BEST

■
Am lUBLE
COPY
 DOCUMENT CONTROL CaTA • R i D
• **.«r r I

I * »t r ■ t < L i. A .>.•» 4« A '

state Utiiverslty of Hew York, At Biiffalo UnclaajIfleC

ih c«*ou»>

I n» T tlc

Desl^ of Tapered Meobere

4 rt %C «>•*>' w K NO T r « I 7 t / >■ tf ffp >rt fni-r>« tv

rinal Report Fel 1S£6 - Lee 1971

» A ^ T mO • h tt\t fA.Ttr, ri.Im *t ,

Georce C. Lee; M. L. Morrell; Robert L. Ketter

H •f»C«*L4Tt 7m. yormi. ho o» ^Acr» fh. HO 0» Mt » S

Dece.ler 1971 __ 52 _____ 74

IC'NTWA^.' ^^ant Si- *•. OH'ClSa TON'S ni« OHT SuMI«| StS»

IGy-63?S4
ft. C T so
None
Y-ROD9-03-03-S04
O T ». L s ML • C T SC . '*» Attv tfvf . ttimt ■
fht* tmptitt'

Hone
I . t * i - «• f . T lOS • • A * t v*» *, •

Lietributlcn of this docLLio;;t is unlir.ited

• ••NL » VIST AN» SO*LT •j c s SO r s VI k. ’ A M » AT’

I!c.val Facilities liiclr-i^erin,, Co:.;oand

Asst. Co-._ar.:lcr for r.ecearch U Dcvc3 o; '.•
;.ashtr. :toa. L. C. ________________
Thi; report covers all rajor aspects of t».c dcsijn of tapor^l fra;.es; frsu toUl
ira-ie atiali'sic to stress analysis of Inllvidiuil r.o. bers - to the dcvcl.j: or.t of
i.Tperci deoijn fov.-.ulrs. t;.2 prl’-ary' cbjootive of the report is to present the
iisnale behind :he develc; .cnt of the desi^;n fon.alas.
The dcsicpi for.nulas developed dcocrilc the ^pross bcliavlor of isolated r-c ices.
.1-“ nor-.dl ir.icrliinto as3"r.ptlcn fat :.cd:ers are adtqna;,:ly prrprrticr.vj r.^i.noo
buclLLir.s is jide. Ais .-> contained 'n fac report is r. brio'* dlscriptica of
w ---- il of the fia-.e ai;alysis and stress ara.ly.-is :;othsc;s applicable to l.'^psred
f*«- ■■•f,. Tho problen of effectivo lor.^th of a tapsred c:lvv-an ia rliid fjar.-.os is
al;.c Ir-atod. I'ho basic approach .’ssd follr.r ; f.ioorctioal soluti-.ns ar. first
c'..1.0J, thoa usia,-: these solutio.io - the A.J.3.C. prisratic i.cihcvc dcsi 71 fc:v.v.lor
ar-.; 1 -.liflel to effect the so._o s-lutic.is bj ’f-e lat.rrduc*.'ca of iippropr'a;c •;.ultl--
if factors vhich are dcp-r.lcnt only on f-.o taporlr.y jo.-;.otry. Jince these
r... : 1 .r-acc 00 Lr.ity vhaa there is no taper, it ::izt \o iccopoiii ci thao thi.o
Appii- *a oosur.es that the current A.I.3.C. c.Uovablc stress forT-ulas for y.Tls.oatic
t\r V... rt.... aicqv.ite.

-n fo '
V3 •' in.CL_-Lrif
!
 <

t •

DESIGN OF TAPERED MEMBERS

(A Summary Report Including Design Recommendations) !

By

G. C. Lee; M. L. Morrell; and R. L. Ketter

B
This research study hat been carrl-.n out as
part of au Investigation Jointly sponsored by the
Naval Facilities Engineering Command, American
Institute of Steel Construction, American Iron and
Steel Institute, and the Metal Building Manufacturer's
Association. Reproduction of this report in whole or
in part la permitted for any purpose of the United
States Government.

December 1971

Department of Civil Engineering

State University of New York at Buffalo

Ns Security Obj*ctisa
te open Publication,
(Orte» \ ‘ 1
this document has been approved
Office el the Chiel ef
for public rçleaae and sale;
Naval Ojentiee*
Its distribution is unlimited Department if Ike NlVf
 1
u ^
1, z'

TABLE 0^• U .'TCNT

1. IKTl-OiMTia;

2. rv.AKK AKD .*:tk5:‘:S A.NAL\M:J

2.1 Kranu’ '.nalj? Is
IlirlliT Pro;>o:.cJ IVHinds
The .Slopc-n.fU-clior.
Tilt I'inile EUr^ciu Met hod
Axi;il And llcxurnl Sii‘et>.ics
2.3 Torsiotial Slrcsr.es

3. siAr.Tf.TT^- ar;'.iD);K/.Tia;s
3.1 Cmi< r<il Analy.si;. of Siiicl*: Mff..-'ts
3.2 Inirij) Yield Solution^.
3.3 El.islic Axial .siablJity So] uL Ions
3.4 Elastic LaltTil Scnbilicy Soliit ions

4. I'Ks^Cl.OPMEN’T 0|- DESiCN FORMliLAS

«*
4.1 tcsnpresslon
«>
4.2 Bending
4.3 Axial Coripression A:.d lier.dinK
4.4 Cowiiarlslon With iment.il

S. CFFKCTn'E LENGTH
5.1 Slopc-D«*flectIon K<|uatIons
5.2 Sid^-Swav Prcvfiued
5.3 Side-Sway Permitted
5.4 Goiaa/nts on The Msc o£ EfCecllvr Length
6. Sll.'PtAKY AND DESIGN EXAWl.E
7. Ar.i:»!a.'Li:r)rH! NTS
e. RLFLDCNCLS
9. TABLES AND FILHKES
10. APPJTJDIX - A suijmary of the propo:,» d
design speclUeatlf.as

r 7^.
r
i. introduction

The use of "tepered structural cleaeots," having tratad depths sad/or

widths, was first proposed by Awlrlklen (1) for reasons of economy In 1952.

In view of the lack of basic understanding of the behavior of these types

of nimbers, in particular the lack of Information having to do with dtslgn

against Instability, the Naval Facilities Engineering Command, the American

Institute of Steel Construction, the American Iren and Staei Institute, and

the Metal Building Manufacturer's Association have Jointly spenaored since

1966 a research project at the State University ef New York at Buffalo to

carry out such studies. As part of those studlee, both analytical a'5

experimental investigations wert conducted on the behavior of auch tapetad

osrbore and the reeulte were used to develop prepoeed design formula».

Thla paper, which ewmariraa many of the conclua loos of thoaa loveatlgetiono,

la concerned primarily with a> the overall analysis of t.sonred mevStx framing,

and b) the development of the design.Consul*» regarding rbe proportioning

of taparad members. .

A general treatment for any arbitrarily taparad been would be very

ambitious aud, in general, impractical due primarily to tha difficulties

associated with the coupling o? fleaurel and torsional deformations. For

thla mason the moat comaom taparad I section now being used was choe.n for

axamlnation, and design formulae for euch members were developed. It la to

be recognised, however, that tha formulas suggested may also be applied (with

caution) to othor sactlona vhlch am "sufficiently braced" to prevent tor¬

sional deformation.

J
 •< o

-2-

à U» coat.ln.d »» thU p*p«r !• • '«»crlptUn of ••»•«! of th.

1
ftn. .od .tl... «thod. appucbl. to Mpered fr»ln,.

It «tao 1» to b. notlid that » Ut.r ..ctloü of th. report 1. dovotcd to

th. probt« of effacttv. tontth of . t.pir.d coluM ln rigid fr«M.
Ih» ba.lc approach oa.J ln tha daralopiaeot of dcalgn formlo« for
taparad Mobera «aa as follow.) thaoratlcal .olutloo. ar. flr.t obtalnod,
than - ».los tb... aolutton. - th. h.l.S.C. prla-tl* ^ar da.lgo fonriao
ara «ditlad to tffact th. aolutlo». bp th. lBtrod.wtloa of appropriât.
1 nultlplylos factor, which ara d.p«od«t only oo th. taparing gao«try. Sloe,
tha sa factor, (by dafloltloo) r.doc. to ualty whaa thar. U no tapar lo th.
anAer, it Buat ba racogulaad that thl. approach aaauaaa that tha currant
a.I.S.C. .llowabl. atr... foraol.. for pritMllc Bwhar. aro odacuata.
 "3 "

2. fRAMS AND STRESS ANALYSIS

from th« structural analyst', point of tiev, fraaaa co^>rt..d of tapaxad

neinbero posuesa unique propartlea %d»lch can be utlllaed In both analysis

and design. Consider, for example, the moment diagram of a continuous pris¬

matic beam under arbitrary loading. In general, the mcr.^at maxima vlll

occur near the center of each span and at the supports. If the cross-section

Is decreased at mld-spen and Increased at the supports, and If the loading

Is kept constant, the moment diagram changas: graater valuea art realited

near the supports and smallar one« are observed at mld-spen. The limiting

condition le reached when a real hingt la specified at tha «Id-span. Thuc

as the shape of the member la changed from Ita originally as»weed outline

of constant cross-section to that of a wedge, the bending moment dlagiem

changes from the Initial Irregular outline of reversed flexure to that o£

single flexure corresponding to two simple, articulated cantilever hecao.

Tapered beee framing will result In weight revlog for many structural

and loading altuatlono. The characteristic redistribution of stresses permits

better utilisation of structural material than does conventional, prismatic

framing. It should be recognised, however, that thia weight sevlug does

not neceeeerily meen thet the resulting structure will be less expenclve,

though in many caaes this will be true.

2.1 Freae Anjlysle

The la-plane enalysla of eerteln, particular type« of single-story

t,r«r.i (no. ha. Wa .hom to fc. .i^>l.r th. «uly.1« ®<

convenrtonalt .rtnotto tltU trnet <1). »or oxt^U. thou*h coah .te.10
frtioe ehovn ln Fig. 2,1 l« •tttlcnlly Indttcmínate, only two unknown ttresa
resultant* are asetned to be associated wit!» each Interconnection. Morecver, *
these unknowns can be determined from the condition that two adjacent simple
frames must have the same deflections at the common link. The standard pro¬
cedure for the analyela of such statically Indeterminate structures would
be as follows: a) The framing assembly Is subdivided Into Ita component
simple frames, each frame subjected to unknown horizontal forces and verti¬
cal forces at each of the connections, In addition to Ita own applied loading,
b) The horizontal and vertical deflections of each frame at each of the lin¬
kage poluta ore expressed In terms of the applied load oed the unknown U and
V forces. In gensral four equations are written for each frame, c) Finally,
by equating the Joint displacements, a set of almultaneois equation» Is ob¬
tained, the solution of which yields the unknown hinge reactions. It ie to
bs recognised that this method la wall aulted for solution by hand fov
small assemblies, consisting of two to four simple frames, or by digital
computer for largar aasembliea.

Since each sambar of each atmpla frame haa a rigid joint at one of ita
ouda and a hinged connection at the other and, the bending moment diagram
in that member la similar to that of a cantilever beam. Aoauming first-order
theory, thla peralta computation of the deflection of a simple frame by
eunmlng the deflections of the component members. A number of deflection
solutions ara presentad In Anlrlklan's paper for tapered cantilever beateo
of various proportions for typical losding conditions. The reader la toferred
to the original paper (1) for a number of examples of on?J..'.ration of the
method of analysis Just deecrlbad.
 '5 *

Other s tenderá «ctaode oí enelyU. beeed on either the flexibility

0, the etiffn... epproech, euch « -o«.t di.trlbntio«. ce» b. edapted elro

t. tepered fr« eeeerf-Uee. For hlfhly «dandent fre»., ho«v.r. . »r.

eutooeted epproech i. deeireble. end «h. finit. el«nt -thod he. dUtlnct

edventege* Ih. r«d.r i. referred to reference 2 for . diecoeeion of the

ue. of finite .l«nt epproeche. to the elution of fr«, coo.i.tlng of

both prlonetlc end tepered «ember, of thln-..U.d .ectlone.

On., the .tree, reenitent. ere obtelned b, en, of the frene enely.U

method, deecrtbed ebove, the exiel force, bending moment, end ehc.r force

varierions .long the metier ... bo determined .. if the »-ber. mere prie-

OAtlC.

2.1 Axial and ?Uxuittl_gtte?.£ea

computet Inn of .tre.ee, in tepered «««. 1, «mpUoeted by the

feet thet et leeet on. no« pnr«tor «t be introduced into the oolution

t0 deecrlbe the verletlon of croo.-eectlonel propertie. .long tho length of

the member. lUmel.t»« vUhln th, nomelly ...»od epproeimotione of .true-

tumi tnec'nenlce, th. uell-huom for«!« for «1.1 -

.till apply -- -Itb the exception thet th. croeo-eoctlonel propertie. very

(3.4). Aeeumlng thet the oentroldel «le et •«“« •U° ““ ^

of eymwtry et thet eectlon. the nononl .tree., e. , due to «o «1 1

ï, «ting et th. centroid of th. cro..-eectlon, located et « dietenoe «long

th* nember, of !•

2.2-1
_
de(*)

o
 6-

The ronnal «treu, ab , located at the sane dlatance s fron the origin,

and at a dlatance c from the centroldal axle, measured in a perpendicular

direction from the longitudinal axle, and due to a bending moment, m, la

oh - 2,2‘2
b K*)
where 1(a) is the moment of inertia about the axle of bending at the section

in question.

The shear stress, , also located at a distance r from the origin,

et a point (x,y) on the cross-section, and due to a shearing force V(z) is

. V(*)Q<i)
2.2-3
V l(*)t

where Q(t) is the first ooinont of area about the centroldal axis. It is

to be underetood that the Q(r.) will very with x and y at a particular crosa-

eectlon.

