Available online at www.sciencedirect.com

ScienceDirect

Transportation Research Procedia 14 (2016) 3943 - 3952

Procedia

www.elsevier.com/locate/procedia

6th Transport Research Arena April 18-21, 2016

TRANSPORT RESEARCH A.RENA

Jointless bridge design at the Virginia Department of Transportation

Edward Hoppe Keith Weakley b, Park Thompson c

aVirginia Transportation Research Council, Charlottesville, USA bVolkert, Inc., Springfield, USA cVirginia Department of Transportation, Staunton, USA

Abstract

In an effort to reduce bridge lifecycle costs, the Virginia Department of Transportation (VDOT) has developed new design method aimed at the elimination of expansion joints. Various structural details have been experimented with, including full integral abutment with a moment relief hinge, semi-integral abutment, and deck extension. In addition, a new type of abutment was developed for long-span applications. The problem of excessive roadway approach settlement was addressed with the use of elastic inclusion on bridge backwall, select crushed stone backfill, and buried approach slabs. Currently, VDOT considers jointless design as the primary choice for new bridges.

©2016 The Authors. Publishedby Elsevier B.V. Thisis an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of Road and Bridge Research Institute (IBDiM) Keywords: Bridge; expansion joint; jointless; integral

1. Introduction

Conventional bridge designs typically include expansion joints whose function is to allow unimpeded thermally induced displacement of the superstructure elements. Over the service life of a bridge, expansion joints become significant maintenance items. Theoretically, joints are protected from environmental influences through various

* Corresponding author. Tel.: +1-434-293-1960; fax: +1-434-293-1990. E-mail address: Edward.Hoppe@vdot.virginia.gov.

2352-1465 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of Road and Bridge Research Institute (IBDiM) doi:10.1016/j.trpro.2016.05.486

types of waterproof seals. In practice, it is not uncommon for the sealing of a typical joint material to degrade well in advance of its intended design life. This results in surface water and de-icing chemicals penetrating expansion joints and causing extensive damage to beam ends, bearing assemblies, beam seats, and substructures. Figure 1 shows an example of serious bridge deterioration caused by a leaking joint. This type of damage is repetitive in nature, requiring substantial maintenance funds to rectify and causing serious disruption to the travelling public. Deck closures (link slabs) are the only possible fix for structures capable of handling the change in structural behavior, otherwise there is no permanent fix available.

Fig. 1. Deterioration at an abutment and over a pier on a conventional bridge.

Recently there has been a trend toward constructing jointless bridges. Although thermally induced stresses must be accounted for in this new type of design, the absence of joints results in substantially lower construction and maintenance costs (Soltani and Kukreti, 1992). In addition to reducing maintenance expenditures, the absence of joints significantly enhances the ride quality. Although the use of expansion joints is still prevalent, the general design trend in North American practice appears to be moving toward jointless bridges.

Nomenclature

Jointless Bridge Bridge with no traditional expansion joints for accommodating thermal displacements

Full Integral Abutment cast monolithically with the bridge deck and beam ends, allowing shear and moment

transfer, supported on a single row of foundation piles Semi-Integral Backwall cast monolithically with the bridge deck and beam ends, allowing only vertical load

transfer and interacting horizontally with the adjoining backfill material Deck Extension Bridge deck extending over the backwall and interacting with the adjoining roadway pavement Virginia Abutment Hybrid abutment with an integral backwall (semi-integral configuration) and a static retaining wall that isolates the integral backwall from soil-structure interactions

2. Purpose and scope

The purpose of this paper is to demonstrate various structural and construction details that are representative of the current bridge practice implemented by the Virginia Department of Transportation (VDOT). Field observations and the results obtained from recent research projects are also presented. The focus is on practice ready, field-proven approaches.

