As the nation's infrastructure continues to age, the cost of maintaining it at an acceptable safety level continues to increase. In the United States, about one of every three bridges is rated structurally deficient and/or functionally obsolete. It will require about 80 billion to eliminate the current backlog of bridge deficiencies and maintain repair levels. Unfortunately, the financial resources allocated for these activities fall extremely short of the demand. Although several existing and emerging NDT techniques are available to gather inspection data, current maintenance planning decisions for deficient bridges are based on data from subjective condition assessments and do not consider the reliability of bridge components and systems. Recently, reliability-based optimum maintenance planning strategies have been developed. They can be used to predict inspection and repair times to achieve minimum life-cycle cost of deteriorating structural systems. In this study, a reliability-based methodology which takes into account loading randomness and history, and randomness in strength and degradation resulting from aggressive environmental factors, is used to predict the time- dependent reliability of aging highway bridges. A methodology for incorporating inspection data into reliability predictions is also presented. Finally, optimal lifetime maintenance strategies are identified, in which optimal inspection/repair times are found based on minimum expected life-cycle cost under prescribed reliability constraints. The influence of discount rate on optimum solutions is evaluated.

A systems-based approach for evaluating the utility of structural health monitoring methods in bridge engineering applications is presented. The vehicle for this evaluation is a systems management model developed within a computationally tractable decision support framework called a Partially Observable Markov Decision Process (POMDP). A major component of this work was the investigation of the complex relationship that exists between 'damage' as inferred from global measurements (e.g., natural frequency and mode shape) and the actual resistance or capacity of a structural system. The concept of a relationship matrix is presented as a means of representing this relationship mathematically. In addition, a simplified example is offered to illustrate how relationship matrices can be assembled via simulation. Finally, the potential implications of this research are discussed.

Until recently, attempts to update Finite Element Models (FEM) of large structures based upon recording structural motions were mostly ad hoc, requiring a large amount of engineering experience and skill. Studies have been undertaken at LLNL to use state-space based signal processing techniques to locate the existence and type of model mismatches common in FEM. Two different methods (Gauss-Newton gradient search and extended Kalman filter) will be explored, and the preliminary results from a simulated 10 degree of freedom (DOF) building model will be discussed.

A novel method for evaluating concrete permeability at an early age is being developed for use in performance-related specifications where the material durability is specified as a performance parameter. In developing the method, the fundamental relationship between ultrasonic pulse velocity (UPV) and permeability of a porous medium is derived. An experimental relationship between UPV and concrete permeability is also established which strongly correlates with its theoretical counterpart. This experimental relationship is developed from UPV and permeability tests conducted on specimens made from a series of concrete grades. The experimental relation utilized the data collected from specimens made from a total of 20 bridge deck normal concrete mixes corresponding to five w/c (water-cement ratio) groups of 0.35, 0.40, 0.45, 0.50 and 0.55. The implementation procedure developed is the application of the 'paste efficiency' principle. In implementing 'paste efficiency' principle during the casting of a bridge deck, standard specimens are prepared in the field and cured in the laboratory. UPV measurements are obtained at an early age both from the deck and the standard specimens. The decrease in UPV from standard specimens indicates paste quality loss (PQL) and is proportional to the increase in permeability of deck concrete.

Effective detection of structural damages has been a challenging issue in the health monitoring of highway bridge structures. Most currently used nondestructive detecting techniques, though, rely heavily either on extensive measurements of local structural behavior or on analysis based on linear dynamics using simple structural models. Their applications, therefore, are often limited by the experimental costs and the complexity of real bridge structures. This research explores the possibility of applying an energy index approach in general nonlinear finite element analysis for damage detection in highway bridge structures. The nonlinear behaviors of the bridges under dynamical loading conditions due to material inelastic deformation and crack damages have been considered. It is shown that by utilizing the energy balance rule for a general, dynamically loaded nonlinear solid body a spatially indexed, scalar energy parameter can be formulated based on the generalized J integral used in fracture mechanics. The evaluation of such an energy index can be implemented into general 3-dimensional nonlinear finite element analysis procedures in computer simulations to detect and locate the structural damages within the highway bridge structures. The effectiveness and feasibility of the proposed energy index approach are illustrated in numerical simulation studies.

This paper presents the results of an ongoing investigation into using broadband vibration data to monitor the structural integrity and health of an all-composite bridge. Bridge 1 - 351 on Business Route 896 in Glasgow, Delaware, was replaced with one of the first state-owned all-composite bridges in the nation in the fall of 1998. The bridge consists of two E- Glass/vinyl ester sandwich core sections (13-ft X 32 ft) joined by a longitudinal joint in the traffic direction. Each sandwich core section consists of a 28-inch deep core and 0.4 - 0.7-inch thick facesheets. Vibration data were obtained from a mesh of 1050 test points covering the upper and lower surfaces of the bridge. From the modal information and the visualization of the data, several aspects of the structural behavior of the bridge were obtained. These characteristics include the interactions between the bridge and abutments; the effectiveness of the longitudinal joint to couple the deck sections; the effectiveness of the core to couple the face sheets; and the structural integrity and dynamic consistency of the entire structure. In addition, mode shapes and natural frequencies were determined and are correlated with theoretical calculations and vibration analyses conducted for this bridge. A novel algorithm using the vibration data is being developed that enables local perturbations sensitive to the state of the material (e.g. manufacturing defects, material degradation or service damage) to be detected and spatially located in the bridge. This technique has been successfully validated for locating damage in 1-D beam structures and is being extended to the 3-D sandwich structure. Applications for quality assurance/quality control and health monitoring of large composite bridge structures using this technique will be discussed.

The Delaware River Port Authority, whose mission is to manage, plan and construct transportation facilities and provide transportation services to maximize the safe and efficient movement of people and freight within the Delaware River Valley, located in southwestern Pennsylvania and southern New Jersey, is a self-financing, bi-state Authority, formed by a compact between the Commonwealth of Pennsylvania and the State of New Jersey and approved by the Congress of the United States. The Delaware River Port Authority is firmly committed to the strategic and integrated use of advanced transportation technology to improve traffic flow, operational efficiency and safety on DRPA's four bridges. To this end, the Delaware River Port Authority has initiated a program, appropriately named 'Smart Bridges.' The Delaware River Port Authority has recognized that this type of program is essential to the advancement of the DRPA's mission as an efficient, customer- friendly transportation and regional development agency. Under the Smart Bridges program the Delaware River Port Authority is introducing new technology into its aging infrastructure and transportation systems to ensure that the facilities continue to serve the region into the 21st century and beyond. Initiatives introduced under this program include EZ Pass, video surveillance systems, computerized traffic control systems and partnering with local universities to investigate the application of various innovative technologies to assist in the maintenance of the bridge facilities.

