J Mater Sci(2007)42:1162-1168 DOI10.1007/s10853-006-1445-1 Effects of fiber orientation on the acoustic emission and fracture characteristics of composite laminates Nak-Sam Choi Sung- Choong Woo Kyong- Yop Rhee Received: 17 November 2005/ Accepted: 2 May 2006/Published online: 14 January 2007 o Springer Science+Business Media, LLC 200 Abstract The effects of fiber orientation on acoustic Introduction emission(AE)characteristics have been studied for various composite laminates. Reflection and transmis- Fiber-reinforced plastic composites(FRPC)have been sion optical microscopy were used to investigate the extensively used in the aerospace industry, automobile amage zone of specimens. AE signals were classified industry, sports utilities, electronic industry and archi through short time Fourier transform (STFT)as tecture due to their advantages of superior strength-to- different types: AE signals with a high intensity and weight and stiffness-to-weight Damage occurrence in high frequency band were due to fiber fracture, while composite materials under in-use conditions may, weak AE signals with a low frequency band were due however, reduce their mechanical performance. Thus, to matrix cracking and/or interfacial cracking. Char the fracture behavior of composite materials has ofte teristic feature in the rate of hit-events having been examined with the aid of acoustic emission(AE) amplitudes showed a procedure of fiber breakages, non-destructive evaluation methods to get usef which expressed the characteristic fracture processes of information in order to improve their structural integ notched fiber-reinforced plastics with different fiber rity and reliability orientations. As a consequence, the behavior of frac Using single-fiber-reinforced and short-fiber- ture in the continuous composite laminates could be reinforced composites, many researchers [1-3 monitored through nondestructive evaluation(NDE) reported on the correlation of AE data with the using the ae technique individual fracture processes of FRPC although the fracture processes were rather complex: Fiber fracture was liable to emit strong waves with high-frequency bands of 250-450 kHz. matrix fracture. as well as debonding and/or friction between fibers and matrix. might generate weak waves with low-frequency band. For safety estimation of FRP tanks/vessels, ASTM N.S. Choi(<) Department of Mechanical Engineering, Hanyang standards [4, 5] recommends an evaluation based on University, 1271, Sa-1dong, Ansan-si, Kyunggi-do 426-791 high-amplitude aE events since those events are often Korea indicative of the major structural damage associated e-mail: nschoi@hanyang.ac ki with fiber breakages. The present author [6, 7 inves- tigated short fiber reinforced thermoplastic (SFRP) Department of Mechanical Design, Graduate School, composites by utilizing the frequency analysis with Hanyang ersity, 17, Haengdang-dong, Sungdong-ku band-pass filters, and proposed an experimental model Seoul 133-791 Korea for ae characteristics of the stable fracture process of K.Y SFRP: AE characteristics were largely affected by the School of Mechanical and Industrial System Engineering, initial notch-tip radius which was influential in the 449-701, Korea damage initiation and fracture processes. It was shown 2 Springer
Effects of fiber orientation on the acoustic emission and fracture characteristics of composite laminates Nak-Sam Choi Æ Sung-Choong Woo Æ Kyong-Yop Rhee Received: 17 November 2005 / Accepted: 2 May 2006 / Published online: 14 January 2007 Springer Science+Business Media, LLC 2007 Abstract The effects of fiber orientation on acoustic emission (AE) characteristics have been studied for various composite laminates. Reflection and transmission optical microscopy were used to investigate the damage zone of specimens. AE signals were classified through short time Fourier transform (STFT) as different types: AE signals with a high intensity and high frequency band were due to fiber fracture, while weak AE signals with a low frequency band were due to matrix cracking and/or interfacial cracking. Characteristic feature in the rate of hit-events having high amplitudes showed a procedure of fiber breakages, which expressed the characteristic fracture processes of notched fiber-reinforced plastics with different fiber orientations. As a consequence, the behavior of fracture in the continuous composite laminates could be monitored through nondestructive evaluation (NDE) using the AE technique. Introduction Fiber-reinforced plastic composites (FRPC) have been extensively used in the aerospace industry, automobile industry, sports utilities, electronic industry and architecture due to their advantages of superior strength-toweight and stiffness-to-weight. Damage occurrence in composite materials under in-use conditions may, however, reduce their mechanical performance. Thus, the fracture behavior of composite materials has often been examined with the aid of acoustic emission (AE) non-destructive evaluation methods to get useful information in order to improve their structural integrity and reliability. Using single-fiber-reinforced and short-fiberreinforced composites, many researchers [1–3] reported on the correlation of AE data with the individual fracture processes of FRPC although the fracture processes were rather complex: Fiber fracture