Manufacture of backings for TOFD ultrasonic transducers Nilka ACEVEDO, Carlos CORREIA and Roberto OTERO Foundation Engineering Institute Caracas, Venezuela; e-mail: [email protected], [email protected], [email protected]

Abstract For the construction of backings in piezoelectric transducers, it has been proposed different combinations of polymeric resins with metallic particles. With the purpose of optimizing the performance of transducers that are going to be used with the TOFD technique, there were made different composite samples of bakelite-tungsten with a relation of 0, 10, 20, 30, 40, 50 and 60 % of tungsten. The main acoustical properties were determined using the techniques of pulse-echo and transmission. It was studied the microstructures of samples, using Scanning Microscopy Electronic technique. The composite morphology was evaluated, specially the shape and the distributions of tungsten rich regions. A relation of bakelite-tungsten with good acoustic properties for TOFD ultrasonic transducers was obtained. An experimental dependence of the acoustic impedance and the attenuation coefficient with the volumetric percentage of tungsten was obtained for 1 MHz.

Key words: time of flight diffraction (TOFD), ultrasonic transducers, backings.

1. Introduction The piezoelectric transducers are devices used in ultrasonic for the emission and reception of mechanical waves. These systems transform the electric power into mechanics waves and vice versa. They present principally three parts: the piezoelectric ceramics, the coupling layer and the backing. For the ideal performance of transducers, the backing material should have specific acoustic properties. In general the acoustic impedance should be similar to that of the piezoelectric ceramic (matching impedance condition) and at the same time it needs a high attenuation coefficient. Generally, the materials used as backings are polymers, but in the last decades, with the aim to improve the performance of different transducers, different authors have studied the acoustic properties of composites. Among the composites that have showed better results it has been proposed different polymeric resins combined with metallic particles of silver, gold, oxide of lead or tungsten (W), among others [1, 2]. Among the most important studies in the field of optimization of backings, we found the previous works of Grewe et al [1], Faguaga et al [2] and Wang et al [3]. In the study realized by Grewe et al [1] it was determined the properties of acoustic impedance and attenuation of a great diversity of simple and composite materials. Several materials were obtained combining polymer resins and metallic particles. In this paper it was used tungsten and alumina. The acoustical properties were determined using the transmission technique in immersion for 5 MHz. Faguaga et al [2] designed the transducer backings using particle and polymers composites. It was studied the acoustical properties of two composites using the same resin and with metallic particles of oxide of lead and alumina. They obtained the material with the best acoustic properties.

In the paper realized by Wang et al [3], they measured the properties of acoustic impedance and attenuation for two composites using the same polymer resin and metallic particles of alumina and tungsten in a range of frequencies from 25 to 65 MHz. Generally is very difficult to find the appropriate technical information to develop specific ultrasonic transducers because it is part of the “know how” of commercial companies. The information available in the technical literature is mainly devoted and focused to academic purposes or other applications. In our work, we decided to develop ultrasonic TOFD transducers using piezoelectric composites with an acoustic impedance of 10 MRayls approximately. The main goal of the present work is to determine the volumetric proportion of bakelite-tungsten that allows us to obtain a relatively simple and cheap material with good acoustical properties for transducer backings to be used in TOFD ultrasonic transducers. For that reason, we will study the properties of acoustic impedance and acoustic attenuation of these samples using the ultrasonic techniques of pulse-echo and transmission, with a frequency of operation of 1 MHz.

2. Experimental Procedure 2.1. Initial characterization of materials The initial materials were powders of bakelite and tungsten. For both components there were determined the density and the range of particles average size. For tungsten, both properties were obtained from the manufacturer. The density was 19.3 g/cm3 and the range of particles average size was 0.6 – 1 µm. In case of bakelite, both properties were established experimentally obtaining an average density of 1.53 g/cm3 and a range of particles average size from 45 to 75 µm. In principle, we could have a serious mismatch to obtain homogeneous composites because the difference of particles sizes from both components. 2.2. Manufacture of the samples The samples were obtained mixing mechanically the dry powders using both components, in quantities previously calculated according to the volumetric percentage. In the manufacturing process it was applied an axial pressure and an increase of temperature up to reaching to 120ºC. For this it was used a hydraulic press with a stove (see the Figure 1). The face on which the pressure was exercised on the sample was identified as the top part and the other end as the bottom. Seven cylindrical samples (base diameter: 30 mm and height: 25 mm) were obtained with the following volumetric proportions: 0, 10, 20, 30, 40, 50 and 60 % of tungsten and the rest of bakelite.

Figure1. Hydraulic press with a stove used to manufacture the backings of bakelite and tungsten.

