.
v3
O.K.
,
,
212
.$r"
be
7'
DESIGN OF'BEAMSFO
gn the floor stringers, noting full lateral support for the compressi
= 250 MPa. Use the AASHTO specifications.
no twisting occurs due to the eccentricity of the brick. Use A36 steel.
Design the interior floor stringers of Fig. P419 using composite desip and unshored cons
Limit liveload deflections to L/800. Use f; = 28 MPa.
~ollowingare miscellaneous beam problems for laterally unsupported spans and
considerations.
4a A column load of 160 kips is carried across an open work area as in Fig. P423. The
unsupported length is 39.5 ft. Select the lightest W shape with the deflection limited to L
any grade of steel if A36 is not adequate. Note that you must assume your end conditio
Answer: W36 X 300. (Simply supported.)
F,
,7 +(I
417 Redo Example 416 using shored construction instead of the "unshored" construction of he
example.
46 and see if a workable solution can be obtained.
For Probs. 419 to 422, refer to the cross section shown in Fig. PC$19. The bridge span will be
assigned by the instructor (36 to 46 ft or 11 to 14 m). If not assigned, use 40 ft or 12.5 m for the span.
For exterior beams the maximum possible load due to truck is onehalf of the truck load. The dead
y6
P=
&
T
*
[II
30 5 '
i .; :. 111
t>OZ
 'lo
L\

B J . ~ pi~r:,
il
'1
i
/i
iI
I ._.I'
,
Figme P419
,~
' ,
',
, ,
218
I
I
Squdre bar
J
I
,
n all these uses the tensile' strength of the steel is used. In 'thi
uration plate buckling or warping is not a consideration. In some
s, however, specifications will require a minimum amount. of
ss for esthetic and safety reasons.
enerally, tension members may be categorized as rods and bars,
tural shapes, builtup members, and wires or cables. Several of
bers are illustrated in Figs. 51 and following.
ALLOWABLE T E N S I O N S T R E S S E S
e AISC allowable tension stress of members, except eyebars, is limited to
F,
W shape
0.6 F,
F, = 0.5FU
(5 1)
0.555
(52)
cification further limits this basic stress to the lesser value of
ut the net section is used for both these equations. For steel with
not over 80
i the basic tensile steel stress is governed by Eq. (52) for AASHTO design.
On the net section across the pin hole of an eyebar (see Fig. 53), th
owable AISC stress is
Figure 51 Tenston members. See Fig. 52 for cables used as tension members. (a) Structural shapes
used for tens!?? members. (b) Upset bar. ( c ) Threaded bar and use of a turnbuckle to adjust bar
lei&. Applicable for square and round bars.
Allowable tensile stresses for several steel grades are shown in Table 51,
ere the reader should note that established practice allows rounding of the
ues for A36 steel to the values shown for both AISC and AASHTO/AREA
In all cases, except eyebars, the tension stresses must be computed based o
both the gross and net crosssectional area when using AISC specifications. Ody
the net area is required for AASHTO and AREA specifications. The net area is
the gross (total) area where welded connections are used. The net area is the
least effective crosssectional area for all other cases as where bolt or rivet holes
are used for mechanical fasteners at the ends or where holes and/or
reductions occur along the member.
The effective net area at approximately the root of the thread of thread
ension members using the AISC specification is
$+'7
1
UP
Anchor rod
(strand or rope)
53. Next in simplicity would be some kind of threaded bar or cable. Here
1
2.' STRUClVRAL
STEEL DESIGN
,$4:;
ii.
.
r
I,.
= A*
1%.
'>
., .
4.
5 81
,.. .,...,.
1.33Ab S A , 5 1.5
4
i
..
1.;
67 (and larger)
23 000
158
CCO
DESIGN OF TENSION ~ ( E . M B ~
rl
Welght
in
Ib/ft
mm
kN/m
Area
in2
pu
m2 ( X lo')
bps
kN
13
16
18
0.52
0.82
0.99
0.008
0.012
0.014
19
22
1.18
1.61
0.017
0.023
;
;
f
:
1
10
13
16
119
"22
25
0.15
0.234
0.0968
0.1510
29.0
46.6
129
207
0.284
0.338
0.459
0.1832
0.2181
0.2961
56.2
66.0
89.2
250
294
397
13.0
58
23.0
36.0
52.0
70.0
91.4
102
160
231
31 1
407
286.0
372.0
576.0
1272
1655
2562
3665
4938
3. For all other shapes, including builtup shapes, with at least three fasteners
a line (Fig. 55 has five fasteners in line), A, = 0.85A,.
4. Any tension members with only two fasteners in a line, A, = 0.75A,.
1;
2
2f
44
50
64
3
3f
4
75
90
1,00

5.24
6.85
10.60
0.076
0.100
0.155
1.47
1.92
2.97
0.9484
1.2387
1.9161
6494
. 
. 
228 S p U
C l U R A L STEEL DESIGN
..',...,..
A,
,.,,a,
..4
(0)
T ~ p i i . l ilr u , ,
\.ur<
Figure 58 Lacing and other means of producing a builtup member with access to interior.
fabrication shop practice gives preference to use of batten or perforated cover plates which are
welded. (a) Single lacing. (b) Double lacing. (c) Batten plates. (d) Perforated cover plates.
(f)
Figure 57 Cross sections of several builtup sections. General crosssection configuration is limited
only by designer's need and ingenuity and may include W and S shapes with'lacing and/or plates.
Where additional area is needed, more plates may be added to any of the above sections. ( a ) Four ;,
angles with lacing. ( 6 ) Four angles with both plates and lacing. (c) Two angles and one or two
.
( d ) Two channels with lacing. ( e ) Two channels with flanges reversed from (d) and both plate and
lacing. (A Four plates welded to form box section.
$,
'.i
4.
not occur. It is also necessary that the member(s) be constructed so that painting
of the complete member can be affected. This requirement generally precludes
use of fully enclosed box or other builtup sections with .enclosed cavities.
Instead, the builtup sections are open on one or more sides, with continuity
being obtained on those sides by use of lacing bars or by use of perforated cover
plates. Either of these configurations allows maintenance of the interior as well
as the exterior of the member, as shown in Fig. 5.8.
iccrlon
parts, making the connection sufficiently strong that no relative movement can
take place across the hole. Tension tests using plastic and photoelastic techniques indicate stress concentrations at the edges of the bolt hole which
have a maximum value on the order of two to three times the average s
f, =
DESIGN OF TENSIOK
L:
'
0.066 kiplft
p" .y;,
a<g
. , .. .
,,
it.
'
t,i
!, .". ,),<..
,;;*,.,i
>\ ,
,
,?,$
"
:.
',
14
P= 1
.03~
Vol. 89, 1922, pp. 847848) and is almost universally used. The meth
Figure E51b
Notes:
';y;;,~'~~,,
, , .v
...,
(
'1 .";Tie rod size could be reduced in the lowerhalf group of purlins for a rod
force of 6 X 1.03 = 6.18 kips.
2. A rafter, substantial girt, or other member spanning the 25ft bays will be
required to carry the concentrated rnidspan load of 12.18 kips without
excessive outward deflection. The roofing will undoubtedly reduce this
effect considerably, but the designer will have to decide how much.
3. Computations neglect the benefits of the roof in contact with the top
flange of the purlins, and may reduce the sag rod force 50 percent or
more.
\
'
"hole" on the diagonal, thus canceling that hole reduction on the net
The effective hole diameter (except for slotted holes) is taken as
AISC specifications
AASHTO and AREA specifications
$ in (or
1.5
,t,
""
, I ,
\[I/
Example 52 What is the critical net section of the hole pattern shown
Fig. E52? Take holes for iindiameter bolts. Note that this pattern is rno
academic than practical, but is used to illustrate the method of using s2/4
Use the AISC specifications for the hole diameter.
fy,
(a)
Mgure 59 Critical net section for tension members with holes. (a) One or more rows of holes w
the net section is obvious. (b) Hole pattern where net section must be found by trial.
=:+$=
+ 3 + 2(2)
I.OOin
+ 0.333 + 0.333 =
10.42 in
Pallow= A,F,
P
510 Net section of angle used as tension member with staggered holes in both legs.
A,F,
7.78 in2
13.25(0.75)(0.65)
7.78(0.5Fu)
whichever is smaller.
*
2d"
angle with
F4We E54
SOLUTION
'
D = 0.75
Net width along path ABDFG:
+ 0.125 = 0.875 in
Example 55 Select the lightest single angle section for the vertical
of the side shed truss of the industrial building of Example
desirable to use the same size angle section for all the verticals,
section for all the diagonals, and similarly, top and bottom chords d s s i p e
using constant section sizes. Only the vertical web members will be d e s i g e
in this example. Use F, = 250 MPa and the AISC specifications.
!'
which
S o ~ u n o NFrom inspection of the computer output (Pa*
shown in Example 26), the following values are obtained from the tw
conditions used:
Left side
Right side
Member
4
8
12
16'
20
LC1
@pq
LC2 (kpq
0.0
31.18
62.36
93.54
124.72
Member
74
78
82
86
90
0.0
48.62
97.26
145.88
194.50
LC1 @W
124.73
93.56
62.37
31.18
0.0
157.57
118.18
78.80
39.40
0.0
Since the signs are (+), all the vertical members have only tension Iorces Ior
either load condition. Note that the computer program automatically
output with wind by the factor 0.75 so that all loads are on the same design
basis. Member 20 has the largest axial tension force with wind from the left;
wind from the right would produce this design value in member 74. Member
20 is the longest vertical of the side shed members and has L = 4.5 m. If One
of the other members had been longer, that member length
the ~ / , .  i ~this case both the controlling length and the maximum
force are in the same member. Design P = 194.5 kN.
L/,. = 240 (per AISC, Set. 18.4, and assuming that this is a main member
with such a large axial force). The minimum radius of gyration is
1
1
b*
3i
4.5(1000) = 18.75 mm
240
A preliminaly side computation indicates that the bolt Pattern !&Own
to
~ 5  5can probably be used, as L / r rather than stress is
m i s pattern may be able to use standard gage distances and
economical. For two bolts at the section and using 22mm A325
D = 22 + 3.0 = 25.0 mm
me effectiveangle area using AISC criteria for shear 1% (see Set. 53.2
A, = 0.90An
'mm
i:
't
i1
faa
I"
i
m2
must be
Based on F, = 0 . 5 ~ "(use Table 51 for F,), the effective net area
194 5
A, =
= 0.9725 x 103 m2
200
m i s effective net area A, must be obtained from a gross section of least
0.9725
0.85(0.90)
1.2712 x
2Dt
An,, = A,  Aholes2
0.9725
= 1.081 x
0.90
m'
A,
Ahole,
> 2.670
0 . ~ .
Actual hole diameter = 22 + 1.5 = 23.5 mm
Edge distance = 152  57.2  63.j = 3 1.3 > 28
g2 = 63.5
Use an L152
89 x 6.3 section.
O.K.
LC1 OrN)
LC2 GN)
L, m
39
 63.65
47
+70.18
+ 20.39
 37.60
6.76
8 2 (longest of verticals)
Member
m2 < 1.2967
P,, = 70.18
 (37.60) = 107 78 L N
' 2&
8.2(1000)
= 34.17 mm
240
Use 25mm bolts: D = 25 3.0 = 28.0 mm. Use two L sections with a
12mm gusset plate as in Fig. E56. Assume two lines of holes, since P is
only 70.18 kN, which gives A, = 0.90An.
SOLUTION
From the computer output, we obtaln the following (includes
impact in live loads):
=
+ 80.3 kips
+ 60.2 kips
Dead load
For A, = 0.75An, the gross angle area is at least
=  22.8
kips
From which
Use long legs of unequal leg angles back to back. (Why?). Again set up a
table of double angles using Tables V10 and V11 of SSDD [A, = smaller
of (A, or 0.85Ag) X 0.901:
1,
mm
6.3
7.9
'
Al
0.353
0.442
2.03
2.991
2L127 X 8 9 X 6.3
2L127X 127 X7.9
r,
mm ,A
,
37.5
39.9
(X
P,,
P,,,
= 80.3
2.660
3.910
"
p:'
,
A,
83 klps
We will assume four holes in the flange at any net section, as shown in the
insert of Fig. E57. Use iindiameter bolts, so that the effective hole
diameter D = + = 1.0 in.
Aholcs
There is nothing lighter, than these two double angles. Note that minimum
%v
L / r ' controls the design. Now checking stress range and using A,, we have
f = psr
=
mZ
107.78
= 54.0 MPa
2.03
This value of 63.6 is much smaller than any value of F,, in Table 14 up to
2000 X lo3 stress cycles for "base metal" at mechanically fastened joints.
Use two L127 X 89 X 6.3 sections.
///
The design of bridge tension members is similar to that using the AISC
specifications except that fatigue will have to be considered, as outlined in Secs.
54 and 19. This will be illustrated by the following example.
= 4(l)tj
F,,
16 ksi
We will use lateral and sway bracing across the top and spanning
between the two trusses to satisfy stability, but this wdl not reduce the
unbraced length of the several web members with respect to the Y axis.
(AASHTO requires the effective depth of this bracing to be at least 5 ft or
1.8 m.) Joint fabrication requires orienting the X axis in the plane of the
truss for rolled sections. Note that commonly the transverse floor beams are
242
{ t
,*
;"
r
7.025
7.025 in2
in2
25(1.414)(12)
= 2.12 in (for r,,)
200
We will arbitrarily use the AISC 85 percent reqmrement, so A, is at least
rmn
rmn =
 Aholcs
* $4
Y d ,
Section

+, Ln
Aholcs,
W12x53
W12~50
0.575
0.640
2.30
2.56
ln2
Arcqd,
13.30
12.14
ln2
Arm, m2
2.48 O.K.
The W12 x 53 sectlon is the lightest W12 that is satisfactory for both
area and L / r requirements. This section 1s also selected so that the connection can be more easily fabricated when the vert~calmember (No. 7) is
designed in Chap. 6.
Check the stress range for the W12 x 53 section:
Psr
A , =  PSr
=  =
A,
83.0
13.30
O.K.
tan8
+
=8 h x 4Lh + t a n ~
.
(6)
~2
Since
dr = [ I
+ ($)2]1'2
H
and
T.= cos 0 
we obtain a general equation for the tension force in the cable at any point as
Although fatigue is not a controlling parameter for this member, all the truss
the truss
members should be checked similarly. Use a W12 x 53 section.
///
Noting that all the terms under the square root are relatively imi,onif icant except
the first, third, and fourth, we may simplify to obtain
b t :
H z wL2
8h
xhere = mdspan sag, as shown in Fig. 51 1
w = uniform cable loading/unit of length;
be
.
.
.
 there will always
load caused by the cable weight
L = span length (the cable length is always somewhat longer)
43
J
When the ends of the cable are at greatly dlffenng elevations, Eq. ( 5  7 ) rather
than Eq. (58) should be used, because the tension at the upper end of h e cable
will be considerably different (it is "caqing" the weight of the cable + any
additional cable loading). For horizontal cables T has the same value at b o b
STRUCTURAL
.,
STEEL DESIGN
L
. The
. , , .. cable
..,..... ,.. ... length
T=~
[ + l16($)]
1/2
is approximately given by
S I T ~ I2nd
I > I u * r r L ~ b l pruduic
c
,riiii.ning 6ileir r o ,~s[e111
Id)
i).
Figure 512 Several conflguratlons of cables In building construction (u) Cabies spaced across span
for a rectangular buildlng plan ( 6 ) S~nglecable using a central tenalon nng/;md for clrculr b u i g
(c) Doublecable system for increased stiffness and mbration control For round buridmg p h  ( 6 )
Cables used In roof support system If cables rad~are.the towers may be esamnally selfsupprtug.
With SF
U=2(D + L + WorE)
U = 2.0 X erection loads
200
m+
120 k N
100
mL
I10 k N
Figure E58
S ~ L U T For
~ ~ the
N 100m level, using Eq. (56), we obtain
wL2
=
8h
8(110)
w h
lOd
llOkN
40 x 4
 0.088
(:r
T = H I + 16 ,
shown, and this tension will be [combine Eqs. (56) and (58) to obtain th
+tan208
h = w/0.088
T"
1
2
0 (assumed)
0.272 m
155.6
154.7
389
386
The bottom cable will carry the uniform load, producing the 23.3ft sag
of the upper cable plus the total roof load:
249
e used. Once this is done and the ultimate axial tension load P, is obtained,
k
1 1 Q V V
p,,
The required cable areas using F = 3.0:
Top cable: P,,= 3(74.8) = 224.4 kips + use I findiameter bridge
strand
From Table 52, w = 4.73 Ib/ft.
Bottom cable: P, = 3(224.4) = 673.2 + use 2findiameter bridge
strand
From Table 52, w = 12.8 Ib/ft.
The struts wlll carry a compression load based on q producing an h
value in the upper cable of 23.3 ft. This gives an equivalent uniform strut
"or diaphragm" load of 0.220/2 = 0.1 10 kip/ft. At a 20ft spacing, the strut
load is 20(0.110) = 2.2 kips.
Check the natural frequency of the top and bottom cables: First find
the cable length:
A,+Fy
= 354.13 ft
With dead and live loads in contact with the top cable and using 350 ft
as the contribution span for these loads, the mass is
With two holes out at the critical section, the net area is
Anet= A,  2(;
+ ;)lw
P U
59.6  1.88 in'
A >..=. OFy 0.88(36)
A >
"
59.6  = p u
 dF
1.39 in?
0.74(58)
Using the largest A,, the gross section using the AISC efficiency factor and
the shear lag factor gives the gross area as at least
Try a C7
* x 9.80:
A, = 2.87 in2
3'
ry = 0.625 in
> 0.575
O.K.
250
DESIGN OF m & I O N L
What is the bolt pitch in Fig. P53 so that the critical net section is at [east 205
Answer: 62 mm.
*"510.
g, = 2.50 in
ih
4,:+
,
rl
,':,
,
A > ,
and
< 2.67(7/8)
N.G.
Try a C8 x 1 1.50:
A = 3.38 in2
g , = 2.50 in
O.K.
t, = 0.220 in
O.K.
(f+ ;)
 0.22
A, = 2.940(0.85) = 2.499
Joint efficiency and shear lag:
X
2.940 in2
A , = 3.38(0.85)(0.85) = 2.44
Use a C8
>
1.88
Flgum PS9
O.K.
510 What pitch is necessary in Prob. 59 so that only three bolt holes are deducted from
to produce the net section?
Answer: 32 mm.
511 Select the lightest single angle for a tension load of 50 kips. The length is 6 ft and I bolts wiU be used as shown in the pattern on Fig. P5 11. Use A36 steel and the AISC s
Answer: L6 X 4 X f .
O.K.
11.50 member.
PROBLEMS
.For all problems, assume adequate fastener strength so that only the net/gross section requirement$,
control.
51 Design an eyebar to cany a tension load of 40 kips using a Iindiameter pin. Use A36 steel
and the AISC specifications. Use t 2 in and w in multiples of in.
,
.
52 Design an eyebar to carry a tension load of 200 kN using a 25mmdiameter pin. Use
MPa and the AISC specifications. Use t 2 15 mm and w in multiples of 3 mm.
53 What is the net width of the plate shown in Fig. P53 using the given fps units?
Answer: 6.75 in.
v, +
v,
rl
75
3 0
40
50
2"
1,s"
= 250
Figure P51 1
210 kN. The length is 19 m a d
shown in Fig. P511. Use 5 =. E O
513 Select the lightest pair of angles back to back to c a m a tensile load of 433 kN. Ux
22diameter
bolts, a 12mm gusset plate, and F, = 3 4 5 &{Pa steel. The member lenglh is 3.2 EL
Use the bolt pattern of Fig. P513 and the AISC specifications.
Answer: L102 X 76 X 6.3.
3"
40
F,
Flsurep53
54 What is the net width of the plate shown in Fig. P53 using the given SI units?
Answer: 176.3 mm.
55 What is the bolt pitch in Fig. P53 so that the critical net section is at least 8 in?
Flgure F513
' ~ 2STRUCTURAL
:
STEEL DESIGN
514 Select the lightest pair of angles back to back to cany a tensile load of 92.5 kips.
iindiameter bolts, a fin gusset plate, and F, = 50 ksi steel. The member is 8.375 ft long.
bolt pattern of Fig. P513 and the AISC specifications..'
Answer: P = 154 kips.
515 Select the lightest single angle for a tension load of 68 kips, assuming one
the critical section. The member is 7.5 ft long. Use A36 steel, the bolt pattern
and both the AISC and AASHTO specifications. Assume no stress reversals fo
Answer: By AASHTO, L7 x 4 x
A.
F!
..,. .
,.,
.,,..,.
..
Figure P515
516 Select the lightest single angle for a tension load of 220 kN assuming one 20mmdiameter bolt
at the critical section. The member is 4.3 m long. Use F, = 345 MPa steel, the bolt pattern shown in
Fig. P515, and both the AISC and AASHTO specifications. Assume no stress reversals for
AASHTO.
517 Design the bottom chord members to satisfy tension for the side shed truss of Example 25
using WT (structural tee), iindiameter bolts, A36 steel, and the AISC specifications. Assume two
bolts at the critical section is the web of the tee. The tee is continuous across the critical joint.
Answer: WT9 X 27.5.
518 Design member 5 of Fig. E57 using a W12 section if the computer output (including impact) is
Dead load = 160.5 1 kips
Live load = 77.17 kips (maximum)
Live load =  5.8 1 kips (minimum)
Use the AASHTO specifications and A36 steel.
A m e r : W12 x 53.
5.23 Do Prob. 521 using the AASHTO specifications. Do not consider fatigue.
524 Do Prob. 522 using the AREA specifications. Do not consider fatigue.
Answer: 747.4 kN.
Answer: L3;
Figure P5.20
,,
7/16.
526 Redo Example 510 if the truss member length is 14.5 f t instead of 11.5 ft.
527 Given the bottom chord of a truss using a pair of C2CO X 17.1 1 back to back with a 15mnr
gusset plate between them. Using two 25mmdiameter A325 bolts at the critical section and for a
deadload bar force of 120 kN, what is the maximum liveload bar force that is d o w e d = k g
FL = 1.67? Use F, = 250 MPa and a panel length of 5.1 rn. Assume at least thrrt fastrnrrs i n , ~ e
line of stress and the AISC specifications as applicable.
Answer: LL = 365 kN.
528 Design the bottom chord member (No. 12) of the highway bridge truss oi E w p l e 57 (refer to
Fig. E57) using a builtup section. Use Examples 67 and 83 as a guide in selecting the rolled
sections to make up the cross section. Loads: dead = 336.9 hps (tension); live mimimum == 13.8
kips (tension); live minimum = 0.0 kip. Use the AASHTO splcifications, iindiameter highstrength bolts, and A36 steel. Panel length = 25 ft, as shown in Fig. Ej7.
45 lb/ft2
..&&+ha.%
4,%W '
61 INTRODUCTION
The vertical compression members in a structure are commonly identified as
columns (sometimes stanchions in fore~gnliterature). Sometimes verti
pression members are called posts. The diagonal compression members
ing the top chord of bridge approaches are end posts. The diagonals of a truss
members used in wind bracing may be called stnlts. Short compression m
at the junction of columns and roof trusses or beams may be called knee
In all cases, however, the member under consideration is carrying a compress1
load.
A structural member carrying a compression load is termed a column if
length is sufficiently great. For lesser lengths the member may be called
compression block. The length which divides these two classifications is such
it affects the maximum compressive stress which can be developed under
load. The length is seldom used alone in describing column behavior. Rather,
a n offshoot of the Euler column fonnula developed in the next section, the r
of column length to radius of gyration ( L l r ) is used.
Material
Steel
Aluminum
Wood
60
30
10
S T R U C STEEL
~
DESIGN
AXIALLY
LOAD
COLUSD~S
<I I
1_____
A sin hy
B cos Xy
member.
undary conditions of x = 0 at y = 0, we obtain B = 0 an
x = A
sin hy
= n,
..,,..,...
dZx
EI=
(iy2
Px
'
&
Ql, b
I
This equation for the critical column buckling load P is generally called
Euler .equation (and the load, the Euler load; the stress, the Euler stress).
es of Eq. (62) by the column area A, noting that the ra
ction r =
, Fcr= P,,/A, and that n = 1 gves
(or Fcr), we obtain
7 r 2 ~
.
I
1
Fcr= (L/rI2
we obtain from Eq. (62),
. I
4n2E
Fc, =  .
( ~ / r ) ~
which is equivalent to the column containing two sine waves in the length L.
~ G i is
s called the second buckling mode, n = 1 is the first buckling mode (single
sine
wave),
and from Eq. (62) it becomes evident that the minim m crit'
. ,
b ~ c q load
g (or stress) isiobtained in the first buckling mode.
lnispection of:,Eq. (63) indicates that very large values of Fc;,'kfn b
obtai'ried by using L / r + 0. Implicit in writing the differential, .equatibn
bending [Eq. (a)], however, is stress being proportional to strain. Thus the uvv
limit of validity is the proportional limit, which is often taken as Fcr+ Fy
&,
&
Since dx/dy
cos hy  k sin ky
P
sin k L
0 at y = 0, we have
k M o 1  cos k L
P
sin k L
2 7 and with k
(P/EZ)'/', we
Using the standard method of solution as before and noting that we have simply
added a constant, we obtain
x = A sin ky
1
Mo
+ B cos ky  P
MO 1  cos kL .
sin ky
P
sin k L
x=(
+ cos I+  1
b l K U L 1 UKAL
STEEL DESIGN
.AXIALLY LOADED
F,,
T ~ E
=
(KLI~)~
differentiating obtain
Recotnineniled
d(F,,)
d(KL/r)
Rotation fixed
conditions
Rotation free
.,,.
ce the maximum value of the Euler stress or any critical buckling stress is
.to
let us also define the critical stress in any re,mion where the Edzr
tress is not valid (such as smail K L / r values):
5,
Translation free
Flgm 63 Theoretical and design values of K for columns with end conditions shown.
..
(a>
(KL/~))
Translation fixed
Translation fixed
277'E
.,...,...
F,,
CY "j7)
KL
(6)
,+)
KL
P'
d(KLlt.1
ake the slope d ( F , , ) / d ( K L / r ) = 0 at K L / r
will arbitrarily define a parameter
(c>
=
KL
=
r
Cc
lopes of Eqs. ( a ) and (c) will be equal (i.e., the two curves defined by these
equations will have a common tangent). Also, experimental column test
ta indicate that takingp = 2 is adequate. Now equating slopes at K L / r = C,
d for p = 2, we obtain
 271'~
 2 r n ( ~ , )=~7
( C,)
m which obtain m as
m=
account for eccentricity, residual stresses, and the several other factors that
complicate the theoi'y. It would also be appropriate to use a variable safety
T*E
(dl
c,*
rom rearranging Eq. (b), inserting K L / r = C,,p
.iAnwLIUn,u
.d
bltjEL DESIGN
k
i ~ ~ . for
i c Fcr, we obtain
F; = F, x
If
\k
sv
The AASHTO formulas for axlally loaded columns are derived similarly to the
AISC values, but the SF tends to be somewhat more conservative, since the
members of the bridge structure are, in general, in a more hostile environment
than building members. The AASHTO formulas are as follows:
For K L / r I Cc:
The AISC kas used the following variable safety factor since 1963:
For K L / r
for all values of K L / r I Cc. For K L / r > Cc, use a constant value of safety
factor based on using K L / r = Cc in the equation above to give
When K L / r > Cc, the Euler equation with SF = 23/12 is used to obtain
:he allowable column stress as
> Cc:
4'
F',=
n * ~
The SF = 2.12 for the AASHTO specifications and the values for the lens&
factor K and Cc are computed the same as for the AISC spzc~fications.Tab!e 61
gives Cc for the several values of F, commonly used for columns.
*
..
fps:
Fa =
149 000
fps, ksi
ksi
(KL,'~)~
b
F',
1.03
lo6
Mpa
(KL/~)~
Eqx~tlons(65) and (67) are used for main compression members. Secondary
f 6 10)
SF(KL/~)'
SI, MPa
c,
264
~h~ AREA allowable column design stress formulas are somewhat similar to the
2 CAC~;R
Allowable stress
Limitation
,

