Magnetomechanical damping behaviour in pure nickel and a 20 wt.% copper-nickel alloy

MAGNETOMECHANICAL AND

DAMPING

A 20 WT.% J. T.

A.

BEHAVIOUR

COPPER-NICKEL

ROBERTS?

and

P.

IN

PURE

NICKEL

ALLOY*

BARRANDt

A low frequency internal friction technique has been used to investigate the influence of magnetocrystalline anisotropy on the frequency of Berkhausen jumps. The peak magnetic damping parameter Ana is directly related to the jump number, following correlation of these results with observations on the temperature dependence of the Barkhausen fraction of total magnetization. In pure nickel, the temperature dependence of Am is explained qualitatively by two types of stress induced domain movement; the larger reordering domain boundary movement is strongly dependent on the presence of The “two peek” phenomenon in the 20 wt.% copper-nickel alloy is magnetoorystrtlline anisotropy. possibly due to domains overcoming a second stress peak derived from the ordering of solute in domains; in this case a linear Am-temperature plot was observed. AMORTISSEMENT

MAGNETO-MECANIQUE ALLIAGE 20% PDS

DANS LE NICKEL CUIVRE-NICKEL

PUR

ET

DANS

UN

On a utilis6 une technique de frottement interne zt basse frbquence pour etudier l’influence de l’anisoD’aprBs 18 corr&tion entre ces tropie magn&o-cristalline sur la frkquence des sauts de Berkhausen. r&ultats et l’effet observe de la temp&ature sur la fraction d’aimantation totale de Barkhausen, le param&e d’amortissement magnbtique au pie, Awa, est directement lib au nombre de sauts. Dans le nickel pur, l’effet de la temp6rature sue Ana s’explique qualitativement par deux types de d6plecements de domaines sous l’effet des oontraintes; le d6placement le plus grand de remise en ordre des joints de domaines depend fortement de la p&ence de l’anisotropie magn&o-cristalline. Le ph&om&ne des “deux pits” dans l’alliage 20% pds cuivre-nickel est peut-&re dti B ce que les domaines franohissent un deuxibme pio de contraintes provenant de la mise en ordre du solute dans les domaines; dans ce c&s, on observe une relation lin&ire Am-temp&ature. MAGNETOMECHANISCHES EINER

DBMPFUNGSVERHALTEN VON REINEM 20 GEW.-y. KUPFER-NICKEL-LEGIERUNG

NICKEL

UND

Mit Hilfe der inneren Reibung bei kleiner Frequenz wurde der Einfluss der magnetokristallinen Notch Korrelation dieser Ergebnisse Anisotropie auf die Frequenz von Barkhausenspriingen untersucht. mit Beobmhtungen der Tempereturabhiingigkeit des Barkhausenanteils der gesamten Msgnetisierung steht der Parameter Alra des Maximums der magnetischen Diimpfung direkt mit der Sprungzahl in Zusammenhang. In reinem Nickel wird die Temperaturabhiingigkeit von Am qualitativ durch zwei Typen von spannungsinduzierter Wandbewegung erkliirt ; die grossere ordnende Wendbewegung Das Auftreten zweier Maxima in hiingt stark von der vorliegenden magnetokristallinen Anisotropie ab. der Legierung h&ngt miiglicherweise mit der tfberwindung eines zweiten Spannungsmaximums, welches In diesem Fall wurde eine von einer Ordnung der Fremdatome in Bereichwanden herrtihrt, zusammen. lineare Abhiingigkeit des Am von der Temperatur beobachtet. INTRODUCTION

Stress-dependent

damping

arises from the irreversible and magnetic established

domain

The variation of the magneto-elastic

in magnetic

movement

It is now well hysteresis in such mater-

ials, like its magnetic

counterpart,

irreversible

of 90” domain

Barkhausen proposed placements

jumps.(1,2)

that

of dislocations

boundaries.

that mechanical movement

materials

Sumner

the stress-induced

at peak magnetic

same as the magnetically

arises from

the

boundaries

by

and

damping

induced

point of maximum susceptibility theory it can be shown that

Entwistlec2)

domain

wall dis-

Am were the

movements k,.

at the

amplitude

of

distribution

of internal-stress

domain

vibration

structures.

