J Nondestruct Eval (2015) 34:10 DOI 10.1007/s10921-015-0283-y

Comparative Study of State of the Art Nondestructive Testing Methods with the Local Acoustic Resonance Spectroscopy to Detect Damages in GFRP Christoph Hornfeck1 · Christian Geiss2 · Markus Rücker3 · Christian U. Grosse2

Received: 5 September 2014 / Accepted: 23 March 2015 © Springer Science+Business Media New York 2015

Abstract This paper evaluates and compares the application of current state of the art methods and the new local acoustic resonance spectroscopy (LARS) method for nondestructive evaluation of damages in glass fiber reinforced polymers. The innovation of the LARS is the combination of the analysis of the acoustic signals and the force excitation. Generic plates of a standardized material (Vetronit EGS 619) and segments of rotor blades of wind turbines were tested. The generic specimens, 2 and 6 mm in thickness, were damaged with various impact energies caused by a spherical impactor with a diameter of 16 mm, which generated impact damages ranging from barely visible to clearly visible on the generic specimen as well as on segments of real rotorblades of windturbines. The impacts have been measured to account for damage diameter, form and area, indentation depth and bulge height. In addition, blind holes of different depths have been drilled to assess the depth of penetration of the methods tested. As observed in scientific literature as well as in current research, impact damages exhibit a peanutshaped damage area when impacted with minimum threshold energy. The current research tested several of the specimens using X-ray computed tomography as a reference measurement. These results were compared to the data obtained by ultrasonic methods, LARS and optical lock-in thermography. Finally, all methods have been applied to evaluate rotor

B

Christoph Hornfeck [email protected]

1

Neue Materialien Bayreuth GmbH, Gottlieb Keim Straße 60, 95448 Bayreuth, Germany

2

Lehrstuhl für zerstörungsfreie Prüfung, Technische Universität München, Baumbachstraße 7, 81245 Munich, Germany

3

IABG mbH, Tests and Analyses, Einsteinstraße 20, 85521 Ottobrunn, Germany

blades of wind turbines. The results are shortly discussed in respect to practical applications and accuracy. Keywords Ultrasonic testing · Local acoustic resonance spectroscopy · Computed tomography · Impact damage · Glass fiber reinforced polymers · Lock-in thermography

1 Introduction Among the most widely used fiber reinforced composites are polymers with glass and carbon fiber reinforcement. Fiber reinforced polymers offer a wide range of advantages over conventional materials, especially for lightweight construction. However, non-visible damages have a greater influence on the residual properties of such composites as opposed to metal-based materials [1], which requires the use of nondestructive inspection techniques. Defects can emerge during the manufacturing process, by mechanical loads during service life (e.g. impacts), or other exposures. In this paper, generated impact damages, real damages on rotor blades of wind turbines and blind holes are investigated. Studies have pointed out local acoustic resonance spectroscopy to be a capable method for detecting flaws in fiber reinforced structures. However, purchasable systems are not yet available, and while the acoustic evaluation is promising, it is still the subject of ongoing research. Concerning the failure of fiber reinforced polymers there are two processes that are of particular interest [2]. The first field of interest is that of prediction models and the characteristics of the impact damage [3–6]. The second field of interest is the ability to detect, localize and characterize impact damages using nondestructive testing (NDT) methods [2,7,8]. In order to test the quality of NDT methods, impact damage characteristics were analyzed; the results are dis-

123

10

Page 2 of 14

cussed in this paper. Visible haziness, indentation depth and bulge height were measured. In addition, several specimens were examined with Computer Tomography to gather comprehensive information on damage characteristics. The test samples were examined using ultrasonic techniques, optical lock-in thermography (OLT) and local acoustic resonance spectroscopy (LARS). LARS showed – despite its low technological readiness level—its aptitude for widespread areas. On the other hand, ultrasound testing provided thorough information on depth and area of the defect. Nevertheless, the application on large structures such as wind turbine rotor blades is fairly tedious. Finally, OLT returned an easy-tointerpret image. However the test is time consuming and the excitation reduces the handiness of the method. Further methods of nondestructive evaluation which can be applied for fiber reinforced polymers, such as shearography or Terahertz techniques are not treated in this paper. Furthermore, the results of the comparison are used to discuss a holistic approach to test and monitor rotor blades of wind turbines.

