- Email: [email protected]
fiberglass composites
Polymer Testing 50 (2016) 152e163
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Effects of accelerated aging on mechanical, thermal and morphological behavior of polyurethane/epoxy/fiberglass composites Alexandre de Souza Rios a, b, *, Wanderley Ferreira de Amorim Júnior c, Elineudo Pinho de Moura a, Enio Pontes de Deus a, Judith Pessoa de Andrade Feitosa a , Campus do Pici, Fortaleza, 60440-554, Brazil Department of Metallurgical and Materials Engineering, Federal University of Ceara LMT (ENS Cachan, CNRS, Universit e Paris Saclay, Universit e Paris 6), 61, av. du Pr esident Wilson, 94235, Cachan, France c Department of Mechanical Engineering, Federal University of Campina Grande, Campina Grande, 58.429-140, Brazil a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 November 2015 Accepted 11 January 2016 Available online 12 January 2016
Wind blades, an important application of polymeric composite materials, are subject to natural weathering. This study aims to evaluate mechanical, thermal and morphological behavior during accelerated aging in three thicknesses of epoxy and fiberglass polyurethane-coated composite plates used in wind turbines, in addition to testing with two acoustic emission techniques. An accelerated aging chamber simulated natural weathering mechanisms for 45, 90, 135 and 180 days. This degradation primarily reduced the mechanical properties of the thinner composites, with some damaged specimens exhibiting fiber-matrix debonding. Thermal properties deteriorated. There were no morphological changes on the polyurethaneeepoxy interface; however, degradation occurred in the fiber-matrix interface on the surface exposed to radiation. The degree of chalking indicated coating deterioration on the external surface of the polyurethane. The acoustic wave propagation speed and attenuation coefficient measured prior to mechanical testing indicated the presence of damage areas. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Accelerated aging Acoustic emission Damage Wind turbine
1. Introduction In recent decades, there has been a growing demand to exploit global wind potential, transforming wind blades into one of the most important applications of polymeric composite materials. Some wind turbine components, primarily wind blades, are manufactured using polymeric composite materials [1e4]. Chemical or environmental aging is caused by various agents, such as humidity, loading conditions and ultraviolet radiation (UV), leading to irreversible changes in the molecular structure of these materials [5e7]. The polymeric matrix is usually more susceptible to aging and generally controls the long-term performance of the composite [5,6]. The critical portion of UV radiation causes photo initiation of the polymer due to the absorption of chromophores present in the matrix (hydroperoxides, catalyst residues, carbonyls, unsaturation). Excited chromophores induce photooxidative decomposition of macromolecular chains, leading to basic behavior changes in the polymer and
* Corresponding author. Department of Metallurgical and Materials Engineering, , Campus do Pici, Fortaleza, 60440-554, Brazil. Federal University of Ceara E-mail addresses: [email protected], [email protected], eng.alexandrerios@ gmail.com (A. de Souza Rios). http://dx.doi.org/10.1016/j.polymertesting.2016.01.010 0142-9418/© 2016 Elsevier Ltd. All rights reserved.
its properties. Photoinduced processes usually change the appearance (surface gloss, color) and mechanical properties (strength, strain, flexibility) of the polymer [8,9]. The combined action of all these factors on the behavior and durability of composites is a highly complex phenomenon that occurs at the molecular level. When a polymer is subjected to ultraviolet radiation it releases constituents in the form of thin, loosely adherent dust. This coating defect is called chalking [10,11]. Coatings are expected to be durable and retain their properties over time. As such, resistance to light, humidity and temperature is a general requirement. The environmental impact of coatings can be further reduced by increasing their efficiency and useful life. The most common cause of coating degradation is ambient exposure to ultraviolet (UV) radiation, water/humidity and temperature fluctuation. In many cases, degradation is evaluated based on changes in chemical structure and the presence of foreign chemicals in the system as a function of time. Polyurethane coatings made with an aliphatic and cycloaliphatic isocyanate-based curing agent exhibit excellent resistance to ultraviolet rays as well as good color retention and glow when exposed to natural weathering. Furthermore, they show good resistance to chalking [12]. A controlled artificial weathering test should replicate
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
conditions typically found in the natural aging process through cycles representing day and night, and periods of surface drying and wetting. The advantage of this method is the ability to accelerate the testing of all controlled parameters, obtaining comparable results at significantly lower exposure times [13]. Possible causes of failures in composite materials used to manufacture wind turbines are an abrupt change in blade thickness with tension concentrators and decoupling of the resins or delamination. Decoupling resins may also be associated with processing defects in manual lamination. Both the individual and collective action of these defects can cause premature wind blade failure [14]. Debonding of the outer skin was the initial failure mechanism, followed by delamination buckling which led to the blade's collapse [15]. Tests using a full-scale wind turbine blade to study structural fatigue behavior are costly, and hence few studies have been conducted to date [15]. Another approach to improve the reliability of wind turbine blades is to evaluate the mechanical properties of composite laminates. Non-destructive testing, such as acoustic emission (AE), is one of the most appropriate to characterize composite materials. AEs are transient ultrasonic waves generated by sudden movement in material under stress [16e18]. When a component is subjected to mechanical load, discontinuities in materials may release AE energy. The electrical signals are then amplified and further processed as AE signal data. Accurate processing of the AE signals can identify different damage mechanisms in composites [19,20]. Therefore, the present study aims to evaluate the mechanical, morphological and thermal properties of artificially degraded composite plates. The properties identified in different analyses provide knowledge on the durability and structural applications of composites exposed to environmental conditions. The signals obtained from acoustic emission provide information on the formation of discontinuities or voids arising from degradation. 2. Experimental 2.1. Materials Composite plates (1, 3 and 6 mm thick) manufactured by hand lay-up were used. These consist of 989 Biax Saertex 45 /-45 fiberglass fabric layers, Hexion R1M135 þ 1366 epoxy resin and DM Coating-sprayed polyurethane. For degradation cycles the plates were machined according to the tensile test specimen (ASTM D638) and Izod impact specimen (ASTM D256-10) [21,22]. 2.2. Composite characterization The effects of sunlight were simulated by a system of 8 UVB radiation sources in the 280e320 nm range using an UV-C AdeximConexim® accelerated aging chamber. Specimens were exposed to a UVB radiation cycle at 60 C for 6 h, followed by exposure to 6 h of condensing water vapor at 50 C and 100% humidity according to ASTM G154-06 [23] (Fig. 1). Specimens were removed every 1080 h (45 days) and four degradation conditions were evaluated: 45, 90, 135 and 180 days. A Physical Acoustics Corporation® sensor (model ASP05F0021) was used for the acoustic emission test. A mechanical stimulus was provided by breaking graphite (Fig. 2), according to ASTM E976-10 and ASTM E1067-07 [24,25]. The absolute maximum and superposition methods were used. 2.2.1. Absolute maximum method The maximum peak time values of each sensor were associated with distance (x) to provide propagation speed. Acoustic signal attenuation (a), expressed in dB/m, was measured using attenuated acoustic intensity (I) and non-attenuated intensity (Io), as described
153
Fig. 1. Specimens in the accelerated degradation chamber.
in Equation (1) [26]:
1 Io I ¼ Io$eax /a ¼ $ln x I
(1)
2.2.2. Superposition or cross-correlation method The time-of-flight (t0) of the acoustic wave between two adjacent signals B1 (t) and B2 (t þ t0), can be used to determine the wave speed through the specimen. Thus, propagation speed can be determined with t0 and the distance between sensors [27,28]. Time-of-flight was estimated by the cross-correlation or superposition method. The value of t0 is computed as the value of t, which is maximum in Equation (2):
∞ Z B1 ðtÞ$B2 ðt tÞdt
(2)
∞
The cross-correlation of two functions (B1 and B2) involves shifting B2 in t toward B1 and multiplying them to determine for which value of t (t0) the product is maximized. This means that the maximum of the integral of the cross-correlation between echoes B1 and B2, as a function of time delay, corresponds to their maximum correlation [28]. An EMIC® DL 10000/700 universal testing machine with a 100 kN load capacity was used for tensile testing. The deformation rate was 50 mm/min. Composite plates were tested by Izod impact testing in each degradation condition according to ASTM D256-10 [22]. According to this standard, the minimum thickness is 3 mm. A CEAST® Resil 5.5, 5.5 J test hammer and impact velocity of 3.46 m/s were used in this test, with a type B, V-shaped notch and a 45 angle. Thermogravimetric analysis (TGA Q 50®) was carried out at temperatures ranging from 25 C to 800 C in air, with a heating rate of 10 C min1 and flow rate of 100 mL min1, using about 10 mg for each sample. Differential Scanning Calorimetry (DSC Q 20®) was employed to determine the glass transition temperature (Tg) of the composites. A heating rate of 10 C/min at 50 C to 500 C was applied to 10 mg samples under nitrogen. The sample constituents and their interfaces were viewed under an Olympus optical microscope (model BX 51 M) at magnifications of 50, 100, 200, 500 and 1000. Cross-sections were made for sample preparation using a water jet cutter and water-only cutting, followed by sanding with 300, 600 and 1200 grit sandpaper, and mechanical polishing with diamond paste. Both procedures were performed using a Teclago® PL 02 polishing machine. The polyurethane coating was directly exposed to UVB radiation
154
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 2. Acoustic emission test procedure: (a) graphite to be broken as mechanical stimulus to generate acoustic signals, (b) conducting the test and (c) mechanical stimulus region.
