.jpg)
1. Introduction
Geologic storage of high-level radioactive waste requires a system that restricts the migration of radionuclides to the surface, and it is typically composed of the engineered barrier and the natural barrier [1,2]. The engineered barrier system includes a canister, buffer, backfill, and a near-field rock [3]. The natural barrier includes the natural soils and rocks surrounding the radioactive wastes. The storage system is located in a subsurface more than 500 m deep to ensure complete isolation from any on-surface activities. It is necessary to monitor the structure in real-time to ensure safety during the entire storage period. There are several proposed non-destructive methods which use mechanical and electromagnetic waves and monitor mechanical strains and infrared thermography [4-6].
The acoustic emission (AE) technique is a passive non-destructive testing method that monitors the state of a structure based on elastic waves generated when the material is deformed. Plastic deformation in a material releases energy and generates acoustic waves at a frequency range of 1–1,000 kHz. The AE method detects and monitors the released acoustic waves from a material. The AE method is proposed as an optimal way to monitor a damage process in brittle materials related to a nuclear waste repository such as concrete and rock [11,13].
The AE signal is analyzed in the time and frequency domains and the parameters commonly used are as follows.
-
Threshold: The center red line in Fig. 1(a) indicates the threshold value of the AE signal. It filters the signal which has less amplitude than the threshold. The engineer set this value considering environmental noise or effective analysis as dB (decibel).
-
Count: The number of times above the threshold for one waveform. In Fig. 1(a), four counts are observed. It is sometimes named “hit” and is the most fundamental parameter of the AE signal.
-
Duration: A time interval between the triggering time to the last time when it’s over than threshold. It depends on source magnitude and noise filtering.
-
Max amplitude: The highest voltage of the waveform is named max or peak amplitude. It is related to the magnitude of the source event. Therefore, it can give information about the attenuation and location of the source.
-
Rise time: A time interval between the triggering time and max amplitude of the waveform. It is related to source-time function, and crack mode or noise filtering.
-
Average frequency: This is calculated parameters from count over the duration. It is used to classify crack mode with other parameters.
-
RA value: This is calculated parameters from a rise time over max amplitude showing the reciprocal gradient of the waveform.
-
Peak frequency: A frequency feature that has the largest magnitude in the power spectrum (Fig. 1(b)).
-
Centroid frequency: A frequency feature that is calculated as a sum of magnitude times frequency over a sum of magnitude (Fig. 1(b)).
Previous studies have established the relationship between AE parameters and structural damage progress [14-17]. However, the heat emitted from a canister and the moisture from permeating groundwater can significantly alter the disposal system and change the AE characteristics. There are several previous researches that investigated the effect of various temperature ranges on the mechanical and AE properties of concretes [18002D22]. However, the study which targets the disposal systems is still limited.
Therefore, this study investigates a relationship between the AE parameters and the degree of damage in concrete subjected to environmental changes. Particular emphasis is placed on the environmental effects of storage temperature and groundwater. In this study, we collected AE signals while performing the uniaxial compression tests on concrete specimens. Here, we varied the storage temperature condition from 15℃ to 75℃ and the water-immersion condition from dry to immersed in specimen preparation. The relations between the acquired AE parameters and the damage progression were analyzed, and its implication was further discussed.
2. Experimental Setup
The concrete specimens were prepared using the silo recipe for the intermediate and low-level nuclear waste repository in Gyeongju, South Korea. Table 1 shows the details of the concrete mixing ratio. The concrete specimens were shaped with a diameter of 100 mm and a height of 200 mm representing the most common design, according to ASTM D 7012. All specimens were submerged in tap water and cured for 28 d at 25℃. After curing, the specimens were placed in ovens and water buckets set at 15℃, 45℃, and 75℃ for 32 days. A total of 18 specimens were produced under six different conditions, three specimens for each condition. The specimen was named after its curing condition, e.g., DC45 for dry-45℃ condition and IC75 for immersed-75℃. The specimen surface was carefully polished to achieve flatness of less than 0.02 mm and less than 0.001 radians for perpendicularity, following the recommendation by the International Society of Rock Mechanics (ISRM).
