A New Technique to Create Continuity in Prestressed Concrete Members

30

Maher K. Tadros Ph.D., P.E. Cheryl Prewett Professor of Civil Engineering and Director, Center for Infrastructure Research University of Nebraska Omaha, Nebraska

Joseph A. Ficenec, P.E. Structural Project Engineer

Wells Engineers, Inc. Omaha, Nebraska

Amin Einea, Ph.D., P.E. Assistant Professor University of Nebraska Omaha, Nebraska

Steve Holdsworth Operations Manager

Barsplice Products, Inc. Dayton, Ohio

This paper presents a new technique for creating continuity in prestressed concrete members. The essence of the system is the creation of continuity at interior supports by coupling top end strand extensions. This coupling is followed by introducing com­pression into cast-in-place joints and tension in the coupled strands. This system has all the benefits of a continuously post-tensioned system without actually implementing the full post-tensioning operation. Members made continuous with this system are expected to exhibit enhanced seismic resistance, su­perior structural integrity and substantially lower deflection levels than other continuous and non-continuous prestressed concrete members in current use.

Precast, prestressed concrete girders are widely used in the construction of bridges as well as floors and roofs of commercial buildi ngs th roughout the world .

Depending on the application, spans and loading reql.lire­ments, these girders are constructed as simple spans, continu­ous through cast-in-place conventionally reinforced concrete joints, or continuous through full-length ppst-tensioning.

A new splicing method which offers significant economi­cal and serviceability advantages has been developed. The method was origin ally conceived at the Uni versity of Nebraska at Lincoln (UNL) and jointly developed by Wells Engineers, Inc., and UNL. It has been approved by the Nebraska Department of Roads and implemented into the design of a five-span pedestrian/bicycle overpass in Lin-

PCI JOURNAL

coin, Nebraska, described in a com­panion paper in this issue of the PCI JOURNAL.

This paper presents current girder jointing methods, describes the new method of splicing, presents the test­ing of a spliced specimen, outlines the advantages of the new method and discusses the economy of the system.

CURRENT SYSTEMS To facilitate the following discus­

sion, consider an example of a bridge superstructure consisting of pre­stressed !-girders and cast-in-place deck and joints.

Prior to the introduction of this sys­tem, three primary multi-span precast, prestressed concepts have been used extensively:

1. Simply supported members with­out continuity design

2. Members made continuous via cast-in-place, conventionally rein­forced deck and joints

3. Members made continuous by utilizing full-length continuity post­tensioning (see Fig. 1)

The first system (Fig. la) is the least desirable of the three when consider­ing structural performance, initial cost and maintenance. The structural advantages afforded by continuity are not utilized. Flexural demands at midspan require the use of appropriate prestressing to counter tensile stresses and to provide adequate ultimate flex­ural strength. This usage results ~n higher initial &nd final camber and, hence, more upward deviation of member top surface from the intended final grade - a definite disadvantage in bridge structures where a "roller­coaster" driving surface is not desir­able . Another disadvantage is the rapid deterioration of movement (expansion) joints at the supports due to penetration of w&ter an{! deicipg chemicals.

The second system (Fig. lb) pro­vides for continuity, but only for loads applied after the deck becomes com­posite with the girders (i.e. , superim­posed dead loads and live plus impact loads). Thus, this system has some of the same shortcomings as the first sys­tem, but to a lesser degree. It requires relatively high amounts of prestressing

September-October 1993

CIP deck

a) A series of simple spans

CIPdeck

b) Connected via CIP non-prestressed joints

c) Continuous via post-tensioning

Fig. 1. Various techniques for gaining continuity in precast, prestressed concrete bridges.

which could result in excessive cam­ber. In addition, the lack of interior joint precompression is a problem with this system; without precompres­sion, the deck concrete cracks in the vicinity of the joint. The cracks then propagate and widen , and deicing salts, water and other deleterious materials collect in the cracks. This collection and subsequent penetration results initially in the corrosion of some joint reinforcement, is followed by the deterioration of concrete joint materials due to the spalling initiated by the reinforcement corrosion, and ultimately results in the degradation of all concrete and reinforcing materials at or near the joint.

The third system (Fig. lc) is the most efficient of the three systems, providing relatively low levels of pre­tensioning and initial camber, continu­ity at interior support joint locations for all loads except member self­weight, and precompression of interior joint concrete. This system provides all these advantages but at a premium cost. It requires a specialty contractor to facilitate post-tensioning; it requires

uniform widening of girder webs to accommodate post-tensioning ducts and local widening of girder webs (i.e., end blocks) at all anchorage locations; it requires special reinforcement at all anchorage locations to accommodate stress concentrations; and it requires special construction sequencing.

