Structural Steel Design Bowles

STRUCT

STEEL

DESIGN

Joseph

E. Bowles

Professor of Ciod Engineering

McGraw-Hill Bmk Company

New York St. Louis San Francisco Auckland Bog& Hambur

Johannesburg London Madrid Mexico hiontreal New& Panama Paris Sao Paulo Singapore Sydney Tokyo Toron

S k;lrrLTURAL STEEL DE§IGN

Cv-.: - r;ht 0 1980 by McGrdw-H111, Inc. All nghts reserved

Pr: t.2 m the Unlted States of Amenca. No part of this publ~catlon n t y ~e reproduced, stored In a retneval system, or transm~tted, In any I'm or by any means, electronic, mechanical, photocopy~ng, recording, or

9

2. se, w~thout the pnor wntten permissron of the publ~sher.

-

t

I

&.@! ' 6 7 8 9 0 D O D O 8 9 8 7 6 5 4 3 2 1 0

Thi, '..do'# was set in Times Roman by Science Typographers, Inc. Tfie ..iilurs were Julienne V. Brown and Madelaine Eichberg; thd cover was designed by Anne Canevari Green;

the ?rs.,duction supenisor was Dominick Petrellese.

'The ?..:wings were done by J & R Services, Inc.

K. '8. Donnelley & Sons Company was printer and binder.

'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,

S t r ~ LI ~. 3. Structures, Theory of. I. T ~ t l e 'TA6!' "478 624'.1821 79-18155 1SBW r 37-006765-1

CONTENTS

Chapter 1

1-1 1-2 1-3 1-4 1-5 1-6 1-7 1-8 1-9 1-10 1-1 1 1-12 1-13 1-14 1-15

Preface

General Design

Considerations

Types of Structures Design Procedures

Steel a s a Structural M a t e n a i 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 a n d Electronic Calculators Structural Engineering Computations in SI

Chapter 2

Elements

of

Frame, Truss, and Bridge

Design

2-1 M e t h o d s of A n a l y s ~ s 2-2 Beam Analysis 2-3 Determinate Structures 2-4 Truss Analysis 2-5 h a d F r a m e A n a l y s ~ s 2-6 B n d g e Analysls

2-7 The Computer Program Furnished in the Appendlx 2-8 The P Matrix

2-9 Load Conditions

2- I0 Checking Computer Output 2- 1 1 Design Examples

Chapter

3 Elastic, Plastic, and Buckling Behavior

of Structural Steel

Chapter

4

Introduction

Elastic versus Plastic Design Theory Safety Factors in Elastic and Plastic Design Elastic versus Plastic Design Deflections Length of Plastic Hinge

Elastic versus Plastic Design Load Resistance Factor Design Local Buckling of Plates Post-Buckling Strength of Plates

Design of Beams for Bending

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

4-8 Biaxial Bending and Bending on Unsymmetrical Sections

4-9 Shear Center of Open Sections

4- 10 Design of Laterally Unsupported Beams 4- 1 1 Beams with Nonparallel Flanges

4-12 Design of Bridge Stringers and Floor Beams 4- 13 Composite Beams

4-14 _{Beam Design Using Load Resistance Factor Design} (LRFD)

Chapter

5 Design of Tension Members

5- 1 Types of Tenslon Members 5-2 Allowable Tension Stresses 5-3 General Deslgn Charactenstlcs

5-4 _{Stresses Due to h a 1 Load on the Net Sectlon} 5-5 Des~gn of AISC Tenslon Rods

5-6 Net Sectlons

5-7 Deslgn of Tenslon Members 5-8 Design of Bndge Tenslon Members 5-9 Cable Deslgn

5-10 _{Deslgn of Tenslon Members Usmg LliFD}

Axially Loaded Columns and

Struts

~ntroduction

The Euler Column Formula Columns a l t h End Condltlons Allowable Stresses in Steel Columns Deslgn of Bu~lt-up Compression Members Column Base Plates

Lateral Brac~ng of Columns

Column and Strut Design U s ~ n g LRFD

Beam-Column Design

Introduction

General Considerations of Axial Load with Bending Effective Lengths of Columns in Building Frames

-

Developing the Beam-Column Design Formulas Determination of the Interaction Reduction Coefficient C,,,

AASHTO and AREA Beam-Column Design Formulas Beam-Column Design Using Interaction Equations Stepped Columns and Columns with Intermediate Axial

Load

Control of Sidesway >

Beam-Column Design Using L R F D

Bolted and Riveted Connections

Introduction

Rivets and Riveted Connections High-Strength Bolts

Factors Affecting Joint Design

Rivets and Bolts Subjected to Eccentric Loading Beam Framing Connections

Fasteners Subjected to Tension

Connections Subjected to Combined Shear and Tension Moment (Type 1) Connections

Load Resistance Factor Design (LRFD) for Connections

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

Chapter

10

10- 1 10-2 10-3

Plate Girders

General Loads

Proportioning Flanges and Webs of Girders and Built-up Sections

Partial-Length Cover Plates

General Proportions of Plate Girders Plate Girder Design Theory-AISC

Plate Girder Design Theory-AASHTO and AREA

Appendix Selected Computer Programs

A-1 Frame Analysis Program

A-2 Load Matrix Gcnerator for AASI-IT0 Truck Loading on a Truss Bridge

A-3 Load Matrix Generator for AREA Cooper's E-80

Loading on a Truss Bridge

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 computer-generated rolled section data tables as a supplement to the text. These tables are in ge agreement with the AISC and ASTM A-6 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

not-so-routine) structural members in either fps or SI units and by any

at least in American practice. Specialized problems are not generally addressed 1

in a classroom environment, and for these (as well as for design office practice and other nonacademic work) the reader should obtain a copy of the latest specifications from the appropriate agency.

I use the digital computer as a design aid in a somewhat interactive mode (via batch processing) for the design portion of the steel design course. I have found that the use of the computer in the steel design course is one of the better academic experiences for students, because it helps them rapidly gain experience in structural behavior. This may be by acci'dent (from mispunching data on the modulus of elasticity, cross-sectional area, or moment of inertia of a member) or by iteration of a design problem in which member sizes are changed as indicated by the computer output. In either case, students readily see the effects of member section properties on structural behavior. Using the computer programs pe:;~ii'fs' thisb.with only a modest amount of work on the part of the student-

no program writing.

