Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
13 Laboratory and Field Evaluations of Ground-Penetrating Radar Systems This chapter was prepared by Kenneth Maser and the staff of Infrasense, Inc. The following three ground-penetrating radar (GPR) systems were evaluated in Task 6 of this study: ⢠A 3-GHz horn antenna and a 2.6-GHz âground-coupledâ antenna provided by Geophysical Survey Systems, Inc. (GSSI); ⢠A 1.3-GHz ground-coupled antenna array (MIRA) and a 2.3-GHz ground-coupled antenna provided by MALA; and ⢠A swept frequency (150 MHz to 3 GHz) noncontact antenna array provided by 3d-Radar. The GSSI system focused on the implementation of new high-frequency antennae, which would have the resolution to detect the small changes associated with pavement delamina- tion. The specific advantage of the horn antenna is that it is noncontact and can be used to survey a pavement at much higher speeds than a ground-coupled antenna, which requires contact with the pavement. The horn antenna tested as part of this work was a prototype. The ground-coupled antenna is a manufactured product. The MIRA system from MALA is a 16-channel array that has the advantage of obtaining greater coverage with multiple paths from the array of transmitters and receivers. Its dis- advantage is that it is a ground-coupled system that is deployed at relatively low (walking) speed, and that the antenna frequency (1.3 GHz) is not optimal for delamina- tion detection. To address these concerns, MALA also tested a 2.3-GHz ground-coupled antenna for which a high-speed deployment arrangement was available. The 3d-Radar system uses a 29-channel array (14 trans- mitters and 15 receivers) producing 29 channels of data. The system is operated in a swept frequency mode from 150 MHz to 3.0 GHz, thus producing data over a range of depths. The array has the coverage advantages described above for the MIRA system. In addition, the antenna elements are housed in a single unit that operates about 6 to 12 in. above the pave- ment surface. Two rounds of testing were carried out with the equipment described above. The first round took place November 8â9, 2009, and the second round took place March 7â8, 2010. Laboratory Testing GSSI System Because the GSSI system involved individual antennae, a series of survey lines was marked on each test slab, and each slab was scanned by each antenna along these lines. The survey for each slab consisted of 15 parallel lines of data in the long direction, spaced 3 in. apart. The scanning process is shown in Figure 2.1. In Figure 2.1a, a small distance-encoder wheel was used to trig- ger data collection at regular distance intervals. In Figure 2.1b, the foam block was used to elevate the horn antenna above the slab surface, in lieu of a system for suspending the antenna over the slab. MALA Systems Tests were carried out with the 1.3-GHz MIRA array and with individual scans by using the 2.3-GHz antenna. Each testâs slab was surveyed with a single pass of the MIRA array. Figure 2.2a shows the MIRA setup. The size of the array was enough to cover a good percentage of the width of each slab. The wooden rails were set up on each side of the tested slab to support the wheels of the MIRA cart. In the first round of testing, three passes with a 2.3-GHz antenna pair were carried out by using the cart arrangement shown in Figure 2.2. In the second round of testing, the cart supported a four-antenna array, and a parallel series of seven profiles provided complete coverage of each slab. C h a p T e r 2
14 3d-Radar System The 3d-Radar system was used for slab testing, as shown in Figure 2.3. Note that the system is fairly large for the lab- oratory scale. Therefore, the results were more sensitive to boundary and end conditions. Also, special arrangements for supporting the wheels of the unit were made by using wooden rails outboard off the slab. Testing of Gpr Systems at Test Track First-round testing of the GSSI and 3d-Radar systems took place November 8â9, 2009, and first-round testing for the MALA system took place November 22, 2009. Second-round testing of all systems took place March 7â8, 2010. (a) (b) Figure 2.1. Scanning of slabs with GSSI equipment: (a) 2.6-GHz ground-coupled antenna pair and (b) prototype 3-GHz horn antenna. (a) (b) Figure 2.2. Slab tests with MALA equipment: (a) MIRA system and (b) a pair of 2.3-GHz antenna. Figure 2.3. 3d-Radar equipment setup for laboratory slabs.
15 GSSI Systems GSSI tested the 2.6-GHz ground-coupled antenna pair and the prototype 3-GHz horn antenna at the test track. The ini- tial tests used both systems, and the second round of tests focused on the 3-GHz horn. Initially, the tests were carried out on a series of data lines spaced at approximately 1 to 1.5 ft apart, with data-collection rates ranging from four to 12 scans per foot. Position of the data was registered by using a distance encoder mounted to the wheel of the test vehicle. The 2.6 GHz ground-coupled antenna pair was placed end-to-end on a skid plate, as shown in Figure 2.4. With this arrangement, both antennae were dragged along the ground at a speed of 2 to 3 mph. The horn antenna was suspended about 12 in. above the pavement surface by using a wooden beam for support, as shown in Figure 2.5. The alignment of the data lines was visually maintained by the vehicle driver using spacing markers painted on the pavement surface every 100 ft. The second round of testing was carried out only with the 3-GHz horn. The scanning consisted of 25 lines of data spaced laterally at 6 in., with a data rate of 12 scans per foot. This closer spacing was used to obtain more detailed resolu- tion in the subsequent imaging of the data. MALA Systems At the test track, the conditions were rainy, and the pavement was wet. Testing included a pair of 2.3-GHz ground-coupled antennae and a 16-channel MIRA system using 1.3-GHz antennae. At the test track, the 2.3-GHz antennae were attached end-to-end, and data was collected as a series of parallel survey lines spaced 1 ft apart. The MIRA system, which is about 30 in. wide, covered the full width of the test lane by using 6 overlapping passes. The MIRA data were collected with three transmit-receive antenna configurations: standard (monostatic), endfire array, and common midpoint (CMP) method. The MIRA system was deployed by using a wheeled cart and position of the data was obtained by using a total station (see Figure 2.6a). The 2.3-GHz antennae were deployed with a wheeled cart, and position was recorded by using a linear distance encoder. Figures 2.6a and 2.6b show these two systems deployed on the Figure 2.4. Deployment of the GSSI 2.6-GHz ground-coupled antenna pair. Figure 2.5. Deployment of GSSI 3-GHz horn antenna.
