r AD-AlOO 595 NAVAL SURFACE WEAPONS CENTER ILVER SPRING NO FIG 20/3

MAGNETOSTR ICTIJON AND C HAR ACTERIZ AT ION OF BRIDGMAN GROWN RARE-KA-ElC(U)Nov g M T SA VAGE, R J ABBUNDI, A E CLARK

UNCLASSIFIED NSW C / T R S L Z9229NL

NSWC TR 81-29

MAGNETOSTRICTION ANDCHARACTERIZATION OF BRIDGMANGROWN RARE-EARTH IRON ALLOYS

BY H. T. SAVAGE, R. J. ABBUNDI,

A. E. CLARK

RESEARCH AND TECHNOLOGY DEPARTMENT

1 NOVEMBER 1980

Approved for public release, distribution unlimited.

rT .ECLGTE ffJUN 2 5 1981;

A

NAVAL SURFACE WEAPONS CENTERDahlgren, Virginia 22448 Silver Spring, Maryland 20910

t4-t6

UNCLASSIFIED

.ECU,,ITV .L.ASSIFICATION OF TNIS PAGE ,14hen Data Entered)

R DREAD INSTRUCTIONSREPORT DOCUMENTATION PAGE BEFORE COMPLETING FORM- I REPORT NuMaER 1z. G0VT ACCESSION NO., 3 RECIPIENT*S CATALOG NUMBERI LNS' TR-8-29 / I)

- 4 TIT (and Subtitle) S TY OF REPORT & PERIOD COVERED

MAGNETOSTRICTION AND CHARACTERIZATION OF BRIDGMAN// RPOWN RARE-EARTH IRON- ALLOYS •-RON R6. PERFORMING ORG. REPORT NUMBER

7- 7.AUHOR(a) S. CONTRACT DR GRANT NUMBER(*)T. avage" .I Abbundi- A. l1

9. PERFORMING ORGANIZATION NAME AND AODRESS 10. PROGRAM ELEMENT. PROJECT. TASK

AREA & WORK UNIT NUMBERS

Naval Surface Weapons Center (R4S) 7ro.gram element NIF, ProjectWhite Oak, Silver Spring, Maryland 20910 o. 0, Task Area No. 0, Work

I1. CONTROL.LING OFFICE NAMIE AND ADDRESS 1. . REPORT lATIE

1 NoMW 8' 813. NUMBER OF PAGfS

IA MONITORING AGENCY NAME 6 AOORESS'II different from Controlling Office) fS. SECURITY CLASS. (of this report)

UNCLASSIFIEDIS. OECLASSIPICATION, DOWNGRADING

SCHEDULE

16. DISTRIBUTION STATEMENT (of this Report)

Approved for public release, distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abetract entered in Block 20, if different from Report)

18. SUPPLEMENTARY NOTES

To be published in proceedings of the Fifteenth Rare Earth Research Conference(1981)

19. KEY WORDS (Continue on reverse aide it neceeesay and Identify by block number)

Magnetostriction, Rare Earth

20. ABSTRACT rContlnue on reverse side if necessary and identify by block number)

.We have prepared large samples of Th. ,Dy.73Fe 2 and Tb.,Ho . 5 8Dy . 2 2Fe2 by aBridgman-like technique developed at United Technology. The samples havemuch larger values of saturation magnetostriction than vertically zonedmaterials. The materials have been characterized by metallographic x-rayand TEM studies.

00D ' 1473 0 3F 1 sss c51 IS '61-S.

n L I

UNCLASSIFIED

SCJRITV CLASSIFICATION OF THIS PAGE (Whten Data Entered)

UNCLASSIFIEDSECUIqTY CLASSIrCA"O'I OF 'mIS PAGErW7"em Data Enreed)

NSWC TR 81-29

FOREWORD

The magnetic and magnetomechanical properties reported here aredevelopments of a program to develop magnetostrictive materials forhigh power sonar. The materials are an alloy of terbium-iron anddysprosium-iron Laves phase compounds. The magnetoelastic strains arevery large and very anisotropic.