The Commodore Barry Bridge is a major long-span bridge across the Delaware River connecting the cities of Bridgeport, New Jersey and Chester, Pennsylvania. A Structural Identification (St-Id) study of Commodore Barry Bridge is presented. The objective of this structural identification approach is to characterize the as-is structural condition and the loading environment of the bridge through experimental information and analytical modeling. The attributes that make long-span bridges different for utilization of experimental and analytical applications as compared to short-span bridges are presented. Some of the experimental, analytical and information tools which are utilized for this research are discussed. The details of constructing a preliminary analytical model after conceptualization of the structural characteristics are presented. Development of the 3D analytical model, and the model characteristics such as elements used, boundary and continuity representations are summarized. The experimental techniques that are necessary for the structural identification of a long span bridge are defined and application examples are provided from the Commodore Barry Bridge. Experiences gained during the applications of different forms of dynamic tests, instrumented monitoring and controlled static and crawl speed load tests are presented with example experimental data. Correlation of experimental results and analytical simulations are presented. Immediate and possible future uses of information generated are summarized.

Information technology issues for the continuous health monitoring of the Commodore Barry Bridge will be presented in two parts in this paper. The first part describes data acquisition design and the second part discusses issues related to a proposed database. Currently, the health monitor consists of more than one hundred channels of information. These channels are made up of slow speed strain gages (measuring intrinsic strains due to environmental effects, temperature changes and wind loads), high-speed strain gages (measuring strains and accelerations related to traffic effects) and one camera (recording images of the traffic pattern at the bridge). These gages are hard-wired to a central data acquisition station in which three slow speed data acquisition systems, one high-speed data acquisition system and a data acquisition computer are located. All data acquisition systems are integrated under the LabVIEW platform. It was necessary to utilize various appropriate sampling frequencies for each system due to the differing nature of the phenomena being measured by each system. The data is post- processed subsequent to acquisition and finally the data is stored and archived. Post-processing algorithms are implemented to eliminate any noise component from the data, complete any necessary signal re-sampling, and to synchronize the collection times of the different data collection systems. Once the data is transferred and archived data analysis can begin. The creation of a database for storage of all pertinent information is envisioned as a future add-on to this project. Two possible versions of the database system are currently being investigated. In the end, the database would allow users to retrieve information according to customized query criteria. It would also allow users to navigate raw data, preview data graphs, extract specified information (such as analytical data, reports, images, video clips and CBB CAD models), save the retrieved data to a local computer hard-disk and modify information for authorized users. One of the system designs being investigated would not be Internet based and thus would provide local users efficient means of access. In this design, the database would perform complex computational work without coming into contact with Internet traffic. The second proposed system design would provide flexible and robust access through the Internet. This system would allow the retrieval of information from the database through a conventional Browser. This database design would provide fast service and the rich display capabilities of highly dynamic web pages and would allow multimedia capabilities.

The Drexel Intelligent Infrastructure and Transportation Safety Institute, working in partnership with the Delaware River Port Authority (DRPA), has been investigating the application of various health monitoring techniques to long span bridges. Specifically, the researchers efforts have focused on the Commodore Barry Bridge, a major cantilevered through truss bridge owned by the DRPA. Health monitoring, in the case of civil infrastructure systems, may be considered as measuring and tracking the operating and loading environment of a structure and corresponding structural responses in order to detect and evaluate operational anomalies and deterioration or damage that may impact service or safety reliability. This paper describes the development and implementation of the health monitoring system that has been installed on the Commodore Barry Bridge. Issues that should be considered when conducting health monitoring of long span bridges are also discussed. The health monitoring system takes advantage of in excess of 100 data channels to continuously track the loading environment and numerous structural responses of the bridge. The major attributes of the health monitoring system for this bridge, many of which are applicable for use on other long span bridges, are also presented and described.

In June of 1999, Applied Geomechanics was retained by Hayward Baker to monitor the Laurel Street Bridge in Santa Cruz during compaction grouting of the ground beneath the bridge footings. This work was performed as part of an extensive program of seismic upgrades to many of California's bridges after the 1989 Loma Prieta earthquake. The project specifications for the compaction grouting put stringent requirements on the allowable amount of bridge movement during the grouting process. Hayward Baker recognized that tiltmeters were one of the few instruments that could measure movements small enough to satisfy the specifications. The tiltmeters used for this application are capable of detecting 0.001 inch (0.0254 mm) deflections over a 100 foot (30.48 m) span. Continuous monitoring of the tiltmeters was implemented to provide instant notification of vertical deflections greater than 0.1 inch (0.254 mm). However, the threshold alarming was complicated by the fact that the normal diurnal movement of the bridge due to thermal expansion and contraction is of the same order of magnitude. Therefore, the normal daily movement of the bridge was modeled with a sine wave, and the alarm thresholds were based on the difference between the model and the recorded data. This model is relatively easy to program within a datalogger and results in alarms that respond to grout-induced movement rather than thermoelastic bridge deformation.

Periodic inspection and testing of a structure is necessary to ensure its structural integrity and reliability. Visual inspection alone is not adequate to determine structural integrity. Internal damage and the damage in inaccessible areas are hard to detect in such investigations. Structural identification is a technique that may help to overcome these shortcomings. Developing the right analytical model of the structure plays a key role in ensuring that the model can be used in structural identification. Finite element modeling is the most commonly used tool in structural modeling. Finite element models usually have a high sensitivity to different structural parameters like stiffness, damping etc. Arriving at the correct values for these parameters is an important factor due to the model's high sensitivity to these values. Different kinds of iterative optimization algorithms have been developed to arrive at these values. This paper discusses one such algorithm and shows the application of this algorithm to the finite element model of a six span bridge.