was liable to emit strong waves with high-frequency bands of 250–450 kHz. Matrix fracture, as well as debonding and/or friction between fibers and matrix, might generate weak waves with low-frequency band. For safety estimation of FRP tanks/vessels, ASTM standards [4, 5] recommends an evaluation based on high-amplitude AE events since those events are often indicative of the major structural damage associated with fiber breakages. The present author [6, 7] investigated short fiber reinforced thermoplastic (SFRP) composites by utilizing the frequency analysis with band-pass filters, and proposed an experimental model for AE characteristics of the stable fracture process of SFRP: AE characteristics were largely affected by the initial notch-tip radius which was influential in the damage initiation and fracture processes. It was shown N.-S. Choi (&) Department of Mechanical Engineering, Hanyang University, 1271, Sa-1dong, Ansan-si, Kyunggi-do 426-791, Korea e-mail: nschoi@hanyang.ac.kr S.-C. Woo Department of Mechanical Design, Graduate School, Hanyang University, 17, Haengdang-dong, Sungdong-ku, Seoul 133-791, Korea K.-Y. Rhee School of Mechanical and Industrial System Engineering, Kyunghee University, Yongin 449-701, Korea J Mater Sci (2007) 42:1162–1168 DOI 10.1007/s10853-006-1445-1 123
J Mater Sci(2007)42:1162-1168 for the SFRP composites that fiber breakage ahead of (MISTRAS 2001, PAC). aE waves were detected by the notch tip was a pre-requisite for the initiation of the two transducers(micro30, PAC) having similar sensing main crack. The present author [8 also reported a characteristics(100-600 kHz with peak sensitivity at technique for monitoring the individual fracture pro- 275 kHz). The two sensors were mounted cesses of cross-ply FRP laminates by using the classi- of the specimen(see S1 and S2 in Fig. 1)using vacuum fication of thermo-acoustic emissions grease and mechanical fixtures The distance between In this work, we study the fracture processes of the two transducers and the initial notch tip was kept to typical FRP laminates with continuous fiber reinforce- 10.0+0.2 mm. AE measurement conditions of pre ment on the basis of acoustic emission characteristics. amp 40 dB and a threshold level of 40 dB were For microscopic examination and identification of adopted. fracture mechanisms in FRPC, the petrographic thin The aE detection system determined source loca sectioning technique employing lapping and polishing tions of AE waves based on timed data of wave arrival of both sides of a sample [9 is utilized in combination at the transducers S, and S,, and on a preset wave with a fracture surface observation. This work focuses velocity. To measure the arrival time difference the on the effects that different kinds of fiber arrangement clock rate was 4 MHz. Pencil lead(dia. 0.5 mm, lead have on AE characteristics in conjunction with the length 3 mm, HB) breakage was used as a simulated fracture processes of notched FrPC AE source for the measurement of the wave propaga tion velocity in the direction of specimen length. The Experimental source location so that the output location could be detected accurately from the original location of inpu Composite materials lead breaks. Figure 2 shows the results of ae wave velocity measured for the respective specimens. values Glass-fiber/epoxy laminates with lay-ups of [o s ]s, of the wave propagation velocity for each specimen [90ls and [+454/-454Is as well as satin-weave(Sw) were on a level with the Lamb wave velocity glass-fabric/epoxy laminates with lay-ups of [0 4/904Is +4504-45° dopted for this study. Ur directional(UD)glass fiber/epoxy resin prepreg and (SW-[0/9016) glass-fabric/epoxy prepreg produced by sk Chemicals were used to manufacture the laminate specimens with a thickness of 2-2.5 mm. They wer made in autoclave using a curing cycle recommended by the manufacturer. Flat-type specimens 15 mm wide and 180 mm long were made by sectioning the lami nates using a diamond wheel cutter. Specimen length A directions were kept parallel with the fibers for the laminate UD-o ls and perpendicular to the fibers for 7.5 the laminate UD-908ls, and transverse to the fabric structures for the satin-weave laminate(Sw-[0%/9016) The gauge portion had a single-edge notched with a low speed diamond wheel cutter. The notching direc tion was perpendicular to the specimen length. After that, a sharp notch was introduced by pushing a fresh razor blade with a static load of 300n into the initial notch tip. The notch depth was kept to 7.5+ 0.2 mm. Acoustic emission measurement Tensile tests with a cross-head speed of 0.1 mm/min were performed. Five specimens were tested for each kind of composite. To monitor the fracture processes of the specimens, AE measurement data was recorded Fig. 1 Notched specimen of composite laminates and AE sensor in real time using a two-channel ae detection system locations Sl and $2