Results of the first attempt to prepare large sample crystals ofthe highly magnetostrictive Tb compounds utilizing high temperature-gradient furnaces are reported here. This method has been identifiedas most promising to yield large single crystals (3" x " dia.) forsonar transducer elements. Nine crystals were prepared. Plane frontconditions were identified. Part of these results will be reported onat the Fifteenth Rare Earth Research Conference (Rolla, Missouri).

This study was carried out in the Solid State Branch of the Radia-tion Division. The materials development was sponsored by the NRLMaterial Program under the direction of Howard Lessoff. Magneticmeasurements and fabrication of prototype transducer components utiliz-ing these materials is being carried out under the sponsorship of theNOSC Sonar Transducer Sciences Program. Research on the magnetoelasticproperties of highly magnetostrictive rare earths is sponsored by theOffice of Naval Research and the NSWC Independent Research Fund.

H. R. RIEDLBy direction

1/22

NSWC TR 81-29

CONTENTS

Page

INTRODUCTION............................5

PREPARATION AND SAMPLE CHARACTERIZATION OF UNITED TECHNOLOGY

SAIPLES............................6

~4AGNETOSTRICTION.........................9

CHARACTERIZATION AND COMPARISON OF VERTICALLY ZONED SAMPLES .9

RECOMMENDATIONS.........................16

3/4

NSWC TR 81-29

INTRODUCTION

The cubic Laves-phase rare-earth Fe 2 compounds have huge valuesof magnetostriction constants (A) and anisotropy energy (E ). However,with an appropriate alloying of binary compounds, ternarie of the formRR xxFe2 may be obtained with large values of A but with Ek reducedby two orders of magnitude. Large values of the magnetomechanicalcoupling factor, k33, have been observed in Tb 2 7 Dy 7 3 Fe2 (Terfenol).

The magnetostriction is quite anisotropic with Xl00 << In 1.6xl- 3.Because the large, anisotropic magnetostriction introduces large inhom-ogeneous strains in random polycrystals (RPC), the permeability isquite low. 2 This restricts magnetomechanical activity.

We previously reported efforts in grain orienting to enhancermagnetomechanical activity. 3 Bridgman growth and horizontal zoningmethods were examined during FY78. During FY79, methods were developedto prepare high coupling oriented samples by a vertical zoningtechnique. 4 This material is prepared by passing a vertical moltenzone (VMZ) along a vertical rod. We achieved considerable improvementwith this over polycrystalline material. Coupling factors larger than.7 were achieved .n long rods. Random polycrystalline materials havevalues of .55. Raytheon Corporation is now building a transducer usinglong rods of vertically zoned material.

This report covers our first extensive attempt to obtain singlecrystal material in a manner suitable for large scale production oftransducer rods. The method has been developed by United Technologyfor single crystal turbine blades. We wish to adapt it to the produc-tion of single crystal Tb 2 7 Dy 7 3 Fe2 . The method is basically a Bridg-man growth technique. A boron nitride crucible is lowered through atemperature gradient. Since a large temperature gradient is necessarythe upper part of the crucible is held well above the melting point(called "superheat"). The ampule can be lowered into liquid to enhancethe temperature gradient. In this reoort we compare the VMZ materialto material prepared by the United Technology (UT) method.

1. A. E. Clark, AIP Conf. Proc. 18, 1015 (1974).2. H. T. Savage, A. E. Clark, and J. M. Powers, IEEE Trans. on Mag-

netics, MAG-II, 1355 (1975).3. H. T. Savage, R. Abbundi, A. E. Clark, "Permeabilit, , Magneto-

mechanical Coupling and Magnetostriction ir Grain-Oriented RareEarth-Iron Alloys," NSWC/WOL TR 78-197.

4. H. T. Savage, R. Abbundi, A. E. Clar] "Magnetomechanical Couplingand Magnetostriction in Vertically Zoned Tb.2 DY.7 Fe2,' NSWC TR79-463.