and detachment of constituents was evaluated according to DIN EN ISO 4628-6: 2002-02, which measures the degree of chalking in a polymer [29]. This technique consisted of applying a transparent adhesive on the polyurethane coating. After it was removed, the degree of chalking was determined by comparing visual standards, which can vary from 0 (absence) to five, where five corresponds to a significant amount of chalking. 3. Results and discussions 3.1. Acoustic emission tests The 1 mm-thick specimens exhibited excessive deformation after accelerated degradation and could not be coupled to the sensor. Hence, they were not considered in the results. Fig. 3 shows that the time required for the acoustic signal to travel between sensors was 55 ms (578 ms e 523 ms ¼ 55 ms). Given that the distance between sensors was 130 mm, the speed of the traveling wave must be 2363 m s1, according to the absolute maximum peak. The absolute peak decreased from 980 V (first sensor) to 6 107 V (second sensor). The frequency content of each component of the decomposed signals emitted during the mechanical stimulus is shown in Fig. 3. The results indicate that the frequency range of the decomposed components decreased from one sensor to another and show significant acoustic wave attenuation. Acoustical activity is also observed between 50 kHz and 500 kHz; however, the highest acoustic events occur between 50 kHz and 100 kHz. It should be noted that peak frequency analysis is primarily influenced by the sensor type used, composite thickness, mechanical stimulus, equipment bandwidth and depth position of the source [30e32]. Ideally, the distance between sensors should not be constant for all measurements and may play a role in peak frequency deviations. The aim here is to investigate damage areas from artificial degradation, presenting comparative information and relate the AE energy of the first and second sensors. Fig. 4 shows acoustic wave speeds using the absolute maximum and superposition methods on the 3 mm and 6 mm specimen degradation conditions. The results obtained by the absolute maximum method corroborate those recorded using the superposition method, showing flexibility. The thinner composites exhibited random modifications in acoustic wave propagation speed during degradation. The 6 mm-thick composites showed a decrease in mean values when compared against the non-degraded condition, possibly associated with loss of elastic properties or damage onset in the material during exposure. Different frequency distributions for dissimilar failures may be due to the fact that elastic acoustic velocities and intrinsic frequencies are associated with modulus of elasticity and density, according to Equation (3) [16]:
sffiffiffi E f fCf r
(3)
where f is frequency, C elastic acoustic velocity, r density and E is the modulus of elasticity. The wave speed in the air is lower than in the composite components, and declining propagation speed in the specimens may be related to the formation of voids due to accelerated degradation (matrix cracking, fiber breakage, fiber-matrix debonding, etc.) [33]. Considering significant differences in the position of the acoustic source, a high variation in frequency spectra would be expected when testing different specimens from the same laminate. In this study, at least six specimens per material were tested. The standard deviations found may be associated with the anisotropic nature of this material, which may contain scattered beams and cause variations in propagation speed. Fig. 5 shows the mean attenuation coefficient values in 3 and 6 mm-thick specimens. Except for the specimens degraded for 45 days, there was a slight increase in attenuation coefficients in both geometries. The initiation and propagation of discontinuities generate voids with a higher attenuation coefficient than that of the composite in the “as received” condition, demonstrating damage formation. In plate type structures, stress waves generated by damage events propagate as combinations of different Lamb wave modes [34e37]. The presence of each type of Lamb wave mode and its frequency content is determined by the location and type of the damage event. In composite test specimens, results from previous studies [38e40] indicated the occurrence of AE signals in the frequency range between 100 kHz and 500 kHz. In some cases, higher order modes occur near the higher end of this range. These higher order modes are either too small or undetectable by commercially available AE sensors. Thus, experimental AE signals in composite materials in this range are likely to be combinations of the fundamental modes. Numerous attempts have been made in the literature to classify AE signals according to the likely failure modes that generate them [40e43]. However, definitive relationships between the different failure modes and the features of resulting AE signals have yet to be established. Such correlations would only be feasible if distinguishing features such as the ratio of different modes and different frequency components are preserved in the signals as they propagate along the composite laminates. Should these features not be preserved during the propagation of AE related waves, attempts to relate features of AE signals to composite failure modes would fail. The characteristics of AE signal waveforms in terms of individual wave mode amplitudes and frequency components are important for identifying types of failure modes. Both the amplitude and
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
155
Fig. 3. Signal detected by the (a) first and (b) second sensors after mechanical stimulus.
Fig. 4. Acoustic wave speeds using the absolute maximum and superposition methods on the (a) 3 mm- and (b) 6 mm-thick specimens under accelerated degradation.
frequency content of AE signals are altered by different attenuation mechanisms as the waves propagate in the laminates. Geometric spreading and frequency-dependent dispersion reduce the
amplitude of AE signals. For example, the viscoelastic nature of the matrix in carbon/polymer composites introduces significantly higher levels of attenuation [44].
156
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 5. Attenuation coefficient of the 3 mm- and 6 mm-thick specimens.
3.2. Mechanical properties 3.2.1. Tensile testing Fig. 6 presents the modulus of elasticity, tensile strength and strain at break of tensile testing in 1 mm, 3 mm and 6 mm-thick composites under the conditions assessed. There was a 10.0% reduction in modulus of elasticity between the “as received” condition and after 180 days degradation for 1 mm-thick composites
(8.39 GPae7.55 GPa) and a 4.8% decline over this degradation range for 3 mm-thick composites (9.05 MPae8.62 MPa). By contrast, the 6 mm-thick specimens showed a decrease in modulus of elasticity up to 90 days of degradation, followed by an increase until the final stage of degradation. The deterioration of this elastic property was indicated, but not quantitatively measured, by the acoustic measurements presented and discussed above. These results show that the thinner specimens are more susceptible to accelerated degradation imposed by the chamber, indicating a more pronounced reduction in modulus of elasticity at the surface. The modulus of elasticity is dependent on the crystallinity of the polymer [45,46]. UV irradiation leads to photodegradation of the polymer via a well-established mechanism previously reported by a number of authors [45,47e49]. The degradation of polymers is initiated with chain scission, followed by branching, cross-linking and, finally, the formation of double bonds. One of the irreversible alterations caused by photodegradation is the loss of amorphous regions due to cross-linking, to which can be attributed the change in the modulus of elasticity of samples [50]. The three thicknesses exhibited a decrease in tensile strength after 180 days of degradation: 17.3% for the 1 mm-thick composite (53.71 MPae44.41 MPa), 10.4% for the 3-mm thick composite (52.97 MPae47.44 MPa) and 14.8% for the 6 mm-thick composite (60.78 MPae51.78 MPa). The 1 mm-thick composites were more affected by accelerated degradation, indicating that the decrease in tensile strength was also more pronounced at the surface.
Fig. 6. Mechanical properties of the composites: (a) modulus of elasticity (b) tensile strength and (c) strain at break.(d) Izod impact test results of 3 mm-thick specimens in different conditions.
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
157
Fig. 7. Visual appearance of the specimens after the Izod impact test.
Fig. 8. (a) Delaminated region of a 6 mm-thick specimen degraded for 90 days and not ruptured after the Izod impact test (50 magnification). (b) Cracked and delaminated region oriented at 45 of a 6 mm-thick specimen degraded for 90 days and not ruptured after the Izod impact test (50 magnification).
The opposite effect occurs for strain at break. Thicker composites experienced a greater reduction in strain at break because their modulus of elasticity (stiffness) did not suffer major damage. Overall, the tensile test results obtained in this study indicated that degradation was predominantly superficial, having a greater effect
Table 1 Residual weight and maximum degradation temperature of the composite from TGA measurements.
Fig. 9. TGA thermogram of conditions evaluated.
Conditions
Residual weight [%]
Maximum degradation temperature [ C]
Sample 1
Sample 2
Mean
Sample 1
Sample 2
Mean
As received 45 days 90 days 135 days 180 days
65.22 66.54 67.87 66.84 62.59
67.07 65.23 66.12 66.20 67.46
66.15 65.89 67.00 66.52 65.03
358.4 356.0 337.6 335.4 337.6
348.9 343.3 338.3 337.6 334.7
353.7 349.7 338.0 336.5 336.2
158
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 10. (a) DSC thermogram and (b) Tg of degraded conditions.
Fig. 11. Polyurethane coating under the conditions evaluated: (a) as received, (b) 45, (c) 90 (d) 135 and (e) 180 days of degradation. 100 magnification.
on the mechanical properties of thinner composites. Several authors have shown strain at break to be a very useful measurement of polymer degradation [51e54]. A similar trend was observed for the other polymers tested. This behavior can be related to chain scission in semi-crystalline polymers, which leads to a decrease in the molecular weight of the material and, consequently, a loss in strain break with weathering [51,54e56]. According to the literature, this property follows a decreasing trend proportional to artificial weathering [50]. Some results obtained using this technique showed high
Fig. 12. External cross-section section of the ''as received'' composite. 100 magnification.
standard deviations, dispersion and non-linearity of the mechanical properties as a function of degradation stages. This behavior may be due to composite processing associated with possible nonuniform degradation within the accelerated aging chamber.
3.2.2. Izod impact testing Fig. 7 shows 3 mm and 6 mm-thick specimens after the Izod impact test in each condition evaluated. After 45 days of exposure, the surfaces of samples exhibited a distinct change in color from colorless to dark yellow, and colorless to light yellow in controls, with minor changes in surface roughness. The discoloration established that photooxidation resulted in the formation of chromophoric chemical species, which were absorbed in the visible range of light [57,58]. Fig. 7 shows an increase in deformation with exposure time in the accelerated aging chamber for the 3 mm-thick specimen after Izod impact testing. In the “as received” and 45 -day degradation conditions there was no complete rupture of the specimen, although delamination and permanent deformation were detected, especially after 45 days of degradation. Complete rupture was observed for the other degradation conditions and was more pronounced in specimens with greater exposure in the accelerated aging chamber. As occurred in the 6 mm-thick specimens, accelerated degradation decreased the fiber-matrix bond in 3 mm-thick specimens. The energy provided by a 5.5 J hammer was not sufficient to fracture the 6 mm-thick specimens. However, delaminated regions were detected near the notch of the degraded specimen, which did not occur in the “as received” condition. This defect is usually present in composite failures (Fig. 8). Even without fracturing,
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
159
Fig. 13. Cross-section of the composite at: (a) 45, (b) 90, (c) and 135 (d) and 180 days of degradation (100 magnification).
