Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
This magazine is an archived publication and may contain dated technical, contact, and link information.
|Publication Number: Date: July/August 1999|
Issue No: Vol. 63 No. 1
Date: July/August 1999
Thousands of motorists driving across a nondescript highway bridge on Interstate 10 in Las Cruces, N.M., have no idea that they are participating in a landmark research program. Although a casual observer would never guess it, this bridge has set a world record. Throughout its structure are 120 calibrated Bragg grating fiber-optic strain sensors - more than any other steel bridge in the world.
This Las Cruces highway bridge is the front-line project in a far-sighted Federal Highway Administration (FHWA) research program that is exploring the use of fiber-optic sensors in highway applications, including both structures and pavements. This research is yielding valuable information about highway construction, ranging from the monitoring of stresses in bridge beams during fabrication and transport to the testing of bridge design ratings and the monitoring of deterioration of the structure over time.
The magnitude of the need to monitor the infrastructure is huge. Approximately one-half of the 260,000 kilometers of our National Highway System (NHS) is rated as poor to fair.1 Of the 576,000 bridges in NHS, 187,000 are rated as deficient.2 Bridge inspection in the past has relied primarily on simple visual inspection supported at intervals with diagnostic equipment. Visual inspection cannot evaluate the condition of steel beneath concrete or of steel beams beneath paint, and it often takes four or five professionals a full week just to inspect a single bridge. Furthermore, visual inspection is not only expensive and time-consuming, it can also be dangerous for inspectors.
Although state and local governments have the responsibility for managing highway bridges and other infrastructure assets, FHWA contributes by conducting a research and development program.3 The portion of this program concerned with fiber-optic sensors is located at the Turner-Fairbank Highway Research Center (TFHRC) in McLean, Va., and has been funded at about $300,000 per year. Specifically, applications of the Bragg grating-type fiber-optic sensors are being evaluated under an interagency agreement with the Naval Research Laboratory (NRL).
New Technology Allows Progress
The technology of fiber-optic strain gauges based on Bragg gratings permits as many as 100 gauges to be put on the same fiber, which can be calibrated to form a system that provides far more valuable information than individual gauges.4 The installation of the gauges is simplified; the cabling requirement is reduced; and the cost-per-sensor is lowered. Also, because the Bragg grating system operates by sensing the wavelength of the light reflected back from the grating, it can be more reliable and repeatable than other types of fiber-optic strain gauges that use other measurement methods such as interferometry.
Traditionally, structural monitoring has typically involved networks of only 10 to 20 strain gauges, yielding data that is generally insufficient for detecting early signs of deterioration. The new capability of installing large numbers of sensors in a system suggests a new, much more reliable approach. It could make possible applications in which networks on the order of a thousand sensors, or one kilosensor, may be required.5An extensive network for the very largest structures could conceivably reach megasensor numbers. Such a dense array of sensors would enable the use of advanced computational methods of data analysis.
While the application of fiber-optic sensor networks to all 576,000 bridges in the nation's infrastructure may not be justified from a cost-benefit perspective, the technology presents considerable advantages for three categories of bridges in particular: (1) existing bridges that have severe deterioration; (2) major bridges that form critical links in urban transportation systems, especially in seismically active areas; and (3) bridges made of novel materials, such as polymer composites, for which little information on performance currently exists.
At present, this key FHWA research program is seeking to provide an information base for an effective cost-benefit analysis, investigating issues such as overall monitoring strategy, systems architecture, installation procedures, temperature compensation, durability, compatibility with other materials, sensor standardization, and data analysis. In the short term, fiber-optic sensor systems do involve increased costs; however, FHWA is attempting to discover what increased benefits and cost-saving advantages they may have for the long term.
Applications of Fiber Optics
Fiber-optic systems can be used in infrastructure applications for a wide range of purposes that involve different time and spatial scales. One set of applications primarily concerns the construction period. The sensors could be used to monitor the prestress forces applied to the steel strands in precast concrete components. This would be used initially to verify that the correct stresses were applied. Subsequently, it could measure the development length, defined as the distance along the strand in which the bond between the steel and the concrete is eventually lost.6 Another application would be to monitor the strains produced in the concrete by shrinkage during the initial hardening process.