Several computer programs are listed in the Appendix to the text for those o are not already using the computer as a design aid. These programs are relatively simple, but efficient, and can easily be punched on cards for use on a local computer system. The band matrix reduction method is used so that computer cpu requirements are minimal. I can furnish these programs on tape at the cost of tape, reproduction, and mailing.for anyone who is using the text in the classroom.

I have not attempted to cite, or promote the use of, desktop programmable calculators for simple tasks such as beam or column designs because of the variety of devices available (e.g., HP, TI, Sharp, Casio, etc.), each requiring a different programming method, and because of continuing rapid change in the state of the art. Listing of the multiple programs necessary for use of the various calculators would take too much text space, at the expense of more important topics.

The text attempts to strike a balance between theory and "how to." The topical treatment is not so exhaustive as to obscure the fundamentals but is of sufficient depth that the reader is aware of the source of the design equations in the various specifications. A number of the equations are partially to completely

! derived so that the reader can be aware of the limitations. A reasonably detailed

'

explanation is given of the basic design problems; and the illustrative examples are essentially step by step. With this format students should be able to cope with the more complex design problems on the professional level, and to obtain design solutions for the assigned home problems.

Appropriate references are cited directly in the text for topics for which coverage is limited but which are sufficiently important that the reader may wish to study the subject in greater depth. The inclusion of references will generally be of more use to those in professional practice than to student users. My expcr-icnce in teaching steel design for a number of years is that most students in the ,first design course are primarily interested in learning how to design the various types of structural members they will be assigned for home or laboratory work. At this point in their professional development they are not overly

PXEFA

terested in the theoretical considerations and the extensive laboratory work researchers and theoreticians that has produced the current design equations.

The complexity of semitheoretical and empirical design equations, couple with the nature of structural design and its intimate association with desim specifications and codes, makes i t necessary to take a strong '-how to" approa in teaching steel design. I t is essential to present the user with a set hypothetical (or real) data and by illustration produce a design. Students presumed to have a sufficient background in the basic engineering and sequence to appreciate what has been illustrated and are taught how to dup the steps with a similar problem to gain confidence, and, based on the illus tive problems, to extrapolate to a problem where the desim parameters considerably different, with a minimum of super-vision.

Fabrication and practical considerations are introduced in the exam problems as appropriate. Fastener spacing, edge distances, erection clearanc standard gage distances, thread runout. and maintenance are considered various sections. This should give the user an appreciation of fabricati problems and other practical considerations. In conjunction with this, the te has a large number of photographs, supplemented with line drawings of struc- tural elements and connections, which should be of particular aid to the no The reader should supplement these illustrations by observing steel frames un

construction. The photographs were all taken especially for this text, to disp individual structural features as appropriate to the development of the disc

, sion.

Plastic design is introduced briefly in Chapter 3 together with the basics o plate theory. This is done so that the design equations with origins in

design or.plate theory can be efficiently referenced back to Chapter 3, there saving text space. Plastic design methods are not emphasized, for two basi

reasons: there is not enough time in a first course to adequately treat the subje and elastic design seems to be preferred in professional practice.

I have deviated from the current textbook, trend to reflect the fonna ibcorporated in some of the steel texts published in the 1950s. This ibcludes the use of simple illustrative examples where the design data are stated as well as more realistic design examples. These examples are anaI Chapter 2 using the computer, and selected members are subsequently designed in the later chapters. The use of simple examples gives the reader a quick grasp of the general objectives of the discussion. More detailed design examples are used to generate a sense of realism and to clearly indicate that steel d e s i g is not just a matter of manipulating numbers. The examples are accompanied with a

reasonable amount of discussion of the analysis provided,

Within the framework of classroom time restraints, a steel design course should be as realistic as possible. For this reason the user is encouraged to c a n y

iiny structural design problems assigned in Chapter 2 through succeeding chapters, redesigning members as necessary and recycling the problem one or more times for member sizing before the connections are desiqed in Chapters 8

and 9. A false sense of security regarding the actual complexity of structural design, and even how the design loads are finally arrived at, can be developed if

fl

PttZFACE

$~$he >,ser is 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'lc s;udent and more grading effort on the part of the instructor. This extra

, i ' l

~".ff~:t can be offset somewhat by assigning fewer total problems, but including

t~i;;s In which loads are glven, to bu~ld confidence, and some with design prc

i

!i31:is,

to

bulld des~gn

sk~ll.

7 iie following text sequence might be appropriate in the semester system:

L 1

, dmester hour:, Rapid coverage of Chapters 1 and 3, with Chapter 2 assigned for reading. Reasonable coverage of Chapters 4

to 10, Probably two wceks each on Chapters 4, 7, and 10.

I semester hours Rapid coverage of Chapters 1 and 3. Two weeks on

Chapters 2, 4, 7, and 10, followed by actual design of a building frame and highway bndge truss, or industrial building, based on the analysls in Chapter 2. One struc- ture should be done In fps, the other in SI. A design notebook should be kept, showing computations and computer input/output. It is also suggested that this work be done in groups, each wlth no more than four students.

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-

jgy, Bradley University, who provided me with released teacbng time.

I would'also like to thank Mr. Andrew Lally and Mr. Frank Stockwell, Jr.,

of AISC, who provided me with a prelimmary copy of the new AISC specifica- and took the time to go over the major changes with me. Mr. Lally also ;..-ovided useful ~nformation on maklng the SI conversions. Mr. Robert Lorenz

sf the Chicago Reg~onal Offlce, AISC, was also helpful in providing me with .ast-minute corrections to the preliminary specification changes.

Both Bethlehem and US Steel corporations were most helpful ln providing copres of their new steel section profiles, nearly a year in advance of their becomlng official. This allowed work to proceed early on computer generation

,.; the Structural Steel Deslgn Data Manual tables. Particular appreciation is due

to Mr. Roland Graham of US Steel, who carefully revlewed selected portions of he manuscript and the entire steel data manual and made some very useful sl:ggestions.

