Vibration Monitoring of the Voigt Bridge using Wired and Wireless Monitoring Systems

Transcription

1 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 1 Vibration Monitoring of the Voigt Bridge using Wired and Wireless Monitoring Systes Yang Wang 1, Kenneth J. Loh 2, Jeroe P. Lynch 2, Michael Fraser 3, Kincho Law 1, Ahed Elgaal 3 ( 1 Departent of Civil and Environental Engineering, Stanford University, Stanford, CA 9435, USA) ( 2 Departent of Civil and Environental Engineering, University of Michigan, Ann Arbor, MI 4819, USA) ( 3 Departent of Structural Engineering, University of California at San Diego, La Jolla, CA 9293, USA) Abstract: Structural onitoring systes using wireless sensors have the potential to serve as low-cost alternatives to coercially available cable-based onitoring systes. This paper describes a wireless sensing prototype syste specifically designed for structural onitoring applications. To validate the perforance of the prototype syste, a network of up to 2 wireless sensing units is installed on the Voigt Bridge located on the capus of the University of California, San Diego. The wireless sensor network is installed in about an hour for a short-ter study of the bridge dynaic properties. Prior to the validation test, a peranent cable-based structural onitoring syste has been installed. The acceleration response of the Voigt Bridge concrete box girder is recorded by both onitoring systes. Strong agreeent fro the data collected by the two systes is observed. The wireless sensing units are also prograed to locally process their easureent data in real-tie using an ebedded fast Fourier transfor algorith; Fourier response spectra are then wirelessly transitted to the wireless network server. The easureents acquired using the wireless onitoring syste are shown to be accurate for precise deterination of the priary odal frequencies and operating deflection shapes of the bridge deck. Key words: Structural onitoring, wireless sensing, sensor networks, data acquisition, operating deflection shape INTRODUCTION Structural health onitoring (SHM) has eerged in recent years as an active research area, especially as civil infrastructure systes continue to experience perforance degradation due to aterial aging, iproper usage, and various types of hazardous events (Farrar et al. 23). A structural health onitoring syste collects and analyzes online inforation about a structure so that indications of structural distress can be identified early. Many types of sensors are coercially available for easuring structural response inforation that can then be used for diagnosing structural safety conditions. Traditional structural onitoring systes require the installation of extensive lengths of cables so that data fro ultiple sensors deployed in a structure can be reliably collected. For a typical low-rise building, the installation of a coercial cable-based onitoring syste is estiated to cost a few thousand dollars per sensing channel (Celebi 22). As the size of the structure grows, additional cabling ight result in significant increase in both onetary cost and tie for

2 2 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 syste installation. To eradicate the high costs associated with installing cable-based structural onitoring systes, state-of-the-art wireless technologies can be explored for adoption (Straser and Kireidjian 1998, Lynch et al. 25, Wang et al. 26). Besides being cost-effective, wireless structural onitoring systes offer the convenience of easily reconfiguring sensor locations. Copared to traditional cable-based systes, wireless structural onitoring systes have a unique set of technical challenges. First, wireless sensing units norally use batteries as a cheap and convenient power source. However, the liited energy supplied by batteries poses a scarce resource for power-consuing wireless transissions. Second, wireless data transission is inherently less reliable than cabled transissions. Furtherore, the data transfer rates of wireless coponents are norally uch lower than those offered by cabled systes. Last but not least, clock synchronization for a wireless sensing syste is ore challenging than for cable-based systes, where a single syste clock located at the data server is used. The wireless structural onitoring syste described in this paper attepts to address soe of these technical challenges to offer a level of perforance on par with cable-based onitoring systes. The wireless onitoring syste provides reliable data acquisition capabilities with counication ranges appropriately scaled to the physical diensions of a ediu-sized civil structure. This paper highlights the key features of the wireless structural health onitoring syste. First, the hardware eleents of the wireless sensing units are described. Second, a signal conditioning circuit is designed to itigate sensor noise and to aplify low-level response signals typical in civil structures. Finally, the paper presents a set of field validation tests conducted on the Voigt Bridge, located on the capus of the University of California, San Diego (UCSD). WIRELESS SENSING SYSTEM ARCHITECTURAL DESIGN To offer flexible deployents in civil structural applications, a siple star-topology network is proposed for the wireless structural onitoring syste described herein. The syste includes one wireless network server and ultiple wireless sensing units. Each wireless sensing unit ay collect data fro ultiple sensors, including acceleroeters, velocity eters, and strain gages, aong others. Incorporated with ebedded icrocontrollers, the wireless sensing units are endowed with the coputational resources that allow the to process their sensor data. The units can also wirelessly counicate sensor data or coputation results to the network server. Fig. 1 shows the overall hardware design of the wireless sensing unit. The wireless sensing unit consists of three functional odules: sensor signal digitizer, coputational core, and wireless counication odule (Fig. 1). The sensor signal digitization odule converts analog sensor signals into digital forats which are then transferred to the coputational core through a high-speed Serial Peripheral Interface (SPI) port. The coputational core then buffers the sensor data in its local eory or processes the data with ebedded engineering analytical routines. Through a Universal Asynchronous Receiver and Transitter (UART) interface, the coputational core is able to counicate