5

i r - .__ _. - " " , ,,,, ,',..I-

NSWC TR 81-29

PREPARATION AND SAMPLE CHARACTERIZATION OF UNITED TECHNOLOGY SAMPLES

We have grown nine samples. They are listed in Table I. Thesuper heat required for a high temperature gradient causes rare-earthloss and interaction with the boron nitride crucible. Without a hightemperature gradient, a large amount of Widmanstatten precipitate (WSP)is obtained. Fortunately our previous work has shown how to eliminateWSP. Material from Sample 2 was used to determine adequate heat treat-ment. Most WSP was removed by holding at 1000 C for 15 minutes andthen air or water quenching. The samples showed no damage due toquenching. Unless otherwise noted in Table I, we followed a procedureof starting the growth at 1 cm/hour for about 2 cm of growth and thenaccelerating the rate linearly to 10 cm/hour. We wished to establishhow fast we coild grow and still obtain plane front growth. Our workshows that plane front growth can be achieved up to 3 cm/hour growthrate. Samples 9 and 10 were grown to see if a fast growth rate orhigh Dy content (DyFe2 is much less peritectic than TbFe2) would sup-press WSP. They did not.

Sample No. 5 was our first truly successful effort. It will bediscussed in detail. As mentioned above, plane front growth wasachieved at a growth rate of 3 cm/hour. A 2 cm piece of the tip endof the boule was single by ordinary x-ray analysis. The boule isshown in Fig. 1. The arrow marks the transition to dendritic growth.The plane front end had an orientation of about 9 deg from <111>.Since <111> is the best possible direction we consider this a verygood growth direction when compared with the vertically zoned material.Sample 5 was split in half. One-half was subsequently annealed at650 C for 3 days. The unannealed half has maintained its physicalintegrity. The annealed half flaked gradually away. The 650 C annealhas apparently consolidated defects in regular pervasive ways. Theremay be an undetermined defect structure in the United Technology mat-erial that is not present in the vertically zoned material. The dif-ference in growth kinetics has influenced the defect structure. Themagnetostriction in this sample will be discussed in the next section.

We initiated a scanning transmission electron microscope (STEM)study to identify the defect structure (and look at domain walls).Sample preparation is difficult. By private communication with otherswho have tried, we think we are the first to successfully prepare asample. A tiny hole must be in the sample. At the hole edge, stressescause cleavage. The thin edge will, in general, oxidize away if exposedto air. We succeeded in preparing a sample by chemical attack. Thesample is kept under alcohol and then the alcohol is pumped off in-situ. Although personnel problems at United Technology hindered the

6

NSWC TR 81-29

TABLE 1

SampleNo. Composition* Comments

1 Tb 2 7DY 7 3Fel.9 5 Sample used to try out equipment.Result showed considerable Widmanstattenprecipitate (WSP). Probably some rare-earth loss due to superheat. Inter-action with crucible.

2 Tb.27Dy.7 3Fel. 95 Superheat reduced 100 F, sample used todetermine heat treatment for WSP removal,large amounts of WSP

3 Tb.27DY.73Fel.90 Rare-earth content increased. Evidenceof plane front growth near tip end wheregrowth rate was slow.

4 Tb1 2 0Dy.22 Ho.58Fel. 9 5 Considerable interaction with crucible.Auger analysis showed migration of Hoto surface.

5 Tb.DY.Fe First relatively successful effort -27Dy 7 3 e1 .9 5 discussed in text.

6 Tb.27Dy.7 3Fel. 90 Sample showed dendritic structurethroughout although pulled at constantrate of 1.5 cm/hour. Considerablereaction with crucible.

7 Tb 20Dy 22Ho 58Fe1 .9 5 Did not react with crucible as much as

previous quatenary, but extremelyfragile.

8 No data

9 Tb.15DY.85Fel.95 Solidified at 15 cm/hour, considerableWSP throughout

10 DyFe1.95 Solidified at 15 cm/hour, considerableWSP throughout

*Composition of starting charge. Prepared at Iowa State University.

7

*., _ -L ;.,' l . ' J q - ,-- :-

ni-&

r

NSWC TR 81-29

FIGURE 1 SAMPLE NO. 5. THE ARROW POINTS TO THE TRANSITIONFROM PLANE FRONT TO DENDRITIC GROWTH. THE TIP ENDIS THE PLANE FRONT END.

,i8

- ;.

