Application of Acoustic Emission to Detect Damage in Composites Materials

Abstract

In today’s modern world, the use of composite in the construction of various equipment and parts due to many advantages such as; weight reduction, design durability, flexibility, etc., is increasing. Damages may occur unintentionally in composite materials, either during the manufacturing process or during the normal lifetime of the part. Structural health monitoring (SHM) of composite structures is an industry requirement. The acoustic emission method can be used as an effective nondestructive testing (NDT) method to continuously monitor the development of damages in composites. In this chapter of the book, it has been prepared with the aim of application of AE to detect damage in composites. For this purpose, the literature done in the field of SHM of composite structures using the AE method is reviewed. The content of this chapter shows the importance of using the AE method for SHM of composite structures.

1. Introduction

Damage can occur in parts during the manufacturing process or after completion of construction for various reasons such as; transportation, maintenance conditions, working conditions, etc. Therefore, it is necessary to ensure the health of the structure and identify the damage during the process of its construction and operation. Parts inspection methods are divided into destructive testing (DT) and nondestructive testing (NDT) categories. Compared to DT methods, NDT is a diagnostic method without introducing damage, stress or failure into the test. Because NDT methods do not damage the part, it saves time and cost for testing [1, 2, 3, 4].

Nowadays, NDT methods are very important in the industry. Many standards and technical instructions use NDT methods for a wide range of equipment and devices such as; storage tanks, pressure tanks, pressure pipes, metal structures of tall buildings, bridges, etc., have been considered mandatory [5, 6, 7, 8].

By using NDT techniques, a cost-effective tool can be accessed to test a sample for individual investigation and examination. Sometimes it takes more than one NDT test method to ensure the structural health monitoring (SHM) of a part. Therefore, it is necessary to familiar with the performance, advantages and limitations of NDT methods to ensure the success of the evaluation [9].

There are different types of NDT methods. Each test method has its own scope of application. Therefore, according to the sample material and the type of expected damage, the appropriate method should be chosen. Often, using a combination of several methods has also been helpful. NDT methods also have limitations. For example, improper part design can make NDT methods unable to detect damage. However, due to their simple and cost-effective use of conventional NDT methods, they are often considered important materials testing tools [9, 10, 11].

Various methods are used to ensure the SHM of composite structures. Several NDT techniques are used for SHM of composite, including; radiographic testing (RT), visual testing (VT) or visual inspection (VI), ultrasonic testing (UT), thermographic testing (TT), infrared thermographic testing, acoustic emission (AE) testing, acoustic-ultrasonic, electromagnetic test, optical test, penetration testing (PT) or liquid penetrant testing (LPT) and magnetic particle testing (MT). In Table 1, the advantages and disadvantages of some commonly used NDT methods are listed.

