Our Experience with Single Crystal Piezocomposite Transducer Materials

Piezocomposite materials have significantly improved the performance of ultrasonic testing. When compared to monolithic piezoceramics, the coupling factor is higher and the acoustic impedance is lower. Both of these factors contribute to improved sensitivity and bandwidth. More recently, single-crystal piezoelectric materials are being developed. These new materials, when fabricated into piezocomposites, should produce and even more dramatic improvement in ultrasound testing. This presentation reviews the sensitivity and bandwidth improvements that can be expected from singlecrystal piezocomposites , and discusses our experience with the fabrication and evaluation of ultrasound transducers incorporating them. Performance and cost will be compared to current piezocomposite designs. The potential use of these materials in phased array probes will also be discussed. Introduction: Ultrasonic inspection is used in many industries. Wall thickness measurement of pipes, flaw detection in welds, characterization of material properties, delamination detection in composite materials are some of the uses of ultrasonic inspection [1]. The ultrasound transducer is a critical component in the inspection system. All information gathered about the test piece is obtained from signals passing through a transducer twice – once during transmission and once during reception. In many, not all, situations increased system bandwidth and improved signal-to-noise ratios can yield better test results. Transducer design is an important factor in determining bandwidth and s/n ratios. Those features of the piezoelectric material itself that relate to bandwidth and s/n ratio are thickness coupling factor, acoustic impedance, and dielectric constant. The thickness coupling factor is a measure of the piezoelectric material’s conversion of stored electrical energy into acoustic energy and vice versa. All other factors being equal, increasing the thickness coupling factor increases the sensitivity of the transducer and the bandwidth of the transducer. The acoustic impedance of the piezoelectric material determines the effectiveness of acoustic coupling to other materials. The closer the acoustic impedances, the greater the signal transfer. This also contributes to increased signal and bandwidth. The energy in the piezoelectric element causes it to resonate. As it resonates, some of the energy is transferred into the adjacent materials – the test piece on the front and the damping material on the back. If the match is poor, only a small portion of the energy is transferred on each cycle and the resulting signal in the test piece is relative long with low amplitude. If the acoustic impedance match is good, much more of the energy is transferred on each cycle resulting in fewer cycles, each of higher amplitude. Therefore, given a test piece material, it is desirable to use a piezoelectric material having a similar acoustic impedance, if possible. The dielectric constant, although an important factor in transducer design, will not be considered in this discussion. Because of the limited availability of single-crystal materials, dielectric constant is not really a variable. Transducer models enable researchers to estimate design performance without costly experimentation. Even when a model doesn’t yield exact results, trends can be determined that guide the design of experimental models. We have used the KLM [2] model for several years and will use it in this presentation to review the motivation for considering single-crystal materials. We have selected a 5 MHz, 6 mm immersion design as our test case. It is commonly used, so extensive data exists on the performance of monolithic and peizocomposite versions of this transducer. The predicted performance using monolithic PZT is shown in Figure 1. Figure 1 – Pulse-echo Impulse Response of a PZT Transducer It is apparent from the figure that this design is not acoustic impedance of the PZT element is relatively high, 3 materials with a sufficiently high impedance to damp the pi used quite effectively, is the use of lead metaniobate as the p impedance (20 MRayles), which is more effectively damped metaniobate itself has a lower mechanical Q, which contribu metaniobate into this design yields the performance shown i has improved, considerable signal amplitude has been lost b metaniobate. Still, this has been one of the more commonly resolution applications. As piezocomposite materials gained acceptance, some of th overcome. Piezocomposite materials are combinations of co yield improved piezoelectric performance. These materials, impedances of 10 MRayles, and lower, while still having co The polymer provides internal damping coefficients approac impedance of the piezocomposite, the acoustic impedance o a PZT –based piezocomposite material into this