3. Methodology

3.1. Developmental approach

VDOT is currently responsible for maintaining approximately 13,400 bridges along 92,000 km of roads in Virginia. Jointless construction has been recognized as being a better way to manage limited available resources, by

reducing the life cycle costs through improved design methods. Jointless bridges do not have traditional expansion joints. All such joints have been eliminated.

The design of jointless bridges requires careful consideration of thermally induced movements and resulting stresses imposed on the structure and the adjoining approach embankments. Daily and seasonal ambient air temperature fluctuations cause repetitive expansion and contraction movements. It is important to ensure that these displacements do not damage structural components.

In an effort to optimize the integral design process, starting in the early 1990's VDOT embarked on a number of field and laboratory research projects aimed at evaluating the performance of bridges during their service life. Field data were used to verify theoretical design assumptions, develop new details, and relax use restrictions progressively as new experience and expertise were acquired. Today, VDOT's design policy is to select integral bridges as the primary choice. Conventional abutments with expansion joints are allowed only when approved by the State Structure and Bridge Engineer. Various design guidelines, samples of detailed design calculations and examples of structural details have been developed and are available on the Internet for VDOT engineers and consultants to promote consistent engineering practice (VDOT Structure and Bridge Division, 2015).

3.2. Selection criteria

According to the abutment selection algorithm developed by VDOT, the usual hierarchy is as follows:

1. Full Integral

2. Semi-Integral

3. Deck Extension

4. Virginia Abutment.

The primary design choice is full integral, as shown in Figure 2. The full integral design provides for thermally induced displacements to be transferred into the pile cap and foundation piles. The girders and deck slab extend into the abutment. A characteristic detail of the Virginia practice is to provide a moment relief hinge in the backwall. It allows for thermally induced rotation and results in a significantly reduced end moment at the pile cap. The hinge detail is constructed using closely spaced, corrosion-resistant steel dowels. A continuous strip of high-durometer-value neoprene material is placed at the interface. It serves to transmit vertical loads. Dowels are designed to transmit shear loads. Full-scale laboratory tests conducted on the abutment hinge detail indicated no sign of damage after application of more than 27,000 displacement cycles (Arsoy et al., 2002).

Fig. 2. Full integral abutment.

The construction of full integral abutments requires adequate foundation pile embedment to achieve tip fixity and limit bending stresses. Typically, it involves a single row of end bearing piles with a penetration length of at least 8

m. If such embedment is not feasible, a semi-integral design is considered, as shown in Figure 3. It is important to drive full integral foundation piles in line within a tolerance not exceeding ±75 mm. This may be difficult to achieve when the rock interface is relatively shallow and highly variable with depth, such as found in karst formations.

Fig. 3. Semi-integral abutment.

Semi-integral design is sometimes referred to as integral backwall. The integral backwall overlaps and overhangs the abutment. Any type of abutment can be used, including gravity and cantilever construction, allowing the integral backwall to be a versatile application, readily adaptable to any foundation. Unlike the full integral design, superstructure and substructure elements are not interconnected. Although the full integral design requires a single row of piles, a semi-integral abutment is constructed with two rows of piles to provide lateral stability. The integral backwall is designed to slide laterally on bearings without any moment transfer to the underlying abutment.

An alternative approach to semi-integral design is deck extension, as shown in Figure 4. This design involves a shortened backwall and deck slab extending laterally over the backwall to eliminate expansion joint at the abutment. The main superstructure beams are configured as on a conventional bridge. Beam ends are not embedded in the backwall. This type of design is not classified as integral, but it provides a jointless solution. Buried approach slabs are used with deck extension when needed.

Fig. 4. Deck extension.

The Virginia Abutment, as shown in Figure 5, is typically used in cases where the current design span limits for integral and semi-integral abutments are exceeded and the resulting thermal movements cannot be safely accommodated. It involves constructing a secondary, static retaining wall behind the integral backwall. Thermal movements of the integral backwall are absorbed by a tooth joint placed at the bridge deck. The void space between both concrete walls, typically covered with epoxy coating, is designed to collect roadway drainage and transfer it away from the abutment. The distance between the backwalls is sufficient for future maintenance and cleaning activities. The Virginia Abutment design eliminates variable soil-structure interaction caused by thermal effects.