3. Make a tentative section selection (sometimes use can be made of bb!rs such
I>.
': fi
F , = O . ~ F ,  (1k.2)
#,A,
.7
(6 12)
ssnn
which
O~VPP
fnr r ~ ~ r n r oVT
l
+:\
when Po = P,,,,,.
In the AREA specifications,
F, = ksi
k = 0.75 for riveted, bolted, or welded compression
#.4
q e net area (= gross area  loss for holes) is used in tension member design.
In any connection design using mechanical fasteners (rivets or bolts), it is
assumed that the fastener completely fills the hole. This assumption is very
nearly met in riveted work, where the head fabrication enlarges the rivet shank, .
and very nearly occurs in bolted work since the hole is only about 1.5 rnrn larger
than the nominal bolt diameter. Under axial compression, although there are
stress concentrations at the hole, it can be safely assumed that no loss qf net
area occurs when a mechanical fastener fills the hole. When an intermediate
open hole is in a compression member (as for utilities, erection, etc.), the
designer must exercise judgment as to whether to use the net or the gross area.
The AISC specifications give the allowable stresses as "On the gross 'section of
axially loaded compression members . . ." Undoubtedly, .there will be adequate arching to transmit the load around $he hole if there is only a small
an'iount of area lost at any section due to the holes. The most conservative
procedure would be to use the gross section when the hole has a mechanical
fastener that fills, or nearly fills, it and the net section for all other cases.
1. Commonly, only the,W8, W10, W12, and W14 sections and rectangulrrr tube
and round pipe sections are used for columns, since the critical radiu of
gyration is with respect to the Y axis. These sections have the best 5 vrtIues
(and corresponding r , / r , ratios). In building design. practical consider1+' IORS
often necessitate use of a given nominal column size throughout the buildng.
It is usual in building construction to run a single column t h ? o u ~ at I z a r
two, and often three or more floors to avoid fleld column $ f i w . v ~ n c Iabor
savings more than offset the Increased we~ghtof steel.
2. When KL/r > C, the AISC speclflcat~onreq6lres use of Eq. (67), the
AASHTO specif~cation requlres Eq. (610). and ARE,\ has a somewhat
similar requirement for Eq. (613). In all these equations F, is independent of
F,. Therefore, In column design one should use A36 steel for all cases where
K L / r exceeds C, or the AREA lirmtation, and even if F, > 36 ksi is being
used for some of the other members. For example. if we are using F, = 60 ksi
and a section is found where KL/r > 97.7 (refer to Table GI), we should try
to specify A36 steel instead of the more expenslve 60ksi steel. The sligqtly
larger safety factor in Eq. (67) allows for the transition between K L / r = C,
= 126.1 (for A36 steel) and the lower values of C, for the higherstrength
steels.
266
< C,
60 lhlit! I I
+ 30 iblft'
rooi X 10 = I 8 k 1 0 (1
. ..
64.6 Design Examples
The design of simple axially loaded columns and struts will be illustrated in the
following examples.
Example 61 Design a column to be used in a onestory discount departstore building. Columns are spaced 20 ft on center both ways. The
roof load is taken as 30 psf dead load and 60 psf snow. This gives a column
load of 20(20)(0.090) = 36 kips. Use A36 steel (C, = 126.1).
':.merit
KL
 =14(12)
Y'
136.6
> Cc
,?&
1.23
LI.Yb
Pa = 5.26(7.99) = 42.0
> 36 kips
O.K.
KL,
KL, =
From Eq. (6'5),
///
Example 62 Design a W section for the conditions shown in Fig. E62. Use
F = 50 ksi steel and the AISC specifications.
1
I
16 f t
8 ft
Since
l ji
I
%3
This appears to be the lightest rolled section possible for this lozdisg
situation.
// /
Example 63 Check the pair of angles selected in Ertamplz 56 to be usid 23
vertical members in the main roof truss of Example 26. Use A36 stet1 a d
the AISC specifications.
LC1 (kN)
Member
LC2 (kN)
L, m
SOLUTION
Take K
&
rn'
r,,,
= 37.5 mm
Figure E62

Compute
HI
r, = 1.66ry = 1.66(3.71)
==
KL
rx
16(12)
6.16
3 1.2
6.16 in
Note that a 12mm gusset plate is used between angles at the connection,
Possible angles are:
2L152 x 152 x 7.9: r,,, = 48.0
(required for tension)
A = 4.71 > 0.5504
2 L152 x 102 x 6.3:
r,,,
C, = 107
(Table 6 1)
41.5 mm
3 15 >0.5504
We will check the pair of angles 152 z: 102 x 6.3. slncs they are the !Ijl:est.
Member 47:
8'2(1W0) = 197.6 + F, = 26.3 \(Pa
41.5
Member 39:
KL
=
..
..
(Table VI5)
0.K
O.K.
,
XYIALLY LOADED
COLmfiS .<4ns n m s nl
Sectlon
A ~ , , ~in2
.
ry, I D
Fa. k s ~
W12X53
W12x58
15.60
17.00
2.48
2.5 1
9.35
P,,,,
=A
F*,Iups
The r, value for the W12 x 58 IS 5.28 > 2.5, so the section is satisfactory. Also, this section is the same size as member 9. so the joint ullI be easy
to fabricate. Use a W12 x 58 section.
///
Example 65 Design the struts for the cablesupported roof of Examp!r 59.
The strut load is 2.2 kips based on cables at 4 f t on centers and struts as
shown in Fig. E65. The maximum strut length IS 2('33.3) = 46.6 ft.
Figure E65
SOLUTION
The maximum K L / r for compression rnembsrs is limited to 2W.
According to the AISC specifications, this would require a radius of a ~ a tion r of at least
Try a W12 section, since we have used a W12 x 53 in Example 57 for,
amember 9, w p h frames into the joint on one end of this member. A gusset$
plate can covkr both members with a minimum of filler material if all the
web members have the same nominal depth.
We further note that Eq. (69) always determines the column stress for
members using A36 steel since K L / r I 120 and C, = 126.1 > 120.
Commonly, plpe struts are used with diameters rdnging from 4 to 6 ic. H e i ~
we have a rather large unbraced length. qo euarnindtlon of Table 114
(SSDD) indicates that we can use.
Extra strong plpe: 8in diameter: r = 3.88 > 2.80 in
A
For KL/r
46.6(12)/2.88
F,
P
'
12.8 in'
194.2,
(Table 115)
3.95 ksi
smaller section in the outer onefourth of the span, where the value
L = 2h is less than 46.6 ft.
A builtup section is a more practical design than using a rolled shape in many
situations. This is particularly true when there is a very long unsupported
column length involved such that to meet the L / r requirements would require
one of the heavier rolled shapes. Another factor of primary importance is that
the radius oPgyration of builtup members can be controlled (see Table 62 for
selected examples so that the value of rx can be made more nearly equal to r,, to
produ<cemaximum section efficiency. This efficiency cannot be obtained using
the sdqdard rolled W shapes, where the ratio of r / r y is often 1.5 to 5 or more,
unless bracing is provided with respect to the Y axis.
Table 62 Approximate radius of gyration for several builtUDsham
..
up sections are very commonly'uskd for bridge trusses and..in co.
r towers. Antennas are essentially builtup colu mns, although not
onsidered ,as sudh. In any case, where compre:jsion (and tension)
embers are used in large unsupported spans, a builtup m ember may need to
e considered. Any of the sections considered in Chap. 5 may be used (refer to
igs. 54, 57, and 58) in addition to any other section configuration, which
ay, or could, be made appropriate for the design problem.
It is somewhat more difficult to produce an optimum, or leastweish~,
builtup section since there are several design parameters to satisfy, including:
r..
rb7
= same as
r , = 0.4211
t7

]lI
I.oi2,Ls
/ r
7 7
Figure
. 64 Builtup shapes using cornbinatlons of rolli.,! shpsi. Sc.iti,,n
""
"A:..
,A
".A
1
""
,,
I..
..A
1
....
,I.,
r~.,.L
!iri:lr:d
;;,:,n!ctr)
5
"2
?
'
.L.
.~:1:>
:?j
/.\
"
.I
of using lacing, single and double batten, and perforated
plates (called, collectively, cover plates) are usually used. Where the steel is located inside a building,
the cover plates may be solid and their use could reduce fabrication costs. In
exterior environments, where corrosion is always a problem, it is necessary to
have access to the interior of the section for maintenance and inspection;
otherwise, the interior must be completely sealed. The "open" cover plates and
lacing allow access to the interior of the section for cleaning and painting
without the careful fabrication required to completely seal the interior. Presently, it appears that the economics of fabrication favors perforated cover plates
to lacing, since automatic gascutting methods allow rapid cutting of the plate
openings in a length of cover plate.
The design of lacing and batten plates, in particular, requires attention to
several details: Lack of proper attention to lacing design was believed to have
caused the first' Quebec Bridge in Canada to fail in 1907. It is standard practice
to allocate a portion of the axial load as the shear developed in the lacing or
batten plate when the compression member "buckles," as shown in Fig. 65.
If we assume equal end moments, as shown in Fig. 65c, and use the
differential equation for bending as used to develop the Euler column equation,
and allow for boundary conditions, we obtain
The derivative at y
0 is
d.~
kL
2
 = ke tan 
Now referring to Fig. 65h, we obtain for the $hear in the lacing,
where k
The AASHTO and AREA specifications make the assumption that the end
eccentricity e shown in Fig. 65c is equal and opposite on the two ends of the
column (shown equal and with the same sign in Fig. 65c). With some a d d i c o d
simplification of the preceding equation for V, we obtain
where
fps
(u)
(b)
(c)
Figure 65 Shear development for a laced (or battened) compression member. (a) Laced column. ( b )
Identification of laced column shear V. For lacing on both through faces, divide V equally on both
lacing bars; for fourside lacing obtain 90" as above. (c) Basic concept of shear in lacing of builtup
section.
SI
ARW
fps
SI
End
AISC
C':
L,,
=
< (',
.:I
[?j.
i..
\/F'J i i p s i
C',
2 if,50
5.;: Sl
.oillnrs~~t~~ri
;? L i 1 :
 
.I I ,<>
ZL,L
:,"
2
 110.:ai.,
IML,,I(L:
iI
where Iq,,=moment of inertia of the jth pdrt aith respect to the pciriicl r.ui
I and through the centrold of the j t h part
A,= crosssectional area of the jth parr
d,(,,= perpendicular d~stancefrom the centroid of the jth area to rile i
axis
5. Compute the radlus of gyration wlth respect to both axes.
#.
IS
I,
=o(J) + Z A , ~ ? , )
7. Check P
= AF, > P
,,,
and iterate as necessarq.
8. Design lacing, perforated cover plates, batten plates. and/or tie (or SIZ:~)
plates.
This procedure will be illustrated by the following ex~mples.
z9
Exampie 66 Design a laced section for the end post of the highway bridge
truss of Example 64, which has seven panels at 25 ft each (see Fig. E66a).
The unsupported length 4, = Lx is 24.04 ft. The computer output (with an
impact factor of 0.17 included) is as follows:
Member
LCI I, kN
Assuming that F,
mately

I,,
Both
t/ and
t,
+ 211,d2 + i l p d 2
2(162) 2(8.82)(1.66)'
= 324 + 48.6 + 134.4
Let us try two channels with a solid cover plate aqd lacing as shown in Fig.
E66b. This configuration, with solid cover plate up, will provide some
protection to the interior of the builtup section, and lacing will allow access
for painting and cleaning. The spacing and configuration will be such that a
reasonably easy framing of the W12 web sections can be made, as shown,
using a pair of gusset plates. We note that filler plates will be required, since
the W12's have depths greater than 12 in.
C12 X 30 data:
2I,,
+ 6.375(4.59)'
= 507.0 in4
=
4.59 in
+ 211,d2 + I,
334'9 22.5
=14.9 
O.K.
280
LSD STXbTS
<
15 in
Try t
&
3l
0.288(0.3 13)
109
O.K.
Compute the lacing bar force. The bar force component psrpzndicu!~
to the member axis is computed using Eq. (614):
(also AISC)
100
62.8 + 10
6;ig)]
+ 
3.349(2.06)
6.9 kips
'?
pd
8.3
 9.6 kips
cos 30'
36[1  0.5(109/126.1)']
Fa =
2.12
btFa = 9.6 kips
d/"
10.64 k s ~
q
Figure E66~
L='
5  9.81 in
cos8 300
[ 1"'
r = ,$if)
t=
0.763
Also, 11.1 is less than 0.67 X L / r ( = 41.9). Limit the L / r of the lacing to
130. The radius of gyration of a flat bar is
ry(of channel)
Design the end tie plates (AASHTO calls these stay plates). The
AASHTO requirements are:
= 0.2881
9.8 1
= 0.262 in
130(0.288)
Use a minimum of three fasteners (or equivalent weld) each side. Transllited
into design (Fig. E66d):
t =  l2
=
50
0.24
Weld
Flgure E66d
Example 67 Design a builtup section using perforated cover plates for use
as a..column in a water tower (see Fig. E67a). The unsupported column
length is 24.7 m and the axial load for design is 1350 kN. Use F, = 250 MPa
and the AISC specifications (noting that a water tower is not a "building"
and may be located where a collapse is more of an expensive nuisance than
a hazard to people, so it may not be necessary to use any specifications),
since adherence to these specifications, although not necessary, will ensure a
safe design. There is usually some wind bracing in water towers, but we will
assume that the bracing is not sufficient to develop restraint against column
buckling.
,x = ). =
43.4 mrn
 = 24'7(1000) = 1 10.26
r
Pal,,,
. 224
81(4.184 x 4) = 1356
>
1350 kN required
O.K.
This cross section will be considered adequate and we will proceed to dzsis
the perforated cover plates.
The cover plate design is not so much "design'. as satisfying seIectec
criteria and producing a hole spacing that will f i t the column Ien&
Referring to Figs. E67c and 66, we obtain:
Figure E67a
SOLUTION
The minimum K L / r will be taken as 200 for main members.
Therefore,
Figure S 7 c
"r
48(0.30)
= 14.4
Use 6 rnrn ( 1 / 4 in). Use welding and locate the plate as shown. Welding
may be intermittent. If the welding is adequate to allow the cover plate and
angles to act as a unit, AISC allows a contribution of the cover plate based
on
where S is the sectlon modulus of the rectangular base plate kith respect to the
moment axis. The plate th~cknessfor this case is also shown in Fig. 6Sc.
From Fig. 68a the area of the column base plate is
m2
///

