I ~

RRll71

7305
 ''l

I
I
I
I
I
I AN ALThRNATI VI:;

I By Dann H. Hall' and A. Rlchard LaWln'

I
I
I Far American Institute Of Steel CalstructlOO

I
I
I
July '985
I
I
I
I , Bri<i3e SOftware Del/elopnent International Ltd.
Coo[ersl:urg, Pennsy ll/ania
I
I I':0
'~"
, -i-
'T)

I TABLE OF CINl'ENl'S

I Page

1 . 0 SlMlARY 1
I 1 .1 Introduction
1.2 Problem statement
I 1.3
1. 4
Objective
AWroach
1.5 Alternate ~thod

I 2 . 0 INTOOOOCl'IGl 5

I 2 . 1 Present Practice
2.2 Problems With Present Practice
2.3 Proposed Alternate ~thod
I 2 . 3 . 1 General
2.3.2 Erection of the Arch Ribs
2.3 . 3 Erection of Deck Units
I 2.3 .4 Jldvantages of Alternate ~thod

3. 0 DESIQI EXAMPLE BY ALTffiNATE ME:llICD 17

I 3.1 General
3 . 2 Design Considerations
I 3.2 . 1 Live Load
3 . 2 . 2 Dead Wad
3.2 . 3 Construction Loads
I 3.2 . 4 'lhermal Wad
3. 2 . 5 Post Tensioning

I 4 . 0 Analysis Using Finite Element Models 25

4.1 General
I 4.2 Floor Beam
4 . 3 Placement of Deck Units
4 . 3.1 Arch Rib
I 4 . 3 . 2 Hangers and Dead Wad Tie Cable
4. 3 . 3 Precast Deck Units
4 . 3 . 4 Bearings
I 4. 3.5 Procedure

4 . 4 Integrated Model
I 4 . 4. 1 General
4. 4. 2 Tie Beam
4.4.3 Deck
I 4.4.4 Floor Beams
4 . 4 . 5 Procedure

I
 <:>
I~ -.J -ii-

I Page

5 . 0 RESULTS 38
I 5. 1 General
5 . 2 Floor Beam
I 5 . 3 Arch Ribs
5 . 3 . 1 General
5.3 . 2 Deck Unit Staging
I 5 . 3 . 3 Superimposed Dead Load
5 . 3 . 4 Temperature
5 . 3. 5 Live Load
I 5 . 4 HarY:Jers
5 . 4. 1 General
I 5 . 4 . 2 Deck Unit staging
5. 4. 3 Superimposed Dead Load
5 . 4. 4 Temperature
I 5 . 4. 5 Live Load

5 . 5 Dead Load Tie Cable and Reactions

I 5 . 5 . 1 General
5. 5 . 2 Deck Unit staging
5. 5 . 3 Superimposed Dead Load
I 5 . 5 . 4 Temperature
5. 5.5 Live Load

I 5. 6 Tie Beams
5 . 6 . 1 General
5 . 6 . 2 Deck Unit Staging
I 5 . 6 . 3 Superimposed Dead Load
5 . 6 . 4 Temperature
5 . 6 . 5 Live Load
I 5 . 7 Deck
5. 7. 1 General
I 5. 7 . 2
5. 7. 3
Deck Unit Staging
Superimposed Dead Load
5. 7. 4 Temperature
I 5. 7. 5 Live Load

I 5 . 8 Post Tensioning
5 . 8 . 1 General
5 . 8 . 2 Deck
5.8 . 3 Tie Beam
I 6 . 0 CIN:llJSlOOS 86

I 7 . 0~ 88

8 . 0 BIBLI(X;RAPHY 89
I
I ,~
'p
'n -iii-

I LIST OF FIQJRES

I FIGURE TITLE

2.1 Typical Tied Arch Brici3e

I 2. 2 Alternate Design - Deck am Tie Beam Unit

I 2. 3MB Alternate Design - Arch Erection Schemes

2.4 Dead Load Tie cable Anchorage

I 2.5 Erection Schane for Deck Units

I 2.6 Deck Unit layout

2.7 Arch Rib to Tie Beam Detail

I 3. 1 Design Study Example

I 3. 2 Design Study Arch Rib Cross Section

3. 3 Design Study Deck Unit

I 3. 4 Design Study Deck Cross section

I 4. 1 Floor Beam am Deck /ot:ldel

4. 2 Arch Rib am Staging /ot:ldel

I 4.3MB Staging sequence - Stage 2 - Stage 3

I 4. 4 Integrated /ot:ldel - Overview

4. 5 Integrated /ot:ldel - Details

I 5.1 Floor Beam M:rnent Envelope

I 5. 2 Floor Beam Shear Envelope

5. 3 Floor Beam Elevation

I 5. 4 Floor Beam Influence Line

I 5.5 Accumulated M:rnents in Arch Rib D.le to

Deck Units
5. 6 Accumulated 'll1rusts in Arch Rib D.le to
I Deck Units

I
 - iv-

I LISI' OF FIGJRES (O:xltinued)

I FIGURE TITLE

5. 7 Stress Envelopes in Arch Rib DJe to

I Deck Units

5. 8 Dead Load ~ts in Tie Beam

I 5. 9 Locatioo of Recorded Deck stresses

I 5 . 10 stresses in top of Deck DJe to Superimp:>sed Dead

Load am 'l'e!1;lera ture - Design 1

I 5 . 11 Stresses in top of Deck DJe to Superimp:>sed Dead

Load am 'l'err{lerature - Design 2

I 5 . 12 Stresses in bottan of Deck DJe to Superimposed Dead

Load am 'l'e!1;lerature - DeSign 2

I
I
I
I
I
I
I
I
I
I
I
 ~

I '~'.(1
.", -v-
~

I LIST CF TABUS

I TABLE TITLE

I 3. 1 Design Example CCInparison

4. 1 M:ldel Explanation for Fig 5

I 5.1 Dead Load M:Inents in Arch Rib lAJe to Deck Units

I 5. 2 Dead Load Thrusts in Arch Rib lAJe to Deck Units

5. 3 Accunulated M:Inents in Arch Rib !Ale to Deck Units

I 5. 4 AccuIlulated Thrusts in Arch Rib lAJe to Deck Units

I 5. 5 Accunulated Stresses in Arch Rib !Ale to Deck Units

5. 6 SUmnary of Arch Rib /obrents

I 5. 7 S\mnary of Arch Rib Thrusts

I 5.8 Factor ed /obrents in Arch Rib

5. 9 Factored Thrusts in Arch Rib

I 5. 10 SUmnary of stresses in Arch Rib

I 5 . 11 Hanger Forces for Dead Load

5 . 12 AccuIlulated Forces in Hangers for Dead Load

I 5 . 13 SUmnary of Forces in Hangers

I 5 . 14 Reaction and Tie cable Forces

5 . 15 Tie Beam stresses at Hanger - Design 1

I 5 . 16 Tie Beam Stresses at Hanger - Design 2

I 5. 17 Tie Beam stresses Between Hangers - Design 2

5 . 18 Deck Stresses Top Line "A"

I
I
I
I ·~
... - vi-

I LIST OF TABLES (Ccntirued)

I TITLE

5. 19 Deck stresses Botton Line "A"

I 5. 20 Deck stresses Top Line "B"

I 5. 21 Deck Stresses Botton Line "B"

5. 22 Deck stresses Top Line "C"

I 5. 23 Deck stresses Botton Line "C"

I
I
I
I
I
I
I
I
I
I
I
I •

I
I']).J)

'" 1
.....
1 1. 0 Stmnary

I 1 . 1 Introduction

'!he tied arch brici:Je is canposed of an arch rib 00 each side

I of the roadway , a tie beam associated with each arch rib which
takes the thrust fran the arches arrl a deck system supported
by the t ie beams . The deck system is most <X1i11ccl.y canposed
I of a concrete deck supported by lo~itudinal str~ers in turn
supported by transverse floor beams . Cable ~ers cconected
between the arch ribs arrl the tie beams transfer the vertical
I loads fran the tie beams to the arch ribs . Thus traffic passes
between the arches at the lowest elevation of the arch ribs .
Usually the arches are parabolic arrl braced overhead for
I stability.

Thrust fran the arch ribs is resisted by the tie beams .

1 '!he deck system is isolated fran the tie beams to insure that
tensile stresses are not introduced into the deck when the tensioo
in the tie beam increases. This is dooe by segmen~ the deck
I using stress relief joints. Lateral loads are carried by a brac~
system which works with the tie beams . It is <X1i1'O' to construct
brac~ at both top and bottan flange levels of the floor beams.
I The arch ribs principally resist thrust ; rut bending <X1iip:uents
can be rather large. If the arch ribs are loaded evenly, the
I berrling is minimized . Berrling in the arch rib is reduced for
a given concentrated load if stiffness of the tie beam is increased.
This i s most easily envisioned by thinking of the tie beam as
I a beam on elastic foundations , i.e. the ~ers and arch ribs
act as the foundatioos . The stiffer the tie beam , the !lOre evenly

I distriruted is the force in the ~ers . This leads to !lOre

even loadi~ of the arch ribs . Since the deck has been structurally
i solated fran the tie beams , neither it nor the lo~itudinal

I stringers cootriOOte stiffness to the tie beams .

'!he structures are ruilt fran falsework . Spans range between

I 200 arrl 1000 feet for this type of brici:Je . They are used where
single spans are required . If cootinuity fran adjacent spans
is available , tied arches are at a disadvantage canpared to cooti-

I nuous trusses , cable stayed or even girder brici:Jes.

I
I
I
 2
I
1 .2 Problan Statement I
Tied arch bridJes have becane less p::lpular because the tie
beams are cx:nsidered n::lI'l-redurrlant. AASID'O BridJe Specifications
define a n::lI'l-redundant member as a tension member which, if it
I
fails, is likely to lead to collapse of the structure. Although
few, if any, tied arch bridJes have actually failed, the tie
girders have suffered cracks in one or two instances (Ref,2).
I
It is clear that failure of such a member could be catastrophic.
Further, the cost of tied arch bridJes is high when canpared
to more modem bridges such as the cable stayed bridge and the
I
segmental cx:ncrete box girder bridJe.

There are several reason assigned to the high cost:

I
o 'lbere are teo many parts in tied arches; I
o Field labor is expensive;

o '!be deck does not work with other CXl1ifXll,ents in

I
the bridJe;

o '!be stress relief joints in the deck are

I
expensive;

o Ncn-redundant members are defined as fracture

I
cri tical and must be designed and manufactured to
more stringent requiranents; I
o Falsework is often not needed on other types
of bridJes. I
1.3 Objective I
'!be objective of this study is to examine other p:lssible
means of constructing a tied arch bridge using modem techniques
that WOlld reduce or eliminate the undesirable 11Cl1-redundant
I
members, nake the structure less expensive to cx:nstruct; nake
more of the CXl1ifXll,ents work efficiently, eliminate as many of
the pieces of the structure as possible and to reduce the amount
I
of field labor, particularly, the elimination of falsework.

Several enabling technologies have been developed over the

I
last decade that are believed to meet the above objectives.
First was the advent of inexpensive high speed electrcnic I
cx:mputers which pennit the examination of structural behavior
in detail that was not eccnanical in the past.
I
I
 3

I Secorxl , the developroent of additives am improved techniques

permit the routine manufacture of high strerl3th ooncrete. '!here
I have been bridges Wilt using ooncrete strerqths specified as
high as 8000 psi (Ref. 3 ) •

I '!he anphasis of this study is <Xl the use of the electronic

CCII1p.lter technology by applying a bridge analysis am design
canputer program to the study . High strer¥jth ooncrete is employed
I in the design rut the technology is not examined in detail.

1 . 4 Approach
I An alternate scheme for the oonstruction of a tied arch bridge
has been developed . A design study was then perforrred which
I was based <Xl an existiN3 design and the resulting design cxxnpared
to the original. The analysis was perforrred using a series of
canputer programs called the BR.IIX;E-SYSTElo1sm developed by Bridge
I Software Developroent International , Ltd . '!he canputer generated
rrodel of the tied arch had to be nodified by harxl .

I '!he BRTIXiE-SYSTElo1sm is based <Xl the finite element method

of analysis . It permits the designer to ruild , analyze and design
large ccrnplex steel girder bridges efficiently.
I
1.5 Alternate t-Ethod
I In the proposed scheme , the arch ribs are erected first .
'lhey may be erected usiN3 a high-line or each half of the arch
I rib may be rotated into place fran it bearing . ~ing the erection ,
thrust IlU.lst be taken by the arutments or by a temporary cable

I between the ems of the ribs .

After the arch ribs are erected , permanent cables are placed

I between the ems of arch ribs to carry dead load thrust. '!he
cables must be supported by the haN3ers to prevent sagging am
reductioo of the effective m:xfulus .

I The deck and tie beam are precast ooncrete units . Each unit
of the deck exterxls full width of the bridge . 'lhe deck is cast

I intergrally with the tie beams . Each unit is equal in lBN3th

to the haN3er spaciN3 . '!he deck is supported <Xl CXlI1(XJSite steel
transverse floor beams which frame into the tie beams . '!he units

I are floated urrler the bridge am lifted into place by hoists

oonnected to the arch ribs .

I
I
 4
I
I
When the units are in place and cast-in-place ooncrete c::att>letes
the closure in the center, the units are post tensiooed. Finally,
the ends of the deck are cast-in-place and the deck is post tensioned
I
to the ends of the arch ribs.

'Ihe concrete tie beams resist cnly thrust fran the applied
I
live loads and superimposed dead loads while the thrust fran
the dead load of the arch and deck system is resisted by the
dead load tie cable. Thus the tie beam is ncn-reduroant.
I
'!he deck is an integral unit after post tensioniD:} and the
joints within the span have been eliminated. nus permits the
I
deck to carry lateral loads tD the ahJtments. '!here is no need
for lateral bracing in cases where the deck is wide enough tD
resist these loads.
I
'Ihe entire structure may be erected without falsework. 'Ibis
should speed constructicn and greatly simplify scheduling ccnstruc-
I
tion and obtainiD:} permits to obstruct channels'-

A large anount of the field labor has been eliminated rut

I
the contractDr still has the work of Wilding the deck units.
However, this work is off-site and not subject tD weather and
other undesirable features of field work.
I
Structural steel weight is reduoed by approximately 2.6 millicn
pounds. Atout 350,000 pamds of post tensioning steel is used.
I
The dead load tie cables weigh about 300,000 pourds. Han:.lers
remains unchanged. I
I
I
I
I
I
•
I
I
 - ---------

.p

I :~ 5

I 2.0 INl'RCUJcrICN

2.1 Present Practice

steel tied arch bridges are used for ll'Oderate spans ranging
fran 200 to 1000 feet. Figure 2.1, which was produced try cx:rnp.1ter
graphics, shows a typical tied arch bridge. 'nle tied arch is
roost often used when a single span is needed or where the adjacent
I spans are so short that they would provide little benefit fran
exntinuity. They are oonsidered try many to be aesthetically
pleasirJj because functional lines are evident to even the roost
I casual observer. The tied arch also provides maximum clearance,
therefore, awroaches may be reduced to a minimum.

I 'Ibe structure is cx:mposed of the arch rib, tie beams and

a deck system. 'nle arch rib usually has no hirJjes and tie beams
are usually rigidly connected to the errls of the arch ribs.
I 'nle rib itself is a welded box sectioo although truss type arch
ribs have been also used. The ribs are erected in sections try
field bolting.
II
I
I Arch Rib

I
Hanoer
I
I Tie Beam

I
I
I TYPICAL TIED ARCH BRID<lE

I
I· Fi 0 2.1
 6
I
'llu:ust in the arch is resisted by a tie beam connecting the
I
eros of the arch ribs. 'nle tie beam is also erected in secticns
by field I::x:lltir¥3 and can either be an I- or I::x:lx-shaped member.
'llle two tie beam !IlE!I1tlers are oonnect.ed with a full rrarent axmection
I
to the arch rib and they are subjected to I::x:lth tensile and flexural
loads. I
Hanger cables are used to suspend the tie beams fran the
arch ribs. It is through these han:Jers that the arch receives
vertical loads. 'nle hangers are CCIlIlDnly spaced at aI::olt 40
I
feet. 'llle shape of the arch rib has been develcped for uniform
vertical loadirv:J; and a parabola can be shown to be the IlOSt
efficient shape.
I
Tie beams connecting the eros of the arch ribs support a
series of transverse flcor beams. Floor beams are spaced so
I
they fall at hanger locations to minimize herDing in the tie
beams. Floor beams frame into the tie beams at their tql.
Diagonal bracir¥3 resists transverse loads on the structure.
I
Floor beams support a series of IOr¥3ibrlinal strir¥3ers. 'nle
strir¥3ers rest on either fixed or sliding bearings which isolate
I
the floor beams fran lTOVerrent of the deck. Because the tie beam
must be rigid in the longitudinal direction to resist tension,
the deck lIUst act separately i f it is to be prevented fran
I
develqling tensile stresses. la¥Jitudinal strains occur fran
live loads and thermal loads. 'nle joints in the deck are called
stress relief joints.
I
'nle IlOSt CCIlIlDn method of construction of tied arches is
to place falsework in the span to support the tie beam and arch
I
ribs during erection. 'nle falsework nay be kept in place durirv:!
erection of the deck system and castir¥3 of the deck. 'Ibis is I
done to minimize unsyrrmetrical loads durirv:! construction which
might cause overstresses in the arch ribs.
I
2.2 Problems With Present Practice

Sane engineers have expressed concern aI::x:lut the tied arch

I
bridge described. 'nle tie members are critical to the safety
of the structure. If the tie member should fail for any reascn,
the structure would be likely to collapse. In the days of riveted
I
construction, tie beams were canposed of several thinner plates
which provided redundancy. Presently, they are oanposed of three I
or farr plates welded into a single member. If a crack should
be initiated in the member, it is possible for it to propagate
through the entire tie !IlE!I1tler. I
I
 - -- - --- - - -- ---,

I '!he MSH'ro Brid:le Specifications (Ref.5 ) defines tension manbers

that can cause collapse as Fracture Critical Members (F'Cl>!s ).