for the SFRP composites that fiber breakage ahead of the notch tip was a pre-requisite for the initiation of the main crack. The present author [8] also reported a technique for monitoring the individual fracture processes of cross-ply FRP laminates by using the classi- fication of thermo-acoustic emissions. In this work, we study the fracture processes of typical FRP laminates with continuous fiber reinforcement on the basis of acoustic emission characteristics. For microscopic examination and identification of fracture mechanisms in FRPC, the petrographic thin sectioning technique employing lapping and polishing of both sides of a sample [9] is utilized in combination with a fracture surface observation. This work focuses on the effects that different kinds of fiber arrangement have on AE characteristics in conjunction with the fracture processes of notched FRPC. Experimental Composite materials Glass-fiber/epoxy laminates with lay-ups of [00 8]S, [900 8]S and [+450 4/–450 4]S as well as satin-weave (SW) glass-fabric/epoxy laminates with lay-ups of [00 4/900 4]S and [+450 4/–450 4]S. were adopted for this study. Unidirectional (UD) glass fiber/epoxy resin prepreg and (SW-[00 /900 ]16) glass-fabric/epoxy prepreg produced by SK Chemicals were used to manufacture the laminate specimens with a thickness of 2–2.5 mm. They were made in autoclave using a curing cycle recommended by the manufacturer. Flat-type specimens 15 mm wide and 180 mm long were made by sectioning the laminates using a diamond wheel cutter. Specimen length directions were kept parallel with the fibers for the laminate UD-[00 8]S and perpendicular to the fibers for the laminate UD-[900 8]S, and transverse to the fabric structures for the satin-weave laminate (SW-[00 /900 ]16). The gauge portion had a single-edge notched with a low speed diamond wheel cutter. The notching direction was perpendicular to the specimen length. After that, a sharp notch was introduced by pushing a fresh razor blade with a static load of 300 N into the initial notch tip. The notch depth was kept to 7.5 ± 0.2 mm. Acoustic emission measurement Tensile tests with a cross-head speed of 0.1 mm/min were performed. Five specimens were tested for each kind of composite. To monitor the fracture processes of the specimens, AE measurement data was recorded in real time using a two-channel AE detection system (MISTRAS 2001, PAC). AE waves were detected by two transducers (micro30, PAC) having similar sensing characteristics (100–600 kHz with peak sensitivity at 275 kHz). The two sensors were mounted on one side of the specimen (see S1 and S2 in Fig. 1) using vacuum grease and mechanical fixtures. The distance between the two transducers and the initial notch tip was kept to 10.0 ± 0.2 mm. AE measurement conditions of preamp 40 dB and a threshold level of 40 dB were adopted. The AE detection system determined source locations of AE waves based on timed data of wave arrival at the transducers S1 and S2, and on a preset wave velocity. To measure the arrival time difference the clock rate was 4 MHz. Pencil lead (dia. 0.5 mm, lead length 3 mm, HB) breakage was used as a simulated AE source for the measurement of the wave propagation velocity in the direction of specimen length. The velocity was obtained through confirmation of the source location so that the output location could be detected accurately from the original location of input lead breaks. Figure 2 shows the results of AE wave velocity measured for the respective specimens. Values of the wave propagation velocity for each specimen were on a level with the Lamb wave velocity 7.5 15 S1 S2 1 100 P (t=2) preset zone P 0.1 mm/min Fig. 1 Notched specimen of composite laminates and AE sensor locations S1 and S2 123 J Mater Sci (2007) 42:1162–1168 1163
1164 J Mater Sci(2007)42:1162-1168 range of 40-70 dB were adopted as hit-events in this l:sW-|P/9016 figure. Each histogram of the event-rate shows the 2SW-+45°4445°4l16 number of events obtained during a measurement time 63UD.9 of 3 s intervals. Detectable AE events began to occur 4UD+45°4-45 at a displacement of 0.14 mm and corresponded to bout 63% of the maximum load Pma. as s increased 6: UD-[Osks further up to the Pmax point corresponding to the crack initiation point, the event rate showed an increase with irregular variations(see stage I in Fig. 3). Just after the crack initiation, the load abruptly dropped due to the rapid crack propagation revealing a drastic rise in the event rate(see the early portion of stage II). with increasing displacement in the next portion of stage II the load gradually decreased due to the slow crack Specimen type growth. This cracking process showed a decrease in the average event -rate which might be indicative of a Fig 2 Measurement results of ae propagation velocity depend ing on lay-ups of composite laminates decrease in the size of the dama ge zone formed during each unit length of displacement. Using the short time Fourier transform. it was confirmed that most of the dominantly corresponding to the transverse wave AE hit events had a low frequency band of 40-270 kHz velocity calculated in consideration of the elastic s shown in Fig 4. Thus those AE events are consid- modulus of the corresponding specimen ered to have occurred due to matrix fracture. fiber- For real AE measurement, AE signals coming from matrix interfacial failure and pull-out process of fibers outside of the pre-set zone between transducers were as reported in references [2, 3, 6, 7 disregarded as noises Figure 5 shows histograms of the event rate in a high amplitude range beyond 70 dB for the same specimen in Fig 3. Over 90% of AE events processed by the Fiber orientation effects of ae characteristics and short time fourier transform was shown to have both