The

through a maximum, strains,

since

depends

decrement with on

the

precise

peaks, and on the initial

decrement

always

passes

and tends to zero at very large

magneto-elastic

strains

will cease

to

exist at stresses in excess of those necessary for nearmagnetic saturation.

The peak decrement is indicative

of the state of stress in the material, e.g. Sumner and Entwistlec2) observed that specimen showed a peak

a fully height

annealed nickel with a specific

Using domain

damping value of 0.52, at a surface shear strain of 4.0 x 10-5. A plastically deformed specimen exhibited a much lower level of damping, 0.031, and there

where 1, is the saturation magnetostriction in the easy direction of magnetization, J, is the saturation

the height of the internal stress peaks influences the

was a peak shift to a shear strain of 36 x 10-5.

intensity, and H/CTis the computed ratio of field at k, to the stress at A, required to move a 90” boundary the same distance. This has been confirmed by measurements

on nickel, iron and mild steel.(s)

*

Received March 31, 1967. t Department of Metallurgy, Manchester, England.

University

ACTA

15, NOVEMBER

METALLURGICA,

VOL.

of Manchester, 1967

number

of Barkhausen

jumps occurring,

Since

there must

be a direct relationship between Am and the number of jumps. So far all measurements of amplitude-dependent magnetomechanical damping have been at room temperature, and therefore the effect of crystal anisotropy could not be accounted for. KosteF) observed a damping peak at 200°C in the damping 1685

ACTA

1686

curve

of pure nickel

at constant

This has been interpreted(l) of the hysteresis constant of

magnetocrystalline

strain amplitude.

in terms of a large increase between room temperature

and ZOO”C, and is associated investigation

METALLURGICA,

with the disappearance

anisotropy.

it was thought

that

In

the

present

the influence

of

crystal anisotropy on the ease of domain boundary movement might best be studied by investigating the effect

of temperature

amplitude

dependent

the crystal anisotropy

on the damping curve.

peak in the

The assumption

will provide

is that

a further barrier

VOL.

15, 1967

for the domain boundaries

to overcome,

and this will

be strongly

dependent.

Nickel

particular

temperature

interest since the anisotropy

shows such a great variation in magnitude to the Curie point.c5)

Sn initial study

of alloying on the magnetic also described. EXPERIMENTAL

The nickel was obtained

constant

is of K,

on heating of the effect

damping in pure nickel is

PROCEDURE

as spectroscopically

pure

rod, 5 mm dia. x 15 cm long, from Messrs. Johnson

FIG. 1. Torsion pendulum suspension.

ROBERTS

Mathey

Limited.

Metallic

in total, with ~20

AND BARRAND:

impurities

ppm of oxygen.

MAGNETOMECHANICAL

were ~8

ppm

annealing.

then recrystallized hydrogen

Specimens

~3

and stress relieved

in. long were by heating in

at 1OOO’C for 24 hr. A copper-nickel

containing

20 wt.%

copper, was prepared by melting

spectroscopically

pure copper

frequency

having

unit

prevent

alloy,

oxidation.

and nickel

an

To

argon

facilitate

in a high

atmosphere melting,

the

1687

_I

. Spewnen 0 Specimen X Magnetic

The rod was cold

swaged and cold drawn down to 0.050 in. dia. with no intermediate

DAMPISG

4-

0 X E E f :: n

unmagnetlzed mognetued to saturation damping contnbution

.I

3-

to re-

crystallised alumina boat used as a mould was contained in a larger graphite boat which acted as a susceptor. broken

After casting, the dendritic

down

by

swaging,

followed

by

amleal for 1 week in order to homogenize.

the

The ingot

The actual specimens were subsequently pretreatment

same

The magnetic

measurements

were carried

out using a free-decay technique in an inverted torsion pendulum. ~10-~

The whole apparatus

was evacuated

torr in order to reduce specimen oxidation

to keep the level of background

damping

low.

to and Low

instrumental damping (log. dec <10P5) was achieved using a torsionally-rigid balance arm to support the inertia assembly, men suspension low

intrinsic

in addition (84 in.

damping,

wound non-inductively non-ferromagnetic. were obtained mixture

to a quartz fibre speci-

0.020 in.) having extremely Fig.