The purpose of this study was to investigate the application of modal analysis to ascertain changes in the boundary conditions (or structural damage) of a complicated bridge structure. Reconstruction of the Interstate 15 corridor through Salt Lake City, Utah had provided an opportunity for destructive testing to be conducted on a three-span, continuous curved steel girder bridge. Forced Vibration testing using an eccentric mass shaker was conducted on the bridge in three phases. Each phase represented a change in boundary conditions. The initial testing was done with the bridge in the as-built condition with the continuous deck at the abutments and frozen bronze bearings. The second phase of testing occurred after the bridge deck was cut way from the approach slabs. For the third phase of testing, the bearings at the ends of the girders were replaced with teflon pads and the bearings over the two intermediate piers were jacked up and greased. The results of the study show that modal analysis is capable of determining changes in a structure's natural frequencies and mode shapes due to a change in the boundary conditions. By extrapolation this would indicate that modal analysis would work as an effective non-destructive evaluation tool.

Knowledge of the integrity of in-service structures on a continuous time basis is an ultimate objective for owners and maintenance authorities. The development of a life extension and/or replacement strategy for highway structures is a crucial point in an effective bridge management system. A key component of such a bridge management system is a means of surveillance techniques and determining the condition of an existing structure within the normative and budgetary constraints. Recent advances in sensing technologies and material/structure damage characterization combined with current developments in computations and communications have resulted in a significant interest in developing diagnostic technologies for monitoring the integrity of and for the detection of damages of structures. To identify anomalies and deterioration processes, it is essential to understand the relationships between the signal measurements and the real occurred phenomena. Therefore, the comparison of measured and calculated data in order to tune and validate the mechanical and numerical model assumptions is an integral part of any system analysis. Finally, the interpreted results of all measurements should be the basis for the condition assessment and the safety evaluation of a structure to facilitate replacement and repair decisions.

The monitoring of geotechnical structures like piles, anchors and tunnels requires the measurement of deformations over bases of a few meters to a few tens of meters. The SOFO monitoring system, based on the use of long-gage low-coherence interferometric sensors therefore presents interesting application opportunities in this domain. The SOFO system was installed in a number of piles to monitor their short an long term deformations, to evaluate the lateral friction and to assess their ultimate bearing capacity. The sensors were also installed inside anchor cables to measure the deformations of the rock, in the free and in the anchored parts. Additional sensors were installed directly on single cable strands. This paper presents the sensor installation and the results from selected applications in the monitoring of piles and anchors.

SOFO ius a structural monitoring system using fiber optic deformation sensors. It is able to measure deformations between two points in a structure, which can be from 20 cm up to 10 meters (or more) apart with a resolution of two microns (2/1000 mm) even over years of measurements. The system is composed of optical deformation sensors adapted to direct concrete embedding or surface mounting, the cable network, the reading unit and the data acquisition and analysis software. The system is particularly adapted to precise short and long- term deformation monitoring of large structures. An array of more than 60 sensors have been installed on the pier of San Giorgio Levante in the Genoa harbor. These sensors allow the measurement of the pier displacements during the dredging works, ship docking and in the long term. The sensors measure the curvature changes in the horizontal and vertical planes and allow a localization of settlements with a spatial resolution of 10 m over a total length of 400 m. The sensors can be measured automatically and remotely. This paper presents the sensor installation and the results from the first monitoring period.

In Germany a group of researchers from four universities has developed an advanced experiment supported state assessment method for short and midspan concrete bridges under the term 'EXTRA.' As a background for this work are the deficiencies of many bridges on one hand and the rising disproportions between available funds and costs for demolition and rebuilding on the other. The presented method can be used preferably in such cases, where conventional assessment methods cannot be applied successfully. The fundamental idea of this method is to determine the ultimate testing load UTL of the structure. This UTL is defined as that load, which just not reaches a level which would cause damages in the tested structure. The level of UTL can be considerably higher than the safe load. For identification the UTL special criterions has been developed. From the UTL afterwards the safe load can be derived by safety considerations. A guideline for practical use of the method for building structures will be available shortly; another one for bridges will be prepared in the near future. Methodical and technical conditions and tasks for preparation and practical realization of the test will be introduced. Main subjects are safety aspects, needed theoretical and practical pre-investigations, the test program, different criterions for UTL-identification, the loading regime and corresponding on- line information on the building response In connection with the presentation of a selected application example experiences on the co-operation with owners, authorities and uninvolved colleagues will be discussed and advises on further developments will be given. Acknowledgements finalize the contribution.

State assessment results of conventional or advanced experimental investigations of existing concrete structures (e.g. using the 'EXTRA-'method) are absolutely valid at the time of testing, but usually not longer than a time of at most five to ten years after the testing date. Thereafter the assessment results will be more and more unsure because of progressing (often unknown) deterioration processes. At the latest at that time a repeated (but expensive) test procedure would be necessary. In the most of these cases (but of course also in many other situations) the concept of a supervised lifetime extension by monitoring of a small number of key- measurands would be the more safer and reasonable solution. Since two years the authors have a close cooperation with the aim to develop a remote monitoring system for fields of application like mentioned before. Based on experiences collected on different applications of such a system more general demands on an efficient system configuration could be derived. An overview on the main characteristics is systematized by matrix elements defined by the main demands flexibility, robustness and reliability as well as the main system components hardware, software and application management. Afterwards it will be shown exemplary (system structure, measurement modes, sensor selection criteria, data transmission, data management as well as optional features like solar energy supply or overload truck identification) how the remote monitoring system SMS 2001^R is able to fulfill the demands discussed before and which further developments are under way respectively. Finally some experiences from the first presentation of a SMS 2001^R in the United States will be added.

Nowadays, in order to guarantee the security in passenger and goods railway transport, fixed systems located in rails are used to measure axles, wheels and brake discs temperatures during train circulation so that abnormally high temperatures as a result of a malfunction can be detected. Measurement systems in this kind of application may be affected by different uncertainty sources, characteristic of infrared temperature measurements, which limit the accuracy of the estimated measurement. Uncertainty sources are specially important in these applications due to: (1) Extremely variable emissivity as a result of stain or different paints used on the surfaces. (2) Difficult evaluation of the environment's radiation as measurements are made outdoors. (3) Alarm temperatures are only about 40 degrees Celsius to 80 degrees Celsius above the environment temperature. The paper analyses the effects of these uncertainties. The results show that, in order to get the minimum uncertainty peaks in the estimated temperature, the proper duty waveband is 3 - 5 micrometer. They also show that, with a proper choice of the wavelength, the uncertainty due to solar radiation remains masked by the uncertainty due to the lens emissivity.

