Precast-Prestressed Concrete I-Beam Bridges

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

ped

,ms.
sult
PRECAST-PRESTRESSED CONCRETE I-BEAM BRIDGES
the
for By PAUL H. KAAR and ALAN H. MATTOCK 0

ful
of
ion

mt
of
ge
of
.a- Paul H. Karr, Development Engineer, Portland
nt Cement Association, Structural D evelopment Sec-
tion.
1n
JS
,s
I?
e
lt

Synopisis
This paper concerns the research on grade separation bridges in
progress at th e Portland Cement Association Laboratories. The bridges
consist of separately placed pre-cast-pre- tensioned girders, combined with
a cast-in-place roadway slab. Continuity is estabushed by addition of
normal bar reinforcement over the girder ends acting togeth er with girder
end di aphragms. The research includes studies of component parts as
well as integral behavior of tl1e bridge as a whole.

Continuity in Precast B·riclges


The importance of the typical grade separation bridge to our arterial roads is
quite apparent. It would be difficult to travel 10 miles without crossing und er or
over at least one of these bridges. Most of th ese bridges are of medium span
length-60 ft or so-and many consist of simple spans with roadway joints at the
end of each girder. Some bridges have been improved by making the roadway
continuous, and enclosing th e girder ends in diaphragms. While some degree of
continuity can be obtained by making th e roadway continuous, few designers have
included this advantage in their calcul ations. This subject of continuity is of
major interest for several reasons :

(1) Bending moments and deflections are reduced by establishing con-


tinuity in a mul ti-span girder system. It follows, then, that the girders re-
quired to carry a given load can be smaller in cross-section when conti nuity

-
is effective, or if stand ard sections are used the reinforcing required will be
less. In either case, th e net result is th e same: more bridge per dollar.

Cem·e~t"vAelopment Engineer an d Senior D evelopment Engineer, respectively, Portland


ssocia tton , Structural D evelopment Section, Skokie, Illinois.

85
( 2) The reserve load capacity of a simply supported girder can be
governed by a single cross-sectional region. In a continuous structure how-
ever, overload will lead to moment redistribution so that several sectio~s gov-
ern ultim ate strength. Accordingly, continuity tends to minimize reductions in
overload capacity caused by variability in properties of concrete and reinforce-
ment.
( 3) If the roadway deck is made continuous, the drainage of water and
ice removal salts can be easily controlled. The elimination of the roadway
joint will also improve the riding surface and reduce maintenance.
Why were such obvious adva ntages in bridge design not incorporated into
these structures long ago? Why have so many bridges of multiple simple spans
been constructed? The first reason concerns the fabrication of prestressed concrete
girders. Since labor is expensive in this country, designers and fabricators have
both sought mass production techniques using minimum labor. The pre-
tensioning type of prestressing is best suited for this mass production and has
become immensely popular for this reason. These separate girders can most
easily be used as simply supported spans.
While some designers were quick to .see the advantage of added reinforce-
ment over the girder ends to establish some degree of continuity, thre was very
little information available regarding th e degree of continuity. Because of this
lack of data designers frequently added the improvement without considering it C(
in their calculations.
Precast-prestressed girders used in conjunction witl1 a continuous roadway tl
pl
deck were tl1erefore selected for development study to provide better desigo
criteria.
Forms of Continuity Connection
Various metl10ds have been used for establishing continuity in separately c
placed concrete girders. Probably the earliest solution was developed by e1
Freyssinet by a "cap cable," shown in Fig. 1. By this method cables spanning

"CAP CABLE•
tl
SIMPLY SUPPORTED BEAM SIMPLY SUPPORTED e1
ti
'-ANCHORAGE

Fig. 1-Use of "Cop Cables" to Produce Continuity between Precast Beams. ei


~
b
the junction between simple span girders were post-tensioned so the joint was
prestressed. Anchoring post-tensioned rods in th e roadway deck is another form
of establishing continuity.
The design innovation of adding normal reinforcing bars in the roadway n
over the girder junction seems to be the simplest of on-site construction procedures 0
and therefore tl1e cheapest. Fig. 2 shows the sequence of construction. T~e [
precast girders are first placed on tl1e piers, the ends being spaced about six 0
inches apart. At this stage th ey are merely simple-span girders. Forms for the ff
roadway deck are th en placed, the girders supporting the forms. The normal
deformed bars are then placed in tl1e roadway spanning tile girder ends a~
over the piers. At tl1e time tl1e roadway is cast, diaphragms are also form
which enclose the ends of the girders and fill tl1e space between tllem. Thus,
the dead load of the bridge is supported by a series of simple-span g~~ers,
live loads are supported by an integral and continuous structure. Positive )Ill. -
hr
span moments are resisted by the girders and deck slab acting as a composite
T-section. Negative moments over the supports are resisted by the normal dei
formed bars embedded in the deck, acting together with tl1e bottom flange O
the precast girder as a reinforced concrete section. Over the whole bridge structure,