%.t
. AXUCTURALS n E L DESIGN
287
FP
ips: kipsiin2
S t . >L?:l
7.8
04
ii.25
1.75
0.357;
L o s t h ~ nitill Jrru on
iound~rion
dr
0.j5fc'
5 0.71;
lq.r
n oi rn
,Lf =
d.v
With rnornenr
(ci
S!
Figure 68 General base plate dunensions and other desigo criteria. ( a ) Base plate dimensions. (b)
Allowable stresses F,. (c) Base plate moment to compute base plate thickness.
m
q(m)Z
or
1 unit wide X
r thick, we have
?
!\I = q ( n ) 2
Using the largest value of M (and noting that if m = n they are cqual), we have
M
fb  Fb   =
(for a strip one unit wide)
*
t2
=
where
6M 'I2
3 X q X ( r n 2 0 r n 2 ) 'I2
=[
Fb
section)
Example 68 Design the base plate for a column as shown in Fig. E68. Use
= 250 MPa, f,' = 20.7 MPa, and the AISC specifications.
F,
7s
Example 69 Redesign the column base plate of Example 68 to resist a
bending moment of 265 kN . m in addition to the axial load (Fig. E69a).
Figure E68
SOLUTION
The pedestal floor line dimensions will be the same as the bas
plate. Thus
F, = 0.35f,' = 7.245 MPa (Fig. 68b)
Let us make m n:
From Table V3 obtain d = 360 mrn; bf = 256 mm
(0.205
m = 0.106 m
B = 205
C = 342
+ 2(106) = 417mm
+ 2(106) = 554mm
B X C = 0.417 X 0.554
The actual contact pressure q is
q=
231.0
Check:
O.K.
 7.229 MPa
..
= 50.9 mm
554
52 mm thick.
SOLUTION
We will design the columntobase plate weld in ExampIe 97.
F, = 7.245 MPa 5 q
say 52 mm
750 rnm:
= 656 mm
Figure E69b illustrates data so far including q = 7.245 and tained from using B and C in the preceding equatlon for q.
t
Along line x  x :
q = 7.248  10 22x
ob
70.6 mm
say 75 mm
.*om the other direction (line yy) at point A :
=
q = 7.248
A common and widely used empirical rule for lateral bracing for both
compression flanges of beams and columns is to provide a bracing (z!so z
compression member) element to carry a lateral brace force of
P, = 0 . 0 2 P
L,e the largest thickness of 75 mm. The final column base plate dimensions
nre 780 x 656 x 75 mm thick.
For anchor bolts as shown in Fig. E69c, assume that the bolts will
carry the full moment even though the axial force will reduce the effect of
the moment considerably. This assumption provides some reserve capacity
of the anchoring system to resist a considerable lateral force (colu.mn
shifting laterally).
I
1 :se
F,
where P is the axial force in the compression member being braced; that is,
P = A& for bending members, where A, = area of compression flange and
f, = average (or maximum) bending compressive stress. For columns, use P =
average axial force in column. This recommendation is given by the Structurzl
Stability Research Council in Guide to Stabilip Design Criteria for hferal
Structures, 3rd ed., edited by Johnston.
A series of tests at Cornell University by Winter (see "Lateral Bracing of
Columns and Beams," Transacfions, A S C E . Vol. 125. 1960) inclicates that very
little lateral bracing is required to allow the compression element to develop Pie
allowable design stress. This restraint could generally be developed by t5s
weight of the floor system onto beams when fulllength contact with the
compression flange is made. Because of the variable nature of flooring (metal
deckfine. concretetobeam. woodtobeam. and so on). i t is suooested
., that the 2
percent criterion be used. Winter also de'nved an analytical expression for the
bracing requirement based on both. reitraint and the relative stiffness of the
column and brace. If a SF of appro xi mat el^ 2.5 is used with this derived
expression, the empirical rule of 2 percefft can be obtained
" Z
Example 610 Determine the minimumsize spandrel (or girt) to brric;. tke
W section with respect to the Y axis for the 730kip a..tial load of Example
62. The distance between columns may be taken as IS ft. Fv for column =
50 ksi.
SOLUTIONThe axial force in the channel section used for the brace is
.
P,
0.02 P
0.02(750)
15 kips
Areqd=
l5
3.73
j,02in'
rru
a i n u b i U K A I . blEEL DESIGN
AXIALLY
Simply search Tables 16 and 17 for this combination of A and r, and find
MClO
If bending or other requirements are also satisfied, this section can be used
for the girj (or spandrel).
!
'
///
P,,=$(FdD
+ FLL. . . )
Areqd
P" 54'12
= 3.79 in' > 3.17
OFc,  0.65(21.94)
N.G.
Also,
Here we need Table 31, since the value of @ ranges from 0.86 to 0.65 depending
on q , which in turn depends on KL/r of the column as well as 5.We may also
note that the value of F,, depends on the value of 11 as follows:
26.96 ksi
+ = 0.65
54'12
= 3.09 in2 < 4.30
0.65(26.96)
Use a pipe column 5 in X 14.62 lb/ft.
=
SOLUTION
P,, = l . l ( l . 1 0
+ 1.5S)A
= 1.1 [ l.l(O.030)
O.K.
///
PROBLEMS
6 1 Determine the allowable load that can be carried by a W14
steel and the AISC specifications if:
(a) KL = 16 ft.
(b) KL = 42 ft.
(c) KL, = 68 ft and KL, = 4 4 ft.
21 1 column using
I;, = 50 hi
3 4 . 7 rolled ssc~lonusing
F, = 415 Sip2
md
6 3 What is the lightest square tube section (see Table 115, SSDD) for a column loading of 121
kips and an unsupported length of 12.4 ft? Use the AISC specificarions and '236 steel.
Answer: 6 X 6 x 0.375.
6 4 What is the allowable column load using the AISC specifications and iF, = 315 %(Pa for a
rectangular tube section 300 X 200 X 9.52 mrn wall (Table V16. SSDD) for an u n s u ~ p o n dl.cgr;l
of 4.8 m?
6 5 What is the allowable load using the AREA spssificaiions lor s W l 4 X 145 cset L; a
% STRUCTURAL
STEEL DESIGN
compression member in a bolted end connection for a bridge truss? The member is 15.5 ft long and
cises A36 steel.
Answer: 771.6 kips.
6 6 What is the allowable load using the AREA specifications for a W310 X 178.6 rolled section
used as a compression member w t h a bolted end connection for a bndge truss? The member is 4.75
n? long and uses F, = 345 MPa steel.
Answer: 3 163 kN.
Q 7 What is the allowable load for the builtup section shown in Fig. P67 using the AREA
speclfications and A36 steel? Assume a bolted end connection and a length of 5.25 m.
6 12 Design the lacing for the allowable load found in Prob. 610, allowing a 20 percent i n w e m
('310 X 44 6
6 U Design the perforated cover plates for the allowable load and wcuon w d in Prob. 61 1.
6 14 Redo Example 61 if the contributory column arc3 u 30 X 20 instead ol 20 X 20 but dl the
5.380 X 74 4
Figure P67
6 15 Design a laced column section (refer to Example 65) for a highway bndge truss end post The
length is 7 . 2 ~ 10.18 m. The truss span 1s 50.4 m and the loads are dead load =  1 I20 kY a d
6 8 Whatlis the allowable load for the builtup member shown in Fig. P68, using the AASHTO
Z
speclfications, F, = 50 ksi, and an unsupported length of 18.7 ft?
"
I  p l ~ t eI 2"
$"
x L"
"
.,i:;"
;: '>.
li
"
Figure P68
i 9 Referring to Fig. P67, place a second channel C310 X 44.64 on the bottom of the S shape to
zaake it symmetrical. What is the allowable column load using the AASHTO specifications if
L 6.6 m and using A36 steel?
Answer: 1962 kN.
5 10 What is the allowable column load for the builtup section shown in Fig. P610 if the member
is of A36 steel and the length is 14.5 ft? Use the AASHTO specifications.
>q
 422 kN (wthout impact). Use F, = 250 hfPa and the AriSHTO spsdicatioos.
live load
Answer: Try two C380 X 50.4, 300 x 15 mm cover plate.
6 16 Redesign the truss end post of Example 66 using a perforated cover plate for both sides of the
channels. Note that AASHTO allows use of the net area of the perforated cover plate in computing
the total section area and column capacity.
Answer: 2C12 X 25, A,,, = 23.7, r,, = 4.58 m, ~ncludestwo 12 m X f plates with 3in holes..
6 17 Design a column base plate for the m a m u m capaclty of a W12 x 170 column with an
unbraced length of 12.0 ft. Assume that K, = 5 = 1.0. Use F, = 50 k s ~ /,,' = 4 !GI, and Pis AISC
speclficatlons. The column 1s Interfaced to a concrete psdesral
6 18 Design the column base plate for a W14 X 120 secuon that carnes an m a 1 load of 5 0 ) kips
and a base moment of 200 ft . lups. Use A36 steel. /; = 3 ksi, and the base plate lntsriacrs 'Je
column directly to the footmg, wh~chhas a total depth of 21 In.
A w e r : 24: X 22; X 2;.
6 19 Design a column base plate for the maximum capacity of a W3 10 X 117.6 rolled section wirh
an unbraced length of 4.1 m. Assume that K, = K, = 1 .O. Use Fy = 345 ;LIP&f: = 28 blPa, and the
AISC specifications. The column is interfaced to a concrete pedestal.
6 20 Redo Example 61 1 for the lightest available W8 section.
6 21 Redo Example 61 1 if the loads are as follows: dead load = 35 psf; live load = 75 psf.
Answer: 5in pipe at 14.6 Ib/ft.
6 22 Redo Example 611 using the following data: dead load = 1.75 kPa; hve load = 3.75 LPa;
column contributory area = 6.1 x 7.1 rn, column length = 4.98 m. and K, = K, = 1.0. Use eikrr a
round pipe or a square structural tube for the column and Fy = 250 MPa steel.
6 23 Design member 6 of Example 66 (refer to Fig. E660) if the dead load bar force = 283.9 Lips,
l6"4
4
Figure P610
6 11 What is the allowable column load for the builtup section shown in Fig. P61 I? The length is
5.3 m. Use the AISC specifications and A36 steel. Neglect the contnbution of the perforated cover
plate.
the maximum live load, including impact = 109.4 kips, and the minimum Live load = 0.0 kip.
Compare the section to that obtained in Example 66. Take P, = +(pd DL + pL LL). wkere
4 1.3, 8, = 1.0, and BL = 1.67 (latest AASHTO). Also, P, = 0.85A Fc,, where F<, = Fa from Eq*
(69) or Eq. (6lo), without using the SF = 2.12.
Answer: Use the same section as in Rob. 616.
BEAMCOLUhm
DESIGN
71 INTRODUCTION
When a structural member 1s !oaded In a manner to produce more t#n oce
stress mode, some adjustments must be made in the allowable stres$s Vvnnere
the stresses are produced as a combination of bending about the X,and Y axes
as in Sec. 48, the final stresses used for deslgn are obtained by superposirioa,
Figure W1 Beamcolumns and beams making up an industrial frame. Note alternate orientation
of strong axis of beamcolumns along sides of frame. A closeup of selected joints is shown in Fig.
915.
Since F,, may not be equal to F4, (part~cularlyin the case of FV shapes from
flange geometry) the beam deslgn In Chap. 4 was obtalned by iteration.
Accumulation of compressive (or tens~on)stresses at one edge of one of h e
flanges was used In the following form of Eq ( a )
298
produces a bending moment in addition to the axial loads in the column. In this
case the column moment is not at the column ends. Similarly, wind pressuregon
long vertical members may produce bending moments, since a large distahce
between floors (or ground to roof) may disallow the concept of wind being
carried, analogous to oneway slab action. In Examples 25 and 26 the framing
of the side sheds to the columns of the main .bay produces large column
moments which must be allowed for in their design (to be considered in a later
section).
Other design conditions produce bending in addition to axial forces. For
example, the top chords of roof and bridge trusses are normally "pinended"
compression members, but the weight of the member will produce bending as
yell. Purlins placed between panel joints of a roof truss or rafter as a means of
redu*
both the purlin member size and roofing span will produce bending in
\he chofd or rafter.
In general, compression members are loaded with axial forces and moments.
'The moment(s) may be at the ends of the member, as in rigid framed buildings,
or developed at an interior point from a bracket, local beam, cable attachment,
or other loading. When the moment effect produces single curvature (see Fig.
71) a much more critical design condition is developed than when the
moment(s) produce reverse curvature.
Bending may also be produced in tension members such as the bottom
:herds of bridge trusses where floor beams may frame into them. Bottom chords
of building trusses may be used to attach hoisting devices; other temporary
loads attached to the bottom chords of building trusses will produce bending in
addition to the axial load present.
:'
"'
or
200
20d
40,000 words
CPU requirements:
n g ~ d :100 x 3
3 0
300'
90,020 words
Since each number (or word) requires approximately 4 bytes the requirements
become:
:b
Pinned:
40,000 x 4
Figure 71 Column loading curvature resulting. The single curvature of ( b ) is often the most critical.
(a) Reversed curvature in building frame. ( b ) Single curvature interior loading. (c) Reversed
cwature interior loading.
300
BEAMCOLLhl?i
DESIGN
331
B~NDING
When tension axial load and bending occur simultaneously, the principal of
superposition may be safely applied. This is because (see Fig. 72a) the tension
load tends to reduce the bending effects, so that the value of A is reduced
somewhat, and, of course, the actual stress is also somewhat less so that the
maximum allowable stress conditions computed as
f,+  < fb
I
F,
Fb will be safe. This safety is partially obtained by neglecting the effects of P  A on
fb, which could be computed (to be strictly correct) as
Mc
c
fb = ++
P,AI
I
When Pb is at midspan, we have (with the limitation A 2 0, see Fig. 72)
pbL3
P,~L'
A=+(a>
48 EI
8 EI
and the resulting bending stress is
(b)
The actual bending stress can only be obtained by iteration of Eq. (a) until the A
value used on the right side is sufficiently close to the A obtained on the left side.
This iterative solution may be reasonably practical on a computer, but with
hand computations, in only a few design situations (where the number of
members is limited ) is this approach economically justified. Neglecting the PA
effect in Eq. (b) is an error, but on the conservative side. Inspection of Eq. (a)
indicates that the tension stress reduces the deflection and also reduces the
compression stress due to bending. With the allowable tension stress F, a
constant value and recalling that Fb may depend on the unbraced length of the
compression flange (and possibly reducing the allowable compressive stress), we
see that use of
fb+   f<, I
Fb

FI
Pb
i=ll801n
\ r = 21
Figure EY1
L)O~rl
#
Example 71 Glven the portlon of* b#hway bndge truss w i t h 102:s
an2
members as in ~ l g E71,
.
what is t h i ' ~ & u m tension stress in the bwer
*?"
,
chord?
.SOLUTION
(neglect the PA effect)
0.0.10(25~)(
12)
8(5l 9)
WL?
jb==
8sr
P
J = AF
0 72 ksl
165
fb   +101782k s 1
(mau~rnum)
(m~n~mum)
///
When a compressive axla1 load acts together bb~th a bending morncrt. tks
deflection is ampl~fiedand the compress~vestresb incre~seb S~ncrthe aJ!ow\.ab!e
compression stresses take Into account poss~ble buckl~ng(l~teraldefIecuons),
member design IS more senslt~veto t h ~ sloaciTnc rp&e than to one producing
tension stresses.
Referring to Flg. 726, we note that rhe ax131 load (absumed to be applrcd
last as belng easler to vlsuallze) Increases the deflect~on.The order of Iorid
application does not affect the outcome, honever. as long 3s jrelding is not
produced. The Important concern IS that there IS an Increase In the detlection
due to the PA effect, wlth d correspond~ngIncrease In the bending stresses. T:
value of the deflect~on1s (w~thPb at m~dspan)for t h ~ scase
A
P ~ L P ~A L '
 '+
48 E l
8E l
()
(c)
are
(4
Figure 72 PA effects on tension and compression members.
B U M  C O L L W DESIGN
ch.ljes more deflection), and for members with an I too small or an L too large a
Il..,kling failure can occur.
The PA effect can also develop in tall buildings, as qualitatively shown in
Fig. 73, when wind or earthquake forces or unsymmetrical loading produces
lateral displacements of the upper floors with respect to the lower building
eiencnts. A computer analysis can be made to analyze the PA effect but
requires iteration. The steps include:
!. Vake a conventional computer analysis using the lateral forces.
2. Cjbtain ithstory lateral displacements Xi and for XI+, (next story above).
3. Compute an additional Pmatrix moment as
Mi = ~ I + l ( X I +l Xi)
arid a shear as indicated in Fig. 736 using
which is applied at the top and bottom of the story with due regard to signs.
e. .
+ 1 floor levels
4. Compute new X, and compare to previous values used in step 3. Iterate until
satisfactory convergence is obtained, such as, say, 0.02 ft or 0.006 m (app;oximately
in).
r?ik P<A effects are also called "secondary effects" and have been largely
ignored until more recently. Some designers have arbitrarily increased the design
A. Another
stresses (from loads) by a factor such as 10 percent to a110
eloped from
fwtor that tends to mitigate the Ph effect IS that it is;u'&
stresses that
wind or earthquake analysis where the designer can use a
are increased by onethird. The PA effect would only exce& this in rare cases.
The analysis computer program in the Appendix h
a
gk
! eeily modified to
automatically scan the deflection matns for the appropriate X values, recompute
the P matrix using these values, then storlng them for comparison with the new
values from the current cycle until convergence and exit.
(61)
sin
K L , we obtain
';Y
KL
.' = ,I,
v, = y ,
?'I
s , = A sin
KL
P ) = sin a
Flgle 73 PA effect for tall bulldings. ( a ) Structure with lateral loads. ( b ) ith story with deflections
gt j exaggerated.
8,
=
"TI
IS
304
Substituting,
 we obtain
2
8, =
C o x , where Ga = 2 El, / LC
2 EIb / L,
Similarly,
8,
(z),
A G sinT I
K a
7 KL
8,
  G
(:)iLc (
"Yl
smcos
7i
KL
1
. 77
coss~nKL
K
TY~
..
K 6
KL
Similarly equating Eqs. ( h ) and (d), we obtain
Lu"
 ?I ( t a n 2
K 6
KL
tan
K
Substituting Eq. ( i ) for tan v , / K L , we obtain
Figure 74 Elastic frame for derivation of G, and Gb terms to obtain effective column length KL. (a)
Part of an elastic frame. KL defined as distance between points of inflection. ( b ) Column element
isolated from (a) with terms used in derivation identified. (c) Conjugate beam and moment variation
(assumed).
From the bending moment diagram for the assumed moment distribution along
the beams linearly varying as shown in Fig. 74c, the slope of the beam at the
juncture with the column (and using conjugate beam principles) gives '
where the summation ( c ) is taken because load and moment are coming from
both directions. From the earlier derivation of the Euler equation and summing
"YI
+ tan tan KL
K
77
f, ' A,/
6(Ga + G,)
tan s / ~
We may program Eq. (72) for increments of Ga and G, and Iterate untd a \ d u e
of F = n / K 1s obtalned to sat~sfythe equality. The value of 7 / K thus obtalned
is used to obtain K as
h,
~b
($1'
Gb
tan s/ K
) + _v2 tan
K I
T/
(73)
(i0
100050 0 30.0:
I1
0:
'5.8
8'0 
K
70.0
1 00
5.0 4 o
3 0 
1.0 6.0 
so
2.0 
 CC
100.050 0
30.0 20.0 
4.0
3.0 
3.0:
1.5 
2.0
1.0 
1.0
Go
Gh
0 
5 0 . 0 3 ~
10.0~
5.0 4.0 3.02.0 
10
0.9 
Gh
~ 0 . 0 ~
1005.0 4.0
3.0

0.7 
0.6
0.5 
0.3 
0.6 
963. Since smaller K values had been used in structures that had an adequate
ervice history, a new look was taken of the derivation for G,, Gb and the
resulting K. Yura (see "The Effective Length of Columns in Unbraced Frames,"
A I S C Engineering Journal, April 1971) correctly pointed out that where the
K L / r ratio was less than C,, inelastic buckling should be considered and E,
should be used in Eqs. (74) and (75). The use of E, is equivalent to
0.2 
0.1 
0.1
o
The 'use of K factors obtained in the manner just described has been
required by AISC since 1963 and by A A S H T O since 1974. The K factors tend to
0.3 
0.4 
0.5 
0.4 
0.2 
. When
2.0 0.8 
o
 E,
Gine~asilc E Ge,,st,c
( a ) S ~ d e s w a ypermitted
( b ) No sidesway
Figure 75 Nomographs for the effective length of columns in continuous frames for lateral restraint
conditions indicated.
This equation may also be programmed for values of Go and Gb and to find the
corresponding value of F = a / K to satisfy the equality. A plot of this is shown
in the +iornograph in Fig. 75b.
The use of both nornographs shown in Fig. 75 involves computing values of
2 EIC / LC
2 EIb / L b
 2 EIc/ LC
G
 2EIb/Lb
Go =
and
(75)
ince El = AE and A I 1, it follows that use of G,, ,,,,, ic gives a Kkchric <
nce El is somewhat awkward to obtain and recalling in the derivation of &e
ISC equations for Fa in the inelastic region,
9;.
..
(74)
and in the elastic region
(75)
From the derivation involving Go and G,, it is evident that if we call one of the
values Go, the other end produces Gb (i.e., the values can be used interchangeably).
When E = constant it may cancel from Eqs. (74) and (75); however, when
inelastic buckling is developed, E, should be used for E in the E I c / L c ratio.
Other considerations include:
1. When the column is pinned to the base, the E I b / L b ratio is zero, since the
theoretical value of I + 0 that results is G + co. For this situation it is
Fa
SF(KL/~)'
From these equations it follows (using i
inelastic. r
FO'
G,= G,
elastic) that
(77)
Fa,
is computation neglects the variable SF. which ranges from 1.67 at K L / r = 0
23/12 (use 1.92 for hand computations) at K L / r = C'..For most columns in
e range of K L / r = 40 to 60, the variation in SF is essentially negligible. The
se of Eq. (77) requires values of Fa, for the same K L / r value as F,,, so i t is
3U8
BE.L\ICOLUbN
DESIGN
3@
necessary to compute a table of values such as Table 117 or VI7 (SSDD) and
using Foe= Fd to correspond to the AISC Specifications
F,' =
1 2  7 ~ .' ~
23(~~/r)~
1. Several cycles, or
2. To convergence, or
3. To an arbitrary limiting K, such as 1.2 or 1.5.
Of course, if the section chosen is not adequate, a new section must be selected
and the computations repeated. Disque (see AISC Engineering Journal, No. 2,
1973) proposed that the iterations for K could be eliminated by using fa = P / A
instead of F, to obtain
Figure E72
, ' ,l
3.94
fo
fa =
712E
23/12 x 5 j 2
570
29.1
49.4 ksi
5 5 . For K L / r = 55
The revised
BE.L'.lCOLL3W
Fa = 19.37 ksi
=29.1(19.37)=563.6>520kips
DESIGN 311
O.K.
Si:lce this value is about 40 kips larger than needed, try a W14
90:
Try W250
114.6:
520
KL = 34 (Table 115)
 19.62 ksi
26.50
r
i c2m Table 117 obtain (note that this table is computergenerated and uses
S F = 23/12):
fa=
Fd
129.18 ksi
F'
Fa = 19.64 ksi
Pa
1.25.
(Table 115)
26.50(19.64) = 520.5
O.K.
189.4/4.6
0
Obtain from Fig. 75a:
Gb =
///
Example 73 Given the frame shown in Fig. E73, with sidesway possible,
use the AISC specifications and Fy = 345 MPa to find the required column
size. Use the same section for both.
co
1.2
K = 1.75
for left column
Check the left column first, since K is much larger than the right column:
1.75(4.6 x 1000) = 70,6
r
114.05
Fa = 143.7 MPa (Table VI6)
KL
=
Pa
AFa
> 2OOO
0.K.
F, = 169.7 MPa
Pa
14.58(169.7) = 2474.2
O.K.
SHTO, and AREA) interaction equation with bending about both ~ x e A,,
s
as well as axial load as
fa
fbx
fby
Fa
Fb
Fb
(7 10)
+I<
rior to 1963, the value of F,, . F,, = Fb. Currently, we recall that F, depends
Witeria,
a d
n several factors, including unbraced length and compact section
.C
'6
"
general AISC allows
..>
"$
Fbx = 0.66Fy
or
F,,
or
0.60Fy
Gb = 10
K = 0.72
K = 0.65
Check the right column:
KL
'4
Gb = 1.0
Using Fig. 756, we obtain
=
Currently, Eq. (710) is used only in certain limited stress conditions. For t5z
remaining stress cases, ther more complicated formulas based on reszarch,
plastic design,..,and elastic stability concepts are uszd. These will be psrtidiy
developed in the next several paragraphs to indicare some of the limirations so
that the practitioner will have some idea of how to follow through should tbs
design vary from routine.
Refer to Fig. 76 for a short ( L / r + 0) rectangular section of dimensions
X d that is stressed with both an axial force and a moment sufficient to
evelop a plastic hinge. The plastic moment in the presence of a compressive
0.65(4.6 X 1000)
= 26.5
112.78
rx
Pa = 12.9(190.3) = 2455
Fa
> 2400 kN
190.3 MPa
O.K.
KL  0.75(4.6 X
1000)
= 30.6
Fa = 186.8 MPa
112.78
rx
Pa = 12.9 x 186.8 = 2410 > 2000 kN
O.K.
Use a W250 X 101.2.
///
' I,= b X
:I.,,
Fallow
Dividing this equation by Fa,,,, = Fa, one obtained the widely used (AISC,
Figure 76 Plastic hnge formation in a very short membsr subjected to both an 1G2! force .n\f
moment.
314
'* "
effect as Eq. (b) of Sec. 72. However, i t can be shown [see, for e:c~mp!e,
Timoshenko and Gere, Theory of Elastlc Stablliry, 2nd ed. (New Yock:
McGrawHill Book Company), Sec. 11 I ] that i t is sufficiently accurate (a
1 to 2 percent error) to amplify the moment for P1:
Although the development above has been made for a rectangular cross section,
it is also valid for all (including W, S, and M) shapes. A plot of Eq. (71 1) is
shown in Fig. 77. Also shown is the plot of a linear equation of the form
for k L / r = 40 and for K L / r = 120, the straight lines shown also on Fig. 77
would be obtained. These curves will be somewhat in error, since the PA effect
$as been neglected. We could use an iterative approach to include the PA
Shown in Flg. 78 is a plot of the loading situation u here tf, = .t12 = Jf
and in Fig. 79 the case where M I = itt and itlz = 0. These to plots represent
the extreme range of cases where a column is loaded w t h end moments,
ma
building frame. The curves shown in F ~ g s78 and 79 have been made using a
modification of Eq. (e) proposed by Galambos and Ketter (Transacf~om,ASCE.
Vol. 126, pp 125, 1961), whlch gives, for equal end moments.
F*
78 Influence of K L / r ar
BF.A\iCOLW
2
hlo
DESIGN 317
may note that the reciprocal of Eq. (712b) is used as the C, amplification
or for laterally unbraced beams. Equation (712a) is used as C, in the AISC
s when using t : i ~
! I I ~ / , I I=~ (  )
Ill
\!!
= 11
%,
KL/r
0
20
40
80
120
0.42
0.70
0.99
1.81
3.16
0.77
0.46
0.17
 0.72
2.51
1.13
1.14
1.16
1.19
1.25
1.11
1.18
1.23
1.52
2.53
The plots using Eqs. ( h ) and ( i ) are reasonably satisfactory for all of the cases
with equal end moment, but it is rather conservative for those cases of unequal
values other than at M , / M , = 1.0 and with a rapidity determined by the
/ M , ratio. To the right of the intersection of the two curves, Eq. (d) cor:;:i.!i.
is requires use of two equations in design and using the most critical .of
cm=
0.4Ml
2 0.4
M2
1.75  1.05M,/M2
+ o.~(M,/M,)~
'
it
by
'llis is because it is easier to solve for the most critical value by using the two
equstlons than to make a plot and locate the intersection and then use the
goverli~ngequation. These equations are adjusted for design use by substitution
sf PC,for Py and M,,,,, for M, and with section area and section modulus to
ob , stresses. This gives
a,
+ C , ~ B ~ ~ a,
~ , (P(KL)?
~ )
where
PdliOW
(fps unlts)
P(KL)* = ksl
#
'
fb
,#@
E ~r
= ;;
3
Fa
Fa
Fa
P0 . 6 5 + B, M,yFb.r + B,. M,  I Pal,,W
fa+
Fa
'1
cmxfb~
f
'myfby
(lfa/F&)Fby
< 1.0