Fig. 5. Virginia Abutment.

3.3. Bridge approaches

The repetitive movement of the superstructure can cause excessive settlement at the bridge approaches (bump at the end of the bridge). Following poor approach performance with some early integral bridges, VDOT conducted a number of field studies, resulting in the implementation of elasticized expanded polystyrene (EPS) material at the back of full integral and semi-integral abutments (Hoppe, 2005). This elastic inclusion, in conjunction with a well-graded crushed stone backfill material, was found to be very effective at absorbing thermally induced superstructure displacements and attenuating approach settlement. The design detail including the elastic inclusion with select backfill is shown in Figure 6. The EPS material is designed to operate within the linear elastic range of strains. The minimum specified thickness of the EPS material is 25 cm. Geotextile separation fabric is installed at the interface of the EPS material and the backfill.

The use of concrete approach slabs with integral abutments has been the subject of extensive debates. Although it is recognized that a concrete slab can provide gradual grade transitioning at high volume roads, its universal utility is not evident. Often, it is more cost-effective to patch a section of settling approach pavement at a newly constructed bridge than risk considerable maintenance activities associated with repairing a settled approach slab. Typically, minor approach settlement is observed once or twice during the first 5 years of operation. With the implementation of structural backfill, the need for approach slabs has been eliminated in many cases, resulting in reduced maintenance needs.

The repair of settling approach slabs at integral bridges can be challenging, particularly with concrete slabs providing the driving surface. To address this problem, VDOT has developed design details for buried approach slabs, facilitating pavement overlay in case of settlement. In full integral and semi-integral designs, there are provisions for the approach slab to remain stationary while the backwall moves laterally.

Fig. 6. Elastic inclusion and select backfill.

3.4. Skew effects

The magnitude of earth pressure acting on the integral abutment can be substantial, as the state of stress shifts from active toward passive as a result of thermal expansion. In addition, thermal expansion of skewed bridges causes superstructure rotation in the horizontal plane, most noticeably on semi-integral abutments and deck extensions. This effect has been verified with field instrumentation and observations. Noticeable rotations of the superstructure were detected with skew angles as low as 5 degrees, indicating that little tangential friction is actually mobilized at the soil-backwall interface. Sufficient earth pressure developed to shift superstructure 98 m long and 26 m wide (Hoppe and Gomez, 1997). To resist this movement, VDOT design practice calls for a concrete buttress at semi-integral and deck extension bridges. The buttress prevents the rotation while allowing longitudinal displacement at the interface of two stainless steel rub plates containing shear studs cast into concrete. VDOT has developed design guidance and structural details for resisting horizontal rotation and accommodating the resulting forces.

3.5. Foundation piles

VDOT's typical practice is to use steel H-piles oriented in weak-axis bending for full integral bridges, to minimize bending stresses. Steel pipe piles are not recommended because of their inherent stiffness and the resulting likelihood of increased stress at the abutment. Concrete piles are also not recommended because under lateral loads, tension cracks progressively reduce the vertical load carrying capacity of these piles. These recommendations were developed following a series of full scale laboratory tests (Arsoy et al., 2002).

3.6. Backfill placement

Backfill of integral abutments is not allowed until the concrete has attained full design strength. Individual lift thickness is kept at 15 cm maximum. There should not be more than 15 cm difference in depth from finished grade during fill placement to avoid unbalanced lateral forces.

3.7. Concrete placement

To minimize cracking caused by thermal strains, the portion of full integral abutment above the hinge, or the integral backwall in semi-integral design, should be cast only when minimal thermal movement is expected. Concrete placement is usually performed at dusk or during a uniformly cloudy day. Also, when casting superstructure elements, the contractor is not permitted to attach any formwork to the substructure.

