N an ~JAN Grant No. N J-4077c. of Electroswictors for Sonar TransducersI. Micro-Mechanics SI. Yearly Report.

Transcription

1 IM _DTIC -t S ELECTEr NzONR N an ~JAN Grant No. N J-4077c Micro-Mechanics SI of Electroswictors for Sonar TransducersI Yearly Report Period October 1, 1992-September 30, 1993 Sponsored by Office of Naval Research (ONR) Submitted to Dr. Robert Y. Ting, Head, Acoustic Materials Branch Underwater Sound Ref. Division Naval Research Laboratory P.O. Box 8337 Orlando, FL Submitted by: Sei-Joo Jang Thomas R. Shrout ps" Ogg

2 * Best Avai~lable C,%opy

3 I. Executive Summary The higher order electromechanical coupling "electrostriction" has been proposed as an alternative to the piezoelectric phenomena commonly employed for Navy-type sonar transducers. Based on the results of the parent program "Relaxor Ferroelectrics for Electrostrictive Transducers" (NO J-4077) and the three-year AASERT program on the classification of electrostrictors, it was found that the Type I - Pb(BIB2)03 and Type II - PLZT offered superior overall performance. Specific compositions were selected based on the given criteria: * Large E-field induced strain a 0.03% Operating temperature range 0-30"C Minimal strain E-field hysteresis (<10%) and fabricated into large ceramic rings with dimensions as follows: O.D. = ± in. (1.65 cm ±.013 cm), wall thickness = ± in. (445 cm. ± cm), and thickness ± in (0.31 cm ± cm). Ceramic rings (four each) were silver electroded and delivered to NRL in Orlando, Florida for sonar transducer evaluation. II. Transducer Fabrication and Testing A. Complitional Selection According to Navy requirements, various families of electrostrictive compositions were selected for evaluation as practical Navy transducers, in order to directly compare performance with PZT piezoelectric materials. (1) Pbo.89La0.11ZrO.65,Ti (PLZT-9/65/35), (Typ II) (2) 0.93 Pb(Mgl/3Nb2/3)O PbTiO3 (PMN-PT 93/7), (Type 1) and (3) 0.85 Pb0.9sLa0.02(Mgtj3Nb2/3)O Pb0.9g,0.02TiO3 (PLMNT 2/85/15) (Hybrid I & IU) The compositional selections were based on the criteria for Navy Sonar transducers outlined in previous reports. Dielectric data for the three compositions are presented in Table I. As seen in this Table, the PLZT-Type U electrostrictors offer the broadest region of macro-micro polar switching (Tm - Td = 141"C),but the lowest dielectric constant 1 khz). In contrast, the Type I PMNbased material exhibits much higher dielectric constants I khz); however, their macro-micro polar region is much less broad. The hybrid composition (Types I & If) 2/85/15 offers an intermediate macro-micro region and dielectric permnittivity. The maximum dielectric loss is similar for all three compositions (- 0. I khz). 2

4 Table I. Dielectric data for selected electrostrictive compositions. Tmax ('C) Td Tmax - Td Kmax tanamax Composition (1 khz) ('C) ('C) (1 khz) (1 khz) PLZT (9/65/35) , PMN-PT (93/7) , PLMNT (2/85/15) , Electric field induced transverse strain and hysteresis data for the three representative compositions are summarized in Table IE. Note: Strain measurements are for the PLZT 10/65/35 composition. As seen in this Table, all three compositions exhibit induced transverse strains in excess of 0.03% (20 kv/cm) with less than 10% hysteresis within the 0-30"C temperature range. Table II. Induced transverse itrain data for selected electrostrictor compositions. Temp. Transverse Strain (%) Hysteresis Composition QC) 10 kv/cm 20 kv/cm (%) PLZT (9/65/35) , PMN-PT (93/7) PLMNT (2/85/15) Note: Data given in Tables I and U are for thin disk ceramics. 3