2 Experimental Setup

J Nondestruct Eval (2015) 34:10

sen impactor was circular with a diameter of 16 mm. The impact energy was controlled by a position sensor placed next to the fall line, giving an analogue tension output which was digitalized using a sample frequency of approximately 2 kHz, and was analyzed to determine the speed at the time the impactor hit the surface representing the impact energy. Low velocity damage mechanisms can be assumed as the velocity was below 5 m/s at all times. Figure 2 shows a sketch and photos of the impactor and the two different mountings of the specimens. The indentation depth and the bulge height of the impact were initially measured with a dial indicator. A square with a 30 mm edge length was drawn around the impact. The difference from every side to the center of the impact was measured. The mean value of the measurements yields the indentation depth or the bulge height if measured on the backside (Fig. 3). Compared to opaque carbon fiber reinforced polymers, translucent glass fiber reinforced plastic offers insight on the extent of the damage by visual examination [9]. Examining the specimens under strong light is a common investigative technique [10]. Here, all impact damages were documented photographically, and the area of the haziness was measured by image processing software.

2.1 Description of Impact Test Rig and Specimens 2.2 Description of Impact Damage To ensure reproducibility, the studies were executed with a standardized material of woven glass fiber reinforced polymers (Vetronit EGS 619). This high pressure laminate consists of warps and wefts with a ratio of 17 to 8 in an epoxy matrix. Reference plates with blind holes with a diameter of three and five millimeters were generated. The purpose of the blind holes was to have a reference boundary layer in a defined depth to examine the penetration of different NDT methods. Figure 1 shows the specimens, the reference plates and the wind turbine rotor blade segments. Impact damages were generated by an impact tower which releases a weight onto the specimens. The shape of the cho-

The four modes of failure postulated by Richardson [21] were observed. Low impact energies left a round indentation in the specimens. This indentation was displayed by blurring of the material and is described by the first mode of failure as “matrix mode”. The specimens with 2 mm thickness start to develop peanut shaped damage characteristics at an impact energy of approximately 8 J (Fig. 4). The name peanut shape refers to the form of the damage and was introduced by Liu [10] and other authors [1,6]. Peanut defects include both delamination and fiber breaks as seen in the computed tomography

Fig. 1 Specimens of different thicknesses and sizes and segments of rotor blades

123

J Nondestruct Eval (2015) 34:10

Page 3 of 14

10

Fig. 2 Impact tower and mounting of specimens

Fig. 3 Measurement of indentation depth and bulge height

Fig. 4 Photos of impacts in specimens of 2 and 6 mm thickness

123

10

Page 4 of 14

J Nondestruct Eval (2015) 34:10

Fig. 5 a Linear rise of visible area of haziness with higher impact energies in a 2 mm thick specimen. The dependence of the extent of the damage of the clamping is visible. b Rise of visible area of haziness with higher impact energies of 6 mm thick specimen

images. For the specimens with 6 mm thickness, this characteristic is not as pronounced as for the thinner specimens. Bigger deteriorations can be observed at approximately 80 J. The material described here is relatively resistant to delamination caused by impact, as opposed to the results presented by Chong [11]. Figure 4 illustrates the results for impacts on plates with a thickness of 2 and 6 mm. Figure 5 indicates the dependence of the area of visible haziness on the impact energy. Cantwell [12] describes how the extent of the damage depends on how the specimen is mounted before impact. This relation can be seen in Fig. 5a. Specimens 2 mmS (S = small) and 2 mmM (M = middle) were impacted while clamped to the structure of the impact tower, whereas specimen 2 mmB (B = Big) were impacted while mounted on a wooden plate (Fig. 1). The clamped specimens showed a higher vulnerability against the impacts than the specimens on the wooden plate. 2.3 Local Acoustic Resonance Spectroscopy The LARS was executed using a miniature impact hammer with a sensitivity of 22.7 mV/N. In order to record the acoustic signals, a half inch free-field microphone with a frequency response of 3.15 Hz to 20 kHz of ±2.0 dB was used. The analog-to-digital converter RME ADI 2 transformed the signals of the microphone and the signals of the force sensor with a sampling rate of 90 kHz to be further processed. The data was analyzed with a self-written code determining the

