Non-contact Deflection Measurement at High Speed S.Rasmussen Delft University of Technology Department of Civil Engineering Stevinweg 1 NL-2628 CN Delft The Netherlands J.A.Krarup Greenwood Engineering A/S H.J.Holst Vej 3-5C DK-265 Brøndby Denmark G.Hildebrand Road Directorate, Danish Road Institute P.O. Box 235 DK-4 Roskilde Denmark Keywords: Bearing capacity, non-contact, laser Doppler, non-destructive test, measuring apparatus, deflection, structural condition, high speed. ABSTRACT: Traditional stationary or low speed deflection measurements become problematic on roads with high traffic densities. Traffic safety is decreased when survey vehicles like low speed deflectographs or stationary falling weight devices carry out measurements on highway networks. This paper describes how the Greenwood High Speed Deflectograph applies Laser Doppler Technology to provide information about pavement response to traffic load. The information obtained is the deflection velocity, the vertical velocity of the pavement surface due to a wheel load. The results from the tests show that the High Speed Deflectograph provides repeatable data comparable to similar information obtained with the Falling Weight Deflectometer. 1 INTRODUCTION As traffic densities on many major highways have increased significantly over the last decade throughout most of the world, there is a need to replace traditional stationary or slow moving bearing capacity measurements with high speed measurements. Using stationary equipment like the Falling Weight Deflectometer (FWD) can result in dangerous situations for drivers and the FWD operator, as well as cause congestion. Another limitation of the existing equipment is the low production of data. The production of data can be increased by the introduction of high speed monitoring equipment performing continu- 1
ous measurements. An example of a high speed device for the acquisition of information on pavement condition is the High Speed Deflectograph (HSD). An operational prototype has now been developed capable of performing continuous bearing capacity measurements at driving speeds up to 8 km/h. 2 LASER BASED HIGH SPEED DEFLECTOGRAPH The unique concept of the High Speed Deflectograph utilizes laser Doppler sensors to provide the vertical deflection speed of the pavement surface. While it is important to measure the deflection speeds correctly, it is equally important to continuously register and control the position of the laser sensors using secondary measuring systems like servo systems and inertial units. The High Speed Deflectograph consists of a towing truck and a semi trailer on which the laser sensors are mounted. The photograph in Figure 1 shows the towing truck, the semi trailer and the measuring sensors mounted on a stiff beam in front of the right hand side twin wheel. The wheel provides the load that leads to the deflection of the road surface. The laser Doppler sensors then register the velocity of the deflection. (Hildebrand et al. 1999). Figure 1: Photograph showing the towing truck, the semi trailer and the measuring sensors mounted on a stiff beam in front of the right hand side twin wheel. 3 FIELD OF APPLICATION OF THE HIGH SPEED DEFLECTOGRAPH The purpose of the High Speed Deflectograph is to provide a quick, safe and reliable tool for the evaluation of the structural condition of roads at network level. As opposed to stationary measuring systems the device provides continuous measurements making it feasible to survey a whole network at regular intervals. The device can be used as a screening tool revealing discontinuities in the bearing capacity. Using suitable models for the interpretation of the data, information about the current structural condition can be obtained making it possible to detect problems at an early stage. The data from the High Speed Deflectograph can be used as a supplement to data from other measuring systems for input to pavement management systems, improving planning and cost effectiveness of maintenance work. (Rasmussen et al. 2). 2
4 MEASURING METHOD The High Speed Deflectograph is based on the basic idea of measuring the velocities of deflections rather than the absolute deflections. The concept is suitable for measuring while driving at high speeds because the rate of deflection tends to increase with increasing driving speed, although not linearly. A number of laser Doppler sensors are mounted on a heavy vehicle. The laser rays emitted from the sensors are incident on the road surface and the sensors measure velocities in the direction of the laser rays. The load will cause the road surface to deflect and the sensors register the velocity of the movement. The sensors are mounted on a rigid steel beam in front of one wheel. Each laser Doppler sensor provides information about the deflection velocity at the exact location where it measures. It is therefore necessary to employ a sensor for each location from which data are needed. Figure 2 below show a schematic layout of the prototype. The prototype is configured with two sensors. One sensor is placed inside the deflection bowl and one sensor outside the deflection bowl for reference. 