accelerated degradation decreased the fiber-matrix bond in the 6 mm-thick specimen. Fig. 8 shows the notched region of a 6 mm-thick specimen after 90 days of degradation not ruptured after the Izod impact test (50 magnification). There was more damage in the region close to the notch, a point of greatest stress concentration. Delamination and fiber-matrix detachment are evident. Fig. 8 shows the delamination path propagated between the interstices fibers, oriented at 45 . Fig. 6d shows the mean values and standard deviations of Izod impact test results for 3 mm-thick specimens. The impact strength of the composites varied with accelerated degradation time. Even without rupture of some samples, there was a decrease in impact resistance between the “as received” and 45-day degradation conditions from 1802.0 J/m to 1791.1 J/m. The conditions in which rupture occurred after impact testing (90, 135 and 180 days of degradation) showed decreased impact resistance, hence the aging effect of the aging chamber reduced the Izod impact strength of the composites and modified fiber-matrix bonding. 3.3. Thermal properties Fig. 9 shows the TGA thermograms of samples in all the conditions evaluated. Residual weight was between 60 and 70% and the maximum degradation temperature (Tm) was around 350 C for all conditions. Table 1 shows the residual weight and maximum temperature of two samples for each degradation condition evaluated. The maximum degradation temperature was obtained using derivative thermogravimetry (DTG), where curves are recorded
based on the TGA curve and correspond to the first derivative of weight change versus time (dm/dt), which is recorded as a function of temperature or time. Table 1 shows that residual mass obtained high values due to the presence of inorganic matter (fiberglass), whose decomposition temperature is higher than the maximum temperature used in this test. There were no significant changes in residual weight with progressive accelerated degradation because the same weight percentage decomposes irrespective of the accelerated degradation stage. However, there was a decrease in maximum temperature according to accelerated degradation stages. The reduction was approximately 5.0% at maximum degradation temperature when compared to the “as received” and 180-day degradation conditions (353.66 Ce336.14 C). The decrease in thermal stability with increasing aging time may be due to the decline in molecular weight of the composite after artificial degradation. The decrease in thermal stability of composites during weathering may also be the result of polymer chain scission along with degradation of both the fiber and fiberematrix interfacial bond [59,60]. Fig. 10a shows the DSC thermograms for all conditions. A sample in the “as received” condition indicated a glass transition temperature (Tg) of 83.29 C. Fig. 10b presents the Tg for degraded composites. After 180 days of degradation the Tg was slightly lower than in other conditions. This decrease in Tg is due to the plasticizing effect of moisture absorbed during degradation. These effects may be reversible if exposure time is short. However, when exposure to moisture is associated with temperature changes and occurs for extended periods, the effects may be irreversible [61]. Moisture
160
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
absorption in polymers usually causes plasticization and hydrolytic effects that are either reversible or irreversible. The modulus and glass transition temperature of epoxy polymers often decline after moisture absorption [57,62]. 3.4. Morphological observations The polyurethane coating was the region most exposed to UVB radiation and moisture. Fig. 11 shows a top view of the coating under accelerated degradation, at 100 magnification. In order to better understand its actual morphological characteristics and degradation, the coating was not sanded or polished. Fig. 11 shows no significant changes in the polyurethane coating according to the progress of accelerated degradation under 100 magnification. The irregularities observed in the coating of the “as received” sample (Fig. 11a) may have occurred due its application by spray gun. Spray coating does not guarantee uniform surface roughness in all areas. Another region analyzed was the polyurethaneeepoxy interface. Fig. 12 shows the constituents of the cross-section of a composite in the “as received” condition. Fig. 13 shows cross-sections of the composite in degradation conditions, directly exposed to UVB
radiation and moisture, that is, located in the external layer of the composite. The images were taken at 100 magnification. Accelerated degradation did not change the polyurethaneeepoxy interface in the external cross-section of the composite, nor did the polyurethane coating vary during accelerated aging. These observations indicate stability in these regions in the face of photochemical attack and thermal and hygrothermal aging caused by the chamber. Finally, the fiber-matrix interface was also evaluated (Fig. 14aee). Fiber-matrix interface degradation occurred in the external cross-section of the composite according to the degradation progress. The UVB radiation and moisture cycles affected the fiber-matrix bond, resulting in loss of adhesion. According to Mariatti & Chum [63] and Sethi & Ray [64], this suggests that moisture absorption occurred by volumetric diffusion (absorption kinetics generally follow Fick's law). Therefore, the increase in the attenuation coefficient and decrease in acoustic wave speed as accelerated degradation advanced may be associated with reduced elastic properties and the formation of damaged areas. These were found to be related to loss of adhesion between the fiber and the matrix, as shown in
Fig. 14. External cross-section of composites at: (a) as received, (b) 45, (c) 90 (d) and 135 (e) 180 days of degradation (1000 magnification). (f) Internal cross-section of a specimen after 180 days of degradation (1000 magnification).
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
161
Fig. 15. (a) Degree of chalking of the polyurethane coating during degradation and (b) visual standard for the degree of chalking [14].
these micrographs. The effects of degradation on internal sections of the composite were also analyzed, since degradation can spread through the material. Fig. 14f shows an internal cross-section (10 mm from the external section) after 180 days of degradation. Morphological observations in the internal cross-section up to the final stage were similar to those seen in the non-degraded condition (Fig. 12). This confirms the tensile test results, whereby the effect of aging occurred mainly on the exposed surface. The other degradation conditions also showed no change in the internal region. Fig. 15a shows the distribution of the degree of chalking in the polyurethane coating caused by accelerated degradation. A scale between 1 and 5 was assigned according to the degree of chalking by comparing it with visual standards (Fig. 15b). Fig. 15 shows that the polyurethane coating released constituents in the form of fine, loosely adherent dust, resulting from the accelerated degradation caused by the chamber. Coating detachment increased linearly as a function of degradation. With this measurement technique, the degree of chalking was shown to be complementary to optical microscopy, which did not detect any changes in the polyurethane coating during accelerated degradation. This polyurethane coating detachment is corroborated by the literature [13].
thermal and morphological behavior of the materials. Thinner composites showed a reduction in modulus of elasticity and tensile strength, indicating that degradation was predominantly superficial. Aging reduced the Izod impact strength of the 3 and 6 mmthick composites. The TGA technique resulted in a 4.9% drop in maximum temperature degradation when compared with the “as received” and 180-day degradation conditions. There were no significant changes in residual weight with accelerated degradation. The glass transition temperature (Tg) after 180 days of degradation was slightly lower than that recorded in the other conditions. With respect to morphological behavior using optical microscopy, there were no significant changes in the polyurethane coating or polyurethaneeepoxy interface (directly exposed to the UVB radiation sources) with accelerated aging. However, changes were observed in the fiber-matrix interface of the exposed cross-section, with the internal cross-section showing stability. Measuring the degree of chalking was a complementary technique used to characterize the morphological effects on the polyurethane coating caused by the accelerated aging chamber. The polyurethane coating exhibited an increase in the degree of chalking with accelerated aging, reaching level 4 after 180 days of degradation. Acknowledgments
4. Conclusions The present study evaluated the mechanical, thermal and morphological behavior during accelerated aging in three thicknesses of composite plates used in wind turbines. Prior to mechanical destructive testing, two acoustic emission techniques were performed to characterize damage areas resulting from degradation. The acoustic emission test was used to determine the acoustic properties of the degraded composites. The high attenuation coefficient observed in the multi-component and multi-interface materials was measured using this same technique. Two acoustic emission techniques (absolute maximum and superposition) produced similar values, demonstrating flexibility. Thinner composites exhibited random changes in acoustic wave propagation speed during degradation. Thick degraded composites showed a decrease in mean values when compared with the “as received” condition, indicating deterioration. A slight increase in the attenuation coefficient for accelerated degradation also indicated damage. It was found that accelerated aging influenced the mechanical,
The authors thank Tecsis Company for providing samples of composite materials. References [1] Y. Pourvais, P. Asgari, A.R. Moradi, O. Rahmani, Experimental and finite element analysis of higher order behaviour of sandwich beams using digital , Polym. Test. 38 (2014) 7e17. http://dx.doi.org/10.1016/j. projection moire polymertesting.2014.06.001. [2] J. Jakobsen, M. Jensen, J.H. Andreasen, Thermo-mechanical characterisation of in-plane properties for CSM E-glass epoxy polymer composite materials e Part 1: thermal and chemical strain, Polym. Test. 32 (2013) 1350e1357. http:// dx.doi.org/10.1016/j.polymertesting.2013.08.010. [3] J. Jakobsen, M. Jensen, J.H. Andreasen, Thermo-mechanical characterisation of in-plane properties for CSM E-glass epoxy polymer composite materials e Part 2: Young's modulus, Polym. Test. 32 (2013) 1417e1422. http://dx.doi.org/ 10.1016/j.polymertesting.2013.09.005. [4] V. Trappe, S. Günzel, M. Jaunich, Correlation between crack propagation rate and cure process of epoxy resins, Polym. Test. 31 (2012) 654e659, http:// dx.doi.org/10.1016/j.polymertesting.2012.03.007. [5] A. Mlyniec, J. Korta, R. Kudelski, T. Uhl, The influence of the laminate thickness, stacking sequence and thermal aging on the static and dynamic behavior of carbon/epoxy composites, Compos. Struct. 118 (2014) 208e216. http://dx.doi.
162
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
