MAGNETOSTRICTION WITH THE MICHELSON INTERFEROMETER

PRINCIPLE
With the aid of two mirrors in a Michelson arrangement, light is brought to interference. Due to the magnetostrictive effect, one of the mirrors is shifted by variation in the magnetic field applied to a sample, and the change in the interference pattern is observed.

EQUIPMENT
Base plate with rubber feet HeNe Laser, 1 mW Adjusting support, 35 x 35 mm Surface mirror, 30 x 30 mm Magnetic base Plate holder Beam splitter, 50:50 Lens with mount, f = +20 mm Lens holder for base plate Screen, white, 150 x 150 mm Coil, N = 1200, 4 , Faraday Modulator Metal rods for magnetostriction Power supply, universal Digital multimeter Battery, 9 V, 6 F22 Connecting cord, l = 500 mm, blue

Fig. 1

Experimental Set-up

Page 1 of 13

TASKS
Construction components. Testing various ferromagnetic materials (iron and nickel) as well as a non-ferromagnetic material (copper) with regard to their magnetostrictive properties. of a Michelson interferometer using separate optical

Fig. 1a

Experimental set-up of the Michelson interferometer for the measurement of magnetostriction of different ferromagnetic materials

SET-UP AND PROCEDURE

In the following, the pairs of numbers in brackets refer to the co-ordinates on the optical base plate in accordance with Fig. 1. These co-ordinates are only intended to be a rough guideline for initial adjustment.

Perform the experimental set-up according to Fig. 1 and 1a. The recommended set-up height (beam path height) is 130 mm.

Page 2 of 13

The lens L [1,7] must not be in position when making the initial adjustments. When adjusting the beam path with the adjustable mirrors M 1 [1,8] and M 2 [1,4], the beam is set along the 4th y co-ordinate of the base plate.

Place mirror M 3 onto the appropriate end of a sample (nickel or iron rod) (initially without the beam splitter BS [6,4])and screw it into place.

Now, insert the sample into the coil in such a manner that approximately the same length extends beyond the coil on both ends so that a uniform magnetisation can be assumed for the measurement. Fix the sample in position with the laterally attached knurled screw.

Next, insert the coil Cs shaft into a magnetic base and place it at position [11,4] such that the mirrors plane is perpendicular to the propagation direction of the lasers beam (see Fig. 1a).

Adjust the beam in a manner such that the beam reflected by mirror M 3 once again coincides with its point of origin on mirror M 2 . This can be achieved by coarse shifting of the complete unit of coil with magnetic base or by turning the sample rod with mirror M 3 in the coil and by meticulously aligning mirror M 2 [1,4] with the aid of its fine adjustment mechanism.

Next, position the beam splitter BS [6, 4] in such a manner that one partial beam still reaches mirror M 3 without hindrance and the other partial beam strikes mirror M 4 [6, 1]. Die metallized side of BS is facing mirror M 4 .

Two luminous spots now appear on the screen SC [6, 6]. Make them coincide by adjusting the mirror M 4 until a slight flickering of the luminous spot can be seen.

After positioning lens L [1,7], an illuminated area with interference patterns appears on the screen. To obtain concentric circles, meticulously readjust mirror M 4 using the adjustment screws.

Subsequent to the connection of the coil to the power supply (connect the multimeter in series between the coil and the power supply to measure the current, measuring range 10 AC!), set the DC-voltage to maximum and DC-current to minimum value. Then slowly readjust the current. For the measurements the resulting currents lie between 0.5 and, maximally, 5 A. Count the changes from maximum to maximum (or minimum to minimum) in the

Page 3 of 13

interference pattern. In addition, pay attention to the direction in which the circular interference fringes move (sources or sinks!). Repeat this procedure using different samples and different current strengths I between 0.5 and 5.0 A (I > 3 A only for a short time!). Notes: The materials require a certain amount of premagnetisation; therefore, the current should be run up and down several times for each individual determination before performing the intensity change measurement.

Premagnetization improve the result at a special current value. E.g. set the value to 1 A and do the premagnetization as follows: change the current to 0 to 2 to 0.2 to 1.8 to 0.4 to 1.6 to 0.6 to 1.4 to 0.8 to 1.2 and end at 1.0 A. The experiment still works without premagnetization - just with a bigger error. The blank trial with a copper rod as sample should serve to demonstrate that the longitudinal deformation effect is due to magnetostriction and not to other causes.