Finally, the system could be used when the structure is completed to verify the design calculations. This consists of placing known loads on the bridge and observing the resulting deflections.7
An application for the post-construction period could be continuous monitoring to detect incidents, such as the fracture of a member or a severe impact (for example, a barge colliding with a bridge pier), that would require immediate response.
Instead of detecting a discrete event, the fiber-optic monitoring system can be used to sense a change in the mechanical properties - even a subtle change, such as the aging of asphalt, the microcracking of concrete, or the freezing of bearings due to corrosion, that indicates deterioration of the materials.8 This "health monitoring" approach includes sensing a change in the amplitude, frequency, or damping of transient vibrations that may be excited by either random traffic, wind loadings, or by forced-loading devices.
Another application could be weigh-in-motion (WIM) measurements of vehicles. Although WIM systems have the potential for use as enforcement tools to detect vehicles that exceed legal weight limitations, actual WIM systems have been used for the less demanding task of obtaining spectra of loads.9 These data can be analyzed to estimate the probability that the maximum load rating of the bridge is being exceeded or to calculate the fatigue cycles experienced by the structure, which can be used to predict remaining service life.10,11
Several state departments of transportation have pooled FHWA research funds to conduct a study of the application of fiber optics to the specific problem of weighing trucks in motion. The New Mexico Department of Transportation is the lead state for this study, which will evaluate a variety of fiber-optic methods that have been proposed for this purpose. The study will be carried out by the Vehicle Detection Clearinghouse located at New Mexico State University.
All these fiber-optic sensor applications involve the measurement of strain. Temperature sensing is also of interest, both for the temperature compensation of strain measurement and for needs such as the prediction of ice formation or proper concrete curing. Another area of interest is chemical sensing. For instance, the detection of chlorides in reinforced concrete helps to predict the corrosion of steel, and the measurement of the variation in pH would be useful for monitoring the curing of concrete.
The phenomena that could be monitored by fiber-optic sensor arrays range over many orders of magnitude, depending on the application and the structure being monitored. Strains range from 5 microstrain for vehicle loadings up to a few percent for soil movements. As shown in figure 1, the time scales cover 16 orders of magnitude in frequency.
The ultimate time scale is the service life of the bridge, which can exceed 100 years. On the scale of years to decades are the deterioration processes, including alkali-aggregate reactions in concrete or corrosion of steel. On the next scale, less than a year, are seasonal and diurnal cycles associated with climatic variables. These are important because the natural frequency of the structure can vary significantly with temperature, masking other effects.12 Conversely, changes in the natural frequency with temperature can be used to diagnose deterioration, such as the sticking of bearings.13 Traffic loadings can also show cycles on these time scales, as well as on a weekly basis, as a result of differences in driving patterns between workdays and weekends.
The time distribution of loadings imposed by individual vehicles depends on vehicle spacing and speeds. For example, on rural roads where there may be as few as one vehicle per hour, this amounts to a frequency of recurrence on the order of 10-4 hertz (Hz) (cycles per second). On more heavily traveled roads with closely spaced traffic traveling at the legal speed limit of 105 kilometers per hour, the frequency could approach 1 Hz.
The structure itself will vibrate at fundamental frequencies typically in the range of 1 to 20 Hz, depending on the span length, stiffness of the span, and on the type of design, such as truss or suspension.14 However, higher order modes with frequencies up to 40 Hz are also generated and may be essential for characterizing structural condition.15 Forces associated with moving vehicles have frequencies in the ranges of 1 to 4 Hz for body/suspension motion and 8 to 15 Hz for axle hop.16 Higher frequencies, on the order of 10 kHz, may be applied for evaluation of pavement dynamics.17 Finally, acoustic emissions associated with the propagation of fatigue cracks fall in the range of 20 kHz to 1 MHz.18
In the design of the data acquisition system, the sampling rate must be matched to the time scale of interest. For example, since modal analysis can involve frequencies up to 40 Hz, the minimum sampling rate should then be 80 Hz at a single sensor.19 For correlation among sensors, phase information is also important, and the sampling rate would need to be higher.
