**7. AE Inspection of Miscellaneous Processes**

Some reports have appeared covering various processes. Since these are rare, yet useful, we list them below.

Tank-bottom inspection by AE has been widely practiced. For this testing, Papasalouros et al. [202] presented the statistical distribution data of their many inspection results. Naturally, such outcomes depend on specific sets of tested tanks, but this information should be of value to other inspectors and tank owners.

Kim et al. [203] monitored a blast furnace for steelmaking with AE for detecting cracks in its steel shell and leaks from hot air blower pipes. Burst and continuous signals were detected, respectively, providing a baseline data for devising an SHM system.

Serreti et al. [204] monitored the cold forming of aircraft wing panels using AE and seeking optimized forming parameters. The final aim is to automate the operation to get resultant panels equivalent to those made by qualified manual processes.

Zielke et al. [205] used AE to gain understanding of thermal spraying processes as AE signals are related to coating thickness achieved and to crack density of coatings. Different spray guns and their conditions must be selected properly and AE results provided guidelines to select the processes.

Manthei et al. [206] examined fracture of automobile windshields with AE and evaluated their fracture processes under static and dynamic loading together with finite element modeling. They identified the initiation point of fracture, which is crucial in proper modeling and support design.

Ravnik et al. [207] correlated low frequency AE during quenching of steel components to quenching cracks. Reliability of hydrophone detector was established and the system is ready for real component tests.

#### **8. General Summary and Follow-Up Studies**

Signal attenuation considered in Sections 2.1–2.3 led to the confirmation of the guided wave theories through reexamination of data from large structures as well as from laboratory-scale plate experiments. These theories for attenuation in isotropic media can be applied to predict signal loss in real structures of metallic alloys. For the distance of 10 m or so, the inverse square root distance (or 1/√x) behavior is enough, while at longer propagation distances material absorption effects need to be added in the form of (1/√x)·exp(–αx). The guided wave attenuation coefficient can be calculated when reliable bulk wave attenuation coefficients are available. However, Section 2.5 revealed the lack of such data for most common alloys. For these cases, one needs to conduct attenuation tests as in [33–38]. It is hoped that more bulk wave attenuation data will be accumulated in the future, although shear wave attenuation tests are difficult.

A new approach for attenuation studies using complex elastic moduli was reviewed in Section 2.4. The damping factors for stiffness coefficients were obtained by the Castaings method. Since no direct verification is feasible, obtained damping factors for PMMA (in Section 2.4) and fiber-reinforced composites (in Section 2.6) were compared with damping factor determination from dynamic mechanical analysis and ultrasonic attenuation studies. The comparison showed Castaings' damping factors to typically be twice higher than those from other methods. Since the Castaings method was developed with sound theoretical foundation, some unaccounted factors appear to contribute to overestimation. A possible source is the common assumption made in most ultrasonic analysis, that is, the wave front to be planar. This was apparently a cause of higher bulk wave attenuation coefficients from immersion or buffer-rod methods as discussed in Section 2.5. In these ultrasonic methods, this point needs to be reevaluated. Clearly, the Castaings method is a valuable addition to viscoelastic analysis and making it verified will be beneficial to the field. Note that mechanical damping and ultrasonic attenuation studies are complementary, but these are usually separated. For example, Treviso's review [17] mentioned no complex elastic moduli works, and Castaings and Hosten [48,49] ignored low frequency damping studies. Still, an identical physical mechanism governs vibration energy loss in polymers like PMMA and epoxy [18,19].

Attenuation studies for fiber-reinforced composites considered in Section 2.6 show significant advances made in the past 15 years. Yet, the fact remains that attenuation in composites is higher than in homogeneous metallic alloys. This is especially true at higher frequencies above a few hundred kHz, where signal analysis must be conducted to characterize microscopic fracture mechanisms, as discussed in detail by Sause [208]. Even from practical engineering approach of using pattern recognition analysis, the availability of high frequency components in AE signals is an important element for success. In composite plate design, ply lay-up sequences are crucial parameters and these also affect attenuation characteristics. This part has not entered into AE or SHM consideration, but will eventually become necessary to incorporate it in attenuation analysis. Initial steps were made in examining the directivity of attenuation behavior [102,104,209], but more future works are needed, accompanied by modeling studies to avoid experimentation. Some of the past attenuation studies used a common AE or ultrasonic sensors [104,209] and this must be avoided in future works. As shown in Figure 6 above, a small aperture sensor of 1-mm diameter gives different attenuation behavior. While sensor aperture effects can be partly corrected by analysis [210], the sensor aperture must be smaller than the wavelength to minimize displacement cancellation effects. From the materials research side, high damping in newer high-strength carbon fibers and their composites is an interesting subject

since there is no obvious mechanism for such a behavior. While not touched in this review, NDE of sandwiched structures is an important field by itself. Attenuation problems are far more serious issues and AE inspection has not progressed adequately. A general review [211] and a guided wave study [212] are cited as starters.

Section 3 surveyed algorithm developments for AE source location. Various approaches have been added since Ge's reviews [108,109] and each is expected to have strong and weak points. At this stage, it is hoped that some organizations or individuals formulate a signal generator software or a set of signal waveforms that can provide the basis for standard performance evaluation of an AE source location routine. It can start from a simple data set for testing location accuracy and speed. Eventually, an advanced data set should measure absolute/relative distance errors, robustness against noise and/or outliers, level of signal-to-noise ratio that can provide a certain level of location accuracy, as well as the minimum number of sensors needed for successful location. Without this type of standardized comparison, the objectivity of algorithm performance cannot be attained.

Section 4 covered AE applications to large bridges, which have been relatively unknown among AE engineers. It is hoped that more open reporting of success and failure promotes further technical developments for this crucial part of infrastructures.

Sensors and the rest of signal processing areas were discussed in Section 5. PWAS characterization is a logical next step in sensor calibration. Optical fiber sensor developments should be watched as a technical leap should come sooner or later. AE tomography has been applied primarily to concrete structures, but masonry can be another possible object for its application.

Section 6 highlighted four AE applications to structures, with three gaining commercial success. High temperature stress rupture monitoring is just as noteworthy technical success. With increasing uses of compressed gas in public and private transport—i.e., buses, trucks, and cars—the life extension of gas tanks can be a logical next step in AE applications.

Other AE applications worthy of mention were collected in Section 7. More applications of this type rarely appear in technical journals and one needs to peruse conference proceedings of AE and related fields.