discussions high and low frequency bands of 40-480 kHz(see Fig 6), suggesting emissions from the fiber breakage For a UD-190sls composite specimen, a typical load P- and accompanying matrix fracture [1-3, 6, 7]. Thus the displacement 8 curve and accompanying histograms of hit-events behavior may represent the histories of fibe the ae hit-event rate are shown in Fig 3. A blast type breakages during the fracture processes of the individ- of AE waves with low and intermediate amplitudes in a ual composites.The why hit-events were hardly measured in stage I before the crack initiation point, Pmax, may be that few fibers were broken in front of the initial crack tip under the transverse tensile loading. A Stage I further increase of 8 into the stage II past the Pr Crack initiation point brought about a large amount of hit-event rate during the crack growth. This AE activity seemed to be 63%ofF due to the breakages of fibers which had been bridged Fig 3 Typical diagrams of load P and AE event rate versus displacement 8 for unidirectional [90 sIs composite specimen: accompanying AE event rate data was obtained from signals. Fig 4 A weak AE signal processed by the short time Fourier 2 Springer
dominantly corresponding to the transverse wave velocity calculated in consideration of the elastic modulus of the corresponding specimen. For real AE measurement, AE signals coming from outside of the pre-set zone between transducers were disregarded as noises. Fiber orientation effects of AE characteristics and discussions For a UD-[900 8]S composite specimen, a typical load Pdisplacement d curve and accompanying histograms of the AE hit-event rate are shown in Fig. 3. A blast type of AE waves with low and intermediate amplitudes in a range of 40–70 dB were adopted as hit-events in this figure. Each histogram of the event-rate shows the number of events obtained during a measurement time of 3 s intervals. Detectable AE events began to occur at a displacement of 0.14 mm and corresponded to about 63% of the maximum load Pmax. As d increased further up to the Pmax point corresponding to the crack initiation point, the event rate showed an increase with irregular variations (see stage I in Fig. 3). Just after the crack initiation, the load abruptly dropped due to the rapid crack propagation revealing a drastic rise in the event rate (see the early portion of stage II). With increasing displacement in the next portion of stage II, the load gradually decreased due to the slow crack growth. This cracking process showed a decrease in the average event-rate, which might be indicative of a decrease in the size of the damage zone formed during each unit length of displacement. Using the short time Fourier transform, it was confirmed that most of the AE hit events had a low frequency band of 40–270 kHz as shown in Fig. 4. Thus those AE events are considered to have occurred due to matrix fracture, fibermatrix interfacial failure and pull-out process of fibers as reported in references [2, 3, 6, 7]. Figure 5 shows histograms of the event rate in a high amplitude range beyond 70 dB for the same specimen in Fig. 3. Over 90% of AE events processed by the short time Fourier transform was shown to have both high and low frequency bands of 40–480 kHz (see Fig. 6), suggesting emissions from the fiber breakage and accompanying matrix fracture [1–3, 6, 7]. Thus the hit-events behavior may represent the histories of fiber breakages during the fracture processes of the individual composites. The reason why hit-events were hardly measured in stage I before the crack initiation point, Pmax, may be that few fibers were broken in front of the initial crack tip under the transverse tensile loading. A further increase of d into the stage II past the Pmax point brought about a large amount of hit-event rate during the crack growth. This AE activity seemed to be due to the breakages of fibers which had been bridged 0.0 0.1 0.2 0.3 0.4 0.5 0 30 60 90 120 150 , daoL P (N) Displacement, δ (mm) 0 10 20 30 40 50 60 r t nevE aet Stage I II Crack initiation 63% of Pmax Fig. 3 Typical diagrams of load P and AE event rate versus displacement d for unidirectional [900 8]S composite specimen: accompanying AE event rate data was obtained from signals with low and intermediate amplitudes from 40 to 70 dB 0.1 0.2 0.3 0.4 0 0 50 100 150 200 200 400 600 800 1000 Freq eu nc ( y kHz) Time (µ sec) Fig. 4 A weak AE signal processed by the short time Fourier Transform with low frequency bands 1: SW-[00 /900] 16 2: SW-[+450 4 /-450 4] 16 3: UD-[00 4 /900 4]S 4: UD-[+450 4 /-450 4] S 5: UD-[900 8] S 6: UD-[00 8] S 1 2 3 4 5 6 0 2 4 6 8 01 X 3 ( yti col e V s/ m ) Specimen type Fig. 2 Measurement results of AE propagation velocity depending on lay-ups of composite laminates 123 1164 J Mater Sci (2007) 42:1162–1168
