New Carquinez Bridge

From LIST Wiki
Jump to: navigation, search

The Carquinez Bridge refers to two parallel bridges which cross the Carquinez Strait linking Vallejo, California to the north, with Crockett, California to the south. The bridges are signed as part of Interstate 80 in California. Toll is only charged to eastbound traffic.


History and description

The original bridge, a steel cantilever bridge, was designed by Robinson & Steinman and dedicated on May 21, 1927. It cost $8 million to build and was the first major bridge in the San Francisco Bay Area.

Upon completion of the 1927 bridge, the Lincoln Highway was rerouted over the span. The Lincoln Highway was the first road across The United States. Its original alignment from Sacramento to San Francisco avoided the un-bridged waterways of the San Francisco Bay and the Sacramento Delta by routing itself through the Altamont Pass and the central valley. (Traffic aross the Carquinez Strait was by steam ferry.) But with the bridge built, the Sacramento to San Francisco route was realigned in 1928 to pass along the eastern shores of the San Francisco Bay and San Pablo Bay, then in a northeastern direction.

In 1958, at a cost of $38 million a similar parallel bridge was built alongside the original one to accommodate the ever-increasing traffic. The original 1927 span served westbound traffic while the newer 1958 span served eastbound traffic.

Carquinez Bridge in 2006 with the 1927 span in the center.

In 2003, as a resolution to seismic problems of the aging 1927 span, a new suspension bridge was opened to replace it, at a cost of $240 million. This new bridge was named the Alfred Zampa Memorial Bridge, after an ironworker who worked on a number of the San Francisco Bay Area bridges, including the Golden Gate Bridge. This span features a pedestrian and bicycle path, completing a bike trail which circles the entire Bay Area. The span measures 0.66 miles (3465 feet / 1056.1 m / 1.06 km). The bridge was dedicated on November 8, 2003 and opened for traffic on November 11. (Originally, the plan was to dedicate the bridge on November 15, but complications involving when just-recalled Governor Gray Davis would have to transfer power to Arnold Schwarzenegger resulted in the date being moved. The coins minted to commemorate the event have the old date on them. The 1927 span was dismantled in 2007, after it was temporarily used to hold eastbound traffic while the eastbound span underwent a seismic retrofit, deck and superstructure rehabilitation, and painting to extend its serviceable life. During demolition, the 3,000-pound bronze bell atop one of the Carquinez Bridge piers was removed and placed into storage. The bell will eventually be displayed in a new museum to be built at the Oakland end of the San Francisco – Oakland Bay Bridge.


The new Carquinez Bridge is a suspension bridge with spans of 147 m, 728 m, and 181 m. Built by the Cleveland Bridge & Engineering Company of Darlington, England, it consists of the south anchorage, a transition pier, two towers (South and North towers), and the north anchorage. The towers are each founded on two footings, which are each supported by six vertical, 3-m-diameter steel shells infilled with reinforced concrete, followed by 2.7-m-diameter drilled shafts in rock (i.e., cast-in-drilled hole, or CIDH, piles). The total length of the CIDH pile at the South Tower is approximately 89 m, with about 43 m of drilled shaft in rock. The total length of the CIDH pile at the North Tower ranges from 49 to 64 m, with about 16 to 26 m of drilled shaft in rock. The design parameters used for the South Tower piles were later confirmed by a pile load test. Additional field investigations during construction revealed significant variations in rock conditions at the North Tower, resulting in the redesign of the length of the piles. Major construction challenges encountered during construction of the South Tower piles, and the revised construction procedure (i.e., under-reaming) used by the constructor to mitigate caving.

By September 4, 2007, all of the original 1927 steel structure had been demolished.