Grateful acknowledgment is also made of the very considerable contribu- tions of Dr. Eugene Chesson, Civil Engineering Department, University of Delaware, who carefully reviewed both the preliminary and flnal text manuscripts. Thanks are due Dr. T. V. Galambos, Civil Engineering Depart- ment, Washington Univers~ty, St. Louis, who revlewed the load resistance factor (

design material. a

Fi ,?:re 1-1 The Eads bndge across the Mississipp~ &ver at St. Louis, Missouri. This rallroad and _'.way bndge completed in 1874 represents one of the first uses of steel (and high-strength F, = 50 : ? S ksl steel) in the United States for a major structure. The 192-m (630-ft) hlgh St. Louis

' I t teway" arch, wlth an extenor s k ~ n of stalnless steel, can be seen In the background.

-1

TYPES OF STRUCTURES

e structural engineer wlll be concerned wlth the design of a v a n structures including, but not necessarily I ~ m ~ t e d to, the follo\wng:

Bridges: for railroads, highways, and pedestrians.

Buildings: including rigid framed, simple connected frames, load-beari cable-stayed, and cantilevered. Numerous lateral bracing schemes, incI trussed, staggered trussed, and rigid central core, may be considered or

Buildings may be further classified as to occupancy o r height as

industrial, mill, high-rise, and so on.

Other structures: including power transmission towers. towers for radar an

installations, telephone relay towers. water supply facilities, a n d trans tion terminal facilities, including railroad, trucking. aviation, and mari In addit~on to the foregoing structures, the structural e n p e e r is engaged in the design of ships, a~rplanes, parts of vanous machines a n d mechanical equipment, automob~les, and dams and other hydraulic struc including water supply and waste d~sposal.

This text will focus pnmanly on structural d e s ~ g n usmg metal, an

particular standard structural shapes as produced directly by the several producers or in a few cases use of members that are built u p from steel p

and shapes and fabncated either by the steel producers or in local st

GENERAL DESIGN COP.SIDER%TI 4,

1-2 DESIGN PROCEDURES compression before fallure Other Important cons~deratl

e use of steel include widespread ava~labiiit:, ~ n d durability, particularly wt

Structural design lnvolves application of engineering judgment to produce a odest amount of weather protection

structural system that will adequately satlsfy the client/owner's needs. Next, this Steel 1s produced by refining iron ore and scrap metals together system IS incorporated ~ n t o a mathematical model to obtain the member forces. g agents, coke (for carbon), and oxygen in hlgh-tempe

Since the mathematical model never accurately represents the real structure, aces to produce large masses of lron called "pigs" or "pig iron." engineenng judgment is agaln required to assess the validity of the analysis so is further refined to remove excess carbon and other lm~urities and that adequate a ~ ~ o w a n c e can be made for uncertainties in both deformations and r metals, such as copper, nickel, chromum, man

statlcs. _{molybdenum, phosphorus, sillcon, sulfur, titanlum, columbium and vanadi}

Based on material properties, structural function, environmental considera- to produce the desired strength, ductility, welding, and corrosion-resis ttons and esthetics, geometrical modlf~catlons in the analysls model are made characteristics

and the solutlon process Iterated untll a solutlon is obta~ned that produces a _{The steel lngots obtained from this process are pasaed between two roU} satisfactory balance among material selection, economics, client desires/flnan- pposlte directions to produce a semiflnis cia1 ability, and various architectural cons~derat~ons. Seldom, except possibly in called either a slab, bloom, or billet, depen the most elementary structure, will a unique solution be obtained-unique in the nal area. From thls point the product 1s sent to other 10

sensc that two structural engineering firms would obtain exactly the same ills to produce the final sect~on geometry, Including structural shapes as

8

2 @Qlutlon. ,# 'r _{plates, and pipes. The process of rolling, in addi}

In structural englneerlng practice the designer will have available for p red shape, tends to lmprove the m a t e d properties of to

k ,

w e use numerous structural materials, including steel, concrete, wood, malleability. From these rollln:: mils the structural sh ' posslbly plastics and/or other metals, such as alumlnu _{are shipped to steel fabricators or warehouses on order.}

'occupancy/use, type of structure, location, or other design parameter The steel fabricator works from the englneerlng or architectural dra

"

dlctait: the structural material. In thls text we will assume that the design produce shop detail drawings from which the requlred dlmenslons are proceeded to the point where the structural form has been decided (i.e., as trus to shear, saw, or gas-cut the shapes to sue and to accurately locate g~rder, frame, dome, etc.) and the several possible alternative structural materia drilling or punching. The origlnal draw~ngs also indicate the necessary have all been eliminated in favor of using steel. We will then proceed with an finishes to cuts. In many cases the parts are assembled in the shop to det .?ddA~lonal analysis required, and make the member selection and connecti if a proper f i t has been obtained. The pieces are marked for ease of

deslg~l appropriate to the topic being studied. _{identification and shipped rn pleces or subassemblies to the ~ o b site for erec} Textbook space and classroom t ~ m e limitations will of necessity reduce

the bare essentials the complexity of the design presentations. The reader shou _{general contractor.}

be aware that real design 1s considerably more comp _{Some of the most important structural properties of steel are the follo} than the simplifications presented in the following chapters.

%) Safety as a design concern takes precedence over all other design consid 1. ~ ~of e[astlcrp, d E The typlcal rdnge for all steels (relatively l n d e ~ e ~ l ~ ~

t q8, The "safety" of any structure, of course, de

?#&np. Since the structure is always loaded after it is built and not always i The value for deslg n as 29 000 ksl or 200 000 MPa-

t ~ e ~ k o d e or manner used in the design, the selection of design loads is 2. Shear modulus, G. pioblem in statlstlcs and probability. This part of t

G = E

rather subjective and produce extremely divergent designs had not bu

codes been developed (and in some form or another, almost universally 2(1 + P )

which place minimum required/suggested bounds f where p = Poisson's ratlo taken as 0.3 for steel. Using ,u = 0.3 9ves G =

is an important factor. 11 000 ksl or 77 000 MPa.