3 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 3 WIRELESS SENSING UNIT Sensor Signal Conditioning Aplification, Filtering, and Voltage-offsetting Structural Sensors Sensor Signal Digitization 4-channel 16-bit Analog-to-Digital Converter ADS8341 SPI Port Coputational Core 128kB External SRAM CY62128B Parallel Port 8-bit Microcontroller ATega128 UART Port Wireless Counication Wireless Transceiver : 2kbps 2.4GHz 24XStrea, or 4kbps 9MHz 9XCite Fig. 1. Functional diagra detailing the hardware design of the wireless sensing unit. with a wireless transceiver, which enables the wireless sensing unit to exchange data with the network server. The functional odules are integrated using a copact two-layer printed circuit board (PCB) as shown in Fig. 2a. All of the hardware coponents, including batteries, are packaged within a weatherproof plastic container, which has a diension of c 3. As shown in Fig. 1, the structural sensor signal ay optionally be aplified and filtered by a sensor signal conditioning odule before the signal is fed into the wireless sensing units. The key coponents and the characteristics of the wireless sensing unit design, as well as the design of the offboard signal conditioning odule are described briefly below. Sensing signal digitization odule The ain coponent of the sensor signal digitization odule is a 4-channel 16-bit analog-to-digital (A/D) converter (Texas Instruents ADS8341). Each wireless sensing unit can accoodate signals fro a heterogeneous set of structural sensors, as long as their outputs are analog voltages fro to 5V. The 16-bit A/D resolution is sufficient for ost structural onitoring studies. The highest sapling rate supported by this A/D converter is 1 khz, which is uch higher than the sapling frequencies typically eployed when onitoring civil structures. Coputational core The coputational core of the wireless unit is responsible for executing ebedded software instructions for engineering analyses. A low-cost 8-bit icrocontroller (Atel ATega128) is selected as the principle coponent of the coputational core. The key criterion for this selection is to balance the power consuption and cost of the icrocontroller versus the coputation power needed by software applications. Running at 8MHz, the ATega128 consues about 15A when it is active. The icrocontroller Sensor Signal Input Conditioned Output Signal ATega128 Microcontroller Connector to Wireless Transceiver Sensor Connector A/D Converter ADS8341 Octal D-type Latch AHC573 SRAM CY62128B Switch for Selecting Aplification Factor Operational Aplifier LMC6484 (a) Wireless sensing unit ( c 2 (b) Sensor signal conditioning ) board. odule ( c 2 ). Fig. 2. Photographs of the Printed Circuits Boards.