J Mater Sci(2007)42:1162-1168 are presented in Fig. 8. With a very high strength 46 times larger than that of UD-90 sls specimen, the UD Stage I [os]s specimen showed the maximum event rate around 8=2.3 mm. However. a drastic decrease in he event rate to a minimum was obvious at 15 8=2.6 mm corresponding to about 72% of Pmax(see i A in Fig 8). After that, the event rate showed irregular ncreases with jumps up and down. With further increase of P beyond 82% of Pmax(see B in Fig 8) the rate rapidly decreased again to a very low level which was similar to that of UD-190sls specimen wn in Fig. 5. Reflected optical observation Displacement, a(mm of the specimen in the stage of load below 82% Pmax showed that many fiber brea Fig. 5 Typical diagran beimen as used in Fig. 3: took place ahead of the initial notch tip, which and aE event rate versus displacement 8 for the accompanying AE event was obtained from signals corresponded to the events arisen before A. Crack with high amplitudes beyor advance became totally shifted to a 90 angle parallel to the fiber orientation, which led to the shearing mode (mode II) fracture(see shear crack in Fig 9)showing minor fiber breakages. The main crack finally propa gated perpendicular to the initial notching direction. It is thought for the fracture process that at the lst minimum point(A)in the hit-event rate, the loading state causing the mode I fracture began to change to a different state causing approximately mode I Time(u sec) fracture around the notch tip. The fracture mode change to shearing induced the formation of different Fig. 6 A strong AE signal processed by the short time Fourier damage zones and additional fiber breakages in Transform, showing a blast-type wave with both high and low load stage between A and B. As the load passed B, a complete mode change seemed to be accom- plished together with the propagation of a shear crack across the upper and lower crack surfaces behind the causing additional fiber breakages. The AE activity in crack tip(see broken fibers on the fracture surface in this final load stage seemed to have arisen from the breakage of fibers, which had been bridged across th For notched UD-0sls unidirectional composite specimens, a typical P-d curve and accompanying histograms of event rate in the high amplitude range 82%60fPg 0.00.5101.5202.5303.540 Fig. 7 Scanning electron microscopy observation of the fracture Fig 8 Typical diagrams of load P and AE event rate versus surface of the specimen of Figs. 3 and 5 Arrows indicate broken displacement d for [0 s]s composite specimen: accompanying AE fibers produced during the crack propagation event rate data was obtained with high amplitudes beyond 70 dB
across the upper and lower crack surfaces behind the crack tip (see broken fibers on the fracture surface in Fig. 7). For notched UD-[00 8]S unidirectional composite specimens, a typical P-d curve and accompanying histograms of event rate in the high amplitude range are presented in Fig. 8. With a very high strength 46 times larger than that of UD-[900 8]S specimen, the UD- [00 8]S specimen showed the maximum event rate around d = 2.3 mm. However, a drastic decrease in the event rate to a minimum was obvious at d = 2.6 mm corresponding to about 72% of Pmax (see A in Fig. 8). After that, the event rate showed irregular increases with jumps up and down. With further increase of P beyond 82% of Pmax (see B in Fig. 8), the rate rapidly decreased again to a very low level which was similar to that of UD-[900 8]S specimen shown in Fig. 5. Reflected optical observation (see Fig. 10) of the specimen in the stage of load slightly below 82% Pmax showed that many fiber breakages took place ahead of the initial notch tip, which corresponded to the events arisen before A. Crack advance became totally shifted to a 90 angle parallel to the fiber orientation, which led to the shearing mode (mode II) fracture (see shear crack in Fig. 9) showing minor fiber breakages. The main crack finally propagated perpendicular to the initial notching direction. It is thought for the fracture process that at the 1st minimum point (A) in the hit-event rate, the loading state causing the mode I fracture began to change to a different state causing the approximately mode II fracture around the notch tip. The fracture mode change to shearing induced the formation of different damage zones and additional fiber breakages in the load stage between A and B. As the load passed over B, a complete mode change seemed to be accomplished together with the propagation of a shear crack causing additional fiber breakages. The AE activity in this final load stage seemed to have arisen from the breakage of fibers, which had been bridged across the r F eque cn y( kHz) 1 2 3 4 5 6 0 0 200 400 600 800 1000 50 100 150 200 Time (µ sec) Fig. 6 A strong AE signal processed by the short time Fourier Transform, showing a blast-type wave with both high and low frequency bands Fig. 7 Scanning electron microscopy observation of the fracture surface of the specimen of Figs. 3 and 5. Arrows indicate broken fibers produced during the crack propagation 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0 1 2 3 4 5 6 7 , daoL P ( Nk ) Displacement, δ (mm) 0 50 100 150 200 evE nt r ate I II III A 82% of Pmax B Fig. 8 Typical diagrams of load P and AE event rate versus displacement d for [00 8]S composite specimen: accompanying AE event rate data was obtained with high amplitudes beyond 70 dB 0.0 0.1 0.2 0.3 0.4 0.5 0 30 60 90 120 150 oL , da P (N) Displacement, δ (mm ) 0 5 10 15 20 25 30 r t nevE a et Stage I II Crack initiation Fig. 5 Typical diagrams of load P and AE event rate versus displacement d for the same specimen as used in Fig. 3: accompanying AE event rate data was obtained from signals with high amplitudes beyond 70 dB 123 J Mater Sci (2007) 42:1162–1168 1165