/ /

3. coefficlenf of expnnsron, cu T h e coefficient of expansion may be as

1-3

STEEL

A S A

STRUCTURAL MATERIAL

a = 1 1 25

x

per " C

AL = a(T, -

T,)

L (ft or m depending on length L ) S i ~ e l is one of the most important structural matenals. Properties of particular

*-ce In structural usage are h ~ g h strength, co BP%:t:rlal, and ductil~ty. Ducabiy 1s the ability

m m d ; z

y j 9 Q

0 0 o m - w

3 $ 2 & Q

.

.

m .L

"

u Q) p

"

%

3

.L V1

-

B

E

V1 M I 3 aJ Z;

GENERAL DESIGN CONSID

in degrees Celsius. To con eit to Celsius, use

C = ; ( F - 32)

and ultimate strength. Table 1- 1 gives the yield poin

es of steel of interest to the structural designer that are

er properties o/ some interest. These properties include the mass de

(1 t = 1OOO kg); or in terms o

76.975 kN/m3. The specif conversion of fps units its of kN/m and kg/m is accomplished as follows.

Given: lb/ft and required to convert to:

m

Note that lb mass and lb weight or force have been used interc ably in the fps system because the acceleration-producing 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

re rolled into plates of varying round, square, and rect Most of the rolling is done on hot steel, with the p

steel." Sometimes the thinner plates are further rolled ' steel products. Several of following sections.

k?

m 0 N

; a a

-8

.- 9 9

4

g

3

g g m z

3 5 ) a ?

-

rd

*

0 o "0 9' + 0 m m + ? Y r :

2

.O m w - 0 0

Z

m

$ 2

5

2

8

$ 8 9 8

k;" 8 . I 2 m m o m - o Q W W t -

2

S:

, 8 2 3 , * P N N N m m . d 2 8

% Q % Z

a 2

.z

b a

<.g

_{a 8 % 3 $}

k

4-44

$i

nh

g

3 g 7 ;

"

-E

6 s

E

8 5

b S l RUCTURAL STEEL Ukbiut4

1 4 . 1 W Shapes

X,e most commonly used structural shape is the wide-flange or W shape. This is

a doubly symrnetncal (symmetrical about both the x and y axes) shape consist- ing i;f two rectangular-shaped 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. 1-1.

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.

& o r 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~able rate 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 dimen- sion (varying from theoretical or design values) can be produced. American Soclety for Testing and Materials (ASTM) specification A-6 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. 1-4.2

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.

--- -

i

---r

[

-.

C - to 330 MPa and refer to Figs 1-30 and 1-36. Similarly, A-44

'V shapes S \hapes C shape, L shape point of 345 MPa, will have a yield strength on the order of

hd

:;;gc Anierrian Stdndard Chdnnel kqiral leg angle bean1 (I-beatri)

guaranteed values converge.

i

"_,;id

U Reitdllgular

8

Square

I L-shdpe 7 slldpe

0

R o u n d s

C 1 1 ~ q u d l leg dtlgle Strut tirrel Tec

i l l ! Iron1 \V.stidpc I h r t

1

Plate

k + $ i s 1-2 ktmctural shapes as drrectly produced by Be steel producen. _{ed and designated A-272 (described ~n ASTM specification A-272). Sp}

on ASTM A-440 was wntten in 1959 for another h ~ g h strength steel

t--:,.j

L

Shapes _{1960 with application to weld~ng. All three of these steels have a yield}

i i . e , ~ shapes are either equal or unequal leg angles. All angles have parallel at is dependent on the thickness of the metal, as shown In Table 1-1. flal~ge faces. Angle leg dimens~ons can vary on the order of +- 1 mm in width.

An L6 x 6 X _{IS an equal leg angle wtth nominal dimens~ons of 6 m and a}

2thLE~ness of

$

In.

An L89 X 76 x 12.7 IS an unequal leg angle w ~ t h leg dimensions of 89 and

76 k:m, respectively, and a leg thickness of 12.7 mm (L3f

x

3 x +).

t

1-4.';

T Shapes

St: LC ~ r a l tees are structural members obtatned by splitting W (for WT), S (for

S"

.

.

M (for MT) shapes. Generally, the spltttlng is such to produce

W I .:1 one-half the area of the parent section, but offset splitting may be

d

.:

-r tze section is required. Published tables of T shapes are based slA.$~netncal splitting. No allowance is made for material loss from splitting p21 t .:t shape by sawing or flame cuttmg.

4 WT205 x 29.8 is a structural tee with a nominal depth of 205 mm

m: c*c of 29.8 kg/m and is obtained by splitting the W410 x 59.5 section (f R'IG x 40).

Exera1 rolled structural shapes are Illustrated in Fig 1-2. 1-3

STEEL STRENG??-I

A l l me1 design takes into consideration the yield strength of the mate Figure lJa Typical stress-stram curves for E;lgure 1-36 Erllargement of lmti structural steel. stram curve for two grades of

yleld strength of several grades of steel available for design is given in Table 1- _{Note that the plasbc r a g e} u yield strength is that minimum value guaranteed by the steel producers an

.- ...---

.

_{G E N E R U DESIGN CONSIDERATIO}

Since about 1964, specifications for several other high-strength (low-alloy) "C x 100

steels-hzve been- incorporated into ASTM specifications as A-572 and A-588. Table 1-1 shows that the steel covered by the A-572 specification covers several $yield strengths, termed grades, such as grades 42, 45, 50, 55, 60, and 65 for the corresponding guaranteed minimum yield stress in ksl. Generally, the yield {strengths of these newer steels are also thickness-dependent, as shown in the table under the heading "plates and bar thickness." The steel producers have designated the several W shapes into five groups, depending on flange thickness (and as shown in Tables 1-1, 1-2, V-1, and V-2)t, compatible with the steel grade. The designer merely has to check these tables to see if the shape is available in the raquired/deslred yleld-stress grade. For example, in the 450-MPa grade, only shapes In group 1 qualify from flange thickness. W18 shapes are available in group 1 only from 35 to 60 Ib/ft inclusive (the five smallest sections and with a maximum part thickness of 0.695 in).