2.3. Obtaining the microstructures using the Scanning Electronic Microscopy (SEM) technique Four surfaces were prepared at different heights, with the purpose of evaluating the distribution of tungsten inside the bakelite. On every surface, micrographs of SEM were taken in the center and at the edge in every sample. The magnifications used were 20X, 200X and 15000X. For 200X it was applied the Manual Point Count technique [4]. 2.4. Measurement of acoustical properties The acoustical impedance, the density and the ultrasonic velocity were measured for each sample. The measurement of ultrasonic velocity in each sample was obtained using a classical transmission technique setup. To determine the attenuation coefficient the immersion pulse – echo technique was used following the same experimental procedure introduced by Graciet et al [5]. The working frequency was 1 MHz. The samples and the ultrasonic experimental setup are shown in Figures 2a and 2b respectively. The acoustical measurements were made using the Acoustic Intensity Measurement System (AIMS) from ONDA Corporation.

Figure 2. The samples mounted (a) in the System AIMS (b).

3. Results and Discussions From the micrographs obtained a qualitative and a quantitative analysis could be done. From the qualitative analysis were observed two regions: a tungsten rich region, with a round shape of agglomerations and another region, which surrounds these agglomerations. The agglomerations of tungsten show a spherical morphology. The region around the agglomerations was constituted by both components: bakelite and dispersed tungsten. It was experimentally demonstrated by Energy Dispersive Spectroscopy technique. From the quantitative analysis, it was obtained the axial and radial distribution of tungsten and it was homogeneous in all samples. The differences in tungsten percentage composition from the top to bottom were not statistically representative. We obtained the same behaviour comparing the microstructures from the center to the edge in all samples. In Figure 3 it is shown the microstructures taken from the central region of every disk surfaces (A, B, C and D) of a sample with 30 % of W.

Figure 3. - Microstructures of different surfaces in the sample with a proportion of 30 % of tungsten.

The acoustical properties of impedance and attenuation present different trends related with the volumetric percentage of W in the samples. These results are presented in Figure 4. The acoustic impedance shows an increasing trend with the percentage of tungsten in the material (see Figure 4(a)). On the other side, the attenuation coefficient does not show a definite trend increasing %W in the samples (see Figure 4(b)).

Figure 4. - Acoustic properties related with the volumetric percentage of tungsten in samples, (a) Acoustic Impedance and (b) Attenuation Coefficient.

From the previous results we decided to use the composite material obtained for 30 % of tungsten as a backing to fabricate a TOFD transducer. In the process of fabricating a TOFD transducer other parameters needed to be optimized too. A good damped and a high sensitivity TOFD

transducer was obtained. In Figures 5a and 5b it is shown the propagated ultrasonic pulse (and his FFT) in steel, using this transducer.

Figure 5. - Ultrasonic pulse and his FFT obtained in steel (a) using the developed TOFD transducer (b).

4. Conclusions A relatively simple and cheap material with good acoustical properties was obtained to fabricate a TOFD transducer. A relationship between the acoustical impedance and volumetric percentage of tungsten has been experimentally obtained for the fabricated samples. We propose for future studies, a most precise determination of random and systematic experimental errors obtaining the attenuation coefficient for these materials. At the same time it should be studied other possible materials with different grain sizes distributions. This manufacturing process allows us to obtain an axial and a radial homogeneous distribution of tungsten particles inside the bakelite, without taking into consideration the important differences between the particles sizes of initial materials. The composition with 30 % of tungsten was chosen as the best backing for constructing our TOFD transducers. As an example, a good damped TOFD transducer was presented and in the near future more experimental results will be presented. Acknowledgements This work was supported by the Project No. 1-41-07001 at the Centre of Materials Technology from the Foundation Engineering Institute which belongs to the Ministry of Science and Technology, Caracas, Venezuela. References 1. M Grewe, T Gururaja, T Shrout and R Newnham, ‘Acoustic Properties of Particle/ Polymer Composites for Ultrasonic Transducers Backing Applications’, IEEE

Transactions on Ultrasonics, Ferroelectrics and Frequency Control, Vol. 37 (1990) 506514. 2. M Faguaga, G Machado and A Moreno, ‘Diseño, Fabricación y Caracterización de Transductores Piezoeléctricos de Ultrasonido para su Aplicación en END’, http://www.cori.unicamp.br/jornadas/completos/UDELAR/ND8001-FAGUAGA.doc 3. H Wang, T Ritter, W Cao and K Kirk, ‘High Frequency Properties of Passive Materials for Ultrasonic Transducers’, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, Vol. 48 No. 1 (2001) 78-84. 4. ‘Standard Test Method for Determining Volume Fraction by Systematic Manual Point Count’, ASTM E 562 – 02, Vol. 3.01, (2004). 5. C Graciet and B Hosten, ‘Simultaneous Measurement of Speed, Attenuation, Thickness and Density with Reflected Ultrasonic Waves in Plates’, IEEE Ultrasonic Symposium, (1994) 1219-1222.