123

full width at half maximum and characteristics of the acoustic spectrum. Jüngert [13–15] introduced the LARS technique and examined applications on wind turbine rotor blades. The LARS technique is a further development of the well-known “tap testing” to a technique suitable for automation and unmanned inspection. A portable test hammer apparatus to make traditional “tap testing” more reliable has already been invented and patented by Bruce Pfund [16]. The invention includes a housing with an impactor head, sensors and an acoustic spectral analyzer. Another remarkable invention in the field of acoustic nondestructive evaluation is by Olson et al. [17]. He describes a scanning apparatus which applies automated hammer taps and also analyses the acoustic signals. A device which interprets the signals of the force excitation is already commercially available [18]. The combination of both methods is new in the area of nondestructive evaluation. The LARS technique is suitable and was tested in this article in particular to detect delamination with a larger extension on widespread areas. 2.4 Ultrasound Ultrasound inspections were done using a mobile ultrasound echo device with a single test probe and phased array techniques. Probes with center frequencies ranging from 2.5 MHz to 5 MHz were used. Ultrasound testing has been utilized in various applications in civil and mechanical engineering, medicine and

J Nondestruct Eval (2015) 34:10

Page 5 of 14

10

Fig. 6 Different sounds on plates of different thickness. The sound of the thicker specimen is sharper and shorter in time

other areas [19]. The method was developed over more than 60 years, and optimized test systems are available. Current research investigates air coupling [20,21], phased array [22,23] and automated applications [15]. 2.5 Optical Lock-In Thermography OLT techniques were applied using a standard microbolometer based uncooled infrared camera together with the lock-in system developed by the company Edevis and 2 kW spotlights. Thermography typically provides good results for shallow depths or on thermoconductive materials such as carbon fiber reinforced polymers.

3 Results 3.1 Evaluation of Impact Damages Using Local Acoustic Resonance Spectroscopy In a first step, the signal differences produced by three specimens with different thicknesses were investigated. For delamination characterized by a splitting-up of different layers as described by Richardson et al. [6], it was assumed the signals measured on thin specimens would be similar to those measured on delaminated thick structures.

During inspection, signal differences caused by the three different thicknesses of the specimens were perceived. The thinner plates generated a lower and bumper sound when excited with the hammer, while the thicker plates generated a sharper and higher tone. The technical characteristics of the sound of the hammer excitation were also different on the three thicknesses, as shown in Fig. 6. Jüngert [13,14] uses the shift of the acoustic spectrum to display different characteristics of the specimen. In this study, a tendency of a shift of the acoustic spectrum to lower frequency on a defect spot could be observed as illustrated in Fig. 7, however the phenomena was not distinct at all defect spots. Further examination of the sound signal interpretation is necessary. Additionally to the sound, the LARS analyses the force excitation of the hammer. The thinner plates yield a bigger full width at half maximum of the force excitation and, consequently, a lower maximum force, whereas the thicker plates yield a smaller full width at half maximum and a higher maximum force. The mean value of the normalized force excitation of five measures on 2 mm thick specimen yielded 79.2 µs, on 6 mm thick specimen 62.5 µs and on 11 mm thick specimen 43.73 µs. Figure 8 shows three force excitations signals measured on the different plates. In this study, there were five hammer taps on flawed spots and five hammer taps on intact spots. As illustrated in Figs. 9

123

10

Page 6 of 14

J Nondestruct Eval (2015) 34:10

Fig. 7 Differences in the spectrogram of intact and defect spots. Defect spots have a lower amount of high frequency

