The new bridge on Maryland Rte. 140 over the Maryland Midland Railroad, Rte. 27 and West Branch in Carroll County, Md., may prove real-time effectiveness of high-performance steel (HPS) by using innovative distributed wire sensors. The old dual structures were built in 1952 and were six spans each with a clear roadway width of 37 ft.
They needed replacement due to deterioration. The piers and abutments were severely deteriorated with cracking throughout and had experienced repeated repairs. The sufficiency rating of both structures was 60.3. The new bridge was partially funded by the Federal Highway Administration’s Innovative Bridge Research & Construction (FHWA-IBRC) Program as an application of an HPS bridge. The replacement bridge is one structure with two additional lanes to make it an eight-lane bridge of total clear roadway widths of 61 ft northbound and 50 ft southbound with an additional 5-ft-wide sidewalk on each side of the bridge.
The installation of steel girders was divided into three phases over two construction seasons. The first phase, which covered the middle strip of girder numbers six to 11, was constructed during summer 2004. The second phase, which consisted of girder numbers one to five, was constructed during winter 2004. The last phase, which consisted of girder numbers 12 to 15, was constructed during summer 2005. Within each phase, there were three pouring sequences starting from the longer-span positive moment area to the short-span positive moment area. The last was the negative moment pour. For phases two and three, closure pours were added and the girder dead-load deflections at adjacent girders had to be verified. The bridge behavior during the construction was monitored and evaluated.
Load tests and structural monitoring are commonly used to gain information regarding the health and performance of an existing structure. For structures using relatively new materials, such as HPS, the use of load tests can prove the structure’s performance and capacities.
Long circuit
One goal of this study was to aim at demonstrating, theoretically and experimentally, the feasibility of utilizing the new class of distributed wire sensors. These sensors are used in monitoring the linear, angular and twist deflections of bridge decks under static and dynamic conditions. The study also aimed at using the proposed sensor to provide distribution maps of the deflection and strain fields over the entire bridge. Furthermore, the sensors and the obtained measurements were utilized to construct the power-flow maps (a plot of a family of curves for different power levels), which served as excellent means for monitoring the health of the bridge. Also, the power-flow map provides a metric that uniquely identifies the structural health in a manner that mimics biological systems that tend to redistribute the load and redirect its path away from the injured sites.
The wire network formed a Cartesian grid with a cross section of the plate-like bridge in the x-z plane, indicating the arrangement of a single wire oriented along the x-axis. The flexibility allowed the research team to place the sensors strategically inside the concrete deck or mount them on the surfaces of the steel beams. The distributed wire sensors were then integrated with an advanced telemetric wireless communication system provided by ATI, Spring Valley, Ohio. The transmitter circuit board of the ATI system was programmed to convert the sensor output signals into a serial data stream and transmit these signals via wireless radio link to a remote processing station for monitoring the bridge performance and health.
In the system, each wire segment was coupled with a miniature Wheatstone bridge connected to a channel of the transmitter via a multiplexer. The output of the transmitter was then telemetered to the receiver via a radio link. This approach improves the practicality of the sensor by eliminating the use of any wires connecting the sensor to the monitoring station and thereby enhancing the reliability of the entire monitoring system.
The wireless system used is small in size, extremely low in power consumption, reconfigurable by the end user, offers a complete multichannel data acquisition front end and delivers wireless digital data with error checking. It is based on an radio-frequency transmitter with surface acoustic wave (SAW) and a narrow band with carrier frequency of 2.4 GHz. The system is programmable over the serial port (RS-232) of any PC. It has gains that are programmable and digital filters that may be set from 1 to 250 Hz. Note that with the filter set at 1 Hz, a 22-bit resolution can be achieved theoretically; with the filter at 250 Hz the user can attain a 10-bit resolution. The sampling rate depends on the A/D converter resolution out of the module and the number of channels that are sampled. The system can provide a sampling rate of 160 Hz for 32 wire sensor segments. The transmission range of the system is in the order of 1.5 miles.
Data pours in
The research team coordinated with the Maryland State Highway Administration (MSA) and its construction crew over their pouring schedule. Prior to the pouring, the research team was on site to mark the steel plate girders to show the intended position and orientation for each sensor; place sensors to the steel girders; connect the readout and observe the reading; make a final check of the sensor reading; consider protection for the sensor; and take a set of readings immediately before and after the concrete is poured. After hardening, the concrete became sufficiently stiff to transfer its strain information to the sensors. Coordination is essential to the success of the sensor installation. The work plan was drawn up once the research team received the construction schedule and plan.
A prototype of the sensor was embedded inside a concrete slab (3 in. x 3 in. x 12 in.). The slab was tested until failure.
A prototype of the proposed sensor was bonded to one of the second phase girders (girder numbers one through five) of bridge Rte. 140. The bridge is 300 ft long. The sensor was then used to monitor bridge deflection during concrete pouring on the bridge deck.
The next step involved the true monitoring of the bridge behavior during concrete pouring, which started in September 2005. Girder numbers 12, 13 and 15 of phase three were selected and sensors were placed.
A truck was used for the controlled load test. The load test was conducted in November 2005 with a vehicle provided by SHA. Data was recorded continuously for each run and processed.
The final stage of this study covered the monitoring of the last portion of the bridge, which included both dead- and live-load effects. In this stage the last third of the bridge deck, which covers girders 12 through 15, was concrete poured on three separate days. Data also were collected from the field in three different days. The first two consecutive days were for concrete pouring and the third day was for truck load tests after the concrete had hardened. The testing equipment was prepared and connected to the steel sensors at the bridge site. The data acquisition equipment consisted of two systems: The first system was at the bridge side and connected directly to the sensors; the second was at a reasonable distance from the first system that kept it in a range to receive the wireless signal from the first system.
The first-day pouring results of the finite-element analysis and the test comparison are shown in Figure 1.
The live-load test was based mainly on a three-axle dump truck of a gross weight of 64,000 lb. Figure 2 shows one of the live-load test results with deflections from finite-element analysis and the sensor measurements.
The test truck was placed at the far end of the bridge such that the truck line of symmetry coincides with the girder in consideration. The truck moved along the bridge to the near end in two different speeds, slow and fast. Test results were recorded and analyzed and then the static test results were compared with those of the finite-element analysis. The load-deflection curves of the static load test were plotted at two different locations of the girder in consideration, specifically, at 0.4L from each end.
Comparison of the field test results with results obtained from theoretical analysis using finite-element models showed very good agreement in most cases. Based on these results, it shows that the HPS bridge behaves well within the designed value even with the complicated phased construction.