Fig. 2 shows specimen design with AE sensor location and a schematic diagram of mechanical and AE measurement systems. The AE sensors were placed in a spiral shape with a height interval of 20 mm, only two sensors at both ends with an interval of 30 mm, as shown in Fig. 2. This ensures the acquisition of various AE signals while minimizing the effect of the location of crack occurrence within the specimen. The uniaxial compression test (UCT) used the load frame with a servo-control system (Model UTM-97001, Myung-Do System, Korea), which has the maximum allowable force of 1,960 kN. A biaxial strain gauge (Model KFG-10-120-D16-11L1M2S, Kyowa, Japan) measured both x and y-axis strains during the whole test. The AE sensors (Model Nano30, MISTRAS, PAC, U.S.A) had a diameter of 5 mm, and these sensors had a resonant frequency between 120 and 750 kHz. The AE signals were then amplified by the preamplifier (Model ILxx-LP, MISTRAS, PAC, U.S.A) with a gain of 26 dB. In the software AE win, the threshold for the signal acquisition was set to be 35 dB. The analog filter in the data acquisition system filtered unwanted components with a bandpass filter from 20–400 kHz. The peak definition time (PDT), hit definition time (HDT), and hit lockout time (HLT) were set as 10 μs, 500 μs, and 500 μs, respectively, to improve the acquisition process.
Fig. 2
(a) Dimension and sensor locations for the specimen, (b) a schematic diagram of the testing and measurement system. All dimensions are in millimeters and the numbers in squares or circles indicate the channel number of the AE sensor.
All uniaxial compression tests were performed until the peak stress appeared, and this was determined as the yield stress. The used loading rate was 0.2 mm·s−1. For each type of experiment, three specimens were performed, and a total of eighteen specimens were loaded until failure. The acquired load and AE data were analyzed by MATLAB software. The AE parameters (i.e. count, energy, RA value, and peak frequency) were calculated. After AE signal acquisition, the data were analyzed only in channels 3–6 to consider the seating effect and signal quality.
3. Results
3.1 Uniaxial Compression Test Results
Fig. 3 shows the compilations of stress-strain responses and specimen images after failure. Table 2 summarizes the test results. In a dry condition, the yield stress ranged between 25 and 32 MPa. In an immersed condition, the yield stress ranged between 20 and 24 MPa. The immersed specimens showed a compressive strength lower than the dry specimens by ~26%. Among the dry specimens, those exposed to 75℃ showed a slightly higher compressive strength than the specimens treated at 15℃ and 45℃ only by ~1.6–1.9 MPa. On the other hand, the compressive strength of the immersed specimens treated at 75℃ was slightly lower than those treated at the other temperatures by 0.7–0.8 MPa. These results indicate that the concrete is vulnerable to groundwater exposure but there is only minimal effect of temperature on the strength.
Fig. 3
(a) Strain-stress curves obtained from the UCT tests, (b) digital photos of the failed dry specimens, and digital photos of the failed immersed specimens. The strain gauges were installed at the center of the specimen height to measure the axial and lateral strains.
3.2 AE Monitoring Results (I)
– Energy-related Parameters
Fig. 4 shows the changes in the normalized stress, normalized cumulative AE energy, and AE count over time. There was a steady increase in the AE energy until ~80% of yield stress, as shown in Figs. 4(a)–4(c). In the range of 80–100% of yield stress, the AE energy began to dramatically rise compared to the previous stage. The AE counts increased when the specimen state was close to the failure because crack growth and coalescence are most abundant in the failure stage. This is consistent with previous studies using concrete, rock, mortar, and glass fiber [15,17,23-25].
Fig. 4
Variations in the normalized stress, normalized cumulative AE energy, and AE count over time: (a–c) dry specimens treated at 15°C, 45°C, and 75°C (DC15-2, DC45-4, and DC75-4, respectively), and (d–f) immersed specimens treated at 15°C, 45°C, and 75°C (IC15-6, IC45-2, and IC75-4, respectively).
Amongst these results, the AE count results depict that the greatest AE energy was observed with the dry specimens treated at 75℃ (DC-75). In the dry specimens treated at 75℃ (DC-75), more than 90% of total AE energy was emitted after 80% of yield stress. By contrast, in the dry specimens treated at 15℃ and 45℃, more than 50% and 80% of total AE energy were emitted after 80% of yield stress, respectively. This result indicates that less AE energy is released during the early stage of the loading and more AE energy is accumulated for the later damage stage as the storage temperature increases.
The immersed specimens also showed a similar trend to the dry specimens, as shown in Figs. 4(d)–4(f). It is worth noting that the energy-related AE parameters, including AE events, AE counts, and AE energy, acquired from the immersed specimens are significantly less than those from the dry specimens. In the AE parameters, total AE energy measurement is proportional to the number of AE events and the AE signal (time and amplitude) area. Therefore, as the attenuation increases, the amplitude of the signal decreases, and as the number of signals below the threshold increases, the number of collected events and their area of them also decrease. As a result, the total AE energy becomes relatively smaller in the immersed condition than in the dry condition. In the same context, the storage temperature effect appeared minimal for the immersed specimens. In all the immersed specimens, more than 95% of the total AE energy was emitted after 80% of yield stress.
