A recent study of Seattle's Interstate 90 floating bridge proved the feasibility of adding a light rail transit system atop the structure. The exercise went beyond the typical analytical study, involving a full-scale load test to ensure that the 17-year-old structure spanning Lake Washington could support the additional weight of a transit line.
Initiated by regional public transportation authority Sound Transit and carried out by the Washington State DOT's Bridge and Structures group (WSDOT) and Seattle structural and civil consulting engineering firm KPFF, the study involved the use of eight flatbed trucks that were loaded to approximate the weight of light rail vehicles. Executed over two weekend nights in September 2005, the test gave the public agencies the information needed to seek voter approval for a transit funding package that would include light rail across the lake.
The I-90 corridor serves as the main connection between downtown Seattle and the region's “Eastside” of Lake Washington—a collection of cities including Bellevue, Kirkland, Issaquah, and Redmond, the home of Microsoft's headquarters. The test subject, known as the Homer Hadley Floating Bridge, is one of three concrete floating bridges over a mile long that span the lake to provide general purpose and high-occupancy vehicle (HOV) traffic lanes, eastbound and westbound. Opened in 1989, the test bridge supports westbound general-purpose traffic, reversible HOV traffic, and a pedestrian/bicycle lane. WSDOT owns and maintains the bridge.
In 2001, Sound Transit requested that WSDOT perform a preliminary analytical study to assess the feasibility of adding a light rail system to the bridge's reversible roadway on the south side of the floating structure. While the preliminary study performed by KPFF indicated that adding light rail may be feasible, it was determined that more advanced computer modeling would be required to confirm the bridge's ability to support a new light rail system. An option to perform a full-scale load test to evaluate bridge response in lieu of the more intense analytical study was also suggested.
Concrete floating bridges are unique structures in that they not only have to carry traditional vehicular traffic, but they also must remain watertight. In essence, floating bridges are permanently moored marine structures rather than conventional fixed bridges. The analytical methods used to predict the response of more conventional bridge structures typically are not applicable to floating bridges. The key advantage to performing a load test is the elimination of unknowns and assumptions inherent to any analytical study. Given the magnitude of future decisions that will be based on the findings of the second stage of the study, Sound Transit and WSDOT decided to perform a load test.
RECREATING REAL LOADS
The key objective of the load test was to simulate the computer modeling performed in the previous study by KPFF to predict bridge response. The eight flatbed trucks that were used to simulate light rail cars were loaded to approximately 148,000 pounds each, with two four-truck combinations each simulating a four-car light rail train. For comparison, the typical legal load limit for highway travel is 80,000 pounds. Sensitive instrumentation, including state-of-the-art GPS surveying equipment, was installed on the bridge to capture data as the bridge responded to the weight and motion of the trucks. Both static and dynamic tests were performed, with data collected and reported in real-time. Truck positioning on the bridge was selected to match the critical load conditions that would be encountered if a light rail system were installed.
The decision to move forward with the test was made in mid-June 2005, with a target for execution by the end of September. The schedule allowed only 3½ months to develop the test procedure, install instrumentation, coordinate traffic control, prepare the test vehicles, and schedule bridge closures. The target date was largely based on Sound Transit's desire to have preliminary results available by mid-November and to take advantage of reasonable weather conditions. High winds and excess rain could skew results, not to mention pose safety hazards. It was determined early that performance of the test would require complete closure of the bridge over multiple days. The window of opportunity to perform the test was further reduced by limiting bridge closure to weekend nights only. Due to previously scheduled major public events in the Seattle area, only one weekend was identified as being ideal for performing the test. If the team was not prepared to perform the test on that weekend, it would likely have been necessary to postpone the project until the summer of 2006. Doing it right the first time was crucial.