中厚煤層采煤機截割部設計,中厚煤層采煤機截割部設計,煤層,采煤,機截割部,設計 titanate 84-4 Electrical property The coefficient while a change The (a across is material engineeri positioning, vibration the electric domain orientation 2. Domain switching can occur with a high applied stress and the consequent elastic strain 3,4. Because of domain switching in PZT ceramics, the material proper- ties, e.g., piezoelectric constant, can be altered 5. Recently, Shindo et al. have examined the damage characteristics in PZT ceramic numerically and theoretically. It appeared from their work that the localized switching near the crack tip significantly affects the 2.1. Specimen preparation The material selected for the present work was a commercial bulk lead zirconium titanium oxide ceramic (PZT), produced by Fuji Ceramics Co. in Japan. The nominal grain size of this ceramic is about 5lm in diameter. Silver based electrodes 10 lm thick were plated on to the specimen surfaces by the following process: silver-metal powder with glass frit was coated on to the PZT surface; then the coated metal was fired in air at 973 K for a few hours 8. After the electrode attachment, the sample was polar- ized between the two electroplates. Two types of specimen were * Corresponding author. Tel.: +81 184 27 2211. International Journal of Fatigue 31 (2009) 14341441 Contents lists available E-mail address: okayasuakita-pu.ac.jp (M. Okayasu). the PZT ceramic for a long period of time, it is necessary to under- stand the material response to the application. The efficiency of the piezoelectric property in the ceramic can be changed if an over- load is applied, due to material damage in the ceramic. There have beenseveralpossiblekindsofdamageinPZTsreportedinpublished papers, e.g., microcrack, grain sliding and domain switching (polar- ization). The material damage occurs when the electric fatigue crack initiates from a porous region of the PZT 1. The damage in PZT ceramics can also be detected if the electrogeneration is trapped at a defect in the sample, which leads to the change of ing damage characteristics in PZT is indispensable for understand- ing their material properties. The main purpose of this paper is, therefore, to investigate the effects of material damage on the elec- trical and mechanical properties during the loading process. In addition, an attempt is made to reveal directly the damage charac- teristics in the PZT ceramic via unique experimental techniques. 2. Material and experimental procedures 1. Introduction Thesignificanceofpiezoelectricceramics duced voltage difference that appears the ceramic as the shape of the ceramic nating stresses. Using this unique ceramics have been utilized in various including memory devices, precision cal actuators, power transducers and 0142-1123/$ - see front matter C211 2009 Elsevier Ltd. All doi:10.1016/j.ijfatigue.2009.04.002 isthetransferofanin- two of the surfaces of subjected to high alter- characteristic, these ng applications electro-mechani- sensors. To employ fracture mechanical parameters, such as stress intensity factor and energy release rate. In addition, these parameters can be chan- ged by the crack growth length 6. It appeared from the above literature survey that there are sev- eral damage characteristics in PZT ceramics, and these can change the material properties. However, details on how to induce damage characteristics in the material properties have not been clarified. One reason is the technical difficulty of revealing the microscopic defects in PZT during the loading process 7. Information concern- Domain switching Material damage mic are discussed in the present work. C211 2009 Elsevier Ltd. All rights reserved. Damage characteristics of lead zirconate cyclic loading Mitsuhiro Okayasu * , Nozomi Odagiri, Mamoru Mizuno Department of Machine Intelligence and Systems Engineering, Akita Prefectural University, article info Article history: Received 1 January 2009 Received in revised form 31 March 2009 Accepted 1 April 2009 Available online 10 April 2009 Keywords: PZT ceramic Fatigue test abstract The effects of the damage characteristics ined during cyclic loading. (PZT). The electrical properties, ing cyclic loading. The k 33 with a high applied stress, a low applied stress. Such damage in the PZT ceramic. lightning-like phenomenon International Journ journal homepage: rights reserved. piezoelectric ceramic during Ebinokuchi, Tuchiya-aza, Yurihonjo-city, Akita 015-0055, Japan on the material properties of a piezoelectric ceramic are exam- material being examined is a lead zirconate titanate piezoelectric ceramic such as the electromechanical coupling coefficient (k 33 ), are changed dur- decreases rapidly to a low level as the sample is loaded cyclically the k 33 value decreases slowly or does not change when loaded with of material degradation is influenced by the severity of the material material damage in the PZT occurs, and this occurrence is related to a bright flash with a click). Details of the damage characteristics