Fig. 8 Three different force excitation signals on specimens of different thickness. The force excitation of a thinner plate is wider, whereas the signal of the force excitation of a thick specimen is thinner and sharper

Fig. 9 Comparison of five taps on the specimen with 2 mm thickness on intact structure (top) and five taps on a defect structure (bottom) shows the bigger value of the full width at half maximum of the force excitation signal on the defect structure

and 10, the results show that there is a difference in the full width at half maximum of the force excitations. The results are illustrated in a color scheme: Blue tones stand for a small full width at half maximum value, a sharp and narrow force excitation signal and therefore for an intact spot. Red and yellow tones stand for a high full width at half maximum value and therefor for flawed spots.

123

Figure 11 shows the grid for a measurement on specimen 2 mmM damaged with 8 impacts with an impact energy E I = 2–14 J. Every measuring point with reference to Fig. 11 was tapped with the hammer. The software acknowledged the position of the grid displayed in Fig. 11, generating a colorcoded image. As illustrated in Fig. 12, the results of the full

J Nondestruct Eval (2015) 34:10

Page 7 of 14

10

Fig. 10 Comparison of five taps on the specimen with 6 mm thickness on intact structure (top) and five taps on a defect structure (bottom) shows the bigger value of the full width at half maximum on the defect structure like in Fig. 9

excitation does not grow with a growing spatial scope of the damage. Moreover, although identical measurements were executed, the impacts caused during the second measurement could not be as clearly detected as those caused during the first measurement. 3.2 Evaluation of Impact Damage Using Ultrasound Fig. 11 Grid of measurements on 2 mmM specimen

Fig. 12 Color plotted measurements of the full width at half maximum with the local acoustic resonance spectroscopy: a first measurement. b Second measurement. c Artificially created ideal image of the measurement

width at half maximum in show that the LARS is capable of detecting damages in specimens of 2 mm thickness if they were impacted with an impact energy of at least E I = 6 J. However, the taps were executed directly on the impacts, which is an optimistic constraint for tests under real conditions. The color plotted measurements depicted in Fig. 12 indicate that the full width at half maximum of the force

Ultrasound tests with water as coupling medium were conducted on reference plates with blind holes of different depths as shown in Fig. 13. These tests showed that the proper detection of boundary layers in glass fiber reinforced plastics is possible, considering material used and the constraints. Boundary layers that lie deeper in the material are partially hidden, and as a result the measurement of the remaining wall thickness becomes difficult. In the impacted specimen, the damage did not show extensive delamination. As a result the characteristics of the damage did not show a clear boundary layer, making detection with ultrasound testing difficult. However, the presence of damage is still visible in the signal. In the 2 mm specimen, impacts with an impact energy of at least 5 J could be detected. The weakening of the echo in the rear panel as observed in several instants, indicates the existence of damage as seen in Figs. 14 and 15. Impacts with an impact energy of at least 50 J were visible in 6 mm specimens with Ultrasound testing. Testing with an Olympus Omniscan MX2 phased array ultrasound device showed the capability of phased array devices to detect impact damage greater than 60 J. On the cutaways of the real rotor blades, the testing with the single probe ultrasound did not provide usable results. However, with the phased array test probe, impact damage for E I > 60 J was revealed. Figure 16a shows a photo of the tested area. Figure 16b shows the image generated by testing with the phased array, which illustrates a hidden damage. 3.3 Results of Optical Lock-In Thermography The specimens of glass fiber reinforced plastics used in this study offered an easy insight into the material by putting them against strong light, as described in [10]. However, internal

123

10

Page 8 of 14

J Nondestruct Eval (2015) 34:10

Fig. 13 Measurement of boundary layers of blind holes in a 2 mm specimen. a Photo of the testing of the reference plates. b Signal at half a millimeter remaining wall thickness. c Signal at one millimeter remaining wall thickness. d Signal at one and a half millimeter remaining wall thickness

Fig. 14 Changing of the signal of the ultrasound by the presence of a defect. a Signal of intact spot with the echo of the rear panel at 2.02 mm. b Signal of flawed spot with no rear wall echo

Fig. 15 Comparison of an intact and flawed spot in the 6 mm specimen. a Intact spot. b Flawed spot with weakened rear wall echo

damage characteristics on coated specimens could not be analyzed by this simple visual testing. For this, optical thermography was used. On opaque, and more thermoconductive carbon fiber reinforced plastics, the OLT also reveals non visible damages.