4 4 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 also contains 4kB static rando access eory (SRAM) for storing stack and heap variables. The 4kB SRAM is often insufficient for the execution of ebedded data interrogation algoriths. To address this issue, an external 128kB eory chip (Cypress CY62128B) is also incorporated within the wireless sensing unit design. Wireless counication odule The wireless sensing unit is designed to be operable with two different wireless transceivers: 9MHz MaxStrea 9XCite and 2.4GHz MaxStrea 24XStrea. This unique design feature is intended to allow users to eploy the legal open-use frequency band in their regions. Pin-to-pin copatibility between these two wireless transceivers akes it possible for the two odules to share the sae hardware connections on the PCB of the wireless sensing unit. Table 1 suarizes the key perforance paraeters of the two wireless transceivers. As shown in the table, the data transfer rate of the 9XCite is twice as fast as the data rate of the 24XStrea; however, the 24XStrea provides a longer counication range but consues uch ore battery power. Signal conditioning odule For field applications, a wireless onitoring syste ust be able to record both abient and forced structural vibrations. Most abient vibrations in civil structures are characterized by low-aplitude accelerations. Recording these low-aplitude signals can be challenging because the A/D converter is vulnerable to electrical noise in the circuit. A signal conditioning odule is designed to aplify signals, filter out noise, and shift the range of sensor signals. The filtering circuit consists of a high-pass resistor-capacitor (RC) filter with a cutoff frequency of.2 Hz and a low-pass fourth-order Bessel filter with a cutoff frequency of 25 Hz. The linear-phase shift property of the Bessel filter ensures a constant tie delay for signals in the pass band, thus aintaining the signal wavefor in the tie doain. Fig. 2(b) shows the coplete signal conditioning circuit odules that support the filtering, offsetting, and aplification of sensor signals. FIELD VALIDATION TESTS AT VOIGT BRIDGE Laboratory and field validation tests have been previously conducted to verify the perforance of the wireless structural onitoring syste (Lu et al. 26, Lynch et al. 25, Lynch et al. 26). Field tests are particularly helpful in assessing the liitations of the syste, and providing valuable experience that can lead to further iproveents in the syste hardware and software design. The following sections present an overview of the validation tests conducted on the Voigt Bridge located on the Table 1. Key perforance paraeters of the wireless transceivers*. Specification 9XCite 24XStrea Operating Frequency ISM MHz ISM GHz Data Transfer Rate 38.4 kbps 19.2 kbps Counication Range Up to 9 indoor, 3 outdoor Up to 18 indoor, 5k outdoor Supply Voltage 2.85VDC to 5.5VDC 5VDC (±.25V) Power Consuption 55A transitting, 35A receiving, 2µA standby 15A transitting, 8A receiving, 26µA standby * For details about the transceivers, see

5 1.8 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 5 UCSD capus. Up to 2 wireless sensing units are deployed in the field to siultaneously collect the abient and forced vibration response of the bridge for operating deflection shape analysis. Voigt Bridge Voigt Bridge is a concrete box girder highway bridge that carries traffic over Interstate 5. The two-lane bridge is about 89.4 long and consists of four spans (Fig. 3). The bridge deck has a skew angle of about 32º, with the concrete box-girder supported by three single-colun bents. Over each bent, a lateral diaphrag with a thickness of about 1.8 stiffens the girder. The thickness of the concrete lateral diaphrags poses substantial challenges for the transission of wireless signals within the box girder. Longitudinally, the box girder is partitioned into five cells running the length of the bridge (Fig. 3b). Coparison between cabled and wireless sensor data Girder cells along the north side of the bridge are accessible through four anholes on the bridge sidewalk. As a testbed project for structural health onitoring research, a sophisticated cable-based structural onitoring syste has been installed in the northern-ost cells of the Voigt Bridge (Fraser et al. 26). The cable-based syste includes acceleroeters, strain gages, therocouples, and huidity sensors. For the purpose of validating the proposed wireless structural onitoring syste, thirteen acceleroeters interfaced to wireless sensing units are installed within the two iddle spans of the bridge to easure vertical vibrations. One wireless sensing unit (associated with one signal conditioning odule and one acceleroeter) is placed iediately below the acceleroeter associated with the peranent wired onitoring syste. While the wired acceleroeters are ounted to the cell walls, wireless acceleroeters are siply ounted on the floor of the girder cells to expedite the installation process. The installation and calibration of the wireless onitoring syste, including the placeent of the 13 wireless sensors, takes about an hour. The Maxstrea 9XCite wireless transceiver operating at 9MHz (allowed by US governent regulations) is integrated with Lateral diaphrag A Longitudinal diaphrag Abut Bent 1 Bent 2 Bent A Abut N Wireless network server One pair of wireless and wired acceleroeters (a) Plan view of the bridge illustrating sensor locations of wired and wireless onitoring systes. Wired acceleroeter Wireless acceleroeter 1.7 Section A-A (b) Elevation view to section A-A. (c) Side view of the bridge over Interstate 5. Fig. 3. Voigt Bridge on the capus of the University of California, San Diego.