To create a system in which the sensors work effectively together, the spacing of sensors along the fiber and the placement of the fibers around the structure are critical factors. The conventional approach for load rating measurements has been to locate strain gauges so that the spacing between measurement points is between 5 to 15 meters for spans in the 20- to 60-meter range. In contrast, installing a single optic fiber containing 100 sensors over the same range of lengths would provide strain measurements on 30-centimeter to 60-centimeter centers. This much higher spatial resolution would improve the accuracy of the observed deflection curve and increase the number of vibrational modes that could be detected.
This is crucially important because bridge health monitoring operates by sensing natural frequencies of vibration. For an idealized simple beam, the maximum amplitude of the fundamental vibration mode would be at the midpoint, and for the second mode, at the quarter points.20 However, in the complicated three-dimensional geometry of a real bridge, a more advanced analysis would be required involving the use of a finite element model.
Data Acquisition Considerations
The inefficiencies of a bulk-storage approach to raw data acquisition could be avoided through appropriate signal processing. For example, in WIM monitoring, there may be long stretches of dead time in which no vehicles pass by. This could be handled by analog triggering circuits, which turn the system on only when a vehicle is detected. After the data are collected and digitized, digital data compression methods such as those used in audio cassette disk (CD) recording could be used to minimize storage requirements.
For a large structure, the transmission of the digital electronic signals is itself a concern. At a sampling rate of 100 Hz and an analog-to-digital resolution of 16 bits, a 1-kilosensor system would have a digital data rate of 1.3 Mbits per second. Because long runs of cable are prone to damage and susceptible to electromagnetic interference, use of radio telemetry on a local network is receiving serious consideration.21 Another possibility would be to convert the digital electronic signal back into a digital optical signal to permit the use of fiber optics as a communications system. As optical computing advances, it may become possible to develop a completely optical system.22
Landmark Results in New Mexico
The first-phase interagency collaboration between FHWA and the Optical Sciences Division of NRL included tests of prototype instrumentation designed by NRL in concrete beams and deck panels at TFHRC. The concrete beams and deck panels with 35 embedded sensors were then loaded to failure. This demonstrated that the sensors were rugged enough to survive the process of pouring and compacting the concrete. They showed strain sensitivities comparable to, or better than, conventional strain gauges, and the sensors remained intact after the concrete itself cracked.
In the next phase, a 32-sensor system was designed, built, and used in field tests at New Mexico State University (NMSU). An outcome was the development of a set of preliminary guidelines for installation of Bragg grating fiber-optic sensors on existing bridges to investigate practical issues in the full-scale application and regular operation of fiber optics by civil engineers rather than optical scientists.
The New Mexico project, co-funded with the National Science Foundation, resulted in the installation of a system of 67 calibrated, fiber-optic sensors on an existing steel bridge on Interstate 10 in Las Cruces. An additional number of sensors, bringing the total to about 120, will be installed this summer. The bridge was built in the 1970s and was known to suffer from numerous fatigue cracks. In August 1998, the system was completed by calibrating the sensors using a truck of known weight. Even before calibration, preliminary results showed that the instrumented bridge is capable of detecting and counting all standard-class vehicles and measuring their speed. Now, it is also capable of measuring vehicle weight. The continuously operating system transmits data by cell phone to the NMSU Civil Engineering Department where civil engineers such as NMSU's Rola Idriss answer the critical questions: How can we better inspect our bridges? How can we save time and money?23
The FHWA Las Cruces project has already demonstrated that fiber optics can effectively replace conventional strain gauges in field situations. The fiber-optic gauges have stayed fastened in position better than traditional wire gauges while performing the functions of both a data collector and data transmitter. This system will continue to operate for several years to investigate long-term performance issues. To date, the Las Cruces project has achieved notable success in its primary purpose of investigating practical issues in the full-scale application and regular operation of fiber optics by civil engineers rather than optical scientists.
Other FHWA Research Initiatives
In another field application, in collaboration with the Swiss Federal Institute of Technology, Bragg grating sensors are being installed on the Viaduct de Vaux, a box-girder steel-and-concrete bridge under construction in Switzerland. An initial set of 32 sensors developed for FHWA by NRL will be used to monitor the strains in the steel girders during erection. Subsequently, a second set of 32 sensors has been embedded in the concrete deck, using Bragg grating sensors in casings designed by the Swiss for long-term monitoring.