Method	Advantages	Disadvantages
VT	Cheap Easy to train Portable Minimum part preparation	Surface indications only Generally, only able to detect large flaws Possible misinterpretation of flaws
PT	Wide application range for different genders and shapes Very low cost of implementation Completely portable Direct appearance of works on the surface Simple application Regardless of the shape and dimensions of the piece Full testing of the part is possible It requires relatively little operator training. Very sensitive to surface damage Low investment cost Can be automated	Only the ability to identify surface defects Not to be used on porous, sanded and machined surfaces Application limitation for hot objects The scope of work should be available The galvanized and colored layers on the surface of the part must be removed. Cleaning work and surface operations are necessary Determining the depth of damage is rarely possible Larger cracks are not visible
MT	Low cost of implementation The ability to identify subsurface defects Direct appearance of works on the surface The ability to perform tests despite the color coating The method is fast and reliable It is possible to identify the most subtle surface defects It is almost independent of the shape and dimensions of the piece It requires relatively little operator training. The defect is directly visible	Application only for ferromagnetic materials The need for several inspection steps in different directions A demagnetization step is required High current requirement Only defects close to the surface can be proven It is not possible to assess the depth of defects Thick surface coatings should usually be removed The test range should be available
RT	Application for all materials permanent test record obtained not only for surface methods but also covered parts can be tested Transcription of the findings in the form of a film The size and shape of the damage can be seen Direct access to the test area is not necessary Especially suitable for three-dimensional damages such as; holes, heterogeneities, heat cracks and welding defects	Decrease in sensitivity with increasing thickness, high degree of skill and experience required for exposure and interpretation need full access to the part Only limited thicknesses can be tested The ability to detect damage is relative to the thickness of the radiation. Sensitive to flaw orientation Determining the depth of damage is difficult. It is a costly and time-consuming method There is a risk of radiation
UT	Immediate results Identification of surface and subsurface defects Minimal surface preparation Determining the depth of the crack Wide range of materials and thickness can be tested. Commenting on the shape, type and size of the damage is possible Fully automatic testing is possible Sensitive to very small discontinuities	High degree of skill required to set up and interpret It is difficult to visit uneven, irregularly shaped, very small and very thin materials The need for reference standards to calibrate equipment and crack characteristics Weakness in identifying linear defects parallel to the sound beam Surface must be accessible to probe Couplant usually required It is used only purposefully and for certain damages and a certain range The test range must be available (audio connection).
TT	It is fast and safe. Can identify of a defect in the part before it completely fails. Large areas can be scanned fast It does not need to contact the surface of the piece.	It only shows surface defects Need an experienced expert The results are not very accurate.
Leak testing (LT)	They are cheaper than some non-destructive testing methods such as; RT and UT. It is safe.	It only shows surface defects Need an experienced expert The results are not very accurate.
Guided wave testing (GWT)	Discovery of surface and subsurface defects in the part The test data is completely recorded. It is fast It is portable. Minimal surface preparation required	We need an experienced expert. Difficult to find small pitting defects.
Eddy current testing (ET)	Fast method Sensitivity to surface defect Can detect through surface coatings Little pre-cleaning required The penetration depth is a function of variable frequency Fully automatic testing and fully automatic evaluation are possible Portability	Only electrically conductive materials can be tested Very small damages are difficult to identify Large-scale testing is time-consuming Detection is affected by changes in shape and conductivity The damage is not directly visible It needs a lot of experience Complex designs are not inspected with accurately. Heat loss is more. No permanent record (unless automated)

Table 1.

Advantages and disadvantages of NDT methods [9, 12, 13].

2. Acoustic emission method

When a solid material is under stress, the defects in it cause the creation of high frequency sound waves. These waves are emitted in the material and they can be received by piezoelectric sensors. By analyzing these waves, it is possible to identify the type of defect, its location and severity. This process forms the basis of the AE method. Therefore, accurate and correct identification of the collected signals and their correct analysis are very useful in making important decisions regarding the continuation of a process or stopping it and taking the necessary measures.

AE is a passive method. In this method, first the elastic waves produced in the material are received. Then, they are analyzed in order to communicate between the received waves and the changes made on the source. According to the propagation of waves from the source to the surface of the material, they can be recorded by sensors and in this way information about the existence and location of the source of propagation of waves can be obtained. These waves can have frequencies up to several megahertz. In this method, frequencies usually are in the range of 150–300 kHz. The propagation of waves continues to the surface of the piece, that is, where the sensors are installed. Then, they are recorded by sensors and they are converted into electrical signals. The AE system processes these signals and turns them into information packets. Finally, statistical information such as; the characteristics and location of resources are calculated and they are displayed in the form of graphic and numerical charts to be interpreted (Figure 1) [15, 16].

Figure 1.

Creation of elastic waves due to stress in the material [14].

Among the applications of the AE method for SHM the following can be mentioned; detection of friction [17], detection of friction and the presence of wear in bearings (bearings, roller bearings) due to improper functioning of grease or lubricant, detection of impact in rotating mechanisms due to excessive looseness of the bearing that causes impact [18, 19], detection of turbulence in the inspection of tanks under it is the pressure that the presence of excessive leakage causes turbulent flow [20], detection of defects in manufactured parts [21, 22], detection of defects in welding [23] and estimation of the life of the structure [24, 25]. Also, this method can be used to reveal and locate partial voltage discharges in large transformers [26], research and investigate the properties and characteristics of materials [27], geology [28, 29] and research on micro-vibrations [30]. The advantages and disadvantages of the AE method are listed in the following Table 2 [31, 32, 33].

