HAL Id: jpa-00218886https://hal.archives-ouvertes.fr/jpa-00218886

Submitted on 1 Jan 1979

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Magnetostrictive underwater sound transducersR. Timme, S. Meeks

To cite this version:R. Timme, S. Meeks. Magnetostrictive underwater sound transducers. Journal de Physique Colloques,1979, 40 (C5), pp.C5-280-C5-285. �10.1051/jphyscol:19795103�. �jpa-00218886�

https://hal.archives-ouvertes.fr/jpa-00218886https://hal.archives-ouvertes.fr

JOURNAL DE PHYSIQUE Colloque C5, supplément au n° 5, Tome 40, Mai 1979, page C5-280

Magnetostrictive underwater sound transducers

R. W. Timme and S. W. Meeks

Underwater Sound Reference Detachment, Naval Research Laboratory, P.O. Box 8337, Orlando, Florida 32856, U.S A.

Résumé. — On envisage l'utilisation de certains alliages de terbium, dysprosium, holmium et fer dans des trans-ducteurs de son sous-marins. Les propriétés intrinsèques d'alliages sélectionnés sont présentées ainsi que leur importance dans la construction de transducteurs. La conception et les caractéristiques de fonctionnement d'un transducteur utilisant Tb0,27Dy0)73Fe2 sont décrites et comparées aux résultats du même transducteur utilisant une céramique piézoélectrique. La conception et les essais de ce transducteur ont clairement montré la nécessité d'amélioration des alliages, en particulier en ce qui concerne l'orientation de la texture, l'abaissement du coût, et leur production en série, avant qu'ils puissent être considérés comme compétitifs avec les céramiques piézo-électriques.

Abstract. — Certain alloys of terbium, dysprosium, holmium, and iron are currently receiving consideration for application in underwater sound transducers. The material properties of selected alloys and their importance in transducer engineering are presented. The design and operating characteristics of a transducer utilizing Tb0 27Dy0 7 3Fe2 are described and compared with results from the same transducer using piezoelectric ceramic. The design and evaluation of this transducer have clearly identified several needs for further development of the alloys, especially in the areas of grain orientation, lower cost, and quantity production, before they can be consi-dered competitive with piezoelectric ceramics.

1. Introduction. — Before the 1950's the choices of terbium (Tb), dysprosium (Dy), holmium (Ho), of energy-conversion materials available to designers and iron (Fe) have been investigated. of underwater sound transducers were limited to The objectives of this paper are to discuss why magnetostrictive metals such as nickel and its alloys those rare earth iron alloys have possible applications and natural or man-grown piezoelectric crystals such in underwater transducers, to describe the design and as quartz, rochelle salt, ammonium-dihydrogen-phos- evaluation of such a transducer, and to suggest what phate, and a few others. The magnetostrictive metals further improvements of the alloys are needed. were superior in ruggedness and reliability but were limited in efficiency and available strain. With the discovery in the 1950's of the piezoelectric qualities 2. Properties of the materials. — How does one of the barium-titanate and lead-zirconate-titanate decide the relative merits of different materials for polarized ferroelectric ceramics, use of the magneto- application to transducers ? Efficiency, power output, strictive metals declined. Thus, in recent years the receiving sensitivity, size, cost, and a dozen other use of ceramics in sonar transducers has been dominant considerations determine for a transducer engineer to a degree that it is now used almost exclusively, the usefulness of a material. Which considerations However, recent developments in magnetostrictive are most important generally depend on the specific materials have prompted further investigation of application. Most of these considerations however their application as acoustic elements. can be related to one or a combination of several

In 1963 and 1964 the rare earth elements terbium material characteristics such as piezoelectric or magne-and dysprosium were discovered to possess extremely tostrictive constants, dielectric or magnetic permeabi-large magnetostrictions, but these magnetostrains lity, elastic moduli, coupling constant, and strain were not immediately useful for application to levels. These characteristics depend on bias field, underwater sound transducers because they occurred bias stress, temperature, and hydrostatic pressure. only at cryogenic temperatures. Then in 1971, Clark Thus, a simple comparison of different materials and Belson [1] and Koon, Schindler, and Carter [2] often becomes difficult. discovered that the magnetostrains of binary alloys of Probably the most universally accepted charac-certain rare earth elements and iron at room tempe- teristic for comparing transduction materials is the rature are 10-100 times that of nickel. The possibilities electromechanical coupling constant, because this of application to sonar were immediately obvious. is a basic index of energy conversion capability. Since then, binary, ternary, and quaternary alloys The square of the coupling coefficient k is defined as