I '!he Specification requires them to be designed to lower fatigue

stresses . '!he material is subjected to lI'Ore stringent toughness
and inspection requirements duri1'XJ' fabrication . '1hese cautionary

I measures increase the cost of the brid:le rut the structure i s

still non-reduOOant . Sane designers choose not to design non-
reduOOant structures if they can be avoided.

I Another undesirable characteristic is the many stre ss

relief joints in the deck. 'll1ese joints:

I o Present problems for maintenance ;

I o Introduce a rough riding surface;

o Add significantly to the initial cost.

I
'!he steel weight of the tied arch brid:le is not significantly

I lower than other types of brid:le construction am fabrication

am erection costs are alI'01'XJ' the highest.

I 'll1e need for falsework increases the cost of construction

by not only adding cost directly rut also by increasing the time
to bJild the bri~e . Falsework may provide an obstructi on in
I shiWi1'XJ' channels .

If the tied arch bridge is to remain a viable option, ways

I must be found to make it more canpetitive.

I 2 . 3 Proposed Alternate Method

2 . 3 . 1 General
I '!he proposed method involves the use of precast deck units
am a dead load tie manber connect ing the eros of the arch ribs
I acting as the tie beam for dead loads . It also assists the tie
beam in carrying live am superiroposed dead loads . '!he arch
ribs remain similar in awearance to conventional tied arch
I designs . Calstruction differs significantly fran normal in that
the alternate structure may be erected without falsewark . '!he
deck is made of precast concrete units . '!hese units include
I segnents of the two tie beams cast integrally with the deck as
sOOwn in Figure 2 . 2 . '!he units are post tensicoed to overCOlle
tensile stresses in them due to thrust fran the arch ribs am
I local effects .

I
 8 I
Reik and Hansel describe a similar bric'kJe in Gennany. 'Ihe I
Gennan structure utilized a p:>St tensiC11ed cast in place deck .
'Ihe tie beams were steel (Ref . 4) .
I
Each precast deck unit is equal in length to the hanger spacing .
'Ihe deck units are hoisted into place using the hangers and other
stabilizing lines as shown in Figure 2. 3 . Transverse floor beams I
are spaced to permit the deck to span across the floor beams
withaJt longitudinal stringers . 'lllree floor beams per precast
unit is usually sufficient. Floor beams are CUllfXJSite with the I
deck for all dead and live loads . steel has been used for the
floor beams to minimize weight and to reduce forming . 'Ibere
are no dia~<:JllS between floor beams . 1hls reduces ooncern I
for secxn::lary web bending which may cause fatigue problems .
'Ihe deck is able to resist lateral loads since it is an integral
element without stress relief joints. 'Ihere shwld be no need I
for a lateral bracing system.

'Ihe deck units are to be match cast. A standard type of

I
shear key 00 the interfaces will provide for shear transfer.
I
2.3 . 2 Erectioo of the Arch Ribs

'Ihe alternate method is best suited to sites where the span

I
is over navigable water. A small cable is stretched across the
span to take the thrust of the arch under its own weight . 'Ihe
Ernst Equatioo or sane other method of det:ermini.ng equivalent
I
stiffness for a sagging cable can be used to ccrnpute the cable
trodulus . It is necessary to prestress the cable to obtain a
reasooable rrodulus .
I
'Ihe arch can be erected fran each pier using a high-line
as shown in Figure 2. 3A. 'Ihe center sectioo of the arch can
I
be erected fran the high-line as shown. Alternatively , each
half arch rib can be rotated into place and spliced at the
center as shown in Figure 2 . 3B. At least ooe concrete arch has
I
been successfully I::uilt in Gennany using a similar technique .
(Ref.1 ) . Wind bracing rrust be added between the arch ribs in
either case prior to placing the deck units 00 the structure.
I'
Q1ce the arch ribs are erected, the initial rranents and
I
thrusts in the arch can be adjusted by tensiooing the temporary
cable. 'Ihe temporary cables are replaced by permanent dead load
tie cables .
I
I
I
I •~ 9
~.

I
I
- ~ __ HQngers

I + ... + ~

I ...
Rolled cr Welded Steel Floor Beams
I
I
t
I Concrete Tie Beam
Typ. Post Tensioning Ducts
I
I Match
face _ Hangers
I inn

I
I
I
I
I
I Hanger s paci ng

I
SECTION A-A
I DECK AND TIE BEAM UNIT

I Fig 2.2
 10 I
Line
I
I
I
splices I
Temporary Tie cable~
I
F.ig 2.3A
I
I
I

r bl

•
Tie Backs

/ Fiel d splice
I
I
Sec 4

Sec 3
I
2 I
I
Temporar), · Tie Cable
I
I
ARCH ERECTION SCHEME I
I
Fig 2,38
I
 :.;
I ...
'n
11

I '!bese pennanent cables are cx::mp::lSed of parallel strands which

are suworted by the hanger cables and oormected to the ends
I of the arch ribs . Intermediate supp:>rt of this cable will insure
that the rrodulus of the tie cable is fully effective . 'Ibis cable
is anchored to the arch as shcMn in Figure 2. 4. '!be deck units
I are manufactured off-site and delivered under the arch ribs.

I 2.3 . 3 Erection of the Deck Units

I Figure 2.5 shows the schena for erection of the deck units.
'!be deck units are precast off-site and barged to the bridge
site where they are lifted into place by hoists attached to the
I arch ribs. Olril'B this process, the arch thrust is resisted
by the dead load tie cable .

I As the units are lifted into place , they are joined to the
previoos unit with minimal post tensi~. 'Ibis is necessary
to set the joints and to insure that wind loads duril'B construction
I ooy be transferred to the ends of the span . When the first units
are lifted into place , they are connected to the arch ribs by
shear pins which insure that the units will be laterally stable .
I Units are lifted into place as shcMn in Figure 2. 5 . Note that
liftil'B lines are oonnected at the outside floor beams in the
units provi~ stability duril'B erection . '!be hoists could
I be attached at different ~er positions on the arch to provide
further stability. By usil'B separate anchors for lif~ and
for pennanent support, the cost of ~ers is doubled . An ~tion
I would be to use the lifting anchors as permanent ~ers as shcMn
on the right of Figure 2. 5 . SUch a configuration would increase
the ~er forces , particularly on the ends where the sl~ of
I the ~ers beccrnes less vertical . Horizontal force oust be
considered in design of the post tensioning of the deck and tie
beam units .
I When the units are in place small donut type gaskets are
inserted to seal the ducts at the joints. A few strands are
I installed to seat the joints and to insure that the units act
as an integral unit during the erection prooess. '!be ranainil'B
p:>st tensi~ is installed when the center closure section
I of the deck , as shcMn in Figure 2.6, has been cast.

Post tensioning of the oc:mpleted p:>rtion of the deck causes

I it to shorten free of restraint fran the arch ribs. It ooy be
desirable to let the deck creep at this tiJre under the post
tensioning load before it is oormected to the arch ribs.
I
I
 Anchor
12
Plate
I
/ I
I
I
Dead load Tie Cable I
P LA N
I
I
Anchor Plote I
Arch Rib
I
I
II
I
I
I
..... 'to.t"
I
+- .......
....... t+ ..
I
I
I I
SEC TI ON A-A I
DEAD LOAD TIE CABLE ANCHORAGE
I
Fig 2.4
I
 :;'
I "1
.",
13

I Alternative hanger
arrangement
I
I
I
I
I
I
I
I
I Bracing

I 1 _ - - Arch Rib

I
I Cable
I I
I_----Hoist
I

I i-----........,.r-- Deck Unit

I ERECTION SCHEME FOR DECK UNITS

I
I
Fig 2.5
I
 14 I
Typ Precast Deck Unll
I
. . .
/ Typ Hanger
I
• • • • . . • .' • . • • .
~ ~ I
~
%
. . • . . . • • • . . . . . • .~
Cenl.r closure
I
Arch RI b.
I
DECK PLAN I
(Shaded orea 10 be casl-In-place)

I
DECK UNIT LAYOUT

Fig 2.6
I
I
Next the deck. is to be oonnected to the arch ribs, The end
sections of the deck are now cast. '!he end sections are cx:rnposed
of portions of the deck. between the arch ribs and portions of I
the tie beams between the end of the precast units and the arch
ribs as shown in Figure 2.7. Final post tensioni~ is then drne
to form a full manent oonnection between the arch and the tie I
beams. 'lhis will cause a slight decrease in the post tensic:t'lirg
stress already in the deck..
I
To minimize weight, a concrete str~ of 8000 psi is suggested
where good aggregate is available. Sane research iIrlicates that
high str~ concrete tends to creep less. I
Tie beams are designed for vertical beIrli~ induced by the
floor beams framing into than. They rust also resist berrling I
induced by the arch ribs and ~ers.

Since the deck. is integral with the tie beams, it resists I

berrli~ forces as part of the tie beam. Design of post tensicni~
in the deck rust consider local berrling stresses as well as overall
berrli~ and thrust stresses . The force fran the arches and tie
I
beams is transferred into the deck through shear. To aC1:UlIIodate
shear, the deck. rust be thicker at the edges than it need be
for local berrling.
I
I
I ';:
')0
15

."
I
I
I
I r Dead

Tie
load

Cables

Containment
I
'1- - - - - - -- - - - - - --
I I
~--------------

I
Post Tensioning Ducts

I Precau Deck and Tie Beam

I Cast -In - Place Precast

I AI

I ~~. Arch to be field bolted

+tt- to Cantalnment Plate.

I '+f"

- t:t..!
I I
l-

I
Arch Erectlan Pin

/ '\
I
I END ELE VAT'I 0 N

I ARCH RIB TO TIE. BEAM DETAIL

I
Fig 2.7
I
 16 I
2.3.4 Advantages of the Alternate /oEthod
I
'l1le proposed alternate method of <XlI1Structioo eliminates
the earlier outlined disadvantages of the tied arch. 'l1le problem
I
of noo-redundancy in the tie beams is eliminated by the Illlltiple
wires in the dead load tie cable and the post tensiooir¥j strand
in the deck units. Instead of a large sir¥jle element tie beam,
I
multiple small wires are used to resist thrust fran the arch
ril:s.
I
'l1le oostly stress relief joints in the deck are eliminated
by post tensiooir¥j the deck. Falsework is also eliminated.
'lhls is nade possible by the dead load tie cable which provides
I
support durir¥j <XlI1Structioo of the deck. HqJefully, <XlI1Structioo
will be nuch faster and navigatioo channels will be un-obstructed
except for short periods.
I
Reduced steel weight should oontribJte to lo./er oosts through
both less base naterial and oonnectir¥J naterial such as splice
I
plates and bolts. Fewer splices also lead to reduced field labor.
I
'l1le use of steel and ooncrete appears at first to be in
reverse of good practice in that steel is used for the arch which
is viewed as mainly a canpressioo member while the deck coocrete
I
is used as part of the tie member which is viewed as nainly a
tension member. H<:7ft'ever, the arch is also a flexural member
subjected to rather large be!rlir¥J stresses. D..irir¥J <XlI1Struction,
I
loads nay induce net tensile stresses in the arch rib. 'l1le deck
is used as part of the tie member which rrakes the overall design
IOClre efficient. 'l1le tie member is actually canposed of high
I
strength steel wire strand with respect to dead load applied
prior to post tensiooir¥j. Tensile forces have been isolated
fran the flexural loads. 'lhls permits the use of efficient high
I
st.rer¥jth strand.
I
I
I
I
I
I
 17

I 3 . 0 DESIGl EXAMPLE BY ALTERNATE ME1lKD

3 . 1 General
I 'll1e best way to determine the feasibility of the proposed
metlxxi i s to perform a design study . 'll1e objective of this
I study is not to develop a o:xnplete design , rut to examine the
obvious probl ems in sufficient detail to determine feasibility
of the method.
I An existing design was selected as a basis for the design
study . 'll1e original design was chan3ed only where necessary
I to aCCUlllodate the alternate approach.
interesti\'¥j a:mparisons of member sizes
This permitted sane rather
and design forces .
AASH'ro Load Factor provisions were used in the study .
I Table 3 . 1 shows a a:mparison of the original and alternate
designs . Figure 3 . 1 shcMs an isanetric view of the design example.
I TABLE 3 . 1 - DESIGl EXAMPLE a:MPARISCN

I ITEM CRIGINAL ALTERNATE - DESIGl 2

I SPAN 620 . 5 ft Same

WIDl'H BE:IWEEN 91. 0 ft Same

I AROI RIBS

AROI RIB HEIGffi' 128 . 0 ft Same

I 8" thick 15" to 9" thick
fc ' = 4000 psi fc '= 8000 psi
I Ul'lGI'lUDINAL 33" deep @ 9 ' -3 " Spa
(steel)
I S'ffiINGERS

9 ft deep steel 5 i t deep steel

Spacing = 36 . 5 ft
I DIA<l:NAL Box sections steel
spacing = 12 . 17 it

None

I BRACING top and bottan of

floor beams

I JOINTS IN
DEn<
Every 36 . 5 ft 620 . 5 ft

I HANGER SPACING 36 . 5 it

11 ft deep (steel)
Same

9 ft deep (concrete)
TIE BEAM

I
 18 I
'!he arch ribs rises 128 feet above the deck. '!hey are
unchanged fran the original design except that the plates are
I
kept at a cx:nstant thickness across the entire span. In the
original design, the plates are thinner in the center of the
span. '!he cross sectioo of the arch ribs is shawn in Figure
I
3.2.

Figure 3.3 sl'x:lws the deck units. '!he floor systan in the
I
original design utilizes floor beams at 36.5 feet which supports
stringers spaced at 8' -0" and an 8.0 inch thick ooncrete cast-in-
place deck. 'lbere are lateral bracing menobers in planes of both
I
top and l:ottan fl.an:3es of the floor beams. 'l11e alternate design
utilizes floor beams spaced at 12' -2" with 00 loogitudinal string-
ers or lateral bracing systan. External to the deck units, the
I
dead lced tie cables take thrust fran the arch ribs due to the
dead load of the deck units. I
'l11e deck is 9.0" thick in the center 50 feet in the alternate
design. 'Ibis thickness is sufficient to span between the floor I
beams. At the tie beams, the deck is 1'-3" thick to provide
adequate shear strength for transfer of shear fran the tie beam
to the deck. I
\,1>• I
Honger
/ I '--;--.--r--...L
Arch Rib
I
I
I
I
I
I
I
DESIGN STUDY EXAM PLE
I
Fig 3.1
I
 19

I 3.2 Design Considerations

3.2.1 Live Lead

I MSIrro Bridge Specifications are applicable to bridges with
spans up to 500 feet. Although this example exoeeds 500 feet,
I 00 nodification was nade to the loading since the original
designers apparently nade 00 such rrodificatioo. 'Ihe original
design was based on HS20 plus Interstate. 'Ihe same live loading
I was used in the design study.

MSIrro specifies when four or !lOre traffic lanes are loaded

I to obtain the maximun load, the . resultant forces are to be reduced
to 0.75 times the cx:JTq)Uted value to acoount for the low probability
of the maximum load occurri~ Simultaneously in all lanes. 'Ibis
I factor is applicable to nearly all live loads reported in the
study.

I Lane load controls !lOSt loo:;Jitudinal members. Floor beams

are cootrolled by Interstate 1oadio:;J for strength, HS20 vehicle
load controlled fatigue. Har¥jers are controlled ' by the HS20
I lane load. Allowable fatigue stresses are l::ased 00 Roadway Case
I and redundant oorrlitioo. Interstate loading is oot considered

I for fatigue.

Figure 3.4 shows the deck cross section. Although there

I is a tarrier in the center of the deck, it is igoored when examin-

ing lane positioni~ for critical oorrlitions because the tarrier
cc:W.d be r€lTOVed at sane future date.

I
I
I . --. ---,
I ;'6 01 .2 " . t.
/ Top 0.0 boll ...

I c:::::=====::5
I Are a • 324 1,,& I . . . 1310001,,4

ARCH RIB CROSS SECTION

I
Flo 3.2
I
 20 I
I
i
36' - 6"

I ,'c • 8000 p.s.i.

=
I
I
=o
<D
I
I
I
4'-0" 12'- 2" 12'-2" 8'- 2"
I
SECTION A-A I

, -- Hangers

I
72 sq in
Tie Cable
Dead Load Deck tape rs over
I
I
20'-0" length

,~ .1\... ... p.,, r-..l>

v
I
i.
.(
-
L: ~ ... ... ...
If"': -
I
- ----
-t
-
--- -- - T

/
/+

I' Typ . Poat T8n.ionin~ Ducts

I
:1
I'
O"l

-CD
I
Floor Beam I
I I
I

~ ... ~ - - - I
I r-...l>
v

DESIGN STUDY DECK UNIT

I
FI g 3.3 I
 21

I 91' - 0"

I j
I 3 Lones ~ Lones

t t
I nn 11. •
,...J
I
I
I
,. CROSS
DESIGN EXAMPLE

SECTION OF DE CK UNIT

I FiO 3.4

I
I
I
I
I
I
I
I
I
 22 I
Six 12-foot traffic lanes are pennitted 00 the structure. '!he I
live load is six feet wide am may rrove within its 12 foot lane
as lol'¥3 as it stays two feet fran the edge of its lane . Each
12 foot lane may be !rOved transversely , rut must not override I
an adjacent lane.
I
3. 2.2 Dead Laid

Dead loads are oanputed based on a unit weight for CXX1Crete I

of 150 pounds per cubic foot am steel of 490 pounds per cubic
foot . A weari1'¥3 surface of 30 pounds per square foot of ridiI'¥J
surface is added to the superinp:>sed dead load. Parapets am
I
the center barrier are coosidered in the superinp:>sed dead load.
The weight of the parapets am barrier are placed 00 the structure
as a series of concentrated loads at the nodes over floor beams .
I
The barrier am parapets oould be precast with the units . In
that case they would be oonsidered part of the dead load of the
deck units .
I
3.2.3 Coostructioo Wads I
'!he feasibility of any large bridge is dependent 00 its const-
ructability . Although a canplete examinatioo of constructioo
stresses is beyood the scope of the study , it is extended to
I
evaluate, in an elerrentary manner, coostructioo of the deck units .