123

The penetration depth of the OLT depends on the lockin frequency; the lower the lock-in frequency, the deeper the position of the displayed plane. On the other hand, a lower lock-in frequency prolongs the test. With four periods of measurement and a lock-in frequency of 0.001 Hz, gener-

J Nondestruct Eval (2015) 34:10

Page 9 of 14

10

Fig. 16 The phased array UT displays a hidden damage. a Photo of tested area. b Yielded image by the ultrasound phased array

Fig. 17 Comparison of OLT with four different lock-in frequencies by the example of the three reference plates with the blind holes. On the right there is the reference plate with 2 mm thickness, in the middle the

reference plate with 6 mm thickness, on the left there is the reference plate with 11 mm thickness. At low frequencies the OLT has deeper penetration depths, so that more blind holes are visible

ating one image takes more than one hour for one measuring process. This is a limitation concerning the practicability of the method. Figure 17 shows four phase images of the reference plates with 11 mm in thickness on the left side, 6 mm

in thickness in the middle and 2 mm in thickness on the right side. Felt pen printings are visible on the phase image showing the lock-in frequency of 0.05 Hz in Fig. 17, the boundary

123

10

Page 10 of 14

J Nondestruct Eval (2015) 34:10

Fig. 18 Detection of impact damages in the specimen 2 mmB of 2 mm in thickness. a Photo of the specimen. b Phase image of the specimen, tested with OLT with a lock-in frequency of 0.005 Hz. The different gray shades in the image stand for different phase angles

Fig. 19 Phase images with different lock-in frequencies and a photo of the damaged specimen. A delamination is visible with OLT. This delamination is not visible in the photo of the specimen on the upper photo on the left. (Dotted line)

layers are not evident. By reducing the lock-in frequency, the boundary layers of the blind holes become visible. With a lock-in frequency of 0.001 Hz the 2 mm specimen becomes almost clear, and blind holes with boundary

123

layers of up to three millimeters appear. This indicates that the penetration depth with the settings in this study is roundabout three millimeters at a lock-in frequency of 0.001 Hz.

J Nondestruct Eval (2015) 34:10

As Fig. 18 shows, the detection of the impacts with an impact energy of at least 10 J in the 2 mm specimen was possible with the OLT. The figure shows a photography and a phase image of the specimen. Testing a segment of a real rotor blade showed that the optical lock-in thermography is capable of showing hidden defects. Figure 19 shows a surface-near separation of the layers, which is not visible. By comparing the lock-in frequencies, Fig. 19 also shows that the flaw increases in depth towards the left-hand side. 3.4 Results of the Computed Tomography The computed tomography (CT) gives a good insight into the inner structure of a material and is also a common method for developing and comparing other techniques of nondestructive testing. In this research project, the computed

Page 11 of 14

10

tomography was executed by a phoenix nanotom m with a resolution of roundabout 15 µm. Figure 20 shows the rising defect of the cutaway view for the 2 mm specimen. Figure 21 shows three tomographies from the center of the impact showing the higher resistance of thicker specimen against impacts. The initiation of the damage occurs from the rear panel for thinner laminates and from the front side for thicker laminates as described in [1]. This can be seen in Fig. 21. In the upper layers, failure of the material occurs by shear stress induced rifts, whereas in the lower laminates failure occurs through tensile strain.