1.

The

using nichrome

Temperatures in a cooling

a narrow

was

dioxide-acetone

jacket

of the free decay

by recording

furnace

wire which is

as low as -40°C

using a solid carbon

contained

the furnace. Measurement made

x

surrounding

of amplitude

slit of light,

Max.

shear

strain

10-3

amplitude

2. Relation between damping and surface shear strain for pure nickel at -40°C. Frequency -1 c/s.

FIG.

given

as the pure nickel specimens.

damping

10-4

C5

a 1000°C

was then swaged and drawn down to 0.030 in. dia. wire.

,

structure was

was

ment.

There was a good correlation between damping

peaks reported by Sumner and Entwistle,‘2) and those measured damping

using the present

The magnetic was

technique,

for a high

material such as pure iron.

obtained

contribution by

to the overall damping

making

measurements

in

the

presence and absence of an axially applied saturating magnetic field. RESULTS

Figures 2 and 3 show typical damping peaks for the same specimen tures.

of pure nickel at differing

All the damping measurements

tempera-

were made at a

frequency ~1 c/s. The “magnetic peak” is obtained by subtracting the damping contribution due to dislocations

only, i.e. in a saturating

t’ot’al damping

curve.

Figures

field, from the

4 and

5 show

the

reflected

from a concave mirror, attached to the inertia system, on to the photocell movement metric

paper.

accurate

of a photodyne

was converted This

determination

provided of both

and the shear strain amplitude. decrement

was plotted

recorder.

The slit

to a wave trace on millia

method

for

the

the log. decrement In all tests, the log.

as a function

of maximum

surface shear strain, 4, at each amplitude of vibration. This is given by 4 = ry/l, where r and 1 are the radius and length of the specimen and y is the angle of twist in radians. By varying the distance of the recorder from the vibrating system a shear strain amplitude range from 1 X 10--6 to 4 x lop4 was possible. It can be shown that the maintenance of constant strain amplitude during the measurement of log. decrement, as is obtained using high frequency oscillator techniques,c2) is not imperative as long as the amplitude is not allowed

t’o decay

SlOO/o during each measure-

Max. FIG.

sheor

strafn

amplitude

Relation between damping and surface shear strain for pure nickel at 214°C. Frequency -1 c/s. 3.

ACTA

METALLURGICA,

VOL.

15, 1967

Curie point

at 360°C.

A further interesting

is that the maximum peak is approximately In

addition

feature

shear strain amplitude independent

Ki%ter’sc3)

work

at the

of temperature. on

the

constant

amplitude

temperature dependent damping of pure nickel was studied. As Fig. 6 shows, a peak is observed at 15O”C, this being similar to KBster’s peak which was observed

at 200°C.

Measurementi

on specimens

having various prior annealing treatments, low strain amplitudes, the range 0.2-2.0

at different

and at several frequencies

c/s, revealed the following

in

features,

some not included in Fig. 6. Specfmen unmagnetired x Specimen magnetized to saturation

12--

l

IO -~

Max. shear sham FIG.

4.

variation

omplifude

Magnetic damping peaks in pure nickel.

in peak damping,

tures during

heating

Am, at several tempera-

to the Curie point.

The main

features are firstly the rapid fall in peak log. decreto 200°C and then a ment on heating from -40”

1

0

plateau from 200” to 300°C followed by a fall off to the

;I

I

200

100

300

Temperature,

400

“C

Fra. 6. Damping curve for pure nickel at a constant shear strain amplitude of 6 x 10-O with associated dynamic modulus variation (0). Frequency 2.0 c/s. The peak annealing strain

16 -

height was dependent both on prior treatment, and the operative shear

amplitude,

i.e. fully

annealed

and the highest strain amplitudes,

specimens,

independently

gave the highest peaks. The peak was broader relaxation independent

\

than expected

remove damping

was

of frequency.

