The transducer materials presently used for the probe design are limited concerning these demands. Indeed, there are high resolution probes on the market which are built by using lead metaniobate piezoceramic or piezoelectric polymers. However, these probes have low sensitivity as most of the acoustic energy is absorbed in the backing and the transducer materials efficiency is low. There are also probes of high sensitivity which are built by using standard PZT piezoceramic. However, these probes have minor resolution according to their design principles. In summary, the needs of high resolution and high sensitivity are antipodes for traditional probes, which often results in a compromise for the probe construction.
Second, a short pulse duration is necessary for good axial resolution, i.e. two signals in a short distance should be detected without interference. The task is e.g. to detect a small reflector near to the surface or backwall of the test object, as the inspection should scan the whole volume including the near surface regions.
The calculation shows us that the piezoceramic properties have been a limiting factor in transducer design. We will see that modifications of these properties will yield dramatic improvements in transducer performance. For immersion probes the ideal ceramic should have the following parameters:
A coupling coefficient as high as possible (kt = 1.0), as this is a measure for the conversion of electric into acoustic energy, which is the efficiency of the piezoceramic.
A low acoustic impedance near to water (Z = 1,5 Mrayl), to reduce the reflection losses between probe and water or avoid them completely.
An adjustable dielectric constant (10 < e< 5000), to match the electric impedance of the probe to that of the flaw detector depending on crystal size and frequency. This will improve the pulse form as well as the amplitude. Fig1. Sketch of 1-3 piezocomposite
Several types of piezocomposites have been investigated but today the 1-3 type piezocomposite is mostly used to build ultrasonic probes. Parallel orientated piezoceramic rods are embedded in a polymer matrix as shown in figure 1. The 1-3 type is manufactured by the dice and fill technique. A piezoceramic plate is diced into an orthogonal array by cutting 80% of the plate thickness. The pitch must be very small, typically e.g. 0,1 mm at 5 MHz, to avoid radial modes and grating lobes. Then the notches are filled with epoxy or other polymer material. When the epoxy is cured the uncutted side of the ceramic is grinded off. Afterwards, both sides are covered with electrodes followed by a polarizing process. This method makes it possible to fabricate piezocomposite plates with dimensions of 50 x 50 mm²or bigger, which are used to produce multiple transducers of smaller size by known methods as mechanical or laser cutting or dicing.
Piezocomposite ceramic can be tailored to our needs in contrast to conventional piezoceramic. The first parameter we can modify is the ceramic volume fraction. Fig. 2 indicates that the thickness coupling factor of the 1-3 composite is higher over a wide range of ceramic volume fraction between 15% and 95% than the coupling factor for PZT of about 0.52. Between 25% and 70% of volume fraction it is higher than 0.65. Within this range we can modify the volume fraction without losses in the coupling factor and thereby can adjust the acoustic impedance to our demands.