16 test track. The second round of tests was conducted March 7â8, 2010, under more favorable field conditions and using the same antenna systems. However, the 2.3-GHz antenna cart deployed four antennae side-by-side. 3d-Radar System In the first round of testing, 3d-Radar used a 2.3-m-wide antenna unit housing 29 antenna elements and producing 29 channels of data. The lateral coverage of this array was 2.25 m. The elements produced data by using a swept frequency with a range from 150 MHz to 2.69 GHz. A photograph of this equipment as it was deployed at the test track is shown in Fig- ure 2.7. Signal generation and data acquisition were controlled by a unit called the âGeoScope,â which was mounted on top of a tool cabinet in the bed of the test vehicle. Data location was registered by using a linear distance encoder mounted to one of the antenna support wheels. Three acquisition configura- tions were used to vary the density of the data in the x and y directions as well as the corresponding speed of data collection. Full coverage of the test section was obtained by using two overlapping longitudinal runs of the system, one on the left half of the lane and one on the right half. For the second round of testing, 3d-Radar used a smaller, 21-channel unit that had a fre- quency sweep range of 140 MHz to 3 GHz and a lateral coverage width of 1.5 m. For the second round of tests, the test area was scanned with five parallel, overlapping longitudinal runs. analysis of Gpr Data The data analysis has been presented as time-depth slices, showing amplitude variations for the multiple antenna data lines within a particular time range (slice). The time slice was (a) (b) Figure 2.6. Testing of MALA systems at the test track: (a) MIRA system and (b) 2.3-GHz antennae in a cart.
17 converted into a depth slice by using an assumed dielectric constant, which for asphalt is typically between 5 and 6. The depth slices presented by each organization were usually accompanied by supporting B-scan samples for individual lines of data. Laboratory Evaluations Round 1 laboratory testing was carried out on the slabs as described in the construction report (see this volume, Chapter 1). For Round 2 laboratory testing, water was introduced into the delaminated areas to see whether its presence would affect their detection. The results of the first- and second-round testing on Slab A are shown in Figure 2.8. Only the GSSI 3-GHz horn antenna depth slice for Round 1 was able to detect a significant anomaly in the delaminated area. However, after the water was introduced in Round 2, each of the three systems detected an anomaly in this area. Figure 2.9 shows the results obtained for Slab B. Note that in the Round 1 result, an anomaly was detected in the depth slice near the 4-in.-deep debond for both the GSSI and 3d-Radar systems. No anomaly was detected in the stripped area. This was surprising, because (a) the 4-in. debond should be harder to detect than the 2-in. debond would be and (b) the stripped area should be much easier to detect with GPR than the debond would be. This observation suggests that the property discontinuity at the 4-in. debond is more pronounced than one would expect at an interface that is simply debonded. Note also that, as with the 2-in. debond, the detection of the 4-in. debond was enhanced by the introduction of moisture. Field Evaluations at the Test Track Figure 2.10 shows the time-depth slices obtained for each GPR system in Sections 4, 5, 6, and 7 of the test track. Note that all systems were sensitive to the stripping condition located in Section 6 and to the presence of moisture in debonded areas, as noted by the standpipe locations. Other than by detecting the presence of moisture, the GPR systems were unable to detect the extensive presence of debonding at a 2-in. depth in Sections 4 and 5. Figure 2.11 shows similar time-depth slices for Sec- tions 8, 9, and 10. Once again, all systems were able to detect the stripping condition located in Section 8 at a depth of 4 in., as well as the presence of moisture in the 4-in.- deep debonded areas in Sections 9 and 10. None of the systems appeared capable of detecting extensive areas of debonding in Sections 9 and 10 where moisture was not present. (a) (b) (c) Figure 2.7. 3d-Radar equipment: (a) GeoScope, (b) Side View, and (c) Rear View.
18 Figure 2.8. Time-depth slices at 2 to 3 in. for first- and second- round testing of Slab A. Round 1 Evaluation Round 2 Evaluation North GSSI Debond @ 2" Sound Debond/Moisture @ 2"Sound boundary effect MALA Round 1 image not provided 3d-Radar -0.5 0.5 0 0 0.5 1 1.5 2 2.5 W id th (m ) -0.5 0.5 0 0 0.5 1 1.5 2 2.5 Figure 2.9. Time-depth slices at 4 to 5 in. for first- and second- round testing of Slab B. boundary effect 3d-Radar Debond at 4" Stripped @ 4" Debond/moisture at 4" GSSI Stripped @ 4" MALA Round 1 image not provided Round 1 Evaluation Round 2 Evaluation 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 0 W id th (m ) 0.5 1 1.5 2 2.5 -0.5 0.5 0 0 0.5 1 1.5 2
19 Figure 2.10. GPR time-depth slices for Sections 4, 5, 6, and 7. 3d-Radar slice at 3â Figure 2.11. GPR time-depth slices for Sections 8, 9, and 10. = standpipe location = possible overshoot of RAP 3d-Radar slice at 5.3 MALA slice at 5" GSSI slice at 5" Sect 8 Sect 9 Sect 10