Carquinez Bridge in the media

New Carquinez Bridge Long-Term Monitoring Study

The New Carquinez Suspension Bridge (Fig.8.1) is a major suspension bridge built in 2003 over the Carquinez Straights and connects Vallejo with Crockett, CA. The bridge carries four lanes of westbound I-80 traffic. The bridge is constructed from a steel orthotropic box girder supported by two major suspension cables and two reinforced concrete towers (referred to as the south and north towers). The total length of the bridge is over 1,000 m (3,500 ft) and the main span between towers is over 700 m (2,300 ft). As the bridge is located in a high seismic region, the bridge owner, Caltran, was especially interested in the assistance in the decision making procedures provided by the cyber-physical system. A long-term wireless structural monitoring system was installed starting in the summer of 2010 with the system in continuous operation now for over 2 years [4, 5]. The wireless monitoring system installed is based on the use of a low-cost wireless sensor node developed at the University of Michigan termed Narada [6]. Narada is a wireless data acquisition system specially designed for monitoring civil infrastructure systems where low power consumption (i.e., rechargeable battery operated), high data resolution (i.e., 16-bits or higher), and long communication ranges (i.e., 500 m or longer) are all required system capabilities. On the New Carquinez Suspension Bridge, a total of 33 Narada nodes each capable of collecting up to 4 independent channels of data were installed over a 2 year period with various sensing transducers interfaced: 23 tri-axial accelerometers (to measure deck and tower accelerations), 3 string potential meters (to measure longitudinal movement between the deck and the towers), 33 battery voltage detectors (to monitor the battery level on each Narada), 9 thermistors (to measure the ambient and girder temperatures), 2 wind vanes, 2 anemometers and 6 string gages(to measure deck bending strains). A total of 124 sensing channels have been deployed on the bridge as summarized in Fig.8.1. The Narada nodes (Fig.8.2a) are all powered by rechargeable batteries that are continuously charged by solar panels installed on the top deck of the bridge (Fig.8.2b).

Figure 1: Wireless sensor network deployment on the New Carquinez Suspension Bridge
Figure 2: Installation details ofthe New Carquinez Suspension Bridge long-term monitoring system: (a) typical Narada Wireless sensor node beneath the deck girder in a magnetically mounted Weather-proof enclosure; (b) small solar panel installed on the bridge top deck to power a Narada node; (c) string potentiometer at the deck-tower interface to measure longitudinal displacement.

The monitoring system is supported by a custom-designed cyber environment that seamlessly integrates the wireless sensors with the Internet where data can be stored in remote database systems and processed using powerful analytical tools. At the bridge, the wireless monitoring system is divided into three sub-networks: one sub-network includes all of the wireless sensors on the girder centered near the north tower (sub-network #1), another sub-network includes all of the wireless sensors on the girder centered near the south tower (sub-network #2), and the third sub-network included the nodes on the top of the two towers (sub-network #3). Each sub-network is managed by a Linux server implemented on an inexpensive single-board computer. The functionality of the server is to: (1) time synchronize the server clock using the network time protocol (NTP); (2) time synchronize the wireless sensors in each sub-network using a beacon time synchronization protocols previously developed for Narada; (3) command the wireless sensors to collect data; (4) locally store data on the on-board hard drive; and, (5) communicate data to the Internet via a cellular modem. The servers can collect data on either a schedule or upon user-demand; their default data collection occurs every 4 h with 8 min of bridge response data collected.

A massive database server termed SenStore has been created off-site to store the data and to expose application programming interfaces (APIs) to data clients seeking to use bridge data in their data processing algorithms. SenStore is a data management system designed for structural health monitoring (SHM) applications [4]. It uses a relational database to store bridge meta-data including structural component definitions, geometric details, sensor network data such as sensor type and installation location, and structural analysis information (e.g., definitions needed for finite element modeling). Because the massive amounts of raw time history data from sensors are ill-suited for storage in a traditional relational database, a secondary database system based on the HDF5 file system is implemented in SenStore for the storage of sensor data. The relational database links directly to the HDF5 file system so that queries for sensor data can be handled by the SenStore server through the relational database. The server–client model implemented in the design of SenStore allows secure access to bridge data through a well-defined set of APIs. Computational tools are designed to perform autonomous system operation and to undertake data analysis. Tools for tracking sensor operation (i.e., fault detection), data visualization, and modal parameter extraction, among others, have been already created [7].