Wheels, hubs and brake discs in a train during its circulation are under mechanical strains that make its temperature increase above the environment temperature. Mechanical defects in those elements produce an excessive friction and, as a consequence of it, an important increment of its temperature in relation to normal values. Detecting these anomalies is essential to avoid accidents and it is performed by fixed systems located next to rails which make infrared temperature measurements of hot points and send them to a supervisory station that takes the proper steps. The paper introduces the most important problems which must be dealt with during the designing stage of the measurement system. It also explains the solutions taken by the authors in order to assure the minimum operative aims demanded by the application. These problems includes: the choice of the detector and measurement method, communication with the supervisory station, and the environment conditions. Finally, the research lines followed by the authors in order to improve and extend the system's capabilities are explained.

Hanshin Expressway Public Corporation (HEPC) is carrying out the daily inspection to detect damages in early stage on the service routes, approximately 220 km in length, in 1999. Moreover, HEPC has been carried out the regular inspection for all highway structures every 5 - 7 years to make an occurrence mechanism and a cause of damages clear. Then HEPC has judged the necessity of measures for repairing and strengthening based on these inspection results, and has made sure of the third person's safety. Thus, HEPC has maintained the highway structures healthy. However, there exists some structures in which it is impossible to make the causes of damage clear and to grasp the development of damages. So considering these structures as the monitoring structures, the field investigation (monitoring) has been carried out to make the causes of damage clear and grasping the development of damage basically. HEPC has made all of the field investigation results into database as structure's medical records of Kartell, and has made good use of the control and the management for Hanshin Expressway. This paper presents the results of monitoring for highway structures damaged by Alkali Aggregate Reaction (mainly Alkali Silica Reaction, hereafter called ASR).

With the goal of diagnosing the soundness of the world longest Akashi Kaikyo Bridge with a center span length of 1,991 meters, advanced monitoring system was installed; these include seismometers, accelerometers, anemometers and so on. First, the overall monitoring system of the bridge is described. One characteristic of suspension bridges is that temperature constitutes the major factor affecting their configurations under normal load conditions. Temperatures and the positions at 3 locations measured by GPS (global positioning system) has been recorded since September 1, 1998. Through continuous measurement of bridge positions and temperatures, along with the use of statistical methods, possibility to evaluate the configuration of a suspension bridge as a whole from the temperature at representative points is shown. Through this method using GPS, possibility to identify the location when deformation is caused by an earthquake or typhoon or when the measured values are off by normal observation is indicated.

Monitoring of a 720 m span suspension bridge which is located at windy and seismically active area in northern part of Japan, opened in 1998 is introduced. The purpose of the monitoring is to understand the actual dynamic behavior and loading conditions, and to develop a health monitoring scheme using ambient vibration measurement which is readily available under service conditions. Data processing scheme based on Ibrahim's time domain method is constructed for identification of natural frequencies, mode shapes and modal damping ratios. Because measurement of the input force is not available for ambient vibration measurement, statistical treatment of measured data is required to obtain reliable results. This scheme is applied to analyze the actual ambient vibration measurement of the bridge under various environmental conditions. In the analysis, amplitude dependent softening of bearing supports which is not expected at the design calculation is identified. This preliminary analysis shows that the current monitoring system is directly applicable for identification of damage at the bearing supports which is one of the hot spots for fatigue and seismic damage, and also has the potential applicability to the health monitoring of the entire bridge.

In Japan, some monitoring systems have been introduced to evaluate conditions of the railway structures. In this work, we introduce the monitoring systems of railway bridges. Here, for example, we show the abstraction of fatigue damage monitoring system. The crack propagation properties of a side notched SUS304 thin sheet, which is affixed to a fatigue test specimen, is studied to develop a fatigue damage monitoring sensor. When cyclic loads are applied to the fatigue specimen, the side notched thin sheet behaves as a displacement controlled condition. The stress intensity factor (Delta) K is expressed by the following equation; (Delta) K equals (Delta) (sigma) (root)GL/2 where GL is the gauge length of the thin sheet. The thin sheet is fatigue pre-cracked, stress- relieved and affixed to a specimen. The results of fatigue crack propagation tests show that the crack propagation rate is dependent on stress range and gauge length, but not on crack length. Under sufficient mean stress conditions, the fatigue crack propagation rate is well expressed by a power low of (Delta) K even at such a low stress range as 20 MPa. Fatigue damage accumulated during a monitoring period can be estimated from crack propagation during the period. The sensitivity of the sensor is controlled through the gauge length of the sensor.

An efficient monitoring system is essential to avoid catastrophic structural failure and reduce retrofit cost on structures at aging time. The objective of this paper is to develop a diagnostic system to monitor structural integrity of bridge by stain sensing using fiber optic sensor (FOS). A number of laboratory tests were carried out in the process of this research for steel and concrete members with attached and embedded type fiber optic sensors. Strain data obtained from cracked area as an estimation of crack propagation and survivability performance of fiber optic sensor due to cyclic loading experience are concerned. Electrical strain gages co- located with fiber optic sensor to verify functioning of fiber optic sensor and finally compared results obtained from both sensors with numerical calculations. Our research results revealed that, fiber optic sensor and strain gage respond same way at transition zone and significant variation in strain response observed when crack propagates from notch tips. In this series of investigation by bringing new experimental system about fatigue crack evaluation and wireless health monitoring we could develop smart system to apply to field assessment.

Structural health monitoring works, which measure key structural parameters systematically, provide valuable information in current evaluation of structural integrity, durability and reliability. The application of new structural concept in design and construction of three cable-supported bridges in Tsing Ma Control Area (TMCA) of Hong Kong and complexity and size of cable-supported bridges have called for implementation of such monitoring works to ensure cost optimal maintenance planning and safe bridge operation. A bridge monitoring system has therefore been deployed and implemented in monitoring and evaluating the structural health of these bridges, i.e. Tsing Ma (suspension) Bridge, Kap Shui Mun (cable-stayed) Bridge and Ting Kau (cable-stayed) Bridge. This system is named as Wind And Structural Health Monitoring System (WASHMS) and is composed of five sub-systems, namely, Sensory Systems, Data Acquisition Systems, Data Processing and Analysis Systems, Computer Systems, and Cabling Network Systems. This paper outlines: (1) objectives and scope of monitoring, (2) identification of key structural monitoring parameters, (3) types of sensory systems required, (4) global layout design of WASHMS, (5) development of WASHMS, and (6) implementation of WASHMS.