SYSTEM CONCEPT The new system accomplishes all

that a pretensioned/post-tensioned pre­cast concrete system does, but in a simpler, more direct manner. The sys­tem utilizes pretensioned precast members with strand profiles opti­mized to "balance" external loads; top strands at adjacent, interior member ends coupled to provide continuity prior to deck placement; and precom­pression of interior cast-in-place con­crete joints. Two methods may be used to achieve the interior joint conti­nuity and precompression.

The first method involves the fol­lowing steps (see Fig. 2a):

1. Erect pretensioned precast mem­bers with top strands extending be-

31

Step (1)

Step (2)

concrete

Step (3)

Non-shrink

Step (4)

Step (5)

a) First technique b) Second technique

Fig. 2. Two equivalent techniques for splicing precast, prestressed concrete girders.

yond member ends at interior joints. Strand extensions must be of sufficient length to permit their appropriate cut­ting for staggered splicing (see Figs. 3b and 3c).

2, Splice the strand extensions using mechanical splices supplied with con­tiguous hardware to allow for slack recovery. Depending on the type of mechanical connector used, it is possi­ble to partially pre-install the splice in the precast concrete plant. Uniformly tighten all coupled strands via the slack recovery hardware.

32

3. Using appropriate jacking appara­tus, push the ends of joined members outward to simultaneously introduce appropriate tensile forces into all cou­pled strands in the joint.

4. Form, pour and cure concrete joint materials. Concrete with high early strength and low shrinkage prop­erties is desirable. The jacking appara­tus must maintain appropriate tensile force levels during this stage.

5. Once joint concrete has attained satisfactory strength, introduce pre­compression into joints by releasing

jacking force . Remove all jacking apparatus, including brackets and mis­cellaneous hardware.

The second method involves the fol­lowing steps (see Fig. 2b):

1. Erect pretensioned precast mem­bers with top strands extending be­yond member ends at interior joints. Strand extensions must be of sufficient length to permit their appropriate cut­ting for staggered splicing.

2, Install appropriate brackets and space struts at both the top and bottom of girders to maintain end positions

PCI JOURNAL

Strand extensions - I field cut as required for staggered splicing

Cross section I. 10'.{)"

a) Precast girder, side view

Bolt Sleeves Splice See Fig 4

""' ~ ~ ~ ' . .

0 0 .., 0 0

I '

I

I b) Strands spliced and all slack removed, side view

c) Jacking operation, top view

B~

Section A-A SectionB-B

d) Joint is complete, side view

Fig. 3. Pre-bid test specimen showing splicing detail, jacking operation and completed joint. Note: Other reinforcement is not shown for clarity.

during strand tensioning . Easy joint access during tensioning and concrete casting is essential.

3. Splice and prestress individual strands using appropriate mechanical splices in an approved manner to insure the introduction of appropriate tensile forces into coupled strands.

4. Form, pour and cure remaining concrete joint materials. Concrete with

September-October 1993

high early strength and low shrinkage properties is desirable.

5. Once joint concrete has attained satisfactory strength, introduce pre­compression into the joint by remov­ing spacer struts, brackets and other temporary items.

Both of these methods provide the intended continuity and precompres­sion of concrete at the joint. The selec-

tion of either of them will depend on preference of the contractor. The first method requires the use of a relatively large jack. It ensures an accurate over­all prestress force while individual strand forces may vary depending on the type of mechanical connector used. The second method requires a smaller jack, a wrench, or another individual strand tensioning technique. It may be perceived to be a safer oper­ation by contractors unfamiliar with standard post-tensioning operations. However, it is a slower process and it may require debonding of part of the strand being spliced to minimize anchorage seating losses.

The first method was successfully used for the laboratory testing des­cribed later in this paper and the sec­ond method was employed for the pedestrian overpass described in the companion paper in this issue of the PCI JOURNAL.

CONCEPT TO REALITY This new method was implemented

in the design of the pedestrian/bicycle overpass at Tenth and V streets in Lin­coln, Nebraska. The bridge consists of two 90 ft (27 m) exterior spans and three 125 ft (38 m) interior spans . With a strict limitation on superstruc­ture depth, a continuous precast, pre­stressed concrete girder system off­ered a competitive alternative to structural steel.

The following girder alternatives were evaluated for use in the above­mentioned bridge: continuous steel !­girders, simple span prestressed con­crete !-girders, prestressed concrete !-girders made continuous via post­tensioning, and the proposed new girder system. The new girder system offered economical and functional advantages over the other systems, hence, its concept was developed and refined for use on this structure. A companion paper in this issue of the PCI JOURNAL describes the details of the Tenth Street Overpass.

