《Handbook of nondestructive testing of concrete：Acoustic Emission Methods.pdf》由会员分享，可在线阅读，更多相关《Handbook of nondestructive testing of concrete：Acoustic Emission Methods.pdf（17页珍藏版）》请在三一文库上搜索。
1、16-1 0-8493-1485-2/04/$0.00+$1.50 2004 by CRC Press LLC 16 Acoustic Emission Methods 16.1Introduction 16-1 16.2Historical Background 16-2 16.3Theoretical Considerations. 16-3 16.4Evaluation of Acoustic Emission Signals. 16-4 16.5Instrumentation and Test Procedures 16-6 16.6Parameters Affecting Acous
2、tic Emissions from Concrete . 16-8 The Kaiser Effect Effect of Loading Devices Signal Attenuation Specimen Geometry Type of Aggregate Concrete Strength 16.7Laboratory Studies of Acoustic Emission 16-9 Fracture Mechanics Studies Type of Cracks Fracture Process Zone (Crack Source) Location Strength vs
3、. Acoustic Emission Relationships Drying Shrinkage Fiber Reinforced Cements and Concretes High Alumina Cement Thermal Cracking Bond in Reinforced Concrete Corrosion of Reinforcing Steel in Concrete 16.8Field Studies of Acoustic Emission 16-14 16.9Conclusions . 16-14 Acoustic emission refers to the s
4、ounds, both audible and subaudible, that are generated when a material undergoes irreversible changes, such as those due to cracking. Acoustic emissions (AE) from concrete have been studied for the past 30 years, and can provide useful information on concrete properties. This review deals with the p
5、arameters affecting acoustic emissions from concrete, including discussions of the Kaiser effect, specimen geometry, and concrete properties. There follows an extensive discussion of the use of AE to monitor cracking in concrete, whether due to externally applied loads, drying shrinkage, or thermal
6、stresses. AE studies on reinforced concrete are also described. While AE is very useful laboratory technique for the study of concrete properties, its use in the fi eld remains problematic. 16.1Introduction It is common experience that the failure of a concrete specimen under load is accompanied by
7、a considerable amount of audible noise. In certain circumstances, some audible noise is generated even before ultimate failure occurs. With very simple equipment a microphone placed against the specimen, an amplifi er, and an oscillograph subaudible sounds can be detected at stress levels of perhaps
8、 50% of the ultimate strength; with the sophisticated equipment available today, sound can be detected at much lower loads, in some cases below 10% of the ultimate strength. These sounds, both audible and subaudible, are referred to as acoustic emission. Sidney Mindess University of British Columbia
9、 16-2Handbook on Nondestructive Testing of Concrete: Second Edition In general, acoustic emissions are defi ned as “the class of phenomena whereby transient elastic waves are generated by the rapid release of energy from localized sources within a material.”1 These waves propagate through the materi
10、al, and their arrival at the surfaces can be detected by piezoelectric trans- ducers. Acoustic emissions, which occur in most materials, are caused by irreversible changes, such as dislocation movement, twinning, phase transformations, crack initiation, and propagation, debonding between continuous
11、and dispersed phases in composite materials, and so on. In concrete, since the fi rst three of these mechanisms do not occur, acoustic emission is due primarily to: 1. Cracking processes 2. Slip between concrete and steel reinforcement 3. Fracture or debonding of fi bers in fi ber-reinforced concret
12、e 16.2Historical Background The initial published studies of acoustic emission phenomena, in the early 1940s, dealt with the problem of predicting rockbursts in mines; this technique is still very widely used in the fi eld of rock mechanics, in both fi eld and laboratory studies. The fi rst signifi
13、cant investigation of acoustic emission from metals (steel, zinc, aluminum, copper, and lead) was carried out by Kaiser.2 Among many other things, he observed what has since become known as the Kaiser effect: “the absence of detectable acoustic emission at a fi xed sensitivity level, until previousl
14、y applied stress levels are exceeded.”1 While this effect is not present in all materials, it is a very important observation, and it will be referred to again later in this review. The fi rst study of acoustic emission from concrete specimens under stress appears to have been carried out by Rsch,3
15、who noted that during cycles of loading and unloading below about 70 to 85% of the ultimate failure load, acoustic emissions were produced only when the previous maximum load was reached (the Kaiser effect). At about the same time, but independently, LHermite4,5 also measured acoustic emission from
16、concrete, fi nding that a sharp increase in acoustic emission coincided with the point at which Poissons ratio also began to increase (i.e., at the onset of signifi cant matrix cracking in the concrete). In 1965, however, Robinson6 used more sensitive equipment to show that acoustic emission occurre