Specification ASTM A-588 allows

F,

= 345 MPa for a high-strength low- @lo$ steeltwhich may be up to 100 rnm (4 in) thick. The steel covered in this @$hPecifl,rttion is primarily for welding and is corrosion-resistant.

in terms of cost/unit of mass, the A-36 steel is most economical. High

stre2g:h steels have principal application where the stresses are primarily 14 Effect of elevated temperatures on either y e l d or ult~mate tensile strength ucpresse

High-"irength steel beams may deflect excessively, owing to reduced f strength at room temperature of approximately 70°F

rnodul:.~. The high-strength steel columns may be less economical than steel tf the slenderness ratio ( K L / r ) is large. Hybrid girders that use

strength steel in the flanges, or built-up columns using high strength steels provide better solutions where member sizes are restricted. In a given cas necessary to perform an economic and availability analysis to determi suitability of using high-strength steel.

TEMPERATURE EFFECTS ON

STEEL

/

1-6.1 High-Temperature Effects

Steel is not a flammable material; however, the strength is heavily temperature- dependent, as illustrated in Fig. 1-4. Both the yield and tensile strength at 1000°F is about 60 to 70 percent of that at room (about 70°F) temperature. The drop in strength 1s rather marked at higher temperatures, as shown on the figure, where the strength at 1600°F is only about 15 percent of that at room tempera-

ture. 1-6.2 IAV-Temperature Effects

Steel frames enclos~ng rnater~als that are flammable will require fire protec-

tion to control the ternperature of the metal for a sufficient time for the Brittle fracture is a failure often associated with low temperatures. Essentially, occupants to seek safely or for the fire to either consume the flammables or be brittle fracture is failure that takes place without material yielding. The

extir~guished before the building collapses. In many cases the building does not stress-strain curves of Fig. 1-3 indicate that m the usual failure of a tensile specimen, considerable elongation takes place. As a matter of fact, a minimum

t

See footnote a of Table 1-1 In J. E. Bowles, Structura/ Steel Desrgn Data Manual, McGraw-Hd, _{percent elongation is specified for steel in the ASTM standard tensile test.}

Fireproofing materials

I

Unit w e ~ g h t

Pcf k ~ / r n ~

Cinder concrete l 1 0 17.3

Gypsum board 30-40 4.7-6.3

plaster, cclnetit and sand 1 0 0 15.7 Expanded shale concrete

Vertnicul~te Perlite

I I , . ,

Usually use 35* nim o f flreprotection for 2-h fire rating; obtain specific thickness values from either tests o r from the producers o f gypsuln, perlite, etc.

rigure 1-5 Methads of producing fireproofing of structural steel members. ( a ) Sprayed fiber. (6) Lath and plaster. (c) Lightweight concrete (formed). ( d ) Gypsum board-use boards to build 'hick;us. ( e ) Corner detail of sprayed on fireproofing. Thickness is built in several sprayings.

(fl

16 STRUCTURAL STEEL DESIGN GENERXL. DESIGN CONSIDERATI

any abrupt change in cross-sectional area-to inhibit lateral contraction tructural design. On the one hand, it sometimes takes

tension situation. ials and methods; on the other hand, th

fast." If the local .building code is care minimum design requirements met, or exceeded, and a catastro brittle fracture. This may initiate as a crack that propagates to a member f of exists that good engineering practice has been followed.

ng codes are supposed to reflect that part of the structural unique for that locale, such as temperatures, earthquakes, ntity, frost depth, and average wind velocities.

list gives several design codes and/or specifications occurring is of little aid in settling the resulting damage claims that are sure may have occasion to use:

follow. Brittle fracture can be controlled in several ways:

Code, published by, and available from, the

1. Detail niembers and their connections to minimize stress concentrations. _{iation, 85 John Street, New York, N.Y. 10038.} 2. Specify the fabrication and assembly sequence to minimize residual ten _{Building Code by International Conference of Building}

stresses. _{Workman Mill Road, Whittier, California 90601.}

3. Use steels that are especially alloyed for low-temperature environments _{lding Code, Building Officials and Code Administrators Inte} East 60th Street, Chicago, Illinois 60637 (formerly BOCA).

itute (ANSI), Minimum Design

ngs and Other Strucrures, ANSI 58-1, 1430 Broadway, New Yo tures are encountered. . .

5 , If possible, machine (or grind) the notch into a sm _{titGte of Steel Construction (Specifications), Steel}

6. P-educe the rate of tensile strain application. _{th ed. (1979), 101 Park Avenue, New Y0rk;N.Y. 1001} lding Society (AWS), Structural Welding Code, 2501 iami, Florida 33125.

1-7 STRUCTURAL DESIGN CODES

.

and Steel Institute (AISI), 1000 Sixteenth Street,

.

Publishes various specifications for using iron and ste Lo.~,d.l building departments almost always require structural de

Association of State Highway and Transportation Of l.ITO), Specifications for Highway Bridges, 341 National Press Bu

Railway Engineering Association (AREA), Speci/ica The various state departments of transportation _{ay Bridges, 59 East Van Buren Street, Chicago, Illinois 6060} generally use the specifications put forth by AASHTO,

several railroads generally use specifications put forth by

The structural designer doing highway or railroad work national and city codes use specification standards as applic closely the design specifications of these publications, partic zations, such as AWS, AISC, and AIS

government is involved with any of the financing. cies for the other construction materials.

nd AISC, as well as the steel producer tables of structural shape design data as well as data on other steel

, wire, and bolts. Certain of these data together the owner/client may require a more stringent design than the building co AISC, AASHTO, arid AREA, have been pre- c,piteria. Only in rare cases can the designer get a variance from the local tnrctzlrnl Steel Design Data (SSDD) man governing body to deviate in a less conservative manner from the code. Vari- by McGraw-Hill Book Company and used as a supplement with

18 STRUCTURAL ST GENLRAL DESIGN CONS1

carry the dead 1 fps: R = 0.0008 x area (when area

> 150 ft2)

SI: R = 0.0086 x area (area

>

11.2 m2) include:

Ceiling materials, including duct work for environmental co (some codes limit R

<

0.40 for horizontal membe supplies.