4 Valuation of the Methods of Nondestructive Evaluation 4.1 Local Acoustic Resonance Spectroscopy

Fig. 20 Cutaway view of impact damages in 2 mm specimen. The picture shows rising impact energies from the top to the bottom

The coin tapping method is already being applied in professional practice. Hammers with different peaks belong to the basic equipment for surveyors. Experience combined with a sharpened ear give the worker an insight into a material. The information is based on the tactile response of the material, such as for the commercially available Mitsui Woodpecker, as well as the audible tapping sound. The LARS analyzes not only the force excitation, but also the sound of the tapping. However, the LARS has a low technological readiness level. There are no established systems, therefore the hardware has to be put together, and the software has to be developed. Impact damages in this study were difficult to detect, if they were not hit exactly by the hammer. Damages can only be detected, by hitting the center of the impacted spot. An advantage of the LARS is the good handiness of the system: the hammer fits on a worker’s belt. Evolutions of future applications should maintain this handiness as it enables testing of widespread areas with little effort. In comparison to thermography and ultrasound testing, LARS is not vulnerable against the high damping of sonic waves and requires no high thermoconductivity as opposed to OLT. 4.2 Ultrasound Testing

Fig. 21 Comparison of different damage characteristics in different thicknesses. In a very thin specimen the damage initiation is at the rear side of the specimen, whereas at very thick specimen the damage starts from the front side

Ultrasound testing has been developed in the last 50 years. In contrary to the local acoustic resonance system, ultrasound testing gives information on the depth of the flaw. Outlines of the flaw can easily and precisely be drawn on the surface. A disadvantage of ultrasound testing is the coupling of the method. Consequently the method is not suitable for widespread areas.

123

10

Page 12 of 14

J Nondestruct Eval (2015) 34:10

Table 1 Evaluation of the used methods of non-destructive testing

Tecnique

Practicability

Technological Readiness Level

Depth of information

Possible field of work

LARS

++

−−

−

Quick first evaluation of structural integrity on bigger components

UT

0

++

+

If more information of a defined spot is needed

OLT

+

+

+

Gives an easy to interpretate overview

Fig. 22 Vision of automatized Inspection after assembly

4.3 Optical Lock-In Thermography The OLT does not require coupling and produces an easyto-interpret image. Moreover thermographical images can be taken from afar, using the right lenses. This gives the method a great flexibility and potential for various applications. However, excitation requires a lot of hardware with high power supply and cannot be applied from afar. Passive thermography uses natural thermal gradient to bring forth internal structures, and could be a solution to avoid elaborate and costly excitation. 4.4 Computed Tomography Computed tomography delivers by far the best insight of the damage, but is also the most sophisticated and expensive method. Computed tomography in this work showed that effect of defect is more severe on thinner specimen.

5 Discussion The results from the experiments in this study uncover the weaknesses and strengths of the methods of nondestructive evaluation used. The strengths of each method should be combined to develop efficient holistic processes. Table 1 summarizes the examined methods.

123

In reality, each field of nondestructive testing in automotive, aeronautical and civil engineering has its own constraints. In this study, the nondestructive testing of wind turbine rotor blades is discussed. This can be separated into two main areas: one is the testing of the rotor blades after manufacturing, and the other is the testing of mounted rotor blades. Flaws in manufacturing are the main reason for severe and expensive accidents in service [24]. For this reason, the quality of wind turbine rotor blades is important, making it essential to check all rotor blades after manufacturing. Flaws in newly manufactured rotor blades include debonding between the half shells and between the half shells and the stringer, and wrinkles/waviness in the fabric in the laminate. To detect these defects, methods of nondestructive evaluation are applied. Size and individuality of the rotor blades and economic constraints have hindered the development of automatized systems. In the author’s opinion, the development of a linear gantry system, which checks the rotor blades thoroughly after their manufacturing, would save a lot of money in the lifecycle of rotor blade. Figure 22 shows a possible scenario for the evaluation of a rotor blade after manufacturing. On mounted rotor blades, the constraints for non-destructive testing are different. The main challenges are the high altitude and difficult access of rotor blades in turbines. Currently, the inspector checking the condition of the rotor blade

