top of page

I-91 Interchange 29 Exit Ramp Flyover Bridge

I-91 Interchange 29 Exit Ramp Flyover Bridge

Medium Span

For a long time, Interchange 29 in Hartford, Conn., was notorious for congestion.

The interchange connects northbound I-91 with Route 5/15, the latter being the major connector between I-91 and I-84 in East Hartford. The original ramp was a single-lane ramp with a steep grade and a significant traffic weave at the intersection with Route 5/15 and saw significant daily back-ups on I-91 that led to numerous accidents and delays. Improvements to the interchange were one of the top priorities of the Connecticut DOT (CTDOT), and the reconfiguration of the interchange resulted in a new high-speed two-lane ramp that crosses over southbound Route 5/15 in a weave configuration.

The new ramp is a straight ramp that crosses a curved roadway at a very flat angle, resulting in significant geometric impacts on the roadway below. The vertical geometry of the roadway below the bridge limited the ability of vehicles to pass under the proposed hammerhead pier caps due to low vertical clearance at the hammerhead piers. There were three potential solutions: raising the bridge, lowering the roadway, or reducing the pier cap’s width. The first two options weren’t feasible, so the team moved forward with the plan of reducing the pier cap’s width and implementing trapezoidal box girders. This solution allowed the design team to locate the bridge bearings closer to the centerline of the bridge, thereby reducing the width of the pier cap by 8 ft. The reduced cap width also reduced the cost of the piers by reducing the volume of concrete and the bending moments acting on the shorter cantilevers. The geometric layout of the bridge also improves its look, as the trapezoidal box girders without exterior stiffeners produce clean lines. When compared to vertical webs, the sloped webs have historically been the look of choice for bridge aesthetics, as the sloping webs draw the eye toward the single columns supporting the pier caps, demonstrating a flow of forces from the superstructure to the ground.

Another major factor that makes this bridge stand out is its innovative use of straddle bents. The goal was to design a redundant beam, and the team incorporated the load path redundant members (LPRM) approach. The team accomplished this by converting a typical single-cell box girder section into a three I-girder member. Plate diaphragms were designed using finite element analysis (FEA) to distribute forces equally to each girder and transfer the load should one girder flange fracture. The team also developed an “integral,” or framed-in, straddle bent concept and a “stacked” straddle bent scheme with the superstructure on top. There was adequate vertical clearance at the straddle bent location to stack the members, leading to a simpler and more cost-effective design. The design team has developed similar details for an integral “framed in” design. Therefore, the triple I-girder design can be adapted to virtually any steel bridge configuration.

The straddle bent approach used for this project represents a game-changer in the world of steel bridges. To date, all steel straddle bents—again, typically single-cell box sections—have been classified as fracture-critical elements, which has significantly precluded the use of steel for straddle bents. The triple I-girder configuration can provide load path redundancy, thereby eliminating the fracture-critical designation and the related long-term inspection requirements. In addition, the girders can be designed for infinite fatigue life, essentially eliminating the potential for a fatigue crack to develop, let alone a

fracture. The triple I-girder design also provides options to the contractor for shipping and handling. The straddle bent can be shipped and erected as one, two, or three pieces, which allows the contractor to achieve maximum efficiency when it comes to truck size and crane size, potentially eliminating an overweight permit, which can lead to reduced costs. This proved to be the case on the Interchange 29 ramp bridge, as the contractor chose to ship the straddle bent girder in two pieces. Once on-site, the two pieces were bolted together on the ground and erected as one piece.

The triple I-girder straddle bent concept offered another surprising benefit: It’s a very economical section to fabricate. During design development, when considering fabrication costs, the design team initially felt that the fabrication of three members might be slightly more expensive than the fabrication of a single box girder. The idea was that while the total flange areas of the triple I-girder would be similar to the box girder, the triple I-girder would have three webs as opposed to two, which might increase costs. But the team moved forward with the triple I-girder option since the long-term savings in reduced fracture-critical inspections would offset the perceived initial cost.

Surprisingly, the design team was wrong. The fabrication cost for the triple I-girder turned out to be substantially less than the equivalent box girder, and the fabricator identified a couple of reasons why. Box girders typically require full-penetration groove welds between the webs and the flanges. In addition, some designers specify bolted connections for these locations to provide internal redundancy and obtain a fatigue Category B member. Groove welds and bolting can be very expensive and time-consuming to execute in the shop. Conversely, welding a web to an I-girder flange is a common shop process using conventional beam fabrication equipment, making it very cost-effective. Secondly, welding stiffeners and connection plates on the interior of a box girder is costly due to confined space work that is time-consuming and comes with increased safety risks. While the triple I-girder beams do require interior diaphragms with bolts, modern CNC machinery can quickly cut and drill the plates and holes for the diaphragm. The result of these factors is that the triple I-girder straddle bent can be as much as 50% less than the cost of an equivalent box section.

When it came to corrosion protection for the ramp’s superstructure, CTDOT chose uncoated weathering steel. The department has a long history with uncoated weathering steel, dating back to the early 1960s, and recently completed a study of its performance. It found the performance of weathering steel bridges with quality details to be very impressive. In addition, some of the oldest uncoated weathering bridges are still in very good condition after more than 55 years in service, further reinforcing the state’s commitment to this corrosion-protection option.

Word has spread about this design. The Texas and Georgia DOTs, two entities that traditionally use concrete straddle bents, have both agreed that the triple I-girder bent is acceptable for widespread use. In the case of the Georgia DOT, steel straddle bents were previously not even allowed for use. Their reversal on this matter is a testament to the design’s significance and impact on the steel bridge industry, and these two states and others are looking to make this design a key tool in their steel bridge toolboxes.

Project Team

  • Steel Team

    • Fabricator: High Steel Structures, Lancaster, Pa. *AISC CERTIFIED*

    • Erector: Hartland Building and Restoration Company, East Granby, Conn. *AISC CERTIFIED*

    • Detailer: ABS Structural Corporation, Melbourne, Fla. *AISC CERTIFIED*

  • Owner: Connecticut Department of Transportation

  • General Contractor: O&G/BHD, JV

  • Structural Engineer: CHA Consulting, Inc.


Year Awarded:


Year Completed:



Hartford, Conn.

Award Class:

Medium Span

Award Type:

National Award


Structure Type:

Coating System:

Weathering Steel

Span Length (ft):

140 ft, 215 ft, 215 ft, 170 ft, 140 ft

Structure Length (ft):


Average Deck Width (ft):


Steel Weight/Deck Area (lb/ft²):

Amount of Steel (tons):


bottom of page
<--feathr--> <--end feathr-->