Figure 6 shows that an axially

x

for a single

process, and the peak temperature

the peak,

applied

field will

but at low temperatures

in the presence

of the magnetic

the field

rises rapidly. A feature in all these tests was the peak in the

\

dynamic modulus curve (plotted as f2), shown in Fig. 6. The peak was located at the Curie point

4-

x*--*--.

*\ 2-

\: -50

I

0

I

50

I

I

100 IS0

I

/

I

200

250

300

Temperahrre,

FIG. 5. The effect of tevtr;aF;e

Curie point \’ 550

400

OC

on the peak damping

(360 f 10%) and was independent of frequency and grain size. Beyond the peak, on the low temperature side, the modulus curve suffered a sharp fall off to a minimum point, at 160 f 10°C; this was followed by a further rise in modulus lower temperatures.

at

ROBERTS

AND BARRAND:

l

Max.

shear stroin

MAGNETOMECHANICAL

1689

DAMPING

Specimen unmognetlzec

omplttude

FIG. 7. Relation between dramping snd surf&e shear strain for a 20 wt.% copper-nickel alloy et -39°C.

A similar series of amplitude-dependent damping m~suremen~ was carried out on a specimen of 20 wt.% copper-nickel alloy. Figures 7 and 8 show typical damping peaks at two different temperatures. It appears that at higher temperatures, and to a lesser extent at low ~mperatures, the magnetic damping peak separates into two distinct components ; the peak shear strain again appears to be independent of temperature. This trend is illustrated in Fig. 9. In Fig. 10 the variation in peak log. decrement, Am, with temperature is plotted for both peaks. The peak at lower strain amplitudes bears a linear relationship to temperature, and extrapolates to zero at the Curie point, 140%. The variation in Am with temperature for the second peak is more difficult to measure as it is not very prominent. It does not however, appear to vary linearly.

‘

Max.

Fm.

sheor strain

amplitude

9. Magnetic damping peaks in 8 20 wt.% nickel alloy.

copper-

DISCUSSlON

Prior to this investigation Stierstadt, and Boeckhc4) measured the size distribution of Barkhausen discontinuities (magnetic moment H > 5 X lo-’ cgs), of a high purity, hard drawn nickel specimen, from the curve of initial magnetization and the hysteresis loop, in the temperature range -170°C to +36O”C. As Fig. 11 shows, the temperature dependence of the Barkhausen fraction of total magnetization, J,IJ,, follows the same trend as the peak magnetic damping

x First

* Spectmen unmagnetized 0 Specimen mognetized to saturation X Mognetlc domping contribution

l

3-

peak

Second peak

‘.

2-

I-

“\$.i -50

Max.shearstrain

amplitude

Fro. 8. Relation between damping and surface shear stmin for s 20 wt.% copper-nickel slloy st 73°C.

!

I 0

r IO0

so

Temperature,

FIG. 10. Effect of t;;reritz

4 ,

OC

on the peak damping

ACTA

METALLURGICA,

VOL.

15,

1967

feature of the curve is the occurrence of three different stages

during

heating

to the Curie point.

region where magnetocrystalline ficant

a greater

number

In the

anisotropy

of Barkhausen

taking place at the peak stress;

is signi-

jumps

this number

are

being

rapidly reduced on heating to a temperature where the anisotropy is destroyed. The nickel wire used in the present investigation

was fully stress relieved, and

so the internal stress pattern would not be expected to alter appreciably

over the temperature

Thus the number of Barkhausen

stress peaks would remain constant. observed,

which is operative

crystalline

anisotropy

over internal Temperature,

*C

stadt & Boeckh’4’).