Becauoe of the variabie cross-secticncl perametera, the points of maxi¬

mum stresses will not necessarily occur at points of maxim«', atreva résul¬

tants. For a particular casa differential calculus may be used to rigorously

determine the maxima of Pq. 2.2-1, -2, and -3. Altamatively, a »imple

trial and error approach may be ««ployed.

2•3 Torsional Stresses

When transverse loads are applied tc prismatic elerîic beams at points

other than the shear center, these loads can be replaird by a atatlcally

equivalent load system assumed to act about the nhear center. This enables

the flexural and torsional stresses to be computed separately and then readily

superimposed to obtain the complete stress distribution. However, such a

 *7-

pvocedura cannot bo directly followed for on orbittoril¡' toporod bon. For

euch meobore the stieor center oleo ny be o function of the longltudlnol
coordinate. Ae on exanple, consider o channel section which hoe o linearly
varying depth end o constant flanged width (Fig. 2,J-o). Here the shear
center convergea toward the veb, in the plane of ayeaetry, oa one moves fro»
the smaller to the larger end of the member. If this channel la cantilevered
at the large end end a concentrated fore« la applied parallel to the web at
the shear caotar of the •■nil end, (Fig. 2.1-h), then the channel both herds
and twiste. (It should be noted that tha exact» coupled flexurel-tnrcionivl
solution for such beams is not available). If the force la applied perpen¬
dicular to the web at the sheer center (Fig. 2.2-c), the te channel will
bend without twisting.
It is possible to uncoup)« the flexural-tovtional b*!.svior of tapered
beams for selected croea-scctlouai gcooatriea and loading», in general,
In ordar to uncouple these deformations, e tapered oeem t'»3t poaseae a
linear ccntroidal axis (which wist also coincida with the longitudinal,
coordinatu exit) and a locua of ehear cantors which lie In a plane parallel
to tha centraídal axis. Furthermore, the loads must lie in a plana parallel
to tha ahear center plane. Seme examples of cross-seetiona which eatiafy
these conditions ere sections which heve two axes of symmetry (I-beam and
rectangular beams), and seetloua which have one axis of aymmetry and the
loads era applied parallel to that axis (channels and T-boame).
Torsional stressas in tapered beams which satisfy the above covwJltlone
and bava small anglas of taper can be computed using the normally recognized
prismatic formulas for thin-welled beams. The St. Vcnent's shearing streee

«
 8

êt a location s along tha Nabar la

2.3-1
T.t * * g

whara G la tha ahaar modulua, t la tha thlcknaaa, and8 la tha angla ot

twiat.
The varplng ahearlng atraaa la

2.3-2

whara g la tha alaatlc aodulus and ^(r.) la the warping statical acoeat.

Tha varying nonaal atraaa la

2.3-3

whara ^(s) Is the harping constant.

Tha noxml and ahaarlng stresaas datemiuad ln tha Ntraar daacrlbad
above can be coablnad to fora tha total atrasa distributions et a given
croos-anctlon. The Nilauri atrasa and Its location along tha »ewbat can
than ba datemlnad ualng dlfCarencial calculua.
▲ norc accurate toretonal analjala haa batn reportad by Loa awl
Subo (5). Datamation« aud atrseaea cau ba fully calculated ualng thalr

theory..
 -ÿ-
• •

3* SIABTUTÏ CONSIDERATIONS

In general, th« result« /«ported In this »«ctíon er« concerned with

the ana ly« ia of linearly tapered beaar-columna of doubly «ynsBetric I-tectlon.

Both "In-plane" and lateral buckling atreogtha are determined ualag the

Raylelgh-Rltt procedure. The end-loading condition« examined are those of

bandtrig and bending plus axial throat«. In both casas, the plane oi "leading"

io presumed to be the plane of the veb.

3• I General Annlyala ?f Siaclc K-*.d)er«

It 1« aosumed that the tapered crosa««ectlonta quels i.on ha» a doubla

«ynmetrlc "I" ahape ao that the ahear-center axis coincide* with the cen-

trnidal t.xla of the member, 'tho flangaa are prosiaaed to bo of constant

width "V and thlcknoa« "t{" al'wis the entire length ot the member, lhe

»♦•b thlcVnens lo also f.te3’aacd to be ualfurm. having a valúa of "t^’. for

tha linearly tiperad member, the depth at any d tat anca 'V from the «wollt r

end can he expressed aa

dt

wl\ora "d^" representa tho exiallest depth at a • 0, and "y M repreeant« the

"tapering ratio." In teres of tbt depth« at the ende c£ che «ember, the

tapering ratio la deflrod by

«
 -10-

Thu*, for • prUMtlc ttcmb«rt V • 0. For • mcaber whose depth at the

large end la three time* thnt of Ifn snaller end, ^ « 2. '’hi* gQunetry

la defined In Fig. 3.1.

The deforwatlonal paveantore fur the «ember In question in the "x"

direction (l.e. perpendicular to the veb) and In the "y" direction (In

the plane of the veh) will be denoted ae "u" and V, respectively.

Twisting about the longitudinal axis, "a", will be defined ae 'V'*

From the principle of virtual diapUcevant,

W • W • 0 3.1-3

where "$u" representa the flrat variation in the strain energy, end "gw"

Is the work dona by the externally applied forcee moving through . the

Assumed virtual displacements, "gu", "ftv" end "ftp"* It 1» to be remembered

that ther« virtual displacements miut »stlafy the prescribed geometrical

boundary conditions for the particular pvobiem In question.

Assuming that the binding rigidity of the meab«r about the y-e.ls Is

a constant, the bending eompou'.nta of rhe flrat variafct n in the strain

energy ere (6)

M* * .

and

The atraln energy for non unifom torsion «ay be written aa (7)

v * £(t* I1 . (.. |i •*,]•. «<s*>*}.s

"diere the flrat two tema rapreaent the flange bending strain energies dun

to warning end the last term ie the pure torsional strain energy. Given

>
 11-

tha ■pacified linear vaxUtlon in depth of tba aenber defined in equation

3,1-1, the flret variation component of the non-unifora tor*Iona1 «train

energy may be written as

♦ 3*<uvJ>V‘+ j^*"} <i*

where, by definition

ío * - 6f

For the loading condition ahown in Fig* 3,2 It it aaauoad that the

applied and-bendlng ffomento can be related .by a non-d.üs&a*fonal paramoter

"m", wliera • The primary bending Domctit af. uvy distance "*M

aoatured from the moeller end of the nenober la therefore

Hjjft) - H^Cor + (1 - of)| ] 3.1-4

whora • 1 ff or 3 + l « The externai work done by the applied loado during

the virtual displacements "du’', "6vM and "Ík?" may be written aa

W • J* {p(u,«u••K,ívVv,^^>,) * M|t(<p6i»M4u"dq»+evM)}d«

where
*. rw»>
4(,)

uubatltutlng these axpraaalonn In the appropriate flret variational erpres¬

st one Just defined, equation 3.1-3 can bo rewritten in the t’ollovins wre

usable form:
 -12-

+ EIä(e)v,'6vn + «IjtÇtyV9 + 5.(j)(l*h^)<PN ♦

P*'*+jm>u«vjw +

3.1-5

to obtain a solution to the tn-plane defoxnatlooal and lateral stability

behavior of tapered vemboro, Bq. 3.1-5 ¿ould be used to astabllah the cor¬
responding differential equations and the aoproprUte boundary condition?.
However, due to the complicated coo-Reometrlcal (i.c. statical) boundary
conditions that are required to handle the torsion preblen, and also because
of the «any difficulties that frequently are encountered when attempting to
obtain direct solutions to cheae types of differential equation», this
approach was not usad (5). Rathar Bq. 3.1-5 was solved using the Kayleigh-
Rlts procedure, which presuma.*, o series of geometrically appropriate dis¬
placement functions. As will b« demonstrated later, the follwirg ansumed
displacement iuoctlooe ylnld aufficlantly accurate answura with compara¬

tively rapid convergence.

m

'*(*) - ¿ >
«"0

3nl-b

where
v, v „d -V «• «>• vnrtr Mrl“
3.1-7
«,,(*) - fi*)*"

vtt(«) - BC*)*“

i ,SÏ
the function* ¿C») "» TS **o -«*)” -*
B' ; ., _
*h* "““ttual
c~cvai obvlout fl.•„»v.ralc
fcgeDíre;<.■
cowltttoM íot th. ..«.iUd ,«««.. »«•

loro, «.« «« «l«“ ^ ,l,> ,„b ..

^ « A 4«fA Va 3.1-5 «n<* 00ti®í t',#t * Ä
Introducing S,.. î.l- • tlOT, «.oit:
.%c ■. «rbltrtry, «h. foU«i^ .«-l»*«- ^
' JO

3.1- Ô
$ H«*» • ‘V> + 1 "8
3.1- 9
? V^%c * ^kÀ1 *'
¿0 „ 3.1"10
? c»;*» • »ï») + £ ^ "4
m-O W«0

vhete
b„, - ^
. » £ «y »ä ^

' J4 nn(,)ïïv»4*
»«n * J4 **

g K tf^eKdx
<Ib * «
nu
3.1-11

»«n * f 4^ ^^
 I»*^ «* < *'«4* > 'Wl * £ «*<*>< ®0a*

Per "its-plW bending, Eq. 3.1-9 provides the complete eolution. Equa¬

tion* 3.1-8 and 3.1-10 ere eitnu'itaneeus equations that asist be eolved to

determine the lateral-torsional buckling load.

For *imply-supported and-condicione the displacement functionr may

be ttKen as
ti£«) • v^t) • ç^a) * * (* « l)t.

Or using the non-dimensional^ indepet.doot pácewàterQ * * , these functions

become
VJÇ) - VC) * VO " C(C * OC* (3.1-1;
In tama of this new parameter, Sqa, 3.1-8, -9, and - JO mfiy be lavritten

ia the following non-dimcncional form;

3.1*13

3.1-14

3,3.-1 r

whnre
 13-

0*0 " ow^ulJdC ♦ J^(l+a)C<Pw<;<*C

ïm - fl»«) »ï < JC . Fra * ]\ < <C

3.1-16
*n * J‘ M»<CK *C . K® * «tan
m

‘ C {c/ (r >*+ )] u¿ <

* J (jy“ [2-/(1 + v{)(<f.; <>• + <i + yoa< »;]} k

^ * £ «o «; »i «t

RIy * î-ending rigidity «bout the y-jurla» asawaad to be constant.

ix(C) - x^d +VC)2(1 +ÍÍVC)

I “ ncroent of Inertia «bout the x-*x1*t ft the «maller end of

XC-
fchft nipmbcr ( ^ " 0).

Vo*
 tapering ratio of tha member

1' * axial Meld load corresponding to uniform yielding of tho member

y° I
at ft* aaaller end ■ oy * A0

- yield atroes level of the material

A. * area of eroea-sactlon at amaller end of member (at C ” 0)

o

- Euler's buckling load about the x-axls, presuming the crcae-

”
sectional properties of the smaller end of tha swmber a
I?tu

ty0 « weak axis radius of gjrstlon at the smaller end of the member *

« constant

• end moment ratio “ » cad ecceutviclty votie for eccuntrically

' «s

loaded members * ~3.

H

Hf iO m moment about the x-esls at tloe general location "Ç" * M*J> + (l-ar)Ç)

GK0 m St, Venant's torsional rigidity et the smaller end of the mead**?

< C - 0)

GK<0 - GK0(l + )

1¾. •'

r(C) - 1
 -17-

A • Area of vcb at C M 0
wo

Ualog thaaa daflnltioaa at wall as thoaa Hated In Eqa. 3.1-12 awl 3.1-if.,

It la pcsalbla to defino freo E<*a. 3.1-14 the In-plane, boi»-eolvrun action.

Correspondingly, the lataral-torslonal buckling strength la obtained by

considering Eqs. 3.1-13 and 3.15.

It la to ba recognised that for tapered œenbers the "useful Haiti"

of the and-aonent ratio 'Vr" will be different froa these values thst ere

Manlngful for prisaetle beeae. In this pepur the uewful limits «rill be

deflnsd ee « »fk, where "k" correrpomls to that part'.cv.lsr end-moment

ratio vhara the max Lun bending stresses at both ends of the metaber ere

«quel. Mk" velues vary with taparlM tatioc "y". Equating the artvono

fiber bending aerees«« yields

Tîno
. «H
3.1-17

or
3.1-1«
• TnvTTîW

Per prismatic members y " k “ •!.

In this Investigation, digital computer piogretae wexn develwpcd to

solve Eqs. 3.1-13, -14, end -15. Specific nuamricel solutions were obtained

to Eqs. 3.1-13, -14, and -15 fot severa; typical tepere:! maobe^s.* The

cross-sectional properties at th* -■•'wllur ifa<5i» of tVi.-* t“(*;*ced rt.vulrt'i •md

ddit lone 1 atiSiUitrlc snd statical pvvpcrtias fox thei c msebeta ^te given lu

lg. 3.4.
The actuel numerical solutions err. omitted from thlfl pspur due to tie
voluminous tables concerned. (Over 7500 iodivldual casos vera considered.)
The Solutions may bs obtained upon vaqueet freon the authors.

n
3.2 Initial Yield-Solution
To obtain io-pUna JiapUcawnta and 'Vtre^tha/» R¡. 3.1-W, which

is hara again raptattd, auat bn »olvfd.

1er each tem in the aertea tha aalue ei\ can ba aacertained, providing

the deteradnat# of tho bracketed ter* ia non-vaniahinf.