Extensive structural health instrumentation systems have been installed on three long-span cable-supported bridges in Hong Kong. The quantities measured include environment and applied loads (such as wind, temperature, seismic and traffic loads) and the bridge response to these loadings (accelerations, displacements, and strains). Measurements from over 1000 individual sensors are transmitted to central computing facilities via local data acquisition stations and a fault- tolerant fiber-optic network, and are acquired and processed continuously. The data from the systems is used to provide information on structural load and response characteristics, comparison with design, optimization of inspection, and assurance of continued bridge health. Automated data processing and analysis provides information on important structural and operational parameters. Abnormal events are noted and logged automatically. Information of interest is automatically archived for post-processing. Novel aspects of the instrumentation system include a fluid-based high-accuracy long-span Level Sensing System to measure bridge deck profile and tower settlement. This paper provides an outline of the design and implementation of the instrumentation system. A description of the design and implementation of the data acquisition and processing procedures is also given. Examples of the use of similar systems in monitoring other large structures are discussed.

A structural health monitoring system has been installed in the cable-supported bridges located in the West of Hong Kong, i.e. the Tsing Ma Control Area. These cable-supported bridges are the Tsing Ma (Suspension) Bridge, the Kap Shu Mun (Cable- Stayed) Bridge and the Ting Kau (Cable-Stayed Bridge) Bridge. The monitoring system of Tsing Ma Bridge and Kap Shui Mun Bridge has been operated since May 1997, whereas the monitoring system of Ting Kau Bridge has been operated since November 1998. In past years, data received from the monitoring systems have been processed, and analyzed and archived. This paper first briefly outlines the operation of the data processing and analysis, and then presents: (1) the load effects monitoring results such as wind, temperature and traffic (highway and railway), and (2) the bridge responses monitoring results such as displacements, stresses/strains, accelerations and cables forces. Comparisons between monitoring results and design parameters and assumptions for the cable-supported bridges are also presented.

This study addresses model-based structural damage simulation and identification of the Tsing Ma Suspension Bridge through modal sensitivity analysis. For this purpose, a precise three- dimensional finite element model has been developed with the attributes: (1) the spatial configuration of the original structure remains in the model; (2) the geometric stiffness of cables and hangers has been accurately accounted for in the model; (3) the mass and stiffness contribution of individual structural members is independently described in the model, so damage to any structural member can be directly and precisely simulated. The model was validated using the measured modal data obtained at different erection stages and after the bridge completion. It is then used as a baseline for structural damage simulation and modal sensitivity analysis. Due to the intensive distribution of natural frequencies of the bridge, modal assurance criterion (MAC) is first utilized to check the correlation of mode pairs between the damaged and intact structure. Ten damage cases are simulated and the sensitivities of various modal parameters including natural frequency, mode shape and modal flexibility to different types of damage are evaluated. The goal of this study is to analytically determine which modal parameter is most sensitive to damage for a large-scale suspension bridge. The analysis results show that, in most cases, the frequency sensitivity to damage is low, while the modal flexibility method clearly indicates the damage locations only using a few lowest frequency modes.

This paper addresses the identification of damage region and location in the Tsing Ma Suspension Bridge deck using modal data. A two-stage identification method is proposed and implemented through numerical simulation for damage detection of the bridge deck. In the first stage, the main span deck of 1377 m length is divided into seventy-six segments and the target in this stage is to determine the deck segment that contains damaged member(s). An index vector derived from mode shape curvatures in both intact and damaged states is presented to identify the damage region (segment). In the second stage, the specific damaged member(s) within the damage region is identified by means of a neural network technique. The combined modal parameters in terms of natural frequencies and a few incomplete modal vectors are adopted as input vector to the neural networks. Two back-propagation networks are trained for the damage location detection. The simulation results show that despite very low modal sensitivity of the bridge to deck member damage, the developed method can still locate the damage at longitudinal structural members such as bottom chords, top chords, and diagonal members.

Namhae Bridge, completed in May 1973, connects Namhae Island with mainland Korea in the province of Kyongsangnamdo. It is a three-span suspension bridge with a main span length of 404 m, and side spans of 128 m each. A geometric survey and static loading test was carried out and the deck geometry obtained from the survey was compared with the intended geometry of the original design. On the other hand, a long-term monitoring system for Namhae Bridge was employed in December 1996. The objective is to monitor the structural responses of the bridge with a view to identifying the deterioration rate over a long- term period. The monitoring system consists of 110 channels of both static and dynamic sensors such as strain gages accelerometers, tiltmeters, jointmeters and anemometers. A vehicle loading test was carried out in order to obtain the initial values of Namhae Bridge. An ambient vibration test for the whole bridge was carried out in 1999. Mode shapes as well as natural frequencies were found to give a more precise description of the current stage. These results will be used as a reference for the future monitoring.

This paper presents an iterative constrained optimization scheme for the finite element (FE) model updating of long-span bridges. The objective is to minimize the differences between the calculated and the measured frequencies by changing some selected structural parameters in the FE model. An eigenvalue sensitivity matrix is first obtained from the first-order Taylor series expansion of the eigenvalues with respect to these selected parameters. A set of linear equations relating the perturbation of parameters to the differences between the calculated and the measured frequencies is then established. The selected parameters are assumed to be bounded within some prescribed regions according to the degrees of uncertainty and variation existing in the parameters based on some engineering judgement. The changes of these parameters are found in an iterative fashion by solving a quadratic programming problem. This model updating scheme is applied to both a 1/150 scaled suspension bridge model in the laboratory and an actual cable- stayed bridge in the field. The results show that the natural frequencies calculated from the FE models after this updating process can be quite close to those of the measured values.

Based on the real-time monitoring data, a fatigue damage model using continuous damage mechanics and a methodology and strategy for evaluating fatigue damage and possible critical locations of local fatigue on bridge-deck sections are developed. The proposed local damage model to evaluate the fatigue damage of bridge-deck sections allows fatigue damage to be considered at a continuum scale as a deteriorating process with its physical mechanism such as fatigue crack initiation and growth. Therefore the fatigue damage variable associates not only cycles of stress range on the accumulating process of fatigue, but also directly associates the state of deterioration and the mechanics behavior. The effective stress range for the variable-amplitude stress spectrum due to traffic load is used to evaluate the fatigue strength of the bridge-deck section at different locations, by which possible location of critical fatigue failure can be primarily determined. As a typical case to use the method and strategy proposed in this paper, fatigue damage assessment and the detection of possible critical location of fatigue in a deck section of the Tsing Ma Bridge are carried out by using strain-time history acquired by the bridge structural health monitoring system.