J Nondestruct Eval (2015) 34:10

Page 13 of 14

10

Fig. 23 a Vision of automatized testing of mounted rotorblades of wind turbines. b Passive thermography of a tip of a rotorblade, source [14]

is rappelled down from the top. This method is inherently risky and offers little room for automation. Passive thermography offers great insight into the structure of rotor blades for wind turbines [25,26]. Additionally, the application of drones is becoming more and more varied. With this in mind, the inspection of rotor blades of wind turbines with thermography drones would be suitable. The sight check could be captured by a drone’s video capabilities and then assessed by a technical expert. However, this idea is not entirely new, there are patents on the idea of detecting flaws from the air with thermography [25], as well as the use drones for this task [27,28]. To the author’s knowledge, these concepts have not yet been combined in reality. GPS enables software to steer and track drones; as a result, automatized scenarios that inspect rotor blades on wind turbines, and wind farms as depicted in Fig. 23, are imaginable. A hindrance in the development of this technology is the same as for newly manufactured rotor blades: high costs.

6 Conclusions 1. In effort to add to this body of research, this study found that in order for an impact hammer to detect damage, it needs to make contact in precisely the same spot of an impact damage. Additionally, this research found that the greatest advantage of LARS is its usability. Consequently, the method is recommended for big structures and for the detection of widespread flaws. 2. Ultrasonic testing gives information on the depth of the flaw and provides an image of the defect that can be sketched on the surface which provides information on size and shape. However, coupling makes the method more tedious. Therefore, ultrasonic testing is recommended after a method used for widespread areas has encircled the area of the damage, or if the area of required testing is known. 3. OLT demonstrated its potential in non-destructive testing. Advantages of this method include that impacts

were clearly visible, and the penetration depths can be easily adjusted by different lock-in frequencies. Its handiness, simple interpretation and the variation of methods of active thermography as well as passive thermography make thermographical testing attractive. Moreover, thermographical systems offer great potential for automatized applications. 4. Cost pressure in the wind energy industry is a stumbling block for the development of new automated holistic approaches for testing of rotor-blades of wind turbines. However, thorough testing of newly manufactured rotor blades pays off in the life cycle of the component. 5. Impact damages could be detected by all mobile methods of nondestructive evaluation starting from the approximate impact energy that produces the peanut shaped haziness in the specimen. From that energy on, CT uncovers characteristics like fiber breakage and air pockets in the material. Future investigations should concentrate on the development of automatized systems of nondestructive evaluation. Several automatized systems can be combined to a holistic concept of nondestructive evaluation of fiber reinforced components. Acknowledgments The authors thank Prof. Dr. Ing. Bernhard U. Seeber, M.Sc. Gaetano Andreisek, Dipl. Ing. Benedikt Rauh for the support in execution of the testing and Mr. Otto Lutz for his advices from fields. The research has been financed by and executed at IABG mbH.

References 1. Abrate, S.: Impact on Composite Structure. Cambridge University Press, Cambridge (2005) 2. Meola, C., Carlomagno, G.M.: Impact damage in GFRP: new insights with infrared thermography. Compos. A 41, 1839 (2010). doi:10.1016/j.compositesa.2010.09.002 3. Caprino, G., Lopresto, V., Langella, A., Leone, C.: Damage and energy absorption in GFRP laminates impacted at low-velocity: indentation model. Eng. Proc. (2011). doi:10.1016/j.proeng.2011. 04.380