3.2-1

The in-plane diaplaceweot "v " 1« therefore

▼ (C) • «i Lr*’“1 ’ vl {sa}<C*>,-C#l) 3.2-2

vhcre TU. *m ni '*« *r* <•««“< ^ “d

rer .erlttt .ml cu be reeltty iDt«frHte«l «ml íumíj, ueint * ccx-

putar.
Horml atreaaea in tka taperad aeubei ace deterwintd 1’ctu a coaaidara-

tioti of the eq-JUtiJn

m,(o4
O *
Ib«>

where
VO *• -WlfvC>* (^0^(0
To facilitât« th« .oluticn proc«*«, aal to xoora generaliM the tMult«, thla

«quation can bi t’.andi.wcn^iooali'icd»

(%) * (^Xtl^r) ? % <5 + VCKl + Y)<1 +

1 !v - (^)¾) j}_:
where denote« cho inltUl yl^ld »c«Be«t the lergcv end of the ««ober

( Ç» 1.0) with P • 0.
Initial yield eolutiow can be obtaiaed by aquetiog to "t.O" the ratio

(o/oy) In aquation 5.2-3. 11»« corre spoiling relntloachlpe between <V/Py0)

and define the "ini«raetIon" at Initial yielding. For «am ccom-

taction (nee Fig. 3.4(e)), the Interaction valuer were obtained tor (hy/Fe^o

0, 0.3, 1.0, 1.3 end 2.0; .- k, 4l. +0.1, 0, -0.1 and -V; aid walwe. of Y

frou «ato t. ats. It ie to be noted rhar r?r thin ««lection of vMicbl««

(•’P" referenced to tbe aoailer end of U*u eftrbor, end "H1 rafoteuced vo

larger end), the UorcliufteT defined, cciwu.ee no^ dlieenetonal ' d.ieign pare-

natore" were found to be aecei»ticU, equal 1er all of the four cioHa-eectloaa

(««e footnote (Ai p. I?)«

Tea term. In the power aerim «ere oaed In tbc calculatloi:» leadioi to

the Inreractlno eslutlor,«. Th« accuracy of the rare)ta vi» cxinlncd by

cctcparlng value« eouiciated wl»h nl »« tern«, and then vi.’. Un To «il

caaes the deviation« cere lese than 0.1X.

Finally, U wuet be reenyhanUed thee the ■eximma atvae« i« »coerid

«ember« of th« type eon«lder«d h«r«ln dooa not oece«aerlly occur at che rad«

of the ~»bcr, even wl.en lafcel aM «alai !"«d« «f* ThU U du*
to the non unlfora eectloa molulue piopertto« of «uch tepared »«Rbera.
3.3 Kxlal St2bUit^iíll^S)-Í£^icn-i
ihen th« dttCcru:inftx»t of th« br.cV.eUd ten, la Bq. va.Uheu.

serait le • «et of homogencoue equation»* end eau0Ot *

* X t in*d P which rnahee the determinei %onleh I»
The particular aulal load, P, wmcn «•
_ hA ^ 4 n

ctltlul (uler W. buckll«« 1--1. p*,‘cr- • 1,1

3.3-1

tu «... 0« ,. «a ——• l- ’alW ^ P*» 4 '7

m «rttt«.i m,. - > - «"m“£Ät cto“';"'

ta nj. 3-.(.) ..a «K«. -i«- o* - • <-* l~“8M * P“‘ *’’•

3.4 .-.«.r.1
A. «» „«a «.rlUr, bucV-ll-s

^,, ..i.,, «- ... ^ 31*u -4 ,

. „ .u>, e,n be rwiltW. l> «!•* to1’"
coovanlcnce In calculation, thee» can
f {(AX* - WW * # 3 . t

f {-«,1* λ + (••• " *®

vhate
AI*» ■

tin *
m
01U *
 21

IW t^tW

X ■ Th« mg«nv«l««

No

Î*-

rl,t l«t«xal-tot«lowfc1 fcuckUn* sr lutlo») 1® «Ä»t«l««^

the clßtnvelu« (l.ft.
by aolviog th« fUr.ct«ri«tU o-iuot^ou cort^pooJiûg te

[\îwn cl [ *W Ii ^c
3.4-î
|[o tcj ' X [bW nJI

vh«ra - 1, M .... ÎI- «>“ 0t ** “U1 ^ Cr

'er ’ Xer
m •• i«r2#r ®tj<J cf thi1 vî*«b«t» 1«
Th« corrtM>ondi«.| "erltlcâl «r.d-«»mnt, *C U* -ûrg»

(>V)tr - vr.r *i

Î1
 -22-

Tos th« appelai ctae whcro no and occeotrlcltlae «»Ut, and th»* »v«h«r

la aubjactad to pure axial thruot, Bqa. 3.4-1 can ba tmcouplad, ylcldina

<A1«i * XBW *» * 0 3.4-3

and

« XXW<V> • 0 3 4.4
Equation 3.4-3 glvaa Xer. • +1.0, or ?ev - P« . Squatlnu 3.4-4 yialda
'O
the toratonal buckling load for axially loaded, tapered columns.

In the alganvalua calculadone for lateral-toralcnal buckling,, aix

torau oi the power »erica were used for oil caaaa where (f* 0. Seven toms

were included fox those cases where er X 0. There reoulted in a wore-or-less

uniform conv.rgence cricsri<m of 99Z < Xn+1/ Xn< 1001. Two snb-aeta of solu¬

tions (pure bending and bending plus axial thrust) for Section I are given

in Fig. 3.5 to illustrate the rate of convergence of the solutions. F^r all
ceses Listad In the figure, ths length of the member ves presumed nquel to

144 inches. Solutions to tue othar cases and lor the ether sections are

comparable tc those Mlurerated.

For all of the beam-column cacos t-on*id«rcd in this report it hsa been

presumed thet the end-conditions at both end« of the tapered members ero

"simply-supported." This vas true for both bending end for warping torsion.

It must be understood, however, that other cace* could h*^e teen equally well

examined, using appropriate algebraic form; tor erampîo, chose given in Fig.

3.3. The csss chosen should give the mere conservative solution« and heneo

is more desirable for design purposes.

o
 -23-

Ufeval-torsionâl buckling colutloc« were obtained for the four someted

oectlcne prevlo-ialy defined. For er.ch, four lengtha of »eaber:,, various V

velue, and four values of *<+1.0, +k. 0 nud -k) were co«.ldernd. In addi¬

tion, five value» of e/d0 wore exe»lned (0.5, i.0, 2.0, 5.0 and •). Solu¬

tion. for (P/Py)o end <M/Wy)t net« obtained using the hereinbefore de.crlbed

procedure. Unlike the lw-plane ca*es the extreme vertat to In the solution,

tend to suggest that these »elected parameters, in and of theoaclvc*, do rot

constitute a unique Invariant set Insofar as lateral-torsional buckling 1«

coaceroed. (Por the detailed uuwvical rvsulta, ace footnote on p. 17).

 -24-

4. DEVELOP MENT OP DESIGN FOBMUUS

Several avenuas of approach ara available for the development of design

approximations to the solutiop referred to in Section 3. j?l*at, using iculti-
variable curve fitting techniques, and eaaentlaUy starting from "Bcratch’',
polynomial expre.aiona containing all the variables could be developed. A
second, basically different, approach could «tart from the assumption thut
adequate design allowable stresses aro now available for prismatic membors,
and these could be modified to handle the tapered problem by the introduction
of certain "to be determined" factors into the prismatic formulas. The««
factors could be calculated from the general condition that

Strength of tapered metubcr___— “ 1 O^o»1

Strength of prismatic «amber baoed ou the emaUcr cross*«ectlon ^ ^

with the restriction that when \ 0 (prlamatic), f * 1.0.

Although both approaches require curve fitting techniques, the second
approach, the one adopted in thie study, offers two advantage« te tha
designer: first, he still will he using the familier A1SC code formulas
fnr prismatic beams - though modification factors will be Introduced: and,
secondly, the«? factors should give the designer *n intuitive fceliop, of
the increase in strength of the tapered member over its prismatic counter¬
part. This increase in strength may help in deciding tbu potential economy
of the structure using euch tapered tmr.bprs. It ia important to recognize
that in the subsequent discussion and development of the i,* per «ri btíÃra -orD1
»nao there ia contained as « multiplier the same factor oí c.nety tnat is
incorporated in the pticmotic formulas. Turthernwta, it be undevotood
that they are applicable to ciau.b'ir« with small tepexi'ig angles.
 ••25“

Bo ley (4)» uelng a aeries solution for tapered ractangtilar bettes and the

Bernoulll-Eulcr theory, found that for tapered angles less than 15 the

error In normal stress computed by the general methods Vcieinbeiore des¬

cribed was leas than a few per cent. Thus, for

4.0*2
Tan 9

and 0 restricted to less than 15° , y ihoul<s be le8g than

y - O.Î68 £- 4,0‘3
ao

Reference to ’'small taper angles" will refer to tapered beams which satisfy

equation 4.0-3. From practical considérât Ion equation 4.0-3 has been fur¬

ther limited to 6. Therefore, when considering the design of tapered members

using the information contained in this report

Y S 0.268 á 6 *

4.1 Compression (4.,I.S.C. 1-5,1.31

Ta this Section the maximum ailowcble stress due to au applied axi&l

load is defined. There are several failure modes of concern: (1> yielding

at the smaller end, (2) streng axis buckling, (3) weak axle buckling, and

(4) torsional buckling.* Since the effective length, Kyt, will be treated

separately in Section 5, the formulas in this Section are baaed on pin-anded

aeraba ra.

* Pure torsional buckling gene roily dee* not govern the design of T.-fheped
columns. Baaed on the procedures definen in Section 3, one may rocdllv
compute tho torsional Imculirg lend of a tapered column when nocaaoary.
 -26“

Considering the elastic case first, a function f (equation 4.0-1) is

sought such that

^tapar " °prlsiaatlc f<Y. ¿0 b, t, w, l) 4.1-1

where
. .EÜJL.
apriamatic " (/(/ro)3 .

Since buckling can occur about cither the strong or weak axis of the member,

the function f will be different fot each case. Observing that the varia¬

tion of the weak axis radius of gyration along the length of a tapered mem¬

ber with constant flange width aud thickness is small, no modification factor

is considered to be necessary. Thus

t£s_ (weak axis) 4.1-2

ataper
<¿/ryo>8

If the member ia braced agsinst weak axis buckling, then the hackling

strength about the strong axis must be doterained. For this case, it will

be assumed that f • so that

*
,, . JÙL— 4.1*3
izUtJ-
Equation 4.1-3 implies that the buckling stress for a tested column of

length I is equivalent to that for a prismatic column of length gt having n

cross-section equal to that of the smaller end of tha tapered column.. See

Fig. 4.1.
 -27

Seeking « Mg" function of simple form with comparatively email error.

^taper

theoretical solutions for the first four sections ¿«finad in Vlg,3.4(a) aud f«r

various lengths were substituted into the equation for "g". After examining

several possible approximating functions, the following was chosen:

g - 1.000 - 0.327 y + 0.0649 (1.00 - 0.0752 y)

This function is tabulated In Fig. 4.2. To indicate the degree of error

involved. Fig. 4.3 contains values calculated by the "Exact" stress method

described in Section 3 divided by equation 4.1-3. It- in to be recognized

that equation 4.1-3 is conservative for valúas greater tV*n unity.

In general, elastic formulas ere applicable for slctvierusss ratio

greater than a certain limiting ratio. For inelastic buckling, that is, for

slenderness ratios less than that limiting value, a transition curve between

the elastic curve and the fully yielded crocs-section (zero slenderness) has

been used in A.l.S.C. prismatic column design. This curve has the form

»taper * °y 4.1-5
 -28-

aa la the case for prismatic members # the limiting slenderness ratio,

Cc , is determined by

0taper (Cc)a

Solving for Ce gives

Cc 4.1-6

Returning to equation 4.1-5, the arbitrary constant "ô" Is determined

by requiring that(¿~) « Cc when cr*^^ ■ % (7y . Thus, substituting the&e

values in equation 4.1-5 giveu

B -

The transition curve, therefore. Is given by

taper (1.0. SÉÍ&L

Kt
)
7
4.1-7

Inclusion of the factor of safety for axial compression In equations

4.1-7 and 4.1-3 yields:

For (g£/r0) < C€

2(¾ 4.1-8

ac i?;
For (gt/r¿) > Cc

4.1-9

where
 -29-

and
weak axia: g » 1,0

streng axis: g « equation 4.1-4

Equations 4.1-d and 4.1-9 are plotted in Fig. 4,4 for Pv ® 42, and 50 ksi.
•'

4.2 Bending (A.l.S.C. 1.5.1.4)

Often a désigner attnirpts to select a particular tapered matute* such

that the máximum stress i* nearly constant along the length of the berra. In

the ea&e of a sloyly-oupportcd bnaa sublocted to end-aontents aleño, ore pos¬

sible optimum design would require equal extreme fiber bending stresses at

both ends, i.e. (M/Sy)0 * . This implies that ,r Sx^/S;. .

•which iß the “liiaitiog’' a.ct«eni: gradient o fk discussed .in Section 3.1

(Equation 3.1-18). llewiu'er, due to the variable mwent of inertia, it must:

be recognized that the maximum ctrase. under this "limiting” movieut gradient

dosa not necessarily occuc at the en.aw w£ tho bupœ. Non?».'nolesJ, for tha

ranges of variables being considered, this condition "■ I« «.n

adequate design eatimatiou <.er tapered beuma subjected to end-memento at

both endti producing single corvativa deferentiora.

In uosi: cowmen practice « structure is designed such vb«t negligibly

small amounts of bending momea1 exists at the «nailer end cf a taperad

member. To accommodate this, c moment gradient function has been incorpor¬

ated in the design foraulca for or *• 0. A uicdlflcatlon factor to the A.l.S.C.

formulas is first sought for the limiting moment gradient case '* - -l-k; n«d

then a moment gradient function is developed to veíate that crliicrti streso

to one corresponding to or =• 0,
 30-

l'rocecding as outliaed at the begtmilng of this 8î ctien, the critical

lateral buckling stress for a prinwtlc beam subjected to candína wowente

is given by

IneteaU of using equation 4,2-1 directly, a ciocifying iactor can be

developed which can be applied to the length of tha member an vea donc for

the pure A/val load case in Section 4.1. Ir.un, the critical bucklin« stress

for e tapered bear, can be enpreseed by

M«> ri4ETyST.^ 4.2-2

(^cr^
“(htF" + ^(ht)* **

where b ^ h (y, do» t.» v,t /.