Modal Analysis Results of New Carquinez Bridge

Figure 3. Sample acceleration time-history data (left) and corresponding frequency domain representation (right).

Modal properties of an invariant system are fixed in theory. However, for operational bridges, the operational environment changes thereby introducing some variance into the system; therefore the observed modal properties are not fixed. The New Carquinez Suspension Bridge, as one of the largest bridges in the Bay Area, experiences continual heavy traffic loads and large temperature variations. With the implementation of the long-term structural monitoring system, automated modal analysis tools are used to extract modal properties at six different time periods over a day. Figure 8.3 is a typical plot of vertical acceleration data of the deck in the time and frequency domains; this type of data is used to extract modal properties using SSI. The relationship between modal frequencies and environmental variables is studied. In detail, the relationship between modal frequency and temperature, modal frequency and time of the day at different days in the week, and the relationship between horizontal movement of the bridge deck and temperature are all studied.

Relationship Between Modal Frequencies, Temperature and Traffic

Temperature change in the Bay Area varies in the range of 8–40 C during a typical day in the summer. This large temperature change will impact the material properties of the steel girder box of the bridge. The orthotropic steel box girder of the New Carquinez Suspension Bridge is connected to the concrete footings at the wind tongues located at the towers. It is allowed to move in transverse, longitudinal and vertical directions to compensate for imposed loads. To evaluate and monitor the performance of the bridge, the behavior of the deck movement is studied for the testbed structure. Data is collected from the string potentiometers installed longitudinally at the wind tongues (Fig.8.2c) of both towers every 4 h. As the initial position between the girder and the wind tongues is impossible to know, only the relative deck movement is obtainable as temperature varies (Fig.8.4). This data verifies deck contraction with reducing temperature. The relative displacements from both sides of the bridge (at north and south towers) are also linearly related with a slope of 1 (Fig.8.4c). As the way the string potentiometers are installed, the positive direction is defined as north for data from the north side and south for data from the south side. These facts indicate the bridge girder expands in the longitudinal direction when temperature increases and shrinks when the temperature decreases, which obey thermal expansion rules as expected. The contraction behavior due to temperature changes is symmetric when looking at the north and south sides of the bridge.

Figure 4: Relationship between the bridge girder longitudinal displacements and temperature

To evaluate the performance of the bridge with the change of temperature, studies on the modal frequencies and temperature are performed. As a large scale structure, the New Carquinez Suspension Bridge has very low natural frequencies as shown in Fig. 8.3with the first mode near 0.19 Hz. Due to the design of the structure, the second modal frequency is very close to the first mode, roughly 0.01 Hz a part. As a result, the second mode is not always autonomously detected by the automated modal analysis tools. In addition, the sixth mode is a torsional mode which has less energy and thus more difficult to detect. Therefore, the first, third, fourth, fifth and seventh modal frequencies are used for most of the analyses presented herein. Figure8.5shows the distribution of modal frequencies as a function of temperature. Modal frequency data presented in Fig.8.5are obtained using the SSI method and least square linear fitting is employed to present the trend of the data. It is shown that the 4th and 7th modes have stronger trending behavior of decreasing modal frequency as temperature increases as compared to the other three modes. However, decreasing modal frequencies as a function of temperature are generally observed for all of the modes. The modal frequency floating range is within 2 % of the mean value. As the observed data is not evenly distributed over the temperature range, it is hard to make a direct conclusion of the variance of the data for different temperature ranges.

Figure 5: Relationship of modal frequency and temperature: (a) lst mode, (b) 3rd mode, (c) 4th mode, (d) 5th mode, and (e) 7th mode.
"Figure 7: Mean of the modal frequency distribution as a function of time of day."