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19795103

http://www.edpsciences.orghttp://dx.doi.org/10.1051/jphyscol:19795103

MAGNETOSTRICTIVE UNDERWATER SOUND TRANSDUCERS

the ratio of the converted stored energy to the stored input energy and in a rather simplified manner can be written as

where E is Young's modulus and S is the strain. The numerator is the portion of the input energy that has been converted and stored as mechanical energy. The denominator is the input energy which is stored in the forms of mechanical work, the magnetic field, the magnetocrystalline anisotropy energy K, the internal strain energy y, and the demagnetization energy 6 due to end effects and internal inhomoge- neities. The coupling will have a value between zero and unity. Since the last three terms in the denominator represent loss mechanisms from the viewpoint of useful energy conversion, one can easily see the desirability of striving for low anisotropy, small internal strains, and homogeneous samples. The square of the coupling coefficient is related to the basic material properties by

where d is the magnetostrictive constant relating strain to magnetic field, pT is the free incremental permeability, and sH is the elastic compliance modulus at constant field. These symbols are consistent with the IEEE Standard on Magnetostrictive Materials.

For transducer application the attractiveness of the rare earth iron alloys can be seen from figure 1 where the coupling constant is displayed as a function of bias field for the best compositions of each ternary and the quaternary systems. The coupling constant is, of course, a strong function of alloy composition as reported previously [3, 51. All the alloy samples reported upon here were polycrystalline ; those desi- gnated (R) contained random grain orientations, and the two designated [l 1 l ] and [l 121 were grain oriented such that those crystallographic axes were parallel to the sample length and thus also to the applied magnetic field. The properties of these alloys are summarized in table I for the bias field producing

1; ?-oxide Annealed Nickel 0 . 2 b 1 ' 20 I 40 I 60 I I J

80 100

Magnetic Bias Field ( k ~ / m )

Fig. 1 . - Comparison of the magnetomechanical coupling constant as a function of magnetic bias field.

the maximum coupling constant. For comparison the coupling constants for a lead zirconate titanate (PZT-4) ceramic and nickel are also shown and are seen to be surpassed. The effect of grain orientation is obvious from the very large coupling constants thereby obtained. The values of 0.76 for

Tb0.23H00.77Fe2

and 0.82 for

Tbo.27D~o.73Fe2 [l 121

are the largest ever reported. The enormous strains obtainable from these alloys

are apparent from figure 2. The nickel saturates quickly to a negative strain of 36 ppm, whereas the rare earth iron alloys exhibit positive strains of hundreds of ppm at fields of 150 kA/m and, except for Tbo,23Hoo,77Fe2 [l 1 l], are still far from saturation. The magnetostrain and its dependence upon magnetic field are extremely important. The magnetostrain to the first power determines the sound pressure and to

Table I. - Materialproperties under the biasjeld condition for maximum coupling constant.

Material

Property k3 3 Hbias S*, d.3 e 3

(kA/m) ( P P ~ ) (PO) ( X 10-l1 m 2 / ~ ) - -

TbO .27D~o.73Fe2 0.82 10.6 145 5.3 3.68 Tbo.23H00.77Fe2 [l 'l] 0.76 10.7 145 5.0 2.85 T ~ o . ~ o H ~ o . ~ ~ D Y o . ~ z F ~ ~ (R) 0.62 15.5 145 5.9 2.82 T ~ o . ~ D Y o . , F ~ ~ (R) 0.60 23.2 208 4.4 2.81 Tbo.z5Hoo.75Fe2 (R) 0.43 25.0 137 2.9 1.09 Oxide annealed Nickel 0.27 6.0 20 30 0.51 PZT-4 0.70 - - -

CS-282 R. W. TIMME AND S. W. MEEKS

1000 -

900 -

Oxide Anneoled

- 50 ---------- I I I I I I 0 20 40 60 80 100 120 140 160 180 200

Magnetic Bias Field ( k ~ / m )

Fig. 2. -Comparison of the magnetostrain as a function of magnetic bias field.

the second power determines the acoustic power produced. Large strains at low magnetic fields are essential to avoid the engineering problems of pro- ducing and controlling larger fields.