Equation 4.2-? presura a prismatic beam or length ht: %hosn t;r^a-3ccc;cnal

properties are those o£ tha amllvr end oi the tapersd h?üs Ws* 4.S),

Solving for Î? yields

hk * „Í£h2s— [ i + -\j 1 T
cóc.v
-“t;- 4.2-3
(GXo)a
i5«>cr)* l ’

Tne above equation was used to calculafe value of h £ct tea f;irr. fom

cross-aectlono defined In Fig. 3.4(a). Various valuer, of length vero pre¬

sumed. In total H» such pointe were calculated. While the resulting

n
 -31-

values, plotted as a fimrtion of che tepcrioi ratio, were quite scattered,

there was Indicated a strong dependence cl* the tapevina ratio

Sines a modification factor to the A.I.á.C. prismatic beam formulas

la desired. It la useful to fltfcC examine the basic K&suirptions of the

current A.I.S.C. formulas.

310 X 103^
?or (i/rj) ?• V*
(A.I.3.C. 1.5-6b)

170 X IQ^Cp
(A/rt)a
or
12 X lO^Cb
Wkt (A.I.S.C. i.5-7)

These two equations were arrived at by cnrsiderlng the tvro limiting cásea of

equation 4.2rlî either complete St. Ocrant ruai»tance, or total 'wrplns

rcaiscancc. For occtions which ars thl*. and doe;: ( 1.o. those n.ivlcji hich

warping tea ¡.stance) the tiret formula will generally ¿o/em. Oa th« other

hand, for sections wnich arc thick anti «bellow, th.o saaond formula i» pret-umsd

to hold. Thu«, in developing n mcdificHtion i'aotor ?c tVcsft rufuntíoc.s, it

î.n reasonable to presume two dlatiectly diff -cut fa’t.'vs; one foe those

oasao where the smaller end is thin and de.'-p, and a ¡i'.c .vrrl om when the

smaller end Is thick end shallow.

.Fcr tha four sections prca'jmed in Section 3.C, it turns out that »II of

the sections are basically thick nnó ahelio*T, and thetetn.•« the second for¬

mula g viveras. This la tena, oven though soctioui I oud III ate basically

beams, and eecticus II and VI nr-j basically to’<moa. Qccause cf the general

nature of the problem being studied, sections 1 and 7.ÏT. were usad to Si.nd d.e

o
 -32-

modlflcatlon factor to be used with the A.l.S.C. fonnule 1.5-7. An entirely

different thin, deep section wee selected for use In developing a modifica¬

tion to A.l.S.C. formula 1.5-6b. This specially selected section (section V)

la also defined In Fig. 3.4(a). Typical results of critical elastic lateral

buckling stresses for each of the three sections are given In Fig. 4.6.

For thin, deep sections, substituting the data detsratt^.ed for section V

into equation 4.2-3 and curve fitting the resulting points as a function of

tapering ratio and (t/rT ) gives

o

h,w 4.2-4

For thick, shallow sections, using the data corresponding to sections 1 and

III and fitting the points as a function of tapering ratio and (^d0/A|) gives

4.2-5

The allowable elaetlc stress In a tapered member for a • +k Is therefore

the larger of the two numbers calculated from

170 X 1(^
F, h.l-t
\ "

and
12 « 10?
(h,W0/Af) 4.2-7

To arrive at a moment gradient coefficient for the o ■ 0 case, the

following Is assumed:

*teper I f
•taper „ + k
 -33-

th. p.rtlcuUr eritieêl .tr...« O.«»* 1» *•« ^ r,Ut1"*

[h... to o«.»...«. «1«. for » “ Ò yUM. «U« «« «by ^ *-“U)r'
lov.ti.ly yroporttOT.1 to l.o*th »«*• AU. U U t. h.
Mt^ th.t & l. illlhcly .r«t.r f« M.tUo V (thlo »4 4.^) th» It U
(or ..Clon UI (thUh .o4 *.11»). Hotlo. th.t A.t.S.C. *..1«U. . con-

•unt (C,, • 1.75) »UlplUi for prl«.tl« h*»« «Irt “d

tim to kMp th. curv. fit .Iml* “A t,K! for b0t''

,^l.„t o«fflcl..t ». obt.lo.4 vlth y .. th. »ly «tuvu. Th. rviultlnf

•xpr««slernt arc
170 » 10*Cy
N * (hyt/tl.)’

or
12 * itfC
f\ " hBtd0/Af 4.2*3

«her« 1.0; er - ♦ k
S 1.75 ; et
1.0 + 0.25 WT

ft eeverel different rang«» of parameter« the ratio of the exect

.tt... to th. «lu.. pr.4lct«4 h, «h. um« .1 4-2-* “

(«ft.r rtmovlm th. f.ct.r of ..f.ty) U t»hul.te4 ln Ft». A.7. U r«l»lm
 -34-

the «tror* listed, cooper1«ou should be mede between the errors for V * 0

(which the current A.l.S.C. epcclflcetlon contain»)( end that for oth^r

taper ratios.
Since, as was pointed out in the preceeding eectlc:., there 1» always

in design, especially design involving welded mcabers, e possibility that

inelastic lateral buckling will occur, a "transition curve" Is necessary.

Thus, equation 4.2-8 Is applicable for values greater than some llinitlns

ftr ) . This Halting value, C,, occurs id»en the critical stress is

equal to ½ Fy. Reeiovlng the fretor of safety fro« equation 4.2-8 and laq'or. la¬

the condition that " ¾ Fy at yialds

* -i /TlO * 10? Cl 4.2-10

V“
Observing the slailarlty between equtcloos 4.2-8 and 4.1-8 (colusm buckling),

a transition curve for the allowable Inelastic lateral buckling stress «y be

written as:

for(hw&/r<j^) i C^

Equations 4.2-8 ard -11 ara plotted in Fig. 4.8 for Fy ■ 33, 42, and

50 ksl. Equation 4.2-9 is plotted In Fig. 4.9. The horirontal paits of

the curves In Figs. 4.8 and 4.9 represent 0.6 Fy, the presuaed Halting

values of nonaal stress contained In the current A.l.S.C. fomilas.

 Tb. t^rlnr. ratio ha- a nucb «mullcr effect on U.o allowuMe sfres»

lot tliIn, Jeep beam thin tt J«>e« for thick, •hollow beams. ThU can be

reen from equntien. 4.2-A anl A.2->. It U Impoitant to keep In mind that

there h&M bvm prcMimH thron, hear oil of these derivation« « limiting

moment grtdlont cr • +k which is i fonction of the loperln* rotin V (equa¬

tion 3.1-ltO.

4.3 fecial Compro*«ion and l'endtrp, (A Li

The »r4*lliodK tar deUrmlnlnt '’exatt" eolutlona for *.x •! cemprckston

aed hn-dlnk. necosaaty to tau::« Initial yield ineglectluR risiduil »traaaos)

were diucas sed in Section 3.2. TJ « purpose of tMi Section U to lit to the

Juta referred to In Section J an Interaction curve nt the form.

4.3*1

vht rn

•V,
■ V»
- ^rt”»
O, )y
o

• fi.netIon to be detemlned
cr.
Ih. r»rrlcul.r ImcMon C, tlMlty ...Iftul - •»«* » '-«l« "Ul* * h“

tho (oroi r^•a

C. •
z r1 * TTT ^
I '‘»‘c Vi
where
4.3-2
* l-° ^ # Tifo ' k
 -36-

where

b «• 0.Î5 [ + (J)- 3.0 (¾)9]

R 1* 'Jeflned In e^iuitlon 4.1-4, k lé defined tn equttion 3.1-1C nnd * ^ M0/Ht

lu the mouint grtdlent. Flp. 4.10 contains tabulated values of (¾

a« a function oi end ar/k. Ao an Indication of the error Involved

when u-jlng e-iuatlon 4.3-1, tabulated valúen of the "«xect*' Urge -nd cooicnt

divided by the largo and Muiaent predicted froa aquation 4.3-1 ere plven In

Fig. 4.11 La a- +k.

E juatiou 4.3-1 i» only valid vhen Lh« mxIb'a; axial load In nnd

the uvuiniuw bending mwant ¿a hy . liicae w.xlaia are the Inltlnl .’eld ^iailta

cf. tha lóale. Uovrevcr, a beap.-colum'<i strength «*7 gov*rn*».J 5v lT.-p5.nne

lastaMllty cr by lateral bnckürg cuba*: ot whlct nay •¡cr.ne befere 'ne Initial

yield stT.*.'-* la reached. Titus, tullcwi-ig the il .1,3.0, int ora.* t Ion pattovn,

conation 4,3-1 can be vxivter. morn generally ’n t.'TTfj cf. the mxls.ua e ^al

load and naxlnun. bcnuirg ctóor.nt ’¿'.lût o be«»-co tumi ten entry wlthont faliure.

Equally truc, foi even greater conu is tinny, equaticc 4.3-t can be written lit

terws of etrrcees instead ot load». Kokins theue /;dlu»ifcr<unta give*

CB 4,.1-3
-'i. » 1.0
\\
\
11 * ÍÍ—1
1 <*;>, J
whera
 37-

« «qua£ic'.^ 4.3-2

• computed tixíal cotrpresslve stresn at etcal’er arid

* computed bending streue at larger ond

« allowable axial etress (Section 4.1^

r allowable bending stress (Section 4.2)

12^¾_
ÎSlsVr.T0’
ac

Ao noted abova the form of equation 4.3-3 la similar to A.l.S.C. torraula

(l.C-1 a) except for the introduction o<: the modification tem Cg. For a

priataClc be«-column with equal, an) ttomenr.« CK «bould be unity. However,

the A.I.S.C. avuUficatlcn fur tor (l-fA/F» ) 1» au aprrcr.íaAtion and if- good

for small f4/?¿ . If 1® *a>öU, nod tbercSove .V> dost, to unity,

the ionaiaetlon equation corn ha approxiunted by the ¡'xïuus a ion

4.4 Cornnar in on with Experime-pt^Hl Pats

(Vo major experimental nrngrans have lean carried out on welded, steal

rapered cectlons: au ®*cllec one at Criumbia University and a mn'-e receñí

one at st"*10 Univnraity cvf Nov York at Buffalo,

 -3S-

initially, interest was ir. the dôtemiinatiun of the elastic stability

of linear tapered I-beams and channels. The testa were carried out at

Columbia (9, 10, Jl). In general, the test members were cantilevered sud

were subjected to ax*al and transverse (at single and multiple points) loada.

The tension flange was braced at the free end, which was always the smaller

end for the non-prlsmatic members. (See Fig. 4,.12). The tests were conducted

20 that several members with the same nominal dimensions were lowdnii differ¬

ently, ringing frew pore axial, load - to axial load plus bending - to bending

aiont. However, due to th» variety of loading and specimens tented, only a

few actually corresponded to the assumptions made in the analytic studies.

Results of two test* aeriet are plotted in Fig. 4..12: one corre spotting

to ft taper ratio of z>ro (prf sovatic) and -he other corresponding to a taper

ratio of four. It Is most important ta noto that the 'actors of safety hav?.

been removed tv afford a truer coop¿riron <.{ fch*~ reaulta. Kv**n though this

has beca done the Interflctlon curve ■-.pderpredlcl a the In^d currying ertnatity

two-to threefold. This I« due primarily to tha and capport conditions, and

the. fact that Xy ia very difficult to predict. The important chnervation to

•»
bs meda fr*>m ?ig, 4.12 ia that the inttruction curva equally undc’^redict?

the otrength of the prismatic nriuber. lh.. tents should not. renlly be compared

with t.ht developed design formulas, since the formulaa are based on simply-

supported meiabers. The toct reaulti clearly ’*afl«:t the affect cf end res¬

traints.

Later, a second program of testing was initiated at StJNY at Puffrlo (12)

with tbs emphasis directed toward t>o inolautio Inter«), stability of t.p'-rad

beams, Recognising that tapered ueu&era arc inefficient ayial load corriera.
ch®. test progr/uu was sat up to eiwulate high bending strc-is* a ad lev e^lal

stross. Thus the test **,suits do not cover the. entire range oí vavisMeß.

Also of interaat, the apneianna «era loaded and '»upperted piroiiav to gable

roof beams: the axial force conpouent varies ;»p the beam deflects. Results

fron tore2 specimens are plotted tr Fig. 4.13 along vith their corcennondiag

interaction curves. The apeeiojens are essentially the samt.1 except ¡to. utir

r»*tío.
In contrast Co Fig. 4.12, Fig. 4.13 aungeats that there in w.ce (conser¬

vative^ error in the design formulas as the taper Increases. This la bcea-ise

all Ibrun ¿pscimens foiled by lateral bvckiiiig very naer the ■‘•ov.pou yield

stress. Equation 4.3*3, it shoe.VJ he resruabeted, was derived Irrmj initial

yiald confiderations.
 -AO-

5. rPFSCTtVE LENGTH

The eolutions obtained in Section 3.°. vers for pln-auded v'.anil'trr., Tf

the column in question ia restrained at either one or «t both ai: its enõf,

then ths buckling load will fcr different eud ;-i differont solution ».«ist be

fourv.1. A. rant rained coJunn cf height i, can be com ido rod as a pin* ended

enlmnn of height lyt, where Kv is the i;o-c«Med effective length vv.ctor.

