Steady Work

July 24, 2006

The road to a successful seismic rehabilitation strategy is often long and arduous. The seismic design of new bridge structures, and the related component of seismic retrofit, demands an extensive engineering effort and an unusually significant amount of decision making throughout the process. However, this approach, and the necessary time involved, yields numerous practical advantages.

Examining candidates

The road to a successful seismic rehabilitation strategy is often long and arduous. The seismic design of new bridge structures, and the related component of seismic retrofit, demands an extensive engineering effort and an unusually significant amount of decision making throughout the process. However, this approach, and the necessary time involved, yields numerous practical advantages.

Examining candidates

Initially, the engineer must determine which code, or combination of codes, is most appropriate. The seismic performance category must then be selected and the structure defined in terms of critical versus essential versus regular versus unusual versus other classifications that are defined in the applicable codes. This is followed by determination of suitable design-response spectra.

Next, computer analysis programs must be evaluated with an understanding that the type and complexity of the structure will often dictate the software requirements. Generally, several program iterations are required to arrive at reasonable outputs that make sense. When working on a rehabilitation project, the next step typically involves the calculation of capacity and demand ratios for the critical members. This is followed by identification of the elements that require modifications in order to comply with either force or displacement demands.

At this point in the process, if the designer has any energy remaining, the focus shifts to the determination of actual rehabilitation strategies. The following discussion will concentrate on these strategies and be limited to bearing and other superstructure retrofits.

Supporting the dead, the live and the lateral

On a recent project on I-80 for the New Jersey Department of Transportation, seismic retrofit was a major component in the goal to upgrade an important section of the interstate to meet current standards. The project consisted of the widening, redecking and rehabilitation of three very complex bridge structures. These bridges could generally be described as multi-span, severally skewed and both short- and long-span steel girder structures. Each structure contained one or more transverse steel-box cross girders arranged in a stacked condition relative to the longitudinal girders. The interstate route, and the structures within the project limits, was declared by the owner to be "essential."

In accordance with the 1995 version of the Federal Highway Administration's (FHWA) Seismic Retrofitting Manual for Highway Bridges, essential bridges must remain fully operational after a large seismic event. This means that any seismic retrofit should be designed and configured to support not only the vertical dead loads and live loads after an event, but it also must be prepared to resist any of the combinations of lateral loads that may be imposed. This requirement was to be in effect until the construction of permanent fixes.

Rocker bearing location, several alternatives

As was very common for many other structures built during the same era as those on I-80, these bridges were constructed with pedestal and rocker bearings. Experience shows that rocker bearings almost never provide the displacement capacity required for even a moderate seismic event. Therefore, the first goal was to determine the best approach for dealing with the consequences of failure, or toppling, at the rocker bearing locations.

One of the options considered was total bearing replacement with a substitute device that could resist all normal forces, as well as seismic loads and movements. However, replacements would have to be completed while maintaining traffic and all other in-service loads. The additional costs associated with this requirement made the bearing replacement option less desirable than other schemes.

The strengthening or replacement of individual components of existing bearing elements also was considered as a method to improve seismic capacity. However, increasing section properties of existing plates, anchor bolts, pins and other bearing elements would have been awkward and costly.

An alternative approach was developed that avoided strengthening existing elements and eliminated the need to temporarily support external loads during construction of the retrofits. This involved the design and detailing of a seismic catcher system that was nearly independent of the existing superstructure, both during and after construction. Any work that was required on the existing structure to accommodate the catcher system could be performed under traffic.

A simplified sketch of this system is shown in Figure 1. Note that both expansion and fixed bearings were modified. However, fixed bearings were only altered when seismic lateral load capacities were exceeded.

Catcher system

Fixed bearings were stabilized for a seismic event by transferring the lateral loads directly into horizontal keeper bars and then into the independent catcher system. At the expansion rockers, the catcher system was configured to allow normal movements until bearing rollover occurred. At that point, the horizontal bars attached to the girders would come to rest on the catcher assembly, and the vertical elevation of the bridge would be maintained. More importantly, the bridge would remain in service and the functional requirements and maintenance of traffic goals would be achieved.

The length of the assembly at the expansion bearings was a function of the displacement demand at each bearing. Note also that the system at a double bearing could be developed as a single, larger device and function for both the fixed and expansion bearings. Perhaps an obvious point, the catchers would need to be anchored to the bridge seats. Special precautions were taken to avoid conflict between anchor bolts and the design steel in the top of pier caps. Providing anchor bolt patterns parallel to the main pier cap rebar minimized potential interferences. The contract documents also required the contractor to locate reinforcement prior to drilling in the caps.

