490 C.K. Lau et al. is + 835mm. Figure 8 and 9 illustrate the longitudinal movement and range of longitudinal movement of Tsing Ma Suspension Bridge at the expansion joint at the Tsing Yi Abutment. It is noted that the trend of the movement, when converted into effective temperature of the bridge, can be readily represented by a straight line. The gradient of the fitted straight line for Tsing Ma Suspension Bridge is 24.4 mm per ~ while the range of movement as revealed from Figure 9 is between i 450mm only. This also illustrates that the longitudinal movement of the bridge is well within the allowable tolerance. Vertical displacement of the bridge deck of Tsing Ma Suspension Bridge is monitored by means of a level sensing system which, by means of a pair of fluid conduits, measures the change of fluid pressure at various locations of the deck and derives the corresponding reference datum. Vertical displacement of the deck is basically a function of temperature and live load. Figure 10 illustrates the measured vertical movement of the bridge deck at mid-span in the past two years. It can be revealed from the figure that the maximum downward deflection of the bridge deck due to combined traffic and temperature effect is within a range of 1.4m, which is well within the designed values of 6m (respectively 4.7 m for live load and 1.3 m for temperature effect). TENSILE LOAD MONITORING ON SUSPENDERS Figure 11 illustrates the measured load on the suspenders of Tsing Ma Suspension Bridge derived by field measurement. The field measurement was conducted in 1998 and a total of 2 x 95 = 190 No. suspenders were measured. Portable accelerometers were used to measure the ambient vibration of the suspenders. The time series data were then converted to frequency domain (spectrum analysis) to give the natural frequencies of the suspenders under ambient condition. These frequencies were then used to derive the tensile force of the individual suspender during the period of measurement. (Note : Tension in a taut string/wire is a function of material/sectional properties and fundamental frequency) According to the design information, the self-weight of an 18 metres deck unit for the bridge (Dead Load + Superimposed Dead Load) is about 516 tons. This weight is to be taken by two groups of suspenders (south and north) during the erection stage, i.e., about 258 tons per group. The increase of the suspenders' tensile force since erection represents the presence of other superimposed dead load, including railway slab, servicing and live load on the bridge. However, the result of the field measurement indicated that all the tensile load now taken by the suspenders are within the Serviceability Limit State for material strength, i.e., 448 tons per group of suspenders. The typical range of tensile load for the suspender is between 300 tons and 400 tons. The corresponding current maximum tensile load taken by each 76mm diameter hanger strand is 75 - 100 tons. VEHICULAR TRAFFIC LOAD MONITORING The intensity of vehicular traffic loads on long-span bridges is govemed by the effects of groups of vehicles in traffic jams and is derived by statistical simulation from local vehicular characteristic data and future prediction of change. The required loaded length used in the design is based on number of daily traffic jams, locations of the jams, duration and distribution of vehicular types and traffic flow during the jams. The design HA Lane Factor for Tsing Ma Suspension Bridge is 3.6. In order to assess the validity of the above design parameter, it is necessary to have a corresponding traffic load monitoring for the bridge. Weigh-in-motion sensors are provided on the Lantau Fixed Crossing so that vehicular number, axle- weights, speeds and type of vehicles crossing the Tsing Ma Suspension Bridge can be measured. The measured data are then used to formulate a database to derive the percentage of goods vehicles and to Structural Performance Measurements for Tsing Ma Suspension Bridge 491 compare with the HA Lane Factor used in traffic load design. Figure 12 illustrates the monthly daily average percentage of goods vehicles crossing the bridge. It is noted that the current percentage of goods vehicles using the bridge is about 34% of the total vehicle. This value is well below the design percentage of 60%. RAILWAY LOAD MONITORING Railway traffic load is one of the most important parameters affecting the structural design of the bridge. Loads due to railway traffic