(7 13)
(7140)
*84
. Fby
and
AISC Eq. (1.61b)
fa +  +   < I
fbx
fby
IIIUIIJ
but ,$A
the actual column load P and AF, = maximum allowable column
*(bad ${anbis
not the same value shown on Figs. 78 and 79 and used to
devcin:; ,.the curves shown). Also noting that /, = M/S, define A / S = B,
niult~ply/./F: by A/A, and take FiA = P, = 0.149 X IO~/(KL/~)'ksi (in fps).
u u r r ~ r t rr
of column sections and for several assumed column lengths based on kTL/ry.
Similarly, the terms ax, a,, B,, and By can be computed and tabulated. These a x
shown in Tables 114 ( 5 = 36 ksi) and VI4 (Fy = 250 hipa) for the W shspcs
commonly used as columns. The AISC manual also has these tabulations for
F. = 50 ksi steel and includes use of S shapes and tube and pipe sections.
Y
320
BEh?fCOLL'4N DESIGN
331
1.0 + 7
Fe
+la
(715)
,,*$
(lfo/~;;)~bx
CmJby
( 1 I;/F;)F~~
' "
+ b   +  fh
I1
Fb, Fb
fo
0.472F;
F"
_< 1
(717)
(7 15)
n ' ~
F
2. I ~ ( K L / ~ ) '
Generally use C,,, = 0.85 for end conditions of Fig. 71 1 b and c; use C, = 1.0
hen the interior moment is greater than end values or with an interior moment
The AREA equations are similar to the AISC equations. For,f;/F,
fbx
Figure 711 C, reduction factor for beamcolumn interaction equations. (a) No sidesway: C
, = 0.6
 0 . 4 M , / M 2 . (b) Sidesway: C
, = 0.85. (c) Column with transverse loading: C, = 1 + +fa/F,'. ( d )
Several cases of transverse loading and factors shown.
1 a
/F
.b
rb."
( 1  .fa / 1";') Fbv
<
V'E
F;" =
is computed using Eq. (712a):
1.431(~L/r)'
lso, at points braced in the plane of bending:
fa
with attention to signs (single curvature =  M,/M2). Note also that M, is
defined as the smaller of the two end moment values. When sidesway (Flg.
71 16) is possible, AISC specification allows:
o.55Fy
fbx
jby
Fbx
Fb,y
++I1
fo
fbx
fby
Fo
fby
Fg,
++
1
I
Cm = 0.85
(sidesway present)
> 0.15:
(7 19)
BEAUCOLL3c.F
value of KL' to
The design of beam columns using the interaction equations is essentially a trial
(i!.i.;:tive) process. A section is tentatively selected and analyzed and if the
i. se;.:ien is too small, a new section is selected and the process repeated until a
sz:i:i'actory section (both strength and weight) is obtained. The steps may be
ou: ,.edas follows:
,.
Fb = 0.66Fy
gA
1.
nr
C,BM
AP
BMFa
Fb
Fa
,B
Fb
0.2 to 0.3
Bx Fa
Fbx
0.1 to 0.2
,B
estimation for
ax
ax  P(KL,)~
...
, '
,
:,
..
+ A P = P + (0.2 to 0.3)M
= P + (6.5 to 9.O)M
: *6
(M in in . kips)
( M in kN m and SI)
ihk value of P, enter a Table such as 114 or YI4 (or in AISC Manual
o i Yables which have been prepared for design use using a computer) with the
K4, value and obtain P,,,,.
If Kx/K; I rx/ry, the KL/r,, values control; if
P.',/K; > rx/ry, the KLx/rx ratio controls and one must use an adjusted
r'
use a 1 a ~ ~ e r s e c t i 6to4sa:~sfqdesign
With care and some preliminary scrutlny ofv Table 114 (or Y I J ) , an
adequate section can be obtained In one to three inals. Thls IS possible. since the
tables show little change in r,/r, and B for the two or three sect~onson either
side of the selected section.
':!
L,nb,acedI L,
computed from Eq. (423), (426), or (427)
if
~ h t values
:
for bending about the Y axis are considerably larger, but a AP
+w,(;&
Fb
0.6Fy
,
P,, = P
fbx
;~ntl
Fb
if LUnbrace,
I LC
i f L ,,r,,
> L,
The unbraced length IS the actual and not the KL value and IS taken with
respect to the bending 2x1s.
7. Compute K L / r cntical, obtaln or compute F,, and compute Ju = P / A .
Compare fa/ Fa I 0.15. If the ratio IS less. use Eq. (715) to see ~f the section
is adequate. If fa/Fa > 0.15, ~t will be necessaryfto :heck barh Eqs. (7130)
and (7140). In this case compute:
P ( K L ) ~ use KL with respect to the bendlng a.uls. ~ h l c hmay be different
from the cntical K L / r used to compute the allonable avlai stress
Fa
8. Check both Eqs. (713a) and (714a) to sat~sfq
AP
" #*?
e tables, obtained as
KL
KL' = I
rr/ rv
It is necessary to use r,/r,, slnce the ratlo IS fixed for a section but is cor
known until a section is tentatively selected.
6. Record Ptable,A 1 ry, r,/ry, LC,L,, B,, and a,. L, and L, are needed, so a rspid
determination of Fb can be made:
,(erminethe axial. force and column moments. We note that this step is also
1 yitive,, since indeterminate frame values are not found until a tentative
.,.!~tlon is used in the analysis.
''2.::>&ipute K to obtain KL. It may be necessary to determine both Kx and Y,
!':bending on column end conditions and lateral bracing.
2 . .'?.,timate the moment contribution (and we will use bending about a single
axis for illustration) as an equivalent axial load AP:
.*
'
'
3.44
DESIGN
By Eq. (713):
fu
' d b r
+ 5 1
Fu
PFb,
By Eq. (714):
i;+f L <
1
0.65
Figure E75
7.48
22
>
1 1.80 1n2
LC = 8.5
<
L,
.
12
L,,
19.6 > 12 ;
.
Fb = 0.6% = 22 ksi
16
&=z==
12'ft + Fb
F, = 15.98 k s ~
1.92(36.2)'
0.6F,
22 ksi
42.5( 12)
jb= 9.83 ksi
5 1.9
.112~
'"
1.92(36.62)
1 1 I .3 ksl
9.32
I I 1.3
= 1    0.92
By Eq. (714):
0.58
a 2 ~
74.2
15.98
7.48 ksi
1.92(~~,/r,)'
KL
+  =
f,   =9.32
0.58>0.15
F,
Fel =
>
F'
lo
14.70
9.32 Lsl
S, = 51.9 in3
l lo
1 1.80
=  =
Y'
ju
'7
x'
= 2.64
x 40.
W 12
ry = 1.94 In
7.88
22
+  = 0.70 << 1 .O
Use C, = 0.85 with sidesway. After some study of Table 114 (with P
1 lo), let us try W12 X 50:
Fb,
= 113.8 ksi
+ 0.85(9.83)
= 0.991 <
0.92(22)
O.K.
1.0
By Eq. (714):
ru
0.6%
932
+ 9'83
22
22
Use a W12 x 40 column.
+<
F, 
1.0
 0.87 < 1 0
O.K.
3 3
326. 5
E;.awple 76 Given the column and bending moments shown for a building
frame, with sidesway for K, restricted by use of bracing and shear walls, use
the AISC specifications and F, = 250 MPa steel to select a tentative column
section.
SOLUTION
Refer to Fig. E76 and assume that
1
Kx=1.25
k;=1.0
Figure E76
+ pequv
pgwen
. 445.. + 7(41)
= 872 kN
Sc2.1 Table VI4 and select W310
x 59.5:
P,,,,,
A = 7.61
l o v 3 m2
rx
=
114.1
I 14.1 = 753
(445) + 8.94(61) _
150
130
Using Eqs. (7 13) and (7 14) would give:
Eq. (713):
< 868.3 kY
860.5 kN
2.64
'Y
1.25
>I
O.K.
Use as tentative section, W310 x 59.5.
Fa = 1 14.1 MPa
LC = 2.58 m
Fb = 0.6%
4.87 m
I~O'MP~
F ( x L ~ )=~445(1.25
fa
Actual
7.61
=
(Table VI5)
L,
> 3.45
58.5 MPa
AFa
  58'5
Fa
114.1
0.51>0.15
P + G,,,,BxMx
\*@
Fa
Fbx
445 +,0.95(8.94)(6133
///
Example 77 The top chord member (No. 6) of the truss used in Examples
64 and 65 will be designed to include the member w e i a t and a temporary
concentrated force of 2.2 kips that will be applied to the center of the chord
during maintenance operations (see Fig. E77a). The bridge may be temporarily closed to traffic if the maintenance load is too large to be carried
safely with traffic (live loading). Other data:
Dead load =  283.75 kips
Live load =  109.44 kips
P =  393.19 kips
O.K.
ax
< 'allow
ax  P(KLX) 
132'8
150 132.8  8.27
865.3
< 868.3 kN
A = 14.70 in2
O.K.
I,
11.00 in'
3%
A ~ C T U R A ISTEEL
.
DESIGN
e used to support the siding and for lateral bracing. Caution is nec&sary,
owever, that girts should be continuous and, should building repair require
emoval of a girt, that it be done in only one bay at a time so that the Isteraf.
support is not lost.
3. There will be a column moment at the crane n l n w q l a e l due to the IongirudinoI
thrust of the crane starting or stopping suddenly. This will also produce a
moment at the base plate even if the analysis is made in such a manner as to
ignore the moment at the crane level. This force will also produce a column
shear that must be resisted by the anchor bolts at the base.
The general design of a column of this type proceeds as follows:
. As in Examples 25 and 26, tentatively analyze the structure and revise until
i&
';62
(c) 1 .
bo.,. l:;e X and Y axes. Now if we take the X axis of the main column member
' %Pic.r~~~d
for bending in the plane of the bent and the Y axis for bending out of
)iac,: (with respect to building length), we have the following considerations:
1. 5 x e d or pinned against rotation at the roof truss level. The roof truss of the
ii~dustrialbents of Examples 25 and 26 provides rotation fixity (at least
nearly so), but translation may take place.
2. Sidesway control. The side sheds in Examples 25 and 26 plus any bracing in
the plane of the first and last bents will act to control sidesway. If it is still
excessive, knee bracing may be required from the column to the roof truss.
3. There wiN likely be moments in the column at the roof truss. There will possibly
be moments in the column at the crane girder level due to lateral thrust of the
crane trolley against the rails making the track. There will be a column
moment at the base due to assumed base fixity. These moments produce
beamcolumn interaction for which Eqs. (713) and (714) must be used.
pinended column free to buckle and with a load Poand an interior Ioad Pi a
in Fig. 713 is in a state of unstable equilibrium if the loads are sufficientlq. IrrrgeIf we use the differential equation
and allow for boundary conditions of different loads in the lower sement, no
lateral displacement at the column ends, change in I for the lower segment and a
common slope at the junction of the upper and lower sezgnents. we obtain
'
n*
Y Axis
i
**
$1
d&iyi4!yss
:. :
I
" E I X ~or plnned at roof level. If we put some cross bracing in the plane of the
chord and vertical cross bracing in one or more of the bays on
'op;osite s<des;'the major amount of sidesway can be controlled.
i . &!&Iintermediate bracing. This is necessary to reduce K L / r , of the main
!::4ccr.;mn and of the column segment above the crane girder point. Girts may
where J = 1 + Po/ P,  a
a = L,/(Lo+ L , )
This is a solution as given by Sandhu, "Effective Length of Columns nich
Intermediate and Axial Load," A ISC Engirieeririg Journal, October 1977. Ths ki
32
BE.L.fCOLL?Ci
DESIGS
33
Example 78 Make a tentative design for the crane column for the industrid
building in Example 26. Refer to Fig. E78a and assume the foliokg:
Figure 713 Figure for derivation of effective length coefficient for a stepped column.
Rearranging yields
'1
I'
but
and equating this load to the Euler load, we obtain
SOLUTION
A single W690 x 264.9 was used to obtain prclimin~ryou:put
shown in Example 26. This size section (area and moment of ~ n e r t ~ ais)
necessary to reduce the lateral deflections at maln truss roof level acd at
side shedtocolumn intersection to tolerable values. The builtup section
will require this or larger values to provide satisfactory lateral displacements. From the computer output (and for a tentative initial builtup section
iteration) for LC = 1, the axial loads and moments (for left column) are:
Member 26 (uppermost):
= 318.74 kN
moment
120.59 k3
c'
Member 25 (intermediate):
P = 387.1 1 kN
Member 24 (lower section):
moment
338.99 kN . m
P=543.01kN
moment=313.6kNm
base moment =  190.1 kN . m
Let us somewhat arbitrarily try a section made up of one W360 x 314 main
column and one W360 X 101.2 crane column, as shown in Fig. E786.
.
A.
considered stable by adequate use of girts and siding). Use the largest force
in the upper column, 387.1 1 kN.
I'
W360
314 data:
PI
I, = 1107.2 x
m4
A = 40.0 X lo' m2
W360
Zy = 428.7 X
=
399 mm
bi
m4
PI =
543.01
= 401 mm
L~
Also,
101.2 data:
1, = 300.9 X l o v 6 m4
A = 12.9 X
m2
ly = 50.4 X
d = 357 mm
m4
bf = 255 mm
+ 1 2 . 9 ) ~= 910(12.90)
X
910(12.9)
= 22 1.9 mm
52.9
The actual K
1.2.
Ke
Kc,,
,,,,,,,, .,
K
c,,,,,,,
0.79(1.2) = 0.95
From Table VI5, obtain F, = 134.6 MPa From Table VI7, obtain F,'=
9
732.8 MPa. Check the Interaction equation [Eq (713)].
A
736 STRUCTURAL
'
STEEL DESIGN
BE.+(COLWW
DESIGN
337
Use ,C = 0.85. (We will not check for bending about the X axis at this
time, since computations are very preliminary. After the next computer run,
if this section is still satisfactory, we would check bending about both axes.)
I
#
U.
By inspection, Eq. (7.14) will be satisfied for the lower column segment.
The upper column segment should be checked for interaction to make
sure that the W360 X 314 is adequate as a column for the full height of the
bent.
We can now reprogram this example with the new column sections and
any revised sectlons for the truss members and see if the lateral deflections
far the several load conditions is satisfactory and that the bending moments
and axlal forces are compatible with the section being analyzed. This should
be done prior to refining the final design, to keep the engineering calculatlons to a minimum and maxlmize use of the computer.
///
in
zcl<crsd bay.
7 9 CONTROL OF SIDESWAY
It is evident that the most efficient column design results when the frame is
adequately braced against sidesway. With no sidesway:
1. C,,, can be less than 0.85 (but may be more in selected cases).
2. The effective length factor K is not greater than 1.0.
A rigidly framed structure can translate laterally sufficiently to undergo
"~idesway."Note also that the development of the equations for the G factors is
based on a common slope at a joint that can only be obtained for a "rigid" joint.
It is therefore necessary to provide specific resistance to sidesway to obtain the
a ~ obtained (see Fig. 714) via:
most efficient columns. ~ h i s ' m be
1. Shear walls (use rigid vertical walls of brick, tile, or concrete block to contain
the lateral movement). If masonry walls are used, a close contact with the
column should be provided so that the column cannot translate in the
construction void between materials.
where
Diagonal bracing is usually designed only for tension. It is assumed that 13bracing member is so small and flexible that it will buckle (with stresses we:!
below
under a very small compression load. For oppositedirection loading,
the buckled member straightens with no damage due to small buckling stresses
and prevents sidesway from occurring by carrying the necessary tension load.
6)
Example 79 Given the story of a tiered building shown in Flg. E79, design
the diagonal bracing for the intenor bay. Assume that this wll be placed in
alternate bents in the outofplane direction.
BE.k\fCOLb\fX
DESIGN
339
4%
P,, 5 0'.pu
When there is bending about only one axis we have [he following:
and
I n these equations = 0.86 (for beams) and o = value given in Table 3! i ~ i
columns a n d noting
varies from 0.65 to 0.86 d e p s n d ~ n gon k Z / r . The C,
terms are as in AISC. The values of P, are
+,
Pu=AFy(l0.25r12)
P,
AFy
4
(730)
~i\'>
Example 710 Given the beam column and loading shown in Fig. E7IC.
select the lightest W310 shape using F, = 350 bIPa and LRFD.
Figure E710
SOLUTIONUse F,
Mu,
P,,
1 . 1 [ 1.1(400)
+ Ii(6Cn'l:
42.2 L
S
i  rn
= Id7!
'K"
BMUCOLUMX DESIGN
With bending only about the X axis, Eq. (728) applies, so that
PROBLEMS
x = A rin!iy
show that K
+ B cosky + Cy + D
x = Oaty = Oandy =
M=Oaty = O
Z, = 2.687 X
138.9
r,
0aty
"
Q
72 Determine the effective length coeific~entsK lor the colurnns In the frame shown In Fig.
Note that the far end of the girder of column A is pinned.
Answer: AB: 1.75; FG: 0.78.
=d&
0.7 for the column of Fig. 6,3b. Note the boundary conditions:
,
A = 20.13 X
ry = 79.0 mm
and
controls
m3
from which
qic =
Figure W2
73 Determine the effective length coefficients K for the columns Ln the frame shown in Fig
///
Figure W3
ill
t:
Figure P75
Figure W8
I!I
79 Design the top chord of a railroad truss with a panel length of 27.583 ft for the flollowmg
conditions:
Live load = 867.2 h p s (E80 loading)
Impact = 473.6 kips
Dead load = 402 kips (estimated)
.@
use a bulltup section somewhat as shown in Fig. P79. Use the AREA specdications and A36
steel. The chord member ends will be either riveted or bolted.
A m e r : Two S24 X 100 and one S24 X 79.9 with a 30 x cover plate.
710 Check the section of the column above the crane runway girder of Example 78 and redesign as
required.
711 Using Example 78 as a guide, make a tentative redesign of the main column of Example 25.
Use A36 steel and the AISC specifications.
712 Redo Example 72 with the column sue limited to W12.
A m e r : W 12 X 96.
713 Design the diagonal bracing for the bent shown in Fig. E73 to inhibit sidesway. Use the
lightest pair of angles with a 12mm gusset plate.
Annoer: Two L63 X 51 X 4.8.
714 Check the exterior columns of Example 23 using the computer output and resize the columns
if necessary. Note that both exterior columns are to be the same sue. Use a single column (no
sphc*for
full building height. Use A36 steel and the AISC specifications.
715 CHeck the interior columns of Example 23 using the computer output and resize the columns
as necessary. Note that basement columns are not necessarily the same size as the upper column
wbch is to be used for full building height. Use A36 steel, the AISC specifications, and not aver
W 10 columns.
Answer: W8 X 48 bottom; W8 X 40 upper.
716 Use the computer output of Example 24 for the exterior columns as outlined in Prob. 714.
Limit column size to W250.
A m e r : W250 X 67.0.
717 Use the computer output of Example 24 for the interior columns as outlined in Prob. 715.
Limit column size to W250.
718 Venfy w t h computahons that the W310 X 178.6 column section of Example 710 is adequate.
$f@ 719 R & ~ OExample 710 d ML = 210 kN m and D = 425 kN. AU other data are the same.
Answer: W310 X 178.6.
720 Redo Example 710 if My moments are present: M,, = 50 kN . m; MLy = 75 kN . m.
Answer: W310 X 282.8.
81 INTRODUCTION
A steel structure is produced as an assemblase of the structura1 members
making up the framework. Connections are required where the various member
ends must be attached to other members sufficiently to allow the load to
continue an orderly flow to the foundation. Since the connection serve5 ru carry
a d from or to adjoining members, i t must be adequately designeti. c
:
;
I
onnection design involves producing a joint that is safe, economical of materiIs, and capable of being built (it must be practical). The more practical
onnections are usually more economical. since fabrication costs greatly affec:
economy of both connections (or joints) and the members thsmsc.!v:s, 2.i
ustrated earlier, particularly concerning builtup tension and compression
embers. Several structural connections are illustrated in Fig. 81.
Connections (or structural joints) may he classified according to:
FIgure Vm1 Highstrength bolted joints. ( a ) Splicing smallertolarger column using filler plates.
(b) Splicing samesize column. ( c ) Diagonal bracing.
tions using bolts are further classified as bearing orjricriontype cop. nsctions.
2. Connection rigidity, which may be simple, rigid (as produced by an indeterminate structure analysis), or of intermediate rigidity. The AISC, in See.
1.2 of the specifications, classifies joints based on connection rigidity as:
Type 1: rigid connections that develop the full moment capacity of tilt
connecting members and retain a constant relative ang!: between tI.5
connected parts under any joint rotation.
P
(el
r'
Figure 83 Several modes of joint resistance. (a) Bolt shear. (b) Plate shear or tearout. (c) Bolt
bearing. (d) Plate bearing. (e) Bolt tension failure. (A Tension on net section.
where
2 A, F,
the connection
For many years rivets were the sole practical means of producing safe 2nd
iceable metal connections. The process required piinching or drilling ho!:s
roximately 1/16 in (1.5 mrn) oversize, assembling the parts using drift pins to
ign the holes, and using one or more bolts to hold the parts togetbet
mporarily. &vets were heated in a furnace (portabie for onsitz use) to a cherry
ed color (approximately 980C) and inserted into the aligned hole through tLe
everal parts to be connected. One member of the riveting crew then applied a
ng bar with a head die to the manufactured rivet head to hold the rivet 1I1
and to shape. Another crew member used a pneumatic driver with a h a 3
to forge the protruding rivet shank to produce the other head. The forg.ing
ation simultaneously reworked the rivet metal and caused a shank enluget to very nearly fill the oversized hole. This reworking and shank enlargeent, together with the shrinking of the hot rivet, produced a substantial joint
of the time. The rivet contraction during cooling is resisted by the joint
rial and develops tension in the rivet so that a riveted joint is intermediate
een a friction and a bearingtype connection (a bearing type is commonly
ed). This joint transmits the design load primarily by friction between the
ed plates making up the joint. The riveted joint has had a long history of
cess under fatigue stresses as in railroad bridges. Only recently has ,ARE.:
llowed use of highstrength bolts and welds in joints for railroad bridgts.
355
41
t the procedure for the design of a riveted connection is exactly the sao;,e a
for a bolted connection. Figure 84 illustrates several sizes of undriwn rivers 2nd
structural applications using rivets.
A325"
A490b
Figure 84 (a) Several sues of undnven structural nvets. ( 6 ) k v e t s in structural apphcations
Turnofnut
Specified torque
No
Yes
Yes
Yes
F',.'i;?,
BOLTED &\Q R
w C O h > % ~ O ~ 359
S
,J
r proper installation
;':'
re p= Slip coefficient (usually can use 0.35 for clean mill scale; most 0~~~~
surfaces are less than this value. and i t may be necessctry to deternine
the value by test)
m = number of slip surfaces
N = number of fasteners
T = proof load of each fastener (as in Table 52)
i.
p
4.
,U =
0.35
Psli, = rn,uNT
iTU=
k11(141)
SF =
one
Y
Prrslsting
Pallowable
~dequacy.
A bolt tension of approximately 0.7Fu gives adequate reserve strengt
the bolt be somewhat overstressed (say, 3/4 turn instead of 1/2 turn).
The belt tension acts as a mawive spring in tension to hold the fastened parts in
relati' ~ositil~n...
This clamping effect also tends to hold the joint against nut
loosening in fatigue load situations, SO that most of the time a locking nut is not
If A325 bolts have not been excessively overstressed (not more than
3/4 turn of nut) they may be reused one or more times. Tests on reuse
indicate that A490 bolts should not be reused in any situations.
Pallow
= F,Ab(nornina~)
SF
0.35(141)
120(0.7854 X 0.0'0@)I @
1.31
The reader should note that this safety factor is against slip and is not
safety factor of the joint, which is on the order of 1.67 (for a tension
f//
t and may depend on tension on net section).
0.85  C , ( L  C2)
fps: C, = 0.007
SI:
L 2 Cz
C, = 0.00275
C, = 406 mm
C, = 16 in
d =4
p
,Is
.'
4
y
. . ,,
, ., ,,,
Oi*
,,. ....
..
Rolled edge:
D, = 1.4
>(
in Or 3 mm)
30
I'
3 x diameter
,
smin= 3
diameter
:
F b#:. Several modes of joint failure. ( a ) Several joints. (6) Bolt shear failure. (c) Tension on net
section iL.:re. ( d ) Tearout failure due to bolts being too close to end in the direction of the stress.
D, = 1.75 X diameter
Rolled edge:
De = 1.25 X diameter
?*A
842 Bolt Distribution and Gage Distances
It is necessary to ensure a reasonably compact joint and one where the
cted material is in reasonably good contact so that the developed friction
is uniform between the parts. If the bolts are too close together,
is obtained, since the maximum coefficient of friction is p 0.35,

AISC specifications require that all connections. except those in rrusses, c2rrying
calculated stresses be designed for the design load but not less thrin 6 kips (or 77
kN).
AISC specifications require that truss joints in either tension or compression
be designed for the design load but not less than 50 percent of the effective
strength of the member based on the type of desi,on stress.
The AASHTO specifications require that connections be designed on the
basis of the average of the design load and the effective strength of the member
but not less than 75 percent of the effective strength of the member. This is
because many of the structural members in U S H T O design are controlled by
factors other than stress, such as L / r . At least two fasteners are :equired in any
AASHTOdesigned connection.
364
tion.
Pi
SD
1 percent f o r each
where
of
L = 450
Same as M S H T O
For most welldesigned sections, Eq. (83) should give a value of e , from 85
A
to 90 Percent. The AISC reductions for shear lag were presented in Set. 53.2.
I n ~ t ~ transfer
dl
of some flange

'fastener
Ptotai
 number of fasteners
is assumpllon made for ellher fastener shear o i bcanns. BeannZ i s ccriided In some speclflcatlons as beans: of fastener on ha$ nlelal 37J G~ C a t
+* *
,"cur
Table 83 Table to determine required bolt length.Fased on grip and thread
length (use table to determine if threads will fall idshear
plane; see example below)
L,,,
gnp + L; round Lrcqdto next larger f in or 6 rnm
Bolt sue, m
I
I
'
i
Thread L
L, m
Thread La
I1
16
12.5
25
If
1;
7
8
15
20
30
35
Grip=;+
Fiwi. the table, L
%:
tqr
L,,,
& + ;=
1&1n
1 In:
* R ~ u n to
d nearest larger
= I&
+ 1= 2 1
16
ble 84 Allowable bolt and rivet stresses for AASHTO specific~fions
stress mm normnal wit area except l u r r \  ~ v iwiu 211
(use glven
b!
on)B
r ~ i d w a b ~shear
e stress. F
Lreqd= 2f In
*"\% .
2.25  1 375 = 0 875 In
ineta1 on fastener. The shear area is obtained using the nominal fastener
diameter. The bearing area is the projected fastener diameter x plate thickness:
the fastener load Pi is
b.
P, = A , x F,
(shear)
Pb = D X t X Fb (bearing)
No allowance is made for the hole size being
in (1.5 mm) larger than the hnl
shank in bearing.
4
$
n the fastener group is unsymmetrical or the load does not pass through
oid of the fastener group, the fasteners are not equally
 * stressed. This
ii?,ibn is considered in detail in the next section.
.We note that the allowable shear stress F, depends on the design assu
,,tie: (:.at the joint is either a friction or a bearing type, and whether the
 1 0
$?
 