transducer d Figure 3 – Pulse-echo Impulse Response of a PZT Piezocomposite Transducer These designs have gained wide acceptance in many nondes Research has shown that single-crystal piezoelectric materia lead titanate, or lead zinc niobate and lead titanate, have an material, coupling factors as high as 0.9 should be achieved Figure 2 – Pulse-echo Impulse Response of a Lead Metaniobate Transducer most appropriate for high resolution testing. The 2 MRayles, and it is difficult to fabricate damping ezoelectric element. An alternative, which has been iezoelectric element. This has a lower acoustic by the same damping materially. Additionally, lead tes to improved pulse ringdown. Incorporation of lead n Figure 2. Although the resolution (pulse ringdown) ecause of the lower thickness coupling constant of lead used methods of producing transducers for high e deficiencies of lead metaniobate’s performance were nventional piezoelectric materials and polymers that with volume fractions around 30%, can have acoustic upling coefficients superior to monolithic PZTs [3]. hing lead metaniobate. Because of the low acoustic f the damping material can be reduced. Incorporation of esign yields the performance shown in Figure 3. tructive te ls having interesting [4] [5]. M Figure 4 – Pulse-echo Impulse Response of a Single-crystal PMN-PT Piezocomposite Transducer sting applications. compositions of lead magnesium niobate and property – when used as a piezocomposite odeling such a material yields the impulse response shown in Figure 4. To appreciate the improvement, note that the vertical scale has been changed by a factor of 5. These results are not necessarily optimized, that is, further changes could be made in each of these models that may improve performance for particular applications. We believe it is clear, however, that singlecrystal piezoelectric materials should provide significant improvements in ultrasonic transducer performance and are worthy of further investigation. Results: It is not the intention of the authors to rank the performance of materials from various suppliers because much of the final performance will depend on material processing. Therefore, suppliers will not be identified by name. The cost of these materials is quite high compared to more conventional materials. Several years ago the material was cost $500 $700 per square cm in small quantities. More recently, prices have decreased to approximately $ 200 per square cm. As available supplies increase, costs should decrease further. Single-crystal material was acquired from four suppliers. The physical appearance of the material differed greatly among the vendors. Some pieces were very transparent while others were only translucent. One vendor indicated that the differences in physical appearance were due to differences in manufacturing conditions but it seems that such differences might also affect piezoelectric performance. The reader is referred to other presentations to review the fabrication of 1-3 piezocomposite materials [2] but it involves dicing a plate of the piezoelectric material into an array of square posts, and then surrounding those posts with a polymer. This process is reasonably reliable and has been used at our facility for many years. The first attempt to dice the single-crystal material was unsuccessful. Even dicing in a single direction fractured the single-crystal material disqualifying it for further use. After reviewing the fabrication procedures presented by Michau [6], changes were made to slow the dicing process and increase the volume fraction from 30% to 60 % and success was finally achieved. Unfortunately, the resulting probes cannot be compares to PZT piezocomposite probes because they are not manufactured at this facility. Also, the material having a translucent, rather than transparent appearance was not diced successfully. It may be possible to do so with alternate dicing parameters, but this material seems much more fragile. The diced ceramic was backfilled with an epoxy having an acoustic impedance of 3 Mrayles. After curing the epoxy, the monolithic portion of the original ceramic plate was removed by grinding. A photograph of the composite is shown in Figure 5. Figure 5 – 60% volume fraction PMN-PT piezocomposite using slower dicing speeds Sputtered electrodes were applied to the samples and they were poled in a liquid bath at 25 degrees C with a poling field of 28 kvolts/cm. The relative dielectric constant and the thickness coupling coefficient are given in Table 1. Although the coupling coefficient is superior to PZT/epoxy piezocomposites, it is not as good as expected. TABLE 1. Properties of monolithic PZT, PZT piezocomposite, and PMN-PT single-crystal piezocomposite PZT(5H type) PZT Piezocomposite PMN-PT Piezocomposite Coupling coefficient (kt) .55 .61 .77 Frequency constant (kHz-cm) 202 155 108 Dielectric constant 380