Advantages

Disadvantages

It is able to detect very small defects in the range of 1 μm. Inspection with this method is very fast and as a result work efficiency is very high. This method is less sensitive to the part geometry. Compared to other conventional NDT testing methods, this method has a high investment value due to its higher efficiency. Can detect damages in defects that are difficult to access with conventional non-destructive testing techniques Deep effect Location of damage is possible It is simple to use Can be conducted remotely A non-invasive method

In this method, static defects, defects that neither grow nor move can be recognized, although this limitation has been partially overcome by applying external stress. The damage should emit a sound The part must be loaded Sensitive to disturbing sounds Can be slower than other non-destructive testing techniques

Table 2.

The advantages and disadvantages of the AE method [31, 32, 33].

3. Composite defects and their detection

In today’s modern world, the use of composite in the construction of various equipment and parts due to many advantages such as; weight reduction, design durability and flexibility, etc., is increasing. The percentage of composite components in commercial aircrafts is shown in Figure 2. For example, In the 1970s, only 7% of Concorde airplanes were built with fiber-reinforced polymer (FRP) composites. Today, 80% of the Boeing 787 Dreamliner is made of FRP composites [35]. Composites are materials that are created from the combination of at least two materials with different physical and chemical properties. In this composition, one of the components plays the role of background and the other plays the role of reinforcement. Composites are produced in different compositions. Composites are widely used in various industries. Composites have many applications in various industries such as; aerospace, construction, dentistry, electricity and power plants [3].

Figure 2.

Percentage of composite components in commercial aircrafts [34].

Damages may occur unintentionally in composite materials, either during the manufacturing process or during the normal lifetime of the part [36]. Porosity, foreign bodies, incorrect fiber volume fraction due to excess or insufficient resin, debonding defects, fiber misalignment, ply misalignment, incompletely cured matrix due to incorrect curing cycle or faulty material, wavy fibers, delamination, fiber defects, contamination, improper cure, resin rich/poor, voids, cracks, missing adhesive impact damage, thermal damage, thickness variation, dimensional problem and interface integrity are examples of damage in composites [9, 37].

4. SHM of composite structures using AE method

SHM of composite structures is an industry requirement. The AE method can be used as an effective NDT method to continuously monitor the development of fatigue damage in composites [38]. Fatigue damage caused by stress or thermal cycling can affect the mechanical integrity and safety of a composite structure. Therefore, it is necessary to identify the damage early using NDT methods [39]. Diagnosing damage in composites requires the improvement of old NDT methods as well as the development of new NDT methods. The AE method basically includes the detection of stress waves with low intensity that is caused by the occurrence of damage. For example, it can refer to warping and breakage of fibers. Stress waves are formed from the strain energy that is released during a failure and are transferred to the levels that are detected by acoustic sensors [40, 41]. One of the advantages of the AE method is the ability to monitor the onset and accumulated damage in real-time, which is not possible with most other NDT methods. Another advantage of the AE method is the ability to distinguish between some types of damage and, if a number of sensors are used, to determine the location of damage events. The AE method is often used to monitor the formation and growth of fatigue damage in composites [39, 42].

Previous studies have shown that inspection of composites using the AE method can provide valuable insight into failure mechanisms that occur at different stages during fatigue life and ultimately lead to failure. This is because different fatigue failure mechanisms create different AE signals. In composite materials, if changes occur due to loading, such as; delamination, separation of resin, breaking of reinforcing fibers, etc., AE signals are produced. In addition to the above, defects such as; corrosion and leakage of ultrasonic signals are also produced. For example, corrosion on the bottom of oil tanks produces explosive signals that propagate through the oil liquid to the wall of the tank. By installing sensors on the wall, these signals can be identified. The signals created by leakage can also be explosive and continuous. Basically, explosions occur at high pressure. When the pressure is low or the current is calm, continuous signals with low amplitude and small propagation distance are created. The ability of the AE test to identify these defects has caused to expand the use of this method in the inspection of equipment such as; composite pressure tanks, storage tanks and piping systems [43].