org/10.1016/j.compstruct.2014.07.047. [6] B. Yang, J. Zhang, L. Zhou, M. Lu, W. Liang, Z. Wang, Effect of fiber surface modification on water absorption and hydrothermal aging behaviors of GF/ pCBT composites, Compos. Part B 82 (2015) 84e91. http://dx.doi.org/10.1016/ j.compositesb.2015.08.056. lito, A.H.M. Gonz [7] P.H.P. D'Alpino, M.V.G. Vismara, L.M.M. Mello, V. Hipo alez, C.F.O. Graeff, Resin composite characterizations following a simplified protocol of accelerated aging as a function of the expiration date, J. Mech. Behav. Biomed. Mater. 35 (2014) 59e69. http://dx.doi.org/10.1016/j.jmbbm.2014.03. 015. [8] H. Makki, K.N.S. Adema, E.A.J.F. Peters, J. Laven, L.G.J. van der Ven, R.A.T.M. van Benthem, G. With, A simulation approach to study photo-degradation processes of polymeric coatings, Polym. Degrad. Stab. 105 (2014) 68e79. http:// dx.doi.org/10.1016/j.polymdegradstab.2014.03.040. [9] K.I.M. Silva, M.J.L. Santos, M.P. Gil, Dependence of the photodegradation rate on the crystalline portion of PE films obtained through in situ polymerization in the presence of TiO2 nanospheres, nanoribbons and microspheres, Polym. Degrad. Stab. 112 (2015) 78e85. http://dx.doi.org/10.1016/j.polymdegradstab. 2014.12.004. [10] V.C. Malshe, G. Waghoo, Weathering study of epoxy paints, Prog. Org. Coat. 51 (2004) 267e272, http://dx.doi.org/10.1016/j.porgcoat.2004.07.007. [11] V.C. Malshe, G. Waghoo, Chalk resistant epoxy resins, Prog. Org. Coat. 51 (2004) 172e180, http://dx.doi.org/10.1016/j.porgcoat.2004.07.018. [12] S. Bhargavaa, M. Kubotaa, R.D. Lewis, S.G. Advani, A.K. Prasadb, J.M. Deitzel, Ultraviolet, water, and thermal aging studies of a waterborne polyurethane elastomer-based high reflectivity coating, Prog. Org. Coat. 79 (2015) 75e82. http://dx.doi.org/10.1016/j.porgcoat.2014.11.005. [13] F. Fragata, C.C. Amorim, A.P. Ordine. Desempenho de tintas de acabamento s radiaço ~ es ultravioleta, UVA e UVB, e condensaça ~o de umidade. frente a ~o Protecç~ Corrosa ao Mater., 29(3), 91e98. Retrieved September 03, 2015, from http://www.scielo.mec.pt/scielo.php?script¼sci_arttext&pid¼S087011642010000300002&lng¼en&tlng¼pt. ~ as, Study of damage and repair of blades of [14] J.C. Marín, A. Barroso, F. París, J. Can a 300 kW wind turbine, Energy 33 (2008) 1068e1083, http://dx.doi.org/ 10.1016/j.energy.2008.02.002. [15] H.G. Lee, M.G. Kang, J. Park, Fatigue failure of a composite wind turbine blade at its root end, Compos. Struct. 133 (2015) 878e885. http://dx.doi.org/10. 1016/j.compstruct.2015.08.010. [16] N.Z. Karimi, G. Minak, P. Kianfar, Analysis of damage mechanisms in drilling of composite materials by acoustic emission, Compos. Struct. 131 (2015) 107e114. http://dx.doi.org/10.1016/j.compstruct.2015.04.025. [17] R.K. Miller, P. McIntire, Nondestructive Testing Handbook, in: Acoustic Emission Testing, Vol. 5, American Society for Nondestructive Testing, 4153 Arlingate Plaza, Caller# 28518, Columbus, Ohio 43228, USA, 1987, p. 603. [18] R.H. Bossi, V. Giurgiutiu, 15 e Nondestructive testing of damage in aerospace composites, in: P.E. Irving, C. Soutis (Eds.), Polymer Composites in the Aerospace Industry, Woodhead Publishing, 2015, pp. 413e448. [19] J. Summerscales, Non-destructive Testing of Fibre-reinforced Plastics Composites, Springer, 1990. [20] M. Zoghi, The International Handbook of FRP Composites in Civil Engineering, CRC Press, 2013. [21] ASTM D 638M e 81, Standard Test Method for Tensile Properties of Plastics (Metric). [22] ASTM D256 e 10, Standard Test Method for Determining the Izod Pendulum Impact Resistance of Plastics. [23] ASTM G154 e 06, Standard Practice for Operating Fluorescent Light Apparatus for UV Exposure of Nonmetallic Materials. [24] ASTM E976 e 10, Standard Guide for Determining the Reproducibility of Acoustic Emission Sensor Response. [25] ASTM E1067 e 07, Standard Practice for Acoustic Emission Examination of Fiberglass Reinforced Plastic Resin (FRP) Tanks/Vessels. [26] X. Wang, M.X. Su, X.S. Cai, Effects of Material Viscosity on Particle Sizing by Ultrasonic Attenuation Spectroscopy, Proc. Eng. 102 (2015) 256e264, http:// dx.doi.org/10.1016/j.proeng.2015.01.141. [27] Nondestructive Testing HandbookUltrasonic Testing, Vol. 7, ASTM, New York, 1991. [28] P.G. Normando, E.P. Moura, J.A. Souza, S.S.M. Tavares, L.R. Padovese, Ultrasound, eddy current and magnetic Barkhausen noise as tools for sigma phase detection on a UNS S31803 duplex stainless steel, Mater. Sci. Eng. A 527 (2010) 2886e2891, http://dx.doi.org/10.1016/j.msea.2010.01.017. [29] DIN EN ISO 4628-6:2002e02. Paints and varnishes e evaluation of degradation of paint coatings e Designation of intensity, quantity and size of common types of defect. Part 6 Rat. degree chalking by tape method (ISO 4628-6 1990). [30] M.G.R. Sause, S. Horn, Influence of specimen geometry on acoustic emission signals in fiber reinforced composites, in: Fem-simulations and Experiments, Conference Proceedings: 29th European Conference on Acoustic Emission Testing, Vienna, Austria, 2010. [31] M.A. Hamstad, On lamb modes as a function of acoustic emission source rise time, J. Acoust. Emiss. 28 (2010) 41e57. [32] R. Gutkin, C.J. Green, S. Vangrattanachai, S.T. Pinho, P. Robinson, P.T. Curtis, On acoustic emission for failure investigation in CFRP: pattern recognition and peak frequency analyses, J. Mech. Syst. Signal Process 25 (2011) 1393e1407. [33] R. Li, T. Natsuki, Q.-Q. Ni, A novel dynamic stress analysis in bimaterial composite with defect using ultrasonic wave propagation, Compos. Struct. 132 (2015) 255e264. http://dx.doi.org/10.1016/j.compstruct.2015.04.046.
[34] R.K. Miller, E.K. Hill, P.O. Moore, Testing ASfN. Acoustic Emission Testing, American Society for Nondestructive Testing, 2005. [35] M.R. Gorman, Plate wave acoustic emission, J. Acoust. Soc. Am. 90 (1) (1991) 358e364. [36] M. Eaton, M. May, C. Featherston, K. Holford, S. Hallet, R. Pullin, Characterisation of damage in composite structures using acoustic emission, J. Phys. Conf. Ser. (2011) 012086. [37] D. Aljets, A. Chong, S. Wilcox, K. Holford, Acoustic Emission Source Location in Plate-like Structures Using a Closely Arranged Triangular Sensor Array, Uni Of Glamorgan, Pontypridd, Wales, UK CF37 1DL, 2010. [38] A. Oskouei, M. Ahmadi, M. Hajikhani, Wavelet-based acoustic emission characterization of damage mechanism in composite materials under mode I delamination at different interfaces, Exp. Polym. Lett. 3 (12) (2009) 804e813. [39] T.H. Loutas, V. Kostopoulos, Health monitoring of carbon/carbon, woven reinforced composites. Damage assessment by using advanced signal processing techniques. Part I: acoustic emission monitoring and damage mechanisms evolution, Compos. Sci. Technol. 69 (2) (2009) 265e272. [40] V. Arumugam, S. Sajith, A.J. Stanley, Acoustic emission characterization of failure modes in GFRP laminates under mode I delamination, J. Nondestruct. Eval. 30 (3) (2011) 213e219. [41] Q.-Q. Ni, M. Iwamoto, Wavelet transform of acoustic emission signals in failure of model composites, Eng. Fract. Mech. 69 (6) (2002) 717e728. [42] C.R. Ramirez-Jimenez, N. Papadakis, N. Reynolds, T.H. Gan, P. Purnell, M. Pharaoh, Identification of failure modes in glass/polypropylene composites by means of the primary frequency content of the acoustic emission event, Compos. Sci. Technol. 64 (12) (2004) 1819e1827. [43] W. Haselbach, B. Lauke, Acoustic emission of debonding between fibre and matrix to evaluate local adhesion, Compos. Sci. Technol. 63 (15) (2003) 2155e2162. [44] K. Asamene, L. Hudson, M. Sundaresan, Influence of attenuation on acoustic emission signals in carbon fiber reinforced polymer panels, Ultrasonics 59 (2015) 86e93. http://dx.doi.org/10.1016/j.ultras.2015.01.016. [45] T. Lundin, S.M. Cramer, R.H. Falk, C. Felton, Accelerated weathering of natural fiber-filled polyethylene composites, J. Mater. Civ. Eng. 16 (2004) 547e555. [46] A.I. Mourad, Thermo-mechanical characteristics of thermally aged polyethylene/polypropylene blends, Mater. Des. 31 (2010) 918e929. [47] L.C. Mendes, E.S. Rufino, F.O.C. de Paula, A.C. Torres, Mechanical, thermal and microstructure evaluation of HDPE after weathering in Rio de Janerio city, Polym. Degrad. Stab. 79 (2003) 371e383. [48] M. Philip, J. Attwood, A. Hulme, G. Williams, P. Shipton, Evaluation of Weathering in Mixed Polyethylene and Polypropylene Products, Report (UK), The Waste & Resources Action Programme (WRAP), 2004. ISBN: 1-84405092-0. [49] B.Y. Zhao, X.W. Yi, R.Y. Li, P.F. Zhu, K.A. Hu, Characterization to the weathering extent of LLDPE/LDPE thin film, J. Appl. Polym. Sci. 88 (2003) 12e16. [50] S.M. Al-Salem, G. Abraham, O.A. Al-Qabandi, A.M. Dashti, Investigating the effect of accelerated weathering on the mechanical and physical properties of high content plastic solid waste (PSW) blends with virgin linear low density polyethylene (LLDPE), Polym. Test. 46 (2015) 116e121. http://dx.doi.org/10. 1016/j.polymertesting.2015.07.008. [51] K. Rajakumar, V. Sarasvathy, A. ThamaraiChelvan, R. Chitra, C.T. Vijayakumar, Natural weathering studies of polypropylene, J. Polym. Environ. 17 (2009) 191e202. [52] D. Briassoulis, A. Aristopoulou, M. Bonora, I. Verlodt, Degradation characterisation of agricultural low-density polyethylene films, Biosyst. Eng. 88 (2004) 131e143. [53] A.I.A. Torik, H. Shirakawa, S. Nagaya, Photodegradation of polyethylene: factors affecting photostability, J. Appl. Polym. Sci. 40 (1990) 1637e1646. [54] Li Chang Sheng, M. Zhan, X. Huang, H. Zhou, Degradation behaviour of ultrahigh molecular weight polyethylene fibers under artificial accelerated weathering, Polym. Test. 31 (2012) 938e943. [55] T. Ojeda, A. Freitas, K. Birck, E. Dalmolin, R. Jacques, F. Bento, F. Camargo, Degradability of linear polyolefins under natural weathering, Polym. Degrad. Stab. 96 (2011) 703e707. [56] M. Sanchez-Soto, A. Rossa, A.J. Sanchez, J. Gamez-Perez, Blends of HDPE wastes: study of the properties, Waste Manag. 28 (2008) 2565e2573. [57] N. Shanmugam, I.A. Hussein, A. Badghaish, A.N. Shuaib, S.A. Furquan, M.H. AlMehthel, Evaluation of oil fly ash as a light stabilizer for epoxy composites: accelerated weathering study, Polym. Degrad. Stab. 112 (2015) 94e103. http://dx.doi.org/10.1016/j.polymdegradstab.2014.12.016. [58] R.S.C. Woo, Y.H. Chen, H.G. Zhu, J. Li, J.K. Kim, C.K.Y. Leung, Environmental degradation of epoxy-organoclay nanocomposites due to UV exposure and moisture attack: part I, Photodegr. Compos Sci. Technol. 67 (2007) 3448e3456. [59] M.D.H. Beg, K.L. Pickering, Accelerated weathering of unbleached and bleached Kraft wood fibre reinforced polypropylene composites, Polym. Degrad. Stab. 93 (2008) 1939e1946, http://dx.doi.org/10.1016/ j.polymdegradstab.2008.06.012. [60] M.M. Thwe, K. Liao, Durability of bambooeglass fiber reinforced polymer matrix hybrid composites, Compos. Sci. Technol. 63 (2003) 375e387. Ne ve, M.E.R. Shanahan, Water absorption by an epoxy resin and its effect [61] B. De on the mechanical properties and infra-red spectra, Polymer 34 (1993) 5099e5105. [62] B.G. Kumar, R.P. Singh, T. Nakamura, Degradation of carbon fiber-reinforced epoxy composites by ultraviolet radiation and condensation, J. Compos
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163 Mater 36 (2002) 2713e2721. [63] M. Mariatti, P.K. Chum, Effect of laminate configuration on the properties of glass fiber-reinforced plastics (GFRPs) mixed composites, J. Reinf. Plast. Compos. 24 (2005) 1713e1721.
163
[64] S. Sethi, B.C. Ray, Environmental effects on fibre reinforced polymeric composites: evolving reasons and remarks on interfacial strength and stability, Adv. Colloid Interface Sci. 217 (2015) 43e67. http://dx.doi.org/10.1016/j.cis. 2014.12.005.