In an application as a transducer the alloy will be placed in a bias field, since the magnetic remanence is too small, and the dynamic strain will be produced by an applied alternating field. With this dynamic strain S,,, and the elastic compliance coefficient at constant field sf3, the energy density E available is given by

A comparison is made in table I1 of the strains and energy densities that can be obtained under the condition of maximum power drive. All of the rare earth alloys shown have an energy density capability greater than the PZT-4 ceramic and very much greater

than nickel. The Tb-Dy-Fe alloys, either random or oriented grain, have the greatest promise because of their large coupling, large magnetostrain, and large energy density. More details on the properties of these alloys hav,e been previously published [6, 71.

3. Application of the materials. - Even before the properties of the rare earth iron compounds were completely known, plans were being made to incor- porate the new material into experimental transducer designs. The first designs were restricted to low-power and low-frequency operation because the alloys were available only in small quantities in the form of short, nonlaminated, arc cast rods with a random grain orientation. Terbium dysprosium iron alloys were used exclusively in the experimental transducers because of their large coupling and magnetostrain. To this date there have been projects within four groups in the U.S. to develop underwater sound transducers utilizing Tb-Dy-Fe : Raytheon Inc. [8], Honeywell Inc. [9], the Naval Ocean Systems Center [10], and the Naval Research Laboratory (NRL) [l l]. The effort at Raytheon concentrated on a segmented ring transducer while the work at the other three laboratories was on a longitudinal vibrator type. In the following, only the development at NRL will be described in detail because the conclusions reached are common to all the development efforts.

Figure 3 is an illustration of the NRL experimental transducer. This design was chosen because it is a well understood configuration which should serve

LEAD EPC

"O"RII

FRONT- MASS

Material

PZT-4 Oxide annealed Nickel Tbo.25Hoo.75Fe2 (R) Tb0.23Ho0.77Fe2 [l 111 Tbo.2oHoo..5*D~o.22Fe2 (R) Tb0.3Dyo.7Fe2 (R) T"0.27D~o.73Fe2 [l 121

F RdlRE EARTH lRON ROD

Fig. 3. - Experimental magnetostrictive transducer with Tb0.27D~0

ROD IS ' 1CK MASS

Table 11. - Material properties under the conditions for maximum power drive.

Property H,, (kA/m)

-

500 (kV/m) 4

28 14 23

113 33

MAGNETOSTRICTIVE UNDERWATER SOUND TRANSDUCERS C5-283

to demonstrate the capabilities of the Tb-Dy-Fe material. The design objective was to obtain a low- frequency, low-power, broad frequency range device. It was expected to be less than optimum because of the nonlaminated construction. The transducer was also to be adaptable for use with identically dimen- sioned piezoelectric ceramic rods for comparison of the two transduction materials. The active elements of the transducer in figure 3 are three rods of Tbo,27Dyo.7,Fe2, each 1.24 cm in diameter by 5.06 cm long. They are spaced at 1200 intervals on a 9.2 cm circle surrounding the center stress bolt. The stress bolt along the center axis of the transducer serves the double purpose of providing a compressive bias stress, which prevents the rods from fracturing under high dynamic drive, and provides a bias magnetic field to the rods from the solenoid wrapped upon it. Each of the rare earth iron rods is within a solenoid which can apply the alternating field as well as addi- tional bias field. A typical test condition could be obtained by applying 0.4 A to the center solenoid and 1.2 A to the rod solenoids. This produced a 0.6 T bias flux density (a 14.3 kA/m bias field) with an ohmic loss of 35 W. The circular piston at the front of the transducer is the vibrating mass and is constructed of magnetic iron. The size of this front mass (2 cm thick and 12.7 cm diameter) was determined by the desire for a low resonance frequency with a reasonable sound pressure output. The back mass of the trans- ducer has been made approximately 10 times the mass of the front piston to insure that practically all radia- tion is from the front.

