27 Evaluation of Fatigue Damage in RC Slabs Considering Water Infiltration …
329
for 100 k times for the first step. Then the load is increased to 127.4 kN and repeatedly
applied for another 100 k times. Finally, the load of 156.8 kN is applied until the
limitation of fatigue failure. The eventual load cycles in this experiment is 214,194
times. In this study, static loads are each applied just before going to the next step
i.e., after load cycles of 100 k times of 98 kN, the load is unloaded and reloaded
again to 98 kN with static manner. The same goes for the further cases of 127.4 kN
and 156.8 kN. The specimen of concrete slab is 3.0 × 2.0 × 1.6 m with a mixture
proportion as in Table 27.1. The compression strength is 20.7 N/mm
2 . A central
area of the top surface (see the bold blue rectangle in Fig. 27.4) is kept saturated
by continuous water pouring to acertain height. Strain gauges are embedded in the
specimen. Arrangement of AE sensors and excitation locations are demonstrated as
in Fig. 27.4. 36 AE sensors of 60 kHz resonance are arranged onto the specimen,
namely 10 for top, 18 for bottom and 8 for sides. For elastic wave tomography,
excitations are made by a 35 mm diameter. Steel ball for 18 points on the top as
avoiding the loading area depicted by a green broken rectangle, and 28 points on the
bottom. Measurement of elastic wave excitations are conducted three times: before
loading, after 100 k load applications, after 200 k load applications, and after failure.
Detected signals by the sensors are amplified by 40 dB at preamplifier and acquired
by the monitoring system of 48-channel Express-8 (Physical Acoustics Corp) with
a 1 MHz sampling rate.
27.4 Results and Discussions
Density of the wave paths in expected aggregation area becomes smaller because
excitation points cannot be set under the wheel guide plate. So the lower of the
resolution resulted by the elastic wave tomography. To understand this fact as well
as to know the analytical resolution of the tomography, elastic wave/ray paths formed
by combinations between excitation points and sensor positions are depicted as in
Fig. 27.5. Accordingly, to suffice the density of wave paths beneath the loading area,
AE sources obtained during the static loading applied after every wheel loading
cycles are utilized in addition to the data derived by the elastic wave tomography.
Tomograms by each step are shown in Fig. 27.6. In the case before loading
(see 0 cycles), areas of slightly small velocity of 3300 m/s are obtained, whereas
overall velocity exhibits 4400 m/s, and therefore the specimen is appeared to be
intact at this moment. In the case of 100 k cycles, the areas of small velocity less
than 3000 m/s become more distinctive and continuously emerged in the transversal
direction, resulting in the average velocity of 4000 m/s. In the case of 200 k cycles,
areas of small velocity less than 3000 m/s widely evolve to the right sides of interest,
and damage areas evaluated by the velocity less than 2700 m/s, which is concluded
by our past paper [2], are observed. As areas demonstrating large velocity of more
than 4000 m/s are still remained in places at this step, this appears to be the stage
where damage and intact areas are mixed. After the fatigue limit, middle and upper
areas are damaged widely of interest. In consideration of the velocity decrease with
329
for 100 k times for the first step. Then the load is increased to 127.4 kN and repeatedly
applied for another 100 k times. Finally, the load of 156.8 kN is applied until the
limitation of fatigue failure. The eventual load cycles in this experiment is 214,194
times. In this study, static loads are each applied just before going to the next step
i.e., after load cycles of 100 k times of 98 kN, the load is unloaded and reloaded
again to 98 kN with static manner. The same goes for the further cases of 127.4 kN
and 156.8 kN. The specimen of concrete slab is 3.0 × 2.0 × 1.6 m with a mixture
proportion as in Table 27.1. The compression strength is 20.7 N/mm
2 . A central
area of the top surface (see the bold blue rectangle in Fig. 27.4) is kept saturated
by continuous water pouring to acertain height. Strain gauges are embedded in the
specimen. Arrangement of AE sensors and excitation locations are demonstrated as
in Fig. 27.4. 36 AE sensors of 60 kHz resonance are arranged onto the specimen,
namely 10 for top, 18 for bottom and 8 for sides. For elastic wave tomography,
excitations are made by a 35 mm diameter. Steel ball for 18 points on the top as
avoiding the loading area depicted by a green broken rectangle, and 28 points on the
bottom. Measurement of elastic wave excitations are conducted three times: before
loading, after 100 k load applications, after 200 k load applications, and after failure.
Detected signals by the sensors are amplified by 40 dB at preamplifier and acquired
by the monitoring system of 48-channel Express-8 (Physical Acoustics Corp) with
a 1 MHz sampling rate.
27.4 Results and Discussions
Density of the wave paths in expected aggregation area becomes smaller because
excitation points cannot be set under the wheel guide plate. So the lower of the
resolution resulted by the elastic wave tomography. To understand this fact as well
as to know the analytical resolution of the tomography, elastic wave/ray paths formed
by combinations between excitation points and sensor positions are depicted as in
Fig. 27.5. Accordingly, to suffice the density of wave paths beneath the loading area,
AE sources obtained during the static loading applied after every wheel loading
cycles are utilized in addition to the data derived by the elastic wave tomography.
Tomograms by each step are shown in Fig. 27.6. In the case before loading
(see 0 cycles), areas of slightly small velocity of 3300 m/s are obtained, whereas
overall velocity exhibits 4400 m/s, and therefore the specimen is appeared to be
intact at this moment. In the case of 100 k cycles, the areas of small velocity less
than 3000 m/s become more distinctive and continuously emerged in the transversal
direction, resulting in the average velocity of 4000 m/s. In the case of 200 k cycles,
areas of small velocity less than 3000 m/s widely evolve to the right sides of interest,
and damage areas evaluated by the velocity less than 2700 m/s, which is concluded
by our past paper [2], are observed. As areas demonstrating large velocity of more
than 4000 m/s are still remained in places at this step, this appears to be the stage
where damage and intact areas are mixed. After the fatigue limit, middle and upper
areas are damaged widely of interest. In consideration of the velocity decrease with