7 km/h Figure 2: Schematic layout of the High Speed Deflectograph prototype. The wheel load causes the road surface to deflect and the laser Doppler sensors register the velocity of the movement. The sensors are mounted on a rigid steel beam in front of the wheel. The sensor in front, outside the deflection bowl, is used as a reference sensor. The laser sensor inside the deflection bowl provides information about the deflection velocity at the location where it measures. Apart from collecting deflection data from the laser sensors a number of other measuring systems need to be implemented. The laser sensors measure the difference in velocity between the sensors themselves and the road surface. Thus, the movement of each sensor has to be measured and subtracted from the deflection velocity output. Furthermore, since the velocity is measured in the direction of the laser rays, the incident of the laser rays on the road surface should ideally be exactly perpendicular. If the incident is not perpendicular, a component of the driving speed will be registered as part of the measurement. Calculation of the magnitude of the component of the driving speed requires the angle of incident to be measured accurately along with the driving speed. The movements of the laser sensors and the angle of incident are measured with an inertial unit 3
comprising three accelerometers and three fiber gyros. The driving speed is measured using an odometer. Movement of the laser sensors is limited and controlled by a servo system on the mounting beam to ensure that the laser sensors are focused at all times. The inertial unit and one distancemeasuring laser in each end of the mounting beam provide input data for the servo system. 5 INTERPRETATION OF DATA The minimum number of sensors required is two. On the existing prototype only two sensors are present. One sensor is positioned 25 mm in front of the load center where deflection data are to be obtained. The other sensor is a reference sensor and is placed further away in a position where no deflection occurs. Sudden changes in deflection velocity reveal discontinuities in the structure. Thus, when equipped with two laser Doppler sensors the High Speed Deflectograph can be used for screening purposes. To obtain more detailed information the device could be fitted with additional sensors. When a larger number of sensors are employed the deflection velocities are measured in several points in the deflection bowl. By integration the absolute deflections can be obtained giving access to existing interpretation methods used with the FWD. 6 PRESENTATION OF DATA In August 21 a test of the HSD prototype was carried out on the Danish highway M3. All measurements were conducted at driving speeds between 7-8 km/h. The purpose of the test program was to document the ability of the measuring concept. Several sections have been measured two or three times to allow for investigations into repeatability. FWD data from June 2 are available for the test sections and have been used to assess the reproducibility of the results. M3 is a typical Danish highway with two lanes plus a paved shoulder in each direction separated by a median. The pavement structure consists of 23-28 mm asphalt concrete on top of more than one meter of granular base and sub-base. The natural subgrade is moraine clay. 6.1 Repeatability Examples of data obtained with the prototype are presented in the following. At this time the post processing of the data is not fully complete. Calibration parameters used for the post-processing of the data has yet to be determined accurately. As a consequence the overall levels of the deflection velocities are not accurate. However, deviations in deflection velocity can easily be observed. By comparing data from subsequent measurements on the same road the repeatability can be investigated. Figures 3 and 4 show examples of repeated measurements. Subsequent measurements in each example were conducted approximately 1 hour apart. The sample-rate of the Doppler sensors is approximately 1 samples/s corresponding to a sample spacing of 2 mm when the driving speed is 2 m/s (72 km/h). The data are filtered using a moving average of 5 samples corresponding to about 1 m. The location is shown on the x-axis in km s and the deflection velocities are shown on the y-axis in m/s. Both examples show good repeatability as the curves show the same features. Differences in levels of deflection velocity are due to differences in driving speed between the measurements. They appear partly because the calibration parameters are not accurate and partly because a higher driving speed causes higher deflection velocities. The assessment of repeatability is therefore based only on the deviations. 4
Two measurements on M3.25.2.15.1.5 146 146.5 147 147.5 148 148.5 149 149.5 15 Figure 3: Example showing two HSD measurements on the same road. The difference in the overall levels is due to influence on the measurements of different driving speeds.. Figure 4: Another example of two HSD measurements showing good repeatability. Two measurements on M3.25.2.15.1.5 152.5 153 153.5 154 154.5 155 155.5 156 5