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Effects of accelerated aging on mechanical, thermal and morphological behavior of polyurethane/epoxy/fiberglass composites Alexandre de Souza Rios a, b, *, Wanderley Ferreira de Amorim Júnior c, Elineudo Pinho de Moura a, Enio Pontes de Deus a, Judith Pessoa de Andrade Feitosa a , Campus do Pici, Fortaleza, 60440-554, Brazil Department of Metallurgical and Materials Engineering, Federal University of Ceara LMT (ENS Cachan, CNRS, Universit e Paris Saclay, Universit e Paris 6), 61, av. du Pr esident Wilson, 94235, Cachan, France c Department of Mechanical Engineering, Federal University of Campina Grande, Campina Grande, 58.429-140, Brazil a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 November 2015 Accepted 11 January 2016 Available online 12 January 2016
Wind blades, an important application of polymeric composite materials, are subject to natural weathering. This study aims to evaluate mechanical, thermal and morphological behavior during accelerated aging in three thicknesses of epoxy and fiberglass polyurethane-coated composite plates used in wind turbines, in addition to testing with two acoustic emission techniques. An accelerated aging chamber simulated natural weathering mechanisms for 45, 90, 135 and 180 days. This degradation primarily reduced the mechanical properties of the thinner composites, with some damaged specimens exhibiting fiber-matrix debonding. Thermal properties deteriorated. There were no morphological changes on the polyurethaneeepoxy interface; however, degradation occurred in the fiber-matrix interface on the surface exposed to radiation. The degree of chalking indicated coating deterioration on the external surface of the polyurethane. The acoustic wave propagation speed and attenuation coefficient measured prior to mechanical testing indicated the presence of damage areas. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Accelerated aging Acoustic emission Damage Wind turbine
1. Introduction In recent decades, there has been a growing demand to exploit global wind potential, transforming wind blades into one of the most important applications of polymeric composite materials. Some wind turbine components, primarily wind blades, are manufactured using polymeric composite materials [1e4]. Chemical or environmental aging is caused by various agents, such as humidity, loading conditions and ultraviolet radiation (UV), leading to irreversible changes in the molecular structure of these materials [5e7]. The polymeric matrix is usually more susceptible to aging and generally controls the long-term performance of the composite [5,6]. The critical portion of UV radiation causes photo initiation of the polymer due to the absorption of chromophores present in the matrix (hydroperoxides, catalyst residues, carbonyls, unsaturation). Excited chromophores induce photooxidative decomposition of macromolecular chains, leading to basic behavior changes in the polymer and
* Corresponding author. Department of Metallurgical and Materials Engineering, , Campus do Pici, Fortaleza, 60440-554, Brazil. Federal University of Ceara E-mail addresses: [email protected], [email protected], eng.alexandrerios@ gmail.com (A. de Souza Rios). http://dx.doi.org/10.1016/j.polymertesting.2016.01.010 0142-9418/© 2016 Elsevier Ltd. All rights reserved.
its properties. Photoinduced processes usually change the appearance (surface gloss, color) and mechanical properties (strength, strain, flexibility) of the polymer [8,9]. The combined action of all these factors on the behavior and durability of composites is a highly complex phenomenon that occurs at the molecular level. When a polymer is subjected to ultraviolet radiation it releases constituents in the form of thin, loosely adherent dust. This coating defect is called chalking [10,11]. Coatings are expected to be durable and retain their properties over time. As such, resistance to light, humidity and temperature is a general requirement. The environmental impact of coatings can be further reduced by increasing their efficiency and useful life. The most common cause of coating degradation is ambient exposure to ultraviolet (UV) radiation, water/humidity and temperature fluctuation. In many cases, degradation is evaluated based on changes in chemical structure and the presence of foreign chemicals in the system as a function of time. Polyurethane coatings made with an aliphatic and cycloaliphatic isocyanate-based curing agent exhibit excellent resistance to ultraviolet rays as well as good color retention and glow when exposed to natural weathering. Furthermore, they show good resistance to chalking [12]. A controlled artificial weathering test should replicate
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
conditions typically found in the natural aging process through cycles representing day and night, and periods of surface drying and wetting. The advantage of this method is the ability to accelerate the testing of all controlled parameters, obtaining comparable results at significantly lower exposure times [13]. Possible causes of failures in composite materials used to manufacture wind turbines are an abrupt change in blade thickness with tension concentrators and decoupling of the resins or delamination. Decoupling resins may also be associated with processing defects in manual lamination. Both the individual and collective action of these defects can cause premature wind blade failure [14]. Debonding of the outer skin was the initial failure mechanism, followed by delamination buckling which led to the blade's collapse [15]. Tests using a full-scale wind turbine blade to study structural fatigue behavior are costly, and hence few studies have been conducted to date [15]. Another approach to improve the reliability of wind turbine blades is to evaluate the mechanical properties of composite laminates. Non-destructive testing, such as acoustic emission (AE), is one of the most appropriate to characterize composite materials. AEs are transient ultrasonic waves generated by sudden movement in material under stress [16e18]. When a component is subjected to mechanical load, discontinuities in materials may release AE energy. The electrical signals are then amplified and further processed as AE signal data. Accurate processing of the AE signals can identify different damage mechanisms in composites [19,20]. Therefore, the present study aims to evaluate the mechanical, morphological and thermal properties of artificially degraded composite plates. The properties identified in different analyses provide knowledge on the durability and structural applications of composites exposed to environmental conditions. The signals obtained from acoustic emission provide information on the formation of discontinuities or voids arising from degradation. 2. Experimental 2.1. Materials Composite plates (1, 3 and 6 mm thick) manufactured by hand lay-up were used. These consist of 989 Biax Saertex 45 /-45 fiberglass fabric layers, Hexion R1M135 þ 1366 epoxy resin and DM Coating-sprayed polyurethane. For degradation cycles the plates were machined according to the tensile test specimen (ASTM D638) and Izod impact specimen (ASTM D256-10) [21,22]. 2.2. Composite characterization The effects of sunlight were simulated by a system of 8 UVB radiation sources in the 280e320 nm range using an UV-C AdeximConexim® accelerated aging chamber. Specimens were exposed to a UVB radiation cycle at 60 C for 6 h, followed by exposure to 6 h of condensing water vapor at 50 C and 100% humidity according to ASTM G154-06 [23] (Fig. 1). Specimens were removed every 1080 h (45 days) and four degradation conditions were evaluated: 45, 90, 135 and 180 days. A Physical Acoustics Corporation® sensor (model ASP05F0021) was used for the acoustic emission test. A mechanical stimulus was provided by breaking graphite (Fig. 2), according to ASTM E976-10 and ASTM E1067-07 [24,25]. The absolute maximum and superposition methods were used. 2.2.1. Absolute maximum method The maximum peak time values of each sensor were associated with distance (x) to provide propagation speed. Acoustic signal attenuation (a), expressed in dB/m, was measured using attenuated acoustic intensity (I) and non-attenuated intensity (Io), as described
153
Fig. 1. Specimens in the accelerated degradation chamber.
in Equation (1) [26]:
1 Io I ¼ Io$eax /a ¼ $ln x I
(1)
2.2.2. Superposition or cross-correlation method The time-of-flight (t0) of the acoustic wave between two adjacent signals B1 (t) and B2 (t þ t0), can be used to determine the wave speed through the specimen. Thus, propagation speed can be determined with t0 and the distance between sensors [27,28]. Time-of-flight was estimated by the cross-correlation or superposition method. The value of t0 is computed as the value of t, which is maximum in Equation (2):
∞ Z B1 ðtÞ$B2 ðt tÞdt
(2)
∞
The cross-correlation of two functions (B1 and B2) involves shifting B2 in t toward B1 and multiplying them to determine for which value of t (t0) the product is maximized. This means that the maximum of the integral of the cross-correlation between echoes B1 and B2, as a function of time delay, corresponds to their maximum correlation [28]. An EMIC® DL 10000/700 universal testing machine with a 100 kN load capacity was used for tensile testing. The deformation rate was 50 mm/min. Composite plates were tested by Izod impact testing in each degradation condition according to ASTM D256-10 [22]. According to this standard, the minimum thickness is 3 mm. A CEAST® Resil 5.5, 5.5 J test hammer and impact velocity of 3.46 m/s were used in this test, with a type B, V-shaped notch and a 45 angle. Thermogravimetric analysis (TGA Q 50®) was carried out at temperatures ranging from 25 C to 800 C in air, with a heating rate of 10 C min1 and flow rate of 100 mL min1, using about 10 mg for each sample. Differential Scanning Calorimetry (DSC Q 20®) was employed to determine the glass transition temperature (Tg) of the composites. A heating rate of 10 C/min at 50 C to 500 C was applied to 10 mg samples under nitrogen. The sample constituents and their interfaces were viewed under an Olympus optical microscope (model BX 51 M) at magnifications of 50, 100, 200, 500 and 1000. Cross-sections were made for sample preparation using a water jet cutter and water-only cutting, followed by sanding with 300, 600 and 1200 grit sandpaper, and mechanical polishing with diamond paste. Both procedures were performed using a Teclago® PL 02 polishing machine. The polyurethane coating was directly exposed to UVB radiation
154
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 2. Acoustic emission test procedure: (a) graphite to be broken as mechanical stimulus to generate acoustic signals, (b) conducting the test and (c) mechanical stimulus region.