on Tsing Ma Suspension Bridge are monitored by means of strain gauges installed at waybeams which support the railway trackform. The monitoring works include the conversion of the recorded waybeam strains into bogie load data and train load data and subsequent derivation of the train weight, passing rate and rainflow counts for fatigue life estimation. Figures 13 illustrates the vertical acceleration record on Tsing Ma Bridge during the past two years of operation. It can been denoted that the average value of vertical acceleration is in the range of 100 to 150 mm/sec 2. The results also reveal that the derived maximum accelerations (on a monthly basis) run closely with the desirable operational upper limit for train running but are still well below the allowable maximum value. Figure 14 is a train load monitoring plot showing the load configuration of a typical 7-car train (i.e., 14 No. bogies) passing the bridge. It can been seen from the derived bogie loads that they are all within the designed envelop (tare load and crush load) and is in agreement with the designed load pattern for the designed train. STRAIN/STRESS MONITORING OF VARIOUS STRUCTURAL MEMBERS Strain gauges are installed at a number of critical locations on the bridge to measure the change in strain of the structural members under different loading conditions. The instrumented locations include chord members of the longitudinal trusses, cross-frame chord members, plan bracing members, deck trough and rocker bearings at Ma Wan Tower. The measured strains are recorded and then used to derive axial, shear and bending stresses of the members and the corresponding loading effects. Figure 15 illustrates the measured strain results of the chord members on the outer longitudinal truss at Chainage 24662.5. Figure 16 shows the load monitoring results for the outer rocker bearing at Ma Wan Abutment during the past two years. The measured stress values are used to compare with the designed values of the members at both the Serviceability Limit State and Ultimate Limit State, whilst the measured strain can be used to establish the rainflow counts for fatigue life estimate. Again, it can be revealed that the current stress levels of the critical structural members are well below the designed limits. DYNAMIC RESPONSE MONITORING The dynamic characteristics of a structure can be represented by its mode shapes and frequencies. Accelerometers are installed at various strategic locations of the bridge deck and main cables so that their dynamic characteristics or response under vibration can be measured and monitored. Table 4 shows the measured results of the first four frequencies and their corresponding mode shapes of Tsing Ma Suspension Bridge. The corresponding values computed by the designer of the bridge and others are also illustrated for comparison. It could be revealed that the measured results are in general higher than those derived in the design and the measured values obtained during bridge construction. This implies that the stiffness of the as-built bridge-deck is stiffer than that of the designed values and that during construction stage. 492 C.K. Lau et al. TABLE 4 COMPARISON OF COMPUTED AND MEASURED FREQUENCIES OF TSING MA SUSPENSION BRIDGE Type and Order of Mode Shape Lateral Mode 1st 2nd 3rd 4th Vertical Mode 1st 2nd 3rd 4th Computed MMHK 1 (Designer) 0.065 Hz 0.164 Hz 0.112 Hz 0.141 Hz __- __- Computed FNp2 (Checker) 0.064 Hz 0.149 Hz 0.266 Hz 0.455 Hz 0.112 Hz 0.133 Hz 0.179 Hz 0.233 Hz Measured HKPU 3 ( without paving ) 0.069 Hz 0.164 Hz 0.214 Hz 0.226 Hz 0.113 Hz 0.139 Hz 0.184 Hz 0.241 Hz Rotational Mode 1st 0.259 Hz 0.235 Hz 2nd 0.276 Hz 0.268 Hz 3rd 0.409 Hz 4th 0.533 Hz MMHK1 _ Mott MacDonald Hong Kong Limited HKPU 3 - Hong Kong Polytechnic University HyD 5 - Highways Department 0.267 Hz 0.320 Hz Measured THU 4 (with paving ) 0.069 Hz 0.161 Hz 0.242 Hz 0.246 Hz 0.114 Hz 0.137 Hz 0.183 Hz 0.240 Hz 0.265 Hz 0.320 Hz 0.485 Hz 0.591 Hz Measured HyD 5 (as-built) 0.070 Hz 0.170 Hz 0.254 Hz 0.301 Hz 0.114 Hz 0.133 Hz 0.187 Hz 0.249 Hz 0.270 Hz 0.324 Hz 0.486 Hz 0.587 Hz FNP 2 - Flint & Neill Partnership. THU4- Tsinghua University CONCLUSION The measured results on wind, traffic and temperature loads indicate that the loads acting on the bridge are far less than the design load values. The measured/derived results on bridge responses indicate that the current stresses and displacements at critical locations are far below the design response values. It is therefore concluded that the bridges are currently under healthy condition. ACKNOWLEDGEMENT The authors wish to express their thanks to Director of Highways, Mr. K.S. Leung, for permission to publish this paper. Any opinions expressed or conclusions reached in the text are entirely those of the authors. References 1. Lau, C. K. and Wong, K.Y., "Design, Construction and Monitoring of the Three Key Cable-Supported Bridges in Hong Kong", Proceedings of the Fourth International Conference on Structures in the New Millennium", 3-5 September 1997 in Hong Kong, A.A. Balkema, Rotterdam, Netherlands. 