Three ceramic rods of volume equal to the rare earth iron rods were constructed by stacking thickness- polarized PZT-4 discs (0.25 cm thick by 1.27 cm diameter) in order to compare the results of the rare

Frequency in kHt

Fig. 4. - Comparison of the transmitting voltage responses of the rare earth iron and ceramic transducers.

earth iron transducer with the results of the same transducer powered by ceramic.

The transmitting voltage response (TVR) at a bias flux density of 0.6 T, or a bias field of 14.3 kA/m, is shown in figure 4, and also compared to the response with the Tbo,2,Dy,~7,Fe, rods replaced by the ceramic rods. This figure shows the advantage of the rare earth iron transducer in terms of a greater sound pressure out per driving volt at the lower frequencies. The TVR of the rare earth iron transducer is 15 dB above that of the ceramic transducer at 500 Hz and 2.5 dB above at the resonance of 2.65 kHz. The effective coupling constant of the rare earth iron transducer (i.e., the coupling constant of the assembled transducer) at 0.6 T was measured to be 0.25. The reduction from the value of 0.6 for the alloy itself was primarily due to eddy currents and demagnetization. The measured effective coupling for the ceramic version was 0.5.

Figure 5 shows the free-field voltage sensitivity of the two versions. Here the ceramic transducer was superior to the rare earth iron transducer. In addition to being more sensitive the ceramic transducer was independent of frequency far below resonance which makes the ceramic transducer more desirable as a receiver.

Frequency in kHz

Fig. 5. - Comparison of the free-field voltage sensitivities of the rare earth iron and ceramic transducers.

At resonance both transducer versions showed a departure from linearity at high drive levels, but it appeared to be related more to the mechanical design than the transduction materials. The acoustic power output of the Tbo,27Dyo.7,Fe2 transducer at resonance was 9 W with a 400 V drive, which corresponded to a sound pressure level of 184 dB reference 1 pPa at 1 m. The ceramic transducer only produced 3 W under the same conditions.

C5-284 R. W. TIMME AND S. W. MEEKS

This transducer development and evaluation was successful in pinpointing problem areas. The Tbo,27Dyo,73Fe2 transducer was inferior to the cera- mic as a sound detector. It was marginally better as a projector, but further improvements are necessary. The transducer had large eddy current losses because the rare earth iron rods were not laminated. The large magnetic fields required will lead to a saturation of the magnetic return path without careful design. The conclusion is that ceramic transducers are not yet threatened by competition from the rare earth iron alloys.

4. Improvements for the materials. - The question, then, is : What is required of a rare earth iron alloy before it would be accepted by engineers as a direct competitor with ceramics in underwater sound trans- ducers ?

The biggest problem is the requirement of large magnetic fields to produce the maximum magneto- strain, hence requiring large volumes of copper which reduce the effective energy density. The energy density was defined above as the elastic energy per unit volume available to do work under the proper condi- tions. The effective energy density differs in that the volume of conductor required to excite the magneto- strictive material is now included. Table 111 compares the simple energy density and the effective energy density for the rare earth iron all6ys and ceramic. The assumption involved in this calculation is a volume of copper solenoid 8 times that of a Tbo,,Dy0.,Fe2 (R) rod is required to produce the dynamic magnetic field for maximum dynamic strain of that alloy. This has been borne out by the trans- ducer development work. Thus, the effective energy density for Tbo,,Dy0,,Fe2 (R) is not greater than for the ceramic. In the case of the Tbo,23Hoo~77Fe2 [l1 l ] alloy the maximum power drive is attained with a magnetic field 8 times less than for the Tbo,,Dy0,,Fe2 (R). This means the copper volume requirement is reduced about 92 X . The result is that the effective energy density of the Tbo,,3Hoo,7,Fe2 [l 1 l ] is larger than that of the Tbo,,Dy0,,Fe2 (R) even though its simple energy density is much less. Only the grain oriented Tbo~27Dyo.7,Fe2 [l121 alloy has an effective

energy density much larger than the ceramic, but even there such a large field is required that the simple energy density is greatly reduced. The conclusion to be drawn is that a large saturation magnetostrain cannot be the sole reason for choosing a magnetostrictive alloy as a transduction material. What is needed is a material such as at the bottom of table I11 that has a magnetostrain increasing very rapidly with field to a saturation of about 1 000 ppm at 40 kA/m. It appears this may be possible by grain orientation of the appropriate alloys.