'!he stresses in the arch ribs duril'¥3 their erectioo are not
I
reported . However, oanputatioos were made to determine that
erection stresses in the ribs were not critical with practical
length stiff legs am tie backs . Coostructioo stresses in the
I
arch ribs just prior to placanent of the deck units rust be kn<7,.m
so they can be added to stresses fran subsequent loads. These
stresses are given.
I
Placanent of the deck units is examined in sane detail. Pairs
of deck units are assumed placed simultaneously on the structure.
I
Each deck unit is canposed of three steel floor beams am their
portioo of deck slab am two tie beam segments. I
At this point, the dead load tie cable is the ally member
resistil'¥3 thrust of the arch ribs . The deck units are oonnected
to the eros of the span by shear pins to resist lateral loads
I
am insure that the deck does not sway . Each unit is sufficiently
post tensioned to the prior unit to set the joint am resist
wind loads duril'¥3 erectioo.
I
I
I
 ,'"
'!l

I ,.,
''1
'0 23

I When all units have been erected and the center closure portien
has been cast, the deck units are finally post tensiened.

I Portions of deck and tie beams subsequently cast at the erxl.s

of the bridge. Final post tensicning of these portions to the
precast units insures a full rocment connection to the arch ribs
I and integral action by the entire deck. 'llle final post tension-
ing terxl.s to unload the stress in the main portion of the deck.
Post tensioning of one em to the arch will simply translate
I the deck slightly tcMard that em. However, post tensioning
of the last em will be resisted by the arch rib. Tensioning
will terri to unload the tie cable and increase thrust in the
I arch. 'llris behavior must be considered when designing the post
tensioning •

I 3.2.4 Thennal Loads

I 'llle possibility exists that solar energy could heat up the

arch ribs, hangers and tie cables more quickly than the massive
deck and tie beams. If this happens, the post tensicning stress
I will be reduced. We have estimated a temperature difference
to enable us to considered this effect in the design of the post
tensioning. Cbnversely, the arch may cool more quickly than
I the concrete. 'llris should also be considered in an actual design
rut is not considered in this study. Additional deck stresses
are induced due to the increased thrust in the arch ribs. 'llle
I effect en the arch ribs is also examined. The results of the
thermal analysis are designated as "Temperature".

I 3.2.5 Post Tensicning

I 'llle anount of post tensioning required in the deck and tie

beams is determined by the amount required to overcx:me the
largest tension stress in the ooncrete. Tensioning prior to
I losses is limited by the canpressive strength of the ooncrete.
Two cases are examined in this study:

I CASE I 1.3(D + 5/3(L + Ill;

CASE II 1.3(D + T + (L + Ill.

I Allowable stresses for CASE II are increased by 25 percent.

I M:lxiroum tension and canpression are determined for each

corrlition. It is assumed that post tensioning in the deck is
at the center of gravity of the deck for overall effects and
I for simplicity at one-sixth the thickness of the deck away fram
the center for local berrling.

I
 24
I
IDsses due to =eep and shrinkage as well as anchorage and
I
friction are oonsidered in the design study. '!he age of the
units, relative humidity and several other factors are not
oonsidered but sOOuld be oonsidered in a detailed design.
I
Preliminary post tensioning of the precast units is perfonned
iJlInediately after they are lifted into place. '!he majority of
I
post tensioning is perfonned after all of the units are ccnnected
and the center closure section is cast. '!he norent oonnecticn
between the deck and the arch ribs is made by post tensionir¥J
I
the deck to the cast in-place em sections. Final tensioning
terrls to unload the tensioning in the other and is acoounted
for by over tensioning the precast units.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 '!l

I ....·~
'0
25

I 4. 0 ANALYSIS USING FINITE ELEMENI' M'DETS

I 4.1 General

'Ille majority of the analysis in the design study is performed

I using the finite elerrent metrod . FESAP, the finite element CCJJp.lter
pro:JLam that is utilized in the study, is licensed to BSDI by
BaJxxx:k & Wilcox.

I ESDI has developed a number of interactive cx::mputer programs

that are utilized to Wild the finite elerrent m:Jdel am a similar
I number of pro:Jrams that process the analysis produced by FESAP .
'Ille results of the analysiS appear to the user in the form of
nanents and shears rather than as raw finite elerrent stresses .
I '!here is also a cx::mputer pLo:JLam which places a specified live
load 011 an influence surface that has been developed fran the
FESAP analysis . In the cx:urse of the analysis, hundreds of
I influence surfaces are subjected to the loader . 'Ille cx::mputer
pro:JLams are part of the BRILGE-SYSTEMsm developed for girder
bridges rut rrodified to analyze the tied arch in this study.
I '!he m:Jdel ruilt to evaluate the placement of deck. units is
a two-dirrensional (2D) m:Jdel. A 3D m:Jdel is used to analyze
I the floor beams . Finally , a fully integrated 3D m:Jdel is used
to analyze the entire structure for superimposed dead load,
thermal am live load.
I 'Ille integrated 3D m:Jdel has 3490 elerrents am an equal nunber
of oodes . '!he minimized barrl width is 485 . 'Ille integrated struc-
I ture has been run 011 a VAX 11 /780 CXlTputer . 'Ille interactive
programs are run 011 a Victor 9000 micL<XXlUputer.

I 4. 2 Floor Beam

I 'Ille floor beams are designed first so that they can be properly
m:Jdeled in the integrated m:Jdel am so the proper deck. thickness
is kmwn . '!his permits the weight to be rrore closely estimated
I when examining staging of the deck. units .

'Ille floor beams are rrodeled by considering entire deck. unit

I of three floor beams with its corresponding width of deck., 36' - 6" •
Design of flcor beams was predicated on simple span behavior .

I '!he webs of the girders are m:Jdeled using a single plate

elerrent over the girder depth . '!his assumption requires a further
assumptioo that shear is constant over the girder depth. 'Ille
I top am Ix>ttan flanges are =deled with beam elerrents.

I
 26
I
'nUrteen nodes are evenly spaced over tte 91.5 foot width . The
I
deck was lTOdeled using eight-rode solid elanents. Six elanents
were used across tte 36 '-6" deck sectioo. The deck is connected
to the steel girder elanents with very rigid beam elanents in
I
the vertical orientatioo . 'nUs insures full CCJ!ilOSite actioo.
The rrodel is sl1o.m in Figure 4.1. The rrodulus of elasticity
for the ooncrete and steel were input based on MSHro.
I
The ooncrete rrodulus is adjusted by a factor of three (3.0)
to account for creep and shrinkage when analyzing for dead load.
I
The analysis is performed by hand since the BRIlXiE-5YSTEMsm loader
works only in tte longitudinal directioo. The design is performed
in interactive design programs USing the properties described. The
I
design program permits the user to m:Jdify plate sizes and check
stresses using tte MSHro wad Factor Design criteria. I
I
I
I
I
I
I
,Iemenll
8 Node lolid I
~
elemenll

b.
Plole
elemenl I
\ 6 • 6 ( SBS)
r.action Illmlntl
Top and bottom
beam ellmlnta
I
I
FLOOR BEAM AND DECK MODEL
I
Fig 4.1
I
I ; 27
'"
'D
.,n

I 4. 3 Placement of Deck units

4. 3 . 1 Arch Rib

I Placement of the deck units 00 the arch structure is examined

USil'¥J a 2D finite elerrent lOOdel . 'Il'e arch ribs , dead load tie

I cable and tie beam 00 ate side of the bridJe were n'Odeled . 'Il'e
IOOdeI i s shown in Figure 4. 2. Since the arch rib is actually
IOOdeled using 30 elerrents , a third reaction at the t:q:I of the

I rib and preventlr¥J lateral translation is required .

A series of beam elerrents are used to OOild the arch .

I 'Il'e ooordinates are cx:rnputed using the paralxllic equatioo :

y = 0.001288(616 . 91Ox-x2 ).
I 'Il'e elanents for the arch rib extend be~ har¥;Jer locations
aloog the parabolic shape of the arch . Figure 3 . 2 shows the
I box cross sectioo and properties of the arch rib . In the design
stoo.y, the plate thicknesses are held constant over the entire
span. '!here are 34 elerrents in each arch rib. All elerrents
I are straight .

I 4. 3. 2 Hangers and Dead II:lad Tie Cable

'Il'e cable I!leI1i:lers in the structure are IOOdeled using spar

I elerrents . 'Il'e har¥;Jers are the same as in the original design ,
Area = 6 . 67 in2 . 'Il'e area of the dead load tie cable is 72 in2 .
'Il'e Jrodulus in the cable elerrents is 28000 ksi . 'Il'e dead load
I tie cable is connected to the nodes at the ends of the arch rib .

I 4.3 . 3 Precast Deck units

'Il'e pr ecast units are lOOdeled as beam elerrents . '1l1e stiffness

I of the beam elerrent is cx:rnputed as the canbine stiffness of half
of the entire deck unit . Yoor¥3 I s m:XIu.lus for live load is used
for the CCflCrete . AlthcAJgh the entire deck is not effective
I due to shear lag , it is believed that the error introduced I::tt
this asSlllllptioo is small . 'Il'e manent of inertia about the horiz-
ootal axis of the tie beam used is 500 ft4 . 'Il'e area of the
I tie beam used is 55 ft2 .

Tie beams are not connected to the arch ribs at this time.
I 'lllree beam elerrents are used to represent the tie beam and deck
between each har¥;Jer . Fach node in the tie beam represents the
locatioo where a floor beam is CXXlI'Iected . 'Il'e weight of each
I deck unit is applied as three (3 ) coocentrated loads at the nodes
of the units being placed . '!his arrar¥Jerrent is shown in Fig 4. 2

I
 26 I
Arch Rib divided Into
34 Beam
I
I
Lateral
reatlon
I
I
lS", EI.m", ,,,
dead load Tie Coble
I
I
I
Beam E.lemenTS - 51 for
Unl t - I
/he complete Tie Beam I
t4--.----.,.-+_~
----Applied loadli
I
116 kp Each I
77 kp
DETAIL" A" I
I
I
H2 / "",,-2.8 I
--.--+-~-
I
116 kp Each
I
DETAIL" s"
ARCH RIS AND STAGING MODEL
I
Fig 4.2 I
 ':!>

I -0
~
, , 29

I 4. 3 . 4 BeariI'¥Js

I Bearir¥Js are trodeled usir¥J a special element that connects

beb¥een the ground and the structure. It permits the user to
specify a spriI'¥J stiffness for any of 6 degrees of freedan .
I 'n1ese elements are called Six-By Six e l ements (SBSs ). An SBS
is placed at each end of the arch rib . 'lbey were connected to
the same nodes as the tie beam. 'n1e SBSs are specified rigid
I in the vertical and transverse directions and given zero rotational
stiffness . '!he longitudinal stiffness is zero at one end of
the arch rib and cx::mpletely rigid at the other.
I
4. 3. 5 Proced.lre
I '!he first analysis is for the arch rib supportir¥J its self
weight . '!he results of this load are accumulated to the results
I of subsequent stages of placing deck units .

Stage 1-
I '!he secoro
analysis is m3.de by adding the first four beam
elements on each end of the structure. '!hey are connected to
I the end of the arch rib and to the first hanger . l'IcIr¥Jer are
connected to the arch ribs . '!hese units are sho.In in place in
Figure 4. 3A and B. '!his analysis represents the first pair of
I deck units placed on the structure. Although the units would
actually be lifted singly , they are treated as if lifted in pairs
and placed synnetrically . Units shoold be placed synnetrically
I in or der to minimize bendir¥J in the arch ribs . '!he resultiI'¥J
analyses includir¥J nanents and thrusts in the arch rib , har¥Jer
tension , tie cable tension and reactions are reported.
I Stage 2-

I In the secoro stage of construction , the stiffness of the

beam elements representir¥J the first pair of units is increased
to their full live load stiffness . I£>ads representir¥J weight
I of the first pair of units is then resroved to avoid multiple
oountiI'¥J . 'lbe first units are effectively stiffened by increas-
ir¥J the nanent of inertia, rut they are not oonnected to the
I arch ribs . '!he secoro pair of deck units are added to the trodel
in a m3.n0er similar to the first . 'n1e weight of the second pair

I of units is placed on the oorrect nodes and their stiffness is

small . '!his stage of analysis provides thrust and nanent in
the arch rib as well as serre additional effects on hangers supp-

I ortir¥J the previous units .

I
 30
I
Shear connection only
I
between deck unit and

arch rib
I
I
I
Unit - 2
I
\ I
STAGE 2 I
No bending 5tiff neS5
Fi g 4.3A
between deck and arch
I
during conUruct Ion
I
I
I
I
I
Unit - 8

~ I
I
STAGE :3
I
STAGING OF DECK UNITS
I
Fig 4.38 I
 ,"~
I ''l
'0 31
I

I Bending in the tie beam at this stage must be accounted for

by installing sufficient post tensioning across the joints between

I units. ~ts in the tie beam are not reported rut IoKlUld be
used to cnnpute stresses in the tie beam and deck during erection.
They are not cnnputed because the roodel is crude and the value

I of moment is small.

Stage 3-

I In this stage, two units are placed in the center of the

span to reduce the negative moment that had accumulate in the
I center of the arch ribs . The procedure for analysis is the same
as for the previous stages.

I stage 4-

In the analysis for stage 4, the units placed during Stage

I 3 are ignored and their loads are raroved. The two units in stage
4 are adjacent to the units in Stage 2.

I stages 5 and 7-

These stages are treated in the same manner as previous stages.

I The center units remain unconnected to the other units.

In an actual design , the engin.e er may wish to provide a full

I 3D analysis for each stage of deck unit placement. study is
required to detennine the error pennitted in placing the weight
of the pair of units unevenly.
I
4. 4 Integrated Model
I 4.4 .1 General

I The integrated nodel is a full 3D nodel of the a:mpleted

structure . The roodel is used to analyze for superimposed dead
load, live load and thermal loads . The nodel is the same as
I the 20 staging roodel. with regard to the arch and tie cable and
bearings.

I 4.4. 2 Tie Beam

The cross section of the tie beam nodel is a:mposed of one

I plate element representing the web and two beam elements represent-
ing top and totten flanges . All elements in the tie beam and
deck roodel are 12 '-2" long, which is the distance between floor
I beams. All tie beam properties are for concrete . Figure 4.4
slx:lws the nodel of the tie beam and deck elements.

I
 32
I
I
Detail "A"
I
I
I
I
ELEVATION OF ARCH
I
Arc·h Rib I
12' - 2" Ty p
Deck Element

Floor B eams
--- Hanger

Deck
I
I
Beam

E Ie men ta \ I==::;;~===:;:::;-;L=====l
/ I
, I
""
Reaction . I
Floor Beam
Spacing
Dead load Tie Coble
I
Tie Beam

I
DETAIL "A" I
INTEGRATED MODEL- OVE.RVIEW
I
I
Fig 4.4
I
I '0'~ 33
.",

I '!be tie beam web is g'-O" deep . Table 4.1 provides properties
and sizes of the two designs examined. Design 1 was perfOI:lTB:l

I first and fo..ux'l to be too heavy. Design 2 was based en the

results fran Design 1.

I Arches are camected to the tie beams at the midheight of

the tie beams as shcwn in Figure 4.4. '1lle nodes at these points
are connected to the top and rottan of the tie beams with stiff

I beam elanents to insure that a full m::ment camectien is provided.

'Ihls connection is slightly eccentric with respect to the center
of gravity of the tie beam. Bearir¥Js were also placed at these

I nodes and 'Nere the same as for the 2D staging rrodel.

I 4.4.3 Deck

'1lle top of the deck is positioned 1 foot ab::lve the top of

I the tie beam web in roth designs. 'Ibe deck is rrodeled by a series
of 12 eight-node solid elanents across the bridge between tie
beams as shcwn in Figure 4.5. 'Ibe portion of ~ deck outOOard

I of the tie beams is rrodeled with another eight-node elanent as

shcwn in Figure 4.5. There are a total of 728 deck elanents
in the integrated rrodel.
I
4.4.4 Floor Beams
I '!be floor beams and deck are connected at four locations
across the bridge. The oonnections are made with stiff beam
I elanents. At the ends of the floor beam, the connection to the
deck is the same elanent as that used for the tie beams. '!be
four connections insure that the floor beam works CCIIlJOsitely
I with the deck for rather uniform loads. It does not accurately
represent the behavior for such loads as the barrier do..m the
center of the bridge. 'Ibe barrier was placed midway between
I connectors so the deck aweared rore flexible than it should
have appeared. Likewise, ccnoentrated loads for live load
analysis do not act correctly in the local region. lio.Hever,
I they are rore than sufficient for predictir¥J overall structural
behavior.