Consistent load path

The most important feature of this catcher system was the fact that the load path from superstructure to substructure did not change during a seismic event. Simply put, fixed bearings remain fixed and expansion bearings remain the same after a major event. In other words, seismic displacements were accommodated, rather than prohibited. Since the integrity of fixed- and expansion-bearing lines was not altered under earthquake loading, structural modeling for seismic analysis was the same as for any other group loading combination. This can be very advantageous on structures exhibiting large differences in strength capacity between fixed and expansion substructure units. With this system, potentially expensive substructure retrofit costs can be avoided.

The catcher system shown in Figure 1 is located completely above the bridge seat and could be considered generally "out of site." This approach is especially suited for localities where appearance could be a sensitive issue. Where this is not the case, and other factors outweigh the visual impacts, a system with catchers anchored to the vertical faces of the substructure units might be more desirable from a cost and constructability perspective.

Stacked Cross-Girders

On the I-80 project, the superstructure framing system included several "stacked" cross girders that required special attention. A stacked system, by definition for this article, consists of longitudinal steel beams or girders with conventional bearings placed on top of transverse steel cross girders. The cross girders were, in turn, supported on conventional steel pedestal and rocker bearings placed on concrete columns. Given this "hinged" support configuration, the entire system would be considered unstable under any extreme event. During a seismic episode, the cross girder would likely "roll over" and result in total collapse of the structure. Therefore, a decision was made to stabilize these cross girders.

Given the tight constraints at the bridge sites, the restraining system shown in Figure 2 was considered the most feasible. As shown in the detail, high-strength rods were loosely fitted to the existing cross-girder bearing stiffeners. The loose fit allowed for normal thermal movements in the transverse direction of the bridge but restrained toppling in the longitudinal direction by engaging the stabilizer rods. The sharp inclination angle resulted in large loads and the need for high-strength rods to provide the necessary strength.

Transverse rotational freedom near the base was permitted through a clevis mechanism. The clevis also permitted adjustments in the length of the anchor rods. Anchorage to the pier was accomplished with anchor bolts embedded in a reinforced concrete encasement wrapped around the existing substructure elements. To achieve composite action with the existing columns, tie bars were detailed to be drilled into the existing concrete. This system that anchored the encasement concrete also provided confinement reinforcement for the concrete substructure elements.

Catcher is behind it all

On the Norfolk Southern Railroad Truss Bridge over I-76 in Montgomery County, Pa., retrofit strategy was actually utilized in the design of a new structure. Given the normally large vertical reactions on long-span railroad bridges, combined with a philosophy of not straying from a standard that has worked for many years, large rocker bearings were proposed at the expansion ends of the truss spans. However, similar to the I-80 project, this bridge was declared essential. As such, rail traffic was to be maintained on the structure both during and after a seismic event.

To comply with the traffic maintenance criteria, a catcher system was incorporated into the original design and construction. This system consisted of a large concrete pedestal located between the two bridge bearings and placed directly on the concrete bridge seat. The pedestal and end floor beam of the truss were designed to support the entire load of the bridge in the event of a rocker bearing failure. The floor beam and pedestal were separated by a small vertical gap to allow normal in-service movements.

Common collaboration

The retrofits described in this article consist of common components and offer advantages in ease of design and detailing, constructibility and economy. Through providing restraint, while accommodating movement, the retrofits provide adequate seismic connection between superstructure and substructure, yet allow the bearings to function normally in transmitting static and other forces.

The designs described in this article are generally in compliance with the FHWA’s Seismic Retrofitting Manual for Highway Bridges, May 1995.

Please note, in June or July of this year, a new and revised document will be published by FHWA titled New Seismic Retrofitting Manual, Part I: Bridges and Part II: Retaining Structures, Slopes, Tunnels, Culverts, and Pavements.

The availability of this new manual can be checked on FHWA’s Turner-Fairbank website ­at www­.tfhrc­.gov.

About The Author: Miller is a project manager for Gannett Fleming Inc.

Sponsored Recommendations

The Science Behind Sustainable Concrete Sealing Solutions

Extend the lifespan and durability of any concrete. PoreShield is a USDA BioPreferred product and is approved for residential, commercial, and industrial use. It works great above...

Powerful Concrete Protection For ANY Application

PoreShield protects concrete surfaces from water, deicing salts, oil and grease stains, and weather extremes. It's just as effective on major interstates as it is on backyard ...

Concrete Protection That’s Easy on the Environment and Tough to Beat

PoreShield's concrete penetration capabilities go just as deep as our American roots. PoreShield is a plant-based, eco-friendly alternative to solvent-based concrete sealers.

Proven Concrete Protection That’s Safe & Sustainable

Real-life DOT field tests and university researchers have found that PoreShieldTM lasts for 10+ years and extends the life of concrete.