AWL. The parameter J,

measure

of

the

evidence

for this proposal

crystal imposed

loop and vp = 3.53 x lop3

of the specimen.

This supports

that Am is related to the number

curve obtained

from internal

walls

based

the

of Bark-

anisotropy.

was associated

friction

measurements

of

They

At -100°C

to the temperature

at

of domain

which magnetocrystalline anisotropy vanishes. Brailsford(a) has shown that between -200°C and +ZOO”C

~30-60 areas.

there

amplitude

negative

anisotropy

form was super-

wave variation

whereas the amplitude

changes only decreased slightly. man.n(@ observed two clearly Barkhausen

of the onset

On this wave

with the internal-stress

small jumps ~10

rise in the

model.

of E which energy.

On

heating to + 100°C the long wave changes disappeared completely

their curve and the

In the latter case the temperature

is a sharp

the movement

of the wall, x.

a low amplitude

scale.

corresponds

Experimental

on a potential

of the position

lies in the position of the plateau on the temperature of this plateau

involved.

comes from two sources.

the free energy, E, for the movement of a wall in a direction normal to the wall, as a

function

where Ni is the overall number of jumps of moment

hausen jumps occurring. The main difference between

number

Stierstadt and Boeckhc4) considered

Mi along the hysteresis postulate

a steady-state

they observed a long wave change of energy, i.e. with large amplitude, which was related to the energy of

VP

cm3, the volume

The domain jumps

stress peaks will form

the value of Am at the plateau being a

plotted domain

NiMi

J,=C-

only when the magneto-

is present.

contribution,

domain

is given by

Since this is not

a second type of Barkhausen jump must be

invoked,

FIG. 11. The temperature dependence of the Barkhausen fraction of the total magnetization, J&7, (0) and of the maximum differential susceptibility, KgBX (0) (Stier-

range studied.

jumps over internal

of the short wave Haacke distinct

jumps in nickel single crystals,

and Jautypes of namely,

,u, which derived from the pinning

walls at Neel spikes,

and larger jumps

p which related to the reordering of larger Therefore it can be assumed that the low jumps

are isolated

movements

of single

constant, K,, and a smaller drop in the positive K, value ; at temperatures in excess of 200°C the aniso-

domain walls over internal stress peaks and the large

tropy is absent.

several

Similarly,

related the temperature strong

decrease

in

Stierstadt

dependence

and Boeckh(*) of J,/J,

magnetocrystalline

to the

anisotropy

amplitude

jumps

walls,

are

co-operative

resulting

in

the

movements

reordering

of

of the

domains. It follows, from the present results and from Stierstadt and Boeckh’s work, that the large jumps

with increasing temperature. However, they showed that the anisotropy disappeared at 100°C. Two

become

reasons can be put forward to explain this apparent discrepancy: firstly, the different techniques used to

only small isolated jumps are occurring. This reasoning is supported by SchaueG who found that, for a

obtain the two sets of data, and secondly, the fact that the anisotropy constants K, and K, are known to be influenced by crystal purity and may therefore

nickel specimen, a temperature ~-50°C was necessary for the production of a domain configuration in the energetically lowest state. This means that the

vary in different materials. Regardless of this anomaly,

magnetocrystalline anisotropy gives rise to an energy barrier. Overcoming this leads to the formation of a

the most

significant

increasingly

anisotropy,

rare

with

decreasing

crystal

and in the absence of’ crystal anisotropy

ROBERTS

stable domain configuration blocks

of domains.

BARRAND:

AND

MAGNETOMECHANICAL

by the reordering of large

If the anisotropy

is too low the

DAMPING

1691

changes in magnetic ordering arising from the presence of solute atoms in the domains.