An effective approach for fatigue analysis of bridge-deck sections under traffic loading is proposed. It allows fatigue damage and predicted life of bridge deck under variable amplitude stress history due to traffic to be evaluated on the basis of health monitoring data. Based on the measurement of bridge traffic and the analysis on the property of strain-time histories recorded by strain gauges set at a bridge-deck section, the variable-amplitude strain-time history can be modeled as block repeated cycles. A representative block used as a standard of the blocks of repeated strain history is defined and obtained from the statistical analysis of many samples of daily strain-time data measured by the health monitoring system. Fatigue damage rate generated by a representative block of cycles is then derived, and computational approach for fatigue analysis based on the damage model and Miner's law is developed, in which the updating of the representative block is included for considering the increase of traffic volume in the future. Finally, as an example to show how the approach proposed in this paper is applied to revaluate the fatigue damage of a long-span suspension bridge, some results calculated for the critical location in the deck section are given.

In this paper, we study the feasibility of vibration-based damage identification methods for the instrumented Tsing Ma Suspension Bridge with a main span of 1377 m. Emphasis is placed on how to deal with the noise-corrupted/uncertain measurement data and how to use the series data from the on- line monitoring system for damage detection. Numerical simulation studies of using the noisy series measurement modal data for damage occurrence detection with the auto-associative neural network and for damage localization with the probabilistic neural network are presented. Five neural network based novelty detectors using only natural frequencies of the intact and damaged structure are first developed for the detection of damage occurrence in the Tsing Ma Bridge. The noisy/uncertain measurement data are produced by polluting the analytical natural frequencies with random noise. Numerical simulations of a series of damage scenarios show that when the maximum frequency change caused by damage exceeds a certain threshold, the occurrence of damage can be unambiguously flagged with the novelty detectors. A probabilistic neural network using noisy modal data (natural frequencies and incomplete modal vectors) is then constructed for the localization of damage occurring at the Tsing Ma Bridge deck. The main-span deck is divided into 16 segments and the damage in each segment is defined as a pattern class. The analytical modal data for each pattern class are artificially corrupted with random noise and then used as training samples to establish a three-layer probabilistic neural network for damage localization. A preliminary investigation shows that the damage to deck members can be localized only when the level of the corrupted noise is low.

A procedure based on the use of artificial neural networks for the identification of dynamic system is developed and applied to the bridge structure under earthquake excitation. This neural network-based approach is also applied for the detection of changes in the characteristics of structure- unknown system. Based on the vibration measurement from a linear/healthy system to train the neural network for identification purposes, then the trained network is fed comparable vibration measurements from the same structure under different episodes of response in order to monitor the nonlinearity of the system. The learning ability of the network is examined for the use of multiple inputs. The effects of the network parameters on learning and accuracy of predictions are discussed. Based on this study it is found that the configuration of neural network model is the same as NARMA model and has the potential for structural damage detection.

Recent research has been conducted at Utah State University regarding the ability to assess the condition of a reinforced concrete bridge bent. Three full-scale, in-situ, reinforced concrete bridge bents were tested and modeled through varying states of damage. Each bent was initially tested in an undamaged state using a horizontal sine-sweep test. The forced-vibration testing was achieved using an eccentric mass shaker. The structures were tested in the frequencies ranging from 1.0 Hz to 20.0 Hz in increments of 0.05 Hz. A known amount of damage was inflicted upon each bent for two separate states. The sine-sweep test was re-administered for each damage state. The changes in dynamic characteristics, such as frequencies and mode shapes, were noted from state to state. Detailed finite element models were constructed to match the changes in dynamic characteristics of each bent for each state. This was achieved by matching the severity and location of the structural damage of the model to the field structures. Decreases in structural stiffness were detected with modal analysis. The models were consistent in matching the changes of the field structures. A method was devised for locating possible regions of structural damage.

Forced Vibration testing of four different damage states of a full scale, six span, reinforced concrete bridge, were conducted by Utah State University. These tests were performed to characterize the bridge based on its dynamic characteristics and to determine any correlation that may exist between the dynamic properties of a structure and the location of the inflicted damage. The test structure was a mature bridge being prepared for demolition as part of the Interstate 15 reconstruction through Salt Lake City. An eccentric mass shaker was utilized to excite the structure at several different natural frequencies for which data was collected through an array of seismometers. This data was processed to determine the natural frequencies and mode shapes of the structure. Investigation showed a decrease in the natural frequencies of the structure as well as noticeable changes in the mode shapes of the structure as a result of localized damage to the structure.

One challenge in performing modal testing of structures at low frequencies is obtaining deflection sensors that provide reasonable output in this frequency range. Deflection sensors that require a fixed datum are generally not practical on full-scale structures. Sensors that use an inertial reference such as accelerometers and geophones are necessary. Velocity transducers are typically used at frequencies above their resonant frequency where their amplitude and phase response is flat. They can be used at frequencies near and below their resonant frequency. Thought, in this region, their response is frequency dependent. The advantage of using velocity transducers at low frequencies is that velocity is equal to displacement times frequency, the output of a velocity transducers falls off more gradually with decreasing frequency than the output of an accelerometer. Using velocity transducers at low frequencies requires characterization of the low frequency response of the transducer. A velocity transducer's response can be modeled accurately as a single- degree-of-freedom, damped resonator. Laboratory methods of calibrating velocity transducers to determine the coefficients that describe their behavior are presented. Methods of correcting the amplitude and phase measurements to account for the transducers response are demonstrated. These procedures are illustrated on modal testing data from highway bridges.