1166 J Mater Sci(2007)42:1162-1168 hindering the crack advance were broken, which generated the high event rate. In the beginning of stage II shortly after that, the cracking was arrested on account of large hindrance effects by crossed fibers in the neighboring layers. As 8 increased further in stage il, the event rate was temporarily minimized before rapidly increasing again. This behavior of the event rate exhibiting irregular jumps up and down may indicate that the cracks advanced further inducing additional fiber breakages and various partial delam inations between layers. Figure 11 shows a typical P-s curve and accompa- racktip nying histograms of the ae hit-event rate in the high amplitude range for a Sw-[0%/90]16 specimen.AE event rate data was obtained from high amplitude ignals beyond 70 dB. The behavior of the de distribution for the specimen of this figure is exhibited in Fig. 12. A value of0 uv in peak amplitude Fig.9 Reflected optical micrograph of notched UD-108] s in this figure corresponded to 40 dB. Around specimen in load stage Il of Fig. 8 δ≡0.62mm, the maximum hit-event rate was observed with the generation of an AE wave having Crack initiation one of the biggest amplitudes. Then it decreased to a minimum at 8:0.66 m considerable decrease in the amplitude was also ascertained. At this displacement corresponding to about 92% of Pmax. it could be observed using a traveling microscope that the main crack was initiated. a thin polished section with a thickness of about 60 um obtained from a different satin weave [0/9016 specimen loaded to 40- 92% Pmax was examined under polarized light. After observing the transmitted light image of Fig. 13b, it was confirmed that many fiber breakages were obvious in front of the main crack initiation site seen in the 0.3 eflected light image of Fig 13a. The main crack under Displacement, d(mm) Fig. 10 Typical diagrams of load P and AE event rate versus displacement 8 for [+454/-454Is composite specimen. Accom- 800 panying AE event rate data was obtained with high amplitudes Crack initiation beyond 70 dB 92%ofP upper and lower crack surfaces behind the crack tip, as s [+451/-45 41, the event rate in the high amplitude 3 range increased greatly as the load approached the maximum(see stage I in Fig. 10). This might be on account of fiber breakages in front of the initial crack Just after passing the maximum load, a large load drop with a very high event rate arose due to the existence Displacement, 8(mm) of rapid crack propagation. The cracking progressed parallel to the fiber length direction in the skin and Fig. 11 Typical diagrams of load P and AE event ra displacement 8 for a satin weave [0/9016 composite core layers, i.e. at +45 or-45 angles to the notching. AE event rate data was obtained from high amplitud During this process, a substantial number of fibers beyond 70 dB
upper and lower crack surfaces behind the crack tip, as similarly observed in Fig. 5. For cross-ply composite specimens with layup of [+450 4/–450 4]S, the event rate in the high amplitude range increased greatly as the load approached the maximum (see stage I in Fig. 10). This might be on account of fiber breakages in front of the initial crack. Just after passing the maximum load, a large load drop with a very high event rate arose due to the existence of rapid crack propagation. The cracking progressed parallel to the fiber length direction in the skin and core layers, i.e. at +45 or –45 angles to the notching. During this process, a substantial number of fibers hindering the crack advance were broken, which generated the high event rate. In the beginning of stage II shortly after that, the cracking was arrested on account of large hindrance effects by crossed fibers in the neighboring layers. As d increased further in stage II, the event rate was temporarily minimized before rapidly increasing again. This behavior of the event rate exhibiting irregular jumps up and down may indicate that the cracks advanced further inducing additional fiber breakages and various partial delaminations between layers. Figure 11 shows a typical P–d curve and accompanying histograms of the AE hit-event rate in the high amplitude range for a SW-[00 /900 ]16 specimen. AE event rate data was obtained from high amplitude signals beyond 70 dB. The behavior of the AE amplitude distribution for the specimen of this figure is exhibited in Fig. 12. A value of 0 lV in peak amplitude in this figure corresponded to 40 dB. Around d * 0.62 mm, the maximum hit-event rate was observed with the generation of an AE wave having one of the biggest amplitudes. Then it decreased to a minimum at d * 0.66 mm, where a considerable decrease in the amplitude was also ascertained. At this displacement corresponding to about 92% of Pmax, it could be observed using a traveling microscope that the main crack was initiated. A thin polished section with a thickness of about 60 lm obtained from a different satin weave [00 /900 ]16 specimen loaded to 92% Pmax was examined under polarized light. After observing the transmitted light image of Fig. 13b, it was confirmed that many fiber breakages were obvious in front of the main crack initiation site seen in the reflected light image of Fig. 13a. The main crack under Fig. 9 Reflected optical micrograph of notched UD-[00 8] S specimen in load stage II of Fig. 8 0.0 0.2 0.4 0.6 0.8 0 200 400 600 800 oL da , P (N) Displacement, δ (mm) 0 100 200 300 400 500 Crack initiation Stage I II evE nt rate 92% of Pmax Fig. 11 Typical diagrams of load P and AE event rate versus displacement d for a satin weave [00 /900 ]16 composite specimen. AE event rate data was obtained from high amplitude signals beyond 70 dB 0 20 40 60 80 100 120 0.0 0.3 0.6 0.9 1.2 1.5 1.8 0.0 0.3 0.6 0.9 1.2 , daoL P ( Nk ) Displacement, δ (mm) evE nt r ate I III II Crack initiation Fig. 10 Typical diagrams of load P and AE event rate versus displacement d for [+450 4/–450 4]S composite specimen. Accompanying AE event rate data was obtained with high amplitudes beyond 70 dB 123 1166 J Mater Sci (2007) 42:1162–1168