-xterior walls supported by the frame, including windows, doors, and ba here R = reduction factor used as ( 1 - R ) x L,,,,,

Interior walls Qat are permanently placed. D = dead load, psf or kPa (kllonewtons/m)

'iechanical eqmpment (heating, air conditioning, ventilati L = live load, psf or kPa, but L 1s limlted to not over 100 psf or

' (such as elevators, including cage, cables, motors). generally, values larger than this are not reduced r .reproofing.

Beams, girders, and columns, including the footings making

lic assembly (such as auditoriums), garages, and roofs. frame.

I From this list it is evident that any part of the building w

nstalled contributes to the total dead load. Dead loads can

Example 1-1 A port~on of an office (multistory) floor plan is shown El-1. The floor is 4-111 concrete on a metal deck over steel bar joists. W the reduced live load for the floor beams and for an exterior column floors down from the top floor?

prescribed by building codes based on occupancy and 1 anow, and earthquake loads are considered. In addition to !oads include:

People, as in auditoriums, assembly halls, and classrooms. Movable room partitions.

Office equipment and production machines if they are m Warehouse products.

Furniture.

Building code values of live loads tend to be based on

SOLUTION Estimate the dead load on the contributory (centered ber) floor area as:

Concrete floor and finish: 4 x 144/ 12 = 48 psf computational convenience and because the actual buildin

:lot known. Ceiling, metal deck, steel bar joists = 12 psf

GENERAL DESIGN COKSID

From Table IV-4 of SSDD, the live load = 5.00 kPa. The r r for a grder based on a contnbutory area as shown is El-1, 18 X 22 (area ABCD): _{R = 0.0086[(12}

₊

_{8)/2 X 91 = 0.774}

_>

_{0.40 (and also 0.}

R = 0.0008(18 X 22) = 0.32

<

0.60

D

+

L 3.703

+

5.00 + 6o + 8o

-

0.40

<

0.60 R = --- = R = --- =

-

-

_{4.33 L 4.33(5.00)} = 0.402

< 0.60

4.33 L 4.33(80)

Use the smaller value of R computed, 0.32. The reduced live loa Since the problem statement limlts live-load reduction to not mor 0.40, the reduced l~ve load is

L' = (1

-

0.32)80 = 54.4 say 55 psf

Compute R for the column; the contributory area is centered on the column L ' = ( 1 - 0.4) x 500 = 300kPa of 9 X 22 (AEFD); but for the accumulation of three stories, we have

R = 0.0008(3 x 9 x 22) = 0.475

< 0.60

O.K. R =

---

60 + 80

-

-

0.40

<

0.60

4.33(80) d loads have been extensively studied in recent years, particularly for larger Using the smaller value of R, 0.40, the reduced live load on -rise structures. Generally, for tall structures ~wnd-tunnel studies should bz

L' = (1

-

0.4) x 80 = 48 psf or smaller regular-shapsd

We note that O.$0 is the maximum R for the column and is 30 m, the wind pressure

uppermost floor level. ory to use. The Nation

Example 1-2 A meeting/banquet room in a hotel has 22 X 27 m. The floor is 125 mm of concrete with a tile surf

15 0.75

d this pressure times

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 live-load reduction factor. First, estimate the dead load using Table IV-3 of S

weight of concrete = 23.5 kN/m3. o allowance is commonly made for the shielding effect of adjacent structure

Weight of concrete: 0.125

x

23.5 r from ground cover.

ge ground level at

th

The wind pressure is commonly computed between floor levels and prorate he adjacent floors using simple beam theory if the vertical distance compare

The sever31 wind values are shown in Fig. El-36. The data display is convenient for computer programming for frame stresses using the com- ~uter-gr,ogram discussed in Chap. 2.

_{/ / /}

Wind pressures can be approximately computed as

where V is mi/h or km/h. This equation is readily derived as q =

f

mu2, where

the mass density of air is approximately 0.00238 lb

-

s2/ft4.

Since wind is a transient load, the building codes usually allow a one-third increase in the allowable design stresses with wind included as a part of the load condition as long as the required section is not less than required in the load condition of dead

+

live loads alone. For example, if a stress of 20 ksi is allowed, then with the wind load condition a stress of 20 X 1.33 = 26.6 ksi could be used.

1-8.2 Snow Loads

Snow loads are live loads acting on roofs. Snow and any other live loads are taken with respect to the horizontal projection of the roof, as illustrated in Fig.

1-8. Figure 1-9 is a map illustrating snow loads that may be used in the absence of spzcific load building code requirements. Even in areas where snow loads are minir~~al, a minimum roof live load should be.used. The NBC stipulates the larger of the snow load or 20 psf or 1.0 kPa. Since 10 in of snow approximates 1 in of water, a 20-psf snow load corresponds to a roof snow depth of nearly 40 in -easily obtained where snow drifting occurs. When rain later falls on snow, however, the saturated snow weighs considerably more and the unit weight can approach that of water.

Snow and other live loads

l i l l l l l l l l l l l l l l l I I 1 1 1 1 1 1 1 1 1 1 '

Figure 1-8 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~iding is 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 live-load 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

Erection loads are not directly considered in biiildiiig c o d ~ s . 'rhese loads may control the design of certain members, particularly very high rise buildings, cantilevered bridges, or cable-supported structures. The engineei responsible for esy phase of the erection may be held legally responsible for damages or loss of life resulting frorn a structural failure during erection. Most structural failures (at least that are reported) tend to occur during erection rather than later. Erection methods andequipment tend to vary from project to project; thus ii is ' '

not practical in textb~oks to do more than point out this very important design area. The engineer musf+d$termine what equipment will be used, where it is placed, loads to be lifted; quantities of material, and the storage locations,so tLat the affected individual steel structural members can be checked for ade- quacy using princip

chapters on design.