in PZT cera- at ScienceDirect al of Fatigue /locate/ijfatigue employed in the present work, as illustrated in Fig. 1: (a) a round rod; (b) a rectangular bar. All specimens were obtained from the same manufactured lot. The compressive strength of the round 3mm 3mm 3mm 7.5mm (a) (b) Electrode Fig. 1. Dimensions of the tested specimens: (a) a M. Okayasu et al./International Journal bar was about 750 MPa, and the bending strength of the rectangu- lar rod was about 80 MPa. 2.2. Fatigue and bending tests Low cycle fatigue and bending tests were carried out using a screw driven type universal testing machine with 10 kN capacity. Using the round rod specimen, a compressioncompression fatigue test was conducted at an R ratio of 0.05 and frequency of 0.05 Hz 7,8. The maximum cyclic load, r max , was determined on the basis of the compressive strength (r B ) of this ceramic, where r max is de- signed to be less than 67% ofr B 5,7. Using the three point bending specimen, a bending test was executed at elevated temperature with a loading speed of 1 mm/min to final fracture. A muffle fur- nace with an accuracy of better than 0.1 K was employed for the high temperature bending tests. The furnace was designed origi- nally to be fitted into the testing machine. At all times during the test, the actual temperature of the specimens was controlled. The electrical properties of this PZT ceramic, e.g., electromechanical coupling coefficient, k 33 , and piezoelectric constant, d 33 , were examined during the cyclic loading. In this approach, anti reso- nance frequency f a , resonance frequency f r and electrostatic capac- ity C T are measured during the tests using an impedance analyzer in advance. In this measurement, the parameters are examined as the applied load is removed to zero. With f a and f r values, k 33 can be obtained by the following equation 5: k 33 1 a fr faC0fr b s 1 4.0E-10 max (MPa), Nf (cycle to fracture) 0.0E+00 5.0E-11 1.0E-10 1.5E-10 2.0E-10 2.5E-10 3.0E-10 3.5E-10 0 20 40 60 80 100 120 Number of cycles P i ezo e l e ct ri c co n s t a n t , m / V 500(89) 450(68) 350(99) 300(410) 200(3,709) 100(45,000 ) 50 50MPa 100MPa 200MPa Fig. 2. Variation of piezoelectric constant d 33 as a function of the cycle number for the specimen loaded at various applied stresses. where a and b are the coefficient depending on vibration mode. On the other hand, piezoelectric constant, d 33 , can be described as: d 33 k 33 e 33 c E 33 s 2 where e 33 and c E 33 are dielectric constant and elastic coefficient, respectively, and those are assessed by the following formulas: e 33 C T t A 2a c E 33 2lf r 2 q for round rod sample 2b where t is the distance between the two electrodes, and A is the area of electrode. l and q represent the length of the round rod and the density of PZT, respectively. Details of the method for the above cal- culations can be found in Ref. 5. 3. Experimental results 3.1. Material properties during fatigue test Fig. 2 shows the variation of the piezoelectric constant (d 33 )asa function of the cycle number for the round rod specimens loaded cyclically at various compression loads (r max 50500 MPa). It should be noted first that the number in the legend indicates the maximum applied stress (r max ) and the cycle number to the frac- ture (N f ). The samples for r max 50 MPa and 100 MPa were loaded cyclically about 50,000 cycles and the cyclic load stopped, where those samples were not fractured completely. A high value for d 33 (3.1 C2 10 C010 m/V) is obtained in the sample during the cyclic loading at the low applied stress of r max 50 MPa, whereas a low d 33 level, less than 1.3C2 10 C010 m/V, is found for applied stress of more than r max 200 MPa. In contrast, the d 33 value decreases inter- mittently with increasing cycle number in the 100 MPa sample, and its value settles after 20 cycles to a similar level to that found 40mm 30mm Electrode round rod; (b) a rectangular bar specimen. of Fatigue 31 (2009) 14341441 1435 for the samples tested at more than r max 200 MPa. To further investigate the change of the electrical properties during the fati- gue test, the electromechanical coupling coefficient (k 33 ) vs. cycle number was investigated for the samples cycled at r max 50 MPa, 100 MPa and 200 MPa. The results obtained are shown in Fig. 3. As with the experimental results of Fig. 2, high and low values of the k 33 coefficient are obtained for the 50 MPa and 200 MPa sam- ples, respectively. Also, the k 33 value decreases with increasing cy- cle number for the 100 MPa sample. This result is convincing evidence that the material properties of the PZT ceramic are al- tered by the loading condition. The change of the material property in this case might be affected by the material damage in the sample during the cyclic loading. To clarify this, a direct observation of the sample was conducted. Fig. 4 displays the SEM images of the spec- imen surface (round rod) before and after five cycles at r max 450 MPa. Note that both pictures were obtained from the same location. From Fig. 4b, the damage (or collapse) of the sample surface is observed. It is considered from this result that the mate- rial damage in the PZT ceramic occurs during the fatigue test, and this may affect the electrical properties. 