Modal testing requires imparting energy to a structure over a range of frequencies. This energy can be applied to the structure one frequency at a time using a monochromatic source, or a broadband energy source can apply energy at many frequencies simultaneously. An example of a monochromatic source is a rotating eccentric-mass shaker, and an example of a broadband source is an impact. A 250-kg, instrumented pendulum was developed at Utah State University to apply impulsive forces to bridge structures and measure the applied forcing function. This paper presents the basic design of this device. Data measured on three highway bridge bent structures are presented and the procedures used to analyze this data are presented. The results of testing with an impact source are compared to the results obtained using an 89-kN rotating eccentric-mass shaker. This paper shows that the rotating eccentric-mass shaker provides very high-quality data over a limited frequency range, while the 250-kg impulsive source provides somewhat poorer quality data, but over a much wider frequency band. The impulsive source also requires less testing time, and simpler data analysis than the rotating eccentric-mass shaker.

Vibrations from construction activities can affect infrastructure projects in several ways. Within the general vicinity of a construction site, vibrations can result in damage to existing structures, disturbance to people, damage to sensitive machinery, and degraded performance of precision instrumentation or motion sensitive equipment. Current practice for monitoring vibrations in the vicinity of construction sites commonly consists of measuring free field or structural motions using velocity transducers connected to a portable data acquisition unit via cables. This paper describes an innovative way to collect, process, transmit, and analyze vibration measurements obtained at construction sites. The system described measures vibration at the sensor location, performs necessary signal conditioning and digitization, and sends data to a Web server using wireless data transmission and Internet protocols. A Servlet program running on the Web server accepts the transmitted data and incorporates it into a project database. Two-way interaction between the Web-client and the Web server is accomplished through the use of a Servlet program and a Java Applet running inside a browser located on the Web client's computer. Advantages of this system over conventional vibration data logging systems include continuous unattended monitoring, reduced costs associated with field data collection, instant access to data files and graphs by project team members, and the ability to remotely modify data sampling schemes.

The feasibility of detecting defects in concrete beams using Lamb waves is investigated in this paper. The Lamb wave can propagate a long distance along the specimen as the guided wave and is sensitive to defects that are smaller than its wavelength. The traditional ultrasonic methods for inspecting defects in concrete use reflection, transmission and scattering of longitudinal waves by internal defects. In traditional techniques signal amplitude and time of flight measurements provide information about the internal defects in concrete. These methods are time consuming and often fail to detect a variety of defects, such as internal corrosion, honeycombs, closed cracks and small inclusions. In this paper Lamb waves are used to detect those defects in concrete beams with and without reinforcement. The Lamb wave technique is found to be reliable for detecting such defects. The effect of separation or delamination between concrete and reinforcing steel bars on the Lamb wave propagation characteristics is also investigated. Corrosion of rebars can cause this delamination. It is found that the cylindrical guided waves propagating along the steel rebars are very sensitive to the degree of delamination between the concrete and the rebars. This investigation shows that the Lamb wave inspection technique is an efficient and effective tool for health monitoring of concrete structures.

We report on the development of a measuring system for dynamic long-term monitoring of highway pavement; i.e. the relative vertical displacements within the pavement layers. Beside the knowledge of weight and frequency of vehicles it is more and more of interest to know the vertical deformations within the pavement layers due to the traffic loads. Our displacement measurement is based on the magnetostrictive principle. Several positions along one measuring axis corresponding to the different layers of the pavement are monitored. The development of the measuring system, the calibration, the embeddment in a test pavement and first test results will be presented.

Among the methods to tackle corrosion of steel reinforcement in highway transportation infrastructure, using corrosion inhibitors has been identified as the most easily and economically applied technique. This study used Electrochemical Impedance Spectroscopy (EIS) to evaluate four corrosion inhibitors in simulated pore solution (SPS) and saturated calcium hydroxide solution (CHS). Three promising inhibitors were identified. It was also found that the electrochemical laboratory test was practical to evaluate corrosion inhibitors quickly and effectively in simulated concrete solutions. A simulated field concrete repair method was devised in order to verify the developed electrochemical laboratory test result. Sixty-three concrete short beam specimens were used. The embedded steel rebars were exposed to chloride environment and electrochemically monitored in accordance with the ASTM G109 procedure. After active corrosion of the upper rebars was detected, the chloride- contaminated concrete was removed. The three aforementioned promising inhibitors were applied to corroded rebars, and new concrete was cast. These rebars were electrochemically monitored to evaluate the effectiveness of corrosion inhibitors for corrosion control. It was found that there was a good correlation between these two test results, and the most effective inhibitor was finally identified.

Fiber grating strain sensors offer a means to monitor the health of highways and bridges as well as a means to monitor vehicular traffic patterns and critical data such as speed, weight, and classification of vehicle types. This paper overviews recent results associated with employing very high speed demodulation systems with capabilities in excess of 10 kHz and strain sensitivities on the order of one microstrain. It is clearly shown that this type of system can be used as an effective traffic monitoring tool as well as for health monitoring purposes.

Ground Penetrating Radar (GPR) can be an effective technique for assessing internal damage levels in concrete roadways. Damage to concrete roadways, particularly those on bridges, can have large economic consequences. Damage often takes the form of corrosion of reinforcing bars, the promotion of internal cracking, eventually large-scale spalling, and the formation of deep potholes. This damage usually initiates internally and does not appear on the surface until it is at an advanced state. The use of asphalt overlays further exacerbates this problem. One of the most important, yet difficult to identify, defects is a delamination, which can be due to expansion associated with reinforcing bar corrosion. The GPR reflections from a delamination can be relatively weak, whereas the reflection from a reinforcing bar can be fairly strong. Identifying the damage levels at an early stage can be used as a guide for efficiently planning maintenance activities. This paper presents the results of a laboratory and field study that focused on GPR methods of detecting delaminations in concrete roadways. The measurement technique used 0.5 to 6.0 GHz air-coupled waves to probe the roadways. Delaminations as small as 0.5 mm were simulated and detected in the laboratory. Field measurements are suggestive that this technique can be effective for field use.

The magnetic characteristics of ferromagnetic steels, such as hysteresis loops, total permeability and differential permeability, are dependent on mechanical stress and temperature. This dependency can be the basis for a sensitive and non invasive sensing method for measuring the active stress in steel tendons and cables. This paper describes the current status of a stress sensor which can reliably monitor stress in tendons and cables. A transient magnetic field generated by a solenoid is utilized to bring the material to technical saturation. The induced voltage, which is affected by the presence of the ferromagnetic material, is measured and related to material characteristics. In order to obtain a stable and linear calibration curve between permeability and monitoring stress for a given material, an optimal applied field H₀ must be determined. Initially, this field is estimated from the measured relationship between permeability and input voltage. The optimal working point is determined by searching for a linear relationship between permeability and applied stress at different temperatures. Temperatures from -20 C to 40 C were used. Experimental results for two common tendons, at two sizes (0.5' and 0.6'), were investigated using this technique. The results were demonstrated to match the accuracy and repeatability of a reference load cell for loading up to 70% of the yield stress. Additional tests were conducted on multi-strand cable assemblies. The experimental results indicate that the calibration procedure can be extended without significant error to accommodate such assemblies.