THEORY AND EVALUATION

If two waves having the same frequency but different amplitudes and different phases are coincident at one location, they superimpose to

= a 1 . sin(t 1 ) + a 2 . sin(t 2 ).

The resulting wave can be described by the following:

= A. sin(t )

with the amplitude

A2

2 = a1 + a2 2 + 2a 1 a 2 . cos

(1)

and the phase difference

= 1 2.

Page 4 of 13

In a Michelson interferometer, the light beam is split by a half-silvered glass plate into two partial beams (amplitude splitting), reflected by two mirrors, and again brought to interference behind the glass plate (Fig. 2). Since only large luminous spots can exhibit circular interference fringes, the light beam is expanded between the laser and the glass plate by a lens L. If one replaces the real mirror M 4 with its virtual image M 4 ,which is formed by reflection by the glass plate, a point P of the real light source appears as the points P and P of the virtual light sources L 1 and L 2 .

As a consequence of the different light paths traversed, and using the designations in Fig. 3, the phase difference is given by:

2 .2.d. cos

(2)

is the wavelength of the laser light used.

According to (1), the intensity distribution for a 1 = a 2 = a is:

I A2

= 4.a 2 cos2

(3)

Fig. 2

Michelson arrangement for Interference. S represents the light source; SC the detector (or the position of the screen)

Page 5 of 13

Maxima thus occur when is equal to a multiple of 2, hence with (2)

2.d. cos = m. ;m = 1, 2,

(4)

i.e. there are circular fringes for selected, fixed values of m, and d, since remains constant (see Fig.3).

If one alters the position of the movable mirror M 3 (cf. Fig.1) such that d, e.g., decreases, according to (4), the circular fringe diameter would also diminish since m is indeed defined for this ring. Thus, a ring disappears each time d is reduced by /2. For d = 0 the circular fringe pattern disappears. If the surfaces of mirrors M 4 and M 3 are not parallel in the sense of Fig.3, one obtains curved fringes, which gradually change into straight fringes at d = 0.

Fig. 3

Formation of circular interference fringes

Page 6 of 13

On magnetostriction:
Ferromagnetic substances undergo so-called magnetic distortions, i.e. they exhibit a lengthening or shortening parallel to the direction of magnetisation. Such changes are termed positive or negative magnetostriction. The distortions are on the order of l/l ~10
-8 -4

to 10

in size. As is the case

in crystal anisotropy, the magnetostriction is also ascribable to the spin-orbit mutual potential energy, as this is a function of the direction of magnetisation and the interatomic distances.

Due to magnetostriction, which corresponds to a spontaneous distortion of the lattice, a ferromagnet can reduce its totalanisotropic and elastic energy.

Inversely,

in

cases

of

elastic

tension

the

direction

of

spontaneous

magnetisation is influenced. According to the principle of the least constraint, this means the following:

In cases of positive magnetostriction (in the case of iron (Fe), under tensile stress the magnetisation is oriented parallel to the stress; in cases of compressive stress the magnetisation orients itself perpendicular to the pressure axis. In nickel (Ni) the situation is exactly reversed.

A true metal (ferromagnetic material) consists of small uniform microcrystals in dense packing, in all whose crystallographic directions. axes are however irregularly are

distributed

spatial

The

individual

crystallites

additionally subdivided in Weiss molecular magnetic fields consisting of many molecules which form the elementary dipoles (*).

If the material has not been magnetised, all six (in nickel all eight) of the magnetic moment directions possible within a crystallite are present with equal frequency and consequently neutralise one another externally as a result of this irregular distribution. The magnetisation of the Weiss molecular magnetic fields is a function of temperature and occurs spontaneously below the Curie temperature.

However, as a consequence of the application of an external magnetic field this non-uniform distribution of the directions of magnetisation can be altered by the transition of a large number of Weiss molecular magnetic fields in the

Page 7 of 13

preferred light magnetisation directions, which have the smallest angle to the direction of the external magnetic field.