I Floor beams are autanatically connected to the top and rottan

of the tie beams in the BRIrx;E-SYSTEMsrn. Since the tie beam
is g'-O" deep, the floor beams are g'-O" deep at their ends instead
I of 5' -0". '1lley became 5' -0" deep at third points. 'Ibis causes
sane additional end stiffness in the floor beams. 'Ihls did not
affect design of the floor beams which are designed as sin'{lle
I •
spans. It would, however, tend to lead to the integrated J1'Odel
underestimatir¥J the stresses at mid-span of the floor beam CCIlPll"ed
to the simple span assumption used in their design.
I
 34
I
TABLE 4 . 1 - KDE:L EXPLANATlOOS F(R FlmRE 4.5
I
Property
I
Location Description Material Design 1 Design 2

A Top flange of tie beam element O::x1crete . 1 ft2 Same

I
B Web of tie beam Plate elan. Concrete 1.0 ft . 75 ft
C
0
Botton flg of tie beam element O::x1crete
Botton flg of floor beam an elan Steel
6 ft2
30 in2
2 ft2
Same
I
E Web of flr beam plate elan. Steel .5625 in Same
F
G
Top flg of flr beam plate elan.
Joint between floor btl elems .
Steel 10 in2 Same I
H Deck Nodes 8-node solids O:xlcrete See Dtl .
I&J Arch Rib Beam Elements Steel 324 in2
131000 in4
Same
Same
I
K&L HanJer spar Elan. cable 3.6 in2 7 . 2 in2
M
S
Tie cable Spar elan.
Stud conn . an. Elan.
cable
Steel
72 . in2
1 . 0 ft2
Same
Same
I
MJI 4000ft4 Same
I
MATElUAL PROPER1'IES I
,{CXlNG'S r-a:xJLUS POISSrn 's RATIO

STEEL 29000000 psi 0. 30

I
CXNlmTE
LIVE LOAD
SUP DEAD LOAD
5400000 psi
1900000 psi
0 . 15
0 . 15
I
CABLE 28000000 psi 0. 3
I
I
1-2-3-4
3-7-4
Connectivity
Connectivity
of end of 8-node solid deck element
of beam stud element
I
7-9
8-10
7
Coonectivity
Oonnectivity
Oonnectivity
of top flange of floor beam
of botton flange of floor beam
of botton of hangers
I
I
I
I
 ''':>

I '.0'~ 35

I J
4
I
I
I ......... K ./"
L

I H 5
/5 / l / 5 /
I ....
-
.-" - -- :...-'

B/
'\ \\ \
\
I "- C
\
o E F G

SECT ION THROUGH

I INTEGRATED FINITE ELEMENT MODEL

I
I
~
2' - 0" ABT 7' - 6"

I ,I .1 6

I
I 2 7 9

I 5

I -o

I
I
I 8

INTEGRATED MODEL- DETAILS '

I
Fig 4.5
I
 36 I
4.4.5 Prooedure I
Superimposed dead load was analyzed first using the integrated
model. In this case the wearing surface is considered by introducing
an artificial density to the ,8-node deck elements. nus results
I
in a load equal to the weight of the wearing surface being applied
rather unifonnly to the structure. Parapets are modeled with
longitudinal rows of concentrated loads along the bridge at each
I
node where parapets or barriers are located. 'nle concrete modulus
is decreased to ale-third the nonnal value for this analysis
to a=unt for creep and shrinkage. 'nle tie cable is considered
I
effective.

In the live load analYSiS, the modulus was changed to the

I
full value. A series of load cases are examined where each case
considers a single concentrated load applied to the deck at prescribed
locations. 'lbere is a load case for a total of 18 lines of loads
I
applied along the span. A line of loads is applied at each hanger
line and reactioo line. Each load line consists of 9 loads. I
'nle results of these 162 load cases are saved fot' a large number
of respcnses including arch ncnents, deck stresses, hanger stresses,
reactioos, etc. Influence surfaces with 162 values each are wilt I
fran these respc:nses.

Each influence surface is then subjected to a searching technique I

to determine the locatioo of the specified vehicles which cause
the maximum and minimum respc:nse within the prescribed limitations
of AASHro. A CXXl'{)arisoo with the original design live load values I
is made for certain cases in Section 5 - Results.

Local bendir¥;j in the deck due to live and dead loads is CCIll{:Uted I
by hand according to MSHro-3.24.3.2. nus method dces not allow
for the flexibility of the flcor beams. It also is tOOu,ht to
be rather conservative. Influence surfaces based 00 the deck
I
and flcx:>r beams is expected to yield lower bending stresses.

'nle tanperature effect is analyzed using the same integrated

I
!!'Odel as for live load. ~ature of the arch ribs, hangers
and the dead load tie cable is increased 100 degrees Fahrenheit
above the concrete.
I
'!be aIOOunt of {X>St tensiooing in the deck and tie beams is
determined by findir¥;j the sum of the critical stresses in the
I
deck and in the tie beam based 00 the above analyses.
I
I
I
 37

Preliminary examination of the c:x:n::rete stresses in Design 1

irrlicated that the Il'Odel should be m:xllfied to ooosider a thinner
I deck. at the tie beams arC thicker in the center . '!he Ix>ttan flange
of the tie beams as well as the webs of the tie beams were found
to be too large. It was also learned that cnly half of the oorrect
I han:Jer cable area had been used . A m:xllfied integrated rrodel was
then ruilt arC the same runs performed again . Results of Design
2 were reported under Secticn 5 - Results .
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 38
I
5.0 RESULTS
I
5.1 General I
nus section reports the results of the analyses described
in Section 4. '!he floor beams results are based on the simple
span analysis which was dene by hard. Results for the arch ribs,
I
~ers, dead load tie cable, tie beam, reactions an:i deck were
all determined fran the 20 an:i 3D analyses. '!hey are reported
in Sl.IIlIllarY fonn with results canbined acc:ord.i.n3' to AASIfIO Cases.
I
Results are first reported in response tenns such as m:::ments
an:i thrust for each load condition where appropriate. 'lhese
I
respcnses are then CClwerted into stresses an:i canbined in the
appropriate load cases. Case I is AASIfIO CASE I an:i Case II is I
AASIfIO CASE IV fran Table 3.22.1.A for Load Factor Design. Appro-
priate cases are also ccnsidered for determining the ano.mt of
post tensionin:] according to AASIfIO Section 9.16.2. I
5.2 Floor Beams
I
Figure 5.1 shows the m:::ment envelope for the simple span
analysis. '!he oontrolling live load m:::ment for strength is caused
by the Interstate load; and the controlling loading for fatigue I
is the HS20 vehicle loadinJ. The ccntrollinJ load conditien
was all siX lanes of traffic placed as close as possible to the
point of consideratien. I
All dead load is placed en the ocmposi te section based en
the asslIllptien that the floor beams be fully shored during castirx] I
of the deck unit. Further, there has been no consideration for
lateral buckling of the top flange. '!his pe:rnti.ts the top flanges
of the floor beams to be made rather small. I
Figure 5.2 shows the shear envelopes. Shear connector design
is oontrolled by fatigue which is related to shear range. I
Figure 5.3 shows the floor beam. '!he weight is 16.2 pounds
per square foot of deck a::rnpared to 36 pounds per square foot I
for the steel in the floor system of the original design.

'!he maximum live load bending stress in the lxlttan of the I

floor beams based on the simple span analysis was 13 ksi inc1~
impact. nus a::rnpared to 7 ksi based on the integrated !lOdel.
The integrated !lOde1 results were based on loading the influence
I
surface for the axial stress in the beam element lxlttan of the
floor beam at the center of the center element. '!he integrated
!lOde1 was not developed to properly IlOdel the floor beams.
I
I
 40
I
I
100
I
I
I
80

L .L . ( Interstate)
I
/ I
60
L.L. ( H S 20)
I
I
-...
Q.
40 I
~

a:
I
'"
UJ
:r D.L.• /'" I
VI 20

I
I
0 10 20 30 40 I
I
HALF SPAN f ee t )
I
FLOOR BEAM SHEAR ENVELOPE
I
'fIg 5.2
I
I ':!l
' ,.,~
.
.'!j
.,) 41

I With only foor oonnecting ocrles per floor beam between the deck
and the steel, the ooncentrated unit loads are not distri.bJted
I to the floor beam oorrectly. When unit loads are placed over
the ocrles with beam stud oonnectors to the floor beam, the load

I is transferred directly to the floor beam, rut when coocentrated

loads are placed between these studs, the load is distri.bJted
to adjacent floor beams as well as to the one uriler oonsideratioo.

I Figure 5.4 shaws the influence line for the stress in the oottan
of the floor beam based on the 3D rrodel. Also, for cx:roparisoo
the influence line based on a simple span is shown. '!be influence

I line fran the 3D rrodel is truncated in the center because there

is no stud in the center of the span.

I
I 2- 4
u
• 0'875" Stud sheor connector. Plr row
-
at lOll spoc ln g . •
I 110 Rows 220 Stud.

t. 10".0'75" l 91 1 -0" ( A36)

I
I
t. 60",0'75" (A36)
I ,
...
0
30" 3 at 90" 3 al 90" JO"

I
11
4"" 0 '312.5 Stiff nor I y p -

I
I ( A36)
r. IS", 1'375",60'- 0" (.4572)
(A36)

I No eoarlng Sllffnoro requlrad Wab Sllffna" ana oldo onl1

I
TYPICAL FLOOR BEAM

I
I Fill 5.~
 42
I
I
I
I
I
"
I \
\
I
I \
20 t I
I
IS
I
-
•
Q.
I
I
.>t:
~

I-
I
z 10
w
~
0
3D Mod e l I
~

I
5
I
I
0 I
SPAN
I
FLOOR BE AM IN FLUENCE LI NE
I
Fi g 5 .4 I
I
I ·n
~
.;p
.,!) 43
."
I 5.3 Arch Ribs

I 5.3.1 General

Since the arch ribs were IT'Odeled as a series of beam elenents,

I the narent and thrust responses CCAlld be reported directly.
FUrther, since the arch rib IT'Odel did not change between the
20 and integrated m:x:lel, no transformations were required before
I the results CCAlld be c:anbined.

5.3.2 Deck Unit staging

I Arch ribs narents at hanger locations are presented in Table
5.1 for each stage of deck unit placement. CClnparable thrusts
I are presented in Table 5.2. Section 4.3.3 describes in detail
the sequence of deck unit placement.

I TABLE 5.1 - DEAD LOAD t-n1ENI'S IN ARCli RIB WE TO STAGING

(FI'-K)
I Wcation 0 1 2 3 4 5 6 7 8

I ARCli
DL
oo.y
UNIT 1
0
728
378
10139
514
6700
470 326 137 -49 -194 -273
3755 1300 -664 -2140 -3120 -3611
DL UNIT 2 -752 5600 12380 7194 2872 -584 -3176 -4905 -5770
I OL
DL
UNIT
UNIT
8
3
418
-116
-3986
1710
-9155
6476
-10400 -9324 -5938
11778 5527
-240 7772 18097
526 -3224 -5724 -6975
DL UNIT 4 108 -135 1044 4576 9398 3040 -1726 -4905 -6495
I DL
DL
UNIT
UNIT
5
6
108
117
-1603
-2871
-1987
-4327
-1020 1892 7099 1470 -2282 -4160
-4390 -3030 222 6520 2305 198
DL UNIT 7 131 -3789 -6066 -6808 -6009 -3533 1237 9102 6828
I
TABLE 5.2 - DEAD LOAD TIiRUSTS IN ARCli RIB WE TO STAGIN:;
I (KIPS)

Wcation 0 1 2 3 4 5 6 7 8
I ARCli oo.y -531 -504 -482 -464 -448 -437 -428 -423 -421
DL UNIT 1 -124 -129 -133 -137 -140 -142 -143
I OL
DL
UNIT
UNIT
2
8
-333
-418
-768
-119
-402
-727
-218
-754
-227
-754
-234
-750
-242
-740
-247
-724
-250
-701
-251
-672

I DL
DL
UNIT
UNIT
3
4
-484
-573
-506
-570
-488
-577
-328
-566
-340
-432
-349
-444
-357
-454
-536
-361
-460
-543
-363
-462
-545
DL UNIT 5 -654 -647 -639 -636 -632 -524

I DL
DL
UNIT
UNIT
6
7
-717
-762
-711
-756
-702
-748
-693
-739
-685
-729
-683
-721
-601
-711
-609
-657
-612
-660

I
 44 •
TABLE 5. 3 - J\£Xl.MJIMID DEAD I.OI\D IOlENl'S IN AROi RIB WE 'ro STAGING •
Locatioo 0 1 2
(F'I'-K)

3 4 5 6 7 8 •
AROi CtlLY
'lUI'AL UNl
0
728
378
10517
514
7214
470 326
4225 1626
137
-527
-49
-2189
-194
-3314
-273
-3884 •
'lUI'AL
'lUI'AL
UN2
UN8
741
755
16276
10471
19586
10437
11408 4486
1048 -4776
-1125
-6979
-5378
-5513
-8235
-341
-9668
8564
•
•
'lUI'AL lin 755 13162 17037 12783 698 -6513 -8803 -6136 1517
'lUI'AL UN4 755 13132 18575 17453 10039 -3538 -10601 -11117 -5056
'1UI'AL LN5 755 10853 15887 16235 12145 3675 -9006 -13267 -9079

•
'lUI'AL UN6 755 6842 9953 10468 8634 4509 -1894 -10334 -8235
'1UI'AL UN7 755 1656 1817 1617 1307 932 553 75 -62

•
•
TABLE 5.4 - ACXl.MJLATED 'IHRusr AROi WE 'ro srAGING
(KIPS)

•
Locatioo 0 1 2 3 4 5 6 7 8

AROi CtlLY -531 -504 -482 -464 -448 -437 -428 -423 -421

•
'lUI'AL UNl -864 -623 -606 -593 -581 -574 -568 -565 -564
'1UI'AL LN2 -1282 -1025 -824 -820 -815 -816 -815 -815 -815
'1UI'AL UN8 -2050 -1752 -1578 -1574 -1565 -1556 -1539 -1516 -1487

•
'1UI'AL LN3 -2534 -2258 -2066 -1902 -1905 -1905 -1896 -1877 -1850
'lUI'AL UN4 -3107 -2828 -2643 -2468 -2337 -2349 -2350 -2337 -2312
'1UI'AL UN5 -3761 -3475 -3282 -3104 -2969 -2873 -2886 -2880 -2857

•
'lUI'AL UN6 -4478 -4186 -3984 -3797 -3654 -3556 -3487 -3489 -3469
'1UI'AL UN7 -5240 -4942 -4732 -4536 -4383 -4277 -4198 -4146 -4129

•
J\ccunulated mcments and thrusts for each stage are plotted
in Figure 5. 5 and 5. 6, respectively. It is evident that the
•
•
mcment in the arch rib in the vicinity the deck unit being placed
is positive while in other locations it is negative . In order
to keep mcments in the arch nore evenly balanced, the two center

•
deck units were ad1ed as the third stage. 'lllese units would
not be lifted into place, rut susperoed fran the arch.

•
•
1= 45
A CCUM UL ATED ARCH MOMENTS ( kp ft )

1
1
1
21000
Arch only
1 14000
1.000

1 -7000
o

-14000
1 21000 2 U nl t 8

1
14000
1000
o
/
1 .....a.
~
-1000

1 .. -1 4 000

21000 Un i t 3

... 14 000

1 c:
•E
7000
o
o
1 ::I -1000

-14000

1 2rOOO
14.000
1000

1 o
-7000

1 -14000

1 / U';' 7
1
7000

-7000
o
I ( Fi nol )

1
1 Fi g 5,5
 ACCUMULATED TRUST IN
46
ARCH RIB ( kp)
I
,........--- r--,........., I
/,/ "t'--
I
~ ~
I
a
- 1000

-2000 /
I

I
-3000 t Arch on ly

I
o I
~
-10 0 a
-2000
P-unl-t 27-1-/- - \ u - n l t - . I
- 3000
I I

- 1000
0
I
- 2000

-3000 I
- 4000 Uni t ·3 U nl t 4
I
0
- J

-2000
00 0
Unlt5 \ U nl t 6 I
- 3000
-4000
I
-5000

0
I
- 1000

- 2000
- 3000 / FlnOI
" 4000
- 5000

Fig 5,6
I :;
.",
.ct> 47
'n

1 Resulting stresses in the arch rib for each stage of loading

are shown in Table 5. 5. '!hey were determined by SU1'IIlin:J the stresses
1 fran the nanents and thrust in Tables 5. 3 and 5. 4.

1 TABLE 5. 5 - ACXlMJLATED smESS IN TOP OF AROI RIB

OOE oro STAGING

1 (KS1)

1 I.ocatioo 0 1 2 3 4 5 6 7 8

AROI rnLY -1.6 -2 . 4 -2 . 6 -2 . 5 -2 . 1 -1.6 -1.2 -0 . 9 -0.7

1 ror lM
ror W2
-4 . 3
-5.6
-25 . 1
-39 . 0*
-17 . 7
-45 . 6
-11.1
-27.6
-5 . 4
-12.4
-0 . 6
0.0
3. 1 5. 5
9.0 15. 6
6. 8
18.8
ror ~8 -8 . 0 -28 . 4 -27 . 8 -7 . 2 5. 7 10 . 6 7. 4 -3 . 9 -23.4*
1 ror UN3
ror ~4
-9 . 5
-11 . 2
-35 . 9
-37 . 6
-43 . 9
-49 . 0*
-34.0
-46.0*
-7.4
-29 . 3
8. 4 13.5 7. 7
0. 5 16 . 1* 17.2
-9.0
4.0
ror UN5 -13 . 3 -34 . 6 -45 . 1 -45 . 3 -35 . 9* -17 . 0 10. 9 20 . 3* 11.2
1 ror UN6
ror UN7
-15 . 5
-17.8*
-28 . 0
-18.9
-34 . 2
-18.6
-34 . 7
-17 . 6
-30 . 3
-16 . 4
-20 . 9* -6 . 6 12 . 0
-15.2 -14.2 -13 . 0
7. 4
-12 . 6

1
J\CCl.MJIATED smESS IN OOl'lCM OF AROI RIB
1 OOEoro~
(KSI)

1 I.ocatioo 0 1 2 3 4 5 6 7 8

ror AROl -1 . 6 -0 . 7 -0 . 4 -0 . 4 -0 . 7 -1 . 0 -1.4 -1 . 7 -1.9

1 ror
ror
UNl
UN2
- 1. 1
-2 . 3
21.2
32 . 6*
14 . 0
40 . 5*
7. 5
22 . 6
1. 8 -2 . 9
7. 4 -5 . 0
-6 . 6
-14 . 3
-9.0
-20 . 6
-10 . 3
-23 . 8
ror UN8 -4 . 7 17. 6 18 . 1 -2 . 6 -15 . 3 -20 . 2* -16.9 -5 . 4 14 . 3
1 ror
ror
UN3
UN4
-6 . 2
-7 . 9
22 . 0
20 . 2
31.1
32 . 7
22 . 3 -4 . 3 -20 . 2*
30 . 8* 14 . 9 -15 . 0
-25.2
-30.6*
-19 . 3
-31 . 7
-2 . 4
-18.3
ror UN5 -9 . 9 13 . 2 24 . 8 26 . 1 17 . 6* -0 . 8 -28 . 7 -38 . 1* -28 . 8*
1 ror
ror
UN6
UN7
-12 . 2
-14.5*
2.1
-11 . 6
9. 6
-10 . 6
11.3
-10 . 4
7.7
-10.7
- 1.1
-11.1
-14.9
-11.7
-33.5
-12.6
-28 . 8*
-12 . 9

1
smESS = (1OIENl' (FT-K) X 12 ) X ( C=24) / 1=131000 IN**4) +
1 ('lllRUST (KIPS) / ARFA=324 SO. IN . )

1 Asterisks identify maxiJI1l.I1l stresses .

1
 48 I
Figure 5.7 sl1a.Is the stress envelope of maximum tensile and
I
maximum CXlI1Pressive stresses in the arch rib during placanent
of the deck units. The envelopes are based on Table 5.5. They
daronstrate the critical locations. The unfactored stress in
I
both top and botton of the arch rib approach the yield stress
so the sectioo of the arch rib co..Jld be increased, or the erectioo
sequence might be IrOdified to lower stresses. A factor of safety
I
of at least 1.25 would be expected. Stresses in the arch rib
after all deck units have been placed are plotted on the sane
figure for CXJI1Parison. The thrust is seen to be about 15 ksi
I
while the bending stress does not exceed 10 ksi. DJring staging,
rranent daninated the loading over thrust. The parabolic shape
of the arch ribs keeps bending to a minimum when loads are applied
I
unifonnly along the span.