PRE-BID TESTING To evaluate and improve specific

details and to provide an actual demonstration for the bidding contrac-

33

Fig. 4. The pre-bid specimen before concrete was poured into the joint.

tors to evaluate, a full-scale test of the girder splicing concept was carried out at the University of Nebraska. Two 10 ft (3 m) long !-girder sections, with identical cross section to the Tenth Street pedestrian/bicycle overpass girders, were cast. Except for their lack of bottom reinforcement, the sec­tion ends were reinforced identically to the ends of the overpass girders. Seven steel sleeves for 1!4 in. (32 mm) diameter bolts were placed through the top flange of each girder for side mounting of steel jacking brackets (see Fig. 3).

After curing, the girders were appropriately placed. The strands were spliced using swaged couplers with zero anchorage set loss. The jacking brackets were mounted and the joint formed. A jacking force of 500 kips (2200 kN) was slowly applied and high early strength concrete was poured in the joint. The jacking force was maintained while joint concrete cured. The jacking force and brackets were removed after 24 hours when the concrete reached a compressive strength of 4200 psi (29 MPa).

For this test, the brackets were mounted on both sides of the flanges, as shown in Fig. 3c, to allow the appli­cation of jacking force as close as pos­sible to the geometric center of the strands in an effort to minimize eccen­tricity. This was required in the test because of shortness of the spliced girders and their lack of dead load to resist the couple that would result

34

from eccentric jacking. For an actual girder, however, it may be preferable to have a single jack in order to have better control of field labor. It should be noted, moreover, that accuracy of placement of a single jack is more crit­ical than that of a double jack arrange­ment, as weak-axis eccentricity should be avoided.

Swaged mechanical couplers with two types of slack recovery hardware were used to splice strand extensions. A mix of two different types of hard­ware was used to investigate the prac­ticality of installing each type. One type utilized threaded rods and a tum­buckle while the other utilized two end plates with a pair of threaded rods and nuts (see Fig. 4). In both cases, the swaged sleeve was installed in the plant. The balance of the hardware was installed in the structural labora­tory for this specimen. It would be field installed for an actual structure. The splice with the turnbuckle hard­ware proved more practical than the other type, given the limited available space (see Fig. Sa).

The spliced 1-girder specimen was tested by simply tying down each of its ends with two 1Y. in. (32 mm) dia­meter prestressing bars to a structural floor and applying an upward concen­trated force at the joint using two hydraulic jacks. The spliced girder resisted the slowly applied concen­trated force until it reached 390 kips ( 1700 kN). At this load level , the spec­imen failed in a classic shear mode.

Fig. 6 shows the test setup and the failed specimen.

The estimated load for flexural fail­ure was 580 kips (2600 kN), while the estimated load for shear failure was 360 kips (1600 kN). There was only one narrow flexural crack at the top of the joint. This crack first appeared at a load level of 340 kips (1500 kN). After the shear failure at a load level of 390 kips, the load was released and the flexural crack com~letely closed. This behavior demonstrates the effi­ciency of the strand splicing and the joint concrete precompression.

OTHER POTENTIAL APPLICATIONS

In addition to using the described splicing and tensioning technique for creating continuity at negative mo­ment areas in bridge girders, the method can be used for several other applications, including the following:

1. Splicing bridge girders near inflection point locations. In long-span applications, this approach provides members with transportable lengths (see Fig. 7a).

2. Splicing pretensioned building beams at locations of negative moments (at support locations) to create continu­ity in multi-span beams (see Fig. 7b)

3. Splicing pretensioned floor and roof joists at locations of maximum negative moment (at beam or girder support locations) to increase their structural efficiency (see Fig. 7c)

Whether or not precompression is introduced at the joint by the previ­ously described method will depend on economic benefits. Providing strand continuity between precast con­crete elements is , by itself, an excel­lent contributor to structural efficiency regardless of whether or not precom­pression is introduced at the joints.

ADVANTAGES OF NEW SPLICING TECHNIQUE

Relative to either simply supported pretensioned members or pretensioned members made continuous via con­ventionally reinforced cast-in-place joints, the new system offers the fol­lowing advantages:

1. Continuity is provided at all inte-

PCI JOURNAL

Threaded rod, left-hand thread

1!2", -7 wire strand extension

a) Swaged/turnbuckle splice, Barsplice Products, Inc.

* Varies as required

b) Z Anchor, VSL Corporation

c) Ring Anchor, Dywidag Systems International, USA, Inc.

d) GRABB-IT Cable Anchor, Florida Wire & Cable Inc.

Fig. 5. Strand splices for gaining continuity.

rior supports for all loads except self­weight. This results in less bending near midspan, less girder deflection (i.e., greater bending stiffness), lower amounts of prestressing and smaller erection camber.

2. The same strands, which are depressed at midspan to satisfy posi­tive bending demands, can be elevated near the supports to satisfy negative bending demands. This optimization of strand placement results in the use of fewer strands, which translates into a cost savings.

September-October 1993

3. Lower levels of prestressing and camber result in less long-term up­ward creep. This reduces or, in some cases, eliminates camber and results in a smoother deck surface with less of a "roller-coaster" profile.