2. Lantau Fixed Crossing Project Management Office, Highways Department, "Structural Health Monitoring System", Highway Contract No. HY/93/09 - Electrical and Mechanical Services in Lantau Fixed Crossing, The Hong Kong Government, 1993. 3. Highways Department, "Structures Design Manual", The Government of Hong Kong Special Administrative Region, 1997. Structural Performance Measurements for Tsing Ma Suspension Bridge 493 494 C.K. Lau et al. Structural Performance Measurements for Tsing Ma Suspension Bridge 495 496 C.K. Lau et al. WIND CHARATERISTICS AND RESPONSE OF TSING MA BRIDGE DURING TYPHOON VICTOR L.D.Zhu 1, Y.L.Xu 1, K.Y.Wong 2 and K.W.Y.Chan 2 1 Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China 2 Lantau Fixed Crossing Office, Highways Department, Hong Kong, China ABSTRACT On 2 August 1997, Typhoon Victor just crossed over the Tsing Ma Bridge in Hong Kong. The Wind and Structural Health Monitoring System (WASHMS) installed on the Bridge timely recorded both wind and structural response time-histories of seven hours duration. The recorded wind and structural response data are analysed in this paper for evaluating wind characteristics and acceleration response of the Bridge. The result shows that during Typhoon Victor, both mean and turbulent wind characteristics varied considerably due to the change of wind direction and the upwind terrain. Larger turbulence intensities and gust factors are obtained during Typhoon Victor, compared with those due to seasonal trade wind. It is also confirmed that the wind excitation mechanism of the Bridge in the lateral direction is different from that in the vertical direction or the rotation. The alongwind acceleration response of the Bridge is approximately proportional to mean wind speed square while the vertical acceleration and torsional angular acceleration are almost proportional to mean wind speed cubic. Furthermore, the natural frequencies identified from the acceleration response spectra are consistent with those obtained from the ambient vibration measurement or the numerical analysis carried out before. KEYWORDS Typhoon Victor, Suspension bridge, Wind characteristics, Acceleration response, Natural frequency. INTRODUCTION With the increase of span length of modem suspension bridges, the prediction of bridge response to strong winds becomes more and more important for the bridge constructed within a wind-prone area. To this end, some analytical methods, computational fluid dynamics technique, and wind tunnel test technique have been developed in the past two or three decades. To verify these analytical and numerical methods as well as wind tunnel tests, the field measurements of wind characteristics and bridge response play an important role. However, field measurement data, especially those during severe storms such as typhoons, are very limited up to now. 497 498 L.D. Zhu et al. Figure 1: The moving path of Typhoon Victor Figure 2: Schematic diagram of the topography ofHong Kong On 2 Aug. 1997, about three months after the opening of the Tsing Ma Bridge in Hong Kong, Typhoon Victor just crossed over the Bridge and made landfall over the western part of the New Territories. The WASHMS installed on the Bridge by the Highways Department of Hong Kong Special Administrative Region timely recorded wind speed and bridge response time-histories of seven hours duration (Lau et al, 1998). These recorded wind and structural response data are analysed in this paper to evaluate wind characteristics and acceleration response of the Bridge and to provide a basis for the verification of the currently used analytical or numerical or experimental methods at a late stage. Before the presentation and discussion of the measured wind characteristics and bridge responses, a brief introduction of Typhoon Victor, the Bridge and its surroundings, and the measurement instrumentation is given first. TYPHOON