Alth0J9h it is not practical to lift two units at owosite

I
eros of the span simultaneaJ.sly, no analysis was perfonned to
consider single units placed in an unsyrrmetrical manner. It
is thought that sane error shculd be permitted in simultaneaJ.sly
I
lifting two units. The effect of this error can- easily be
determined and shoold be considered when determining a factor
of safety. Another (Xlssibility is to join two units prior to
I
lifting. 1hl.s would permit using the hanger locatioos to attach
the hoist cables without the units beccming unstable or introducing
a different han)er arrangement.
I
5.3.3 SUperinp)sed Dead Load
I
'!he superinp)sed dead load is CXJT1POsed of the wearing surface
and barriers. Results for rranent and thrust are presented in
I
Tables 5.6 and 5.7 , respectively. 'furust is large with respect
to the rranent CXJI1Pared to the staging results. '!his is expected
for loads applied unifonoly along the span.
I
I
I
I
I
I
I
I ;;
'~ 49
.-:> ARCH CONSTRUCTION STRESSES (ksl)

I
I
I
I 50

40
I /MaXlmum tension

30
I
20
I
10
I
0
I -10 /
Final bottom

I -20 Final top /

I -30

I -40

I -50
Maximum compression /

I
I
I r i g 5.7

I
I
 50 I
5.3.4 Tarpmiture I
'lhe results of the thental analysis are presented in Tables
5.6 and 5. 7. In this case, the force is caused by the arch attan- I
pting to expand and tension in the ~ers attanptil'¥1 to restrain
the arch fran rising. 'lhese actions cause a positive rranent in
the arch ribs . I
I
TABLE 5.6 - Sl»lARY OF ARCH RIB t-nlENl'S
(Fr-K) I
Han:Jer I£x:: -) 0 1 2 3 4 5 6 7 8
I
DL 755 1656 1817 1617 1307 932 553 75 -62
original DL 1048 765 924 970 940 934 936 842 863
I
SIMP 874 132 64 73 68 54 39 28 22
originalSUp 314 124 102 110 117 125 132 136 139
I
TEMP 3151 3174 3177 3176 3173 3170 3167 3163 3159

IL+I (+) 799 202 409 596 648 644 574 443 386
I
IL+I final 961 585 880 1111 1222 1201 1057 836 693
orig IL+I 556 777 1229 1442 1509 1444 1254 1002 842
I
IL+I (-) -74 -238 -461 -555 -596 -561 -468 -345 -234
IL+I final
orig IL+I
-377
134
-461
-455
-754
-899
-993 -1115
-1084 -1167
-1112
-1111
-983
-945
-759
-693
-598
-476
I
TABLE 5. 7 - Sl»1ARY OF ARCli RIB 'lllRusrs
I
(KIPS)

Han:Jer I£x:: -) 0 1 2 3 4 5 6 7 8
I
DL
orig DL
-5240 -4942 -4732 -4536 -4383 -4277 -4198 -4146 -4129
-4736 -4712 -4516 -4131 -4169 -4033 -3925 -3846 -3800
I
SIMP
orig Sup
-1215 -1172
-476 -476
-1125
-458
-1082 -1046
-440 -423
-1017
-409
-998
-398
-983
-390
-981
-385
I
TEMP -93 -88 -84 -81 -79 -77 -75 -74 -74 I
IL+I -907 -878 -845 -801 -785 -763 -749 -739 -736
LL+I final
orig IL+I
-425
-836
-410
-836
-393
-805
-378
-773
-366
-745
-355
-720
-349
-701
-343
-686
-342
-677
I
I
 51

I 'n1e appropriate temperature range to be used is a qualitative

matter to be oonsidered in design which is beyood the srope of
I this study. However, it is s~gested that 100 degrees is excessive.
'n1e range would depend 00 the oolor of the metal parts and 00
the env:irorJrent.
I
5 . 3. 5 Live Lead
I 'n1e live load responses were determined for both nonent and
thrust in the arch ribs . Both maxinun and mininun values are
I given for each in Table 5.5 and 5 . 6 respectively . In this study,
ooincident values were not detel:mi.ned even though they are required
for design . In this case, ooostruction oontrolled the arch rib
I design so it may be unnecessary to examine ooincident resp:xlSeS.

'n1e BRIIXiE-SYS'l»1sm is capable of performir¥J such an analysis

I of ooincident loadir¥J if necessary. It simply saves the load
positions for ooe Ioadil'¥) and awlies than to another influence
surfaoe .
I Values of maxirm.Jn nonent and thrust fran the original design
are also presented for canparisoo . 'n1e final run was made usirv;
I a thinner deck at the tie beams , 1 ' -3", and thicker in the center ,
9. 0" • '!be han;jer cables were doobled in size and the tie beams
were IIOdified as shown in Figure 4. 1 . 'n1e first analysis will
I be referred to as Design 1. 'n1e other as Design 2 . Positive
nonents in Design 1 are generally lower than in the original
design , whereas in Design 2, they were nearly the same . 'n1e
I reasoo is that in Design 1 the deck systan is llllch stiffer and
distribJtes load more evenly CNer the han;jers . nus is ooosistent
with the observatioo of the stagir¥J loads which produoe positive
I nonents only above the unit . Cbnversely , Design 2 permits more
ooncentration of loads since the deck is more flexible . Design
2 closely awroximates the original design values in both senses.
I 'Ilu:usts are similar between Design 1 and the original design.
nus result is understandable in that the same surfaoe of deck
I shalld be loaded in either case and the results are not dependent
00 deck stiffness . 'Ilu:ust in Design 2 is less than in Design

I 1 because the tensioo area is Significantly less and the arch

ribs elcogate more under load . nus is oonsistent with the observed
larger arch nonents in Design 2 .

I 'n1e nonent results are ocrnbined in Table 5.8 for each case
to detel:mi.ne the oontrollir¥J oorditioo which is used to canpute

I arch rib stresses . Table 5. 9 is a similar canpilatioo for thrust .

I
 52 I
TABIE 5. 8 - FACImEIl fo01ENl'S IN AROi RIB I
(K-FT)

Hanger Loc -) 0 1 2 3 4 5 6 7 8 I
DL
SIMP
755
874
1656
132
1817
64
1617 1307
73 68
932
54
553
39
75
28
-62
22 I
TEMP 3151 3174 3177 3176 3173 3170 3167 3163 3159
LL+I final
LL+I final
961
-377
585
-461
880
-754
1111 1222 1201
-993 -1115 -1112
1057
-983
836
-759
693
-598 I
CASE I +
CASE I final
3833
4200
7686
3593
8222
4351
8088
4603
5331
4435
1862
3883
-433 -1540
3060 1945
-1944
1450 I
CASE I -
CASE I final
1941
1302
6733
1325
6337
811
5593
46
2635
-628
-748 -2689
-1127 -1361
-3246
-1510
-3288
-1348
I
CASE II + 7236 11637 11997 11700
CASE II final 7463 7211 7719 7770
8894
7501
5425
6964
3187
6261
2188
5332
1828
4956
I
CASE II - 6101
CASE II final 5724
11065 10866 10203
5851 5595 5035
7276
4463
3859
3958
1834
3608
1164
3260
1021
3277
I
CASE I
CASE I final
3833
4200
7686
3593
8222
4351
8088
4603
5331
4435
1862 -2689
3883 3060
-3246 -3288
1945 1450
I
CASE II 7236 11637 11997 11700
CASE II final 7463 7211 7719 7770
8894
7501
5425
6964
3187
6261
2188
5332
1828
4956
I
I
TABLE 5.9 - FACl'CRED 'ffiRUSl' IN AROi RIB
(KIPS) I
DL -5240 -4942 -4732 -4536 -4383 -4277 -4198 -4146 -4129
SIMP
TEMP
-1215
-93
-1172
-88
-1125 -1082
-84 -81
-1046
-79
-1017
-77
-998
-75
-983
-74
-981
-74
I
LL+I -907 -878 -845 -801 -785 -763 -749 -739 -736

CASE I -10357 -9851 -9445 -9039 -8759 -8535 -8378 -8269 -8238
I
CASE I final -9312 -8836 -8465 -8122 -7851 -7652 -7511 -7412 -7385

CASE II -9692 -9204 -8822 -8450 -8181 -7974 -7826 -7725 -7696
I
CASE IIfinal -9065 -8595 -8234 -7900 -7636 -7444 -7306 -7210 -7184
I
I
I
 53

I '!be stresses in the arch rib are sumnarized for all cases
in Table 5.1 0 . Subsequently , they are factored and canbined

I into the two cases described earlier. It is seen that the arch
rib is slightly overstressed at several locations when the
teIJllerature CCI1ditioo is oonsidered in Design 1. Temperature

I acoounts for the overage in every case . Without temperature ,

the arch is satisfactory with regard to final stresses in this
case . '1be temperature CCI1dition , as stated earlier, is believed

I to be very oonservative . 'Ihere is no overage in the final case.

'1be reasons were discussed earlier.

I
TABLE 5 . 10 - S!M11\R¥ OF S'IRESSFS IN 'lQP AND OOI"ltJl1 OF AROl RIB

I
S1RESS'lQP (KS1)

I cr.sE I -41 -48 -48 -46 -39 -31 -20 -18 -18
CASE I final -38 -35 -36 -35 -34 -32 -30 -27 -26

I cr.sE I I -46 -54 -54 -52 -45 -37 -31 -29 -28
CASE II final -44 -42 -42 -41 -40 -38 -36 -34 -33
I
S1RESS 00l'KM (KSI)
I cr.sE I -23 -13 -10 -9 -15 -22 -32 -33 -33
cr.sE I final -19 -19 -17 -15 -14 -15 -16 -19 -20
I cr.sE I I -14 -2 -0 0 -5 -13 -18 -20 -20
CASE II final -12 -11 -8 -7 -7 -8 -9 -11 -11
I
I 5 . 4 Hangers

5 . 4 . 1 General
I '1be hangers are vertical in both analyses . In Design 1 the
hanger area is 3 . 5 in2 , whereas it is twioe that in Design 2 .
I '1be original design has hanger areas of about 7 in2 . '1bere are
actually foor hanger cables in the actual arran::jement . If the
alternate arran::jement of inclined hangers is used , there will
I be ooly two hanger cables per attachrent.

I
I
 54 I
5.4.2 Dead Load I
'!he deck \IDits were placed 00 the stnx::ture as described
in Sectioo 4. '!he force in the first harger is the weight of I
half of the deck \IDit. Table 5.11 sOOws the harger forces for
each stage of loadirq. sane of the forces cha.n:Je as subsequent
loads are added because of the stiffness that was assured I
effective between the \IDits after they are placed . stage three
is out of the ncmnal sequence because it is the center two deck
\IDits which were placed to balance arch rib lOCIlleIlts . Table 5. 12 I
sOOws the accumulated I'la.rqer dead load forces . In Table 5 . 13
they are carpared to the original dead loads . It is interestirq
that the dead load of the deck in Design 2 is nearly balanced I
by the additiooal steel weight of the original design.

I
TABLE 5 . 11 - HANGER FCRCES FeR DEAD LOAD STlIGING
(KIPS)
I
Har;;er I.oc - > 1 2 3 4 5 6 7 8
I
DL UNIT 1 366 o
DL
DL
UNIT 2
UNIT 8 -5
347
5 347
o I
DL UNIT 3 1 1 347 o
DL
DL
UNIT 4
mIT 5
1 1
1
346
3 345
o
o
I
DL UNIT 6 4 344 o
DL mIT 7 -1 8 341 o I
TABLE 5 . 12 - ACXl.MJLATED HANGER FeRCE FCR DEAD LOAD STAGING
I
( KIPS )

Har;;er I.oc - > 1 2 3 4 5 6 7 8

I
DL
DL
'lUI'AL
'lUI'AL
UN1
lN2
366
366 347
o
o
I
DL 'lUl'AL UN8 361 352 347
DL
DL
'lUI'AL
'lUI'AL
lN3
UN4
362
362
353
354
347
348 346
347
347
I
DL 'lUI'AL UN5 362 354 349 349 345 347
DL
DL
'lUI'AL
'lUI'AL
UN6
1N7
362
362
354
354
349
349
349
349
349
348
344
352 341
347
347
I
I
I
 55

5 . 4 . 3 Superinp:>sed Dead Load

I 'nle hanger forces are nearly equal exoept for the ooe closest
to the arch rib which is slightly less than the others because
the berrling stiffness of the tie beam transmits sane of the load
I directly to the supports .

'nle superimposed dead load is different fran the original

I design because the barriers were not considered in the superim-
posed dead load of the original design .

I 5 . 4. 4 Temperature

I 'nle results are equal for all rut the first hanger . 'nle
force is a result of the arch attempting to lift upward . 'nle
upward novanent is approximately proportional to the stiffness
I of the deck unit. 'nle force in the first hanger is larger for
the sane reason that the first hanger foroe was smaller in the
superinp:>sed dead load case .
I
5 . 4 . 5 Live Load
I Live load was found to be oontrolled by lane load rather
than vehicle load. 'nle triOOtary area may be larger in the
I alternate deck than for the original design where the stress
relief joints tend to limit the trirutary area . Also , the tie
beam in the original design was less rigid . However , the
I differences bet.....een the alternate am original designs are not
large.

I
I
I
I
I
I
I
 56 I
TABLE 5. 13 - SlMlARY CF F'CRCES IN HANGE&S I
(KIPS)

Hanger Loc -) 1 2 3 4 5 6 7 8
I
FINAL DL
Original
362
275
354
295
349
297
349
297
348
300
352
303
341
303
347
303
I
SIMP DL
Original
74
31
90
35
93
36
93
36
93
36
92
36
92
36
92
36
I
10 7 7 7 7 7 7 7
I
lL+I(+) 61 72 74 74 74 74 75 75
Original 55 62 64 64 65 65 65 65 I
CASE I 699 733 735 735 733 737 725 733
CASE II 658 680 680 679 678 682 669 677 I
AREA = 6. 9 SO IN . I
S'ffiESS (KSI)

CASE I 101 106 106 106 106 107 105 106

I
CASE II 95 98 98 98 98 99 97 98
I
5 . 5 Dead IDad Tie Cable and Reactions
I
5. 5. 1 General I
'!he area of each pair of dead load tie cables is 72 inches
square . '!hey are used mainly to carry the thrust of the dead
load of the structure . '!he axial stiffness of the deck and tie
I
beams are large canpared to that of the cables . '!he tensioo
due to live load is small in the tie cable. I
Results are reported for each of the loadir¥Js. It is assuned
that there is ally a tensile force in the cable . It is assuned
fully effective when the first deck unit is placed 00 the structure.
I
'Ibis can be accc:rrplished ally i f it is supported and tightened
under the dead load of the arch rib . Support is supplied by
the harY:)er cables . '!hus its full rrodulus is assuned in all
I
analyses .

'!he results are presented in Table 5. 14 .

I
I
I ...
'~

.~

.
~
57

I 5.5.2 StaginJ of the Deck. Units

I '!he individual and aocumulated results are reported for each

stage of loadirg. '!he results are linear and additive in the
dead load tie cable and reactions.
I
5.5.3 Superimposed Dead Load
I Superimposed dead load thrust is carried by both the tie
cable and the deck units. }b,o,1ever the COl'ICI'ete is assumed to
I be ally Cl'le-third as stiff as for the live load and ~ature
analyses. '!he reactiCl'l in:licates that the alternate design is
slightly heavier than the original design.
I
5.5.4 Temperature
I '!he ass~tiCl'l for this load caused the arch and tie cable

I to beoane lCX'¥jer. '!his caused a reductiCl'l in the dead load force

in the tie cable. '!his load is shown as an increase in the tensile
force in the deck and as increased bendinJ nanent in the arch

I ribs. '!he ~ature difference of 100 degrees Fahrenheit was

rather arbitrary, rut not unreasonable. I.engthenirg of the hargers
also occurred due to the ~ature charge. '!his had a tendency

I to rroderate the uplift Cl'l the deck due to the raisirg of the
arch ribs.

I 5.5.5 Live Load

I Live load causes maximum thrust in the tie cable when the
thrust in the arch ribs are maximum and positive nanent in the
arch is maximum. '!here is no live load that can cause a reductiCl'l

I in the cable tension. Cable thrust is also maximum when the

reaction is naximum. '!hus, the same live loadinJ is applied
for each response. It is interestirg to note that the ratio

I of cable thrust to reaction force is greater for superimposed

dead load than for live load. '!he reaSQ'l for this is that the
tensile stiffness of the concrete in the deck units is greater

I for live load than for superimposed dead load. '1hus, the deck
is assigned a larger portion of the live load than it is for
the superimposed dead load.

I It is interestinJ to note that the reaction is the same as

in the original design.