6 6 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 each wireless sensing unit. Two types of acceleroeters are associated with each onitoring syste. At locations #3, 4, 5, 9, 1, and 11 in Fig. 3a, PCB Piezotronics 381 acceleroeters are used with both the cabled and the wireless systes. At the other seven locations, Crossbow CXL1LF1 acceleroeters are used with the cabled syste, while Crossbow CXL2LF1Z acceleroeters are used with the wireless syste. Table 2 suarizes the key paraeters of the three types of acceleroeters. Signal conditioning odules are used for filtering noise, aplifying and shifting signals for the wireless acceleroeters. The signals of the wired acceleroeters are directly digitized by a National Instruents PXI-631E data acquisition board (Fraser et al. 26). Sapling frequencies for the cable-based syste and the wireless syste are 1, Hz and 2 Hz, respectively. The bridge is under noral traffic operation during the tests. Fig. 4(a) shows the tie history data at locations #6 and #12, collected by the cable-based and wireless onitoring systes when a vehicle passes over the bridge. A close atch is observed between the data collected by the two systes. The inor difference between the two data sets can be ainly attributed to two sources: 1) the signal conditioning odules are used in the wireless syste but not in the cabled syste; 2) the wired and wireless acceleroeter locations are not exactly adjacent to each other, as previously described. Table 2. Paraeters for acceleroeters used in the cabled and wireless sensing systes. PCB381 CXL1LF1 CXL2LF1Z Maxiu Range ±3 g ±1 g ±2 g Sensitivity.7 V/g 2 V/g 1 V/g RMS Resolution (Noise Floor).5 g.5 g 1 g Acceleration (g) 5 x 1-3 Wired # x 1-3 Acceleration (g) 5 x 1-3 Wireless # x 1-3 FFT Magnitude Wired # Tie (s) 4 2 Wireless # Tie (s) (a) Coparison between wired and wireless tie history data. Wired # Wired # Frequency (Hz) FFT Magnitude.5 Wireless # Wireless # Frequency (Hz) (b) Coparison between FFT to the wired data, as coputed offline by a coputer, and FFT to the wireless data, as coputed online by the wireless sensing units. Fig. 4. Coparison between results fro two onitoring systes (Location nubers are as shown in Fig. 3a).

7 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 7 Fig. 4(b) shows the Fourier spectru results deterined fro the tie history data. The FFT results using the data collected by the cabled syste are coputed offline, while the FFT results corresponding to the wireless data are coputed online in real-tie by each wireless sensing unit. After each wireless sensing unit executes its FFT algorith, the FFT results are wirelessly transitted to the network server. Strong agreeent between the two sets of FFT results validates the coputational accuracy of the wireless sensing units. It should be pointed out that because the sapling frequency of the cabled syste is five ties higher than that of the wireless syste, the agnitude of the Fourier spectru for the cabled syste is also about five ties higher than those for the wireless syste. Operating deflection shape analysis One attractive feature of the wireless onitoring syste is its easy reconfigurability. To deterine the operating deflection shapes of the bridge deck, the configuration of the original wireless onitoring syste is changed to attain a ore suitable spatial distribution. Twenty wireless acceleroeters and the wireless network server are now ounted to the bridge sidewalks, instead of inside the girder cells (Fig. 5). The counication distance between the server and the farthest-away wireless sensing unit is close to the full length of the bridge. Both vehicle traffic and haer excitations are eployed during the test. Haer excitation is applied during intervals of no passing vehicles. The operating deflection shape (ODS) analysis presented in this paper is based on the data collected during a haer excitation test. DIAMOND, a odal analysis software package, is used to extract the operating deflection shapes of the bridge deck (Doebling et al. 1997). Under haer excitation, the operating deflection shapes at or near a resonant frequency should be doinated by a single ode shape (Richardson 1997). Fig. 6 presents the first four doinant operating deflection shapes of the bridge deck using wireless acceleration data. The ODS #1, #2, and #4 show priarily Abut Bent 1 Bent 2 Bent Abut N Wireless network server Wireless acceleroeter Haer location Fig. 5. Wireless acceleroeter deployent for operating deflection shape analysis. ODS #1, 4.89Hz ODS #2, 6.23Hz ODS #3, 8.1Hz ODS #4, 11.64Hz Fig. 6. Operating deflection shapes extracted fro wireless sensor data.