A second factor needed to make rare earth iron alloys more competitive is a reduction of cost. The price per unit volume needs to be reduced to no more than about twice that of commercial piezoelectric ceramic. This might be accomplished through use of less pure rare earths if these impurities do not degrade performance significantly.

The third improvement needed concerns the pro- duction of the rare earth iron alloy. To date, these materials have only been produced in research labo- ratories in the form of short rods. It is necessary the materials be available in quantity in various shapes and sizes. Thin slabs (approximately 2 cm wide X 10 cm long X 0.2 cm thick) would be useful since they could be used as building blocks for laminated construction of other shapes.

5. Conclusions. - At this point the outlook for widespread use of rare earth iron alloys in underwater transducers is uncertain. The general engineering opinion is that these alloys do have great potential. Many of the material design goals previously esta- blished by the transducer designers and engineers have been met or exceeded. But, as is always the case, the price must be paid. For these alloys that price not only includes cost but also the need of large magnetic fields. In addition, some designers are hesitant to accept magnetostrictive alloys simply because magne- tic fields are inherently more difficult to route and control than electric fields. Thus, these alloys must be significantly better than ceramics to be accepted.

Continued development of the rare earth iron alloys is needed. A threefold thrust for the research is recommended :

Table 111. - Comparison of simple and efective energy density at maximum power drive.

Material -

PZT-4 Tb0.25H00.75Fe2 (R) Tb0.3D~o.7Fe2 (R) Tb0.23H00.77Fe2 [l ~ ~ , . 2 0 ~ ~ 0 . 5 , ~ ~ 0 . , 2 ~ ~ , (R) Tb0.27D~o.73Fe2 [l 121 Desired material

Property H,,ias (kA/m) H,,

- -

500 (kV/m) 48 28

159 113 20 14 36 23 48 3 3 20 14

MAGNETOSTRICTIVE UNDERWATER SOUND TRANSDUCERS CS-285

a) Further characterization of the properties of Although the rare earth iron alloys have not yet grain oriented alloys. been widely accepted as transducer materials, the path

b) Investigation of the effects of impurities on the to improvements is visible. One should not become

alloy magnetostrictive activity. discouraged, but rather remember these materials have not been known very long and, at a similar point

c) Development of methods of alloy production in development, ceramics were also looked upon with in quantity. some doubt.

References

[l] CLARK, A., BELSON, H., Proc. API ConJ 5 (1972) 1498. [6] CLARK, A., J. Underwater Acoustics 27 (1977) 109. 121 KOON, N., SCHINDLER, A., CARTER, F., Phys. Lett. A 38A 171 TIMME, R., J. Underwater Acoustics 27 (1977) 139.

(1971) 413. [8] BUTLER, J., CIOSEK, S., J. Underwater Acoustics 27 (1977) 165. [3] SAVAGE, H., CLARK, A., POWERS, J., IEEE Trans. Magnetics [9] AKRRVOLD, O., HUTCHINS, D., JOHNSON, R., KOEPKE, B.,

MAG-11 (1975) 1355. J. Underwater Acoustics 27 (1977) 183. [4] TIMME, R., KOON, N., J. A C O U S ~ SOC. Am. 58s (1975) S-74. [l01 SMITH, R., LOGAN, J., J. Underwater Acoustics 27 (1977) 175. [5] SAVAGE, H., CLARK, A., KOON, N., WILLIAMS, C., IEEE Trans. [11] MEEKS, S., TIMME, R., J. ACOUS~. SOC. Am. 62 (1977) 1158.

Magnetics MAG-13 (1977) 1517.