I
I
 58 I
TABLE 5.1 4 - REACl'IOOS OOE TO STAGING (F DID< UNITS I
(KIPS)

Loadi1'¥] Seq - > 1 2 3 4 5 6 7 8 I

unit -> unl un2 un8 un3 un4 un5 un6 un7

Df:l'AIL -541 -348 -348 -348 -348 -348 -348 -348 I

J\al.IM.JIATED -541 -889 -1237 -1585 -1933 -2281 -2629 -2977
I
Force in the Dead Load Tie cable D.le to stagi1'¥] of Deck Units
(Kips)
I
Toad; 1'¥] Seq - > 1 2 3 4 5 6 7 8
unit -> unl un2 un8 un3 un4 un5 un6 un7
I
lEl'AIL 143 252 674 364 463 547 614 662
ACXlMJLATED 143 395 1069 1433 1896 2443 3057 3719
I
SuIl1nary of Reactions
(Kips)
am Tie cable 'l11rust
I
---DEAD LOAD--- TEMP LLt-I ----SERVICE---- ---S'I'RENGIH---- I
l\RClI STG SIMP CASE I CASE II CASE I CASE II

RFACrIOO 376 2977 718 595 4666 4666 6581 6066

I
Original Total = 3335 594 3929 3929 5625 5625

TIE CABLE 452 3719 168 -1248 87 4426 3178 5829 4131
I
Tie cable stress (ksi) 61.5 44.2 81.0 57 . 4
I
5.6 Tie Beam I
5.6.1 General

'1lle tie beam results are reported in terms of stresses in

I
the botton of the web. Finite element stresses are ocmputed
at only the center of elements, thus it is necessary to rrodify
the results to obtain values at nodes. Results are reported
I
at both the maximum positive m:::ment area midway between har¥Jers
am at maxinum negative m:::ment at the har¥Jers . Results are
reported f= Design 1 at each location as described in Section
I
4 in Table 5.15 .
I
I
I '~
• !:I
..........
.'f')
59

I 'fue dead load results were recanputed by harrl. 'fue values for
the tie beam are not cx:rnpared to the original design values because

I the dead load tie cable causes significant differences in behavior

in the alternate design cx:rnpared to the original.

I Canplete results for Design 2 are reported in Tables 5.16

and 5.17.

I 5.6.2 Dead Load

The stress for maximum and positive dead load rrarent was

I canp.lted using the IT'Cldel shown in figure 5.8.

I 5.6.3 Temperature

Stresses in the oottan of the tie beam web at the hanger

I are reported as
on each side of
the average of the stresses in the two elements
that hanger. Results in the center element between
angers was used for the center lcx:ation.

I /1:)st of the tie beam is in tension. Ibwever, it is of interest

to note that the beam undergoes c:anpression in the center of
I the span. 'lttis is caused in part by the arch rib lifting the
tie beam enoogh to overoane the tension due to thrust. Since
the arch is rrore vertical near the ends of the span, its thermal
I expansion gives a upward berrl near the ends which introduces
a positive rrarent in the tie beam . 'lttis produces a tensile stress
in the oottan which is additive to the thrust stress. In the
I center of the span the negative berrling rrarent produces c:anpress-
ion in the oottan that overoanes the tensile stress due to thrust.
'fue hanger forces are equal except for the first hanger fran
I the support.

II 5.5 Superimposed Dead Load

I SUperimposed dead load, like the thennal loading, mainly

applies tensile force in the tie beam . At no lcx:ation is there
a canpressive stress reported in the oottan of the tie beam.
I stresses reported have been determined in the same manner as
the thermal analySis.

I
I
I
 60
I
I
I
I
116 kp 116 kp 116 k p 116 kp
I
l l J J I
~ ~ I
Hang. r Hanger

t
L: 36'- 6"
~ I
I
-III PL
I
I
·222 PL
I
I
MOMENT DIAGRAM I
DEAD LOAD MOMENT IN TIE -SEAM
I
I
Fig 5.8
I
I
I
I ::;.....r, 61
'....
I 5 . 6 Live Load

I Live load stresses are also reported at mid ler¥Jth of each

element . Results are determined in the same manner as in the
above t\oIO cases.

I Floor beams not at hangers were not loaded in this study;

therefore , the effect of berrling in the tie beams due to local
1 loads is not evident in the results . The aanent due to this
loading case is protably not significant since the span is cnly
36 . 5 feet and the beam depth is 9 feet . '!he small variation
I beb.>een element stresses beb.>een hangers indicates that local
effects are negligible.

I '!he results show a significant increase in tie beam bending

toward the center of the span . '!he reasoo is that the tie beam
is deflecting more there which increases the m::ment. In the
I secood case the deck system was lighter and the m::ments in the
tie beam becarre larger . Ccnversely , the heavier 24 inch deck
resulted in much smaller tie beam stresses than those reported
I for the 15 inch deck.

I TABLE 5 . 15 - TIE BEAM S'rnESSES roI'KM OF WEB AT

FeR DESI~ 1
~

(PSI)
I Hanger Lee -) 0 1 2 3 4 5 6 7 8

I DL
SIMP DL
-205
140
-205
172
-205
172
-205
165
-205
150
-205
140
- 205
135
-205
120
-205
115
TEMPER 447 260 135 13 -90 -175 -235 -280 -302
I LL+I TENS
LL+I m1P
520
-379
583
-360
836
-576
1018
-765
1092
-836
1055
-795
947
-669
799
-434
687
-364

I TIE BEAM S'ffiESSES AT oorn::M OF WEB BE'IWEEN HANGERS

FeR DESI~ 1
I (PSI )

Hanger Lee -) 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9
I DL 103 103 103 103 103 103 103 103 103
SIMP DL 200 225 210 192 179 169 157 159 157
I TEMPER
LL+I TENS
315
440
201
680
74
1210
-41
1098
- 135
1035
-207
895
-260
802
-292
690
-304
550
LL+I m1P -613 -462 -790 -910 -816 -750 -602 -410 - 322
I
I
 62
I
TABLE 5.16 - TIE BEAM S'rnESSES Wl'ltM CF WEB AT H1INGEltS
I
Frn DESI~ 2
(PSII I
Hanger Loc -) 0 1 2 3 4 5 6 7 8

DL -189 -189 -189 -189 -189 -189 -189 -189 -189

I
SIMP DL 302 344 286 229 183 148 122 104 95
TEMPER
LL+I TENS
996
1211
740 471 461
1136 1590 1939 2077
26 -138 -264 -351
2002 1757 1430
-400
1193
I
LL+I mn> -855 -777 -1130 -1553 -1728 -1676 -1441 -1092 -840

SERVICE I£lAD
I
CASE I (+1 1324 1291 1687 1979 2071 1961 1690 1345 1099 I
CASE II (+1 2320 2031 2158 2440 2097 1823 1426 994 699

POST-TENSICNING
I
---------------
REXm.AR -2474 -2474 -2474 -2474 -2474 -2474 -2474 -2474 -2474 I
MAXlMJM crMPRESSI(]II 0i0CK
-------------------------
I
SERVICE I£lAD
I
CASE I (-) -742 -622 -1033 -1513 -1734 -1717 -1508 -1177 -934
CASE II (-I 254 118 -562 -1052 -1708 -1855* -1772 -1528 -1334
I
-2474 -1855 (HANGER 51 = -4329 ) 3200 PSI NG
= -4825
'I'EMP(EARY -2474 + 20% -1855 ) 4400 PSI NG
I
FACICmD POST TENSI(]IIE[) S'ffiESS = 1.3 (-2474 + 20 %1 = -3859 KSI

FJ\CIUm) I
FACl' P-T -3859 -3859 -3859 -3859 -3859 -3859 -3859 -3859 -3859
I
CASE I (-) -1705 -1482 -2322 -3312 -3752 -3684 -3210 -2477 -1942
CASE II (-I 330 154 -730 -1367 -2220 -2411 -2304 -1987 -1734
I
FINAL I -4179 -3956 -47% -5786 -6226 -6158 -5684 -4951 -4416
FINAL
FINAL
II
I
-2144 -2320
-5565 -5341
-3204
-6181
-3841
-7171
-4694
-7611*
-4885
-7543
-4778 -4461
-7069 -6337
-4208
-5801 I
FINAL II -3529 -3706 -4590 -5226 -6080 -6270 -6163 -5846 -5594

ALLOWABLE STRESS = 8000 * .95 = 7600 PSI I

7611 evceecls 7600 psi NG
I
I '~
....,
63

I TABLE 5.17 - TIE BEAM STRESSES Nr a:YrrQ.1 CF

RR lESIrn 2
1m3 BE'l"IIDl ~

(PSI)
I Hanger I.oc -> 0-1 1-2 2-3 3-4 4-5 5-6 6- 7 7-6 6-9

I DL
SIMP DL
94
376
94
350
94
267
94
236
94
195
94
165
94
143
94
130
94
125

I TEMPER
LL+I TENS
LL+I CXMP
671
1063
-713
606 347 126 -60 - 204
1336 1746 1999 2035 1665 1566
-310

-699 -1360 -1657 -1716 -1566 -1255

-376
1249
-929
-409
1033
- 729

I srnVICE LOAD

I CASE I (+)
CASE II (+)
1535
2406
1762
2390
2127 2329
2474* 2455
2324
2264
2124
1920
1605
1495
1473
1095
1252
643

I POST-TENSIrnING
-------------
RmJLAR -2474 -2474 -2474 -2474 -2474 -2474 -2474 -2474 -2474
I
MAXIK.M a:MPRESSICN rnEO<
I -------------------------
srnVICE LOAD

I CASE I (-I
CASE II (-)
-241
630
-455
153
-979
-632
-1327
-1201
-1427
-1467
-1309 -1016
-1513* -1326
-705
-1063
-510
-919

I -2474 -1513 (HANGER 5-6) = -3967 > 3200 PSI NG

'l'EMPCEARY -2474 + 20% -1513 = -4462 > 4400 PSI NG
I FACKRID POST TENSICNED S'rnESS = 1.3 (-2474 + 20 %) = - 3659 KSI

I FACKRID

I FAcr P-T -3659 -3659 -3659 -3659 -3659 -3659 -3659 -3659 - 3659

CASE I (-) -930 -1370 -2451 -3162 -3343 -3060 -2411 -1721 -1294
I CASE II (-I 619 199 -622 -1562 -1933 -1966 -1727 -1406 -1194

FINAL I -4790 -5230 -6311 -7021 -7202* -6919 -6271 -5560 -5153
I FINAL II -3040 -3660 -4661 -5421 -5793 -5626 -5566 -5267 -5054

ALLOWABLE S'rnESS = 6000 * .95 = 7600 PSI

I 7202 psi less than 7600 psi OK

I
 64
I
5.7 Deck
I
5.7.1 General I
Deck stress results are needed to detennine the arrount of
post tensionin:] steel required. All ooncrete is post tensiooed
to avoid tensile stress. Since the deck varies in thickness
I
arrl the stress levels vary across the width of the bridge, it
is necessary to examine deck stresses in several locatiCX'lS across
the width arrl al.cn3 the length of the span. Sane loads enter
I
the deck fran the arch rib arrl must be transferred into the entire
deck through shear actien. Local effects are not a functien
of shear lag.
I
Total stress in the deck is a <XIl\bination of overall thrust
fran the arch rib; overall berrling in the deck units due to flexi-
I
bility of the arch ribs; local berrling between floor beams; arrl
to sane degree bendiI¥J between han;jers. I
Stresses en I::oth top arrl I::ottan of the deck were recorded
aloo:; the three lines "A", "B", arrl "en shc1Hn in Figure 5.9.
Since cnly stresses in the center of each element length are
I
obtained, cusps near ha.rBers are not observed directly fran the
results. Live load influence surfaces were not developed to
examine local effects in the deck. Canp.ltation of local stresses
I
is presented in section 5.7.2.

Post tensionID) forces c:anp.lted in the Tables are dealt with

I
in subsequent sectiCX'lS.
I
Lin. Idontlfl,atlan
I
I
I
I
LOCATION OF RECORDED DECK STRESSES
I
Fig 5.9
I
I
1 ': :-
, r, 65
1-

I 5.7.2 Local Effects

'llie local beOOing due to live and dead loads is ccmputed:

I o:ntinuoos Span = 12'-2" = 12.17 ft
I Dead lDad /obnent:
Deck 'lhlckness = 9.0 in.
Density o:ncrete = 0.150 k/cuft
I Future Wearing Surface = 0.030 ksf

W= 0.150(0.75) + 0.030 = 0.142 k/ft

I Dead I.Dad /obnent =1 WL =1 (0.142)(12.17)(12.17) = 1.75 K-ft
12 12
I HS20 Live lDad /obnent = 0.900L K-ft for simple span AASHTO
II 3.2.33 (For o:ntinuous Span Use 80 percent of the simple
span value).
M = (0.80)(0.900)(12.17) = 8.76 K-ft
II Secticn M::ldulus of Deck

= 1 bt = 1(1)(0.75) = 0.0351
I I

S
12 12
= I/(t/2) = 0.0351/0.375 = 0.09375
ft

ft
I Dead I.Dad stress

= M/s = 1.75/0.09375 = 18.66 = 130

I f

Live I.Dad stress incllrling 30 percent impact

ksf psi

I f = 1.30(8.76/0.09375) = 121.50 ksf = 844 psi

= 130 = 974 psi

I 'lbtal Local Deck stress

For deck thickness = 1.25 ft:

+ 844

I W = 0.150(1.25) + 0.030 = 0.2175k/ft

=1 = 1(1.0)(1.25) = 0.260
I s
6
bt
6
ft

Dead lDad /obnent =1 WL = 1(0.2175)(12.17) = 2.68 K-ft

I 12 12

Dead lDad stress = 2.68/0.260 = 10.32 ksf = 72 psi

I f Live I.Dad = 1.30(8.76/0.260) = 43.80 ksf = 304 psi

I 'lbtal Local Deck Stress = 72 + 304 = 376 psi

 66 I
Deck stresses are reported in Tables 5.18 through 5.23.
I
'!he discussion below relates to these tables. '!he tables are
broken into two parts so that the locations at the l'la.rBers arrl
between the l'la.rBers may be reported clearly. The tanperature
I
arrl superimposed runs were not observed to be different enough
to warrant the use of different values in positive arrl negative
m::ment areas of the deck.
I
The results for superimposed dead load arrl temperature runs
are shown in Figures 5.10, 5.11 an::! 5.1 2 . Stresses are plotted
I
for the entire bridge width for each oorxlition. In each figure
there is a plot of the stress at the em of the span, the quarter-
point arrl the mid-point of the span. Figures 5.1 0 arrl 5.11 shc7.I
I
stresses in the top of the deck for DeSign 1 arrl 2, respectively.
Figure 5.12 shows stress levels in the bottom of the deck for
Design 2.
I
5.7.3 Dead Load
I
Dead load stress due to the weight of the units is ClCITp.lted
as if the structure was canpleted before the dead load was placed
I
on the structure. '!he deck is not affected by thrust or overall
beOOing due to its own weight since the dead load tie cable carries
thrust fran the arch rib arrl the deck units are not attached
I
to the arch ribs at that tiIre.
I
5.7.4 Superimposed Dead Load

Superimposed dead load is applied after the deck units have

I
becane part of the integrated structure. It may be thought of
as having two ~ts; the thrust arrl the local beOOing.
'!he stresses reported are results of the oenter element between
I
l'la.rBers. '!he one exception is the stress at the arch rib. 'Ibis
value is the stress in the element closest to the arch. '!he
oenter element is used because it is at the max:iJlulm positive
I
m::ment location with respect to local berxling. All values
represent a CXI1lbination of the two canponents; thrust arrl local
effects.
I
The results show clearly that the thrust is the daninate
cullpollent. At the em of the span, the stress is highest where
I
the thrust enters the deck. In the middle of the deck, the stress
is small. At the quarter point arrl at mid-span the stress is
much lower in Line "A" than at the other locations. 'Ibis is
I
due to the thicker deck at these points. Shear has distribJted
the thrust rather uniformly over the deck by the quarter- I
point.