and detachment of constituents was evaluated according to DIN EN ISO 4628-6: 2002-02, which measures the degree of chalking in a polymer [29]. This technique consisted of applying a transparent adhesive on the polyurethane coating. After it was removed, the degree of chalking was determined by comparing visual standards, which can vary from 0 (absence) to five, where five corresponds to a significant amount of chalking. 3. Results and discussions 3.1. Acoustic emission tests The 1 mm-thick specimens exhibited excessive deformation after accelerated degradation and could not be coupled to the sensor. Hence, they were not considered in the results. Fig. 3 shows that the time required for the acoustic signal to travel between sensors was 55 ms (578 ms e 523 ms ¼ 55 ms). Given that the distance between sensors was 130 mm, the speed of the traveling wave must be 2363 m s1, according to the absolute maximum peak. The absolute peak decreased from 980 V (first sensor) to 6 107 V (second sensor). The frequency content of each component of the decomposed signals emitted during the mechanical stimulus is shown in Fig. 3. The results indicate that the frequency range of the decomposed components decreased from one sensor to another and show significant acoustic wave attenuation. Acoustical activity is also observed between 50 kHz and 500 kHz; however, the highest acoustic events occur between 50 kHz and 100 kHz. It should be noted that peak frequency analysis is primarily influenced by the sensor type used, composite thickness, mechanical stimulus, equipment bandwidth and depth position of the source [30e32]. Ideally, the distance between sensors should not be constant for all measurements and may play a role in peak frequency deviations. The aim here is to investigate damage areas from artificial degradation, presenting comparative information and relate the AE energy of the first and second sensors. Fig. 4 shows acoustic wave speeds using the absolute maximum and superposition methods on the 3 mm and 6 mm specimen degradation conditions. The results obtained by the absolute maximum method corroborate those recorded using the superposition method, showing flexibility. The thinner composites exhibited random modifications in acoustic wave propagation speed during degradation. The 6 mm-thick composites showed a decrease in mean values when compared against the non-degraded condition, possibly associated with loss of elastic properties or damage onset in the material during exposure. Different frequency distributions for dissimilar failures may be due to the fact that elastic acoustic velocities and intrinsic frequencies are associated with modulus of elasticity and density, according to Equation (3) [16]:
sffiffiffi E f fCf r
(3)
where f is frequency, C elastic acoustic velocity, r density and E is the modulus of elasticity. The wave speed in the air is lower than in the composite components, and declining propagation speed in the specimens may be related to the formation of voids due to accelerated degradation (matrix cracking, fiber breakage, fiber-matrix debonding, etc.) [33]. Considering significant differences in the position of the acoustic source, a high variation in frequency spectra would be expected when testing different specimens from the same laminate. In this study, at least six specimens per material were tested. The standard deviations found may be associated with the anisotropic nature of this material, which may contain scattered beams and cause variations in propagation speed. Fig. 5 shows the mean attenuation coefficient values in 3 and 6 mm-thick specimens. Except for the specimens degraded for 45 days, there was a slight increase in attenuation coefficients in both geometries. The initiation and propagation of discontinuities generate voids with a higher attenuation coefficient than that of the composite in the “as received” condition, demonstrating damage formation. In plate type structures, stress waves generated by damage events propagate as combinations of different Lamb wave modes [34e37]. The presence of each type of Lamb wave mode and its frequency content is determined by the location and type of the damage event. In composite test specimens, results from previous studies [38e40] indicated the occurrence of AE signals in the frequency range between 100 kHz and 500 kHz. In some cases, higher order modes occur near the higher end of this range. These higher order modes are either too small or undetectable by commercially available AE sensors. Thus, experimental AE signals in composite materials in this range are likely to be combinations of the fundamental modes. Numerous attempts have been made in the literature to classify AE signals according to the likely failure modes that generate them [40e43]. However, definitive relationships between the different failure modes and the features of resulting AE signals have yet to be established. Such correlations would only be feasible if distinguishing features such as the ratio of different modes and different frequency components are preserved in the signals as they propagate along the composite laminates. Should these features not be preserved during the propagation of AE related waves, attempts to relate features of AE signals to composite failure modes would fail. The characteristics of AE signal waveforms in terms of individual wave mode amplitudes and frequency components are important for identifying types of failure modes. Both the amplitude and
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
155
Fig. 3. Signal detected by the (a) first and (b) second sensors after mechanical stimulus.
Fig. 4. Acoustic wave speeds using the absolute maximum and superposition methods on the (a) 3 mm- and (b) 6 mm-thick specimens under accelerated degradation.
frequency content of AE signals are altered by different attenuation mechanisms as the waves propagate in the laminates. Geometric spreading and frequency-dependent dispersion reduce the
amplitude of AE signals. For example, the viscoelastic nature of the matrix in carbon/polymer composites introduces significantly higher levels of attenuation [44].
156
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 5. Attenuation coefficient of the 3 mm- and 6 mm-thick specimens.
3.2. Mechanical properties 3.2.1. Tensile testing Fig. 6 presents the modulus of elasticity, tensile strength and strain at break of tensile testing in 1 mm, 3 mm and 6 mm-thick composites under the conditions assessed. There was a 10.0% reduction in modulus of elasticity between the “as received” condition and after 180 days degradation for 1 mm-thick composites
(8.39 GPae7.55 GPa) and a 4.8% decline over this degradation range for 3 mm-thick composites (9.05 MPae8.62 MPa). By contrast, the 6 mm-thick specimens showed a decrease in modulus of elasticity up to 90 days of degradation, followed by an increase until the final stage of degradation. The deterioration of this elastic property was indicated, but not quantitatively measured, by the acoustic measurements presented and discussed above. These results show that the thinner specimens are more susceptible to accelerated degradation imposed by the chamber, indicating a more pronounced reduction in modulus of elasticity at the surface. The modulus of elasticity is dependent on the crystallinity of the polymer [45,46]. UV irradiation leads to photodegradation of the polymer via a well-established mechanism previously reported by a number of authors [45,47e49]. The degradation of polymers is initiated with chain scission, followed by branching, cross-linking and, finally, the formation of double bonds. One of the irreversible alterations caused by photodegradation is the loss of amorphous regions due to cross-linking, to which can be attributed the change in the modulus of elasticity of samples [50]. The three thicknesses exhibited a decrease in tensile strength after 180 days of degradation: 17.3% for the 1 mm-thick composite (53.71 MPae44.41 MPa), 10.4% for the 3-mm thick composite (52.97 MPae47.44 MPa) and 14.8% for the 6 mm-thick composite (60.78 MPae51.78 MPa). The 1 mm-thick composites were more affected by accelerated degradation, indicating that the decrease in tensile strength was also more pronounced at the surface.
Fig. 6. Mechanical properties of the composites: (a) modulus of elasticity (b) tensile strength and (c) strain at break.(d) Izod impact test results of 3 mm-thick specimens in different conditions.
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
157
Fig. 7. Visual appearance of the specimens after the Izod impact test.
Fig. 8. (a) Delaminated region of a 6 mm-thick specimen degraded for 90 days and not ruptured after the Izod impact test (50 magnification). (b) Cracked and delaminated region oriented at 45 of a 6 mm-thick specimen degraded for 90 days and not ruptured after the Izod impact test (50 magnification).
The opposite effect occurs for strain at break. Thicker composites experienced a greater reduction in strain at break because their modulus of elasticity (stiffness) did not suffer major damage. Overall, the tensile test results obtained in this study indicated that degradation was predominantly superficial, having a greater effect
Table 1 Residual weight and maximum degradation temperature of the composite from TGA measurements.
Fig. 9. TGA thermogram of conditions evaluated.
Conditions
Residual weight [%]
Maximum degradation temperature [ C]
Sample 1
Sample 2
Mean
Sample 1
Sample 2
Mean
As received 45 days 90 days 135 days 180 days
65.22 66.54 67.87 66.84 62.59
67.07 65.23 66.12 66.20 67.46
66.15 65.89 67.00 66.52 65.03
358.4 356.0 337.6 335.4 337.6
348.9 343.3 338.3 337.6 334.7
353.7 349.7 338.0 336.5 336.2
158
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
Fig. 10. (a) DSC thermogram and (b) Tg of degraded conditions.
Fig. 11. Polyurethane coating under the conditions evaluated: (a) as received, (b) 45, (c) 90 (d) 135 and (e) 180 days of degradation. 100 magnification.
on the mechanical properties of thinner composites. Several authors have shown strain at break to be a very useful measurement of polymer degradation [51e54]. A similar trend was observed for the other polymers tested. This behavior can be related to chain scission in semi-crystalline polymers, which leads to a decrease in the molecular weight of the material and, consequently, a loss in strain break with weathering [51,54e56]. According to the literature, this property follows a decreasing trend proportional to artificial weathering [50]. Some results obtained using this technique showed high
Fig. 12. External cross-section section of the ''as received'' composite. 100 magnification.
standard deviations, dispersion and non-linearity of the mechanical properties as a function of degradation stages. This behavior may be due to composite processing associated with possible nonuniform degradation within the accelerated aging chamber.
3.2.2. Izod impact testing Fig. 7 shows 3 mm and 6 mm-thick specimens after the Izod impact test in each condition evaluated. After 45 days of exposure, the surfaces of samples exhibited a distinct change in color from colorless to dark yellow, and colorless to light yellow in controls, with minor changes in surface roughness. The discoloration established that photooxidation resulted in the formation of chromophoric chemical species, which were absorbed in the visible range of light [57,58]. Fig. 7 shows an increase in deformation with exposure time in the accelerated aging chamber for the 3 mm-thick specimen after Izod impact testing. In the “as received” and 45 -day degradation conditions there was no complete rupture of the specimen, although delamination and permanent deformation were detected, especially after 45 days of degradation. Complete rupture was observed for the other degradation conditions and was more pronounced in specimens with greater exposure in the accelerated aging chamber. As occurred in the 6 mm-thick specimens, accelerated degradation decreased the fiber-matrix bond in 3 mm-thick specimens. The energy provided by a 5.5 J hammer was not sufficient to fracture the 6 mm-thick specimens. However, delaminated regions were detected near the notch of the degraded specimen, which did not occur in the “as received” condition. This defect is usually present in composite failures (Fig. 8). Even without fracturing,
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
159
Fig. 13. Cross-section of the composite at: (a) 45, (b) 90, (c) and 135 (d) and 180 days of degradation (100 magnification).