‘■n Section 4,2 9 aoeiewhat. similar apçi*oacn was used to ohtuit, s function p

which wia-ed the btickling lend of a rin-ended prisoatic. columr. to that of

d pin-ended tip a red column Uue to the particular analytic procedura ur.<»d

Jr. the development, this function g has been absorbed Into the effective

li-ngth factor K,^ derived ia this r.ection. Thus , ae will be heroin

derived v is interpreted as relating a rest reined tiper.-iá column of height Í

to a r»>i'.-radi.d prismatic colutwi of height y,.t,

ib«^ effective Itvagth Is determined by coiididcrin.g a rectangular rigid

jrd5it (71g- 5.1/ composed t.f pricmutle becuna and r^peTed columna. The top

ueeis his e monsarf of inertia T.r and the bei.tom fce.&’r, ha» a mos-eTit. of inertia

yB» -:hte KV& prasuT.ed to c:t at the cortroids of the coVsmne. Vie fn».-*

haa two rapport conditions ut th-.. plane •>£ the frame: (1) vecficai sjtpport

äc A and B (side-sway prevented by n horizontal suppôt • -vf 0) ant* U; vertical

support at A and b (aide-sway permitted). Finally, it is to be ctnphaalted

teat Orly b ickling in the plane of the frame Js conRide red in these developments,

5.1 Slopo-Deflnctiovi Hiuat 0011.¾

/he freire is analyzed by using a lope-de flection equation for tapered

 -4I-

Kcrtbers cf the, follwiag torn-. (13), (sea fis, :,.2.),

AB ~ C - V ,. , 4ab 5J.-1
HÏO M Sro -AB + i 4 eAB

J8A + V- r . 4ae -
EIo “A 2I0 ' f-ll» * ~T~ - hf

(tUtmtk <.w, tM, «ul ,.M - CM ete ¿etftnainoj by ccn-siíei ing cb»

AIK«K.et,l «quucior, íor « Uf.r» coi««.., ptosmina all of ,w .,tsurt

1.0 the ctosa-seeUCT U ooncoofratoô to th» fUng«». •¡hese caeiHeimU

Uve iUHtmt f«m> tl*|>««!t% on the ,«nr.Jty ,//. .»hern

q ™ nA P
ri-j
?„ « 7,4 £;o
and i*
£J ta dr áo

C,:o rgng-i ^ < .¾

CAA * 5 q L5 (“) - i 1-4 (a)

joy
'AB -w.
q 'Vj+v* (ra-n) * 3 T-5(8)

c ~ I - -Jc:— {' /'«•xA ,,]

£B q q(v^> Lp Uw 4 5*

o
 -42-

vhcrc

w » (1 + v)e 5.l-7(a)

n - (l + y)mP

For Î:
V*
5.1-4(b)
CAA ‘ ï ' q“ cot tf! h ^q'

C,.B ■ CBA ‘ ^-T 6 «cte lit(l+v)] - i 5.1.5(b)

CDS “ q - 5(¾) COt Ce ln<l'h')3 ' 25(¾)

vhere

5.1-7(1)
8 •

If che member Is prletn&tLc (v * 0)» au¿ there lo no axial force imposed

5.1-8
CAA ‘ CBB ’ 3

5.1-»
CAB ' CBA ’ 6

Stiffnc.ts coefficients for tèpered SïOisiifct.‘j are plotted In Tlfc,9* S.3 to 5.6

£8 a function of P/IV and v • Pnth compression and tension a<i»l thrusts

'’’yÄo
are pruuu...cd.
 •43 "

0.2 Bi de »Sway is Prfcvatitfcd

Tfcê fra.M'î illustrated tn Fig. 5.1(a) is consi.!*red in this Saction.

Equations 5,1-1 and 5.1-^ can be applied to each arsnber in the fräste. The

cceifir.tents for the tapered colu’uns art given by equations 5.1-''», -5, and -6

and for the prismatic bearac by equatlona 5.1-8 and -9, From synanetriC;, gen-

rcetric, and a qui librium conditions the eight slopa-deflcc‘*ion equations reduce

to two.

r
(i-U + *v
5.2-1
’ Sa
L

where ^ lí
^ h

b Ic.

*r. t tt

Noting that the coefflcierto are a functior, ni the applied load, r.£»l

that the set is hmogeneous, the critical load is detennined by forcing the

detenaicate of the coefficient nvitriv to uçwnl /.ero. The cump/exícieí, oi

the resulting equation ret^iire that the critical load be found by Incrementing

the applied load end tasting for a r.cro solution. Once the c.cici^àl load is

found then the effective length factor can be calculated from

 -44-

5,:-2

The calculations for effective length were carried out £jv values ci

R.j, and ranging fron aero to infinity, P:i2p. 5,7 fo 5.15 contain graphs

of the etfective length factor® for the case where. Wside-sway iò prevented.'”

5,3 Sid a-Sway is P^rrnltte «1

Tp.j .tability of the fraaic illustrated in Pig- 5.1(h) it here considered

Proceeding in a similar way -so in S, r.tior 5.2, the eight *iope-ddflactiovi

oquetien can be reduced to four equations involving the five nnknowne•

MÀÎ-’ mbc» ^cr»» 4n<! A fcii'iaticn c.m be found by cons lie ring

i-ree body dlegrars and shear equilibrium oC the daflccted mme. Chuo die

»1

! 0 5.3^1

0
I r,
 Bwu« 5.vt U . ..e Ci
, r- ^ the coefficient ^ívlii W'*LÜ 7-to*
MWt«n «*Ut. «*» ^ °‘ C
, . ..,. COTy,l„lti« involvedi tue crl'.ic»! x») **-
Ai.1». bvcavve of the nuMoUol 0^..1
-v, „piled io*Ä and teotlW! «•* » *••* "lm
oVtnlnnd !... l.t.nren»t.Uni *. -rf
, t. 0n,.a t!ie «itinal load in t*«*. ** '**
fiie dîtevminafo. 0n.-o . lS 5.» . ««1» •*
te calcvlv«.) fvoa oonatto» 5 o-v- »” • H „
1 . *.* M-4- '-»v'aV 15 .u•
/ » „ rr.*r cüSC .oXC t- ^ 7
v.,e ^rcf tive leagth lactoia vor tl-

5 4 Cor^*tfi.iffi.tUe r .
u...tjn it we« assumed tut huckUr-K occ a -
£“i the discusión, it
,rpn, AKi, of the ta-oered member, However, if uv t
only afc-o: the etren¿ &kI> . , *,«
.Ud,. vf. y. velues fir prls^tl- co-ci^ * • •
occurs about the weak axis, • • , t t,.,. v,i0V direction
•a n ... «-b«» t-U» »vS,v.tiaU/ Vxt«..tc d> «**
9h.-.ud be useo * - . . ,atc<i ;;o rhe ler.ßth
« . V a tapered coiuvr. ie *■ t
älnce the effective Hr.¿-.T r
Sv K the A.X.SvC. deat^n fonauus -
,» « priematie pin-ended aMu-u by 1^.. to
. .... .'.nr od Ciov. .'Of ■ 5** t'1 .
ï ni I ’ can be •**» *•>» S l" -** .
® i »vet Lhe tereced col inriO-
neoath abcusjcd that tue
The solution Cor tue efte-.-tK* e ,
♦•»Mir'd at their or,*.» t*> priBíaâtU •
in the rigid «*» ,en ’ 4.-^0-13 'nr the3a
^ ►> ,-r tv. couivalert .1^.10*-
, ^ i h-se beams < re abac tayexen, «.un t.. .
however, Ihwse team- ocnutic
, 4 . rhe tonertd be^e can he thou3n, of %.. ,0
I,cana should be used, i. *
« a Mfciffneis. wtifiness ct A c. h - Η
beans UhvUk Mree cnd 3tlffne . . T, hv
..._«i... ti *„rt ,a .".ivtr ny

H*. KI< 5.Ú'-'5

0
 -46-

„here CM. C85 evj «« Ml»* if M-'*. -5 «* *«• ********

5.4-1 to che .U«n»M o# * prl»â».lt !>eœ «tth the Ht trA hold lUai,

(4FX /<•). and solving for Che equivalent „orient of inertia gives

V' lo r CBB -, 5.4-2

^ “ ZF*“ r

7he motors o? Irertia Ifeq is used In the erpcesalnr Äy in place of ¡t U

th3 rapere'i Seam is on tor. and sin.Ui.viy ver R¿ when euch a equation t>.L^s

ni: the base of the s truc tu ¿o. The stiffnee;. tactor t.w -ß obtai o-d

fctri fig. J.3 co 5.c.

Roprcduced from
best available copy.
 -47'

6., nb-MMARV AND

rnis report bas trieo to cover all major aspects of th* <ls«l*n of

taperuci trauter; from total fciime analysis - to fttre»-; &T,aly^.s of i’iítv-

tdtiai numbers - to the development of tapered design fonrn' though not

In that order, Since dcrcription? and methodology of frame and sî'^ibS

analyali for meokers of variable cross-section can ha fo*-t.d elcewbere, thr

primary objective of the report was to present the ra-Jerale l*\ilnd the

deveiepraent of the design fornm’as.

To facilitate nresentatlon and undeiotandiug, the authors have

reviewed and esttendad the backgrt- i"<i fer rhe* calculrtlon of fheCifticn«.

crideal loads for there types yl aioubtrs. Many f.aa<?« mr« u^ai'JrcJ, aud

using those calculated solutions modliicatInn Motors were deteralncC oy

nonaal curve fitting technliuer vhich alio«/ application ol the current

prismatic (4.Ï.SX..) envigo formula a. Tbs veswltlrg modified d«.t>igu tor-

raulaa vrere compared v.c. ¿»xpwi iaietitb 1 reanl tst and *o bn con a t .at-oo

to iho same degree an Che cvrv'«ot A.I.S,C. design iortwlsv fit prirm'tle

uauberg.

It 1:, to be und etc food that the design formulas dove lop a« d.jcrihc

the gror.rt behavior of isolated mctsb.rs. /hat icf the oiirifal underlying

Assumption that: members ara adequately proportioned against locsl bucklicg

sure bn made. Until lerthor research is performed on this topic, wound

engineering Judgement bas*.d or, ftznerloiK'* usi*ig priamatic mr.mhecs must be

exercised to snsurc adequacy of croes*siectional proportions.

In the interact inn formulo, the computed and all wähl« art»!, strass

are based on thn smaller end prop'-^»83• Vhereua, tho computed and aU^r-

able bending stress are based on the larger end properties. Th*-*_J^d

[ã^duTop»,
consequence of non-dimensionalliing equation 3.2-3 which waa used in

determining the maximum streps.

The uses of the suggested formulas are perhapa best illustrated by

means of an example. It should be recognised, however, that the case sel¬

ected is only for demonJtratton purposes and docs not, in and of itself,

describe a range of suggested application. Thi structure chosen i* a

portal frame with an overhang on one side. (See Fig. 6.1). Moreover, it

is specified at the outae.l that the overhang nmi both of the columns are to

be tapered, built-up members whereas the main beam is to be a prismatic,

wlde-f1 ange section. All members are to be of A-36 steel.

The fra.lie was analyzed by the slope-deflcclion method makinç U'C of

the stiffness and carry-over factors contained in Fig. 0.3 to 5.6. fbe

first assumption made was that all members are prismatic and are of same

si^e. (It should be recognized that for the purposes of this analysis

the overhang can be replaced by a moment and a shear). The slope-dcflection

equations (assuming that oide-eway is prevented) can be solved in lerr.s of

the required moment cf inertia. A 30 V?F 132 section satisfies the design.

Next the columns are assumed to have a taper ratio of, say, on” (th ^

is, y «• 1.0). The frame at.alycLs is repeated and the design formulae checked.

At this point the pruv«ntoU side-uwuy cace wae egoin solved to obtain the cionicnt

diagram shown in Fig. 6.2. Tue use of the design formulas will be illustrated

by checking the member AR. The section properties can » c r »Uulnted from the

dimensions listed in Fig. 6.3.

The axial stress at the smaller end: 3.7 k»1

3380
Thu bending stress at the larger end: U - 13.3 ksl
334
 -49-

Thc Al lownblo Axial Srrc.-a:

• 1.0
V

Cc “ 126 ln*

wink axis (treat as prismatic) r^0 •* 3.06 In.

K ■ 0.70, 44.0
ryo

.strong axis*-wldb*sway allovfd, rxo ' 3.J3 In.

b . bIo . 60 X 471 A e,
UT U X 4V19 " 0,51

.. _ bIo _ 60 « 671 _ ..
’S • Î7; • “uTo- • •

rvorc Fig. 5.15 Ky « 1.5, *«d !xi * 54.0 In.

rxo

Ihe allowable axial (trass Is thfiufarc

1.0
(%rf
• 17.99 ksi
4V r.s.
Tlio Allovohlt» r,opting Str»*:^:
Y
l'ëscd on the 4c:aptoi>bion êud 1/1 oí the vib9 J.l? In.
o
The nx'mcrt grsdinnt Is or 0, thut tha sw-ont gridlnnt roof fie lent Is

given by j 0
o.ro
't>Y 1.0 o.m?
 50.

Tiie division between elastic and inelastic bending bebaviov is determined

from the formula

:t - 106 in.

Tbc lOiicLh modification factors for bending are

hit
w
» 1.0 + O.OOddSy 'VU'rl -003
and
h# - 1.0 0,0230y m\Jido/hi * 1.32

Since h L/rm » 62 < C

w To r

,0 ,~ Jf .- 21.6 kai
-! [>•' 2* (.1
or 'T

12000 C,
Jft - 27.8 test > 0.6 f " 22 ksl
Vo'Af y

therefore, - 22 ksi.

Combined Stress

From the previous calculations

,t V ,fL s

tt\ ° ‘ -M ,nd (él

(r;)o by 4 ' °-n
Since a ’ 0

cms 1.0 - 0.9 (,4-) f 0.6 (^-)0 - o. 94

V
12nsK
where l*1 • - 51.1 ksi
e 23 <u Ur )»
' y xo

Thus

Ak) m .960
0.21
1.0-0,077 (0.71) - 0.93 < 1.0
• +
\i' /
ny o
(>-)
1. V 6
e o
Y
 -51-

•n.~ Mr«.* col»» selected is accept .Mc. Application

of the- design formular, to the otl.cr members yield.", the Moal dcslpn.

as shown in Fig. 6.3.