86
1 PRECAST GIRDER II
I I
PRECAST GIRDER
I
ir
STAGE l - PRECAST GIRDERS PLACED ON BRIDGE PIERS BY CRANE

d
y
11I
CONV ENTIONAL DE FOAMED BAR REl!!.ORCEMENT

II
I I
,r
I
STAGE D - CONVENTIONAL DEFORMED BAR REINFORCEMENT FIXED IN PLACE OVER SUPPORT

s
st
STAGE m- DECK AND DIAPHRAGMS CAST IN PLACE. CONVENTIONAL DEFORMED
BAR REINFORCEMENT NOW BONDED TO PRECAST GIRDERS

Fig. 2-Construction of Precast-Prestressed Continuous Bridge.


is
it composite action between roadway deck and girders is achieved by bond between
the roadway concrete and the rough top face of the girder, aided by stirrups
y
projecting from the girder into the roadway slab.

Experimental Program
The following subjects relating to such bridges were investigated in a
y comprehensive laboratory project directed toward providing data for more
y economical design:
g
(1) Girder Continuity
The first phase of the project inv.olved tests to establish the effectiveness of
the type of continuity connection discussed above. Pretensioned girders without
end blocks and with straight strands were used in the study, since these lead to
the most severe conditions possible at the connection. In this case, the precom-
pression at the bottom flange of the girder will be greater than in girders with
either end blocks or deflected strands. The connection behavior when compres-
sion from prestress and from applied negative moments are combined at the
as bottom girder flange can then be studied to clarify a possible detrimental influence
m of precompression in the bottom flange of the girders. Connections of this type
were tested by joining two six-ft. long girder st ubs over a central support as shown
in Fig. 3. The ends of tl1e girder stubs were loaded, producing maximum negative
es moment over the single support. Variables consisted of three different amounts
le
of deck reinforcing steel combined with three different amounts of prestressing.
ix Deck reinforcing might be termed light, medium, or heavy, corresponding to
0.83, 1.66, and 2.49 per cent, based on girder depth times width of the bottom
Range. The prestress amounts can be described as none, medium, and large,
corresponding to initial concrete precompression in the bottom flange of 0, 2100,
and 3200 pounds per square inch.
The theoretical ultimate strengtl1 of each connection was computed, first
neglecting any influence of the precompression of th e bottom flange, and ,secondly
assu~~g precompression over tl1e full girder flange length. An important char-
te actenstic of pre-tensioned girders is the prestress transfer length in which the
e· steel _stress is gradually transferred to the concrete. There actually is no precom-
of pression at the end of the girder. Then too, precompression from prestress and
ompress from negative bending are not additive, because compression from
e, 8ex11re releases prestress in a similar manner as prestress is reduced in ;i loaded

87
Fig. 3-Test of Continuity Connection.

prestressed concrete column. By comparing test results with the calculated


strengths, it was determined that for the practical range of continuity reinforce-
ment from 0.5 to 1.5 per cent deck steel, the influence of the precompression at
the bottom flange may be neglected in the calculation of the ultimate negative
bending moments. The experimental results from this project are fully discussed
in PCA Development Department Bulletin No. D 34, "Pilot Tests of Con-
tinuous Girders," to be published in mid-1960.
( 2) I-I orizontal Shear Connections Between Girder and Roaclway
In order for the girder and cast-in-place roadway to perform as a structural
unit, the horizontal shearing strength at the girder-slab junction must be
effectively developed. The second stage of our program concerned a study of the
transfer of horizontal shear across a structural joint. The variables studied in
these test specimens were : shear keys, stirrups, roughness of the contact surface,
and adhesive bond.
The first tests involved pushing off a short length of simulated girder flange
from the girder web, as shown in Fig. 4. The next step, shown in Fig. 5, was to
test a complete composite girder, dial gages being placed so that the relative slip
between d eck and girder could be measured. It was concluded from the girder
load-deflection curves that composite action ceased when the roadway slab moved
0.005 in. relative to the girder. The corresponding horizontal shearing stress for
this 0.005-inch movement was 500 psi for a rough contact surface and 300 psi for
a smooth contact surface. Strengths of the concrete were typical-5000 psi for
the girder and 3000 psi for the roadway.
The tests indicated that keys artifically formed in the girder top were of
little help because by the time tl1e roadway had slipped enough to bring the key
into use, composite action had b egun to break down.
Stirrup reinforcement protruding from the girder into the slab was found_ to
be essential for effective composite action. Stirrups required by ordinary design

88
:d
e-
at
1e
:d
a-

al
ie
1e
in
e,

ie
to
ip
er
:d
Jf
Jf
Jf

to
:a Fig . 5-Horizontal Shear Test using Composite Girder.