6.2 Reproducibility By comparing data with results obtained with a FWD the reproducibility of data is investigated. In figure 5 the results obtained with the HSD are compared with the maximum center deflections from the FWD. The left y-axis shows the deflection velocities in m/s and corresponds to the lower curves obtained with the HSD. The right y-axis shows absolute deflection in 1/1 mm and corresponds to the upper curve obtained with the FWD. Figure 5 shows good correlation between high deflection velocities and large maximum deflections indicating that the results from the HSD can be reproduced. Two measurements and FWD on M3.3.25.2.15.1.5 4 3 2 1-1 FWD [1/1 mm] -2 146 146.5 147 147.5 148 148.5 149 149.5 15 Figure 5: High Speed Deflectograph measurements compared with maximum absolute deflections obtained with FWD to show reproducibility. To avoid comparing velocities with distances, the slope of the deflection bowl where the HSD is measuring is estimated using the FWD data. The measured deflection velocities should be directly proportional to the slope. The HSD is measuring 25 mm in front of the wheel load. The slope is estimated from the FWD data using the maximum deflections at 2 mm and 3 mm distance from the load plate. By multiplying the slope with the driving speed the deflection velocities can be estimated: V dfwd = V ds d 2 d d 3 (1) where V dfwd =estimated deflection velocity, V ds = driving speed; d 2 =maximum deflection 2 mm from the center of the load; d 3 =maximum deflection 3 mm from the center of the load and d =distance between d 2 and d 3 (1 mm). In the following examples, Figures 6 and 7, the driving speed is assumed to be 2 m/s. 6
Two measurements and FWD on M3.25.2.15.1.5 146 146.5 147 147.5 148 148.5 149 149.5 15 Figure 6: Comparison between two HSD measurements and deflection velocities estimated from FWD data by calculating the slope and assuming a driving speed of 2 m/s (72 km/h). Two measurements and FWD on M3.25.2.15.1.5 152.5 153 153.5 154 154.5 155 155.5 156 Figure 7: Another example comparing two HSD measurements with estimated deflection velocities calculated from FWD data assuming a driving speed of 2 m/s (72 km/h). In Figures 6 and 7 the HSD results are compared with the estimated deflection velocities. The curves show good correlation and indicate that the results from the HSD can be reproduced. 7
7 FUTURE WORK The calibration parameters used as input for the post-processing of data should be calculated accurately. This will allow for a further investigation into the possibilities of extracting additional information from the data, e.g. comparing the levels of deflection velocities from different road sections. Conducting further tests using recent and closely spaced FWD measurements for comparison will improve the quality of the data available for reproducibility tests. Additional tests should be conducted on different types of roads with different pavement structures including natural subgrades. Plans for the transformation of the prototype into a production version are currently being prepared. The production version will be more rugged, certain unwanted mechanical vibrations in the system will be dampened and some measuring systems like GPS will be added. Additional laser Doppler sensors are also under consideration. It is the aim of the owner of the prototype, the Danish Road Directorate, that the modified and improved HSD will be ready for production testing in the second half of 22 8 CONCLUSION The development of the High Speed Deflectograph is now in its final stages of testing aimed at proving that the fundamental concept is functioning in practice. Results of a testing program on a typical Danish highway are described. By comparing subsequent measurements from the same road sections it has been demonstrated that data obtained with the High Speed Deflectograph show good repeatability. Reproducibility has been demonstrated by comparing the data with results obtained with the Falling Weight Deflectometer. It therefore appears that the measuring concept of the High Speed Deflectograph is feasible and that it will be possible to construct measuring devices for conducting bearing capacity measurements on roads at normal traffic speeds applying the laser Doppler concept of the HSD. 9 ACKNOWLEDGEMENTS The development of the High Speed Deflectograph for roads is supported by the Danish Agency for Trade and Industry, Ministry of Business and Industry. The Danish Road Directorate is the owner of the High Speed Deflectograph prototype. Carl Bro Pavement Consultants, Denmark also supports this project. Greenwood Engineering A/S is applying for a worldwide patent covering the idea of the High Speed Deflectograph for both roads and railroads. References Hildebrand, G. Rasmussen, S. & Andrés R. 1999. Development of a Laser Based High Speed Deflectograph. Nondestructive Testing of Pavements and Backcalculation of Moduli: Third Volume, ASTM STP 1375, S.D. Tayabji and E.O. Lukanen, Eds., American Society for Testing and Materials, West Conshohocken, PA, 1999. Rasmussen, S. & De Man A. 2. Measurement Techniques for Track Performance. Rail Engineering International Edition 2 Volume 1: 13-16. 8