Several nondestructive testing methods can be used to determine the damage in a concrete structure. Linear ultrasonic techniques, e.g. pulse-velocity and amplitude attenuation, are very common in nondestructive evaluation. Velocity of propagation is not very sensitive to the degrees of damage unless a great deal of micro-damage having evolving into localized macro-damage. This transition typically takes place around 80% of the ultimate compressive strength. Amplitude attenuation is potentially more sensitive than pulse-velocity. However, this method depends strongly on the coupling conditions between transducers and concrete, hence unreliable. A baseline test of the linear acoustics of several mortar samples was conducted. These mortar samples have been previously damaged to different levels. Several other testing methods were also performed on the same samples to form a comparison. The focus is in comparing the sensitivity of a new testing method (Non-linear Acoustic NDE) with several more traditional testing methods. Non-linearity of the material stiffness is expressed in non-linear acoustics as the effect that damage and flaws have on the modulation of a signal as it propagates through the material. Spectral (non-linear) analysis is much more sensitive to lower damage states and less dependent on the repeatability of the coupling of the transducers.

The Pacific Northwest National Laboratory (PNNL) has been conducting a multi-year program for the U.S. Nuclear Regulatory Commission (NRC), Office of Nuclear Regulatory Research on the effectiveness of NDE for inspection of nuclear power plants. One task of this program concerns the development of generic flaw density and distribution functions for fabrication flaws in reactor pressure vessels. Reactor pressure vessel material from the cancelled Shoreham nuclear power plant was obtained as a part of a joint program between the NRC, Baltimore Gas and Electric and the Electric Power Research Institute. PNNL has conducted NDE inspections of this material and has estimated the density and distributions of fabrication flaws in this material. This paper discusses these inspections and the results of analyzing this inspection data. More than 4,000 fabrication flaws were detected and are described in this paper.

Pacific Northwest National Laboratory personnel have developed a cost-effective solution for implementing the use of advanced technologies for monitoring the condition and performance of aging industrial facilities. A combination of operations and maintenance (O & M) know-how together with Laboratory technical capabilities have been used to develop and demonstrate the effectiveness of a condition monitoring software system. Already proven in a moderate size pilot heating plant, the system is expected to pay large dividends in the reduction of O & M costs in an aging cogeneration facility. Additional projects are currently underway to develop this technology to its full potential. This advanced architecture was designed to provide each segment of the plant operations and maintenance (O & M) team with understandable information for making safe, cost-effective life-cycle operating decisions. The software will provide plant operators, maintenance technicians, engineering staff and administrators with on-target, on-line information that enables high process efficiency simultaneously with cost- effective, life cycle oriented, capital equipment management. This infrastructure information becomes increasingly critical as the equipment, systems, and the facility itself become older. The result of this research provides the O & M practitioner with the ability to intelligently select the asset management course of action that minimizes both the cost and risk engendered by the operation and maintenance of aging process facilities.

A high-pressure natural gas-pipe-line runs underground a valley of a river in a geological unstable area on the northern outskirts of the Alps. To get the permission to establish and to operate the pipe-line, the local authorities demanded for an information system which makes it possible to detect additional mechanical stresses in the pipe-line due to unspecified bending moments caused by the sudden gliding of the soil stratum. It was decided to use strain-gages as stress-transducers. They were bonded 20 years ago to the surface of the pipe at 8 important areas, protected against humidity and damage, wired to a measuring station on an island in the river and connected with a telephone modem to the control station of the gas company. The strains were monitored continuously, recorded and so available for assuring structural integrity of the pipe-line. When maximum allowable strains (stresses) would be exceeded there was a chance to close two pipeline valves at both sides of the river. The experiences showed, that long-time monitoring of strain-gage- signals can be used as cost-effective inspection tool for older equipment to detect possible environmentally induced damage, to assess the life integrity of the pipeline.

Like all structures, water pressure pipelines have a finite life. Pipelines will eventually begin to fail, leaving the pipeline owner to deal with the quandary: what caused this to happen, can we prevent future failures, must we replace this structure now? The causes for pipeline failure include defects and anomalies which may occur in any phase of a pipeline's life: during the engineering, the manufacture, the construction, or the operation. Failure may simply be the result of environmental conditions or old age. In the past five years, passive acoustic emission detection technology has been adapted to concrete pressure pipelines. This method of inspection is based on the caustic emissions made by the prestressed reinforcing wire as it releases its energy. A recently patented method of using this technology relies on a series of remote, independent test stations to detect, record and time-stamp these acoustic emissions. A low-powered, high- performance embedded processor system makes use of global positioning system time signals to synchronize multiple stations. These methods are re-defining the standard of care of water pressure pipelines. This paper describes pipeline failure mechanisms and a state-of-the-art data sampling system which has been developed to evaluate pipeline structural integrity.

Among the most critical components in the electric power system is the power transformer. As such, a significant body of research has been put forward to attempt to anticipate the needs for maintenance to be performed. Traditional health assessment has required sampling of oil for submission to a laboratory for analysis, but this has been deemed undesirable in light of budgetary constraints on maintenance staffing, and new predictive maintenance philosophies for substation equipment. A number of processes have been developed in recent years for online health assessment of transformers, most of which have focused on dissolved gas analysis. This paper describes a novel optical methodology for on-line transformer health assessment that utilizes an ultraviolet absorption measurement to identify the degradation of the transformer oil. An optical system was selected because of its immunity to the electromagnetic noise typical of substations, and because of the minimal impact that non-conducting materials have on the insulation system design of the transformer. The system is designed to identify deterioration and premature aging resulting from overheating, low level arcing or excessive exposure to atmospheric air. The system consists of a light source, filter, guide and detection components, and a very simple computational requirement. The measurements performed with the prototype system are validated with a high precision spectrophotometry measurement and an independent oil-testing laboratory.