!.XASHTO, and A E X
$0 obta~nthe dIottahis
XISC values s h o i t ~in
,
material, compcisd
No
made for fasteners in either sin&
. distinction
     .. is
.
 .

or doubIr shrzi. T i e
F, = 1.225
nd will generally be limited by fastenertometal stresses. as shown in Tab12 84
der the column headed F,.
The AREA bearing values are stipulated for the fastener type except th2t
ring is not considered. in the
for connections using his!strength WE.
. . design
 ,
ese values are shown In 1 able 86.
MPa
ksi
11.0
13.5
248
248
20
27
36
36
Fo
ksi
MPa
76
93
138
186
Use beanng stress on n y t s : single shear, 27 ksl or 185 MPa; double shear, 36 k s ~or
i 2 5 0 MPa.
Need not cons~derbeanng on bolts In frict~onconnections.
Table 86 Bearing values for rivets and bolts by several specifications
(top part of table is metalonfastener,bottom part is fastenertometala
Material
ksi
AISC
FU
MPa
ksi
MPa
ksi
MPa
AASHTO
ksi
MPa
AREA
ksi
MPa
F=
0.62(825 MPa)
120
250
Not required
4.26
int geometry reduces this value of F to something on the order of 3.3 rot
mpact joints (in tests) and to around 2.0 for joints whose length is in excess of
70 mm (50 in). This value of F compares to the tension value on the cross
ction for A36 steel of
F =   F"  1.72
0.58 F,
A307 bolts
&vets A501 and A502
Power dnven slngle shear
double shear
A325 and A490 bolts
CO
150
servations of joints in long service indicate that a safety factor of F 2 2.0 lor
fasteners gives satisfactory service.
ms are often spliced to produce continuous spans. Splices are usualIy p!aceiI
lose to the location of zero shear in the span. Any use of mechanical fast*Pnzrs
the tension flange will reduce the effective area somewhat. Based on otes:s,
r~ss
ISC and recent AASHTO specifications allow ths designer to use the ,
lange area for stress calculations as long as th? holes (in the tension flange) are
s than 15 percent of the tension flange area. When the area of holes e x c r d s
percent, the flange area is reduced by that portion of holes in excess oi 15
For example, if A, (of tension flange) = 20, A,,,,, = 4. the percent holes = 4
= 20 percent. The excess hole area = 20  15 = 5 percent and, the
ension flange area is 20(1  0.05) = 19, not 16, as would be obtained by t.l.ring
t all the hole area (20  4 = 16).
These computations should also be used for the flnngz spiics plates.
x 100/20
370
BOLTED
STRUCTURAL. STE
P,,,, = 2
I /
11 ,'
For a frictidntype connection:
X A,
x F,
(double shear)
he number of bolts required is N = 442.5/91.2 = 4.85. Use five bolts.
the bolt pattern shown in Fig. E826, .so that the splice plate wihth will
maximum but the maximum net section is obtained for the channel.
r a three bolt line and sheared edges, the minimum edge dis
use 40 mm
1.7(22) = 37.4mm
Minimum bolt spacing = 3 0 = 3(22) = 66 mm
250  2(40)
O.K.
Use spacing =
2
For the forward two bolts:
85
Center bolts as 40 +  = 82.5 nim fronI edges
2
Spacing to edge is less than maximum allowed of 12t = 12 x 10 = 120 m a .
Use the distance from the front bolt to the edge of splice plate at 40
(1.75D for sheared edges) and similarly from the back bolts to the edge of
the W410. Set the gage distance so thatonly two holes are deducted from
the critical net section.
C310 X 44.6
s =\
/
2
1
z
= 46.1 mm
use s
50 mm
(arbitrary choice)
F, = 5(0.01)(27)(1.50 X 400)
660 > 442.5
O.K.
5(0.013)(22)(1.50 x 400)
Ph = iV x ,+Ihx
On web of C3 10: Ph
O.K.
858 > 442.5 kY
Use five bolts for the friction connection.
For a bearingope corlrrecrlorr
Grip = 2 x 5 x 13.0 = 23.0mm
L,,,,=23.0+28=51
iise55rnrn
=
Figure W2a
Figure E82b
Check the net section of the splice plates. With three bolts out, the are
requirements are:
Gross:
442.5
A, =
= 1.475 X
2 x 0.6Fy
m2
Net:
Ap =
442'5
= 1.106 X
OSF,
m2
A >=
"lo6
 0.85
1.301
13 = 18.0 mrn
say 4 bolts
m2
> 1.106
Try two plates 250 x 5 mm with three holes in the critical section.
A,,, = [250  3(25.0)]0.01 = 1.75 X
O.K.
.=
2.6
(also use 4)
II
.3
TODchord
159.0 t 99.3
= 129.2 kips
2
The 75 percent member strength criteria give
pa" =
P,,,
Use A36 steel and the AASHTO specificatiJns. Use A325 highstrength
bolts. P,,, =  99.3 kips (P,, =  40.6 kips). The gusset plate t = 5/16
in (minimum t allowed by AASHTO for a plate).
373
Memno
I'
Po,,
< 129.2
From Table 84, the allowable bolt shear stress F, = 13.5 ksi. Thz nuxSsr
of bolts required in the connection to transfer 129.2 kips is
129.2
P
N=?=
= 15.9 bolts
AbF,
0.7854 x 0.875"
13.5
Use N = 16 bolts for symmetry and since 0.9 bolt is not possible. Tne use of
16 bolts requires four rows. Use a bolt spacing of 3D:
4(2 625)
+ Z(1.5) =
13.5 in
Figure J333
SOLUTION
The fastener design will be based on a frictiontype connection.
AASHTO does not allow a bearingtype connection in a main membe
Stress range does not have to be considered for connection design.
We will try four 7/8indiameter bolts at a section as shown in Fi
E83, the same as assumed in the design of member 9 for tension
Example 57. A deduction for net area does not have to be made f
compression members unless a complete stress reversal occurs (this does not
occur here).
AASHTO (Sec. 17.16) requires that a connection be designed for th
average of the design load and the full effective member strength but not
less than 75 percent of the effective strength of the member. From Example
1 2 j ( 8I )
l $ in (1.125 in)
The distance furnished and based on the standard gags distance = 5.5 in
(see Table 13) is computed as
b,  5.5
10.01  5.5
dlurn=  =
= 2.26 ~n > I . 125
O.K.
2
2
Check the bolt beanng on the gusset plate. slnce I, = 0.313 < 0.640 of
flange of a W12.
O.K.
x 0.3 13)(40) = 175.3 kips > 129.2
ade sufficiently wide that tension on a net section
li be some f~llerplates needed between the gusset
nce the W 12 sections are deeper than I2 in.
///
'
.Jf\
."a
'
,t*'
This design was based on using a 12mrn gusset plate and 25highstrength bdlts.
P,, = 0.5(0.6<)~,
= 0.5(150)(2.66) =
199.5 kN
controls
2t0.7854 x 0.025')(150) l d
375
able 81):
147 kN
199.5
= 1.36
147
Use two bolts for shear, since a fraction is not possible.
Check the beanng:
N
=
The bolts through the angles and gusset plate will be in double shear, as
illustrated in Fig. E84b. Assume that a bearingtype connection (slip can
tolerated) is satisfactory (a designer's prerogative). Check the bolt len
using Table 83 to see if the threads are in the shear plane.
15mrn bolts
1 2 7 X 89 X 6 . 3
I
:I ,.I
V\FJ/v
I
_
I

Figure E846
Bolt grip
6.3 x 2 t 12 = 24.6 mm
24.6 + 30 = 54.6
use 60 mm
(Table 83)
45 mm
One,aQgle + gusset = 6.3 12 = 18.3 mm
Thread runout location = 60  45 = 15 mm
15 mm
<
L,,,,,,
nerally, when the eccentricity of the load on a bolt group is less than sboct
in (60 mm), it is neglected. Joints such as the simple frame connection of Fig
a, which is widely used, are in this category. The bracket connectiori of
g. 812a is loaded with an eccentricity that is obviously too l a r ~ eto be
glected. The framed beam connection may be large enough that the resuf5ng
ccentricity,,isalsd too large to neglect. One may note that the standard frzrnstf
connection angles in the AISC design manual neglect the eccentricity for vduls
to about 3.7 in (one of: the standard framing angle connections with nearly
value of maximum eccentricity is shown in Fig. 812b).
A load to be resisted by a bolt group that is eccentric with respect to the
id of the group pattern can be replaced with a force that has a'lineof
= PP,
through the pattern centroid and a moment with the magnitude Lti
ere e is the eccentricity of the load. This is illustrated in Fig. 8l3a a ~ 5.d
considering that each bolt in a pattern that is centrally loaded carries its
ed share of the total load, we have, for equalsized bolts,
P
p.=
"
!V
where Psi is the shear force on the ith bolt with a vector to resist the a p p l i d
force P .
An additional
bolt force is develo~edbv the eccentric moment ,bf = ?r.
~
.
Assuming a group of bolts acting as an elastic unit. we have a concept similx to
that of beam resistance being developed and as related to the beam moment of
inertia. Referring to Fig. 814, we have a bolt pattern \%ithan applied moment .El
which produces a resisting moment for rotational equilibrium that is equal to

,=n
i l
18.3 mm. ' .threads in the shear plane (see Fig. E84a)
1I
Z ,
1.
Rh  RL

Rv =

R,
COk>~~3
o77~8 ,s
he bolt pattern and plate adequate for the g~venload in a bearing~?cnnection assuming threads In the shear plane?
SOLUTIONSince the bolt pattern 1s symrnetncal (as in most ~ractlclrpeob]ems), the centroid of the pattern is readllq located and marked 3s cg, 3s
shown on sketch
Compute z ( x 2 + y 2 )
Y   MdY  My
R,=R=
I dl
C.d:dl
erld can be interpreted as the polar moment of inertia of a group of unit are
:.ate that if we use the area A with the denominator of either Eq. (A or ( g ) ,
arLJR, are obtained as stresses. Multiplying the numerator by the area A of
zth !~olt produces the force R, or Rh. With A in both the numerator
denominator, it cancels, giving R as a force. For general design, the equat
are
f 18.3
Example 85 Given the bracket connection shown in Fig. E85a and th
the fasteners are 25rnm A325 bolts and the plate is
= 250 MPa steel,
<
1l s.?
Figure E & 5 b
Com~ute
P = l I0 k N
R,,,
0.5088~
Set up a table as follows and omit the signs of I and y (use veciors
previously drawn on Fig. E85b to determine the direction of the ir afid c
vectors):
Placing these values on Fig. E85b, it is easy to see that bolts 1 and 5 are th
most highly stressed (critical). Bolt 3 is loaded the least amount (13.47 kP.,
The resisting force on bolt 1 is computed as
R =\/(31.8
+ 18.33)' + 38.2'
63.03
f, =
0.7854 X 0.025'
Check the plate bearing:
'
1000
63'03
0.025(12)
63.03 kN
kips
O.K.
x = 2.75 in
I
210.1
< 1.5(400)
O.K.
Check the possible tension rupture of the plate along the forward bolt line:
Moment of inertia, I = 0'012(0'270)3
12
 2(0.012 X
J
8
0.025)(0.075)~
887.5
Section modulus S =  =
C
16.308(2)
= 0.12.. . . .,
270
. .,
Since 16.40/0.7854 = 20.88 > 17.5 ksi, the numbzr of bolts at a Iin diameter is too small without computing R.
Try 16 bolts (bypassing 14, since 12 bolts were so hi@Iy stressed):
O.K.
 12.5 < 2 5 0 / f i
887.5
405.75
t
12
The joint is adequate for bolt shear, plate bearing on bolt, and bending.
R,
= 877.0
50
= 3.12 kips
16
//J
Example 86 Given a crane runway bracket that carries a load as shown in
Fig. E86. Use A36 steel and either 7/8 or 1indiameter A325 bolts.
Assume a frictiontype connection. Find the number of bolts and the
bracket plate thickness.
+ 3.12)' + 10.62'
12.15 kips
O.K.
'Wh6
'w
IS
Find the plate thickness for the bearing (assume that the column flang
adequate):
tp==
12.15
87(1)
l + 5
0 625 In
2
ccentnclty (for 101nt performance). it is
e,,,
0.139 ~n
h=7x3+3=24in
tph3
12
M=50(15)=750in.k1ps
2tp(lO.S2 + 7.5' C 4.52
$1
fi
Me
Fb=22ksl
+ 1 S 2 ) = 7745
3.625 
~~~~~l~
S ~ L U T I ~Referring
N
to
750(12) = 0.53 in
t = 774(22)
b<   95  15.8
t  v36
15.0
15.8
 0.95 in
use tp = 1.O in
Figure W7
= 1.0
%&
'$c
eeff= e
I + 2n
 4
1
f i
eeff= e
l + n
2

to sre
87 Redo Example 86 taking into account the reduced ecceL;:::
'6
t=
Fig
 =
750(12)
774%
f b = F b = 2 2 +  = 1
1.
,333
Check the plate bearing along the forward fastener line and neglect bolt
$ales 1/16 In larger than bolt:
Ip=
BOLTED f i RI\ETED.
~
CO~+XXOP.IS
jO(14.25)
C ( x 2 + y 2 ) = 405.75
712.5 In . kips
(Example 56)
R,
R,
1.756(7.5)
R =\/(4.17
13.17 kips
+ 4.83)'
i13.17'
> 17.5
15.95 kips
N.G.
% STRUCTURAL STEEL
DESIGN
x!?
8
e,,, = 17.75   = 13.75 in
2
M = 50(13.75) = 687.5 in . kips
C ( x 2 + y 2 ) = 14(2.75)'
H"
+ 4(92 + 62 + 3')
609.9
6873
i/.t
The AISC Design Manual gives tables based on one, two, and four vertical
columns of fasteners which may be used to design eccentric connections. By
assumlng the number of fasteners in a row, a computation for e,, is made, and
with n and e,,, we obtain a coefficient that is multiplied by the allowable
fastener load to give the total group eccentric load. One may readily derive an
equation for the allowable eccentric load for a single vertical (or horizontal) line
of fasteners. Equations are given in several textbooks which may be used in an
attempt to reduce the computational effort in finding the number of fasteners
for an eccentrically loaded connection. The author suggests that with the
increased computational efficiency available to the designer with the pocket
calculators, it is as easy to "punch it out'' as to try to use an equation developed
by someone else. This is so particularly because no simple equation exists and
most equations require some iteration anyway.
(dl
Q15 Frarmng connecttons. (a) Momentreslstlng connecbon using top and s a t m$a bckd
o column flange. Web angles c a n y shear. View l w k m g down (b) M o m c n r  r e ~ t m gc o c n c c ~ ~~rz, : o
olumn web. Note that top and bottom plates are welded lnto column web a d flrrc&eto ~ ; i
oiumn snlrclici>.(c) F r a m ~ n gfloor system In pgwer statlm Note >hear zufirnen 113 %eb of r ~1.
der. Coping IS shown for small floor beams IJI nesr foreground (d) F r z m g for bndse Siriz,t,s.
F
=A.
8
'
DESIGN
BOLTED LL?)
SOLUTION
From Example 4 15, obtain
V,,,,
78.3 kN
v , , ~=,140.5 kX
Figure W 8
v = w=L
16.0 mm
t,.,
13.8 mm
.'I....
3.
22.77
use two bolts
Nshear =  1.47
15.46
22.77
= 1.29
Nbrg  1.5 X 58 X 0.75 x 0.27
In the web of a W18 x 50:
22.77(2)
Nbrg
= 1.96
2t(0.75)(1.5 X 58)
22.77
2.7
22.77(2)
use four bolts for symmetry
15.46 = 2.95
Use a web angle with a length of 6 in and t to be determined:
For bearing:
Nshear
Nbrg =
Nshear =
280.7
0.022 x 185 x 16 = 4.3 1
280'7
 5.36
52.3
,
*L
ii
tz0.176
.r
(1
'
uset=$in
,I:
 1 1 ; ~
; 1 , ; +(1; 1
T , '77
t = 0.174 in
For shear:
!!
.. ,
![;a
'
'
'
L=5
30
+2X
use410
The angle thickness will be controlled by the bearing of the three
in the outstanding leg (o.s.1.) fastening the beam to the girder; therefore,
t 2 16 mm.
X
2t(0.022)(3)(185) = 280.5
i'
t 2 ll.5^'mm
127 X 11.1 (Table V9):
g = 75 mm (Table V 13)
Effective edge distance de = 127  75 = 52 mm
."
d,,,,,,, 2 1.5D = 33 mm
A L102 X 102 X 11.1 will not give sufficient clearance between the
  W7hO
bolts and the bolts through the girder web in the o.s.1. The small amount of
overdesign is rather negligible in any case. The design is summarized in Fig.
E89. Note that AREA does not require a "bearing" check using A325
bolts. This check established the approximate angle t and is recommended
whether required or not.
Figure 816 illustrates the usual conditions for bolts in tension. When the bolt ;c
tightened to develop the proof load, the shank elongates. Simultaneously tLclamped plates are compressed. When we apply a load to the connection, we
have the free body of Fig. 816c, which gives

$4
j
,$
 Ae,
= Aebolt
he two clhmped plates expand, and taking P' = prorated part of total P
3%
STEEL DESIGN
where terms
62.4
0.601 X 22
=
X
by a palr Of
The load Is
E8 I Oa.
i 817
~ are
.
COh%xmNS 3%
Use A325 bolts and A36 steel; the load is 62.4 kips.
from the bottom of a W33 x 221 beam as shown in Fig.
W33 X 201
2.36
land,
o.K
= 9 + kln
a WT12
47:
tf
2L's
%dge
b, = 9.065
d = 12.15 in
=
0.875
 9.qoc  :5.
=  
t, = 0.515
5
 *I
& \
'
7
k = 153 In
&?\K
\ '??l+.
S>L3)
7
6 2 4k
Required depth
3 = 9 + 1 53 = 1053 > 12.15 ln
Check the bending moment at the toe of the fillet In web.
Figure =lob
Firwe mloo
>
*>
A, =
2.15in2
2.15
0.85
 2.53 in2
A&
= (3'38
+ 1/8)]
~t~
+ 41.92 in  kips
~(0.875)~
= 0.1276L
Fb
0.50 in2
M = )I.*(?)
'=6=
6
s =M
55
1
b = L
= 2.6875 in
2
16
T = 62.4
  3 1.2 kips
2
Fb = 0.75Fy = 27 ksl
62.4
= 2.84 in2
22
A , =  62.4
=29
0.K
O.K,
41.92 =12.161n
useL=12.5in
0.1276(27)
~~~i~~~~the number of bolts to can7 the hanger force. Use 7/8indiameter
bolts for the hanger to connect the WT to the beam:
62.4
= 2.35
N=
0.601(4)
L =
pf
P
1.70
61.1
1.7
s)
=
62.4
= = 35.9 kips > 7
61.1 kips
O.K.
///
Figure WlOc
201:
15.745 in
tf = 1.15 in
g = 5.50 in
Step 4. Check the "prying action" using Eq. (89):
.' ( 70(1.75)(0.875)'
SOLUTION
Design the rod:
1.75 in
+ 21(6:25)(0.875)'
=: I 1.1 kips
0.K.
O.K.

L/
Since this value is only 300 lb/bolt over the allowable, take it as 0.K
WT12 x 47 with four 7/8indiameter A325 bolts to the flange . Use a
of the
W33 X 201 beam.
L = 12.5 in
2L4 X 3 X 1/4 with long legs back to back
Step 5. Check the beam flange for adequacy without using stiffeners.
tb = 1.15 in (tnanee
 of W33 X 201)
t... = 0.715
in
 w
Check the beam for the hanger force. Neglect the torsion produced bythis hanger rod on only one side of the beam flange, since it is located very
near the end of the beam. Take the hanger rod hole at the standard gage
distance of 88.9 mm for a W410 x 59.5 section. Also,
r, = 12.8 rnrn = t,
tw = 7.7 mm
bf = 178 mm
b=
89  7.7  1.5
2
39.9 mrn
= .5 .F in
From which
250(12.8)'[0.389(1
=
+ 44.5/39.9)] + ( lo')
1.7
19.8 kN
< 38.75
N.G.
Since 19.8 < 38.75 kN, the flange of the W410 x 59.5 is too thin and we
must either use a section with a thicker flange or reinforce the flange of this
section. It will be about as economical to use a section with a thicker flange.
///
3%
COMBINED
SH
f, + f, r fa,,,,
A better fit is obtained using a quadrant of a stress ellipse, since F,,,,, has
separate values for tension
and shear F,, whi& give
(8ljj
ere
Tfie ~ S H T O
equation for the combined stress cases are the same as AISC
some additional conservatism and simplification, giving, for friction'
(c)
(d)
,,
F; 5 13.5  0.221
ksi
(8 17)
MPa
(817a)
Fm
F:
93  0.221
3.
2.
'
tension cohnections
AISC
Bearing
c,
,
aastener
.; 502 grade 1
grade 2
c2
f ~ q
SI
30
38
207
260
26
55
180
380
c
3
fps
SI
1.3
1.3
23
29
158
200
1.8
1.8
20
44
138
303
Friction
AASHTO
cqa
c5
fps
SI
0.75
0.75
i(S
A.307
.". .325Nb
17.5
121
Set F,
FigurrF.812
" Standard holes 1/16 in (or 1/5 mm) larger than nominal bolt diameter; see the AISC
~ ~ (Sec.
u l16.3) for slotted or oversize holes.
ILL
For bolt diameters larger than If in (38 mm); F , = 90 ksi (620 MPa).
Only static tensile loads allowed.
Check P
P=1941>>420kN
420 kN.
=
wi.~:..fv and f , are the actual computed stresses and C, is a coefficient from
Table 87.
The AREA currently has no provisions for fasteners in combined shear and
ter:sii:n.
0.13094P
380  1.4(0.13094P)
420 kN
85 MPa
<
useP=420kN
0.K.
* ,.
,:*
4
fl
0.K.
//I
..
) CONNECTIONS
T = M
=   109.5  239.6 kN
d
0.457
number of bolts in tension in the clip angle is based on F,= 3
A, = 0.4908 x lo' m2.
=
1.6
We note that it is good to not have to use more than two bolts, since
Distance xo = L / 2 w ~ t hsolne
XO
des~gners
= N / 2 w ~ t hsome
= 3.25
designers
F i N 824) Critical sections and dimensions for clip (top) and scat angle d e s i a for use in
beams to columns.
frarmng
Figure B 1 3 c
205 mm
use 200 mm
long leg outstanding. The long Icg will
have to be long enough to place two bolts at 3 D + edge distance + vaIuc of
g, from Table 113 of SSDD.
L > 28 + 75 + g, = 28 + 75 + 65 = 168 mm
Try L178 X 102 X 22.2 (refer to Fig. E8 13c):
a = 50  22.2 = 27.8 mm
.P
SOLUTIONW460
74.4 data:
=
Assume that web angles will be used to c a q shear so that the seat
angle carries only compression due to moment and the clip angle carries
only tension.
Fb
0.755
Fb
0.01643 x
lo'
m'
Since this 'is thickest angle in this group, go to the top of the next
since thickness controls. We would not make L > 205 mm, b
extend.outside the column flange and also because the ben&
b i ~ o m equestionable with a very long width and only two fastener
TV L203 X 102 x 25.4 mm.
a = 50  25.4 = 24.6 mm
0.0246
M = (239.6)= 2.947 kN . m
2
S = 200(0.0254)2 = 0.02 15 10 3 m3
6
2.947
f = 137 < 187 MPa
0.K. (and no angle th
0.0215
A routine check for tension shows the section to be adequate. ~~t us
see if it necessary to use a web angle for shear, since this angl
thick. Refer to Fig. E813d for critical dimensions and other da
analysis.
R = 275 k~
///"
1.4
3.
,,"
.'....,, .k
.I/
j; = 
 0 ~'I;I:\
lL

,.'p,lrl
LIL!
j;~)t
[t
Figure E813d
v acts
203  12
= 95.5 mm
2
Since this value is so large and because we are using the long leg out
ing to provide adequate length for two rows of bolts, let us
distance required for the beam and use that value to &termin
in the seat angle. Assume that the reaction is concen
from the end of the beam for computiilg the moment.
e =
(N+k)t,Fb=R
k = 27.8 mm
(N
N = 135.2 mm
e =  =135.2 67.6 mm
2
Re' = 275(0.0415)
[Eq.(45)]
tw = 9.0
+ 27.8)(0.009)(187.5)
= 275
(.OITI~!II:
1:
.I[,
or bolt> t: = 
col:~p!l[e I L
('he.h I n r < r l ~ r l o ne ~ l u ~ [ i o n
e ' = e + 12  kangle=67.6
=

+ 12  38.1 =41.5mm
Example 814 The bracket connection shown in Fig. E814a uses a piece o
WT and two pieces of angle to make a stiffened beam seat. The fasteners
are A325 highstrength bolts and A36 steel. Is the connection adequate for
the shear and moment to be resisted? Use the AISC specifications.
Fi
8 2 5k
+I
&
( "f) i
17.5 1  
17.5 1 
= 11.08
);I
<
13.7 ksi
N.G.
1.?0"+
* . .
1.5"
V)
%'
Q'
1.5"
Figure ES14a
SOLUTION
Assume a frictiontype joint. For 7/8indiameter bolts, T, = 39
kips/bolt (Table 82).