Godin et al. used the AE method as a method to detect different types of failures in glass-polyester composite materials. The main goal of their work was to analyze the signals in order to identify different sources of emitted waves [44, 45]. Özaslan et al. identified the damage mode of composite sheets based on the AE method and digital image correlation (DIC) method [46]. Muir et al. identified the damage mechanism in composites through machine learning and acoustic propagation [47]. Fotouhi et al. investigated the observation of the initiation and growth of interlayer separation failure in composite materials using the AE method [48]. Fotouhi et al. using the AE method and fuzzy classification and wavelet transform, classified the different failure mechanisms that occur during the growth of interlayer separation in glass/epoxy samples under three-point bending [49].

In the study of Yu et al. an AE signal analyzer with a resonant circuit was designed and built to extract the specified frequency of an AE signal. The results showed that the developed AE signal analyzer generally has the same crack detection capability as a conventional AE signal analyzer, under static and dynamic tensile tests of composite materials [50]. Chang et al. found that transverse matrix cracks caused by fatigue loading of carbon/epoxy composites can be identified by detecting high-amplitude sound emissions that have a specific frequency [51]. Barile et al. found that using the characteristics of the AE method, the failure and growth of interlayer cracks in high-amplitude signals can be detected using this test [52].

Hamzeloo et al. studied the degradation mechanisms in the bending of polymer-based composite sandwich panels by the AE method. By analyzing the acoustic data, they investigated and diagnosed the damage mechanisms in polyester/glass composite sandwich panels with polyurethane foam core in the three-point bending test by applying initial attenuation resulting from impacts with different energies [53].

The research of Saeedifar et al. investigated the creation and expansion of interlayer separation in glass/epoxy multilayer composites by the AE method. The results show the optimal performance of the AE method in investigating the behavior of initiation and propagation of interlayer failure and predicting the growth curve of interlayer separation in composite samples [54]. Saeedifar et al. found that damage diagnosis, damage identification and damage localization in multi-layer composites can be detected using the features of the AE method. Also, cracking and breaking of fibers can be detected by having a much higher AE rate than matrix cracks. In some cases, it is possible to differentiate between matrix cracking, delamination and local fiber damage. However, the identification of different fatigue damage mechanisms may be difficult, because the AE signals are often produced by signals caused by friction and wear of the crack surface with each loading. With a careful analysis of the AE spectrum and good sound filter techniques, it is possible to differentiate between certain types of fatigue damage [55].

The research of Šofer et al. showed that the failure of fibers, warping, breaking and cracking of the matrix can be detected using the AE method. Matrix crack failure can be evaluated using the AE test in the frequency band signal between 150 and 200 kHz in carbon fiber composite pipes [56]. Jiang et al. found that the bending behavior and damage evolution based on three-dimensional strain in the reinforced composite can be detected using the AE test [57].

The reliability of the AE method for accurate monitoring of fatigue progress has been proven in numerous studies. Figure 3 [58] shows the relationship between the amount of AE and strain increase for a boron/epoxy composite during 1 load-controlled fatigue test. As seen in Figure 3, the strain gradually increases up to about 1.5 million cycles and then as the composite approaches the final failure, its speed increases faster with more loading. This trend is reflected by the results of AE, which is low at first, but due to the increase in the occurrence of delamination cracking (layering) and fiber breakage, it rapidly progresses toward decrease.

Figure 3.

The effect of the number of tension-tension load cycles on the increase in strain and the amount of AE of a boron/epoxy composite in load control [58].

Romhany et al. predicted the fatigue growth curve of interlayer separation in carbon/epoxy samples by the AE method [59]. The results of the study by Mousavi Nasab et al. showed the optimal performance of the AE method in determining the moment of failure in the sample repaired with a composite patch [60].