3.8. Stage construction

In many cases, integral bridges are built in stages to allow traffic flow during construction. Problems usually arise in the second stage when deflections generated by the non-composite dead loads result in rotations at beam ends. Direct connection of a composite section cast in the first stage with the non-composite section can result in extensive concrete cracking, especially in the concrete end diaphragm, because of torsional effects. The recommended approach is to construct the adjacent stage separately and then cast a narrow closure pour interconnecting both completed composite parts of the structure. The objective is to have sections with compatible strain levels prior to concrete tie-in. This is especially important for full integrals and integral backwalls. Long simple spans with large deflections and end rotations need to be carefully considered.

3.9. Design limits

The total bridge length from abutment to abutment and the total thermal movement at abutment should not exceed the design limits provided in Table 1.

Table 1. Length and skew limits.

Full integral Semi-integral Deck extension

Steel bridges 90 m for 0° skew 135 m 135 m

45 m for 30° skew 30° max. skew 45° max. skew

Concrete bridges 150 m for 0° skew 230 m 230 m

75 m for 30° skew 30° max. skew 45° max. skew

Total movement at abutment 40 mm 55 mm 55 mm

For intermediate skew angles the maximum allowable bridge length is interpolated. Total movement at the abutment corresponds to the estimated full temperature range (expansion and contraction).

Integral backwalls are designed (moment and shear) to resist passive earth pressure resulting from thermal expansion. A passive earth pressure coefficient of 4.0 is typically selected for design incorporating EPS material and structural backfill. If EPS material is not used, a coefficient of 12.0 is required for design. On single span integral bridges, EPS material is placed at the upgrade abutment only. The intent is to provide preferential direction and range of movements while using expansion bearings at both abutments.

A symmetrical layout is recommended, with the goal of achieving similar movement at each abutment in order to balance passive forces. Fixed bearings are usually installed over the center piers. Elastomeric expansion bearings may be used at piers of full integral bridges, provided that the bridge grade does not exceed 1%.

3.10. Jointless retrofit

In addition to the emphasis on the construction of new jointless bridges, there have also been innovative efforts to retrofit existing structures where practical. Figure 7 shows a continuous slab retrofit detail adopted by VDOT. This detail can be combined with the deck extension to transform an existing jointed structure to a fully jointless bridge.

Fig. 7. Continuous slab (link slab) retrofit detail.

4. Discussion

Successful implementation of jointless bridge construction involves continuous assessment of field performance and optimization of design methodology when feedback becomes available. Figure 8 shows a girder being installed at a full integral abutment. This type of construction requires the temporary support of girders. It is accomplished with leveling nuts and steel plates supporting the bottom flange at the required elevation. Bridge girders are installed on the anchor bolts placed in line with hinge dowels.

Fig. 8. Installing girder at full integral abutment.

The design of each structural element of a full integral bridge is relatively simple. The backwall is designed for passive earth pressure. It is treated as a continuous beam. The pile cap is designed to support vertical loads and passive pressures that develop during thermally induced movements. If a bridge is skewed, lateral loads also need to be considered.

Figure 9 shows two photographs of a large skew (45°) semi-integral bridge under construction. The abutment includes a buttress to resist superstructure rotation. The buttress is designed as a vertical cantilever.

Fig. 9. Skewed semi-integral bridge under construction.

Figure 10 shows example field data collected at the bridge in service during a hot summer day (Hoppe and Eichenthal, 2012). Buttress loads, measured by sensors installed at both abutments, follow the ambient air temperature pattern as the structure responds to thermal strains.

Fig. 10. Loads recorded at buttresses of semi-integral bridge.