VICTOR Tropical depression Victor originated in the middle of the South China Sea on 31 July 1997 and its intensity continuously increased afterwards (Lee et al, 1998). The tropical depression Victor first moved northwesterly for 12 hours and then had a sudden turn to near north and remained in almost the same direction during its passage over Hong Kong (see Figure 1). The tropical depression Victor became a real typhoon when it entered the region of 250km south of Hong Kong at 8:00 on 2 August 1997 HKT (Hong Kong Time). At 19:00 HKT on 2 August, the centre of Typhoon Victor moved into the region about 8km east of Cheung Chau Island (see Fig. 2). The lowest air pressure measured on Cheung Chau Island at sea level was 972hPa. Typhoon Victor then crossed over the Tsing Ma Bridge at 20:05 and made landfall over the western part of the New Territories. Victor crossed over the whole Hong Kong within 2 hours at the average translational speed about 25km per hour. After leaving Hong Kong, Typhoon Victor continued moving in the north until it decayed on 3 August in the Southeast of China. The measured highest 10 minute mean wind speed in the wall area of the Typhoon during its passage over Hong Kong was about 110km per hour (30.6m/s) at a 500 m height above the ground, just 2 hours after its landfall (Lee et al, 1998). TSING MA BRIDGE AND TOPOGRAPHY The Tsing Ma Bridge in Hong Kong is a suspension bridge with an overall length of 2160m and a main span of 1377m, carrying a dual three-lane highways on the upper level of the bridge deck and two railway tracks and two carriageways on the lower level within the bridge deck (see Figs.3 and 5). The Wind Response to Tsing Ma Bridge During Typhoon Victor 499 alignment of bridge deck deviates from the east-west axis for about 17 ~ in anticlockwise. The bridge deck is 41m wide and 7.643m high (see Fig. 5). The two bridge towers of 206 m high are made of pre- stressed concrete. The east bridge tower sits on the Northwest shoreline of Tsing Yi Island, called the Tsing Yi tower while the west bridge tower sits on Ma Wan Island, called the Ma Wan tower. Hong Kong is situated in the coastal area of South China. Not only there are many islands in Hong Kong, but also there are many mountains covering most areas of the territory. The topography of Hong Kong thus varies from place to place (see Fig. 2). The local topography surrounding the Tsing Ma Bridge within the dashed circle of 5kin in radius is a typical example. The bridge is embraced by sea, islands, and mountains of 200 to 500m high. If taking the bridge as a centre, the whole surrounding area may be roughly classified into seven types of regions (I to VII), bounded by seven lines R1, R2, R3, R4, R5, R6 and R7 as shown in Fig. 2. The TsinYi Island adjacent to the Bridge is in Region I and VII. The top levels of Tsing Yi Island are 218m in the north (Region I) and 334m in the south (Region VII). The Ma Wan Island adjacent to the Bridge is in Regions III and IV. The top level of Ma Wan Island is 69m only. Figure 3: Elevation ofTsing Ma Bridge Figure 4: Locations of anemometers and accelerometers Figure 5: Positions of sensors on cross section of bridge deck INSTRUMENTATION AND DATA ANALYSIS There are altogether seven different type sensors installed for WASHMS, including, amongst others, six anemometers and 24 uni-axial servo type accelerometers (Lau, et al, 1998). Two digital ultrasonic anemometers (AneU), called Gill Wind Master Ultrasonic Anemometer, were installed on the north side and south side, respectively, of the bridge deck at the mid-span. They are specified as WITJN01 and WITJS01 in Figs 4 and 5. Each ultrasonic anemometer can measure three components of wind velocity simultaneously. Two analogue mechanical anemometers (AneM) were located at two sides of the bridge deck near the middle of the Ma Wan side span, specified as WITBN01 at the north side and WITBS01 at . load monitoring for the bridge. Weigh -in- motion sensors are provided on the Lantau Fixed Crossing so that vehicular number, axle- weights, speeds and type of vehicles crossing the Tsing Ma Suspension. The monitoring works include the conversion of the recorded waybeam strains into bogie load data and train load data and subsequent derivation of the train weight, passing rate and rainflow counts. the Three Key Cable-Supported Bridges in Hong Kong", Proceedings of the Fourth International Conference on Structures in the New Millennium", 3-5 September 1997 in Hong Kong, A.A.