5 B. Powder Fabrication/Synthesis Based on the composition families (Types) given above, large quantities of ceramic powders (> 1 kg) were prepared using powder synthesis techniques developed in the parent program. The following compositions were fabricated: (1) Pbo.9Lao.10(ZrO.65,TiO.55)03 PLZT- 10/65/35 Note: 9/65/35 exhibits excessive hysteresis. (2) 0.95Pb(MgI3Nb2r-0.05)PbTiO3 PMN-PT 95/5 (3) 0.9OPb(MgI/3Nb2/3O PbTiO 3 PLMNT 1/10/ mole% La203 (4) 0.85Pb(MgI/3Nb2/3)O PbTiO 3 PLMNT 2/ mole% La203 Additional formulations were synthesized, 0.93PMN-0.071PT, PLZT-10.5/65/35 and a Srmodified PMN-PT, but were deemed insufficient in terms of overall performance. Specific issues included excessive dielectric aging and hysteresis, both believed to be associated with changes in raw materials during the course of this program. C. Electrostrictive Ceramic Ring Fabrication Powders prepared above were granulated and pressed into large cylinders. Subsequent to binder burnout and densification, the cylinders were machined (Piezo Kinetics, Inc.) to the desired dimensions given in the executive summary. Upon machining, the samples were electroded using a fired on silver. Special Note: The application of the silver electrode must be performed at a temperature (5 600C) to minimize reactive interface which may lead to degraded dielectric performance. Unlike that found for PZT-based ceramics, the role of the electrode interface in high dielectric constant materials is of major concern. D. Electrostricive Ring Testing - Accesion For CRA&I The ceramic rings were qualified based on: (I) capacitance-temperature measurements; (2) 8 pyroelectric behavior, and (3) polarization-e-field (P/E). The results are given in the following Figures 1-4, respectively, and summarized in Table 111. DTO QUALITY IJNPECTZD 8 t., istributio I de Av0dIdlt'y Cod 4 Dist Avdi. E tca wdor

6 I, Table In. Dielectric/polarization data for selected electrostrictive compositions. Tmax AT Kmax P 1 khz Td Tm 1 KV/cm PLZT (10/65/35) -65"C -18"C 8YC PMN-PT (95/5) -17"C -25"C 42"C 27, PLMNT (1/90/10) -17"C -37"C 54"C 23, PLMNT (2/85/15) -17"C -50"C 67"C 20, As presented, the Type I materials had similar Tmaxs 1 khz) but with increasing La were found to have decreased Knax's but expanded AT(Tmax-Td) ranges. As expected,the Type II PLZT (10/65/35) material possessed the largest AT (Regime 1E). The level of induced polarization (Pind) is significantly lower than the Type I materials at 10 KV/cm; however, similar at 20 KV/cm. In terms of hysteresis, the 95/5 samples possessed levels of < 1 Hz, being significantly higher than that for small-thin disks samples of the same composition. Ill. Scale-Up Issues As observed for many of the compositions investigated, the level of P/E hysteresis was found to be significantly higher for the large ceramic rings in contrast to thin disks (1 cm diam.. I- 2 cm thick) used in all the preliminary screening tests, Table IH. In addition to hysteresis, dielectric aging was found to be prevalent. Various processing issues are known to affect dielectric aging and indirectly the P/E hysteresis due to grain-grain boundary inhomogeneity, defects, impurities. etc. The following is a list of processing parameters suggested for next generation of large sonar ring, i.e. 2" rings. * 0.5-1% PbO deficient High purity raw material, i.e. Nb2O5 * Hot isostatic pressing SrCa modification for grain boundary control -Transgranular vs. intergranular fracture 5

7 In addition to processing issues above, the following areas need to be addressed. " Silver electrode interface Note: Silver electrodes used have been developed for low K PZTs. " Uni-polar drive vs. bi-polar drive Note (1): Note (2): Bi-polar drive gives a frequency doubling. Unipolar drive may give rise to strain walk off due to the small remanent polarization tail in the temperature regime of interest 6