NSWC TR 81-29

STEM work, we do have a TEM picture of one sample (Fig. 2). Thesecond phase has cracks which propagate into the parent phase. Thesample is from our second run which had a lot of WSP present. The WSPis apparently quite deleterious.

MAGNETOSTRICTION

Magnetostriction measurements were made in Samples 2 and 5. Sincethe results in 5 are much better, we will discuss them. Figures 3 and4 show the magnetostriction in the plane front and dendritic regionsof sample 5. The large value of magnetostriction is in agreement withthe x-ray data. The dendritic region does not show as high a magneto-striction thus showing that the preferred axis is "wandering." Theslope of the perpendicular magnetostriction is limited due to demag-netizing effects. The slope of the parallel magnetostriction is alsolimited by demagnetizing, but in addition by defects.

After a 650 C anneal the magnetostriction at 400 Oe. was studiedas a function of temperature, Fig. 5. The easy axis of magnetizationis set by stoichiometry. The maximum magnetostriction occurs at mini-mum anisotropy. Minimum anisotropy occurs at the easy axis transitiontemperature. The maximum magnetostriction (400 Oe.) occurs at 40 C.This implies a Dy rich condition, the opposite of our expectations.The reasons for this are unknown. The value of the saturation magneto-striction is 1l500 ppm as compared with %1000 ppm in the verticallyzoned material. Since the figure of merit goes as the magnetostrictionsquared, this is a considerable improvement over the vertically zonedmaterial.

CHARACTERIZATION AND COMPARISON OF THE VERTICALLY ZONED SAMPLES

We have studied the microstructure of the vertically zoned mater-ial (VZM) before and after annealing. A rod with a coupling factor of.6 before annealing was split down the middle. One half was thenannealed. The microstructure that developed after annealing is shownin Fig. 6. A featureless region that looks like plane front growth isinterspersed with a dendritic region that has long black "stringers"(Fig. 7). The "stringers" have been identified as rare earth oxide.A significant difference in permeability was found in different sec-tions of the rod. Different admixtures of the regions could be respon-sible for the changes in permeability. This would result in a lowercoupling factor (CF) than might otherwise be possible. The stringerscould be caused by constitutional super cooling (CSC). CSC can beeliminated by a slower growth rate. X-ray studies show the growthaxis considerably further from <111> than the United Technologysamples. This result is also confirmed by our magnetostriction measure-ments. We finally note that no Widmanstatten is present before orafter annealing, probably because of the high temperature gradient.We apparently need a higher temperature gradient in the United Technol-ogy method.

9 _ _ _ _ _ _1

NSWC TR 81-29

F*II

FIGURE 2 TRANSMISSION ELECTRON PICTURE OFWIDMANSTATTEN PRECIPITATE IN SAMPLE 2. THEPRECIPITATE HAS CRACKS WHICH EXTEND INTO THEPARENT PHASE.

10

NSWC TR 81-29

1400 SAMPLE 5 PLANE FRONT GROWTH END

00

00

1200

00

0 1000a.

M IL

800--0

I-0

I-

o600-

° 0

. 0400- 0

AXI. HE I MAUENTI WITH FIL PEPNILRT

FIGUR THE AMLETAIS.O NPAEFRN EINO SML .TE

11

4., ,- u . . ... " ' - ' , ..... . . . . .. . _

NSWC TR 81-29

SAMPLE 5 DENDRITIC GROWTH END

1000

0

600-

0I-.uj

400 -

200

8000 0

H(kOe)

FIGURE 4 MAGNETOSTRICTION IN DENDRITIC REGION OF SAMPLE 5. THE11 MEASUREMENT IS WITH THE FIELD PARALLEL TO THE LONGSAMPLE AXIS. THE i MEASUREMENT IS WITH THE FIELDPERPENDICULAR TO THE SAMPLE AXIS.

12

.I-

NSWC TR 81-29

7W0PLANE FRONT END

500

ca

2W0

100

-40 -20 0 20 40 60 90 100TEMPERATURE 1C)

FIGURE 5 LONGITUDINAL MAGNETOSTRICTION AT 400 Oe AS A FUNCTIONOF TEMPERATURE IN THE PLANE FRONT REGION OF SAMPLE 5.