17、d at much lower load levels than had been reported earlier, and hence, could be used to monitor earlier microcracking (such as that involved in the growth of bond cracks in the interfacial region between cement and aggregate). In 1970, Wells7 built a still more sensitive apparatus, with which he cou
18、ld monitor acoustic emissions in the frequency range from about 2 to 20 kHz. However, he was unable to obtain truly reproducible records for the various specimen types that he tested, probably due to the diffi culties in eliminating external noise from the testing machine. Also in 1970, Green8 repor
19、ted a much more extensive series of tests, recording acoustic emission frequencies up to 100 kHz. Green was the fi rst to show clearly that acoustic emissions from concrete are related to failure processes within the material; using source location techniques, he was also able to determine the locat
20、ions of defects. It was this work that indicated that acoustic emissions could be used as an early warning of failure. Green also noted the Kaiser effect, which suggested to him that acoustic emission techniques could be used to indicate the previous maximum stress to which the concrete had been sub
21、jected. As we will see below, however, a true Kaiser effect appears not to exist for concrete. Nevertheless, even after this pioneering work, progress in applying acoustic emission techniques remains slow. An extensive review by Diederichs et al.9 covers the literature on acoustic emissions from con
22、crete up to 1983. However, as late as 1976, Malhotra10 noted that there was little published data in this area, and that “acoustic emission methods are in their infancy.” Even in January, 1988, a thorough computer-aided search of the literature found only some 90 papers dealing with acoustic emissio
23、ns from concrete over about the previous 10 years; while this is almost certainly not a complete list, it does indicate that there is much work to be carried out before acoustic emission monitoring becomes a common technique for testing concrete. Indeed, there are still no standard test methods whic
24、h have even been suggested for this purpose. Acoustic Emission Methods16-3 16.3Theoretical Considerations When an acoustic emission event occurs at a source with the material, due to inelastic deformation or to cracking, the stress waves travel directly from the source to the receiver as body waves.
25、 Surface waves may then arise from mode conversion. When the stress waves arrive at the receiver, the transducer responds to the surface motions that occur. It should be noted that the signal captured by the recording device may be affected by the nature of the stress pulse generated by the source,
26、the geometry of the test specimen, and the characteristics of the receiver, making it diffi cult to interpret the recorded waveforms. Two basic types of acoustic emission signals can be generated (Figure 16.1): Continuous emission is “a qualitative description of the sustained signal level produced
27、by rapidly occurring acoustic emission events.”1 These are generated by events such as plastic deformations in metals, which occur in a reasonably continuous manner. Burt emission is “a qualitative description of the discrete signal related to an individual emission event occurring within the matria
28、l,”1 such as that which may occur during crack growth or fracture in brittle materials. These burst signals are characteristic of the acoustic emission events resulting from the loading of cementitious materials. Acoustic emissions from concrete occur over a very wide range of frequencies. The earli
29、est work concentrated on rather low frequencies. Robinson6 recorded acoustic emissions mainly at two frequen- cies: 2 kHz and 13 to 14 kHz; Wells7 worked in the frequency range of 2 to 20 kHz; and Green8 recorded emissions only up to 100 kHz. More modern instrumentation, however, can record much hig
30、her frequencies, typically in the range of 50 kHz to about 2 MHz. At lower frequencies, extraneous background noises from the test equipment of the laboratory environment become a problem; this was the diffi culty faced in the earlier investigations referred to above. On the other hand, at very high
31、 frequencies, the attenuation of the signals is too severe, and thus, the distance from the piezoelectric transducer to the acoustic emission source must be reduced. The precise frequency range that is monitored does not appear to be very important for concrete. Detailed studies by Tanigawa et al.11
32、 in the frequency range up to 400 kHz showed that at low stresses, emissions tended to be in the frequency range below 150 kHz; at higher stresses, the higher frequency components become more signifi cant. However, the relative shapes of the acoustic emission output vs. load curves were much the sam
33、e for all of the frequency ranges studied. FIGURE 16.1 The two basic types of acoustic emission signals. (A) Continuous emission. (B) Burst emission. A B 16-4Handbook on Nondestructive Testing of Concrete: Second Edition 16.4Evaluation of Acoustic Emission Signals A typical acoustic emission signal