8 8 The Proceeding of 4 th China-Japan-US Syposiu on Structural Control and Monitoring Oct.16-17, 26 flexural bending odes of the bridge deck; a torsional ode is observed in ODS #3. SUMMARY AND DISCUSSION A wireless structural health onitoring syste specifically designed for expedient onsite deployent to civil structures is presented in this paper. Robust software and hardware designs enable low-cost and reliable data collection and interrogation fro a network of autonoously functioning wireless sensing units. The paper presented findings fro a field validation test conducted at the Voigt Bridge located on the UCSD capus. Strong agreeent is observed between the data collected by the wireless syste and the data collected by a baseline cable-based onitoring syste. Operating deflection shapes of the bridge deck are successfully References Celebi, M., 22. Seisic Instruentation of Buildings (with Ephasis on Federal Buildings). Report No United States Geological Survey (USGS), Menlo Park, CA, USA. Doebling, S.W., Farrar, C.R., Cornwell, P.J., DIAMOND: A Graphical User Interface Toolbox for Coparative Modal Analysis and Daage Identification. Proc. of 6th International Conference on Recent Advances in Structural Dynaics, Southapton, UK. Farrar, C.R., Sohn, H., Heez, F.M., Anderson, M.C., Beent, M.T., Cornwell, P.J., Doebling, S.W., Schultze, J.F., Lieven, N., Robertson, A.N, 23. Daage Prognosis: Current Status and Future Needs. Report LA-1451-MS, Los Alaos National Laboratory, NM, USA. Fraser, M., Elgaal, A., Conte, J.P., 26. UCSD Powell Laboratory Sart Bridge Testbed. Report No. SSRP 6/6, Departent of Structural Engineering, University of California, San Diego, La Jolla, CA, USA. Lu, K.-C., Wang, Y., Lynch, J.P., Loh, C.-H., Chen, Y.-J., Lin, P.-Y., Lee, Z.-K., 26. Abient Vibration Study of the Gi-Lu Cable-Stay Bridge: Application of Wireless Sensing Units. Proc. of SPIE 13th Annual Syposiu on Sart Structural and Materials. San Diego, CA, USA. obtained using the acceleration data collected siultaneously by 2 wireless sensing units. ACKOWLEDGEMENT This research is partially funded by the National Science Foundation under grants CMS (Stanford University), CMS (University of Michigan), ITR (UCSD), and the Office of Naval Research Young Investigator Progra awarded to Prof. Lynch at the University of Michigan. The authors wish to thank following three fellowship progras: the Office of Technology Licensing Stanford Graduate Fellowship, the Rackha Grant and Fellowship Progra at the University of Michigan, and the fellowship provided by California Institute for Telecounications and Inforation Technology. Lynch, J.P., Loh, K., 25. A Suary Review of Wireless Sensors and Sensor Networks for Structural Health Monitoring. Shock and Vibration Digest, 38: Lynch, J.P., Wang, Y., Law, K.H., Yi, J.H., Lee, C.G., Yun, C.B., 25. Validation of Large-Scale Wireless Structural Monitoring Syste on the Geudang Bridge. Proc. of 9th International Conference on Structural Safety and Reliability. Roe, Italy. Lynch, J.P., Wang, Y., Lu, K.-C., Hou, T.-C., Loh, C.-H., 26. Post-seisic Daage Assessent of Steel Structures Instruented with Self-interrogating Wireless Sensors. Proc. of 8th National Conference on Earthquake Engineering. San Francisco, CA, USA. Richardson, M. H., Is It A Mode Shape Or An Operating Deflection Shape?. Sound and Vibration Magazine, 31: Straser, E.G., Kireidjian, A.S., A Modular, Wireless Daage Monitoring Syste for Structures. Report No. 128, John A. Blue Earthquake Eng. Ctr., Stanford Univ., Stanford, CA, USA. Wang, Y., Lynch, J.P., Law, K.H., 26. A Wireless Structural Health Monitoring Syste with Multithreaded Sensing Devices: Design and Validation. Structure and Infrastructure Engineering, in press.