'
232 ksi
15(11.20)
82.5
= 13.72 ksi
= lO(0.601)
11.20(15)~
= 420.0 in3
6
S = bh2
=
6
M = 82.5(1.47
f , = f ==
447'15
420
 1.06 ksi
The resulting stress diagram is shown in Fig. E8146. The hatched part pf
the M / S diagram is tension due to moment, which must be carried by the
a current form of the LRFD equation for connection design using A325 and
490 highstrength bolts (and with size limited to diameter 2 1.5 in) is
R, = 1.1(1.1D + 1.4L)
is value is compared to the fastener resistance or platetobolt bearing as
Bearingtype connection:
R, = 0.625A,FU
Frictiontype connection:
R, = 0.7rny A,Fu
Combined shear and tension:
2
0.70
+ = 1.00
here Ab=
A,=
Fu=
m=
p=
0.75
Example 815 Given the connection shown in Fig. E815, determine if the
plate thickness shown is adequate and find the number of 20mm A325
bolts using a frictiontype connection, and a bearingtype connection.
Figure B 1 S
406
SOLUTION
+ 1.4L)
= 1.1[1.1(45) + 1.4(75)] = 170kN
R, = I.l(l.10
A,+F,
does not
The A, furnished is
Find the number of bolts required in a frictiontype connection (n
tolerated, but note that after plates slip, !the bolts must still shear for
connection failure). Assume that y = 0.35 (clean mill scale on fay'
surface).
A, = 0.7854(0 
9743)'
(fps)
n
Figure P 81
[Eq. (54)]
qRn
I
= 0.64
F, = 400 MPa
\
L as rzq'd.
>> 170 kN
O.K.
Figure P82
Redo Rob. 82 using F, = 250 MPa and the AASHTO spzcificarions for the full
pacity of the angles. Use A325 bolts and a friction connection.
Answer: 18 bolts, L = 610 rnm just under L for no reduction.
DO rob. 82 for the fps equivalent of the pair of angles and a 3/5in _wset plate for an axid
oad of 240 kips, A36 steel, and A325 bolts in a bearing connection.
Answer: Eight 7/8indiameter bolts.
5 Given the beam splice shown in Fig. P85, use the AASHTO specifications to (a) W i m b o l ~
d w v e r plates for the full moment capacity of the beam. @) D e s i p b l o : and web plzta for the
&
iull web shear capacity of the beam. Use 7/8indiameter A325 bolts for all splice parts, and A3
steel.
Parr~alanswer. M = 1950 In . laps, T = 107 laps.
I
are adequate for tearing but that the bearing should be checked as appropriate. .UI boIt hes
angles are on standard gage distances, as in Table 1 13 of SSDD.
ikfj+~f*\
Figure P85
86 Design the eccentrically loaded bracket connection for the load shown in Fig. P8
25mmdiameter A325 bolts in a friction connection. Determine the plate thickness for both
and tear along the forward row of bolts.
P = 225
kN
9 Redo Example 814 for twelve 7/8indiameter ,A325 bolts.
1 Redesign the connection of Example 88 if the floor beam on the left docs not frame into
8 x 50 section.
the
Answer: Yes.
87 Determine the number and placlng of 22mmd~ameterA325 bolts for a cable connection to a
W360 x 314 column as shown In Fig. P87. Assume any needed T d ~ m e n s ~ oton produce L that is
'~va~lable.
Check the tenslon in the stem of the T so that a large enough sectlon is used. Assume that
the hole for the cable attachment ulll be reinforced so that capacity 1s not limted at that point. Use
and F, = 250 MPa for the T section.
the AISC specif~cat~ons
Answer. 10 bolts.
Figure P87
Answer: P,
217 kips, N
ICO kN acd
412
Electrode covering
Weld f ~ l l e rrneral
Shielding atmosphere
Ih
s:Fcctj
develop the gas shield and to obtain any
n e slag is later brushed away to expose the
g.This welding process is vew similar to the submerged
an electrocondu~ti~e
slag which is held in position between
etal to be joined by watercooled retaining plates (see Figrial is melted and current passed throu& it to maintain the
and filler metal. The filler is obtained from the
ed into the slag. The process is generally done in
the filler melts the retaining plates are slowly
rnpleted, partly cooled weld. which has a thin 'la=u cover in^
must be brushed away.
~
l welding is~ used to shopweld
~
thick
~
plates ~together* I~t ha\ been~
er o p u l a r in bridge work to weld girder plates and floor platesthe order of 20 to 450 mm can be welded by this process in one pass
II
The butt joint is the only joint likely to be ill direct tension. The a110
tension or compression stresses for weld metal is given in Table 9allowable stresses in tension or com~ressionfor the weld metnl mnv a
into this by further limiting the F, of the base metal to 42 ksi for E60 elec
and'to 55 ksi for E70 electrodes in structural grade steel.
The allowable shear stress for fillet welds is limited to
Fo = Oa3Fu(electrode)
in the AISC specifications, but it is always necessary to check that th
sufficient base metal to resist the same shear stresses. Tn penern1 t h e ~ P I A
'13.2Fillet Welds
The fillet weld shown in Fig. 94 is approximately triangular in crdss section.
Care must be taken that the throat dimension shown in Fig. 94c & built ou
adequately. In most cases the legs of the weld D are made eoual hiid t h i ~i s nnt
*
..2
,A
cj
E E
I
computed as
T = D X cos 45" =
I
D ~"0.70711
Figure 94 Critical shear area for fillet welds. (a) Fillet weld for tee joint. (b) Fillet weld for lap joint.
!'c) Throat dimension for minimum shear area.
%
8
:32 u3
.L
0 .2
$*,a%g %
m m m
.xu
is because the quantity of molten weld metal is much smaller and the resultin
shrinkage is much less.
>
< ,'
t ~ ;~ r 7
In1
(0)
Slot
(b)
welds.
here
Part
( 5 1/4
f>1/4
SI, mm'
t i 6
t>6
.,
,,,
.,.. ,.
..
It is possible in the fabrication shop when using the submerged arc welding
process to specifically produce a fillet throat thickness by rounding the wela out
so that for
b
D 5 3/8 in
D > 3/8 in
x 0.7071 1 (in)
use throat = D x 0.7071 1 + 0.1 1 (in)
use throat = D
[,,,I(
Pz F ~ )
""'
rod
a
."<<,A,
1v 5 8"
or betweet,
A I S ( ' S c c . I . I 7.4
Flat bar ro gusset plate
Improieil
Weld both
loin[ p l l l l
llc)l
'i17rl11~l
I O rollins .I\!,
Ii.1
the
reduce
A[SC Set. 1. I 7 . 6
Lap joints
a large weld (or welds from both sides) is imposed on a thick piece b2se
a lamella tear can occur. The tear can occur because the shrinkage strains
om [he welding operation will be large and restrained. T h e restraint may be
he far side or from the member rhickness Or a
tch of electrode and base metal in a f u l l  ~ e n e ~ ~ ~ butt weld tends to increase the possibility of tearing (i.e.. from using an
with A36 base metal). A thin, stiffened
~f
Figure 96
Stnrcturai
94 L M E L L A TEARING
'eaMg tsee Fig. 97) is a phenomenon that may occur in certain welded
Joi'ts It is not a common condition because it involves several
5 ORIENTATION OF WELDS
aboratory tests on small to mediumsize joints show that butt welds
limit joint capacity where the electrode has been "matched" to the base
~h~ orientation of the applied stresses does not have a si~dificanteffect
butt joint strength.
~h~ orientation of stresses for fillet welds is a significant factor
ultimate joint strength. Tests (see Butler and Kulak. "Strength of Filiet
wioU, C O ~ E C X ' I O N S
(?
. ..
jr'
Figure E9 1a
Use D = 15  2 . 0 13
~ mm.
F, = 0.3 x F, = 0.3(415)
124.5 MPa
se a 150mm weld on each side of joint as s h o r n in FigSet. 117*7satisfies AISC set. 117.4). Use 26mm end returns Per
allowable shear strength of the weld.
It
be noted that a butt weld used to resist a moment develops stresse
Figure E9 1b
side'?
a 3/gin
ample 92 ~~~i~~ the welds for connecting an L4 3i
20. Use the AISC specifications7 'I8
r static loading and dbnamic loading
'
 