J Mater Sci(2007)42:1162-1168 116 Crack initation 50 Notch tip Reflection Displacement, S(mm) Fig. 12 Behavior of the AE amplitude distribution versus 8 for the specimen of Fig. 11. The amplitude data was obtained from all AE signals beyond the threshold level(40 dB) Notch tip Reflection “密≡ Transmission transmitted light, respectively Fig. 13 Polarized optical microscopy observation of fracture hindering the crack initiation. Considering that the hit processes in a satin-weave composite specimen:(a) and events with high amplitudes were generated due to micrographs were taken from a thin section with a thickness of fiber breakages, it is understandable also in stage Iof about 60 um under reflected and transmitted light, respectively. Figs. II and 12 that fiber breakages ahead of the initial 92% of Pmax. Many fiber breakages were obvious in front of the notch tip led to the initiation of the main crack. This result was similar to that of references [ 6, 7 based on short-fiber-reinforced plastics in that fiber breakages tensile loading followed the rather straight-forward were a pre-requisite for the main crack initiation. pathway spanned by reinforcing fibers in front of the During main crack propagation, additional fiber break initial notch tip, which induced breakages of the fibers ages occurred depending on the extent of the local Table 1 Fracture process and AE characteristics for Fracture process notched composites UD-o &ls 1. Main crack initiation Minimized ude and event rate similar to the results of ref [ 6, 71 2. Fracture mode change Dual peak distribution, High event rate 3. Shear crack propagation Low event rate UD-90 s]s 1. Until crack initiation Few AE vent rate UD-[+454/-1.Until uIon Low amplitude, Low event rate 454]s n and Low and high amplitudes, High event rate delamination SW-+45/- 1. Main crack initiation Minimized amplitude and event-rate 45 Dual peak distribution similar to the results of ref. 2. Crack propagate on Increased amplitudes, High event rate Sw-[0/90]16 Same as above Same as above
tensile loading followed the rather straight-forward pathway spanned by reinforcing fibers in front of the initial notch tip, which induced breakages of the fibers hindering the crack initiation. Considering that the hitevents with high amplitudes were generated due to fiber breakages, it is understandable also in stage I of Figs. 11 and 12 that fiber breakages ahead of the initial notch tip led to the initiation of the main crack. This result was similar to that of references [6, 7] based on short-fiber-reinforced plastics in that fiber breakages were a pre-requisite for the main crack initiation. During main crack propagation, additional fiber breakages occurred depending on the extent of the local Fig. 13 Polarized optical microscopy observation of fracture processes in a satin-weave composite specimen: (a) and (b) micrographs were taken from a thin section with a thickness of about 60 lm under reflected and transmitted light, respectively. The section had been made after loading the specimen to about 92% of Pmax. Many fiber breakages were obvious in front of the initial notch tip Reflection Notch tip Notch tip Transmission (a) (b) Fig. 14 Polarized optical microscopy observation of fracture processes in a satin-weave -[+450 /-450 ]16 composite specimen around 95% of Pmax: (a) and (b) micrographs were taken from a thin section with a thickness of about 100 lm under reflected and transmitted light, respectively 0.0 0.2 0.4 0.6 0.8 0 200 400 600 800 aoL d, P (N) Displacement, δ (mm) 0 10 20 30 40 50 60 p mA litude (µ ) V Stage I II Crack initiation Fig. 12 Behavior of the AE amplitude distribution versus d for the specimen of Fig. 11. The amplitude data was obtained from all AE signals beyond the threshold level (40 dB) Table 1 Fracture process and AE characteristics for notched composites Fracture process AE UD-[00 8]S 1. Main crack initiation Minimized amplitude and event rate similar to the results of ref. [6, 7] 2. Fracture mode change Dual peak distribution, High event rate 3. Shear crack propagation Low event rate UD-[900 8]S 1. Until crack initiation Few AE 2. Crack propagation Low event rate UD-[+450 4/– 450 4]S 1. Until crack initiation Low amplitude, Low event rate 2. Crack propagation and delamination Low and high amplitudes, High event rate SW-[+450 /– 450 4]16 1. Main crack initiation Minimized amplitude and event-rate, Dual peak distribution similar to the results of ref. [6, 7] 2. Crack propagation Increased amplitudes, High event rate SW-[00 /900 ]16 Same as above Same as above 123 J Mater Sci (2007) 42:1162–1168 1167