It is noted that, if atual force effect *= « be included in the

analysis of frames consisting of tapered mentbors, the slope-deflection

equations described in Suction b may be used in an iterative procedure

by utilising the stiffness and carry-over facrors given in Fig.« 5.3-

5.6. Again, the axial force in tapered columns should be estimated as

a ratio in terms of the Duler's buckling load based on the smaller end

geometry,( r'ex)c.
 7. ACKNOWLEDGMENTS

The results reported herein are based on research studies carried

out at the State University of New York at Buffalo. These were financially

supported Jointly by the Naval Facilities Engineering Command, American

Institute of Steel Construction, American Iron and Steel Institute, the

Metal Building Manufacturer's Association, and the State University of

New York at Buffalo. Technical guidance to the project was provided by

the joint WRC-CRC subcommittee on Tapered Members of which Dr. A. Amirikian

is Chairman. The subcommittee members are: Messrs. A. Amirikian, J. H.

Adams, D. J. Butler, T. R. Higgins, R. L. Ketter, K. H. 'ccpman, W. A. Milek,

L. W. Lu, G. C. Lee, N. W. Rimmer, A. Toprac, I. Viest. Messrs. Ronald Friend,

H. H. Lee and Yasuo Tada, formerly graduate research assistants in the Depart¬

ment of Civil Eng*..eering. State University of New York at Buffalo, maos

significant contributions to the investigaticn.

8. REFERENCES

1. Amlrikian, A., "Wedge-Beam Framing," Trans. ASCE, Vol. 117, 1952,

p. 596.

2. Lee, G. C. and Morrell, M. L., "Finite Element Analysis of Space

Frames of Thin-Walled Members," Civil Engineering Research Report,
SUNYAB, Oct. 1971.

3. Lee, L. H. N., "On the Lateral Buckling of Tapered Narrow Rectangular

Beams," J. Appl. Hoch. V. 26, Sept. 1959, pp. 457-453.

4. Bolay, B. A., "On the Accuracy of the Bernoulli-Euler Theory for

Beams of Variable Section," J. Appl. Mach. V. 30, Sept. 1963,
pp. 373-378.

5. Lee, G. C. and S*abo, B. A., "Torsional Response of Tapered I-Girders,"

J. Struct. Kiv. ASCE, V. 93, Mo. ST 5, Oct. 1967, pp. 233-252.

6. Bleich, F., Buckling Strength of Metal Structures. McGraw-Hill, 1952.

7. Venkayya, V. B., "Lateral Stability of Non-Prismatic Continuous Beams,"

Dissertation Abstracts, V. XXIII:2, Ph.D. Thesis, Univ. of Ill. 1962,
98 pp., Univ. Microfilm Serv., No. 62-2986.

8. Lee G. C. and Kettcr, R. L., "Residual Stresses in Welded Tapered

Shapes," C.E. Research Report SUNYAB, Sept. 1971.

9. Krefeld, W. J., Butler, D. J., and Anderson, G. B., 'Welded Cantilever

Weds*5 Beams," Welding Journal Research Supplement, Vol. 38, No. 3,
March 1959.

10. Butler, D. J. and Anderson, G. B., "The Elastic Buckling of Tapered

Beam-Columns," Welding Journal Research Supplement, Vol. 42, No. 1,
Jan, 1963.

11. Butler, D. J., "Elastic Buckling Tests on Laterally and Torsionally

Braced Tapered I-Beams," Welding Journal Research Supplement, Vol. 45,
No, 1, Jan. 1966.

12. Prawal, S. P., Morrell, M. L. and Lee, G. C., "Bending and Buckling
Strength of Tapered Structural Members," paper to be published in
Experimental Mechanics.

13. Lee, H. H., "Factor of Effective Length for Tapered Columns," Civil
Engl nearing Project Report, SUNYAB, 19/0.
ADDITIONAL LITERATURE

14. Abbassi, M. M., "Buckling of Struts of Variable Bending Rigidity,"

J. Appl. Mech.. V. 25, Dec. 1958, pp. 537-540.

15. Abbassi, M. M., "The Second Approximation for Buckling Loads of

Tapered Struts,", JL Appl. Moch.. V. 27, Mar. 1960, pp. 211-212.

16. Anonymous, "Useful Charts for Tapered Beams," Engineering. London,

V, 186, No. 4871, Aug. 28, 1959, p. 108.

17. Appl, F. M. and Smith, J. 0., "Buckling of Inelastic Tapered Pin-

Ended Columns," Engineering Mechanics Div. j.. V. 94, April 1968.

18. Bazant, Z. P., "Non-Uniform Torsion of Thin-Walled Bars of Variable

Section," Publ's. Int. Assoc, for Bridge & Struct. Engng.. V. 25,
1965, pp. 17-39.

19. Boley, B. A. and Ziranoch, V. P., "Lateral Buckling of Non-Uniform

Beams," J. Aeronautical Sciences. Vol. 19, No. 8, Aug. 1952, p. 567.

20. Botizan, "The Generalization of Takabeya's Method for the Analysis of

Multi-Story Frames with Members of Variable Sect! n," Acier-Stahl-
Steel. V. 33, No. 10, Oct. 1968, pp. 437-444.

21. Cram, A. A. and Hall, A. S., "Influence Lines for Non-Uniform

Continuous Beams," Civil Sngr. Trana.. Australia, V. CF.6, No. 1,
Mar. 1964, pp. 25-28.

22. Cranch, E. T. and Adler, A. A., "Banding Vibration of Variable

Section Beams," J. Appl. Mech.. V. 23, March 1956, pp. 103-108.

23. Culver, C. G. and Preg, S. M. Jr,, "Elastic Stability of Tapered

Beam-Column*," J. Struct. Ply. ASCE, V. 94, No. St2, Feb.1968, pp.
455-470.

24. Cywinski, Z., "Theory of Torsion of Thin-Walled Bars with Variable.

Rigidity," Archly, inzynioru Ladovel. Vol. 10, No. 2, 1964, p. 161.

25. Desay, Praka?h, "Determination of Buckling Load for Columns with

Varying Section - An Approximate Easy Procedure for Certain Cases,"
The Structural Engineer. V. 45, No. 6, June 1967, pp. 205-209.

26. Diwa, A, F. S., "Three Moment Equation for Variable Depth Beams,"
J. Struct. Div. ASCE, V. 90, Ho. ST6, Dec. 1964, pp. 149-169.

27. Fleming, J, F., Podolny, W. Jr. and Rile, T. S., "Drift Study of
Tapered Frameworks," Civil Eng.. ASCE, Aug. 1967, pp. 38-41.
 Fling, R. S., "Tapered Girders Cut Gym Cost," Enfi. News^ccord,
28,
New York, March 19, 1968, pp. 148-150.

29. Fogel, C. M. and Ketter, R. L., "Elastic Strength of Tapered

Columns," J. Struct. Dlv. ASCE, V. 88, ST5, Oct. 1962, Proc. 3301,
pp. 67-106.

30. Fok, T. D. Y. and Mosure, T. F., "Analysis of Non-Prismatic Continu¬

ous Structures," J. Struct. Di.v« ASCE, V. 92, No. ST1, Feb. 196 ,
pp. 1-10.

31 = Fok, T. D. Y., "Formulas for Deflections of Cantilever Beam »• th

Variable Section," Civil Engineering. ASCE, V. 28, No, 10, Oct. 1958,
p. 763.

32. Frisch-Fay R., "The Buckling of Struts with Varying Cross Sections,"
J. Inst, of Engrs.. Australia, V. 31, March 1959, pp. 81-83.

33. Gaines, J. H. and Voltecra, E., "Upper and Lower Frequencies of

T&pered Beams," J. Enpng. Mech. Div.. ASCE, V, 94, No. EM2, April
1968, pp. 465-488.

34. Gatewood, R. E., "Buckling Loads for Beams of Variable Crosa Section
under Combined Loads," J. Aeronautical Sciences, Vol. 22, No. 4,
April 1955, p. 281.

35. Gere, J. M., "Moment Distribution Factors for Tapered Beams,"

Civil Eng.. ASCE, V. 28, No. 7, July 1958, pp. 597-598.

36. Gere, J. M. and Carter, W. 0., "Critical Buckling Loads for Tapered
Columns," J. Struct. Dlv. ASCE, V. 88, ST1, Feb. 1962, Proc. 3045,
pp. 1-11.

37. Ghani, A. F. M. R. A., "Shakedown Analysis of Non-Prismatic Beams,"

Dissertation Abstracts, V. 27:6(B), Sc. D. Thesis, Washington Unxv,,
1966, Univ. Microfilm Serv. No. 66-11, 886, 133 pp.

38. Girljavallabhan, C. V., "Buckling Loads of Non-Uniform Columns,"

J, Struct. Div. ASCE, V. 95, No. ST11, Nov. 1969, pp. 2419-2431.

39. Gordon, A., "Deflections with Varying Moment of Inertia," Civil

Em>. ASCE, V. 25, No. 2, Feb. 1955, pp. 98-100.

40. Hamzanl, H. H., "Plastic Analysis of Non-Prirraatic Members,"

Dissertation Abatracts, V. XVIII; 4, Fh.D, Theels, Stanford univ.,
1953, 146 p., Univ. Microfilm Service No. 58-1290.
41. Hartmann, A. J., "Elastic Lateral Buckling of Continuous Beams,"
J. Structural Div. ASCE, Vol. 93, No. ST4, Aug. 1967, p. 11.

42. Harvey, J. W., "Buckling Loads for Stepped Columns," J. Stjnict^. Dlv.
ASCE, V. 90, No. ST2, April 1964, pp. 201-222.

43. Heidebrecht, A. C., "Vibration of Non-Uniform Simply-Supported Beams,'

J. Eneng. Mech. Div. ASCE, V. 93, No. EM2, April 1967, pp. 1-15.

44. Hope-Jones, E. F., "Deflection of a Non-Uniform Beam,"

Civil Engr. London, V. 31, June 1965, pp. 154-158.

45. Housner, G. W. and Keightley, W. 0., "Vibrations of Linearly Tapered

Cantilever Beams," J. Engr. Mech. Div. ASCE, V. 88, No. EM2, April
1962, pp. 95-103.

46. Jenkins, W. M., "Influence Line Computation for Structures with

Members of Varying Flexural Rigidity Using Electronic Digitai
Computer," Structural Engr. London, V. 39, No. 9, Sept. 1961, pp.
269-276.

47. Jones, G., "Using the Column Analogy to Determine Moment Distribution
Factors for Non-Prismatic Members," Civ. Engr . J* 1:.»
London, V. 60, No. 710, Sept. 1965, pp. 1303-1305.

48. Krishnaswamy, K. T., "Method for Determining Deflections in Beams

of Variable Stiffness," J. Amer. Concrete Inst., V. 60, No. 1,
Jan. 1963, pp. 157-160.

49. Krynicki, E. J. and Mazurkiewiez, E. Z., "Frames of Solid Bars of

Varying Cross Section," J. Struct. Div. ASCE, V. 90, No. ST4, Aug.
1964, pp. 45-174,
•

50. Larnach, W. J., "The Analysis of Vierendeel Frames and Girders

with Non-Prisraatic Members," Civ. Engr. & Publ. Works Rev., London,
V. 56, No. 661, Aug. 1961, pp. 1045-1048.

51. Light foot, E., "The Elastic Analysis of Open Frame Cantilevers and
Girders of Variable Depth," Civil Engr- A Public Works Rey., L^°n»
V. 53, No. 629, Nov. 1958, pp. 1275-1277; V. 53, No. 630, Dec. 1958,
pp. 1411-1412.

52. Lin, K. H., Rossow, E. C. and Lee, S. L., "Inelastic Stability of

Tapered Web Columns," Publ's. Irt'i. Assoc, for Bridge A Struct.
Engng.. V. 28-11, 1968, pp. 113-Î36.

r*
G
*3
53. Massey, P. C., "Th*; Lateral Stability of a Non-Piltic Beam,"
Building Science, Vol. 2, 1967, p. 273.

54. Ondra, 0., "The Determination of Moment Distribution Constants of

Members with a Variable Moment of Inertia," Dissertation Abstracts,
V. XVI:2, Ph.D. Thesis, Lehigh Univ., 1955, 137 p., Univ. Micro¬
film Serv. No, 15,332.

55. Ormerod, A., "Moment Distribution Factors for Members with Change
of Cross-Sections," Concrete A Constr. Engineering^ V. 56, No. 4,
April 1961, pp. 153-158.

56. Ormerod, A., "Calculation o- Critical Loads for Struts of Non-

Uniform Section," Civ. F.ngr. A Publ. WorksRey., London, V. 53,
No. 630, Dec. 1953, pp.‘1387-1388.

57. Ormerod, A., "Bending Moments in Pin-Ended Struts of Variable

Section," Civ. Engr. 6 Publ, Works Rev., London, V. 54, No. 633,
March 1959, pp. 333-334.

58. Otakar, Ondra, "Deflection and Slope of Beams with Varying I,"
J. Struct. Div. ASCE, V. 89, No. ST1, Feb. 1963, pp. 25-48.

59. Peteu, V., "A Method for Determining Deflections in Beams of

Variable Stiffness," J. Amur. Concrete Tnst., V. 61, No. 2,
Feb. 1964, pp. 239-243.

60. Pel, M. L., "Matrix Solution of Beam with Variable Moments of

Inertia," J. Struct. Div. ASCE, V. 85, ST8, Oct. 1339, Proc. ¿U8,
pp. 1-14.

61. Fschunder, R. J., "Tapered Twoera with Variable Wall Thickness,"

J. Struct. Div. ASCE, V. 93, No. STS, Oct. 1967, pp. 515-531.

62. Robertson, J. C., "The Lateral Stability of Beam of Non-Uniform

Cross Section," Diosertation Abstracts, V. 24:3, Ph. D. Thesis,
Stanford Univ. 1963, Univ. Microfilm Serv. No. 63-6444, 97 pp.

63. Rogers, B. G. and Munse, Jr., W. H., "Plastic Analysis and Design
of Non-Prismatic Members," J. Struct. Div. ASCE, V. 90, No. ST5,
Oct. 1965, pp. 299-324.

64. Sanko, F.. and Willcock, B. K., "Computer Analysis of Bridge Having
Varying Section Properties," The Structural Engineer, V. 45, No. 11,
Nov. 1967, pp. 395-400.