Fig. 2: Coupling factor as function of ceramic volume fraction
By variation of ceramic volume fraction and selection of the best fitting PZT material we can as well adjust the dielectric constant of the piezocomposite within a wide range. Therefore, we can choose the best piezocomposite material for each probe type to get optimum pulse form and amplitude. Finally, with the type of polymer material to be used we can modify the acoustical and mechanical performance of the piezocomposite transducer. Depending on the intended application we can use hard or flexible epoxy to influence the internal attenuation of the transducer. In addition, with the selection of the right epoxy we can reduce the planar coupling and thereby suppress radial modes which may deteriorate the resolution capabilities of the probe. By using a flexible epoxy we can bend the piezocomposite into a cylindrical or spherical shape. This allows us to build line or point focused transducers without the need for an additional lens in front of the crystal. Based on these results we have produced piezocomposite transducers and measured their data. In table 1 we compare these data to conventional PZT ceramic and PVDF polymer transducers. Table 1 indicates that the data of the piezocomposite ceramic are very close to our goal for the ideal ceramic as defined before. The acoustic impedance has been reduced dramatically (Z < 10 Mrayl), the coupling factor is higher (kt > 0,50) and the dielectric constant can be tailored to the specified range ( 100 < ( < 5000). In addition, the planar coupling factor is lower than for PZT. Concluding, we gain the following positive results:
|Table 1: Typical Data of Piezoelectric Materials|
| Concluding, we gain the following |
positive results: :
1. Improved electroacoustic efficiency
2. Reduced planar coupling
3. Low acoustic impedance
4. Mechanical flexibility
|Transducer Material||1-3 Composite||PZT (P5)||PVDF || Acoust. Impedance Z [Mrayl]|| 6,5 ||29,3 ||3,9 || Thickness Coupling Factor kt ||0,65 ||0,47 ||0,15 || Planar Coupling Factor kp ||0,20 ||-0,59 ||<0,1 || Transmission Coefficient (Perspex) T° ||0,884 ||0,355|| 0,990 || Pulse Echo Efficiency n [dB] ||16 ||0 ||-31|| Dielectric Constant e ||100 - 5000 ||900 ||12|| Max. Temperature [°C] ||150 ||300 ||80|
The specific features of piezocomposite material are of special interest when the transducer is coupled to material of low acoustic impedance. This is especially true for probes with plastic delay lines or wedges and for immersion technique or medical applications. These probes can be designed to have high sensitivity and high resolution as well. In addition, the effort for an effective mechanical damping can be reduced considerably.
As an experimental example for a 2 MHz angle beam probe we have directly glued a PZT as well as a piezocomposite transducer to a perspex wedge. In both cases we did not use any matching layer between the transducer and wedge and no backing behind the transducer. Fig. 3 indicates the advantages of the piezocomposite transducer: The probe with the PZT generates an long pulse which is unacceptable for most applications whereas the piezocomposite makes a short pulse with an extra 8 dB in amplitude.
Another example shows a 4 MHz longitudinal wave probe WSY70-4 normally used for testing of coarse grained austenitic material. In this application a high pulse amplitude is needed to compensate for the high sound absorption and scattering, and in addition short pulses are necessary to reduce the noise level which results from grain scattering. As shown by Fig. 4, the probe with the composite transducer has a 12 dB higher amplitude than with the lead metaniobate crystal. Additionally, the composite probe has a shorter signal with a bandwidth of 90 % instead of 70 % for the old design.
Fig. 4: Comparison between lead metaniobate (left) and 1-3 piezocomposite transducer (right) for a WSY70-4 probe
For immersion probes we also get similar improvements using piezocomposite transducers as demonstrated by the third example. In Fig. 5 we compare pulse form and frequency spectrum for a 2 MHz probe Z2K with 10 mm transducer diameter. The echo of the composite probe has 11 dB more amplitude and is clearly shorter than for the old design, also indicated by the increase in bandwidth from 45 to 76 %.
Such broadbanded probes with high signal amplitude are especially well suited for the nondestructive testing of plastic materials, composite structures, cast iron, austenitic steel and other highly attenuating materials. Depending on the application, composite probes will provide from 3 to 20 dB more sensitivity combined with a distinctly shorter pulse when compared to standard transducers. Due to their low acoustic impedance, composite transducers will especially improve the features of probes with plastic delay lines, e.g. dual or angle beam probes, and immersion probes. Piezocomposite probes are of exceptional advantage for automatic testing machines as here the combination of high sensitivity and extreme resolution is often required. For arrays and paintbrush probes the construction with piezocomposite transducers becomes substantially easier as here only a light backing is needed to produce high resolution probes. This light backing will reduce the mass of the probe thereby making the probe more resistant against thermal and mechanical stress which typically occurs in testing machine applications.
Today, composite transducers can be produced at acceptable costs when the necessary technological equipment is installed. The higher costs of the transducer are partially compensated by reduced labor needed to finish the complete probe. Therefore, the composite transducer will replace the conventional piezoceramic crystals where the technical advantages are clearly visible. However, not all probes need to be broadbanded. On the contrary, the flaw evaluation using DGS method requires narrow banded probes, i.e. the probes with conventional ceramic will maintain their field of application in future.