4. Discussion 4.1. Material damage vs. electrical properties To examine the relationship between the material damage and the material degradation in Figs. 2 and 3, a further set of tests was carried out. There are several damage characteristics in the PZT including microcrack (grain sliding) and domain switching 5,7. An attempt was made to examine the electrical properties of the specimens after receiving artificial microcrack damage. Instead of a microcrack, a machined slit was created in the round rod speci- men by a thin diamond cutting saw, e.g., 0.35 mm thick. Fig. 5 Fig. 4. SEM images of the PZT sample before and after cyclic loading at 450 MPa for five cycles. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 10 100 1000 Number of cycles Electromechanical coupling coefficient 200MPa 100MPa 50MPa (max ) 50MPa 100MPa 200MPa Fig. 3. Variation of electromechanical coupling coefficient k 33 as a function of cycle number for several specimens loaded at 50 MPa, 100 MPa and 200 MPa. Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 Step 1 Step 2 Step 3 Depth: 0.7mm 1.4mm 2.1mm 1.0mm 1.0mm Case I (a) Case II (b) Machine slit Machine slit Removed electrode (a face) Fig. 5. Schematic illustration showing the specimen materials 1436 M. Okayasu et al./International Journal of Fatigue 31 (2009) 14341441 3mm Removed electrode (both faces) Electrical wire with mechanical damage created by machine slits. 2mm Fig. 7. Pictures of the specimen: (a) before loading; (b) electrogeneration and (c) fracture. shows a schematic illustration of the samples showing the artificial damage. Two types of machine slits were created; (Case I) on the edge of the sample and (Case II) in the middle of the specimen. In Case I, several slits of the same size were machined in the sam- ple, one after another, denoted as Steps 16. In Steps 7 and 8, the electrode was removed by a file except for the area just around the electric wire attached to the sample surfaces. On the other hand, the machined slit was being made deeper at every step in Case II. It should be pointed out that compared to the actual mate- rial damage (microcrack and grain sliding), the size of the machine slit is much larger. However, the machined slit has been used to deliberately induce a greater degree of damage in order to study 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 012345678 E l ect r o mech an i cal co u p l i n g co ef f i ci en t , k 33 (a) Case I (b) Case II Step Depth of slit, mm Fig. 6. Variation of electromechanical coupling coefficient k 33 obtained in Cases I and II; see also Fig. 5. M. Okayasu et al./International Journal the consequent behavior 9. The electrical properties of the spec- imens with the machined slit were examined after every cut. Fig. 6a and b shows the electromechanical coupling coefficient, k 33 , measured after each step for Cases I and II, respectively. It is seen that the value of k 33 did not change in Case I even though the number of machined slits increased. On the other hand, a slight reduction in the k 33 coefficient can be seen machined slit was deepened in Case II, although the rate of reduction of k 33 is much smaller than that obtained in Fig. 3. This result might suggest that material damage, such as crack and grain sliding, does not play an important role in dictating the response of the electrical properties of the PZT ceramic (Figs. 2 and 3). 4.2. Damage characteristics during the loading process To understand the reasons for the reduction in electrical prop- erties as shown in Figs. 2 and 3, the observation of the specimen material during the static compressive loading to fracture was con- ducted using a video camera. From these observations, it was found that an electrical activity in the PZT occurs several times, re- lated to a lightning-like phenomenon and consisting of a bright flash with a click sound. Representative pictures of the specimen obtained in the loading process are displayed in Fig. 7: (a) before loading, (b) electrogenesis and (c) fracture. The intensity of the click occurring during the loading can be identified in the sound wave in Fig. 8. As seen in Fig. 8a, the click is detected eight times before the final fracture in this case. The enlarged wave for a click sound is indicated in Fig. 8b. The point of the electrogenesis is Base PZT specimen (Round bar) Loading direction (a) Before loading (b) Electrogeneration (c) Fracture instantly of Fatigue 31 (2009) 14341441 1437 further indicated on the compressive stress vs. displacement rela- tions (Fig. 9). As seen, a large number of the electrogenerative events are observed at the beginning of the loading process, espe- cially below 200 MPa. Because of the observation of the lightning