"CI'UL.
Ufi,.,,UN
.,,UO,.
Substituting L,
,.
"A,..,.
..
D 5
39.8
(0.707 1 1 X 0.1875)(21)
Use the weld shown in Fig. ~ 9  2 b .
=
Figure E926
Figure E92c
For
loading it is necessary to balance the weld
neutral axis of the angle (AISC, s
Referring to Fig. E 9  2 ~and placing the weld across
reduce the joint length, we have
L1
+ L2 + 4 = 14.5
Ll + L2 = 10.5
L l = 10.5  L,
21 = 2.78 kips/in
heck: 6.62
J
,
"c
we equate PC, to the beam flange compressive force (Pbf = A,&) and Iook
he dimension d,, we obtain the current AISC equation [Eq. (I.152)l:
dc
<
100r2\/1;
(fps)
Pbf
7
dc
10.73t:\jF,
<
(SO
Pb/
(9 1m)
c=
*
,Q
tTpTi:r
as
;+
,"here the additional term (~,/36)"' is used to adjust for other grades of steel.
A forther adjustment to the factor 33 400 to incorporate test results (see Chen
(92)
F,,,
,
I
"
i'
~ U C I T J R N . STEEI;
DESIGN
wELUE;
'.a""r,'.',glq:
'
.
.
.W
j ~. ;. , .: ..,'*"
.
CONNECI~&$~
(
:.<
. . .
. '
I' the w1umn flange thickness tcf is less than that given on the right sides of
eqyations above, flange stiffeners are required. neMSC sp&fications req
column web or flange stiffeners meet the following criteria:
I
"
*a,,)!
"
compute the plate length. Note that the plate will have to be long
enough to allow placing 5 in of weld at D = 0.5 in. I t will also have
the
some length between the end of the beam welds and the butt
column in order to develop adequate strain. BY proportion:
, , ..
.,
1 :
. ,.. ...
b E93a
b o ~ u m oDesigning
~
for full moment capacity:
M = FbS, = 24
I
T = C==
'd
=
16.01
" ..
',Se a fop plate that is 8 in wide at the column weld and tapered to 3 in,
zhown in Fig. E936.
'P
Figure E93c
'
97
=  = 0.73
22 X 6
26.4
& =  =97
7.42
Therefore, a pair of stiffener plates are required opposite the tension flangeu s e plates 3 in wide (2 x 3 > 0.67 x 8 o.K.).
'
Figure E93d
Use a f in
e,
flange for
der web), the 3in leg should be welded to the beam and bolts in the 44x3 QsJthat there will be adequate erecti~nclearance and edge distanceangle is welded to the beam web and fieldbolted to the colu
% = (R:oment + R : ~ ~ ~ ~ ) ~ ' ~
where
:.:\
a:
...
.i
< Rallowable
to
in
~~~~~l~ 94 Design the web ande connections for a W18 x 55 as
framed
beamtocolumn
flange
connection
cqi
~ i ~ g~ 4 .afor a
a 55kip end reaction. u s e two ~3 x 3 x t in x 12 in long. Use the
specifications, E70 electrodes, and A36 steel.
+2
2.5
x 2.5)s = 2 C 2 . 5 ) ~
.X = 0.367 in
4
 L K U C T U WSTEEL DESIGN
Figure E940
R =   = 55=
27.5 kips
2
2
/'
/~ / 3 ) :
: 'le polar moment of inertia is (I of line about base = ' ~
3.11
' a =   14.4
, r ,  Y = 1.617 kips/in
ioint 1 of the weld is critical by inspection (or drawing rays from c.g.)
M = Re = 27.5(3.0  0.367) = 72.4 in . kips
4. = ((0.465 +
5&&,
ere b = seat angle width. The value of r is obtained by trial. We may initialIy
timate the eccentricity (assuming the k distance for any angle is t + 0.375 in
t
9 mm) to obtain
e.1n111al
.. =2
b
R,===
R
L 27'5
12
2.29 kips/in
+ 0.5  ( t + 0.375)
einlt~al
. . . =  2+ 1 2  ( t + 9 )
in
rnrn
3.38
D =
0.70711 x 21 = 0.227 in
2.46
0.7071 1 X 0.3 X 70 = 0.165 in
Check the web shear capacity:
O.K.
e beam seat angle shown in Fig. 912 must be designed for bending stability
must be of sufficient thickness and leg length that an adequate fillet weld
be placed along the vertical legs to carry the shear and.moment due to the
ccentricity of the reaction. The angle is checked for bsnding at the fillet runout
k distance from tables) as in Fig. 912a. The allowable bending stress is taken
1.617)~ 1.308~)"~
= (6.b45)'I2 = 2.46 kips/in
D=
=D
+*
0.257 in
& = R
Example 95 Design a beam seat angle and weld for the conditions shown
in Fig. E95a. Use F, = 250 MPa and E70 electrodes. Data for a W4lO X
STRUCTURAL STEEL
WELD W COWECTI
Use the
Also,
 <  = 335
21.2'
t  m
an inspection of Table V 18 (S
b, = 395.4 mm
r/
Figure E 9 d b
= 30.2 mm
fb=
6M
twd2
.
.,,.
,..
6(15.16)
= 132.9 MPa
18.9(0.19022)
870
0.842
...~.,.
y==
50 040
15.16
L=~=o.lsoo,=~,
57.5 mm
0.~.
Use D = 9 mm.
/li
= 2.4733 D, x 106 m4
ob site. Several situations using CO~UIIIII end plates are sho"n in Fig* 'I4.
base plate may be either butt or filletwelded to the column The decision is
Since d,, furnished = 538.2 > 332.2 required, a web stiffener is rtq&&
~hi'stiffener(a pair with one on each side of the web op&site the
column compression flange) only has to be onehalf of h a m depth, since
the load is only on one side.
33.3
tc/
Is,,"
> T = 7(0.67
SOLUTION
Step I. Design vertical stiffeners 1 and 2.
For stiffener I, opposite the tension flange of the column:
610
:
f b '   M  132.6 MPa
S,
4.60
AfC = 6,$ = 0.398 X 33.3 = 13.2534 x 10
G c tcw tbf
17 m m
329)
2
use tJt = 20 m m
use b,,
120 mm
Pw = ; f b ~ =
/ j(132.6 X 13.2534) = 2929 kN
'bf
bst,ff =
d Z 3 8 7 mm, b f = 3 9 8 mm
As, 2
WELDED C
635
 tan' = 58.64'
387
cos 8
0.52042
= k,
F,,
Use the beam for the "column" dimension and 2.5 instead of 5k, sin
column tension flange is at the end and not centered on the 5k zone
2929  250 X 0.0179(33.3
250
= 10.146 X
m2
As, =
 2.5 x 48.4)
6.2615 x l o p 3m2
A,
6.4 x low3m 9 . K .
t, = 28mm
A , = 0.028 X 398 = 1 1.14 x lo' m2
0.K.
For stiffener 2, opposite the compression flange of the column:
db(,+,.s, = d  2k = 635  2(48.4) = 538.2 mm
A stiffener is required if bb > dbw:
, , ,..
.,.,..,...
dbw
>
>
F,
Using inside fillet welds (see Fig. E99b), we obtain D,, = 8 mm, since
= 25 > 20 mrn. Check the effective D of the beam web for shear:
t,
10,73t & f i
pbf
PbJ = 2929 kN
10.73(17.9)~a
= 332.2 mm
2929
So only 7.4 mm of the weld on the beam web is effective, since the shszr of
the base metal controls. Use D along the inside flange as required and an
WELDED C O h W C
R, = 00.6 FEmA,
Figure E996
The
web and
= 250(0.0179)[33.3
Pweld
+ 2(20 + 48.4)]
= 761.2 kN
" P b f  PWeb
= 2929  761.2 = 2167.8 kN
.
WELDED C O W E
9 ,
,,
W410
..
'sign the welds and gusset plate for a pair of L152 x 102 x 19
for
ctive angle capacity in static tension. Use F, = 250 MPa, Ej'0 electr
pi.,. . : , : a t i ~ t ~
Keep
.
the joint length to a minimum,
9.  30Prob. 99 for a dynamic load. . ,
9.
Design a welded framed simple bea+ connection to ca
22 ).: 50 beam to a W12 x 53 column. Use
ai.2 Li X 3 X t angles with a length of 12 in.
dnnuer:tP5/16in;Dwcb=3/16;Dcol=1/4.
,
9.: : Design the framing angles and weld for
1:

9":
'
YO specifications.
.~.moer:1, = 5/16; D = 1/4 in.
Design the welded end connec
,
Figure P915
(.
. neat
':
.: .
Builtup rolled sections may be used where the overall depth is limited in
section is to use two lighter rolled shapes in parallel. Even where the to& weight
.
PLATE GIRDERS
4g
,,
."+
i
,,
..,
( C)
epth D and clear distance hatio h,rb is a significant parameter in plate girder design. (a) Welded plate girder ( b )
in drsim).,( Girder elevation (side view) illusirating other
plate
the reduction7 or
and larger clear web
A
n exceeds 50 to 00 ft
for spans up to 300
00 m) or more. Roued beams are generally more economical for bridge
less than 50 to 60 ft and are used in a deck stringer confisrationough girders are generally more economical than trusses, the latter are sm
in many situations for esthetics, particularly where additiond lanes
ired and the existing facility is a truss
460
s [ ~ i r ~ Jt
. lr
(t
Jl c
o ri I J I I C
I o ~ I J\$kith
I~~
I O J ~or
L~~~ = .![her
j r ~ n d ~ truck
r d
because of truck length factors and positioning of wheels. When the 'pan is
a line
long, the lane loading with single concentrated load
tvine
to find lhc
enough
and
is
more
convenient
to
use
than
of trucks
number of sucks and their positioning for maximum stress
computer program.
me
ln all girder design the standard Iane/tmck loads are incrcased
,impact factor, which is a function of span length.
Figure 105 illustrates the placing of the standard truck loading On
twolane bridge to obtain the contribution of truck loads to either of ths
lin2'
girders. The designer must make a similar type of 3 n n l ~ for
~i~
using more than two girders.
Example 101 what are the design moment and shear values for the *d'r!
of a highway bridge as shown in Fig. 105?
HS 20 loading, s = 19 it, w = 28 ft, span = 110 f t  D a d load due "
deck, sidewalk, and so on = 3.1 kips/ft.
S o L u n o ~nelane factor L, is obtained by 9"1 about the left
X+ Y
Lf = S
aderFig
PLATE G
I =
L
The moments are
+ 125
50
110
O.K.
Use the equations given in Sec. 19 for L = 110 < 145.6 ft.
103
FLANGES AND WEBS OF GIRDERS
   PROPORTIONING
 BUILTUP SECTIONS
. ..
40
110
The shear values are
= ~ ( 5 9 1 6 . 8 1232.0) = 1703.6 ft
The flange cover plate of a builtup section usmg a rolled section as the bas
may be proportioned
as follows (refer to Fig. 106). Assurmng that the maxim
 bending stress fb is the allowable value of Fb,the average flange value is abo
0.95Fb. At the junction of the cover plate and the rolled section flange the stres
s about 0.90Fb. Therefore, let
M, =
= O.90Fb.Sr
kips
M,,,,,
+ impact + Md
= 1703.6(1.21)(1 + 0.21) + 4689 = 7183 ft . kips
= ML(Lf)
+ Vd
= 65.9(1.21)(1.21) + 170.5 = 267 kips
Vdeslgn
= VL(Lf)+ impact
///
Railroad girders are more difficult to design, since there are 18 wheels in the
Cooper Eseries standard loading. Design aids are available, such as those
shown in Table 12, which gives the approximate (to exact) maximum moments
(c)
PLATE GIRDERS
Mtota~
 sx
( d + C)0.95Fb d + C
..
17 5" )u!d be evident that Eq. (101) does not give a uniqr.. r b ovluLlun a n a the final
f!$P.:; cover plates must be investigated using f, = M ,/ , q.
A
reasonabl
m
e$\iliit' of C = 1 in or 25 .
Af
ZX
'a:
465
We could use flange cover plates 25 X 548 mm; however, let us arbitrarily
use a plate width 52 mm larger than the bf of the beam, which allows a
26mm overhang on each side for welding the plate to the flange. This gives
b; = 418 + 52 = 470 mm
t;==13'82
0.47
29.4 rnm
use 30 mm
&'
r:;t
5..
c'JTION
wt
3.36 kN/in
d = 912 rnm
12 499.4 x
The section modulus is
=
M L  . . .wL2
=
8
Md "
72.6(20)'
8
.,
f b = 2 5 .3855
719=
Mdes~gn
=3800 kN . m
3800
0.025)(0.95
x 0.6F,)
 "'I. =
rrFh,
20
 1
IS
ill
I
<
150
O.K.
13.72
0.912 + 0.025
"
by 418
t;
,'
726 KC 1
'
w 1)J
149.9 MPa
?'he cover plates will be welded as shown in Fig. E102b. This requires
checking by/$ for acompliance with AISC Sec. 19.2.2 (stiffened edge
element) :
72.6 kN11n
i\
 1
ruqina
Fn i ~ 1n1 1 :.
a
y,
(1
(0.912
.m
= 17OkN.b
.... ...
A, =
i
= 3630 kN
3.36(20)~
8
m4
The additional moment due to the flange cover plates is proportional to the
flange area:
,'noments are:
is?
+2~d'
30
Figure E1026
O.K.
9 = 32.0
///
466
us
I
1
v$ =
vQ
I
(kips/in or kN/m)
AASIlTO Set 1 7 12
lLfQ
=
(hips or kY)
I
where F,= total force to be carried by the weld In the cutoff length a
M = bending moment at the theoretical cutoff point
Q, I= terms previously defined
F,
I*
Example 103 What 1s the theoretical cutoff dlstance for the coverplat
beam of Example 102? Also, design the weldlng to fasten the ~ l a t e sto the
beam using E60 electrodes
S~LTJTION
Obtain the cutoff points (Fig E103a) The capacity of the beam
without cover plates is
M, = SxF, = 13.72(150) = 2058 kN . m
b
20
+I
Figure 107 Code requirements for theoretical cutoff d~stancesfor cover plates. Requireme
same for a v e r plates narrower (shown above) or wder than next underlying plate or beam
Figure E103a
a*
M,  k,~'
0 at midspan,
" .
ddS STRUCTURAL.
S'I'EBL DESlG ii
,'
x =
;i :
:$
9
'8
?.
4% x 2
M,  L
With M, = M, = 2058 kN . m at the distance x from midspan where
beam capacity is adequate without cover
the theoretical cut
is
M,
g
:
.
L/2.wehave
=
,, :,,, v :,
(3855  2058)
[
The theoretical length of the cover plate
x =
,:
= 6.828 m
]'I2
= 2x = 2(6.865) = 13.73 m.
470 rnm; I, = 12 499.4 x
3855
L2
l ' o t ~ lengrh
i
CP= I i . l i
771  77.1(3.17)
= AJ = 0.030(0.470)
nl
1,=:0~,
r:=;O;iirn
530 k ~
(Y
 + 15
/I/
= 6.41
m3
12499.4
(Noting that I is
and Q = low3,we obtain 281.6 k N / h and not the
direct calculator reading of 0.28 16.)
The allowable stress using E60 electrodes:
Pi
iVf, =
> 2D),
which will
flSrw= f;7
IMJ
.\/,
= 40(1.4085) = 56.34 kN
'
Me = [0.705(2)
+ 0.418](1.4085)[912 + 2(30)]
Figure 108 ~
i section
~ to dobtain~ an approximate
~
expression for the
flan~
area
hd,
..
472 STRUC.IUK~.L y
i b c ~
L)L!SII,~J
Figure 1010 illustrates the assumed girder web plate loading on a strip
dx(h). From Chap. 3 the critical buckling stress was found to be
$!
.,,,"
~,T~E
Fcr =
12(1  p2)(h/t)2
..l
3
I
,,&
\' '
SI:
where the ratiogh/t has been substituted for b / t as given in Eq. (35)
The force exerted on the web of the web segment shown in Fig. 10lq$
A f f sin @
r A,fb
since (in radians) is a very small angle. The stress f/ must be of suffi
magnitude to overcome any residual stress Fr in the web; thus the web strai
units of F L  2 ) at yield is
f =er+ey=
&
Fr +
E
97 100
5
[F,(F,
F,
.
.
I
+=
d~
&
2 ( F r + F , ) dx
fi/2
+ 114)11/*
, ..
.,. .. , .. .
.
A d tan
AJffl
but
+ in radians, giving
Afffl
MPa
dx
2F,(F, + F')Eh
h
t
<
1.5
259
248
kcT2E
( L ) 2 ( t )dx
12(1  p2) h
a/h
333
284
This value'should not be larger than the critical web buckling force
FCrtdx=
kCnZE A,
24(1  p2)
F,(< + F,)
When A7 steel was used, it was assumed that the residual stress Fr could be
~dequatelyapproximated as Fr = f , / 2 = 33/2 = 16.5 ksi. This value !is cur.ently being',used for all values o f . steel F,. It I is! also',assumed 1
.4,/A, = ,I Poisson's ratio for steel p = 0.3, E = 29 000 ksi, and for
'2"Yjtain
h 13 784
t
[ F ~ ( F+, 1 6 . 5 ) ] ' / ~
P U T E GIRDERS
'47
T
l ii I
i bl
F, according to
vu =
Vb +
(b)
v"j
(4
fi (V,)Fr,
( e>
here
(35)
~ q (106),
.
the critical stress is taken as the mean of
q. (106) and the value of 0.8Fy,, to obtain
(107)
v,
cornpression
FYs = 5
!
V; = Fc,(htw) = FcrAw
ultimate shear force capacity of a girder for web plastification (plastic
capacity) is
., y,
v
= ysAw
k = 4.00
+(108)
P L i T E GLRl
t:;d
''
5, 2
0.6Fy
) .
19'A
i*";,:
<'..
i
Intermediate stiffeners are required for any other shear stress condition.
When stiffeners are required, the spacing is limited to
:J
but a / h I 3.0.
[ I + (o/h)']'/'
YDht,
(10.14)
*&"
P U T E GI1U)EY.S
$4
:.
.'*
AlSC a n d A R E A
.A.iStlTO
4, 2 (%)
(1015)
,.:.the minimum
io should be maintained.
he stiffener must be fastened to the web to carry some vertical force.
! i'ifferentiation of the equation for P, with respect to a / h gives
*":I
P, =
r 0.015~,h~<
"LJsiui ?, = F,/ E, a safety factor of 1.67, and assuming that the stiffener force is
debe1;qxd in the distance h/3 from the compression flange, we obtain
.B;
ha&u
tips
1.67
D o r fillet
h some additi*
obtained as
.I
hfr
C o m p u t e L'lr
L 1 = O75h
r = [11'4~"~
Cornpure L r
h r
~ ~ ( f i Check
:
Use t;
1096.f
in or m.
The allowable bearing stress F,,, = 0.90FY (AISC Sec. 15.1.j.1) based on the
lesser F, if the flange and stiffener are of different y~eldgrades of steel.
w i n g Stiffeners
b
Beanr,, tdfeners are always required in pairs over the reactions. Bearing
stiffeces mgy be required beneath concentrated loads carried by plate girders.
These s:,:feriers must extend the full flange to flange distance and have a close
bearir~,:against the flange delivering the load. The stiffener width must be such
as to extend approximately to the outer edges of the flange or angles.
Bear'ng stiffeners are designed as columns with an area that includes the
stiff n ~ r sand a central weblarea of 12twfor end and 25tw for interior bearing
?
stiffeners (see Fig. 1013). This area is used for computing the radius of gyration
andrfo. checking the coludn stresses. The effective length of the stiffener may
be taken as 0.75h because of being securely connected to the web.
The effective bearing area A: is taken as the area outside the flange angle
'ffiet oi ihe flangetoweb
&~ds.
design requires co
ing L r / r to find the allowable column stress Fa
$$@$$:eking
<
PY
$4
fif
:i r
In the AISC specifications, Eq. (1017) is directly combined with the appropriate
kc term for displaying the design equations.
Ir
Example 104 Design a welded plate girder to support two columns span
ning an auditorium space in a highrise building. Floor loads deliver cequivalent uniform load to the top flange
of 2.8 kins/ft  (not
i..
n..
r.
l.i i d i n a t&he
,"
girder weight): General span and loading is a's shown in Fig. E104a. We
f l n n a ~st, the ends and at the
will assume lateral bracing" of the comnressinn
 rI  concentrated loads. Other design data: E60 electrodes,
.
 ,n,..~.,,,r,,i,.f i r tai n n c
ATSC
s,
A36 steel, and girder depth limited to 84 in.
Figure ElO4b
90k
FO;
h/r
For
162, obtain t
78/162
h
Maximum  =
=
0.5 in.
105'

fi
2R
= 343.4 kips
bh3
I,, = 12 + 2,4d2
$8
760
,It
162 (approximately)
.,
7
0.24(53)~
= 84.3 ft  kips
8
The total bending moment = 2722 84.3 = 2806.3 ft . kips.
 M 2806.3f12'1
JbC=:
19.53 k s ~
S
1724.5
M=
/
,
I
w
.
Check the web shear SO that the plate is not too thin (neglect beam weight at
j'lls pmnt):
1)
1 O(U.3
. .
13)
v.n.
= 14 829.8
+ 48
We must investigate both th; end panel gnd the Intenor (between two
columns) panel smce C, is different for each locat~on.For the end panel:
L = 18 f t (largest value), and m , = 0 and ibf, = 2664 f t . klps
> 2664
s, = 63822
1600 ln3
39.875
0.490
New girder weight =
(3 1.5 29.3) = 0.207 kip/ft
144
'be approximate additional moment due to girder weight is
M=
For this L / r , ratio, F, = 0 . 6 5 = 22 ksi for both the end and interior
panels.
Now check AISC Eq. 1.105, since h / t w = 208 > 7 6 0 / a :
0.207(53)~
MtOw = 2722
= 72.7
2795(12)
"4, = 1600 = 20.96 ksi
Continue the design. Use two flange plates x 18 in and a web plate
3
y X 78.
==
2 1.05
O.K.
 At this point the bending stress and slider proportions are adequaie
unless a later interaction check requlres a revlsion of girder section.
Step 6. Compute the stiffener requirements.
AISC specifications require bearing stiffeners under reactions and the
two concentrated column loads. In the end panel the actual shear stress at
the reaction is

h = 208
tw
= 79.75 m
SX=1600in3
A, = 29.3 in2
Af=15.75in2
f ==
A,
173.25 + 0.207(53/2)
 = 6.11 ksi
0.375(78)
According to AISC Sec. 110.5.3, intermediate stiffeners (other than the end
I]
4'25 36
0.825  0.375 7.67
C" =
45 OOOk  45
q ( h / tJ2
36(20gL)
is limited
36(0.169)
= 2.1 1 ksi << 6.1 1
N.G. stiffeners required
2.89
Try one stiffener at half the distance (noting that if this works for the 18ft
panel it will also work on other end in 17.5ft ane el):
F" =
a
h
108
78
 =
1.38 < 3
22.2
> 22 ksi
O.K.
(Y1
33.01 kips
> 1.38
O.K.
F, = 36(0'170)
2.89
= 2.12 > 1.13 ks,
With intermediate stiffeners and C,
Eq. (109):
< 1, we
Since 7.67 ksi is greater than the actual shear stress at the bearing stiffener,
6.1 1 ksi, and the shear is less at the interior points, it is not necessary to
check the shear stress further for stiffener analysis. Note that Fv could have
O.K.
no st~ffenzrs
required
Step 9. Check the web cnppllng under the compression tlange due ti
uniform load.
Assume that the flange is restrained against rotation (since it carria
uniform load). The load carried in compression to the web is 2.8 kipsift t
weight of top flange. We will neglect flange weight, so that the compressiv
stress is
2.8
jc = 0.375(12)
= 0.622 ksi
F = IO"*[5.5
( h j tWl2
1.4
> 0.622 h i
O.K.
208'
2.69'
Note that if F <f,, we would have to either add stiffeners to decr~asea
'or'inctease t,.
Step 10. Design bearing and intermediate stiffeners.
For bearing stiffeners at girder ends (we will use the same size
column loads) try two 8 X 1/2 in bars, for a width (2 X 8 0.375 =
in) approaching the width of the flange plate of 18 in.
b
  8= 1 6 ~ 95
O.K.
t
0.5
=
 &(5.5
'
In terms of A,, x 100/A, this value could have been obtained from TabIe
118 by using double interpolation. Try two bars 3/8 x 6 in:
A
2 ( : x 6)
O.K.
16
O.K.
0.375
The minimum moment of inertia of the stiffeners is
=
fi
oS(16.35)~
= 182.9 in4
12
The "effective" column area for the radius of gyration (see Fig. 1013):
I =
A = 16 x 0.5
=dF
182.9
9.69
76.5 in
jU= h[(&y]l/'
O.K.
Check the bearing stress; assume a 5/16in weld, so that the effective
bearing area = (8  0.3125)(0.5)(2) = 7.7 in2:
P, = 32.4(7.7)
kips/in
[ Eq. (1016)]
= 4.34 in
O.K.
For intermediate stiffeners, with only one intermediate stiffener use two
plates.
1.0 and Y
1 (A36 steel
For a pair of stiffeners, we have 2.7 kips/in; for each side this becorn&;
2.7/2 = 1.35 kips/in.
For 1/2in bearing stiffeners, use a 3/ 16in weld and E60 electrodes.
F,,, = 0.1875(0.70711)(0.3 X 60) = 2.39 kips/in
Use a 3/16in fillet weld continuous for both bearing and intermediate
stiffeners.
Rationale: Few stiffeners and the weld can be made in one pass. I t 2
too difficult to measure and set up alternating weld distances and gaps.
Step 12. Design welding to fasten the flange plate to the web.
Check the end for maximum shear:
Use a 5/16in (t, > 3/4 in) continuous weld on both sides at F, = 392
kips/in. Note that the weld 1s considerably overdesigned, but for an im
portant girder the use of intermittent welds is not worth the savingsularly if the weld can be made in one pass, as here.
Figure E104c illustrates the design summary for the girder.
490
105~
17 5'
AREA:
When crossties rest directly on the girder flange:
18'
16
?  8 X j t
2  6 x i 8 ' t
Web
X 78"
1 28X+"e
12
6X
:"t
X 18"
AND AREA
Z1
Plate girder design using AASHTO and AREA specifications is very similar but
more conservative than AISC because of the more hostile environment to which
the girder will be subjected. In general, however, the same general considerations apply:
1. The girder is proportioned by the momentofinertia method.
2. No uniq'ue solution is possible.
3. Shear and stiffener requirements are more rigid. Bearing stiffeners are always
required, but AASHTO allows use of longitudinal stiffeners.?
107.1
,,
Fb 1 0.555
However, if the compression flange is laterally unsupported in a length L, the
stresses must be reduced as in Part I11 of SSDD and in the appropriate
specification.
The b/2$ ratio for the flanges is also limited:
AASHTO:
1.19Fb, we
fi
lw
For A36 steel, the limiting h/t, = 171. When the value of fb
compression flange, AREA allows an increase in the h/t, ratio:
< Fb in
the
Alsg
Fb = 0.55Fy.
tw 2 5/ 16 in (8.0 mm)
AREA:
."a
..
PWTE GI&LRS
4 ~ 3
'
1.64, to obtain
fps:
v3
'W
fi
234
h
618
5tw
lw
The limiting h / t w = 68 for A36 steel when f, = F, = 0 . 3 3 ~ ~ .
AREA requirements are slightly more conservative:
h
360
h
950'
fps:  2 SI:  <lw
The limiting h / t w = 60 for A36 steel.
Intermediate stiffeners are not required by either specificati& if the h / t w
ratio is less than that given by Eq. (1020) or (1021) and may not be required
for larger h / t w ratios as given in the following section.
M~'hen h / t w is larger than given by the preceding equations, the stiffener
: '1 spacing is obtained from Eq. (35) by
obtain

<
vx
S1:
ips:
332t
a =  (in)
fi
875r
SI: a=
fl
(mm)
In these equations t has units of in or mm. The stiffener spacing is limited to not
more than 72 in or 1800 mm when using Eq. (1022).
fi
fi
where t has units of in or mm. The previous two editions of the AfSC
specifications included this requirement for spacing the first interior stiffener
from the bearing stiffener. Currently, the XASHTO specifications for intermediate stiffeners is somewhat more indirect and is given in Sec. 107.4 foI10wing under "Intermediate Stiffeners."
The AREA specifications reduce the 348 factor slightly, to obtain