Impact loading can cause local failure in composite structures, which greatly reduces its strength and stability. These failures are usually internal and cannot be detected by visual inspection. Among the different methods of NDT, AE is of great interest for detecting failure due to its high sensitivity to the processes that stress waves create inside the part [61]. The study of James et al. also showed that the AE method is a suitable method for impact damage ascertainment in composite plates [62]. Jang et al. obtained useful information about low-speed impact events and locations with the AE inspection system and the triangulation method at different sensor distances and impact locations in a composite stiffened plate [63]. In 2014, Zarif Karimi et al. studied the monitoring of reduced strength of perforated composite parts by the AE method. According to the results, the advance rate was identified as the most important factor and the frequency range of matrix failure was 62.5–125 kHz, fiber slippage was 312.5–250 kHz and fiber failure was 375–312.5 kHz [64].

In 2014, Ammar et al. studied the mechanical behavior and AE technique in order to detect defects in sandwich panels. Two sandwich panels with different layers of porcelain as a shell and foam made of PVC with a density of 60 and 100 kg/m³. In order to create defects in the samples, the foams were subjected to different number of angular cuts, so that the density of defects in them was different. In the meantime, the AE test was also used to detect and characterize the location of defects in two types of sandwich panels during four-point bending loading [65].

In 2014, Ahmadi Najafabadi et al. analyzed and monitored the onset of failure in aluminum/composite joints using the AE method. Also, the moment of onset of failure was detected, the critical force and, accordingly, the fracture toughness were calculated. Modeling was done based on the obtained fracture toughness values [66]. In 2016, Ahmadi Najafabadi et al. investigated the damage monitoring of T3-2024 aluminum sheet repaired with patching of multi-layer metal fibers with the AE method. They divided the samples into 4 groups according to the crack angle of 0 and 45 degrees (repaired and unrepaired state). By using the total energy of AE waves, the onset and critical growth of separation have been identified [67]. In 2007, Wu and Choi used AE with the aim of studying the analysis of the fracture process in slotted multilayer composites with different layers. Their results showed that the final path of the original crack coincides with the orientation of the fibers [68].

Oskouei et al. investigated interlaminar fracture toughness in polyester/glass multilayers subjected to mode I loading by the AE method [69]. Silversides et al. used the AE method to determine the interlaminar fracture toughness in carbon/epoxy composites subjected to mode I, II and combined mode I and II loading [70]. Allagui et al. investigated the gradual degradation and evolution of damage in the bio-composites during experimental tests using the AE method [71]. For damage area real-time localization of carbon-fiber-reinforced plastics (CFRP) composite laminates, a method based on deep convolutional neural network (CNN) and transfer learning presented using AE signals by Zhao et al. that study showed that the proposed damage area localization method can be applied for SHM of intelligent perception of composite materials [72].

LZ complexity was used to identify the damage modes in a plain weave fabric CFRP specimens tested at an elevated temperature by Barile et al. [73]. This study showed that delamination or debonding can be detected by the AE signals with LZ complexity above 0.6. Open-hole plain woven composites (OHPWCs) under tensile load for damage mode identification, damage initiation detection and damage evolution were analyzed by Liu et al. [74]. Damage participation rate (DPR) based on AE count to characterize the contribution of different damage modes to the overall failure of the plain woven composites presented in this study.

Rubio-González et al. applied the synergistic combination of AE and self-sensing capability provided by the integration of carbon nanotube (CNT) networks to SHM of glass fiber epoxy composites under flexural loading. Damage mechanisms such as matrix cracking, fiber/matrix debonding, delamination and fiber breakage can be identified by this method [75].

Wang et al. applied AE to SHM of the three-point bending process of 3D_C/C_TiC_Cu composites and evaluated the damage change of the materials under different bending stress. The result showed that the AE method is a useful tool for investigating damage evolution of 3D_C/C_TiC_Cu composites [76]. Kucukkalfa et al. applied AE test for damage detection of carbon nanotube (CNT)/ cellulose nanocrystal (CNC)-reinforced foam-cored sandwich composites under flexural load. The test results showed that reinforcements can affect the retardation and/or elimination of core damage, face sheet-core debonding, matrix cracking and fiber breakage in the sandwich composite structures. With the addition of 0.1 wt.% CNT caused the strengthen of core material, as a result, the ratio of AE signals related to fiber breakage and core damage was decreased [77].