1-9 HIGHWAY AND RAILROAD BRIDGE

LOADS

'The American Association of Highway and Transportation Officials (AASHTO stacdard truck loadi

General lucarion of rndxrrnurn momsnr

is oLtained with eith

ti,c i l S truck and span lengths, shears are as follows: , ), ,,, ,*,. ,. .. .

-. - 3 1 kips IJJ kN

24 k ~ p s 108 k N

Bridge span I6 kips 7 3 kS

---

33.8 to 145.6 ft M,., = :[(0.9~ + 4.206)(0.5L

+

2.33)

-

1 l 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).

GENERAL DESIGN COWIDEIW Moments, shears, and floor-beam reactions for Cooper's E-81) loa

al axle lodd nd of train:

sf

S'RUCTURAL STEEL DESIGN ~ a b ~ d 1-2 (Continued) . , 479.6 282.0 128.1 . ! 23 712.0~ 17 990.0 522.0 306.8 ,. 3 35 1 18.0b 27 154.0 626.4 367.3 729.3 426.4 197.9 1225.3 .-. ' 1 48 800.0~ 38 246.0

-

.. .

-

.U1 values shown are for one rail (one-half track load). Axle loads shown in diagram. Obtain _{total uniform loading, however, must not be less than the following:}

;,

.

. :s for other E loads by proportion. ,,,

I. .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.

f ::-cd on the locomotive weight, the Cooper load is designated as E-40, E-50, _{2. AREA wind requirements:}

E.50, E-75, E-80, or E-110 and is directly proportional (i.e., E-60 = $XE-80). The cl:rrent AREA design criteria are based on the E-80 (sometimes E-110) loading

s.:dwn in Fig. 1-1 1. Table 1-2 can be used to obtain the bending moments and Pressure, force/area

s.+c.;;rs at selected locations for girder bridges, with values given for a single rail Unloaded span Loaded span

fbading (based on one-half the axle load shown in Fig. 1-1 I). fps, psf SI, kPa fps. psf SI, kPa

Where multitracks or road lanes are carried by the bridge, the live load is as ici!ows:

-

Percent of live load

Lane or track AASHTO AREA

2 100 100

3 90 2 X 1 0 0 + 1 X 5 0

4 75 2 ~ 1 0 0 + 1 X 5 0 + 1 X 2 5

XI.xe than 4 75 As specified by designer

Other bridge loadings that must be considered include impact, wind, and

longitudinal forces. Impact and longitudinal forces allow for dynamic effects AREA: 0.15 X live load (without impact). 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 Other loadings that may require consideration include differential tempera- >elf-explanatory and in the case,of a loaded railroad bridge, the wind against the tures between top and bottom flanges or chords, ice and snow loads, possibIe train may be'a substantial load. overloads, and for continuous bridges, support (pier) settlements.

. - - -- .- - . llr, ~ L O I U I Y G E N E W . DESIGN CONSIDERATIONS

<..*,,., 6. ..4 ..'i .... , -..,

-

Railroad bridges make a distinction between those bridges which consi nt; must be divided by 100 to use in de een centers of single or groups of longitudinal s which frame into transverse floor beams or girders, or between trusses or girders, ft or m

= length between transverse floor beams or between supports as app cable, ft or m

beneath the ballast. _{e 90 percent of I, computed above for ballasted-deck bridges.}

1-10

IMPACT LOADS

Example 1-4 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

impact load as _{russ, depending on their lengths. For the stringers the impact factor is}

l 5 ₌

0.330

>

0.30 therefore. use

1/

= 0.30

//

Item ilroad bridge consists in two trusses spa

sses are made up of seven panels at 27.60 ft/

Elevator loads 1.00 at is the impact factor?

Macbnery and other moving loads > 0.25

r,unoN Since L > 80 ft,

L = 7(27.60) = 193.2

600 = 25.6 percent 193.2 - 30

The AASHTO impact requirement is Example 1-6 What is the impact for the floor beams of the AREA truss of

Example 1-5? Floor beams are transverse members connecting the two

f PS SI trusses at panel points.

I, =.

-2

< 0.30 I,

-

-

l 5 i 0.30 SOLUTION S = 27.6 ft, L = 17 f t

< 80.

I,+ 125 - L + 3 8 -

-

_3(17)*

+

40

- -

= 43.1 percent 1600

where L is the length of span or portion of span that is loaded, in ft or m. The AREA impact specifications depend on the rolling equipment. For diesel and electric locomotives and tenders:

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 earth- citation based on building geometry, and apply

Figure 1-12 Earthquake zone map for the Umted States. (After Unrform Bu~ldng Code, 19

Example 1-7 A 10-story apartment building with basement is as shown

Fig. El-7a. The exterior is insulated curtain walls and thennopane windo with an estimated weight of 15 psf. The interior walls are generally stud partitions plastered on both sides with insulation between apartments. Use 4-in concrete floors (tiled or carpeted) on corrugated metal pan supports carried by open web steel bar joists. The building site is in Memphis, Tennessee. Estimate the earthquake force and corresponding story load. SOLUTION Estimate the roof and floor dead loads as follows:

...

Alloaf.. . .

Wood sheathing 5-ply felt roofing Ceiling and bar joists

= 3 psf = 7 psf = I1 psf GENERAL DESIGN roof Elevation

. . . -.

.,

.

G E N E W DESIGN CONSID

,,.,.'.b..* " L 9

Any floor:

Partitions in 40 X 30 apartment at 8-ft height and

cross-walls at 20 psf , I 4 .

.

gives: (40 x 2

+

30 x 2)(8)(20)/(40

x

30) / 4 Floor:

-

(144) (concrete) 12 Ceiling (estimated) Bar joists and metal pan Exterior wall at 10-ft height

(2

x

40

+

2

x

90)(15)(10)/(40

x

90)

Total = 93.5 psf Total floor weight = 0.0935(40 x 90) = 336.6 kips

These weights are illustrated in Fig. El-7a. For easi weights of 76 and 337 kips, respectively, for remaining wo Fig 1-12, the Z factor is 0.75. Take I = 1.00; take K = 0. Table 1-3. The total building weight = 76

+

lO(337) = 3446

The earthquake force in the E-W direction is computed D = 4 0 f t

T = : 0'05(100) = 0,7906 Figure El-76

Since the frame is of steel, the alternative computation for period is T = 0.1

x

number of stories = 0.1(10) = 1.0 s

The author will average the two values of T to obtain T = 0.895 s. _{Summing the horizontal floor loads and including the top value of 13.7} 1

C =

---

= 1 _{= 0.0705} 218.59 versus 218.6 kips as a check. These lateral floor loads

1 5 r ~ 15- further prorated to the several bays in the E-W direction for the fr Substitution of this accumulation of factors/weight into Eq. (1-4) analysis load condition(s), which includes earthquake forces.