accelerated degradation decreased the fiber-matrix bond in the 6 mm-thick specimen. Fig. 8 shows the notched region of a 6 mm-thick specimen after 90 days of degradation not ruptured after the Izod impact test (50 magnification). There was more damage in the region close to the notch, a point of greatest stress concentration. Delamination and fiber-matrix detachment are evident. Fig. 8 shows the delamination path propagated between the interstices fibers, oriented at 45 . Fig. 6d shows the mean values and standard deviations of Izod impact test results for 3 mm-thick specimens. The impact strength of the composites varied with accelerated degradation time. Even without rupture of some samples, there was a decrease in impact resistance between the “as received” and 45-day degradation conditions from 1802.0 J/m to 1791.1 J/m. The conditions in which rupture occurred after impact testing (90, 135 and 180 days of degradation) showed decreased impact resistance, hence the aging effect of the aging chamber reduced the Izod impact strength of the composites and modified fiber-matrix bonding. 3.3. Thermal properties Fig. 9 shows the TGA thermograms of samples in all the conditions evaluated. Residual weight was between 60 and 70% and the maximum degradation temperature (Tm) was around 350 C for all conditions. Table 1 shows the residual weight and maximum temperature of two samples for each degradation condition evaluated. The maximum degradation temperature was obtained using derivative thermogravimetry (DTG), where curves are recorded
based on the TGA curve and correspond to the first derivative of weight change versus time (dm/dt), which is recorded as a function of temperature or time. Table 1 shows that residual mass obtained high values due to the presence of inorganic matter (fiberglass), whose decomposition temperature is higher than the maximum temperature used in this test. There were no significant changes in residual weight with progressive accelerated degradation because the same weight percentage decomposes irrespective of the accelerated degradation stage. However, there was a decrease in maximum temperature according to accelerated degradation stages. The reduction was approximately 5.0% at maximum degradation temperature when compared to the “as received” and 180-day degradation conditions (353.66 Ce336.14 C). The decrease in thermal stability with increasing aging time may be due to the decline in molecular weight of the composite after artificial degradation. The decrease in thermal stability of composites during weathering may also be the result of polymer chain scission along with degradation of both the fiber and fiberematrix interfacial bond [59,60]. Fig. 10a shows the DSC thermograms for all conditions. A sample in the “as received” condition indicated a glass transition temperature (Tg) of 83.29 C. Fig. 10b presents the Tg for degraded composites. After 180 days of degradation the Tg was slightly lower than in other conditions. This decrease in Tg is due to the plasticizing effect of moisture absorbed during degradation. These effects may be reversible if exposure time is short. However, when exposure to moisture is associated with temperature changes and occurs for extended periods, the effects may be irreversible [61]. Moisture
160
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
absorption in polymers usually causes plasticization and hydrolytic effects that are either reversible or irreversible. The modulus and glass transition temperature of epoxy polymers often decline after moisture absorption [57,62]. 3.4. Morphological observations The polyurethane coating was the region most exposed to UVB radiation and moisture. Fig. 11 shows a top view of the coating under accelerated degradation, at 100 magnification. In order to better understand its actual morphological characteristics and degradation, the coating was not sanded or polished. Fig. 11 shows no significant changes in the polyurethane coating according to the progress of accelerated degradation under 100 magnification. The irregularities observed in the coating of the “as received” sample (Fig. 11a) may have occurred due its application by spray gun. Spray coating does not guarantee uniform surface roughness in all areas. Another region analyzed was the polyurethaneeepoxy interface. Fig. 12 shows the constituents of the cross-section of a composite in the “as received” condition. Fig. 13 shows cross-sections of the composite in degradation conditions, directly exposed to UVB
radiation and moisture, that is, located in the external layer of the composite. The images were taken at 100 magnification. Accelerated degradation did not change the polyurethaneeepoxy interface in the external cross-section of the composite, nor did the polyurethane coating vary during accelerated aging. These observations indicate stability in these regions in the face of photochemical attack and thermal and hygrothermal aging caused by the chamber. Finally, the fiber-matrix interface was also evaluated (Fig. 14aee). Fiber-matrix interface degradation occurred in the external cross-section of the composite according to the degradation progress. The UVB radiation and moisture cycles affected the fiber-matrix bond, resulting in loss of adhesion. According to Mariatti & Chum [63] and Sethi & Ray [64], this suggests that moisture absorption occurred by volumetric diffusion (absorption kinetics generally follow Fick's law). Therefore, the increase in the attenuation coefficient and decrease in acoustic wave speed as accelerated degradation advanced may be associated with reduced elastic properties and the formation of damaged areas. These were found to be related to loss of adhesion between the fiber and the matrix, as shown in
Fig. 14. External cross-section of composites at: (a) as received, (b) 45, (c) 90 (d) and 135 (e) 180 days of degradation (1000 magnification). (f) Internal cross-section of a specimen after 180 days of degradation (1000 magnification).
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
161
Fig. 15. (a) Degree of chalking of the polyurethane coating during degradation and (b) visual standard for the degree of chalking [14].
these micrographs. The effects of degradation on internal sections of the composite were also analyzed, since degradation can spread through the material. Fig. 14f shows an internal cross-section (10 mm from the external section) after 180 days of degradation. Morphological observations in the internal cross-section up to the final stage were similar to those seen in the non-degraded condition (Fig. 12). This confirms the tensile test results, whereby the effect of aging occurred mainly on the exposed surface. The other degradation conditions also showed no change in the internal region. Fig. 15a shows the distribution of the degree of chalking in the polyurethane coating caused by accelerated degradation. A scale between 1 and 5 was assigned according to the degree of chalking by comparing it with visual standards (Fig. 15b). Fig. 15 shows that the polyurethane coating released constituents in the form of fine, loosely adherent dust, resulting from the accelerated degradation caused by the chamber. Coating detachment increased linearly as a function of degradation. With this measurement technique, the degree of chalking was shown to be complementary to optical microscopy, which did not detect any changes in the polyurethane coating during accelerated degradation. This polyurethane coating detachment is corroborated by the literature [13].
thermal and morphological behavior of the materials. Thinner composites showed a reduction in modulus of elasticity and tensile strength, indicating that degradation was predominantly superficial. Aging reduced the Izod impact strength of the 3 and 6 mmthick composites. The TGA technique resulted in a 4.9% drop in maximum temperature degradation when compared with the “as received” and 180-day degradation conditions. There were no significant changes in residual weight with accelerated degradation. The glass transition temperature (Tg) after 180 days of degradation was slightly lower than that recorded in the other conditions. With respect to morphological behavior using optical microscopy, there were no significant changes in the polyurethane coating or polyurethaneeepoxy interface (directly exposed to the UVB radiation sources) with accelerated aging. However, changes were observed in the fiber-matrix interface of the exposed cross-section, with the internal cross-section showing stability. Measuring the degree of chalking was a complementary technique used to characterize the morphological effects on the polyurethane coating caused by the accelerated aging chamber. The polyurethane coating exhibited an increase in the degree of chalking with accelerated aging, reaching level 4 after 180 days of degradation. Acknowledgments
4. Conclusions The present study evaluated the mechanical, thermal and morphological behavior during accelerated aging in three thicknesses of composite plates used in wind turbines. Prior to mechanical destructive testing, two acoustic emission techniques were performed to characterize damage areas resulting from degradation. The acoustic emission test was used to determine the acoustic properties of the degraded composites. The high attenuation coefficient observed in the multi-component and multi-interface materials was measured using this same technique. Two acoustic emission techniques (absolute maximum and superposition) produced similar values, demonstrating flexibility. Thinner composites exhibited random changes in acoustic wave propagation speed during degradation. Thick degraded composites showed a decrease in mean values when compared with the “as received” condition, indicating deterioration. A slight increase in the attenuation coefficient for accelerated degradation also indicated damage. It was found that accelerated aging influenced the mechanical,
The authors thank Tecsis Company for providing samples of composite materials. References [1] Y. Pourvais, P. Asgari, A.R. Moradi, O. Rahmani, Experimental and finite element analysis of higher order behaviour of sandwich beams using digital , Polym. Test. 38 (2014) 7e17. http://dx.doi.org/10.1016/j. projection moire polymertesting.2014.06.001. [2] J. Jakobsen, M. Jensen, J.H. Andreasen, Thermo-mechanical characterisation of in-plane properties for CSM E-glass epoxy polymer composite materials e Part 1: thermal and chemical strain, Polym. Test. 32 (2013) 1350e1357. http:// dx.doi.org/10.1016/j.polymertesting.2013.08.010. [3] J. Jakobsen, M. Jensen, J.H. Andreasen, Thermo-mechanical characterisation of in-plane properties for CSM E-glass epoxy polymer composite materials e Part 2: Young's modulus, Polym. Test. 32 (2013) 1417e1422. http://dx.doi.org/ 10.1016/j.polymertesting.2013.09.005. [4] V. Trappe, S. Günzel, M. Jaunich, Correlation between crack propagation rate and cure process of epoxy resins, Polym. Test. 31 (2012) 654e659, http:// dx.doi.org/10.1016/j.polymertesting.2012.03.007. [5] A. Mlyniec, J. Korta, R. Kudelski, T. Uhl, The influence of the laminate thickness, stacking sequence and thermal aging on the static and dynamic behavior of carbon/epoxy composites, Compos. Struct. 118 (2014) 208e216. http://dx.doi.
162
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163