Figure 8.6 shows the distribution of modal frequencies as a function of the time of the day. The sensor networks are configured to acquire data 6 times a day separated by 4 h apart. Data from roughly 60 consecutive days are included in this plot and spline of the mean values at each sensing point is fitted to show the data trend. From Fig.8.6b, c, d, e, the data shows the trend that the modal frequencies decrease during the early half of the day and increase in the late half of the day. Besides the temperature variation (Fig.8.6f), change of the traffic load is another important event that happens during the day which would also temporarily change the performance of the bridge because traffic imposes dynamic loads and increases the weight of the structure. It is difficult to acquire accurate traffic load data, but a good estimation could be made based on the relationship of traffic and time of the day. Traffic loads are higher during the day time compared to the night time and the peaks occur at rush hour (8 a.m. and 5 p.m.). In Fig.8.6d, e, the frequencies seem to dip lower at rush hour as compared to frequencies from other non-rush hour periods. In Fig.8.7the mean of the normalized frequency data of each mode and the mean of normalized temperature data at each sampling time are presented on the same figure. The amplitudes of the modal frequency data curves are amplified by 20 times to exaggerate the trend. Modal frequency data are fitted by splines. Only the first mode data behaves differently than the other modes in the trending over time of the day. The rest of the modes (third, fourth, fifth and seventh) have similar general trending with frequency declining while the temperature is rising and vice versa. Therefore, the influence of temperature and traffic loads on the New Carquinez Bridge is certain and not negligible. For a certain mode, the bridge tends to operate in a frequency lower than the natural frequency as the temperature and/or the traffic loads increase and tends to operate in a frequency higher than the natural frequency as the temperature and/or the traffic loads decrease.

Figure 6. Relationship 0f modal frequency and time ofthe day: (a) lst mode, (b) 3rd mode, (c) 4th mode, (d) 5th mode, (e) 7th mode, and (f) temperature


[1] AU2 Shakal A, Huang M (1989) Overview of the strong motion instrumentation program. In: Seminar on seismological and engineering implications of recent strong-motion data

[2] AU3 Cross E, Koo K, Brownjohn J, Worden K (2012) Long-term monitoring and data analysis of the tamar bridge. Mech Syst Signal Process

[3] Ni Y, Hua X, Fan K, Ko J (2005) Correlating modal properties with temperature using long-term monitoring data and support vector machine technique. Eng Struct 27(12):1762–1773

[4] Kurata M, Kim J, Lynch J, Linden G, Sedarat H, Thometz E, Hpley P, Sheng L (2012) Internet‐enabled wireless structural monitoring systems: development and permanent deployment at the new carquinez suspension bridge. J Struct Eng

[5] Kurata M, Kim J, Zhang Y, Lynch J, van der Linden G, Jacob V, Thometz E, Hipley P, Sheng L (2011) Long-term assessment of an autonomous wireless structural health monitoring system at the new carquinez suspension bridge. In: SPIE, San Diego

[6] Swartz R, Jung D, Lynch J, Wang Y, Shi D, Flynn M (2005) Design of a wireless sensor for scalable distributed inNetwork computation in a structural health monitoring. In: 5th international workshop on structural health monitoring, Stanford

[7] Zhang Y, Kurata M, Lynch JP, van der Linden G, Sadarat H, Prakash A (2012) Distributed cyberinfrastructure tools for automated data processing. In: SPIE, San Diago

[8] Peeters B, de Roeck G (1999) Reference-based stochastic subspace identification for output-only modal analysis. Mech Syst Signal Process 13(6):855–878

[9] Van Overschee P, De Moor B (1996) Subspace identification for linear system: theory, implementation, applications. Kluwer Academic, Dordrecht

[10] AU4 Brincker R, Zhang L, Andersen P (2001) Modal identification of output-only systems using frequency domain decomposition. Smart Mate Struct 10(3):441–445

[11] Ni Y, Fan K, Zheng G, Chan T, Ko J (2003)Automatic modal identification of cable-supported bridges. In: SPIE

[12] Kim J, Lynch J (2011) Comparison study of output-only subspace and frequency-domain methods. In: Society for Experimental Mechanics Series

Personal tools