fps:
.I_
From an approximate aqeraging of fi we obtain 348, which if used to backconpute the SF gives 1.37 at a / h = 0.5 to 2.02 at a / h . 1.0. In earlier AASHTO
specifications the stiffener spacing could be computed as
348 t
914t
fps: a =  (in)
SI: a =  (mm)
h <3840
1455 ' S I : 5 
"
>
IJ = 11t1.[24(%)2 0.131
from which
=
kT ~ E
12(1p2)SF

 0
tw
0.5
0.6
0.7
0.8
0.9
1 .o
25.36
18.83
14.90
12.30
10.58
9.34
332.8
344.4
357.2
370.9
387.2
404.0
SI:
b

188
5 
<i
(1 024
Fbrg = 0.80Fy
AREA:
Fbrg= 0.835
feners if
h
<I50
tw
and
f,5Fo
where
I
ti
& [ I+ ( h / a ) 2 ]
where C= I I
~,(h/t,)~
'
B = 222 000 in fps units
= 1 520 000 in SI units
ere
I
AREA:
width 2 2 in
thickness 2 width/ 16
same as for AASHTO
107.5 Interaction
.\ASHTO specifications include a bending stress reduction if the shear
f, I 0.6Fv according to
Fi;rve 1015 Bridge bracing using diaphragms. (a) Diaphragms used with five plate girders. (b) E
6i:.phragm. Must be sufficiently set back from adjacent diaphragm or abutment so that it can
i*.?ected and periodically painted.
Figure 1016 Longitudinal and cross bracing. Note use of verrical and longitudinal stiffeners. Bridge
in background of (c) is older and uses riveted construction where bridge in foreground is alIwelded
The three girders carry a fourlane roadway plus shoulders and walkway for both bridges. ( a ) Cross
bracing. (b) Fabrication details. Note that the end fastener plates ire welded to both web and fiansOn continuous bridges these plates provide flange brace points. (c) Use of both cross bracing and
longitudinal truss bracing. Note longitudinal stiffener along compression flange, where h / is
~
excessive.
Cross b a r n s a t 2 . 3 f r
760 min
Figure E105a
:IR+
16
71
Figure 1017 Top flange bracing using web plate and bracket for the throughgirder ballasteddeck
rallroad bridge shown m Fig. 104.
H/ V
The approximation of the axial force using the sin defined by H and
than the actual length is sufficiently accurate for design.

use 28 percent
(Note that tlphimpact factor is not the same as used in Example 415
design momecf
..
M, = 9984(1.28)
12 779.5 kN . m
V,
1680(1.28)
2 150.4 kN
We will neglect the live load on the pedestrian/semce walkway. With rhis
assumption the deadload estlmate is
0.025&Af
Pkb
+ 27.5  9 = 27.6
v rather
= 10.0 kll
= 14.6
x 4.45 = 1680 k~
=4.1 kN
kN
= 6.2
=2.9 kN
kN
42.2 kN/n
= 4.4
0.300 x I4 59
Total
= 377.4 kips
kN
V = 274.5(%)
> 25 m as
Note that some approximation 1s used for the steel deck to allow for
forming the trough to hold the ballast to g v e an effective length of 6.49 m,
Note also that an estimat~ngfactor of 10 percent is applled for
<
S T R U C T n k L Ki'W:L3
1 )I: , I C ~ ? ;
P'LATE G~RDERS
42'2(27'5)2 = 3990 kN
8
8
Tne dead load shear (at reaction) is
v = WL
=
40.2Q7.5) = 553 kN
2
MdeSi,, = M L


"
hF,
Ihe,maximum
,.,.
, , . ..,...
shear is at the reaction, but with a moving load contributing
.he major effects (with impact) will not change much for some, distance
along the girder, so we will take t,,, = 16 mm.
~
 ==
tw
815 rnrn.
2730.4
2.034(0.35 x 250) = 15.3 mm
NL1:
1:
(O.K. so far)
2034
16
O.K.
'
16 580
32.54
 2034 X 0 . 5 c
6
A 
Flange area
m2
A w  1.994 x 0.016

0.0052 m~
total
= 0.06237
m2
b 
fi
 =  =1994
16
tw
O.K.
bftf3
Ix = Iweb+ 2 ~ # ~
12
720)' +
12
= 10 571 + 121 510
+ 24 = 132 114 x
lOW6m4
815(0,070)~10~
12
fI
jv==
ht
'
2730.4
= 85.6 MPa
1994(0.0 16)
< 0.355
O.K.
0.872(16)
1.5 1 m
<
1.828
O.K.
764
x 1.28 = 1234 kN
290 kN
= 1524 kN
and
1524
jv = 1994(O.Q16) = 47.8 MPa
Figure E105c
9
= 58.88
+
;"I
m3
j
u =  =V Q
2730.4(58.88 X
I
0.132114
, The shear resistance is limited to
I Field bolt
= 1216.6 kN/m
~ I O O IO V J ~ ~r I
ou
A I O U . I I 1.30
s h o p weid
Framing angle, designed ln Example 89
F, = 0.35F,ASh,, = 0.35(250)(0.016>(1rn)
/ x lo3
"
 / \ 
~twnlu/rn~1~10.0
"
Figure E105d
0.K.
8M
L~
.The.dsflection is approximately
A =   =5wL4
384EI
force
5(0' 1773)(27'5)4
= 0.04996 m (say 50 )
384(200 000)(0. 132114)
27.5(1000)
= 43 mm = 50 mm
(take as O.K.)
Amax =
640
Step 9. Design the upper flange knee bracing.
We shall arbitrarily place a brace on every fourth floor beam f
spacing of q0.76) = 3.04 m. Where this coincides with or is "close eno
to a stiffener, we will increase the stiffener thickness from 8.5 mm to 1
For any bracing plate alone, we will use a 12mm plate. Shopwe
stiffener or brace plate to girder web and compression flange as shown i
Fig. E105d. Fieldweld the bottom of the stiffener brace plate to the c
beam. Shopweld the angle for the bracket and punch for three 22highstrength bolts at each end. Take the horizontal component of stiffener
Pa
t,
13.21 m m
O.K.
Annvec 1638 kN m.
&
d!,S
1 10(4.45)(0.25) = 122.4 kN
End panel:
Interior panels:
122.4 +
90
= 143.01 k~
2
ave a separate set of design data of length depth and load. Be sure to w th
175 to 250 kips, L = 40 to 60 ft, and D = 60 to 96 in, as assigned. Use A36 stre
electrodes, and the AISC specifications. Assume lateral support at the ends and load points.
If no specific data are assigned, take P = 2 15 kips, L = 54 ft, and D = 80 in.)
L
*dtr)
bC
(,vt,
D ~ S ~a W
welded plate girder for a singletrack opendeck railr
,rrack contributes 0.25 kip/ft loading, including rail, ties, and so on.
loading and span from 40 to 110 ft as assigned. Use A36 stee
s~.zications.Limit the overall girder depth to L/10 2 D 2 ~ / 1 5 ,
:t. Take the girder spacing S 2 L/15.Design sway bracing and/or diaphragms in
stby flanges, stiffeners, and welds. (Note: If specific problem data are not assigne
E1 i0 loading, L 90 ft, D = 96 in, and s = 6.5 ft.)
&12 Denim a welded plate girder for a singletrack opendeck railroad bridge. A
tr7;k contributes 3.65 kN/m of loading, including rail, ties, and so on. ~~~i~~for
Cl
live loading and span from 14 to 34 m, as assigned: Use the
ad depth D but limit D to between 1.22and 2.45 m. If specific problem
CooPr E80 loading, L 23 m, D = 2.3 m, and S = 1.6 m.
'
. ,.,.
The following steps are required to run the program, with tenninol
matching the computer program listing.
1, Code the structure according to rigid frame, truss, or combination T
account any hinges, as in chap. 2.
1Determine NP. The computer program computes NPPl = NP
S E U C I ED
careful attention to signs. The output signs are interpreted using Fig
e program computes the fixed end moments, the output is automati
corrected for the FEMs to obtain the design values. Member forces do not
to be corrected for any node forces read in other than moments
'
m for mom
of gravity
X 1, X2 = horizontal distance from origin end of member of P1 a
or m
Vl, V2 = column shears from alternate direction for a space frame, V1 =
near end shear; use kips or kN and sign same as P1, P2
3. Put a blank card at the end of the member data.
used for
~LW;
NNZP
SELECTED C O W
IWINRE
IOLPL
UNIT^
UNIT,,
..
..
"
Y9
K OR K N
I N OR
* ,
lrPMIN(I15I
DATA F F U 1 1 2 ~ r 1 0 0 0 . r 1 . ~ 1 0 0 0 . ~ 1 ~ ~ 1 0 0 0 0 0 0 0 1 1 2 ~ t ~ ~ ~
READ CARD
2 R E A D ( ~ , ~ ~ O O ) L Z Z , U N I T I ~ U H I T ~ ~ U ~ ~ ~ Z ~ U ~ ~ T ~
000 F O R M A T ( Z O A ~ , I , A Z S Z X S A Z ~ ~ X ~ ~ ~ A ~ S ~ X ~ )
IF(EOF(lll750~3
3 WRITE ( 3 r Z 0 0 0 ) ZZZ
2000 F O R M A T ( ~ ~ ' J ~ ~ S ~ X ~ Z O A ~ ~ ~ ~
REAO CARO
R E A O ( ~ . ~ ~ ~ ~ ) N P J N ~ ~ ~ N L C P L I S T R , N L Y ~ N N Z P J I ~ E ~ #
~IDLPL,IYINBE~IWRITX~IW~ITP
1 0 0 1 FORMAT(lbI5l
L++*+*
REAO MODULUS OF E L A S T I C I T Y  K S I
READ(lr1002lE
1 0 0 2 FORMAT(BF1O.C)
 *
D E F I N E U N I T CONVER$ION FACTORS
IUNIT
1
IF(IMET.GT.O)IUNIT
2
0 0 4 0
1 . 1 ~ 6
FU(I)
FFU( I U M I T s I I
4 0 CONTINUE
NPP1
NP
OR B P I
***********
I
I
SELECTED CObIP
2 0 4 CONTINUE
DO 2 0 6
1 = l r 6
IFfUPE(I).GE.NPPl)GO
'
.


NS2 = N P E f J l
NPE(II + 1
S T I F F I N S 1 4 NSZ) * STIFF(NSI+NSZ)
2 0 5 CONTINUE
c
c
EASAT(1,J)
5 5 0 0 GO TO 1 0 6
THE B A N D M A T R I X
IS N O W F O R M E D F O R R F O U C T I O N
WRITE BAN0 R A T Q I X I F L I S T 4 > 0
72R IF(I1.GT.O)GO
C
H2
C
IN
ISTIFF
IN
COPE
TO 6 0 1
NSANO*NP
IF(LIST8.LE.O)GO
TO A 9 R 9
WRITE(3r2009)
2 0 0 9 FORMAT(
1 1 0 1 ' T H E RAND " A T Q I X W I T H 1 0 0 0 F A C T O R E D ' n I l
ni = 1
R2
N8ANO
MCOU
NP*NBAHO
On 3 0 5
I = 1.NP
W R I T E ( 3 r Z O l O ) I r ( S T I F F 1 II?.;~.MI,MZ
1
2 0 1 0 F O R M A T ( ~ X I I ~ , ~ X I  ~ P ( ~ F 11
~ " ~5 .X ~~ ) 3~ P ( 9 F l Z . 2 ) ~ / , 5 X 1  3 P ( 9 F 1 2 . 2 ) 1
1FIMZ.GE.NCOUIGO
TO 8 8 4 9
11 = M2 t 1
M2 = M2 t NBANO
r
3 0 5 CONTINUE
8 8 8 9 WRITE(3.20111MZ~NRANO
2 0 1 1 F O R M A T ( I / ~ T ~ I ' M O S T I F F ( 1 I ENTQIES
' ~ I ~ . ~ O X P ' S A NY TOD T H . ' r I 4 t l / l
C
C
C
NOTEDO
NOT READ P  U A T P I X E N T Q I E S COW UNFOQM LOADS OM 3 E I N S
C
DO NOT READ P  M A T Q I X E N T R I E S FOR FEMINOUT
SO COUPUTEQ
C
COHPUTES F E * 5 0 F I N A L MOMEMTS A Q E COQDECTEO FOP FEM * * * * * a *

TO 4 2 6
IF(NNZP.EO.OIGO
0 0 9_91 NN = 1,NNZP
READ( l r 1 0 1 0 l t ! l
C
1 0 1 0 FORMATf1615)
C
READ CARD******
READ(1,1011)(PRfLI~L*l~NLC)
C
IF(II.GT.09GO
TO 6 0 5
203 0 0 204
I = l,b
0 0 204
J = 1.6
EASAT(1.J)
0.
DO 1 8 7 K 9 1 . 3
E A S A T f I # J ) * E A S A T f I p J ) + EAfI,K)*ESATfK.JI
1 8 7 CONTINUE
TO 2 0 6
1011 F O R M A T ( 8 F 1 0 . 4 1
OD 9 9 0 L = 1,NLC
P(n1.L)
= Pfn1.L) + P R ( L )
9 9 0 CONTINUE
9 9 1 CONTINUE
4 2 6 CONTINUE
DO 9 9 2
NS1 9 l r N P
DO 9 9 2 N S 2 9 l r N L C
9 9 2 T F f N S l r N S Z l = 0.
,.
************~*****;~******T~***lOtOOO~OOO.**
4 0 6 URITE(3,2012)UNITl~UNIT2
2 0 1 2 F O R M A T ( / * Z X s ' T H E PMATPIX,
' r A 2 r 1 AN0 ' , A 4 r / )
NS1 = 1
NS2 = 1 0
4 2 7 IF(NSZ.GT.NLCINS2
= NLC
DO 4 0 8
I = ltNP
408 V P I T E ( ~ ~ ~ O ~ + ~ I ~ ( P ( I ~ J I ~ J J N S ~ ~ N S Z ~
5 1%,STRUCTURAL STEEL
DESIGN
i 0 1 4 F O R M A T l T 5 r 'MP
',13~1X,lOFll.2)
IF(NS2.EQ.NLC)GO
TO 4 2 8
N S l = NSZ + 1
I
NS2 9 NS2 + 1 0
URITE13r2016)NSl
2 0 2 6 FORHbT(11,5Xr1THE
PMATRIX CONTINUED BEGINNING U I T H NLC
GO TO 4 2 7
6 2 8 CONTINUE
C'
67
70
66
80
C
C
= P(NtM)
IFfSTIFF(NK).NE.O.)P(NrM)
ENO OF MATRIX
REDUCTION
c
C
C
F(Z)F(Zl*3~/4.
F(3l*F(31*3.14.
IF(JJ.NE.IDLPLIG0
Tn 973
C
PUT 0 t L I N T F  M A T Q I X
00 970
I
193
9 7 0 TF(YEMNO.11
* F(I)

SOLUTION
IS
IN
P(I.JI
1.58 CONTINUE
I F ( 1TRUSS.EQ.O)GO
TO LATER A D O TO Y I V O L O I D COHOIT.
TO 1 8 9
'
4 3 1 CONTINUE
I F AOOEO TO WINO
STIFF(NK)*P(L*KpM)
I C 1 7 F O R M A T ( I I r 5 X r ' T H E XHATRIXp
',A29
OR RADIANS',/)
NS1 = 1
NS2
10
4 2 9 IF(NSZ.GT.NLCbNS2
= NLC
DO 5 0 3
1 = lrNP
503 U R I T E ( ~ ~ ~ O ~ ~ ) I ~ ( P ( I ~ J ) P J * ~ ( S ~ , N S Z )
L O 1 8 FORMAT(6X,'NX
9
*rI3~1XrlUFll.5)
IF(NSZ.EQ.NLC)GO
TO 4 3 1
N S 1 = NSZ + 1
NS2 = NSZ 4 1 0
~ ~ 1 T E 1 3 r 2 0 1 9 ~ ~ S l
2 0 1 9 F O R M A T ( l l r 5 X r ' T H E XMATRIX CON1 D BEGINNING WITH NLC
9 6 CONTINUE
GO TO 8 5
\fi CONTINUE
C
NOTE I F YOU PEA0 DHATQIY ENTRIES FOR FE* THEN OUTPUT *UST 5 E
AOJUSTEO BY HAHO TO ACCOUNT FOP F I X E D END MOMENTS DM 9 E A M ENDS
6 0 7 CONTINUE
Fl2l(F(ZIFEMlI/FU(l)
602 IF(ITRUSS.EO.O.OR.IWPITP.E~~~)WPITE(~~~O~OIJJPUNIT~~UNIT~
2 0 2 0 F O R M A T ( I P 4 x 9 'LOADING CONOITION NO 9 ' r I 3 r l r
5Xr *MEMBERit 4 Y p ' A X I P L
1 37X. 'DESIGN END MOMENTS C0RRECTEO'r/,
7FOWCFs ' r A29 8Yr'FOR FFM AN0 WINO (NEAR FNO F I P S T l r ' , A L P I ~ ,
RFWINO 5
GO T O 5 5 5
6 0 5 P ( N P P 1 r JJ).O.
FEU1 = F M I I J J I
COUNTS NL
IF(NLC.EO.1)GO
TO 1 9 5
IF(F(1I.GT.PMAX(ME*NOIIDMA~(MFRNO)
IF(F(1I.LT.PMIN(MEHNOIIPMIN~~EMNOI
* F(11
TF(HEMNO.JJ)
F(1)
I F ~ I W R I T P . L E . O ) G O TO 4 4 1
WRITE 9AR FORCES FOR NLC
F(11
1 AND r V Q I T P
IF(MEHNO.EO.~IURITE~~PZOZ~~JJ
FORCES [ K I P $ O Q K N I FOR NLC
2 0 2 5 FORHAT(5X,'BAR
'rI3*11
5U)
. .
STRUCTURAL STEEL DESIGN
SELECfED COMPUTEX
JJ = JJ 4 1
GO T O 1 8 9
195 JJ
1
1 9 1 GO T n 5 5 5
1 9 2 I F ( I T R U S S . E ~ . O I G O TO 1 9 3
IF(NLC.EO.l)GO
TO 1 9 3
REWIND I
C
I T R U S S = 1 AND N L C I S L A R G E U S E I W R I T
0 FP
O R M A X AND IN
B A R F O R C E S ONLYUSE
I W R I T P = 1 F O R C O ~ P L E T E L I S T I N G
B A R F O R C E S WHEN N L C IS L A R G E R THAN 1
C
VRIT~(3,2033)
2 0 3 3 F 0 R M A T ( 1 1 ' 8 X * ' T ~ ~n A Y I n U n L I V E L O A D B A R F O R C E S & N O D E A D L O A D v A L U E
1S"lr5Y'qnEf4
N O ' P ~ X Y ' M A X L L 1 , 9 X y * M f N L ~ q , f j x ~ rL O
~ A~D ~ORo L ~ S TN L C
z',/T
0 0 447
.
7
lrNn
~ R I T ~ ( 3 r 2 0 3 5 ) I ~ P n ~ ~ ( ~ ) . ~ n ~ ~ ( ~ ) , ~ ~ ( ~ y ~ ~ ~ )
FORHAT(6X*T ~ ~ X ~ F ~ O . ~ Y ~ X , F ~ O . Z ~ ~ X , F ~ O . ~ ]
4 4 2 CONTINUE
W R I T E B A R F O P C F n A T a I v IF I W R I T P
I
IF( I W R I T P . N E .  ~ ) G O T O 1 9 3
WRITE ( 3 1 z 0 3 z ) ~ ~ 1
2 0 3 2 F O R n A T ( I I * 5 X ~ ' T B~ A~R F O R C E M A T R I X
4 4 3 DO 4 4 4 1 . 1 , y n
S T A R T I N G WITH
NLC
qy13y11,
OF
444 ~ R I T E ( ~ ~ ~ ~ ~ ~ ) I I ( T F ( I ~ J I ~ J ~ N ~ ~ ~ N ~ ~ )
2028 F O R ~ A T ( ~ X ~ I ~ , ~ X Y ~ D F ~ ~ . ~ )
IF(NSZ.EO.NLC)
GO TO 1 9 3
NS1NSZ41
NSE=NSZ+~O
IF(NS2.GT.NLC)NSZ.NLC
URITE(3r2032)NSl
GO T O 4 4 3
1 9 3 GO T O 2
C
7 5 0 STOP
END
Variable Identification
A sample set of data cards are llsted ulth thls program for both an fps and an
bridge truss using three wheel loads, seven truss panels (25 ft or 7 5 m),
heels 5 ft or 1.5 m. The coding is such that the panel NP
are 26,2, 6, . . . and NPPI = 26 here and is not punched
number of wheel loads
number of truss panels
step length, ft or m
increment of wheel movement, f t or m
STEP
0 for fps; = 1 for SI bndges. This is used to proper1
MET
impact factor.
wheel loads. Note the order of using loads on the sample cards. U s h g
8. 32. 32. runs the truck forward on the bndge.
cumulative wheel spaclng as 14. 28. for using 14 ft between each set of
X(1)
wheels. This spacing or the SI equivalent wdl usually be the most
critical spacing.
NP numbers in order from left to nght where the panel loads are
Ml(1)
placed. M1 has the same meanlng here as in the frame analysis
program. Note that NPPl is not used. For a through deck truss, this
section of the program requlres a sllght modiflcatlon to get loads onto
the truss where NPPl may be, since the current method of use omits
the flrst and last Ml(1) values (they are input for ~dentdication).
922
r
S T R U C T U R A L STEEL DESIGN
C
(
r
'
1 FORMATt215r2Fl0.4,15)
i,
WRITEt3v5)NPrSL,NSvSTEP
WRITE(~~~)~IIP~I)~I=~~NP)
b FORMATt///.*
P'.Ilr'
=lrF8.3)
WRITE~~.~)~JIX(J)~J=I~NXI
'
REA0(1.81(MltI)rI*lrNSPl
r//)
R E A O ~ ~ ~ ~ ) ~ P ~ I ) ~ I = ~ ~ N P ) ~ ~ X ~ J ) I J = ~ ~ N X )
'
2 0 1 CONTINUE
I F (XIMPAT.GT.0.30)
XIMPAT0.30
00 1 0 3 I1.MZ
W 1 0 3 J1,NSPl
103 REACT~J~II=REACTIJ,II*~~+XIMPAT)
WRITEt3r4)
4
FORMAT( I / / / / *LOAOm,/*
'
CONOITIOW9.TZO.'
THE R E A C T I O N S ' )
0 0 1 0 6 J=11M2
a
1 0 6 W R I T E t 3 . 3 1 Jt ~ R E A C T ~ I I J ) ~ I  ~ , N S P ~ ) "
*. t
3
FORMATtT5,13qlOt5XvF6.1))
PUNCH M l FROM L E F T TO R I G H T ACROSS fk T R U S S "
C
PUNCH
FORMAT
I
S
COMPATIBLE
W
I
7
H
A'NALYSIS
PROGRAM FOR DIRE$$
C
7 FORMATl///r9
X 1 1 I 1 v ' =',F8.31
SK=O.O
SK2=0.0
DO 1 1 0 M = l v N P
DO 1 0 9 I = l r N S
IPl=I+l
DO 1 0 8 J = l , N L S
RR=P(M)*tJ*STEPSKSK2)/SL
ALR=PtMIRR
I F .LJ*SJEP+STEP .GT.
SLI
SKSLJ*STEP
JZ= tSL/STEPI*tI11
J3=J+J2
TEMPERARY STORAGE OF R E A C T I O N VALUES
TEMPtIvJ3)=A1R
TEMPfIPlrJ3)RR
1 0 8 CONTINUE
1 0 9 CONTINUE
L
PLACE CONSECUTIVE LOADS I N T H E I R PROPER P O S I T I O N
IX=XtM)/STEP
IST=IX*STEP
SK2=Xt M I  I S T
INT=O
I F tM.EQ.1)
GOT0 1 0 5
DO 1 0 7
I=lrNSP1
DO 1 0 7 J = l r J 3
n n i = n1
INT=XfMMl)/STEP
M2=J+INT
C
ADO TEMPERARY STORAGE TO PERMANENT STORAGE
107 REACTII~MZ)=REACT~IIM~)TEMP~I~J~
GOT0 1 1 0
105
00
104 I=ltNSPl
DO 1 0 4 J = l r J 3
1 0 4 REACTtI.J)=TEMP(1.J)
1 1 0 CONTINUE
IFtMET.EQ.1IGOTO
200
XIMPAT=5O./tNS*SL*125.1
GOT0 2 0 1
700 X I M P A T = ~ ~ . ~ ~ / ~ N S * S L + ~ ~ . ~ ~ I
FORMAT( 2 0 1 4 I
00 1 1 2 I=Z.NS
WRITE(2v9)MltII
FORMATt15)
8
9
112 W R I T E ~ ~ ~ ~ O ) I R E A C T I I ~ J I ~ J J ~ ~ M ~ I
10
FORMATtBF10.41
4000
STOP
EN0
****
****
I*
INPUT
,$i
1 FaR M E T R I
ISUIT
NO OF SPANS OR GIRDER SEGMENTS:
OX = INCR. OF LOAD MOVEMENT. F T OR M;
SPAN
LEN
FAC = L O I D R A T I O I F F P = 8 0 LOAD
TRUSS SPAN OR GIRDER SEGMENT:
F A C = 0.75
FOR E ~ W  0 . 9 FOR E721.00
FOR E B O   L . ~ ~F~~
~
~ 1 1 0 NOSPAN
c
c
c
= 0 FOR FPS:
WIOH
IWRIT
= 1 TO WRITE
'
Variable Identification
(see sample data cards listed at end of program listing)
TITLE
up to 80 columns of alphanumeric data for probl
NOSPAN
number of panels
ISWIT
0 for fps; = 1 for SI prob
spacing to SI and to
IPUNCH
0 for not punching O U ~ P U
= 1 to punch output in f
frame analysis program
for no impact factor;
to write selected inter
am; = 0 when pro
crement of wheel movement left to right, f t or m
ISPAN
panel length, ft or m
WIDH
width between two trusses making up bridge an
computing the impact factor, ft or m
FAC
factor to convert E80 to E110, E60, etc.; use F
E80 loading
NPs(J)
NP numbers (same a
established (note that these entries do not inc
cation may be required for a through deck b
w.
@
i
i$
%
"r,Tt,
~ E ~ 0 1 1 ~ 1 0 0 2 1SPAN,UIOHHFAC
0 X ~
'
1 0 0 2 FORMAT(BF10.41 N P S ( J ) = NP N U M B E R S F O R NONZERO
*MATRIX
ENTRIES
NSPN = NOSPAN
1
REAO(lr100611NPSIJI .J=lqNSpNI
0 6 FORHATl1615)
I F 1 1SWIT.GT.OIDIS
= 25.
8 = NOSPAN
TOTSPN = B*SPAN
SET WHEEL LOAOS AND WHEEL SPACING
SUHXL = 0.
0 0 5
I1.18
X I 1 I = FFXI I 1
IF(ISWIT.GT.OIXII
I = x111*0.3048
P I 1 I = FFPI II*FAC
F
I( ISWIT.GT.OIPII
I = PlI)*4.44822
SUMXL = SUHXL + XI T I
5 CONTINUE
7 PAC = 1. + PAC
~ ~ ~ ~ ~ ~ ~ , ~ ~ ~ ~ I T I T L E . N P , H O ~ P A N , O X , S P A N ~ ~ L O A O ~ S U ~ X L ~ P A C ~ F A
2000
F O R ~ A T ( / / / ~ ~ X , Z O A ~ , I , ~ X ~I
' N* PV I ~ V ~ X * ' N O S P
= ~ ' N* 1 5 * 3 X *
1 2 x 1 I D X = o r ~ 7 . ~ l / r 5 ~ , ' S P A N LENGTH = ' . F 8 . 3 r 3 X v ' U N I F
LOAO/LENGTH
2~,~6.l,
/, ~ X , * S U H OF X ( L 1 OF CONC. LOADS a ' t F ~ . Z I ~ X V ' I M P & C T
~ A C T D R= * , ~ 6 . 3 , / , 5 ~ ,
IFACTOR FOR ELOADS1.0
FOR E  8 0 = ' *
465.3,
~ x , S O I S T CTOC
OF TRUSSES
'rF731//)
,145 F O R M A T I S X v ' P l I i
WRI T E ( 3 r 2 0 0 6 )
2906 FOPMAT(lOX.'***
4
I.,
6F9.1.//)
LOAOS M I 0 IMPACTUSE
.TOTSPNIXILI
IFlL.EQ.NP+1.WO.AOXH.GTTOTSPNIXIL)
IFIL.GT.NP+l.OR.XlLIGTTSPANIXIL)
'
j4$!JZSTOP
= TOTIP. + SUMXL + o x
0 0 8 3 K=1,20
DO 7 3 L = 1 . 2 1 0
7 ' PR1K.L)
= 0,
~ ' iCONTINUE
DO 9 5 L * NCIZOO
IFIL .EQ.NP+I.ANO.AOXH.LE
= Z*SPAN
SUMXL
= O I F F + Z*SPAN
SPAN
SWXL
IFlXfL)..LT.O.tGO
TO 1 4 8
IF(XIL).NE.SPANIGO
TO 2 2
0 1 = SPAN
DZ = 0.
2 2 sun1 = sun1 + XILI
IFlSUMl.GE.010.0001)GO
TO 9 6
IFIL.EO.NP+lIZZX
= SPAN
(OZ + 0 . 0 0 5 )
IFILO.NP+1~ANO.SUMl.GE.ZZXlGO
TO 9 6
9 5 CONTINUE
96 C O N T I N U E
SUM3 = SUM3 + SUM1
TOTSUM = SUM3 + SUM2
N C = L * l
02 = 01
RR = 0.
R L = 0.
I F 1 IWRIT.LE.OIG0
TO 5 5 0
u R I T E I ~ ~ ~ ~ ~ ~ ~ T O T S P N ~I MNr C 0~2 A
1 O
O lXt HS~U M l t O Z t NNv L.SPAN.SUH
1 7 SUM21 O I F F t T O T S U M v J J v X 1 1 9 1 ~ X 1 2 0 ) ~ X 1 2 1 1 t X f L ~
3 0 0 5 FORMATl3X.'TOTSPN
= 1 , F 8 . 1 , 3 X ~ e N C = ' 1 1 5 r 3 X , 1 A 0 X X H = *,F7.3,
2 3 X V 1 1 M = l.15,
5 X v e O Z = 1.F7.312X.'01
= '~F7.3,'
SUM1 *@,F7.3.
3/r5X*
'OZ a ' ~ F 7 . 3 r 3 X ~ ' N N  ' ~ 1 5 . 2 X , v L ~ 1 , 1 5 t 2 X T * S P A N = ~ , F 7 . 3 ,
4 3 x 1 'SUM3 = ',F7.3,3X,'SUMZ
=
.F7.3r3X1'01FF~',F7.3,3Xv
'TOT
5'rF6.lr2Xv'JJ~',IZ1.
/r5X1 'XI9 ~ ' ~ F 7 . 3 ~ 3 X ~ e X 2 0 = 1 ~ F 7 . 3 3 3 X ~ * X 2 1 = * 7
6F7.3v3XvQXL='.F7.31II
CONTINUE
ACCUMULATE EFFECTS OF LOAOS I N U I Y SPAN
0 0 9 9 L L = NN1L
IFILL.LE.NP)GO
TO 9 7
P I L L ) = WLOAO*XlLLl
IFILL.EQ.NP+l.OR.NN.EO.L)OZ
= 02
X1LL)IZ.
9 7 RR = O Z * P f L L ) / S P A N
RL = P I L L 1
RR
PTIKC) = PTIKCI
RR
PTIKC21 = PTIKC21  RL
02 = 02
XILLI
SUMP = SUMP + P I L L )
9 9 CONTINUE
SUMPT = SUMPT + P T I K C )
1FlJJ.EO.IM)SUMPT
= SUUPT + P T I K C  2 )
Z = 1M
JJ
OZ = Z*SPAN + TOTSUM
AOXH
NN = L L
SUM1 = 0.
1 0 2 CONTINUE
SO CAN
WRITE ONLY V E R T I C A L VALUES AT PANEL P O I N T S I H O R I Z = 0.)
U S E P R I N T E D OUTPUT TO OESIGN TRANSVERSE FLOOR BEAMS
BASED ON LARGEST PANEL LOAD * FLOOR BEAM REACTIONS
AN0 DEPENDING ON NUCBER OF TRACKS ON BRIOGE
OOES M T
I N C L U D E ANY IMPACT FACTOR *a*+****+
WRITEl3rZOO8)I~lPTIJl~J,Z~NY~2~
2 0 0 8 FORMAT(//.2Xv15v12F10.2v/,
5X,12F10.2,/1
WRITEI~~~OO~ISUMP~SU*PT
2 0 0 9 F O R M A T f 5 0 X , ' S U M TRUSS LOAOS = ' , F 1 0 . 2 . 3 X ,
'SUM NODE LOAOS = * ,
1F10.2,/)
0 0 1 3 8 J = 1,NY
PRIJIII = PTIJI
1 3 8 CONTINUE
1 4 0 CONTINUE
141 U R I T E I 3 ~ Z O l l ) I
2 0 1 1 FORMATI/I~SXI
'ONLY UNIFORM LOAO ON BRIDGE. I = ' , I 3 , / / )
IFlIPUNCH.LE.OIG0
TO 1 4 9
DO 1 4 0
A
I = 1.210
D I F F = 0.
ADX = A*OX
AOXH
AOX
C
******
F I N O NO OF SPANS TO USE
, 5 SPUSO = A O X l S P A N
ZERO
DO 8 6
CONTINUE
MATRICES
ARBITRARY
K = 19.200
AMOUNTS
X I K I = 0.
1
i
i
I
b
4..
0 0 8 8 K = 1,50
T I K I = 0.
CONTINUE
I M = SPUSO
RM = I M
IFI8M.LT.SPUSOIIM
= IM + 1
BM = I M
1FIAOXH.GT.TOTSPN)OIFF
= AOXH
;UM2 = 0.
? F I A B S ( O I F F ~ . L E . O . O O ~ ) G ~ TO 1 5
DO 1 2 MM = l v N P
>OM2 = SUMZ + x t n n )
:~lSUM2.GE.OIFFlGO
TO 1 5
1 2 CONTINUE
t K NC = tin + 1
fin = I M
6 7 = BM*SPAN
AOXH + SUM2
'WR = NC
LUM1 = 0.
5UM3 = 0.
cUHP = 0.
\UMPT = 0.
I~~~~.EO.NP.ANO.OZ.LE.O.)OZ
i!
TOTSPN
3 1 = SPAN
OZ
= 2*lIMJJ)
7 = JJ
KC
+ 4
**
= 0.
DO 1 0 2
C
C
In1 = I 1
142
143
25
2014
145
149
148
2016
150
**
C
...,,,..
**
J E
9
1.05
4
***I,
J E BOYLES
3 . ,...
9 8 .....
o...
3.45
C
/*
.
 "....'C,C.L~B:~$,.;<'~'M,~DRAS
...*...

'
II
<
. c$,.;.!
i
:,
.:. P!=3.
I ,
(,>t,:
 ...
3x0.
.,
,.. ,..F.
. . . .r.7.,
.,....*7.".,~,
BASLER, K.,
474, 479, 48 1
Beam analysis, 57
biaxial bending, 164
Beam analysis:
differential equation, 58
for unsymmetrical bending. 167
Beam columns, 297
design formulas: XXSHTOand
AREA, 311
AISC, 313, 318319
design methods, 322
effective length of, 303, 306
G factors for, 306
inelastic eRect/reduction, 307
K factor chart, 306
Beam framing connections, 385
Beam stresses, 146
allowable: Ar\SHTO and AREA,
15 1
AISC, 149, 174
laterally unsupported, 174
bending, 146
biaxial bending, 164
elastic design, 148
shear, 146, 159
Beams:
compact section criteria, 149
deflections, 146
laterallv unsu~norted.174
Lateral bracing:
.
for beams, 150, 155, 174
for columns, 291
for girders, 496
knee bracing, 497
Live loads, 18
reduction for, 19
Load conditions, 70
Load resistance factor design
(LRFD), 132
table of,@factors, 132
Loads, 18
bridge: AASHTO, 26, 27
AREA, 28
equations for shear and
moment, 26
impact, 32
tabulated shear and moment,
AREA, 29
dead, 18
earthquake, 33
hi:
eccentric, on fasteners, 375
impact, 32
live, 18, 19
ponding, 25
snow, 24
map for, 25
, wind,21,31
Lowtemperature effects, 13
Llr ratio:
compression members, 258, 260,
277
tension members, 231, 237, 238,
242
LRFD, 132
beam column desig*, 338
beam design, 208
column design, 292
connections, 405
tension member design, 248
Lug angles, 3,65..
Mass density, steel, 7
MUNSE, W . ,
364
Net area:
column, 264
in tension, 219, 225, 232
effective hole diameter for;233
at thre'dd root, 219
use of s2/4g, 233
N E W L I N , D. E., 427
NEWMARK, N. M., 185
hlt, 456, 48 1 , 49 1
Llr, 263, 27 1 , 275, 307
Reaction distance for beams, 157
bracing for, 158
Reduced eccentricity of fasteners.
stresses: bendin
AISC specifications fo
S1 conversion factors. 39
Sidesway, control of, 337
Slot weld, 414, 419
AISC specifications for, 4
Spacing of bolts, 363
Stepped columns, 329
K factor for, 33 1
Stiffeners, bearing, 158, 37
493
column flange, 427
web. 4'76
536
INDEX
$.%
$& ', @j "able
($.,&"!;I
&,
d
,
*'
Ult~rnatestrength. 7
of values, 6
Unhraced length, I75
LL'.151, 323
L", 151
Unlt welght of steel, 7
. VALLAERT,
J . , 368