89
for vertical shear in the girders of this type and protruding into the roadway I b
are sufficient for composite action. A complete description of these tests sad
the developed data is included in the forthcoming PCA Development Departm:n0 t
Bulletin D35, "Horizontal Shear Connections."
( 3) Bridge Design Studies
A two-span two-lane continuous highway bridge to carry the AASHO
st~ndard H20-Sl6 loading wa~ ~ext desi~ed, ~ ith dimensions as shown by
Fig. 6. All phases of th e prehmmary design, with the exception of the girder

A
I
f------- ss'-o" -------t--------- 66'-o"

A
ELEVATION

SECTION A-A

Fig. 6-Continuous Precost- Prestressed Bridge Considered in Design Study.

connections and the horizontal shear connections, conformed either to the 1957
edition of AASHO "Standard Specifications for Highway Bridges" or to the ACI-
ASCE report "Recommended Practice for Prestrnssed Concrete." Design criteria
for the exceptions noted were based on laboratory tests. This design emphasized
the need for additional development work as described in the points following.
( 4) Shearing Strength
Shearing streng th of the girders in the region of the continuity connection
was investigated by a series of tests on the type of specimen shown in Fig. 7.
( The blocks suspended from th e girders are not part of the loading but are an
added dead weight to produce the same dead weight stresses in the one-half
scale model as would exist in the prototype.) A half-scale girder, together with
a cast-in-place roadway deck and including a tied-down cantilever were used for
the test specimen. By varying the tie-down force, rotation of the joint was
prevented, thus simulating continuity. The variables were the spacing of the
stirrup web reinforceement and the location of th e applied loads. These loads
simulated the distribution of loads in the AASHO H20-Sl6 design truck. Four-
teen combinations of stirrup spacing and load location have been tested. The
prestress in the girder was found to exert a favorable inRuence on ultimate shear
strength in the specimen.
( 5) Flexural Strength
A fully con tinuous girder and deck slab were tested in flexure as shown in
Fig. 8. The loading simulated the distribution of loads in the AASHO equivalent
lane loading and was arranged to produce maximum bending moments at the
interior support. The specimen behaved in a very satisfactory manner, the p
maximum load sustained being about seven times the design live load plus the g
weight of the girder. The girder section was designed for 1.5 times dead load g

90
,t

)
y
:r

Fig. 8-Continuous Precost-Prestressed Bridge Girder under Test.

p~us 2.5 times live load plus impact with no tension in the bottom face of the
~rder. Moment redistribution occurred to a very large extent, and hence the
girder was in effect over-designed from the point of view of ultimate strength.

91
( 6) Creep and Shrinkage Studies
Two 66-ft. continuous girders fabricated to half-scale have been under long-
time study, as shown in Fig. 9, to better evaluate the effects of creep and

Fig. 9- Continuous Precast- Prestresse d Bridg e Gird ers Subj ect to Sustain ed Load.

differential shrinkage on composite prestressed continuous structures. Tbese mem-


b ers have now been in place almost two years. During this time measurements
have been made of concrete and reinforcing strains, deflections, and support
reactions. Small reverse moments attributable to creep deformations were noted
at interior supports. Anchors protruding into or through the diaphragm may be
needed to insure continuity over long time periods.
(7) Fatigue Tests
A typical connection under repeated load test is shown in Fig. 10. The
loads were arranged to produce maximum bending moment at th e girder junction
by pulsating loads applied at the girder stub ends. The only variable involved
was the maxirmun value of the pulsating load. A composite connection designed
for a static ultimate strength of two and one-half tim es the service load moment
sustained without failure over ten million cycles of one and one-half times the
service load moment.
( 83 Reverse Bending Tests
In continuous bridges of more th an two spans it is possible for small positive
moments to develop at girder connections. Structmal connections joining the
girders at the bottom flange area and therefore resisting positive moment, have
b een tested to develop design criteria for this detail.
( 9) Bridge Test
A c9mplete two-lane two-span highway briqge was constructed to one-ha~
scale after this exhaustive seri es of bridge component tests. The test bridge is
shown in Fig.' 11. The testing of the bridge fell into two classes, service load

92 .
1g.
nd

Fig. 10-Dynamic Load Test of Connection Between two Precast- Prestressed Girder Stubs.

ll·
its
1rt
3d
be

,e
Je
,e

Fig . 11-Service Load Test of Precast- Prestressed Continuous Bridge.

II tests and overload tests In the service load tests 10 000 or 20 000 lb concrete
js blocks were applied se~arately to 50 different lo'cati~ns on th e' bridg; roadway
.d to determine lateral distribution of loads and influence curves. At th e conclusion
of tl1e service load tests, the bridge was loaded to destruction by application of
loads simulating the distribution of axle loads in the extraordinary equivalent
military truck load, for which bridges on ilie Interstate Highway System must
be designed. The mass of data obta_in~d is still being processed and will appear
in a series of Portland Cement Association Development Department Bulletins.
Vol
1
1

11
ll
11
Ii
1:
1:
H
H
l<

94

You might also like