8 PYRO. COEFF. (C/m2K) C, - V, tn U N- q- I 0 m )N ot u, i LiCL rc LLi (E'UD) tv- NIIUZ8UI-

9 LOSS I0% t (U -- 0 I Ow Ln tn z1. c" (U W - Z, tn in CLI- InL CZ. I N IL' N in cc * *. in 00 V, cm IN(SO 00)033i

10 I I I - I IIper I od acq mode delay delayed 1 uolts pk 6588 range 1889iM lkx "C ref a Z.OM98E-85 R ref a 1E 87 AD gain a 2.aueform triangle cursor a x-axis f ield 9-axis charge "C coop R coup a 8 a. -" lcoop a 8 I' o ma' th re-zero disk +r 4 to'~i~., / lo t info auto soc thickness off ;"" WJ/m ± "C,/o p I I, I I I II, t..i I! I I. t. I 10/65/35 RT dia = mm area= 1.663E-04 m 2 thickness= Am Figure 1c. Polarization E-field behavior of PLZT room temperature. Note (1): y axis c/m 2 full scale x axis - 10 kv/cm full scale Note (2): All four rings were measured as indicated by the four plots.

11 LOSS LA LA - a- Ln 16; Z \ \ 0) a a- r' Cho *v N - n La La u x0 LAJ F-- CC0 Iyin 0 'JJ 0 W) U") Inl U") 0 a" 0 U") cncu cu - LNULSNO) DIdiD313ia

12 Pyro Coef. (c/m 2 K) L 44 W.60 a-. z- 0 If)m S-n E~cE 0 L on U- U)/3 lc 4DCRtJP

13 --- "- -! " pe r iod acq mode delayed delay I volts pk range lkx C ref "R ref o Z.GOOOE-05 a IE#08 AD gain a Z waueform triaiiyle cursor a x-axis field y-axis charge C coup a 8 R comp a,. ',,.. '',, :,,.. " I comp a 9 math re-zero disk -r* "ref *-r -. plot info thickness auto scl off w flu/m 2± 80.0m8 mc/um, p i. p I, I p p I I I.. i, I, I, I,,1 )5/ lia = mm area= 1.663E-04 ml thickness= ;4m Figure 2c. Polarization. E-field behavior of 95/5 room temperature. Note: y axis c/m 2 full scale x axis - 10 kv/cm full scale

14 LOSS cr. *l z tn *Zm Z\\ c U, \U \cj &n r-uc u as tn 0 m Li U, a UP. CCL.w in tn cm tn -- 01: m cu tn U, cu ININO I01i

15 PYROELECTRIC COEFF. - U, ~ -,, - Ln. o ~~00 ~ Is? Z >~ ~~C.0 E u 2 tn r- 00 Lo 0. I o 01 oz a-/ IUI8I~

16 ... period acq mode delayed delay 1 uolts pk 625e range 10000' IkX C ref a E-05 fr ref a IE408 AD gain a Z "uaueform triangle cursor a x-axis field Y-axis charge C comp 0 0 :., '.:,' ',.. ': :,;.. -.' :,,..., :,I : I : ', : I ' ', R comp comp a a 8 math disk -r * ref t-r S0plot info auto scl loop area thickness off i...- "_ Il /m, l I, I, I.I.I.I. I.I i I J i I. I. I, I I, 1/90/ dia = mm area= 1.663E-04 m 2 thickness= ;m Figure 3c. Polarization temperature. E-field behavior of 1/90/10 room Note: y axis - x axis c/m 2 full scale 10 kv/cm full scale

17 sso1 In In CC 550"1 N 0 In VVI CDC Sn'n W 00 ' :J In. Lai *..,- 4P IQ *. in onvin raw inuosnod 31IW.D1iaI

18 av 0\0 z 3 I-- 1 oa x. II cc * 100 I- %n~ cu In. *33300OI813130bi

19 0c a CL %n (YIo0 w00 L-- *: E W 0 Lci kn be Ln ~ ~~ V*c "l (atu0) N~iWZd~lE

20 period acq mode delayed delay 1 uolts pk 6598 range 1 lkx C ref a 2.OS8E-85 R ref a 1E*08 AD gain a 2 ueform triangle cursor a x-axis field y-axis charge Ccomp a9 11 R camp a0 I camp as math loop area disk *r * ref *r. ' plot info thickness auto scl off ±,2M..M mcm -... '. 1 I_ I I 1 85\15\2 RT SILVER dia = mmt area= 1.663E-04 m2 thickness= gm Figure 4d. Polarization E-field behavior of 2/85/5 PLMNT. Note: y axis c/m 2 full scale x axis. 10 kv/cm full scale