Throughout the Interstate 10 bridge in Las Cruces, N.M., 120 calibrated Bragg grating fiber-optic strain sensors have been installed to provide an information base for an effective cost-benefit analysis of the use of fiber-optic sensors on highway structures, investigating issues such as overall monitoring strategy, systems architecture, installation procedures, temperature compensation, durability, compatibility with other materials, sensor standardization, and data analysis.
Another phase of the FHWA Fiber-Optics Research Program focuses on the development of strain gauges for use in pavements and in soil. It has been difficult to monitor strains in these friable heterogeneous materials using conventional gauges. To date, several gauges have been built using conventional designs of the gauge housings (plate, diaphragm, and ring-uralite), but with Bragg grating sensors replacing the conventional foil sensing elements.24 The gauges have been evaluated under load in test-pavement sections with a heavy-vehicle simulator at the U.S. Army Corps of Engineers Cold Region Research and Engineering Laboratory at Hanover, N.H.
Research on fiber-optic strain gauges for use in pavements and soils will continue to be a priority, focusing on improved housings for the sensors that effectively transfer pressures and strains from the surrounding soil or pavement with minimum disruption. Research is also planned on the sensors themselves and associated photonic components. This includes different grating configurations, such as long-period or chirped gratings and distributed sensing. Other research directions include: (1) additional work on temperature sensing and compensation; (2) chemical sensing, especially pH and chlorides; and (3) development of hotonic components, including light sources with a broader wavelength range and higher bandwidth opto-electronic converters.
Research directed at options for data transmission and management include advanced wireless systems and all optical systems. Also, computer methods for data visualization and algorithms for structural health assessment need to be developed. Finally, research has also been carried out on the problem of the temperature dependence of fiber-optic sensors, and a method has been developed using Brillouin scattering.25
With the emergence of commercial vendors of Bragg grating fiber-optic monitoring systems and with the publication in the near future by FHWA of guidelines for the application of these systems to bridge monitoring, the installation of fiber-optic sensors should be regarded as a part of the normal design and construction process, rather than a specialized research project.
Gauging the Future of Fiber Optics
The FHWA Fiber-Optics Research Program is succeeding in its primary objective, which is helping highway engineers understand fiber-optic sensors, how they work, and - even more importantly - how they work together. The capabilities and possibilities that emerge from placing and calibrating sensors to operate as a highly sophisticated system have enormous potential for the future of both infrastructure construction and infrastructure maintenance nationwide and worldwide. Identifying and repairing cracks and faults when they are small is vastly less expensive and contributes significantly to the safety of the traveling public.
Fiber-optic strain gauges based on Bragg gratings have proven to be effective replacements for conventional strain gauges for highway applications. They create opportunities for monitoring on time and spatial scales that could not be achieved previously in civil engineering. The potential for a very large number of sensors also requires careful consideration of the design of the individual components of the system and in their integration. The data management and interpretation aspects must also be taken into account in the monitoring strategy. FHWA research focuses on resolving these issues so that fiber optics can enter widespread use by state departments of transportation.
In the future, fiber-optic gauges may play a role in California highway safety, where transportation officials have expressed interest in installing them in critical zones of seismic activity. Planning is also currently underway to install a system of 128 sensors on the Woodrow Wilson Bridge in Washington, D.C., a critical part of I-95 and I-495 (the National Capital Beltway), if Congress authorizes funding. In addition to setting a new world's record for number of sensors, this phase of the FHWA program would allow the investigation of installing and operating a very large-scale network under extremely heavy traffic conditions, including a high count of truck traffic.
Dr. Richard Livingston is a senior physical scientist in FHWA's Office of Infrastructure Research and Development at the Turner-Fairbank Highway Research Center in McLean, Va. He received a bachelor's degree in history from Dartmouth College (1968), a bachelor's degree in engineering from Thayer School of Engineering at Dartmouth College, a master's degree in nuclear engineering from Stanford University (1970), and a doctorate in geology from the University of Maryland (1990). His professional interests concern the materials science and nondestructive testing of construction materials. During his career, he has worked in research positions at the Atomic Energy Commission, Environmental Protection Agency, and the National Institute of Standards and Technology. He has also served as a consultant for the conservation of several architectural monuments, including the Statue of Liberty, the Washington National Cathedral, Colonial Williamsburg, Westminster Abbey, the Taj Mahal, and Hagia Sophia.