Deck extension is essentially a hybrid solution involving a cross between conventional and semi-integral design. The deck slab is extended beyond the top of the backwall and a recessed drip bead is provided on the underside to mitigate water infiltration. A layer of expanded rubber, approximately 1 cm thick, is placed between the top of the backwall and the underside of the slab extension to seal the joint and allow for relative displacement. The exposure to passive earth pressure is limited because of the relatively shallow depth of the deck extension. Figure 11 shows an example of a deck extension in service.

Fig. 11. Deck extension example.

Figure 12 shows the EPS elastic inclusion, geotextile separation fabric, and select backfill material at the back of the semi-integral bridge under construction. Typically, the integral backwall overhangs the stub abutment and a drip bead is formed in the concrete at the protruded base to prevent water weeping along the backwall-abutment interface.

Fig. 12. EPS and select backfill material at the backwall.

The use of elastic inclusion mitigates roadway approach settlement by absorbing bridge displacements. Field measurements conducted periodically at integral bridge approaches indicate that this type of design is effective (Hoppe, 2005). No appreciable settlements were detected on a number of bridge approaches monitored after construction. Sometimes, limited pavement patching is required in the proximity of the bridge backwall shortly after the bridge is opened to traffic. Current VDOT practice is to specify EPS thickness on bridge plans and show placement details. In the case of a single span integral, the EPS material is placed on the upgrade abutment only.

Figure 13 shows three views of the Virginia Abutment in service. This design represents new type of jointless technology. The superstructure girders are embedded in the concrete end diaphragm much like the semi-integral abutment, although there is no direct contact with the backfill material. Unpainted weathering steel can be used, as there are no open joints in the proximity of girders and bearings. It significantly reduces the initial construction cost and future maintenance efforts. A buried or traditional approach slab may be constructed at the secondary backwall.

Fig. 13. Virginia Abutment in service.

Initially, VDOT imposed design restrictions on jointless bridges, recognizing that they behave differently than conventional structures. The restrictions were placed on the maximum span length, skew angle, and curvature. As performance feedback is obtained from field monitoring, these restrictions are being gradually relaxed, resulting in a more widespread use of this type of design. New jointless bridges constructed in Virginia have performed very well to date.

5. Conclusions

• The main justification for the use of jointless bridges is lower lifecycle cost.

• Jointless design requires accounting for thermally induced stresses.

• Steel H-piles oriented in weak-axis bending are optimal foundation elements for integral abutments.

• Design guidelines and structural details have been developed and adopted by VDOT.

• Full integral abutment is the primary design choice at VDOT.

Acknowledgments

The authors wish to acknowledge the engineers of the Virginia Department of Transportation who provided their technical guidance and dedication to implement jointless bridge design in Virginia. Ms. Linda Evans of the Virginia Transportation Research Council assisted with editorial work.

References

Arsoy, S., Barker, R.M., and Duncan, J.M., 2002. Experimental and Analytical Investigations of Piles and Abutments of Integral Bridges.

Virginia Transportation Research Council, Charlottesville, pp. 55. Hoppe, E.J. 2005. Field Study of Integral Backwall with Elastic Inclusion. Virginia Transportation Research Council, Charlottesville, pp. 36. Hoppe, E.J., and Eichenthal, S.L. 2012. Thermal Response of a Highly Skewed Integral Bridge. Virginia Center for Transportation Innovation

and Research, Charlottesville, pp. 39. Hoppe, E.J., and Gomez, J.P. 1997. Field Study of an Integral Backwall Bridge. Virginia Transportation Research Council, Charlottesville, pp. 50.

http://www.extranet.vdot.state.va.us/locdes/electronic%20pubs/Bridge%20Manuals/VolumeV-Part2/Chapter17.pdf. Accessed September 7, 2015.

Soltani, A. A., and Kukreti, A.R., 1992. Performance evaluation of integral abutment bridges. Transportation Research Record No. 1371,

National Research Council, Washington, D.C. VDOT Structure and Bridge Division, 2015. Design Aids and Typical Details, Volume V - Part 2.