I
I ...~ 67
STRESS AT TOP OF DECK DUE TO SOL AND TEMPERATURE
"
I
I
I
I 150 A LIne B

I 100

I 50

O~
Line C /
______________________________ ~

I SUPE'RIMPOSED DEAD LOAD

,

I
LI ne A
I
I
250

200
/
I 150
, ..,...~ Llne 8

I ..
0.
100
Line C - -·\1
I
~

50

I o

I - 50

I TEMPERATURE

I
Fig 5.10
I
 68
STRESS AT TOP OF DECK DUE TO SOL AND TEMPERATURE
I
~-
-:--
r---. . . . . .
I
/<"'"
"',,- I
~
150
Line A Line B
'" I
I
.
~

0-
~
100 \ I

0
LI ne C / I
SUPERIMPOSED DEAD LOAD
I
300
Li ne A
I
250
\ I
-..
~
200
Line. B
I
0-
~
150
I
100 II ne C
I
50
II
0

-50
I
- 100
I
TE MPER ATURE
I
Fig 5.11 I
I
1'....1> 69
STRESS AT BOTTOM OF DECK DUE TO SOL AND TEMPERATURE

1
I
I
I 2 00

I 150 LI ne A LIne B

I .
~
100

-
~

50
I 0
LI ne C

I
SUPERIMPOSED DEAD LOAD

I
L i ne A
I 250

I 200

I 150

I
'"
-
~
100
L i ne 8 LI ne C
/
I 50

0
I
I TEMPERATURE

I Fig 5.1 2

I
 70 I
5.7.5 Tanperature I
'1lle results for this analysis were obtained in the same manner
as for super~ dead load. This run did not produce observable I
variation in deck stresses between hangers since there were no
vertical loads applied. Temperature has t'n'O apparent effects
en the deck: '1lle arch is attempting to lift the deck U~i I
and the arch is attempting to lengthen the deck due to expansion
roth longitudinally and vertically of the arch ribs and l~
of the dead load tie cable. In the analysis, the tie cable actually I
pulled on the deck whereas is the actual structure, it would
anount to a transfer of the dead load thrust fran the cable to
the deck and tie beam units. I
As in super:iJrposed dead load, there is a large difference
is stress across the deck at the ends of the span. This occurs I
because the force in the deck is introduced by the arch rib and
dead load tie cable. In fact, shear lag is so severe that the
center of the deck reverses sign to maintain equilibriun. At
I
the quarter point of the span tensile force is equal across the
bridge width and remains equal over the center portion.
I
5.7 . 6 Live I£>ad
I
'1lle local effects were cc::Ilputed by haOO as was local dead
load stress. The overall stresses were determined fran the loader
which positicns the live load to produce maximum and minimum
I
stresses in each elarent investigated . '1lle reported values
are the average of the t'n'O elarents en each side of a hanger .
'!be influences surfaces are produced by loading only positions
over floor beams at hangers, thus the local effects of bending
over floor beams are not evident in these results. '1lle haOO
calculated local stresses are added to the stresses determined
by the loader.
I
I
I
I
I
I
I .!)
'~
·n
71
.-
~

I TABIE 5. 18 - !lEO< S'ffiESSES 'roP LINE "A"

(PSI)

I Ranger Lac -) 0 1 2 3 4 5 6 7 8

SIMP DL 152 63 72 73 77 78 80 80 80
I TEMPERA'IURE
lL+I TENS
217 236
242 352
167
458
190 220
551 580
243
570
260
508
272 277
410 309
lL+I aMP -68 -299 -453 -555 -590 -565 -492 -391 -299
I SERVICE LOAD WI'llU1l' LCX:AL
CASE I (+) 394 415 530 624 657* 648 588 490 389
I CASE II (+) 611 651 697 814 877 891* 848 762 666

SERVICE LOAD LCX:AL

I DL LCX:AL (+)
lL+ I LCX:AL (+)
72
304
72
304
72
304
72
304
72
304
72
304
72
304
72
304
72
304
SERVICE

I CASE I & II 376 376 376 376 376 376 376 376 376

PQ;T-TENSICNING REQJIRED f"'ffi LCX:AL BENDING CASE 1;1

I -------------------------------------------
RmJLI\R -891 -891 -891 -891 -891 -891 -891 -891 -891
LCX:AL -376 -376 -376 -376 -376 -376 -376 -376 -376
I 'rol'AL
-----------------------------------------------------
-1267 -1267 -1267 -1267 -1267 -1267 -1267 -1267 -1267

I MAXDI.M aMPRESSICN 0ID:l<

-------------------------
SERVICE LOAD WI'I1UJT LCX:AL
I CASE I (-)
CASE II (-)
84
301
-236 -381
0 -214
-482 -513* -487
-292 -293* -244
-412 -311
-152 -39
-219
58

I Case I
'l'EMPCIlARY
-1033 -293 (HANGER 4)
-1033 + 20% -513
= -1326
= -1753
(
(
3200
4400
PSI
PSI
OK
OK
case II -1267 -293 (HANGER 5) = -1560 ( 4000 PSI OK
I 'l'EMPCIlARY -1267 + 20% -293 = -1813 ( 5500 PSI OK

FACItRED PQ;T TENSICNED VAWE = 1. 3 (-1267*1.20) = -1977

I S'ffiENG'lll 0ID:l<
FACr P-T -1977 -1977 -1977 -1977 -1977 -1977 -1977 -1977 -1977
I CASE I (-)
CASE II (-)
50 -566 -889 -1108 -1178 -1122 -963 -743 -544
391 0 -279 -380 -381 -317 -198 -51 75

I (LCX:AL)
CASE I -753 -753 -753 -753 -753 -753 -753 -753 -753
CASE II -489 -489 -489 -489 -489 -489 -489 -489 -489
I FINAL I -2680 -3296 -3618 -3837 -3907 -3852 -3692 -3472 -3273
FINAL II -2075 -2466 -2745 -2846 -2847 -2783 -2664 -2516 -2391
I
 72
I
TABLE 5. 19 - OED< STRESSES a:Jl'l'Oo1 LINE "A"
(PSI) I
~er lDc-> 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9
I
SIMP DL 162 41 49 52 56 59 61 63 63
TEMPERA'lURE
LL+I TENS
217
223
205
305
165 177
387 453
199 214
483 467
226
426
234
352
238
266 I
LL+I mtP -39 -250 -367 -440 -463 -440 -387 -309 -232

SERVICE IJ:W) (NO LOCAL) I

CASE I (+) 385 346 436 505 539 526 487 415 329
CASE II (+) 602 551 601 682 738 740 713 649 567
I
LOCAL BENDING
DL LOCAL (+1
LL+I LOCAL (+)
72
304
72
304 304
72
304
72 72
304
72
304
72
304 304
72 72
304
I
SERVICE
CASE I & II 376 376 376 376 376 376 376 376 376
I
POOT-TENSIOOING
---------------
RmJLAR -891 -891 -891 -891 -891 -891 -891 -891 -891
I
LOCAL -376 -376 -376 -376 -376 -376 -376 -376 -376

'lUl'AL
-----------------------------------------------------
-1267 -1267 -1267 -1267 -1267 -1267 -1267 -1267 -1267
I
MAXDlJM ru!PRESSIOO OlEn<
-------------------------
I
SERVICE IJ:W) (NO LOCAL)
CASE I (-I
CASE II (-)
123 -209
340 -4
-318
-153
-388 -407* -381
-211* -208 -167
-326
-100
-246
-12
-169
69
I
case I
~y
-1033
-1033
-407 (HANGER 4)
+ 20% -407
= -1440 < 3200 PSI
= -1647 < 4400 PSI
OK
OK
I
case II -1267 -211 (HANGER 5)= -1478 < 4000 PSI OK
~y -1267 + 20% -211 = -1731 < 5500 PSI OK I
TENSIONED VALUE = 1.3 (-1267 + 20 = -1977
FACTORED POOT

FACKRED I.OI\OO
%)
II
FAcr P-T -1977 -1977 -1977 -1977 -1977 -1977 -1977 -1977 -1977
CASE I (-)
CASE II (-)
126 -489 -732 -885 -931 -876 -759 -587 -421
442 -6 -199 -274 -271 -217 -130 -15 89
I
(LOCAL)
CASE I -753 -753 -753 -753 -753 -753 -753 -753 -753
I
CASE II -489 -489 -489 -489 -489 -489 -489 -489 -489

FINAL I -2603 -3219 -3461 -3614 -3660*-3605 -3488 -3317 -3151

I
FINAL II -2024 -2472 -2665 -2740 -2737 -2682 -2595 -2481 -2377
I
I ..,..
'~
.!>
73
...
-10

I TABLE 5.20 - DEO< Sl'RESSES TOP LINE "B"

(PSI)

I Hanger Lac - > 0 1 2 3 4 5 6 7 8

SIMP DL 12 87 87 89 90 92 93 93 93
I TEMPERA'IURE
LL+I TENS
7
58
8 149 191
203 334 401
219 242
426 406
259
357
270
285
276
223
LL+I a:MP -10 -246 -377 -453 -473 -453 -396 -319 -256
I SERVICE LOAD (NO IJX:AL )
CASE I (+) 70 290 421 490 516 498 450 378 316
I CASE II (+) 77 298 570 681 735 740* 709 648 592

IJX:AL BENDING

I DL IJX:AL ( +)
LL+I IJX:AL (+)
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844
SERVICE

I CASEI&II 974 974 974 974 974 974 974 974 974

PC6T-TENSIOOING
I ---------------
RmJLAR -740 -740 -740 -740 -740 -740 -740 -740 -740
IJX:AL -974 -974 -974 -974 -974 -974 -974 -974 -974
I 'lUI'AL
-----------------------------------------------------
-1714 -1714 -1714 -1714 -1714 -1714 -1714 -1714 -1714

I MAXIM.JM CDlPRESSIOO
-------------------------
OID::K

SERVICE LOAD (NO IJX:AL)

I CASE I (-)
CASE II (-)
2 -159
9 -151
-290
-141
-364 -383* -361
-173* -164 -119
-303
-44
-226
44
-163
113

I case I
~y
-1714 -383 (HANGER 4)
-1714 + 20% -383
= -2097
= -2440
<
<
3200
4400
PSI
PSI
OK
OK
case II -1714 -173 (HANGER 5) = -1887 < 4000 PSI OK
I ~y -1714 + 20% -173 = -2230 < 5500 PSI OK

FACl'CRID PC6T TENSIOOID VALUE = 1.3 (-1714 + 20 %) = -2673

I FACl'CRID LOADS
FAcr P-T -2673 -2673 -2673 -2673 -2673 -2673 -2673 -2673 -2673
I CASE I (-)
CASE II (-)
-5 -420 -704 -867 -908 -863 -738 -569 -434
12 -197 -183 -226 -213 -155 -58 58 147

I (IJX:AL)
CASE I -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997
CASE II -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266
I FINAL I -4676 -5091 -5374 -5537 -5578 -5533 -5408 -5240 -5104
FINAL II -3927 -4136 -4122 -4165 -4152 -4094 -3997 -3881 -3792
I
 74
I
TABLE 5.21 - OED< STRESSES a::::tI'KM LINE ''Btl
(PSI) I
Hanger I.oc - > 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9
I
SIMP DL 24 59 62 66 70 73 74 76 76
TEMPERAWRE
LL+I TENS
-10
29
184 164
227 256
172 184 195
285 299 294
203
270
208
250
211
223 I
LL+I <nIP -63 -88 -154 -193 -207 -198 -164 -125 -92

smVICE LOI\D (NO LOCAL) I

CASE I (+) 53 286 318 351 369 367 344 326 299
CASE II (+) 43 470 482 523 553 562 547 534 510
I
LOCAL IDIDING
DL LOCAL (+)
LL+I LOCAL (+)
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844 I
smVICE
CASE I & II 974 974 974 974 974 974 974 974 974
I
PCST-TENSICNING
---------------
RmJLAR -740 -740 -740 -740 -740 -740 -740 -740 -740 I
LOCAL -974 -974 -974 -974 -974 -974 -974 -974 -974

'lUl'AL
-----------------------------------------------------
-1714 -1714 -1714 -1714 -1714 -1714 -1714 -1714 -1714 I
MAXlMJM a:MPRESSICN <lim<
------------------------- I
smVICE LOI\D (NO LOCAL)
CASE I (-)
CASE II (-)
-39 -29
-49* 155
-92
72
-127
45
-137* -125
47 70
-90
113
-49
159
-16
195
I
case I
~y
-1714
-1714
-137 (HANGER 4)
+ 20% -137
= -1851
= -2194
(
(
3200
4400
PSI
PSI
OK
OK
I
case II -1714 -49 (HANGER 5) = -1763 ( 4000 PSI OK
~y -1714 + 20% -49 = -2106 ( 5500 PSI OK I
FACTORED PCST TENSIONED VALUE = 1.3 (-1714 + 20 %) = -2673
FACImED LOAC6
I
FACT P-T -2673 -2673 -2673 -2673 -2673 -2673 -2673 -2673 -2673
CASE I (-)
CASE II (-)
-104 -113 -254 -333 -358 -333 -259 -172 -100
-63 202 93 58 61 92 147 206 254
I
(LOCAL)
CASE I -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997
I
CASE II - 1266 -1266 -1266 -1266 - 1266 -1266 -1266 -1266 -1266

FINAL I -4775 -4783 -4924 -5003 -5028 -5003 -4930 -4843 -4770
I
FINAL II -4002 -3737 -3846 -3881 -3878 -3848 -3792 -3733 -3685
I
 .::> 75
I .....-
'0 TABLE 5.22 - DEn< S'IRFSSFS 'lOP LINE "C"
I (PSI)

Hanger Loc -) 0 1 2 3 4 5 6 7 8
I SIMP DL 17 50 74 80 83 84 85 86 86
TEMPERA'lURE -38 -89 130 188 218 241 257 269 275
I IL+I TENS
IL+I a::MP
78 96
-29 -242
193 252 270 256
-324 -381 -396 -371
217 164 131
-334 -280 -246

I smVICE LOAD (NO I.O:AL )

CA'>E I (+) 95 146 267 332 353 340 302 250 217
CA'>E II (+1 57 57 397 520 571 581 559 519 492
I I.O:AL WIDING
DL I.O:AL (+ I 130 130 130 130 130 130 130 130 130
I IL+ I I.O:AL (+ I
smVICE
844 844 844 844 844 844 844 844 844

CA'>E I & II 974 974 974 974 974 974 974 974 974
I FOOT-TENSIauNG
---------------
I RmJLAR
I.O:AL
-615
-974
-615
-974
-615
-974
-615
-974
-615
-974
-615
-974
-615
-974
-615 -615
-974 -974
-----------------------------------------------------
I 'lUl'AL -1589 -1589 -1589 -1589 -1589 -1589 -1589 -1589 -1589

MAXIMJM <n!PRffi'>IGl 0iE0<

I -------------------------
smVICE LOAD (NO I.O:AL)
CA'>E I (-I -12 -192 -250 -301 -313* -287 -249 -194
I CA'>E II (-) -50 -281* -120 -113 -95 -46 8 75
-160
115

= -1902
I Case I
'1'EMl'C:RARY
-1589
-1589
-313 (HANGER 41
+ 20% -313 = -2124
<
<
3200
4400
PSI
PSI
OK
OK
Case II -1589 -281 (HANGER 51 = -1870 < 4000 PSI OK
= -2188
I TEMPCRARY -1589 + 20% -281 < 5500 PSI OK

FACTORED FOOT TENSIONED VALUE = 1. 3 (-1589 + 20 %1 = -2478

I FACItmD I.CllIOO
FACT P-T -2478 -2478 -2478 -2478 -2478 -2478 -2478 -2478 -2478
I CA'>E I (-)
CA'>E II (-I
-41 -459 -606 -722 -751 -695 -613 -494 -422
-65 -365 -156 -147 -124 -60 11 98 149

I (I.O:ALI
CA'>E I -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997 -1997
CASE II -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266
I FINAL I -4517 -4935 -5081 -5197 -5226 -5171 -5088 -4969 -4897
FINAL II -3809 -4110 -3900 -3891 -3868 -3804 -3734 -3646 -3595
I
 76
I
TABLE 5.23 - DEO< S'ffiESSES roI'l'CM LINE "e"
(PSI) I
Hanger I.oc -)

SIMP DL
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9
I
19 38 60 68 73 75 77 78 78
TEMPERA'IURE
ILtI TENS
-5
19
170 136 175 197
270 299 319 328
212
328
224
314
232 235
295 280 I
ILtI a:MP -58 -92 -159 -207 -217 -202 -174 -140 -121

SERVICE l.OI\D (NO I.CX:AL) I

CASE I (+) 38 308 359 387 401 403 391 373 358
CASE II (+) 33 478 495 562 598 615* 615* 605 593
I
I.CX:AL BENDING
DL I.CX:AL (+)
ILt I I.CX:AL (+)
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844
130
844 I
SERVICE
CASE I & II 974 974 974 974 974 974 974 974 974
I
FOOT-TENSICNING
---------------
REX;ULAR -615 -615 -615 -615 -615 -615 -615 -615 -615 I
I.CX:AL -974 -974 -974 -974 -974 -974 -974 -974 -974

'IOTAL
-----------------------------------------------------
-1589 -1589 -1589 -1589 -1589 -1589 -1589 -1589 -1589 I
MAXIM.JM CDlPRESSICN OlEn<
------------------------- I
SERVICE l.OI\D (NO I.CX:AL)
CASE I (-)
CASE II (-)
-39 -54
-44* 116
-99
37
-139
36
-144* -127
53 85
-97
127
-62
170
-43
192 I
case I
TEMPCRARY
-1589
-1589
-144 (HANGm 4) = -1733 <
+ 20% -144 = -2051 <
3200
4400
PSI
PSI
OK
OK I
Case II -1589 -44 (HANGm 5) = -1633 < 4000 PSI OK
TEMPCRARY -1589 + 20% -44 = -1951 < 5500 PSI OK
I
FACTORED FOOT TENSIONED VALUE = 1.3 (-1589 + 20 %) = -2478

FACTORED IJ:lI\OO I
FACf P-T -2478 -2478 -2478 -2478 -2478 -2478 -2478 -2478 -2478
CASE I (-)
CASE II (-)
-102 -150 -266 -361 -375 -339 -277 -203 -161
-58 151 49 46 69 111 165 220 250
I
(I.CX:AL)
CASE I -1998 -1998 -1998 -1998 -1998 -1998 -1998 -1998 -1998
I
CASE II -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266 -1266

FINAL I -4578 -4626 -4742 -4837 -4852 -4816 -4753 -4679 -4637
I
FINAL II -3803 -3594 -3696 -3699 -3676 -3634 -3580 -3525 -3495
I
 .
I ·~
1.:0
.)
77

5.8 Post Tensioning Q:rnp..ItatialS

1 5.8.1 General

Design of p:>St tensioning is based on section 8 of the AASmO

1 Brid3e Specification. It is assurred that no creep or shrinkage
have been raroved fran the units prior to tensioning. This is
a o::nservative assumption with regard to a l.aI:ge brid3e.
1 CASE I includes dead and live loads. CASE II includes dead,
live and terrperature loads. All allowable stresses for CASE
1 II are increased to 125 percent of the allowable according to
AASHrO Table 3.22.1A. (Ref. 5) •

I Post tensioning is designed for local berrling in the deck.

I.ocal berrliD;J includes dead and live load rocrnents ocmputed l::1t
assll1liD;J that floor beams are rigid. This strand is draped to
I follow the local rocrnent envelope. Post tensioniD;J applies maximun
cx:mpression where the local berrling produces tensile stress.
en the other side of the deck, the same p:>St tensioniD;J produces
1 no cx:mpression. 'Ihus, the local cx:mpression does not have to
be added to its p:>St tensioning stress when checki.r¥3 for crushiD;J.

I Overall stress in the deck due to thrust and beroing of the

entire deck unit is assurred to be OCl'lStant across the deck when
designing lX'st tensioning to resist it. '!he maximun value fran
I either the tq;> or botton is used . This p:>St tensioning is p:>Sit-
ioned in the center of the deck.