13

NSWC TR 81-20

FIGURE 6 MICROGRAPH OF VERTICALLY ZONED, ANNEALED ROD,2X MAGNIFICATION.

14

NSWC TR 81-29

75um1

FIGURE 7 "STRINGERS" OF RARE-EARTH OXIDE INVERTICALLY ZONED MATERIAL OFFIG. 6, 200X MAGNIFICATION.

-. 15

NSWC TR 81-29

RECOMMENDATIONS

The goal of this program is to develop mass production techniquesto inexpensively produce large samples of highly magnetostrictive rareearth-Fe alloys. The alloys should possess magnetostrictions ofgreater than 1000 ppm at 500 Oe.

Widmanstatten-free Bridgman growth single crystals offer thehighest promise. These samples will possess the highest possible sat-uration strain (2400 x 10-6). Methods must be found to eliminatedefects which limit the magnetostriction obtained at 500 Oe. Ourrecommendations are: (1) for more complete STEM work to identifydefect structures; (2) determination of the phase diagram to eliminatedefects by controlled growth kinetics, (3) crucible improvement. Amost valuable study is the STEM because it will yield microscopicinformation on defects and structure of domain walls. X-ray topog-raphy will be done under a separate contract. The experiments will

complement each other. STEM looks at microstructure, topography atmore long range defects. The existence of a phase diagram will helpin the choice of starting alloys and growth conditions. It may bepossible to prevent the formation of a defect structure, rather thansubsequently eliminating it by heat treatment. There is a seriousproblem with crucible interaction with the melt. The last severalruns have been damaged by crucible breakdown. Alternate cruciblematerials calcia or calcia-zirconia should be investigated.

16

NSWC TR 81-29

DISTRIBUTION

Copies CopiesDefense Technical Information Center Naval Weapons CenterCameron Station Attn: LibraryAlexandria, VA 22314 12 China Lake, CA 93555

Dr. Howard Lessoff Naval Postgraduate SchoolNaval Research Laboratory Attn: Mechanical EngineeringCode 6820 DepartmentWashington, DC 20375 12 Monterey, CA 93940 1

Office of Naval Research Naval Air Systems CommandAttn: Dr. Robert Pohanka Attn: Code 52031800 N. Quincy Street 52032Arlington, VA 22217 1 Washington, DC 20361 2

David W. Taylor Naval Ship Commanding OfficerResearch & Development Center Office of Naval ResearchMaterials Department Branch OfficeAnnapolis, MD 21402 1 Building 114, Section D

666 Summer StreetNaval Underwater System Center Boston, MA 02210Attn: LibraryNewport, RI 02840 1 Commanding Officer

Office of Naval ResearchBranch Office536 South Clark StreetChicago, IL 60615 1

1

17

HNIIIMswc()I/1 (now. 5.76)

TO AID IN UPDATING THE DISTRIBUTION LIST-FOR NAVAL SURFACE WEAPONS CENTER, WHITEOAK TECHNICAL REPORTS PLEASE COMPLETE THEFORM BELOW:

TO tLL~HODETS OF NSWC/TR 81-29VSijO, TOd6

DO NOT RETURN THIS FORM IF ALL INFORMATION IS CURRENT

A. PACILITY NAME AND AOORI9SS (OLD) (Show Zip Code)

NE9W ADDRESS (Show Zip Code)

Il. ATTENTION LINE ADDRESSES:

C.

SREMOVE THIS FACILITY FROM THE DISTRIBUTION LiST FOR TECHNICAL REPORTS ON THIS SUBJECT.

HUMMER OF COP1ES DESIRED ___________________

DEPARTMENT OF THE NAVYNAVAL SURFACE WEAPONS CENTER

WHITE OAK, SILVER SPRING, MD. 20910 POSTAGE AND FEES PAID__________DEPARTMENT OF THE NAVY

DOD 316OFFICIAL BUSINESS

PENALTY FOR PRIVATE USE, $300

COMMANDERNAVAL SURFACE WEAPONS CENTERWHITE OAK, SILVER SPRING, MARYLAND 20910

1 ATTENTION, CODE R45