org/10.1016/j.compstruct.2014.07.047. [6] B. Yang, J. Zhang, L. Zhou, M. Lu, W. Liang, Z. Wang, Effect of fiber surface modification on water absorption and hydrothermal aging behaviors of GF/ pCBT composites, Compos. Part B 82 (2015) 84e91. http://dx.doi.org/10.1016/ j.compositesb.2015.08.056. lito, A.H.M. Gonz [7] P.H.P. D'Alpino, M.V.G. Vismara, L.M.M. Mello, V. Hipo alez, C.F.O. Graeff, Resin composite characterizations following a simplified protocol of accelerated aging as a function of the expiration date, J. Mech. Behav. Biomed. Mater. 35 (2014) 59e69. http://dx.doi.org/10.1016/j.jmbbm.2014.03. 015. [8] H. Makki, K.N.S. Adema, E.A.J.F. Peters, J. Laven, L.G.J. van der Ven, R.A.T.M. van Benthem, G. With, A simulation approach to study photo-degradation processes of polymeric coatings, Polym. Degrad. Stab. 105 (2014) 68e79. http:// dx.doi.org/10.1016/j.polymdegradstab.2014.03.040. [9] K.I.M. Silva, M.J.L. Santos, M.P. Gil, Dependence of the photodegradation rate on the crystalline portion of PE films obtained through in situ polymerization in the presence of TiO2 nanospheres, nanoribbons and microspheres, Polym. Degrad. Stab. 112 (2015) 78e85. http://dx.doi.org/10.1016/j.polymdegradstab. 2014.12.004. [10] V.C. Malshe, G. Waghoo, Weathering study of epoxy paints, Prog. Org. Coat. 51 (2004) 267e272, http://dx.doi.org/10.1016/j.porgcoat.2004.07.007. [11] V.C. Malshe, G. Waghoo, Chalk resistant epoxy resins, Prog. Org. Coat. 51 (2004) 172e180, http://dx.doi.org/10.1016/j.porgcoat.2004.07.018. [12] S. Bhargavaa, M. Kubotaa, R.D. Lewis, S.G. Advani, A.K. Prasadb, J.M. Deitzel, Ultraviolet, water, and thermal aging studies of a waterborne polyurethane elastomer-based high reflectivity coating, Prog. Org. Coat. 79 (2015) 75e82. http://dx.doi.org/10.1016/j.porgcoat.2014.11.005. [13] F. Fragata, C.C. Amorim, A.P. Ordine. Desempenho de tintas de acabamento s radiaço ~ es ultravioleta, UVA e UVB, e condensaça ~o de umidade. frente a ~o Protecç~ Corrosa ao Mater., 29(3), 91e98. Retrieved September 03, 2015, from http://www.scielo.mec.pt/scielo.php?script¼sci_arttext&pid¼S087011642010000300002&lng¼en&tlng¼pt. ~ as, Study of damage and repair of blades of [14] J.C. Marín, A. Barroso, F. París, J. Can a 300 kW wind turbine, Energy 33 (2008) 1068e1083, http://dx.doi.org/ 10.1016/j.energy.2008.02.002. [15] H.G. Lee, M.G. Kang, J. Park, Fatigue failure of a composite wind turbine blade at its root end, Compos. Struct. 133 (2015) 878e885. http://dx.doi.org/10. 1016/j.compstruct.2015.08.010. [16] N.Z. Karimi, G. Minak, P. Kianfar, Analysis of damage mechanisms in drilling of composite materials by acoustic emission, Compos. Struct. 131 (2015) 107e114. http://dx.doi.org/10.1016/j.compstruct.2015.04.025. [17] R.K. Miller, P. McIntire, Nondestructive Testing Handbook, in: Acoustic Emission Testing, Vol. 5, American Society for Nondestructive Testing, 4153 Arlingate Plaza, Caller# 28518, Columbus, Ohio 43228, USA, 1987, p. 603. [18] R.H. Bossi, V. Giurgiutiu, 15 e Nondestructive testing of damage in aerospace composites, in: P.E. Irving, C. Soutis (Eds.), Polymer Composites in the Aerospace Industry, Woodhead Publishing, 2015, pp. 413e448. [19] J. Summerscales, Non-destructive Testing of Fibre-reinforced Plastics Composites, Springer, 1990. [20] M. Zoghi, The International Handbook of FRP Composites in Civil Engineering, CRC Press, 2013. [21] ASTM D 638M e 81, Standard Test Method for Tensile Properties of Plastics (Metric). [22] ASTM D256 e 10, Standard Test Method for Determining the Izod Pendulum Impact Resistance of Plastics. [23] ASTM G154 e 06, Standard Practice for Operating Fluorescent Light Apparatus for UV Exposure of Nonmetallic Materials. [24] ASTM E976 e 10, Standard Guide for Determining the Reproducibility of Acoustic Emission Sensor Response. [25] ASTM E1067 e 07, Standard Practice for Acoustic Emission Examination of Fiberglass Reinforced Plastic Resin (FRP) Tanks/Vessels. [26] X. Wang, M.X. Su, X.S. Cai, Effects of Material Viscosity on Particle Sizing by Ultrasonic Attenuation Spectroscopy, Proc. Eng. 102 (2015) 256e264, http:// dx.doi.org/10.1016/j.proeng.2015.01.141. [27] Nondestructive Testing HandbookUltrasonic Testing, Vol. 7, ASTM, New York, 1991. [28] P.G. Normando, E.P. Moura, J.A. Souza, S.S.M. Tavares, L.R. Padovese, Ultrasound, eddy current and magnetic Barkhausen noise as tools for sigma phase detection on a UNS S31803 duplex stainless steel, Mater. Sci. Eng. A 527 (2010) 2886e2891, http://dx.doi.org/10.1016/j.msea.2010.01.017. [29] DIN EN ISO 4628-6:2002e02. Paints and varnishes e evaluation of degradation of paint coatings e Designation of intensity, quantity and size of common types of defect. Part 6 Rat. degree chalking by tape method (ISO 4628-6 1990). [30] M.G.R. Sause, S. Horn, Influence of specimen geometry on acoustic emission signals in fiber reinforced composites, in: Fem-simulations and Experiments, Conference Proceedings: 29th European Conference on Acoustic Emission Testing, Vienna, Austria, 2010. [31] M.A. Hamstad, On lamb modes as a function of acoustic emission source rise time, J. Acoust. Emiss. 28 (2010) 41e57. [32] R. Gutkin, C.J. Green, S. Vangrattanachai, S.T. Pinho, P. Robinson, P.T. Curtis, On acoustic emission for failure investigation in CFRP: pattern recognition and peak frequency analyses, J. Mech. Syst. Signal Process 25 (2011) 1393e1407. [33] R. Li, T. Natsuki, Q.-Q. Ni, A novel dynamic stress analysis in bimaterial composite with defect using ultrasonic wave propagation, Compos. Struct. 132 (2015) 255e264. http://dx.doi.org/10.1016/j.compstruct.2015.04.046.
[34] R.K. Miller, E.K. Hill, P.O. Moore, Testing ASfN. Acoustic Emission Testing, American Society for Nondestructive Testing, 2005. [35] M.R. Gorman, Plate wave acoustic emission, J. Acoust. Soc. Am. 90 (1) (1991) 358e364. [36] M. Eaton, M. May, C. Featherston, K. Holford, S. Hallet, R. Pullin, Characterisation of damage in composite structures using acoustic emission, J. Phys. Conf. Ser. (2011) 012086. [37] D. Aljets, A. Chong, S. Wilcox, K. Holford, Acoustic Emission Source Location in Plate-like Structures Using a Closely Arranged Triangular Sensor Array, Uni Of Glamorgan, Pontypridd, Wales, UK CF37 1DL, 2010. [38] A. Oskouei, M. Ahmadi, M. Hajikhani, Wavelet-based acoustic emission characterization of damage mechanism in composite materials under mode I delamination at different interfaces, Exp. Polym. Lett. 3 (12) (2009) 804e813. [39] T.H. Loutas, V. Kostopoulos, Health monitoring of carbon/carbon, woven reinforced composites. Damage assessment by using advanced signal processing techniques. Part I: acoustic emission monitoring and damage mechanisms evolution, Compos. Sci. Technol. 69 (2) (2009) 265e272. [40] V. Arumugam, S. Sajith, A.J. Stanley, Acoustic emission characterization of failure modes in GFRP laminates under mode I delamination, J. Nondestruct. Eval. 30 (3) (2011) 213e219. [41] Q.-Q. Ni, M. Iwamoto, Wavelet transform of acoustic emission signals in failure of model composites, Eng. Fract. Mech. 69 (6) (2002) 717e728. [42] C.R. Ramirez-Jimenez, N. Papadakis, N. Reynolds, T.H. Gan, P. Purnell, M. Pharaoh, Identification of failure modes in glass/polypropylene composites by means of the primary frequency content of the acoustic emission event, Compos. Sci. Technol. 64 (12) (2004) 1819e1827. [43] W. Haselbach, B. Lauke, Acoustic emission of debonding between fibre and matrix to evaluate local adhesion, Compos. Sci. Technol. 63 (15) (2003) 2155e2162. [44] K. Asamene, L. Hudson, M. Sundaresan, Influence of attenuation on acoustic emission signals in carbon fiber reinforced polymer panels, Ultrasonics 59 (2015) 86e93. http://dx.doi.org/10.1016/j.ultras.2015.01.016. [45] T. Lundin, S.M. Cramer, R.H. Falk, C. Felton, Accelerated weathering of natural fiber-filled polyethylene composites, J. Mater. Civ. Eng. 16 (2004) 547e555. [46] A.I. Mourad, Thermo-mechanical characteristics of thermally aged polyethylene/polypropylene blends, Mater. Des. 31 (2010) 918e929. [47] L.C. Mendes, E.S. Rufino, F.O.C. de Paula, A.C. Torres, Mechanical, thermal and microstructure evaluation of HDPE after weathering in Rio de Janerio city, Polym. Degrad. Stab. 79 (2003) 371e383. [48] M. Philip, J. Attwood, A. Hulme, G. Williams, P. Shipton, Evaluation of Weathering in Mixed Polyethylene and Polypropylene Products, Report (UK), The Waste & Resources Action Programme (WRAP), 2004. ISBN: 1-84405092-0. [49] B.Y. Zhao, X.W. Yi, R.Y. Li, P.F. Zhu, K.A. Hu, Characterization to the weathering extent of LLDPE/LDPE thin film, J. Appl. Polym. Sci. 88 (2003) 12e16. [50] S.M. Al-Salem, G. Abraham, O.A. Al-Qabandi, A.M. Dashti, Investigating the effect of accelerated weathering on the mechanical and physical properties of high content plastic solid waste (PSW) blends with virgin linear low density polyethylene (LLDPE), Polym. Test. 46 (2015) 116e121. http://dx.doi.org/10. 1016/j.polymertesting.2015.07.008. [51] K. Rajakumar, V. Sarasvathy, A. ThamaraiChelvan, R. Chitra, C.T. Vijayakumar, Natural weathering studies of polypropylene, J. Polym. Environ. 17 (2009) 191e202. [52] D. Briassoulis, A. Aristopoulou, M. Bonora, I. Verlodt, Degradation characterisation of agricultural low-density polyethylene films, Biosyst. Eng. 88 (2004) 131e143. [53] A.I.A. Torik, H. Shirakawa, S. Nagaya, Photodegradation of polyethylene: factors affecting photostability, J. Appl. Polym. Sci. 40 (1990) 1637e1646. [54] Li Chang Sheng, M. Zhan, X. Huang, H. Zhou, Degradation behaviour of ultrahigh molecular weight polyethylene fibers under artificial accelerated weathering, Polym. Test. 31 (2012) 938e943. [55] T. Ojeda, A. Freitas, K. Birck, E. Dalmolin, R. Jacques, F. Bento, F. Camargo, Degradability of linear polyolefins under natural weathering, Polym. Degrad. Stab. 96 (2011) 703e707. [56] M. Sanchez-Soto, A. Rossa, A.J. Sanchez, J. Gamez-Perez, Blends of HDPE wastes: study of the properties, Waste Manag. 28 (2008) 2565e2573. [57] N. Shanmugam, I.A. Hussein, A. Badghaish, A.N. Shuaib, S.A. Furquan, M.H. AlMehthel, Evaluation of oil fly ash as a light stabilizer for epoxy composites: accelerated weathering study, Polym. Degrad. Stab. 112 (2015) 94e103. http://dx.doi.org/10.1016/j.polymdegradstab.2014.12.016. [58] R.S.C. Woo, Y.H. Chen, H.G. Zhu, J. Li, J.K. Kim, C.K.Y. Leung, Environmental degradation of epoxy-organoclay nanocomposites due to UV exposure and moisture attack: part I, Photodegr. Compos Sci. Technol. 67 (2007) 3448e3456. [59] M.D.H. Beg, K.L. Pickering, Accelerated weathering of unbleached and bleached Kraft wood fibre reinforced polypropylene composites, Polym. Degrad. Stab. 93 (2008) 1939e1946, http://dx.doi.org/10.1016/ j.polymdegradstab.2008.06.012. [60] M.M. Thwe, K. Liao, Durability of bambooeglass fiber reinforced polymer matrix hybrid composites, Compos. Sci. Technol. 63 (2003) 375e387. Ne ve, M.E.R. Shanahan, Water absorption by an epoxy resin and its effect [61] B. De on the mechanical properties and infra-red spectra, Polymer 34 (1993) 5099e5105. [62] B.G. Kumar, R.P. Singh, T. Nakamura, Degradation of carbon fiber-reinforced epoxy composites by ultraviolet radiation and condensation, J. Compos
A. de Souza Rios et al. / Polymer Testing 50 (2016) 152e163 Mater 36 (2002) 2713e2721. [63] M. Mariatti, P.K. Chum, Effect of laminate configuration on the properties of glass fiber-reinforced plastics (GFRPs) mixed composites, J. Reinf. Plast. Compos. 24 (2005) 1713e1721.
163
[64] S. Sethi, B.C. Ray, Environmental effects on fibre reinforced polymeric composites: evolving reasons and remarks on interfacial strength and stability, Adv. Colloid Interface Sci. 217 (2015) 43e67. http://dx.doi.org/10.1016/j.cis. 2014.12.005.