I Strand in the botton flaD;Je of the tie beam is designed based

on tension and cx:mpression in botton of the web element. It
is assuned OCl'lStant over the entire span as is p:>St tensioning
1 stress in the deck. 'lbere is no design presented for strand
in the web.

I '!he following paraneters are used in designing p:>St tensioning:

Cbncrete -
I fc' = 8000 psi
Density = 150 pef
I Young's Modulus = 5700 ksi
Allowable strand anchorage = 3000 psi.

I Post Tensioning Steel -

Ultimate strength = 270 ksi

I Yield Stress = 216 ksi
YOUD3"'s Modulus = 28000 ksi

I
 78 I
5.8.2 Deck I
Post tensioning in the deck is divided into two parts. Q1e
part is used to IX'St tension the deck for local berrling stresses I
due to dead and live load. '1llee seo:xld part is used to post tension
for overall tensioo forces in the deck. CNerall forces are due
to superimposed dead load, tanperature and live loads. Post I
tensiooing for local m::rnents is placed eccentrically ooe sixth
the depth of the deck toward the tensile stress. At floor beams
the local berrli03 m::rnent is negative so the strand would be located I
t/6 above the center of gravity of the deck. By p.1tting strand
at this locatioo, a canpressive stress is irrluced on ooe side
equal to (2t/9)*Force. 'llle other side has zero cx:mpression.
I
'llle ranainin:J strand is placed at the center of the deck.

'llle canpressive strength of the deck ITBy rx>t be exceeded

I
at the time of p:lst tensionin:J. 'lhls limit is based of a phi
factor = 0.95 and load factors appropriate factors plus IX'St
tensiooing stress with a load factor of 1.3. Allowable tensile
I
stress is zero.

Allowable stress of O. 55fc I for the unfactored loads must

I
not be exceeded prior to losses in p:lst tensionin:J. 'Ihis incltrles
the sun of any canbinatioo of local and overall stresses and
post tensioni03 forces.
I
'llle allowable stress of O. 4fc I for unfactored loads must
not be exceeded after losses have occurred.
I
No tensioo ITBy
loads.
occur in concrete under any canbinatioo of
'lhls is simply to insure that the joints do not open.
I
strand stress shall rx>t exceed 0.7*ult1JTBte strength of the
strand when it is installed.
I
strand must not exceed 0.8*Yield of the strand under perJTBnent
loads after losses.
I
Loss of post tensionifB shall incltrle ooosideration for the
following factors:
I
Shrinkage - SH
Creep of Cbncrete - CRc
I
Creep of steel - CRs
Elastic shortening ES I
In addition, friction losses in strand ducts shall be
considered • I
I
I :;,.,'-'
79
I-
Steel Allowables

I Post tensioning must overcane all tension stresses in the

concrete without failin;J the ooncrete in the CCIlpression regicn.

I AASH'IU-Secticn 9 is used to check allowable stresses.

Assumpticns:

I Para. 9.1 3.2.1 Strains vary linearly over depth of secticn.

Para. 9.14 phi = 1 . 0 for factory produced ooncrete.
Para. 9.15.1 TaTIporary stress before losses due to

I creep am shrinkage = 0.70 * stren:Jth of prestress steel

0.70 * 270000 = 189000 psi

I stress at service load = 0.80 * yield strength of post tensioning

steel.

I = 0.80 * 216000 = 172800 psi

Para 9.15.2.1 Post tens oonc (Before losses) . = 0.55 * stren;Jth

I = 0.55* 8000 =4400 psi

I Para 9.51.2.2 Stress at service load

Comp.= 0.40 * strength = 0.40 * 8000 = 3200 psi

I Tensicn = 6*(stren;Jth)**0.5
bomed reinf. case = 6 (8000) **0.5 = 537 psi

I Para 9.15.2.4 Anchorage = 3000 psi

Para. 9.16.1 Friction loss

I Wire galva. metal sheathin;J = 0.0002 K/ft
To = Tx( 1 +KL)
I L = Span/2 Because jack.ir¥J will oc= fran both ends.
To = Tx(1 + 0.0002*310) = Tx(1.062)

I Para. 9.16.2.1

SH shrinkage = 0.80(17000 - 150RH)

I = 0.80(17000 - 150*70)
ES elast. short. = 0.5(Es/Eci)*fcir
= 5200 psi

= 0.5(28/5.4)*1589 = 3896 psi

I
I
I
 80
I
CcXlcrete Creep
I
CRc = 12*fcir - 7*fcds
fcir = concrete stress at the center of gravity of the prestre-
I
ssil'¥3 steel due to prestressil'¥3 force and dead load of beam inmed-
iately after transfer; fcir shall be oanputed at the section
or sections of maximum narent. (At this stage, the initial stress
I
in the tendm has been reduced I:7t elastic shortening of the concrete
and teOOon frictial for post-tensioni~ members. The reductions
to initial tendal stress due to these factors can be estimated,
I
or the reduced tendm stress can be taken as 0.63fs' for typical
pretensiaJed members.) I
= 12*1500 = 18000
Creep of steel
CRc - 7*0
I
CRs = 20000
for 270 ksi
- 0.3FR - 0.4ES -0.2(SH + CRc)
strand
I
= 20000
= 10587
- 0.3(10714) - 0.4(3896) -0.2(5200 + 18000)
psi I
Para. 9.17.4

Allow. Steel Stress in Bonded members

I
fsu* = fls *(1 - 0.5(p*f l s/f lc)
effective prestress after loss not less than 0.5f IS I
say p* = 0.005 or 1/2 percent
fsu* = 270000(1 - 0.5(0.005*270000/8000) = 247219 psi
I
Total loss equals:
I
SH + ES + CRc + CRs

5200 + 3896 + 18000 + 10714 = 37810 psi I

Percent Loss = 37810/189000 = 20 percent
I
Friction 0.062

I
I
I
I
, ....
.... 81

, Design of post tensioniN3 in deck

Phi = 0. 95 ; Strergth = 7600 psi

I strergth

I
, CASE I
CASE II
0. 95*8000 > 1. 3(DL + 5/3 (LL+I ))
0.95*8000 > 1.3(DL + T + (LL+I))

, steel 0. 95*270000 = 256500 psi

,
Service

Percent loss = Estimated at 20%

Before l osses
I CASE I
o::ncrete
I Ccmpression
0. 55*8000 = 4400 > DL + (LL+I) + Post tension*1.20
Tension = 0. 0
I CASE II
a:ncrete
I Ccmpression
0. 55*8000*1 . 25 = 5500 ) DL + (LL+I) + T + Post tension*1.20
Tension = 0.0
I steel 0. 7*270000 = 189000 psi before losses

I
, After Losses

o::ncrete

CASE I Ccmpression 0. 40*8000 = 3200 > DL + (LL+I) + Post tension

I
, CASE II 0. 40*8000*1 . 25 = 4000 > DL + (LL+I) + T + Post tension

II Steel 0. 8*216000 = 172800 psi after losses

I
 82 I
losses.
~termine the level of stress permitted in the stram before I
Final stress must be less than 172800 psi. I
Loss of prestress in steel = 37810 psi.

Before loss limit = 172800 + 37810 = 210610 psi

I
Absolute limit = 189000 psi I
Since 210190 ) 189000, limit = 189000 psi

Stress in stram after losses = 189000 - 37801

I
= 151199 psi

Area of 1-270 ksi 1/2" dia stram = 0.153 in2

I
I
Line "A" t = 15"
I
Local berxling stress = 376 psi top am oottan.
Positioo stram at 15/6 = 2.5 " al:ove center of deck.
Allowable stress in steel = 151199 psi Area ale 1/2" dia
I
stram = 0.153 in2
I
Stress due to ~t tensicnirg placed at 1/6 deck thickness fran
center of gravity en side of tensien stress. nus stram is
to be draped to match local beOOing rrarents. I
f = 2F/t
F required = (t/2)*f = 15/2*(376) = 2820 pounds/inch I
or 2820 * 12 = 33840 pounds/ft

No. stram reqd = 33840/(0.153*151199) = 1.45 I

Adjust for 6.2% frictien loss.
1.45 * 1.062 = 1.54
say 1.5 stram/ft. I
Stram for axial forces in deck
I
Critical case - After losses = 1267 psi (CASE II)
to avoid any tensien in deck.
No. stram = 1267*15*12/(0.153*151199) = 10.06 I
Adjust for 6.2% frictien.
10.06 * 1.062 = 10.68
= say 11 stram/ft I
I
 83

f Line "B" t= 9"

I Ux::al bending stress =

974 psi top am lx>ttan.
Positien strand at 9/6 = 1.5 " above center of deck .
I Allowable stress in steel = 151199 psi
Area = 0 . 153 in2

I Stress due to post tensioning placed at 1/6 deck thickness fran

center of gravity en side of tensien stress . This stram to
be draped to match local bending nx:ments .
I f = 2F/t
F required = (t/2)*f = 9/2*(974) = 4383 pounds/inch
I or 4383 * 12 = 52596 pounds/ft

No. stram reqd = 52596/(0 . 153*151199) = 2. 3

I Adjust for 6 . 2% frictien .
2 . 3 * 1 . 062 = 2.4 ; Say= 2. 5 stram/ft

I stram for axial forces in deck

Critical stress after losses = 1714 psi (CASE II)

I No . strand = 1714*9*12/(0 . 153*151199) = 7.98
I Adjust for 6 . 2% friction .
7 . 98 * 1 . 062 = 8 . 47
= Say 8.5 strand/ft

I Line "e"
"e"
I Since line
same.
is nearly the same as line ''B'', make the

----------------------------------------
I Total stram weight in deck

I Line
A
Ux::al
1. 5
Overall
10
Total
11.5
B 2. 5 8. 5 11.0

I e 2.5
Total
8.5 " .0
33 . 5 strand/ft

I Average 33 . 5/ 3 = 11 .17 stram/ ft

Weight = 0 . 153*3 . 4*11.17*91*620 = 327836 pounds

I
I
 84
I
5. 8. 3 Post Tensionio:; in Tie Beam Botton Flange
I
Maximum required post tensioning stress before losses equals
2474 psi.
I
Force = 2474 * 2 * 144 = 712500 pounds
Allowable strand stress = 151199 psi
I
Reqd No. strand = 712500/(. 153*151199) = 30 . 8 strand
Adjust for 6 . 2% friction .
30 . 8 * 1 . 062 = 32 . 7
I
= Say 33 strand

Since botton flange is overstressed in canpression, the flange

I
must be made larger than 2 square feet. 'lhls should reduce the
stresses to a point that the 33 strand will be satisfactory. I
Required weight of strand in flanges of tie beam

33 * 0. 153 * 3.4 * 620 * 2 = 21287 pounds

I
TOtal strand weight = 327836 + 21287 = 349122 pounds I
Cl1eck anchorage area in deck.
Design anchorage stress = 3000 psi.
I
Maximum force = 151199 * 11 * 0.153 = 270244 pounds/ft. I
or 270277/(9*12 ) = 2502 psi < 3000 psi OK

No extra ooncrete required for anchorage . However , it is

I
recx:mnended that the end sections be thickened to simplify posit-
ienio:; post tensicnin3' anchors . I
Cl1eck cast-in-place end sectiens . %ese sections must be cast I
and post tensioned after the precast units have been post
tensioned . Since sectien to be be post tensionio:; will cause
a tensile force in the remainder of the deck prior tensiooing I
will tend to be relieved. 'Il1e cast-in-place sectien is 10 feet
100:;. 'lhls will cause an increase in the strand stress of 10/ 620
= 1. 5 % if the last stressio:; is of the same magnitwe as in I
the precast units. Additional post tensioning foroe will be
required to overcane thrust and berxlio:; in the arch ribs. 'lhls
additicnal post tensioning will be a larger percentage en shorter I
bridges.
I
I
I,
85

I
other Ccnsideratioos

I It is clear that a lower strergth coocrete is acceptable

for the deck in the study am that it WOJld not have a large
I effect on the tension flanges of the tie beams .

Q1e coold interpret specification on the ~ature condition

I to rrean that the unlikehood of all loads beirxJ maximl.In at the
same tine is 10;1 so that one may reduce the total tensile stress
for CASE II t7t 1/1 . 25 . If that is true, CASE I WOJld control
I post tensioning am the area of stram required WOJld be less
than 9iven al:ove.

I Transverse deck stresses were not examined. If the tie beam

provides significant restraint to the floor beams, it will lead
to tensile stresses in the deck am sane transverse prestressing
I steel may be desirable .

I
I
I
I
I
I
I
I
I
I
I
 86 I
6.0 Conclusions
I
It is the intent of the study to investigate design options
which may lead to more practical tied arch bridges. Based upoo
I
this design study, it appears that the ideas suggested warrant
further consideratien.
I
'!he total weight of the proposed structure is cnly slightly
IOClre than the conventional design. The steel weight of the
alternate is much less than that of the original design. '!he
I
arch ribs increased fran a.i:loot 1.7 million poUIXls to a.i:loot 1.85
million pounds. '!he arch bracing stayed the same at a.i:loot 925
t:OOusaOO p:>unds. The tie beams decreased fran a.i:loot 1.6 million
I
poUIXls to zero. '!he floor systan is decreased fran a.i:loot 2 million
poUIXls to atout 930 tOOusand p:>unds. I
'!he 8000 psi concrete strength is higher than that camonly
used in present practice. However, 8000 psi concrete is being
used on the Fast Huntington Bridge in West Virginia with no
I
rep:>rted prdllems (Ref.3). However, the analyses irrlicate that
6000 psi concrete ~d be adequate for this instance. I
'!he use of 15 inch thick deck at the tie beams appears to
be excessive. Shear lag appears to be critical cnly at the ends
of the span. It seems rea.scnable to expect that a 9 inch deck
I
with a large chamfer could be used over the entire span with
the exceptien of the ends of the span near the arch ribs. At
these locations a deck thickness of 15 inches is recamerrled.
I
'!hese p:>rtions are cast in-place so special deck units are not
required. 'Ibis IOCldification would further reduce weight. '!he
use of a cx::rrp:l5ite steel plate en the bottan of the deck in
I
these regions has been used for shear transfer in Germany (Ref. 4) •
SUch a cllar¥3e ~ further reduce structure weight. I
Erection of the arch ribs by rotating than into p:>sition
am splicing than has significant benefits in erection time am
freedan fran falsework obstructing water traffic. '!he erection
I
procedure was rep:>rted for a concrete arch bridge in Germany(Ref.1).
'!here the engineers felt that the method was econcmical for longer
spans.
I
Another configuration of the deck was examined where the
stiffness was less than that of the cases rep:>rted. An important
I
relationship between deck stiffness am arch rranents was evident.
As the deck becanes stiffer, arch live load rranents decreases.
'!he reascn for this is that the parabolic arch is designed for
I
a uniform load. If the deck equalizes the loads, the rranent
in the arch will be snaller, permitting thinner arch ribs. I
I
I~ 87

I '!be normal tied arch bridge deperrls ooly en the tie beam for
stiffness whereas the suggested method utilizes the entire deck
I stiffness . '!be tie beams were made 9 feet deep ~ to 11
feet en the original design . 'Ihis reductien in stiffness was
rrore than balanced by the aalitien of the deck. A greater depth
I would have reduced the rrcrnent in the arch rut the arch was cc:ntr-
olled by the erectien so no benefit would have resulted . '!be
reduced depth would lead to a decrease in wind loading .
I Since the deck is integral with the tie beam , lateral loads
may be resisted by the deck and tie beams . 'Ihls permitted the
I eliminatien of two levels of diagc:nal bracio:J required in the
original design.

I '!be use of the dead load tie cable and the {XlSt tensioned
tie beams insures that the ties are not fracture critical. '!bey
further simplify construction by reducir¥J the amount of field
I 001 tio;J .

'!be continuous deck provides a much SIllX)ther- riding surface

I and a reductien in maintenance oosts . I t is roost probable that
an impervious wearir¥J surface would be installed to further
protect the joints. '!be oost of this surfacio:J must be consid-
I ered in any econanic evaluation. It appears that a reascoable
amount of {XlSt tensionio;J steel is required .

I
I
I
I
I
I
I
I
I
 88
I
7. 0 Acknowledgements
I
'lhls ~k was initiated at the request of the lvnerican Institute
of Steel Constructioo who fuOOed a portion of the ~k for which
I
we are greatful. '!hey al so provided the design drawings for
the original design fran which the alternate design was derived .
'!he interest aOO encouragement of AISC staff are also appreciated .
I
'!here have been fruitful discussions with several design
consultants concerning this ~k . Particularly helpful was the
I
finn of Al:vid Grant & Associates whan we learned, t:.c:Mard the
em of the ~k , had designed a bridge not dissimilar to that
proposed . Unfortunately , for various reasons not related to
I
the adequecy of the design , the structure was not bJilt.
I
'!he figures were drawn by G. !:avid Brierley~ of ESDI
who also reviewed the entire manuscript aOO made flUII'eI'OUS oootrib-
uticns . Lon B. VanFossen of ESDI produced the carrputer runs, I
made the carrputer rrodels used aOO reviewed the n.anuscript.

I
I
I
I
I
I
I
I
I
I
I
 89

8.0 Bibliography
I
1. "Arch halves fall, making pivotal link in Germany"
I Engineering News Record, June 13, 1985.

2. ''Bridge fixed quickly under traffic", Engineering

I News Record, November 19, 1981.

3. "Hybrid girder in cable-stayed debJ.t" ,Engineering

I Engineering News Record, June 13, 1985.

4. Roik, Karlheinz arrl Haesel, Jochen, "cx:MPa3ITE

I BRII:GES IN GERMANY", Second US~apan Joint
Saninar on canposite arrl Mixed Calstruction,
University Washingtal, Seattle. July 18-20,
I 1984.

5. "Standard Specifications for Highway Bridges,

I 'lhirteenth Etlition", ~ican Association of
state Highway arrl Transportation Officials,
Washingtoo, D. C., 1 983.
I
I
I
I
I
I
I
I
II
I
1:1'"
......
I~

1
1
I
1
1
1
1
I
I
I
I
I
I
I
I
I
I