It might be expected

reordering processes are energetically u~avorable. The fall off in Am to zero at the Curie point reflects

that the latter effect would predomina~ at higher alloy concentrations. This is assumed here, since a

the instability of the domain structure in the temperature region close to the ferromagnetic transition.

linear relationship

It is of interest to see how the constant temperature

dependent

in with this picture.

damping

The damping

Fig. 6, and in K6ster’s(3) with

the

temperature

crystalline that

anisotropy

the

peak

work,

region

from the second peak would be more app~oable to the

peak at 150°C,

influence of anisotropy curve for pure nickel.

where

the

small.

is not

magnetoThe fact

dependent

on

frequency relaxation

movement. boundary

that the peak was probably

rotation

in the hysteresis

constant

in the temperature

where the magnetocrystalline

anisotropy

region

was rapidly

di~p~aring. in dynamic

modulus

ture, Fig. 6, shows a well defined temperature

with tempera-

minimum in the The sharp fall

region of zero anisotropy.

force resisting domain-boundary

Entwistle(Q)

solute ordering in adjacent subsequent

boundary

cause a 90”

will

vector.

Temporarily,

energy a minimum.

Because t*he energy is

thus higher, the boundary could experience an additional force resisting its motion which would reduce the magnetic

permeability.

all

parameters

magnetostrictive tion. The rise

temperatures

of Am, and must

domains to be completed,

the solute atoms in the volume swept by the boundary

domains

parallels the behaviour

a 90”

if

will not be in the preferred sites which would make the

individual

temperature

that,

for some time to allow

movement

of the magnetization

in modulus from the peak at the Curie point, together with the similar fall in damping, is associated with the strains produced by domain formain modulus from 160°C to room

suggested

has been stationary

ma~etic

The variation

as it has a similar shape to the

Magnetic ordering of solute has been shown to give rise to an additional

proves that it is not due to a reversible process. A likely explanation for its occurrence is that offered by C0chardt.d) He suggested due to a large increase

Fig. 10,

In fact, the non-linear curve estimated

of anisotropy.

at 2OO”C, coincides

is negligibly

temperature

amplitude

results for nickel fit

of Am with temperature,

would be difficult to explain in terms of the influence

other

The ordering

being

constant.

the degree of solute

At

higher

ordering

will be

greater and therefore the number of possible Barkhausen jumps will be correspondingly lower. On this

therefore be due to increasing magnetocrystalline anisotropy. The temperature dependence of the

assumption

Barkhausen

Curie point since the domains will be non-exis~nt

y,

fraction

shown

in Fig.

that of the dynamic

of maximum

1 I, compares favourably with modulus. This is hardly surpris-

ing since both magnetic modulus

are known

susceptibility,(4)

susceptibility

to be strongly

and dynamic dependent

internal strains, particularly magnetostrictive arising from domain formation.

on

strains

The rise in damping, in the axially applied magnetic field, at low temperatures, Fig. 6, has previously been observed by MisEk,@) who assigned the effect to microscopic eddy currents. An initial survey of the Am-temperature variation in a 20 wt.% copper-nickel alloy was undertaken to investigate

whether

influence of alloying anisotropy. complicated

this was closely additions

related

to the

on magnetocrystalline

As Figs. 7 and 8 show, the results are by the dissociation of the magnetic peak

into two peaks, the second peak being more apparent at higher tem~ratures. This phenomenon is possibly due to the presence of two different types of internal stress peaks opposing domain wall motion, viz: a magnetocrystalline anisotropy strecJs peak characteri&e of the matrix, and a stress peak modified by

in

will depend on the temperature,

a linear Am-temperature

be expected;

the ferromagnetic

transition.

Fischbach(lO)

has observed

peak in an Fe-17% dependent

plot might well

the line extrapolating

to zero at the

a magnetic

at

ordering

Al alloy which was frequency-

at low strain amplitudes.

In this case t’he

small cyclic applied stress would induce oscillation the domain

boundaries

energy dissipated temperature

held by solute ordering,

passing through

dependent

on the frequency.

in the case of the ferromagnetic a

frequency-dependent

expected operating

a maximum

in the damping-temperature shear strain is sufliciently

peak

the at a

Likewise,

copper-nickel

magnetic

of

alloys,

might

be

spectrum if t#he low.