65. Shou-llng, W., "Flexural Calculation of Non-Uniform Members,"

J. Struct. Div. ASCE, V. 89, No. ST6t Dec. 1963, pp. 271-295.
66. Smith, R. B., "Design Charts for Symmetrically Haunchad Members,"
Civil Eng. ASCE, Aug. 1965, p. 72.

67. Strauss, M. W., "Design Constants for Beams of Variable Section,"

J. Amer, Concrete Inst.. V. 27;. No. 6, April 1956, pp. 839-849.

63. Srinivasan, A. V., "Buckling LoaJ of Bars with Variable Stiffne.na;

A. SimpleNumerical Method," J. Amer. Inst. Acronnotlcs & Astro¬
nautics. V. 2, No. 1, Jan’1964, JTp. 139-140.

69. Sami, S., "Continuous Girder Bridge with Variable M-ment of Inertia,"
J_. Struct. Div. ASCE, V. 86, ST1, Jan. 1960, Proc, 2346, pp. 19-39.

70. Kitipornchai, S. and Trahair, N. S., "Elastic Stability of Tapered

I-Beams," C.E. Res. Kept. No. R166, Univ. of Sydney, March 1971.

71. Vickery, B. J., "The Behavior at Collapse of Simple Steel Framer,

with Tapered Members," Struct. Engr.. London, V. 40, No. 11,
Nov. 1962, pp. 365-376.

72. Wang, Shou-ling, "Flexibility of Axially-Loaded Non-Uniform

Members," J. Struct. Div. ASCE, V. 92, No. ST2, April 1966, pp.
195-205.

73. Wilde, P., "The Torsion of Thin-Walled Bars with Variable Cross- ^
Section," Archiwuia Meehaniki Stosowanei. April 20, 1968, p. 431.

74. Wright, W., "Critical Loads for Non-Uniform Pin-Ended Columns,"

Civil Engr. & Publ. Works Rev.. Feb. 1968, pp. 185-]87.
 Finimi: caption?

Fig. 2.1 Wcdi;e-Beam Framing Kxar.mlep m

2.2 Channel Section witli Linearly Tapered Web

(a) Locus of Shear Centers

(b) Load Anpiied at the Smaller End Shear Center

Parallel to the Web

(e) Load Applied at the Smaller End Shear Center

Perpendicular to the. Web

3.1 Tapered Beam Geometry Used for Analysis in Section 3

3.2 Loading Presume»! for Analysis in Section 3

3.3 Functions which satisfy the Boundary Conditions for

Use with Power Scries Solutions

3.4(a) Section Properties at the Smaller End

(b) Parameters Used to Examine Lateral Stability

(c) Limitinp End-Moment Ratios "k"

3.5 Lateral Buckling Solutions for Section T, ¿ = 14A

and a ~ +k

(a) Pure Betiding; e -», \ • m /p

cr -t ey o
(b) Bending plus Axial Thrust; e « d- , \ * p/p
° cr e
y
Definition of the Lev.gth Modification Factor "g11
for Columns

Values of Length Modification Factors "g”

f omparicon ru "Et-.act" Strong-Axis Bur.¡ ling Stress to

Equation 4.1-3. Values greater then Unity are Con¬
servative

4.4 Allowable Axial Compressive Stress

Dafirition of the Length Modification Factor "h"

for beams _
Reproduced Irom
best available copy.
A .6 Critical Elastic Stresses Due to Bending

a. y Comparison of Accuracy of Results for Lateral

Buckling Stresses of Beams

A. 8 Allowable Bonding Stress tor Thin, Deep Smaller

End Cross Sections c * 1.0
hY
A.9 Allowable Bending Stress for Thick, Shallow Smaller
End Cross Sections c, “1.0
b
Y
4.10 Values of Interaction Equation Term "CL"
fii\ fn )
4.11 Comparison of "Exact"'*^t given by Eq. 4.3-1.
Values Greater tlian Unity are Conservative

4.12 Comparison of Design Formulas to Colombia Test Results

[10]
4.13 Comparison of Design Formulas to Buffalo Test Results
[12]
r •» Stiacc.Ui.al HcJcls Uf.ud for Ll.e Determination of the
J » i.
Effective Length Factors of Tapered Columns in Frames

J, ¿ Sign System for the Slope-Deflection Equation of

Tapered Members

5.3 Stiffness Factors: Small End Fixsd

5.4 Stiffness Factors: Large End Fixed

5.5 Carry-Over Factors: Small End Fixed

5.6 Carry-Over Factors: Large End Fixed

5.7- Effective Length Factors for Tapered Columns: Side-

5.15 5ví*y Prevented - Y " 0, 0.5, 1.0, 1.5, 2.0, 3.0, 4.0,
5.0, and 6.0

5.16- Effective Length Factors for Tapered Columns: Side-

5.24 Sway Permitted - Y « 0, 0.5, 1.0, 1.5, 2.0, 3.0, 4.0,
5.0, and 6.0

6.1 Frame Example Reproduced from

best available copy.
6.2 Moment Diagram

5.3 Dimensions, of Members ior Final Design

r
 lo»:;'»!.■»ry Conditions
i(*), ííOO,
r. « 0 :í - t

Clov-rod Cl¿.uv*ed I S(7. - ;)3

Cl 1 2 (K ' /.)
fhipjíorCot! 1
C i i;.pl y
Suppori ;‘d 1 '¿(* " Jt)
r.Mpyori ".<1

C1 ; M|)w%d !‘rcc
.. . „ -

riß. .3.3 ► Kum limi; v;|>icl» vü I satisfy the brurJ.iry cou»! il ion» for
out with i»nr.'.j sfrif" »flot icr.n
 IT ffi TZ. Y
y*£2>?:2en

r.v r.r¿.xv.: rxssa C7^=l

\ntjcxr~ ti
I
rrxxjrza
H
600 in 6 00 in 600 in 600 in 1200 in
400 1200 400 1200 600
0-25 025 075 075 0.25
010 010 ..1.:0 25 0-25 010
 1^4 in. , 192 in. , 240 in.
+ k •k -1
1.0 ■5, 20 . 3.0
1.Ü 2.0, 5.0 . m

fiR. 3./, (c) - Limiting end - moment ratios, "k"

y Ij I ! jcctiop. II Section LIT Section IV

r~ |~* ——— i_!---
. ...
1 •
0 1.000 1.000 1.000 1.000
1.0 l n.4o0 0.485 0.468 0.488
j !•' fi 0.334 0.382 0.365 0.386
2-° 1 0.284 0.314 0.293 0.318
! 3*° I 0*198 0.2?9 0.2O8 0.233

3.4(1:) t,
 lac»*ral.-Torsional Buckling
3?ci ion I, l ** 14| inches

VAl'iJES CF \

??nro Bending
r-
V « Nufiber of Terms in Power Series

4
__L_| 6 7 8 9

0 +1.Ò 0.8C80 0.7066 0.7790 0.7760 0.7736 0.772C

40.5 1.0476 1.0261 1.0191 1.0152 i.cm 1.0108
1.3510
0 1.:799 1.3655 1.3604 1.3554 1.33 '7
•0 5 1.6967 1.6b/5 1.6812 1.6747 1.6/19 1.6701
-1,0 1.6022 1.5903 1.5754 1.5652 1.5606 1.5577

x.o 41. Ö 1.1831 ï ; 1557 1.1416 1.1337 1.1787 1.1253

40.5 1.7213 i.fca:u 1.6075 1.6580 1.6517 1.6474
0 2.779/ 2.7175 2.7 U58 2.6961 2,6896 2.6853
-0.5 3.7147 3.5236 3.5005 3.4799 3.4695 3.4634
-1,0 2.4694 2.2484 2.1953 2.1711 2.1587 2.1512

B-'.ndinn ?lus Axial 11.rust (c/dc) « 1 0

i"
{> 4-1.0 0.5211 0.5061 0.5013 0.4986 0.4969 0.4958
+0.5 0 6096 0.5934 0.5864 0.5854 0.5835 0. 5323
; 0.7110 U. 6540 0.6894 Ü.6859 0.6833 0.0824

2.0
! n°-¿
-0.
-1.0

+1.0
0.7954
0.7.326

0.6321
0.7769
0.7611

C.fiSi.
0.7706
0.73C3

0.6030
0.7664
0.7512

0.6040
0.7640
0.7485

0.6014
0.7 624
0.7467

0.5997
0.7118
HI 5 0.74cl 0.7286 0.7710 0.7166 0.7128
0 0.8900 4.3o36 0.6554 0.8504 0.3473 0.8451
f -0.3 0.9258 0.9469 0.9364 0.9303 O./W 0.92.40
1 •1.0 0 8952 0 8334 Ci. 8409 0.3341 V \
■» ;
- V 0.8274
L_ _ _ _
L,__

f,S
Fír. 4.2 - V.iiups cf U-nßth modificar.ion t'actorr. tor îîtrong-axl
buckling, "g"

y V
h-
C 1 .000 3.2 0.453
0.2 0.937 3.4 .447
0.4 .87¾ 3.C .436
o.e .&2b 3.9 .427
0.Ò .777 4.0 .418
.0 .733 4.2 .410

1.2 . u96 4.4 .402

Ú.4 • c36 4.6 394

1. C .623 4.3 386

1.8 .393 5.0 .377
2.0 .537 5.2 . 368
2.2 .543 5.4 .358
2.4 .522 5.6 .347

2. b .503 5.8 .334

2.8 .486 6.0 320
3. r ./♦71
 ^1

'•'iß. /i.*í - Coinp.ivJ**pjh*ct*' f.I rou}'. ,'i/. '•> fovckî in»; tu

ci;... ; i ru t, .) ■>. VMv: ; ßr.M.cr fl*-n uc-iv ..n1 vf.i««.i-.11 i.1".*
J
I-7-•>—T—7—T—t:-r——--','"'7 !
/ -ZU
1
f ■'
J .(>;

WÊÊIBSBOBSSSSSSSSSmSBSSÊmÊM
»' • •«
i i
})’}_!. ■'i. i f) - ti int«, l'ücl ioii ri>u;iL ji :i tivi;;ni f j c.U i on Irrr; "(! "

ReprodiKod from
^*’0 'W^iUblp topy.

7S
r

r> <* > . i C> <i U* O i,\ < ' (It I I > t'Os* :.
» i < • » ' C; c • c> O* <> (# 1^0 « * 1• » » l' i » «.* ; j t i o
r. >: • < • .: ) c > c* < > C • #1 -10.1 o ?»
_ ..L
i» ¿»
<> . < ■' f. I f-l
,« I C ' •' » i.' i (i\ tH 1.» . « o. r > r i ri t.» i » o-. . » .-i
ii ' » V. ' C ' l ' Î O f' t < > t«' i'*'O r> tu i t.) f» ^ • t « • « t> < > O
i O /J • <
o i t I .-I r I • I C- 4 O » Vi
o i O C ; O .• • . t O i > « < » • •■< O o o
f . « •• I I ^ <•! .1.10 .> ; » . i • t o r > <» r • «>«
t > I Ci . ^ ' V.' t.< v’ «. * ;> r» o «..> t' o > » r» <.» t' C."
o4
■ *> o r I » • ••! C> O • I • < • «

c> c> i> i •> o o c t\ » ' ■ o o o • r » C* » ■ • i f » t \ ri o o O

r» c.» c> »> o C.I « »•. o <■.; t' o V' i ’ » > ' » i.' . i.» c* i» cj o o
V>
f. t » l> I #-4 . 4
—ÄT.tr; . r.- -rrar*. i
f > O »•>! . * C » CA r*» f.% r.o ví «r> C*% l/“* I . O 4#í »fi
( ^ L' C.* 1 » O» i ) ^ • O O i > o O tn l J C * t «

!1
rx
O c I > ( --4. I » O <

I I ; ii • • I o » ^ f.% I-» c :i o . tj
i I • í V# O C\ V.’ C*' (.• o c» l » c C- Ci t-’’ o
'■V
o ^ rl n o •- 4 .4 r\ c» i C'
r
I r. C»
c.' i’’ « : o (O G% . • r*i f* I-«. r, r* r*: i^*
t i c. c.» C» O' Ci o o o l.* Ca i*» (. i r.\ u i o» t % <.% o < ^
• • •
• •I r-4 «-4 r*4 r> . i 9 A . -4 • 4 » ) a o • 4 o o ri
I *» U.
1^*4 Cl r> r; C' c- iCt c ♦ PN <0 f > . 4 (‘f
r # (. ci ( % to o o O r i 4J ;> o ( > o O O' O O < »

°l i o n •i <1 r O O

( ’ o o ri . r» r-, O c ¿ j r» o »VT ( I o I'* r*% ( • O . i

( > o <. <-> o c» íj t» í C' i * C ! c» i.* i • o c» Ci r.> o o

¡■°í
• •
. I • « i c¿
i •
(^ 1-4 . 4
.kfcVtt—*-.
• • • I • 4 -1 o
má*k~* " •#-
I
r> o ••<
. . . •-. wí-.
0
< # C' f í «*> (0 C-. »-i <*.* 4<i C' 4 • . • ri < # o
- » o O V' o (.. C» -J C* CA Í . Ci Ci ¡o ca c: í> c. E °
I rJ r • • o °
i Ci C» » * • 4.4 O Ci < I o O • «
- r' -1
!. ^ i .> o < í C V a
I «f»!' i . > • r#» e» ! a* o # 4 •• • r 4 I
I o o o
•
I
» •
*4 j
c/
t o
O' < -
r: • > c?
'.N I :*• o C • í *• O C' i* <'
s?
O : ! '< J o ^
•( o 3 1 y I ; c*. »r» r» #> • 4 'i .» c % t f\! l" » (•! tO . • f'j i u d
u < ■> C.» O O i. * (i ! i% í.» o .• f> «.» t. »*« C* i • *» «I

¡o I
u »
:3
O t. . r ? • r. o
o
f -■
o o r.
I i \ r. . »
I c) o
* {»• c*. • í <'#•
d O C:
i• Ci C : f'i
.J « X)

