Use of Magnetic Tomography Technology to Evaluate Dowel Placement
Verifying the accuracy and reliability of MIT Scan-2 was one of the main objectives of this study. The manufacturer-specified measurement tolerances for MIT Scan-2 are as follows:
- Repeatability: 2 mm (0.08 in.)
- Horizontal and vertical alignment: ±4 mm (0.16 in.)
- Side shift: ±8 mm (0.31 in.)
- Depth (cover): ±4 mm (0.16 in.)
The above values are the observed maximum variations, not standard deviations. The standard deviations are approximately one-third of the values listed above. Tests were conducted to verify these values and also to determine the conditions under which the specified levels of accuracy are valid.
The following series of tests was conducted:
- Repeatability - tests to determine the magnitude of random error on repeated testing.
- Effects of cover material and testing conditions - Because MIT Scan-2 operates on an electromagnetic field, the presence or absence of nonconducting material does not affect the results. Tests were conducted to verify that the test results are independent of the cover material and presence of water.
- Operating range - tests to determine the range of testing conditions under which MIT Scan-2 provides accurate results.
Based on the test results, the overall standard deviation of the measurement error and the confidence interval of MIT Scan-2 results were determined.
Most of the laboratory testing was conducted at the MnRoad facility. MnRoad constructed a slab with slots and simulated joints cut into the test slab (Figure 7) for this study. The dowels were placed in the slots at various depths and orientation. Measurements were then taken using MIT Scan-2 and compared to manual measurements. The slots provided a stable platform for holding the dowels in place during hand measurements and during MIT Scan-2 testing (Figure 8). For example, shifting the rail, rather than moving the dowel bar, ensured that the orientation of the dowel bar remained constant throughout the series of testing that was conducted to evaluate the effects of side shift.
Figure 7. Saw cut being made on test slab to simulate a joint at the MnRoad facility.
Figure 8. Dowel bar wedged in a slot of the test slab at the MnRoad facility.
Repeated measurements were taken with the test sample fixed in place to verify the reproducibility of the MIT Scan-2 test results. Both the rail and the test sample were left undisturbed through each series of tests. For each series, 10 repetitions of measurements were taken. The results of these tests are summarized in Table 1. As reported by MIT GmbH, the range of random variations in the test results does not exceed 2 mm (0.08 in.) on repeated testing.
|Series||Trial||Misalignment, mm||Side shift, mm||Depth, mm|
Because MIT Scan-2 operates on an electromagnetic field, the presence or absence of nonconducting material does not affect the results. This characteristic of MIT Scan-2 has several important practical implications, including the following:
- Validation testing can be conducted in open air, which greatly facilitates the ability to verify MIT Scan-2 results with manual measurements.
- The presence of water on the pavement surface does not affect the results. Therefore, testing can be conducted in the rain, if needed.
- The changing moisture content in concrete, as the concrete cures, will not affect the test results. Testing can be conducted on concrete at any age, including concrete during its plastic stage.
Several tests were conducted to verify that MIT Scan-2 test results are unaffected by the cover material and the presence of water. The dowel slots in the test slab were filled with aggregate and/or water for this test, shown in progress in Figure 9. The results are summarized in Table 2 for the test using the aggregate cover and in Table 3 for the presence of water. The results are within the repeatability error of MIT Scan-2.
Further evidence that the cover material does not affect MIT Scan-2 results is provided in the results of a field test conducted by the Ontario Ministry of Transportation, in which pavement joints were exposed after scanning to verify MIT Scan-2 results (Figure 10). The results from this test are summarized in Table 4.
Figure 9. Dowel covered with aggregate for the evaluation of the effects of cover material.
|Cover||Misalignment, mm||Side shift, mm||Depth, mm|
|Trial||Wet/Dry||Misalignment, mm||Side Shift, mm||Depth, mm|
Figure 10. Exposed dowel bars for validation of MIT Scan-2 results.
|Joint||Bar No.||Depth, mm||Horizontal misalignment, mm||Vertical misalignment, mm||Side Shift, mm|
*M = manual measurement; D = difference
The relatively large error in the depth results may be attributable to the effects of variations in the composition of the dowel bar metal. The variation in dowel length also affects the depth results, but dowel bar lengths typically do not deviate by more than 6 mm (0.2 in.) from the nominal length, and the resulting error from this source is less than 1 mm (0.04 in.). According to MIT GmbH, variations in metal composition can cause more significant errors. The key factor is the carbon content, which may not be controlled in low-cost steel and may vary from steel mill to steel mill. The error in absolute depth has a negligible effect on the alignment results, because the alignment results depend mainly on the ratio of the signal amplitudes, not the absolute signal strength. Hardware and software enhancements included in the MIT Scan-2 compensate automatically for the effects of variations in metal composition.
(This section continued on following page.)