Neasure-

ments so far have failed to locate such a peak, perhaps because

the shear strain

Alternatively,

was not

its occurrence

might

sufficiently depend

low.

on the

relative size factor difference between the constituent atoms in the alloy. magnetic

damping

Fischbach(*O) has shown that the peak in the Fe-Al

alloy is associ-

ated to some extent with the Zener relaxation which because

may of

not occur

in the copper--nickel

the

small

very

size factor

peak, system

difference.

ACTA

1692

Obviously,

in

independent

and amplitude-dependent

order

to

correlate

METALLURGICA,

the

amplitude-

magnetic peaks

an alloy must be used where the size factor difference is high enough to ensure that both can be measured. The

evidence

presented

here

frequency

internal friction

providing

accurate measurement

dependent

damping

suggests

techniques

that

low

are capable

of

of strain amplitude-

in magnetic

materials.

It

is

VOL.

15, 1967

5. The “double 20 wt.%

peak” phenomenon

alloy is possibly

different types of internal stress peak, viz : a magnetocrystalline

anisotropy

of solute in the domains. 6. The linear Am-temperature more prominent

anomalous

properties.

is assigned

SUMMARY

anisotropy. 7. It is concluded

frequency

internal

friction

have revealed amplitude-dependent pure nickel and a 20 wt.% 2. The

peak

relationship

magnetic

magnetic peaks in

copper-nickel damping

alloy.

of

the

magnetization,

Barkhausen

J,/Js,

reported

fraction by

total

Stierstadt

and

are explained

anisotropy.

domain configuration; presence

4. Constant measurements

the latter process is dependent

of magnetocrystalline

anisotropy.

amplitude, temperature-dependent on pure nickel verified KGster’s(3)

The “magnetic

damping”

peak observed

here,

and the anomalous behaviour of the dynamic modulustemperature curve are explained in terms of the absence

of magnetocrystalline

and the disappearance ferromagnetic

transition,

anisotropy

of domain 360°C.

of

solute

to

the

influence

the applied

ordering

in

relationship

of magnetocrystalline

that Am offers a reliable measure

stress system,

jumps occurring

under

and that this is strongly

by the presence of crystalline

anisotropy.

ACKNOWLEDGMENTS

The

authors

provision

thank

Professor

C. R.

of laboratory

facilities.

One of the authors

(J. T. A. R.) acknowledges for a maintenance grant.

Trottle

his indebtedness

for

to S.R.C.

in a qualitative

manner by stray field coupling of individual domain wall movements plus larger re-arrangements of the

work.

dependence

of the

in terms of the

The non-linear Am-temperature

of the number of Barkhausen influenced

of

on the magnetocrystalline

3. The observations

of

Am-temperature

Both these effects reflect the influence of

temperature

on the

domains.

for pure nickel parallels the temperature

dependence Boeckh.c4)

measurements

characteristic

relationship

peak is explained

temperature

1. Low

stress peak,

the matrix, and a stress peak due to magnetic ordering

intended to extend the investigation to other materials, particularly nickel base alloys which exhibit magnetic

observed in the

due to the presence of two

at 200°C

structure

at the

REFERENCES 1. A. COCHARDT, Magnetic

Properties of Metals and Alloys, p. 251. ASM (1959). 2. G. SUMNER and K. M. ENTWISTLE, J. Iron Steel In&. 192. 238 (1959). 3. W. K~STER, 2. Metallk. 55, 246 (1943). 4. K. STIERSTADT and W. BOECKH, 2. Phys. 186, 154 (1965). BRAILSFORD, Magnetic Materl:aZs, p. 61. Methuen (1960). 6. G. HAACKE and J. JAUMANN, 2. angew. Phys. 12, 289

5. F.

(1960). 7. A. SCHAUER, ibid. K. MI&K, Czech J. :: K. M. ENTWISTLE, 10. D. B. FISCHBACH,

16, 90 (1963). Phys. 7,247 (1957); ibid. 8, 128 (1958). Metall. Rev. 7, 175 (1962). Acta Met. 10,319 (1962).