F = 0.75(1.0)(0.80)(0.0705)(1.5)(3446) = 218.6 kips

The roof value is _{12 FATIGUE}

F,,, = 0.07TF = 0.07(0.895)(218.6) = 13.7 kips ( T

>

0.7 s)

The story loads are found using hn = distance grou failures which have been attributed t follows: Z W,hn = 337(90)

+

337(80)

+

337(60)

+

.

.

ue. Fatigue failure is a material fracture caused by a sufficiently Iarg

15 1 650 f t

.

kips ulsating stresses, or stress reversals. Th

Wn A, ture of the material at a location where

F = ( F - F )-= (218.6

-

13.7)

lop 2 W, hn oscopic in size) exists. A crack form

For tenth floor nding on stress level, rapidly or slowly (sometimes so slowly that the

F,, = 0.001351(337 X 90) = 40.98 kips (shown in Fig. El-7b) re) progresses to failure of the part

F, = 0.001351(337 X 80) = 36.42 kips _{Most metals tested under repeated or cyclic loadings display stress r}

' F8 = 0.00135 l(337 X 70) = 3 1.87 kips s qualitatively illustrated in Fig. I-

rly, these curves were commonly displayed as stress level versus cycIes.

F, = 0.001351(337 X 10) = 4.55 kips

4(8 STRUCTURAL STEEL DESIGN GENERAL DESIGN CONSIDERATIONS 41

to !@I 100 to 500 500 to 2OCO Over 2

Base metal with rolled surfaces T or R e p 60 (415) 36 (250) 24 (165) 24 (165)

plates and shapes with full or

of girder webs or flanges adja- cent to welded transverse stiff-

T o r Rev 32 (221) 19 (131) 13 (90) l o b ( Figure 1-13 Qualitative plot of stress range F,, versus number of cycles to failure. _{Base metal at end pr partial}

. .-",. , 3. ... '," ....- ,.

-

T or Rev 2 1 (145) 12.5 (86) 8 (55) anically fastened connections

can be defined as

T o r Rev 45 (310) 27.5 (190) 18 (124) 16

T o r R e v 27(186) 16(110) 10(69) 7(48

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 T or Rev 45 (310) 27.5 (1%) 18 (124) 16 (110

Welded Details," AISC Engineering Journal, No. 4, 1977). Typical stress range

values which may be used for all three specifications are given in Table 1-4. The T or Rev 45 (3 10) 27.5 (190) 18 (124) 16 (110

reader should consult the Structural Welding Code, Sec. 9-14, 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 1-4 is generally applicable

to buildings. Lesser values than shown are necessary for reduced sections, T or Rev 21 (145) 12.5 (86) 8 (55)

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. 1-7.2) 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 se 12 ksi or 83 MPa for girder webs.

. . - " . ,...

_ . . .

_

___

-________

._ _. --- -.--.---_I-

N = 10

x

365

x

2 0 = 7 3 OM

_,,

.A STRUCTURAL STEEL DESIGN

Example 1-8 A rolled section will undergo an expected 1 X 106 cycles of stress over the design period of the structure. The stress analysis gives

f,,

= P,,,/A = 16 ksi;

f,,

= - P,,/A =

-

12 ksi. The basic stress'for this member is Fa = 22 ksi (tension) and 16.5 ksi (compression). 'The structural configuration limits

F,

= 24 ksi in the base metal. Is the section satisfactory?

S l u l r ~ p l s b ~ y bent

SOLUTION

f , , = 1 6 - ( - 1 2 ) = 2 8 k s i > 2 4 k s i 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 a l ) c r c d C U I L I I I I I I i n a r i g ~ d i r ~ n ~ r .lnJ hcnt r ~ 1 k r

1-13

STEEL STRUCTURES

Slructures of steel include bridges, buildings, trans

sign supports, and even art objects. The primary focus of this asid buildings, since these are the most common projects invol

Buildings are commonly classified according

:,ti?ry buildings. Little use is made at present of s T ~ L I ~ S - o n soli~rnn

.: zpt in multistory apartments. .

'L-13.1 Industrial Buildings 1-14 Several bents used in steel building frames.

Iildustrial. buildirigs are commonly one- or two-story structures fnr industrial (such as manufacturing, storage, or retail/wholesal and institutional (including schools, hospitals,

Other structures may include gymnasiums, aren tr: rlsportation terminals (land, sea, and air). Thes frame, as illustrated in Fig. 1-14, or have a roo resting on load-bearing walls (see Fig. 2-4). The rnay"8e~"?lgid or pinned; may be a two or th t;uss-on-column system. The truss may be rig frames under construction are illustrated in Fig. 4-1.

A building frame is a three-dimensional skeleton b rigid in only one plane. Some buildings are rigid in both t but this: type of frame will not be considered

resulting' from considering only the principal frame termed a bent and may be one or more stories in iiiusirate; terms defined here and later). The ter vlhether'rigid, truss-on-column, rafters-on-colu to span between columns in the principal pla the third dimension is the bay spacing. Span

GENERAL DESIGN CONSIDEIUTIONS

h type, where traffic passes between the trusses. The deck-type trus preferred if clearance beneath the truss is not a factor, because pr

Many truss bridges combine both types (see Fig. 1-19) 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. 1-20 (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 high-strength

ACCURACY

OF

COMPUTATIONS AND

ELECTROMC

leng:;. uf the diagonal members. .

the 10-in 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

1-13.2 Bridges 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 i t 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