3.3 AE Monitoring Results (II)
– Frequency-related Parameters
Fig. 5 shows the changes in the peak frequency for six storage conditions. The peak frequency was mostly in the range of 80–160 kHz and 240–320 kHz. In these two ranges, the low peak frequency group was more collected than the high peak frequency group except for dry specimens treated at 75℃ (DC-75). The AE signals belonging to the high-frequency band increased with higher temperature, as shown in Figs. 5(a)–5(c).
Fig. 5
Peak frequency characteristics of the acquired AE signals: (a–c) dry specimens treated at 15°C, 45°C, and 75°C (DC15-2, DC45-4, and DC75-4, respectively), and (d–f) immersed specimens treated at 15°C, 45°C, and 75°C (IC15-6, IC45-2, and IC75-4, respectively).
The average frequency and the RA value distribution are represented in Fig. 6. The RA value magnitude differences were observed by dry and immersed conditions rather than temperature variation. The RA value was in the range of ~300,000 μs·V−1 and mostly below 50,000 μs·V−1 in dry conditions. This value in the immersed condition was in the range of ~90,000 μs·V−1 and mostly below 20,000 μs·V−1. The average frequency is in the same range of ~400 kHz for dry and immersed specimens.
Fig. 6
Distributions of average frequency versus RA value of the acquired AE signals: (a–c) dry specimens treated at 15°C, 45°C, and 75°C (DC15-1, DC45-2, and DC75-1, respectively), and (d–f) immersed specimens treated at 15°C, 45°C, and 75°C (IC15-1, IC45-2, and IC75-1, respectively).
Fig. 7 shows the change in the cumulative AE counts and the RA value over normalized stress. The RA value is shown in red marks, and it is the maximum value among the 100 signals. Therefore, the red mark increased when the higher RA value is acquired. In the dry specimens, as shown in Figs. 7(a)–7(c), and Figs. 7(d)–7(f) showed that the RA value increases more gradually than in immersed specimens. The final RA value was in the range of 200,000– 300,000 μs·V−1 and 50,000–100,000 μs·V−1 for dry and immersed specimens. The increase in max RA value increased with AE count growth. The highest RA value was acquired at over 95% of yield stress and immersed specimen’s final RA value was 32% magnitude of dry specimen results. There was no significant difference in max RA value result between temperature variations. This result means the max RA value criteria to monitor the material should be different depending on immersed and dry conditions.
Fig. 7
Cumulative AE count and RA value with respect to normalized stress: (a–c) dry specimens treated at 15°C, 45°C, and 75°C (DC15-1, DC45-2, and DC75-1, respectively), and (d–f) immersed specimens treated at 15°C, 45°C, and 75°C (IC15-1, IC45-2, and IC75-1, respectively).
Fig. 8 compares the yield stress and total AE energy of DC and IC specimens with respect to the treated temperature. Among the AE parameters, the AE energy was found to be the most relevant to the yield strength. In the case of DC, it was confirmed that the overall AE energy emission increased as the temperature increased, and especially it rapidly increased at 75℃.
Fig. 8
Comparison of the yield stress and total AE energy with respect to treated temperature and drying and immersion conditions. Note: dry specimen results are plotted as triangle markers, immersed specimen results are plotted as square markers and calculated in average value.
4. Conclusion
This study investigates the change in AE parameters according to the conditions (temperature, dryness, saturation) of the deep geological disposal system using concrete specimens. The immersed specimens showed a compressive strength lower than the dry specimens by ~26%. The AE parameters value such as AE count, AE event, AE energy, and RA value of the dry specimen tended to be higher than the immersed specimen, and in particular, the dry specimen at 75℃ showed the highest energy release. There were minimal differences in energy-related parameters between the storage temperatures in the immersed specimens. The peak frequency was mostly in the range of 80–160 kHz and 240– 320 kHz. The increase in max RA value increased with AE count growth. The highest RA value was acquired at over 95% of yield stress and immersed specimen’s final RA value was 32% magnitude of dry specimen results. Among the AE parameters, the AE energy was found to be the most relevant to the yield strength, and particularly in the dry conditions, the cumulative AE energy increased as the temperature increased, and especially it rapidly increased at 75℃. Finally, this study shows that the AE method can capture the mechanical damage process of concrete materials, but their AE characteristics can vary with the storage conditions.