4 . Joint precompression reduces and/or eliminates flexural cracking in the joint area. This creates concrete surfaces which are less permeable to deicing salts and, hence, assures more corrosion-free reinforcement in the joint area.

Compared to the pretensioned/post-

tensioned systems, the new system offers the following advantages:

1. Expensive post-tensioning an­chorage hardware at tendon ends is eliminated.

2. Post-tensioning duct sizes usually necessitate the uniform widening of !-girder webs to accommodate the ducts. This results in larger concrete, reinforcement and strand quantities, and greater girder self-weight. The new system, which requires no ducts, saves on web widening and all the resulting quantity increases.

35

3. End blocks, normally required in post-tensioned systems due to stress concentrations at tendon ends, are eliminated. This results in both con­crete and reinforcement savings.

4. Grouting of ducts is eliminated. 5. Strand transfer and development

length in the splice zone is not a con­cern since the strand is continuous across the joint.

ECONOMY OF THE SYSTEM

This system, in the authors' opinion, should be chosen for its structural superiority to other available systems. It offers, for the first time, direct conti­nuity in precast, pretensioned concrete members. Furthermore, it is expected to be an economical alternative partic­ularly as contractors gain experience and confidence in their ability to use the technique.

To obtain an indication of the rela­tive cost of this system, one may com­pare it with the full-length post-ten­sioning system, the closest system in terms of structural performance. Con­sider a 250 ft (75 m) long, two-span AASHTO Type IV 1-girder with 24-X in. (12.7 mm) continuity strands. The proposed system would require 24 couplers at about $25 material cost and $75 estimated splicing labor per cou­pler; 6000 ft ( 1800 m) of continuity prestressing strand at $0.25 per ft; and the use of the jack and spacer bracket cost of approximately $300 per joint. Thus, total joint cost, excluding joint concrete, is $4200 (i.e., 24 couplers x $100 per coupler+ 6000 ft x $0.25 per ft + $300 = $4200).

If the girder is made continuous through full-length post-tensioning, the total weight of post-tensioned strands is 3120 lb ( 1420 kg). In addi­tion, the web must be uniformly wid­ened 1 in. (25 mm), which results in a concrete volume increase of 3.5 cu yd (4.55 m')- the AASHTO Type IV !-girder is 4ft 6 in. (1.37 m) deep . Assuming a cost of $2 per lb of post­tensioning strand for material and labor and $600 per cu yd of concrete, the total cost of post-tensioning and web widening is $8340 . Two end blocks are also required to house post-

36

Two loading rams Structural floor

a) Pre-bid testing setup

Fig. 6a. Pre-bid test setup.

Fig. 6b. Failed specimen, shear failure.

tensioning anchorage hardware and to reinforce the concrete to resist anchor­age stress concentration. These blocks would cost approximately $600. Thus, the total cost of this alternative is approximately $8940.

This comparison may be an overly simplistic one as it does not account for other important factors, including the increase in girder weight with the post-tensioning alternative and its impact on the load carrying capacity of the girder. Nevertheless, it supports

the general idea that the proposed sys­tem can be economically attractive.

CONCLUSION

A new technique to create continuity in pretensioned concrete members has been developed, tested in the labora­tory and implemented in the design of a full-scale pedestrian/bicycle overpass in Lincoln, Nebraska. This technique offers true continuity of pretensioned concrete members without resorting to

PCI JOURNAL

___.,-

a) At inflection points of bridge girders

b) Building beams, at negative moment areas

c) Floor and roof joists, at negative moment areas

Fig. 7. Other potential applications of the new splicing method.

conventional post-tensioning. With the potential for use in bridge

girders and building beams and joists, its use can result in significant sav­ings , reduced structural depth, im­proved serviceability, and expansion of the usable span ranges for mem­bers. Its use allows precast, pre­stressed concrete members to compete in greater span ranges. The system has already gained the approval of the Nebraska Department of Roads.

September-October 1993

ACKNOWLEDGMENT

The authors acknowledge the finan­cial support of tbe Center for Infras­tructure Research, University of Nebraska-Lincoln ; Concrete Indus­tries, Inc. , Lincoln, Nebraska; and Barsplice Products, Inc ., Dayton, Ohio.

The following individuals made sig­nificant contributions to the work described in this paper: Lyman Free-

mon , bridge engineer, Nebraska Department of Roads; Paula B. Wells, chief executive officer, Steve Kneip, vice-president, and Samar Gogoi , structural engineer, Wells Engineers, Inc.; Larry Fischer, vice-president, Concrete Industries, Inc.; John Stan­ton, professor, University of Washing­ton; and graduate students Tony Pick, John Savage, Mike Vigil and Svein Magnussen, all of the University of Nebraska at Omaha.

37