Figure 1 also shows illustrations of sensor installations with one AE sensor and three AE sensors. In the latter case, one sensor is installed at the bottom of the U rib, two more sensors were installed at the left and right side of the U rib respectively to perform crack source location. A 16 channel smart AE monitoring system MicroII made by Physical Acoustics Corporation was used for continuous monitoring of the fatigue test process. The fatigue test lasted for 30 days, the AE sensors were not installed until 7 days or 504 thousands load cycles of the fatigue test. 4. Behavior of acoustic emission in fatigue crack development During the 30 day fatigue test, visual inspections on the deck were conducted from time to time in each day in order to find out if there is crack initiation or crack growing. Specifically, the visual inspection is focused on a). areas that AE has indicated crack activity so that to find out if the observed crack matches with the AE indication; b). if there is a crack in somewhere that AE hasn’t picked it up before it is visualized. In the end of the 30 day fatigue test, the orthotropic steel bridge deck was thoroughly inspected and cracks were found in 12 locations. Among them, 10 of the locations are at the intersections of the U ribs and the lateral webs where AE sensors were installed. 2 of the locations are not in the web area, but in the middle of two webs with crack direction along with the rib length. At all 10 locations where the AE sensors were installed nearby, crack initiations were found by AE well before the visual inspection. However, since there was no AE sensors installed in middle of the webs, the 2 cracks located in the midway of the U ribs were not found by AE because the AE sensors were too far from the crack positions so that the AE signal was greatly attenuated before it reaches to the AE sensor. In order to illustrate how AE behaviors during crack initiation and development, the AE data in the two of the seven crack locations where AE sensors were installed were analyzed here. The two crack positions are labeled as C and G as shown in Figure 2. 4.1 Case study 1: AE activities at position C Take the position C as the first example, the elapsed time and the fatigue cycles along with AE monitoring and visual inspection results at location C are shown in Table 1. As is seen in the table, crack initiation was observed about 142.3 fatigue hours or 513 thousand fatigues cycles by the AE. The crack was not visually found until 182.6 fatigue hours or 657.5 thousand fatigue cycles when a 20 mm crack was observed. So, in this example, the AE was able to report the crack initiation in about 40 hours before visual inspection. On the other hand, the crack initiation and development process can be clearly identified from a time history graph of AE hit activity as shown in Figure 4(a). In this figure, the vertical axis
RkJQdWJsaXNoZXIy MjM0NDE=