123

10

Page 14 of 14

4. Davies, G.A.O., Hitching, D., Wang, J.: Prediction of threshold impact energy for onset of delamination in quasi-isotropic carbon/epoxy composite laminates under low-velocity impact. Compos. Sci. Technol. 60, 1–7 (2000) 5. Hou, J.P., Petrinic, N., Ruiz, C., Hallet, S.R.: Prediction of impact damage in composite plates. Compos. Sci. Technol. 60, 273–281 (2000) 6. Richardson, M., Wishheart, M.: Review of low velocity impact properties of composite materials. Compos. A 27, 1123–1131 (1994) 7. Ramasamy, P., Sampathkumar, S., Prsath, P.: Experimental Investigation of Impact damage on woven GFRP Composite Laminate using acoustic emission and Ultrasonic Technique. In: Proceedings of the National Seminar & Exhibition of Non-Destructive Evaluation. pp. 55–59 (2011) 8. Scarponi, C., Briotti, G.: Ultrasonic technique for the evaluation of delaminations on CFRP, GFRP. KFRP composite materials. Compos. B 31, 237–243 (1999) 9. Schürmann, H.: Konstruieren mit Faser-Kunststoff-Verbunden. Springer, Darmstadt (2007) 10. Liu, D.: Impact-induced delamination—a view of bending stiffness mismatching. J. Compos. Mater. (1988). doi:10.1177/ 002199838802200706 11. Choong, W.H., Yeo, K.B., Fadzlita, M.T.: Impact damage behaviour of woven glass fibre reinforced polymer composite. J. Appl. Sci. (2011). doi:10.3923/jas.2011.2440.2443 12. Cantwell, W.: Geometrical effects in the low velocity impact response of GFRP. Compos. Sci. Technol. (2006). doi:10.1016/ j.compscitech.2006.10.015 13. Jüngert, A.: Untersuchung von GFK Bauteilen mit akustischen Verfahren am Beispiel der Rotorblätter von Windenergieanlagen. Dissertation, Universität Stuttgart (2010) 14. Jüngert, A.: Local acoustic resonance spectroscopy (LARS) for glass fiber-reinforced polymer application. J. Nondestruct. Eval. doi:10.10/1007/s10921-013-0199-3 (2013) 15. Jüngert, A., et al.: Zerstörungsfreie robotergestützte Untersuchung von Rotorblätter von Windenergieanlagen mit Ultraschall und Thermographie. ZFP Ztg. 115, 43–49 (2009)

123

J Nondestruct Eval (2015) 34:10 16. Pfund, B.: Portable test hammer apparatus. US Patent 5,686,652, Appl. No.: 08/707,851 (1997) 17. Olson, L. et al.: Scanning apparatus and method for non-destructive materials evaluation and mapping through use of acoustic waves. US Patent Number 5,404,755, Appl. No.:867,248 (1995) 18. JR Technology Limited. Development of the Woodpecker Tap Tester (2012) 19. Krautkrämer, J., Krautkrämer, H.: Werkstoffprüfung mit Ultraschall. Springer, Köln (1966) 20. Lin, S.: Study on the high power air coupled ultrasonic compound transducer. Ultrasonics (2006). doi:10.1016/j.ultras.2006.05.120 21. Revel, G., Pandarese, G., Cavuto, A.: The development of a shocktube based characterization technique for air-coupled ultrasonic probes. Ultrasonic 54(6), 1545–1552 (2014) 22. Gaun, X. et al.: Post processing of phased array ultrasonic inspection data with parallel computing for nondestructive evaluation. J. Nondestruct. Eval. (2014). doi:10.1007/s10921-013-0219-3 23. Huang, J., Que, P.W., Jin, J.H.: A parametric study of beam steering for ultrasonic linear phased array transducer. Russ. J. Nondestruct. Test. 40(4), 46–53 (2004) 24. Balk, S.: Status und Trends in der Entwicklung von Materialund Fertigungskonzepten für Rotorblätter von Windkraftanlagen. IABG mbH, München (2011) 25. Zell, H.: Verfahren zur Überprüfung des baulichen Zustands von Windkraftanlagen. EP 2 631 476 A2, Mühlheim an der Ruhr (2013) 26. Zell, H., Carl, V.: Wind Turbine Inspection New Methods of Remote Nondestructive Inspection of Rotorblades. Wilhelmshaven (2012) 27. Wobben, A.: Method for monitoring wind power plants, US Patent 6,785,637, Appl. No.: 10/089,774 (2001) 28. Fernandes, R.A.: Monitoring system for power lines and right of way using remotely piloted drone. US Patent 4,818,990, Appl. No.:95,152 (1989)