Kalteremidou et al. used AE to identify the dominant stress/strain component in carbon/epoxy composite materials even before damage mechanisms became evident. The result showed that AE can applied for identification of the dominant stress/strain component at early loading steps [78]. Yang et al. applied AE to identify the damage mechanism of continuous alumina fiber-reinforced alumina matrix composites (Al₂O₃f/Al₂O₃) composites. The results showed that numerous high-energy AE signals were generated with a decreasing of initial load [79]. Wu and Pei applied AE to SHM of carbon fiber-reinforced laminate composites with torsional loads. The test results showed that combining the AE technique, micro-CT and SEM can be a useful method for SHM of composite structures under torsion with different off-axis angle structures [80]. A summary of SHM of composite structures using the AE method is given in Table 3.

Damage modes	Materials	The utilized AE parameters	Reference
Matrix damage and delamination	Pultruded fiber-reinforced composite (PFRC)	Rise time, counts, energy, endurance, amplitude, peak frequency	Jiang et al. [57]
Delamination	Unidirectional Cytec AS4/5276-1 carbon epoxy prepeg	Energy	Silversides et al. [70]
Interlaminar fatigue crack growth	MWCNT/carbon fiber reinforced hybrid composites	Energy	Romhany and Szebényi [59]
Weak adhesion	Carbon fiber reinforced polymer (CFRP)	Cumulative energy	Teixeira de Freitas et al. [81]
Voids	Glass-fiber-reinforced-polymers (GFRP)	Energy	Kosmann et al. [82]
Impact damage	CFRP	Energy	James et al. [83]
Delamination	Glass/polyester composites	Energy	Oskouei et al. [69]
Delamination	Glass fiber/epoxy composite	Sentry function and frequency	Fotouhi & Najafabad [84]
Porosity and fiber waviness	Fiber reinforced composites	Cumulative energy and counts	Qamhia et al. [85]
Debonding	Fiber reinforced polymer (FRP)- and steel reinforced grout (SRG)	Cumulative energy	Verstrynge et al. [86] & [87]
Crack growth	Aluminum sheet repaired with fiber metal laminate patch	Energy	Ahmadi Najafabadi et al. [67],
Debonding	A methacrylate-based universal hybrid composite (Filtek-Z250), a flowable composite (Filtek-Z350 flowable) and a silorane-based composite (Filtek-P90) were investigated.	Cumulative events	Cho et al. [88]
Matrix crack, fiber de-bonding, fiber breakage and foam crack	Foam/glass-polyester sandwich panels	Energy	Hamzeloo et al. [53]
Matrix cracking	Graphite/epoxy	Energy	Prosser et al. [89]
Fiber cracking	Al3Ni fiber reinforced composite	Energy	Harris [90]
Crack	Glass-fiber/epoxy laminates	Events	Wu and Choi [68]
Crack	Fiber reinforced composite	Event count and energy	Yu et al. [50]
Delamination	CFRP	Events, cumulative energy and cumulative events	Barile [73]
Interfacial debonding, matrix cracking, fiber breakage and core failure	Sandwich composite with glass/epoxy skin and foam core	Cumulative counts	Pashmforoush et al. [91]
Fiber breakage	Carbon/epoxy	Events	Chou et al. [92]
Matrix cracking, fiber/matrix, debonding, delamination and fiber breakage	Glass fiber epoxy	Cumulative energy and strength cumulative	Rubio-González et al. [75]
Delamination	Al2O3f/Al2O3	Events, cumulative energy	Yang et al. [79]

Table 3.

SHM of composite structures using the AE method.

5. Conclusion

SHM of composite structures is an industry requirement. AE method can be used as an effective NDT method to continuously monitor the development of damages in composites. In this chapter of the book, it has been prepared with the aim of application of AE to detect damage in composites. For this purpose, the literature done in the field of SHM of composite structures using the AE method is reviewed. AE is an NDT method to diagnose damage in composites. The capabilities and limitations of this method are described in this chapter. Previous studies have shown that inspection of composites using the AE method can provide valuable insight into failure mechanisms that occur at different stages during fatigue life and ultimately lead to failure.

Conflict of interest

The author declares no conflict of interest.