34、from concrete is shown in Figure 16.2.12 However, when such acoustic events are examined in much greater detail, as shown in Figure 16.3,13 the complexity of the signal becomes even more apparent; the scatter in noise, shown in Figure 16.3, makes it diffi cult to determine exactly the time of arriva
35、l of the signal; this means that very sophisticated equipment must be used to get the most information out of the acoustic emission signals. In addition, to obtain reasonable sensitivity, the acoustic emission signals must be amplifi ed. In concrete, typically, system gains in the range of 80 to 100
36、 decibels (dB) are used. There are a number of different ways in which acoustic emission signals may be evaluated. Acoustic Emission Counting (ring-down counting) This is the simplest way in which an acoustic emission event may be characterized. It is “the number of times the acoustic emission signa
37、l exceeds a preset threshold during any selected portion of a test,”1 and is illustrated in Figure 16.4. A monitoring system may record: 1. The total number of counts (e.g., 13 counts in Figure 16.4). Since the shape of a burst emission is generally a damped sinusoid, pulses of higher amplitude will
38、 generate more counts. 2. The count rate. This is the number of counts per unit of time; it is particularly useful when very large numbers of counts are recorded. 3. The mean pulse amplitude. This may be determined by using a root-mean square meter, and is an indication of the amount of energy being
39、 dissipated. Clearly, the information obtained using this method of analysis depends upon both the gain and the threshold setting. Ring-down counting is affected greatly by the characteristics of the transducer, and the geometry of the test specimen (which may cause internal refl ections) and may no
40、t be indicative of the nature of the acoustic emission event. In addition, there is no obvious way of determining the amount of energy released by a single event, or the total number of separate acoustic events giving rise to the counts. FIGURE 16.2 A typical acoustic emission signal from concrete.
41、(From Berthelot, J.M. et al., private communication, 1987. With permission.) Arrival of the longitudinal waves Threshold 200 mV 1 Volt Duration Peak amplitude 200 s Acoustic Emission Methods16-5 Event counting Circuitry is available which counts each acoustic emission event only once, by recognizing
42、 the end of each burst emission in terms of a predetermined length of time since the last count (i.e., since the most recent crossing of the threshold). In Figure 16.4, for instance, the number of events is three. This method records the number of events, which may be very important, but provides no
43、 information about the amplitudes involved. Rise time This is the interval between the time of fi rst occurrence of signals above the level of the background noise and the time at which the maximum amplitude is reached. This may assist in deter- mining the type of damage mechanism. Signal duration T
44、his is the duration of a single acoustic emission event; this too may be related to the type of damage mechanism. Amplitude distribution This provides the distribution of peak amplitudes. This may assist in iden- tifying the sources of the emission events that are occurring. Frequency analysis This
45、refers to the frequency spectrum of individual acoustic emission events. This technique, generally requiring a fast Fourier transformation analysis of the acoustic emission waves, may help discriminate between different types of events. Unfortunately, a frequency analysis may sometimes simply be a f
46、unction of the response of the transducer, and thus reveal little of the true nature of the pulse. Energy analysis This is an indication of the energy released by an acoustic emission event; it may be measured in a number of ways, depending on the equipment, but it is essentially the area under the
47、amplitude vs. time curve (Figure 16.4) for each burst. Alternatively, the area under the envelope of the amplitude vs. time curve may be measured for each burst. FIGURE 16.3 Typical view of an acoustic emission event as displayed in an oscilloscope screen. (Adapted from Maji, A. and Shah, S.P., Exp.
48、 Mech., 26, 1, 1988, p. 27.) Trigger level 4000 points, 2.0 ms 400 points, 0.2 ms 40 points, 0.02 ms NoiseSignal 0.00.150.05 0.0 1.5 0.5 0.00.0150.005 Rise Time Standard deviation of noise () First arrival Mean 0.5 ms 0.35 ms Voltage 1.5 ms 16-6Handbook on Nondestructive Testing of Concrete: Second
49、Edition Defect location By using a number of transducers to monitor acoustic emission events, and deter- mining the time differences between the detection of each event at different transducer positions, the location of the acoustic emission event may be determined by using triangulation techniques. Work by Maji and Shah,13 for instance, has indicated that this technique may be accurate to within about 5 mm. Analysis of the wave-form Most recently, it has been suggested14,15 that an elaborate signals processing technique (deconvolutio