phenomenon, it may be that the generation of electric charge is attributed to a part of the failure (or damage) in the PZT ceramic 2. In order to examine the effect of the electrogenic phenomenon on the material property in the PZT ceramic, the experimental data of Fig. 9 is correlated with the electromechanical coupling coeffi- cient (k 33 ) vs. compressive stress. The results are shown in Fig. 10. It is seen that the k 33 coefficient decreases nonlinearly with increase of applied stress, and its value settles out when the load- ing exceeds 200 MPa. Because many data points for the electrogen- eration are plotted in the region below 200 MPa, it might suggest that the electrogeneration is associated with the material degrada- tion in the PZT ceramic. Moreover, due to the material degradation, the electrogenesis might be attributed to the occurrence of domain switching in the PZT ceramic 2,10. To verify this clearly, another approach was conducted. In the previous studies, it was reported that the domain switching can play a prominent role in the tough- ness and fatigue properties of the piezoelectric ceramics 3, where the poled PZT ceramics have a high fatigue strength compared to the unpoled sample 4. This would be due to the change of lattice structure or an anisotropy effect 7,11. On the basis of previous re- ports, a study was performed to examine the variation of the microhardness during the cyclic loading. The cyclic loading was carried out with the maximum stress of about 60 MPa. The 0 0.01 0.02 0.03 0.04 Time, sec 0 3.0 -3.0 -6.0 6.0 Sound intensity, V (b) 0 10 20 30 40 50 60 Time, sec 0 -3.1 2.8 5.8 -6.1 Sound intensity, V Click noise from specimen Fracture (a) Noise from testing machine Fig. 8. Variation of the sound intensity during the loading test, showing click noise from the sample. 0 100 200 300 400 500 600 700 800 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Displacement, mm Applied compressive stress, MPa Fig. 9. Electrogeneration events during the loading process. 1438 M. Okayasu et al./International Journal 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 100 200 300 400 500 600 700 Applied compressive stress, MPa E l ect r o mech an i cal co u p l i n g co ef f i ci en t Fracture Electrogeneration events of Fatigue 31 (2009) 14341441 obtained data is shown in Fig. 11. It is seen that the microhardness level is apparently increase with increasing the cycle number. Hence, the result obtained would suggest the change of domain orientation occurred during the cyclic loading. Further experimental approach was carried out, where the mechanical properties were examined as a function of cyclic load- ing; and these were then compared to the electrical properties of the PZT ceramic (Fig. 3). The experimental results presented in Fig. 12 demonstrate the variation of both the flexural modulus (E B ) and the k 33 coefficient as a function of cycle number for sam- ples tested at P max 50 MPa, 100 MPa and 200 MPa. It should be pointed out first that the flexural modulus obtained in Fig. 12 was determined from the compressive stressdeflection curves 8. In this case, the flexural modulus was used as a parameter in this assessment, because the E B modulus is very sensitive to mate- rial damage 8. The k 33 coefficient obtained in Fig. 3 is expressed as its negative function, C0k 33 , in order to compare it easily with the E B modulus. As in Fig. 12, a different trend of E B variation is Fig. 10. Relationship between the electromechanical coupling coefficient k 33 and applied compressive stress at the electrogeneration events in the specimen. 300 320 340 360 380 0 10 1,000 100,000 Number of cycles Vi ck ers h a rd n e s s Fig. 11. Variation of the microhardness of the PZT ceramic as a function of the cycle number. 8 10 12 14 16 d u lu s , k N /mm -1.0 -0.8 -0.6 -0.4 co u p l i n g co ef f i ci en t - k 33 -k33 coefficient Flexural modulus (a) max 50MPa M. Okayasu et al./International Journal observed. Low and high levels of the flexural modulus were ob- tained at 50 MPa and 200 MPa, respectively; the E B modulus for the 200 MPa sample, 13.5 kN/mm, is about twice as high than that for the 50 MPa sample. On the other hand, E B for the 100 MPa sam- ple varies linearly with increasing cycle number. Interestingly, the variation of E B for all the samples shown in Fig. 12ac is very sim- ilar to that of the C0k 33 coefficient. The relationship between the k 33 and E B values is further indicated in Fig. 13. As can be seen, the flexural modulus is linearly related to the k 33 coefficient with R 2 = 0.93. We conclude from these results that the mechanical strengths during cyclic loading are associated directly with the 0 2 4 6 1 10 100 1000 10000 Number of cycles F lex u r al mo -1.6 -1.4 -1.2 E l ect r o mech an i cal 0 2 4 6 8 10 12 14 16 18 1 10 100 1000 10000 Number of cycles F l e xura l m odul us , kN / m m -1.0 -0.8 -0.6 -0.4 -0.2 0.0 E l ect r o m ech an i cal co u p l i n g co ef f i ci en t - k 33 -k33 coefficient Flex