J Mater Sci(2007)42:1162-1168 failure zone and caused a rapid increase in the ae classifications, could be utilized for non-destructive event rate(see stage II in Fig. 11). This result was also identification of different fracture mechanisms, which similar to references 6, 71 can help to better understand mechanical toughness Similar behavior of the ae hit-event rate and the and, thus, make more reliable design of fiber rein- main crack initiation mechanisms were also shown for forced composite materials and structures SW-[+45/-45J16 specimens, except that the load at the crack initiation was about 95% of Pmax, 1.8 times larger Acknowledgement This work was supported by grant No ROl-2005-000-10566-0)from the Basic Research Program of the than that of Sw-09016. The main cracking was Korea Science& Engineering Foundation rarely visible when viewed under reflected light(see Fig 14a). However, at this load stage some advance of matrix microfracture and many fiber breakages were References seen in a 45 direction biased from the notching direction under transmitted light(see Fig. 14b)con- 1. Wolters J(1986)In: Proceedings of 2nd internatio/ the firming the high AE activity behavior in advance of the main cracking Plastics Industry, Montreal, pp 29-36 2. Koenczoel L, Hiltner A, Baer E(1987) Polym Compos 8: 109 3. Suzuki m. Nakanishi H. Iwamoto m. Jinen E. Maekawa Z Conclusions 4. Standard practice for acoustic pn 36: 229 Koike K (1987)J Soc Mater Sci mission examinato reinforced thermosetting resin pipe (RTRP), ASTM Desig Table 1 summarizes fiber orientation effects on the nation: E1ll8-00. American Society for Testing and Materi als, West Conshohocken, PA, (December 2000) characteristic behaviors of AE hit-event rate in a high 5. Standard practice for acoustic emission examination of amplitude range above 70 dB for various kinds of fiberglass reinforced plastic resin (FRP) Tanks/Vessels, typical composite laminates. The AE characteristics might represent the process of fiber breakages accord nd Materials, West Conshohocken, PA, July 2001) 6. Choi NS, Takahashi K, Hoshino K(1992) NDTE Int 25: 271 ing to the various loading stages, which expressed 7. Choi NS, Takahashi K(1998)J Mater Sci 33: 2357 characteristic fracture processes for individual fiber- 8. Choi Ns, Kim YB, Kim Tw, Rhee KY(2003)JMater Sci reinforced composite laminates. The feature of the AE 38:1013 hit-event rate, in combination with AE amplitude 9. Choi NS, Takahashi K(1993)J Mater Sci Lett 12: 1718 2 Springer
failure zone and caused a rapid increase in the AE event rate (see stage II in Fig. 11). This result was also similar to references [6, 7]. Similar behavior of the AE hit-event rate and the main crack initiation mechanisms were also shown for SW-[+450 /–450 ]16 specimens, except that the load at the crack initiation was about 95% of Pmax, 1.8 times larger than that of SW-[00 /900 ]16. The main cracking was rarely visible when viewed under reflected light (see Fig. 14a). However, at this load stage some advance of matrix microfracture and many fiber breakages were seen in a 45 direction biased from the notching direction under transmitted light (see Fig. 14b) con- firming the high AE activity behavior in advance of the main cracking. Conclusions Table 1 summarizes fiber orientation effects on the characteristic behaviors of AE hit-event rate in a high amplitude range above 70 dB for various kinds of typical composite laminates. The AE characteristics might represent the process of fiber breakages according to the various loading stages, which expressed characteristic fracture processes for individual fiberreinforced composite laminates. The feature of the AE hit-event rate, in combination with AE amplitude classifications, could be utilized for non-destructive identification of different fracture mechanisms, which can help to better understand mechanical toughness and, thus, make more reliable design of fiber reinforced composite materials and structures. Acknowledgement This work was supported by grant No. (R01-2005-000-10566-0) from the Basic Research Program of the Korea Science & Engineering Foundation. References 1. Wolters J (1986) In: Proceedings of 2nd international symposium on AE from reinforced composites. The Society of the Plastics Industry, Montreal, pp 29–36 2. Koenczoel L, Hiltner A, Baer E (1987) Polym Compos 8:109 3. Suzuki M, Nakanishi H, Iwamoto M, Jinen E, Maekawa Z, Koike K (1987) J Soc Mater Sci Jpn 36:229 4. Standard practice for acoustic emission examination of reinforced thermosetting resin pipe (RTRP), ASTM Designation: E1118-00. American Society for Testing and Materials, West Conshohocken, PA, (December 2000) 5. Standard practice for acoustic emission examination of fiberglass reinforced plastic resin (FRP) Tanks/Vessels, ASTM Designation: E1067-01. American Society for Testing and Materials, West Conshohocken, PA, (July 2001) 6. Choi NS, Takahashi K, Hoshino K (1992) NDTE Int 25:271 7. Choi NS, Takahashi K (1998) J Mater Sci 33:2357 8. Choi NS, Kim YB, Kim TW, Rhee KY (2003) J Mater Sci 38:1013 9. Choi NS, Takahashi K (1993) J Mater Sci Lett 12:1718 123 1168 J Mater Sci (2007) 42:1162–1168
Copyright of Journal of Materials Science is the property of Springer Science Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holders express written permission. However, users may print, download or email articles for individual use