• o. Ci O c c; V C- O <* o &t o «.i C3 C' V % O O o

1 I 4-ii!
5 ¡I i • •
O •-< C- O 0.4 *•< r»4 I I -4 r-« i o o >
-4; .i
li • > <•> O o o o /** ■*■ ; i i n* O fj O o s r* ' i c » f. » • r> o
:■> o 4 . O VI (i O r.' : •.% < » t • O > ! r- (•> i. ; C'
oj i
• ‘ .•< r 4 t-4 • f -4 ..4 ,-| ■:i‘> o
■*—: i3» * TtX ITH ■. rt
rr ‘
.
--i r- *
i c» t* fV -i
Cj C» n -t <■• o f . >ii o tV • ' i t>"> t . o : O
c:‘Í f i v«i i r» o i. w ■ V* * O • V» ' C/ * ■•< o t>
i i.» Z: O c-
Ci il
I O -i o • • . « t - t-^ r »
I 1^
».1 O. r.. í.» "i <.» «.• 'io »r. 11-.11 o . » i i .*
I. *• JI1 Cí t • % fl > r.. i . 11 t* t. C • C -■* • c» c V f> c
j •
i r t. •! . -1 r i j •-« »..» O C» t» r> « .-i t r> o
: c^ «i
r>. -, r. * O ; > . . r. o c. i r• 1 o o o • * í'' • i». r \ 4 ( : c*i
4J V- o I O C» ü O c> ti. c« ¢: ! o. c- V- I < • C*. O c> o
;»
'i * ‘jt . I (1 r < o * O ¡ o o
! o c;
... ■.
./.i, < . t . *» O* c \ • • 4 .-4 ! tv t. . I .< ».I < » o c* f ^ ’
ri C-. C*» C. • c*. O 4.. c * C> o O lo V* C.» 4 » O

' o O o • 1 < • rl c ri
“11 %' » »I C I C i I Ci • i r# r * i o r. < • tii' tn r ' > c--1*. <* l • f>
v J t .* t ■ t> 5 .i ci I
i . » o - >
I ' I '•» C i 1 c . >. t - C« o
o c: rí • •

«..I .•• I ! I < ! f.i • 4 • •! *-4 • -4 o f . 4

^ - * * ■ 1 A :• - i •.**»• if !.*si.-wtr i ■ ' • 4 “W •**
i;
CO O I f'• »/• r . t \i» cv c* I c» o O I « 7 >.•> *0
•*' ... I • • ! t> . i • i O • •
/..1-./ * j* c > » ■• O o ; c. o o C> t - O «i
I Ci I*."' c » c.» . • i o
—* I i I
1

i
t

••

i; o n o O
• i • i
1
i

1
,

■ i C; / • <i
• I • f
•
# 4
I”:
-t I j
-•«i i'

J
 1

Lé....J
 »Hb"
S >

<
\
cU*
I
t'^V-

~~T~ T
Ispecitn^ri Wq_j point F.q, 4.3-¾ | _d<,_ _d/ ,|_'
6 0“ 60 !!6 f;
LÖ-C-7 C* /«
Ol" 110 i I5 *V
i 0-C-9 ui
I* "
6- i :i j 1140'
LB-C-IC

Other data; a=o

b=GO

¿C

h
(a)No sides way (b) With sidesway
'Z'Ji

L
I
o-

J
<» .
./ .
¿4 J » JULpI 'jAr't < u-uyiu'i.iui«<í)lâj
üb 1-5 2 3 4 5ü Ö10 20 5000
Rt

<>• ^
./
ÍM

s’- ff
<>G
P
 b.w>

05- odl
04L* juw-Iu-ujJ. ujjlIm.' a1j ajj.UuLjlUU.Ü LU.«J^UÎlS
0-5 1 1õ 2 3 4 5 6 810 20 50CJO
r

04r^ J.mLJL' UaiJLkÜ

0 05 1 15 2 3 4 56 810 2050oD
Rt
íííí

s.
 (Y)

1B8
30 F? o blo
20 h
Rr*= 4^-°-
10
8b
7
6
5
4
K
S' 3

15

--o
0-5 T- p c|
Q,,4i * i i ! i ..t.-i-iJu.LD-j.iLLi.Li.i.11V ULLs
0 OG 1 15 2 3 4 5 6 810 2050od
Rt
'Z * '•
X '

aMHHAMHiaftlfllÉ
O',
Q 5, y -
0-Ai/ Q
OsL-o- ui-jJL 1J-MI.UJ >.I I.' P :Ulai.'iiiUi! U./,!M
O 0-5 1 15 2 3 4 5 67010 20 5003
Rt
\ : * ">
.1

S. Z?-
 COr Mrrwa

100-
ÖOi-
30'i:-
20-
d-JEl!?
Kb LIb
10h R j°
C>r Kt~LTt
jr*
p-iÆ4-
^r(K7D2
5f
K,

16

or.
O*'*. i. •< $8'
03» -»-JU-o.. .Ü4.U!,' uxl IIV UUJ, 'v1
0 Ob 1 1Ö 2 3 4 £>67810 205000
Rt
V :5
Fis. fe.I - Kráw* frXrtJttplc

40Ó K» U

Mciiiber «V o. *•> 6

AB 12.4 24 t4 0.50 1?1

BC -—30 Vf ne —— -
CO 12.4 24 6 0.50 12.1 075

tB 6.0 24.0 0.42 7.0 0.60

rif’.. 6.3 - PlwrnslrmK «f infabern for fin.il

v.r*
 APPENDIX

This appendix has been Included to provide the designer with a

quick reference to the proposed tapercd-be.rui design fomulns. Since tho

development of the formulas closely follows that ot prismatic beams, this

section paraphrases the appropriate A.I.S.C. provisions.

The following Is a working definition of a tapered member that Is

within the scope of the proposed formulas:

A tepered member Is a member possessing at lea"t

one axis of symmetry, e.g., I-shapes and [-shapes,

which axis of symmetry corresponds to the flexural

axis If moments arc present, where the flanges are

of equal and constant area, and where only the depth

of the member varies In a linear fashion according

to the equation

«*) ■ ^(i ♦vi >

where y la the tapering ratio and Is equal to

(dt-d#)/de.

Section 1.5 ALLOWABLE STRESS

Compression

1.5.1.3.1 On the grosa section of axially loaded tapered compression

members when Kyt/r0, the largest effective slenderness rr.tio of any

unbraced segment, Is less than Cc

tf < r
*►'(
 r (1.5-1)
•v p.s.

1.5.1.3.2 On the groes section of axially loaded tapered columns

"h®0 Kyl/r0 exceed Cc

F . —Hrit_
^ 23 (lV7re)® (1.5-2)

(The allowable axial stresses are tabulated for Fy •• 36, 42,.5Q( 65, snd

100 ksl In tables Al. The effective length factor Ky can be determined

from Figs. 5.7-5.24).

1.5.1.4. Bendlnx

1.5.1.4.6a Compression on extreme fibers of tapered flexural members,

e

having an axis of symmetry In, and loaded In, the plane their web: the

larger value computed by formulas (1.5*6a) or (l.S-6b) and (1.5-7) as appli¬

cable, but not more than 0.60 Fy.

When htfrm < C_

w \ r
(1.5-6«)

When hJ./tr * C.
170,000 Cby
(1.5-6b)
Or, when the cctnpreseion flange is aolld and approximately rectangular

In cross-aectlon and Ita area la not leas than that of tho tenalon flange

nooo cbv
(1.5-7)
hjji-dg / Aj

In the foregoing,

I - distance between crosa-sections braced against twist or lateral

displacement of the compression flange

r_ * radius of gyration of the smaller end cross-section comprising

o the compression flange plus one-third of the compression web
area, taken about an axis in the plane of the web

Af ■ area of the coopresaion flange

i.O , for * ■ ♦ k

{ 1.75
for or
Í.Ò ♦ 0.¿5 {y
where or ■ and k la defined In equation 3.1.18

N. - 1.0 ♦ 0.00385 yf/r"

ha • 1.0 -f 0.0230 y

. i M0.000 cj

(The allowable bending stresses are tabulated for ty ■ 36, 42, 50,

65, and 100 ksl and for or * 0 and In tables AS through A13. Tables A2,

A3, and A4 contain values for functions hy, U(, and respectively. The

heavy solid line in tables A5, A7-A13 indicates the division between for¬

mulât (1.5-6n) and (1.5-6b). Thus the values below this line, i.e.,

formula (1.5-6b),can be converted to tho « « 0 case by s'épie multiplica¬

tions. by Cfey. The same conversion applies to table A6, formulas (1.5-7).

o
This sanplfc C'Mwtrr.íun cirmc»; t<.‘ to '»aim i t.hî ue-ivy lí. 'e,

thus tables A7-A!Vu»ve b«ier> provide*!.

11i~<*i 1 6 coMBrfFp s;i\:-sshs

l.C.i Axial Coripi¿viisn ana Beading

rrpr-cü piwrib«*ih tiuWj*.»rteH ».o Ijrh «x'äI »•OTprer.aio»'. «ud bendlaf,

•:í".*eesc-í lail ‘.e pi,opotc:. >ne'‘ u. j*et>..!ty fo'lwlrg rcquivffeew :

fix
f * ) 4. B i-Ü-.' < 1.0 vi. ■■»•la)
Vfrü\ « ll-i-'i \
‘ 1 -ÜL . i
'e. /.
*bV •
K
a-d
:1.6-1-.)
A. * (ï; . ,,,
C.M Tv'*

«bcn (£j /a\|.v»0 £ >' • ? i. forewlas (t.* iusy le vt.'.l n Ue:

•ovr-u1*.» 'T.r-ía) and (l.(-iL)

•^a N + f , 1,,
{—) i? >•.’)
%v'° K V '.

w >.«

• ut''cí»c- nt ;».? t*Ta 'i¿r and i'*'.*»:, ws*. • d bi '.oii l..-¿1

axi.il fe re-. íleo ; i*v.lrt'j«i

* Cimirr.*“)! .Mr beid>.ij; • tveiv a*" o *d U.ut m*

I• 'nilítd il bandui}. ViOAiCnir. aiviiu fiit-te!
1 *'eW V
- .-I... t vit'sic Ajjís thi 'iftia1 cnl'V.'.e*'.'
IfcRf.l!’ la ihe p’ . ii if i* »¿ and
ti tí.i .'•orrnepí/.vU'.f «Mllcr
c er.d rea lit.) o< ^yrat i sr.

u remp ved i mol 1er rr.n il f.tr». ’'**

• coBi'i ifid iär>;et cr.c ten«’! t, .»Cies«

and
(l.o*0.i (^) *»•>
J \ *>

R
li.o-c.» *«•» (1^’î
'i'"/
\»* t
ft S
V

rt,,tc cr. H,/H '•. fi J-, M*

(ï,.,,u ; „t ,K- M.W.1 of .ayjL&R».B2Sg«!» >>->»**•' "*

cu, b, ..«o for - ffW^ fi f.., «„'«„•>•

in
 TiiMv A?
).011),1 !i M uli f !i 11 nu I', vltu , li
w

[ .• ) •• i 3 'i 9
"1

I . !

! 1
I ,
* > I
« ! ,

I .
\ , •
1 !
> , . 1 . t / '
I *
. I - I V t I
1 1
) . t I I 1.

• . • I . ) ••

I . I . U ’ Iu > I
•. ' I •"» 1 ' i • t ^ < I
t .
I . I ■ • ./ J i .• ■I *
» . , i
} I V. , 1 , w ».'• ï .
1 . 1. ) / ) . / \ »
1 •’
t . t I . r; • I %
1
I . 1 ' . I ) . i- / '» . ■ 1
I . * • . 1 ' ' I . i
I . * I. » . / 7 -■
.tTi '\t I 1. » ‘f » . / /■

I *< i ' 1. • i 1 . ' . I I . / '• >

1 . ' ’ .
• . I « .
] . t
1 . • '
i . ’

I . '• ) >

i .

_ Jj.
R*produc«d **om
be ti eviUbU copYj^jy

lia

J
*1 !' I l! :• t . , ]',(!:.) i n‘i:: n. •!. ■ ( ‘ )

end (t -(.i,) IV: i;

(Reproduced
Jbet* eviUbl« coPYi.

tro
.11.
!'■. I • ( I 7) ii

■ < , ' •

t <i 'i , / 1 0! ». M
HO* 00 • - t ••

. -1 ■1. i '1 ’ > t \:

* • #

►» y /
A■1 . O 1 • •

. k •/ / * » > • . 11
i,. ,1 • . O
. I . -. . oil

• i I k

. i ’/ ■} HO • . / ¡

/0.0' / dO'.'
i,' ï tv-..', f, -,-)0 ...--

(’.i\ 1/ . 1 A / 70iJ .
rn 11 . / C-- 4, -Ok

» i.n i h .(:
K1: 1 1-.1/ .- ■ • A * •’ 1

Vl'O ? ‘0 0 . 1 ■

•r i* I . t’ * / o-, -..0/

i (S>. 1 / .Of -0)00 A . (IO

! f ' 1 ! . * -»OHO ■

1 1 . 1').‘ : M to *, -W

n • t.A ■ ;-<l oo . !

i 10.0 i
0. «.'I -k/Ofl ' . AO
1 /*‘0
1 V ' o . / -- s -UtO

] M> Í* . ll '- tHoo Î. Of

I ' (H* » .0 . *,. o, ’ ^,0-(

¡/.H») I . -(.0-

) H''-'» H.<. . ■ '•.

‘ /. ./

.- - .,.
/ . I *, ----( ./.
I t" '
-V . /*
( Y* I . (

. 1

J t M f • , '• 4,13?

’ . ,. - . ■

I •<.« f«. ! •OVO ¡I ^ . ft**

,’t ' <* I ♦ ' » -k.. >

1 T. r

" *..
NHHHI.

ÏVîht

t 1 I < '•. 11 » « * ! » i«:'' : i

i i.

\ V t ,

/ *

Rtproductd from
btil a»«il«bU copy.

l\h
Reproduced from
best available «py.
9
Ä;

Á
Reproduced (rom
be»t available copy. 9 iy.:<
I