*On magnetic crystal anisotropy:

In monocrystals one observes a marked anisotropy of the magnetisation curve. This is due to the so-called magnetic crystal energy. The source of this anisotropic energy in the transition metals (Fe, Ni and Co) is in their spin-orbit coupling energy, which is based on the relativistic interaction between spin and orbital movement.

In a rotation (directional alteration) of the spin, which is coupled by the mutual exchange energy, the orbital moments experience a torsional moment such that they also experience rotation. In an anisotropic electron distribution (d electrons) this effects a change in the overlapping of the electron clouds of adjacent atoms and hence an alteration of the total crystal energy.

One thus differentiates between the longitudinal magnetostriction, a length change parallel to the field direction and a transverse magnetostriction of the length alteration perpendicular to the external field direction. The relative length change = l/l generally increases with increasing magnetisation and reaches a saturation value s at M = M s (M s : saturation magnetisation). The relative volume change V/V(i.e., volume magnetostriction) is usually considerably smaller, since longitudinal and transverse magnetostriction nearly always have opposite signs and compensate each other to a large extent.

In this experiment only the longitudinal magnetostriction is considered (see Fig. 4). One must take into consideration that the magnetostriction is a function of temperature and that pre-magnetisation is necessary. Additionally, the magnetostriction in alloys is also dependent on the composition of the metals and the appropriate pre-treatment (see Fig. 5).

Page 8 of 13

Fig.4

Magnetostriction of different ferromagnetic materials with their relative change in length l/l plotted against applied magnetic field strength H m

Fig. 5

Magnetostriction of different ferromagnetic alloys with their relative change in length l/l plotted against applied field strength H m

Page 9 of 13

Thermodynamic description of magnetostriction:

Magnetostriction can be described quantitatively and thermodynamically using:

S: elastic tension s: elastic deformation (i.e.: l/l) B: magnetic induction H: Magnetic field strength : magnetic permeability with

H B

E: Elasticity module with

E =

S s

As a result of thermodynamic relationships, it can be shown that the direct and reciprocal magnetostriction effects are mutually linked via

S 1 H = . B 4 s

(5)

For a free rod (unloaded and not clamped in position), the following is true:

s = .

B E

(6)

with the substance-specific quantity

S . B

In other words, the relative longitudinal change is given by

s = ..

H E

(7)

In this context, cannot be a constant as otherwise a linear increase in the relative length with the magnetic field strength would result. This is however not the case, since a saturation value is reached as of a specific field strength.

Page 10 of 13

On the evaluation of the measuring results:

The magnetic field strength of a cylindrical coil is given by:

Hm

N.I
2 4.r 2 + l s

(8)

where Hm: magnetic field strength at the centre of the coil in Am r: Radius of a winding (here: 0.024 m)
-1

l s : Length of the coil (here: 0.06 m) N: Number of windings (here: 1200)

On condition that the field is homogenous, the field strength is by the following for l >> r:

H =

N.I l

(9)

For this measurement we assume, as a first approximation, that the magnetic field strength H m acts on the entire length of the rod (l = 0.15 m). The alteration in length l is obtained from the number of circular fringe changes n; in the process the separation per circular fringe change alters by /2 ( = 632 nm): l = n./2

(10)

Tabulate the results of the measurements on nickel and iron as indicated in Tables 1 and 2.

Table 1 I/A 0.83 1.27 1.6 1.87 H/A/m (see (9)) Ringchanges/n l/m (see (10)) l /l With l = 0.15 m

Page 11 of 13

Table 2 I/A 0.53 0.71 0.94 1.33 2.04 3.28 H/A/m (see (9)) Ringchanges/n l /m (see (10)) l /l With l = 0.15 m

In the measurements, the direction of magnetostriction also becomes apparent: In iron the radii of the interference rings increased with increasing magnetic field strength (sources!); thus, the rod must have become larger (see Fig. 6).

Fig. 6

Measuring results of the magnetostriction of iron (steel) with the relative change in length l/l plotted against applied field strength H

Page 12 of 13

In nickel the rod became shorter (sink of circular interference fringes); therefore, a negative magnetostriction existed in this case (see Fig. 7).

Fig. 7

Measuring results of the magnetostriction of nickel with the relative change in length l/l plotted against applied field strength H

SC Ng July 2005

Page 13 of 13