Measurement of Coefficient of Thermal Expansion by Fizeau’s Interferometer

Sajag Kumar (2011143)1, ∗

1
School of Physical Sciences, National Institute of Science Education and Research (NISER), Bhubaneswar 752050, India
(Dated: September 8, 2021)
We measure the coefficient of linear thermal expansion of a copper rod by Fizeau’s interferometer.
We compute the wedge angle of the interferometer, using the principle of interference of light, at
different temperatures to obtain the temperature gradient of wedge angle. And from this gradient
we obtain α = (50.506 ± 1.703) × 10−6 K −1 , the literature value [2] of α is 16.6 × 10−6 K −1 , which
is more than 19σα away from the obtained value.

I. INTRODUCTION B. The Principle of Interference [3]

The shape of an object changes upon heating. Tem-

perature of an object increases upon heating. It is di- The optical instrument that we are going to use to
rectly proportional to the average molecular kinetic en- measure α is called Fizeau’s interferometer. Interferom-
ergy of the object. An increase in temperature results eters are optical instruments that exploit the principle of
in an increase in the average molecular kinetic energy interference of light to extract information about some
of the object. As the energy of molecules increase they system.
start moving and vibrating faster and due to this the in-
termolecular forces between them weakens which further The principle of interference of light says that when
results in increased intermolecular distances. Hence, an two monochromatic coherent (constant phase difference
object undergoes expansion upon heating, and this ex- δ) light waves of intensities I1 and I2 are incident at a
pansion of an object due to increase in temperature is point their resultant intensity (I) would be:
called thermal expansion. There are some objects which
undergo contraction upon heating, their mechanism is
beyond the present discussion. p
I = I1 + I2 + 2 I1 I2 cos δ . (2)

A. Coefficient of Thermal Expansion

Evidently there will be a maxima of intensity,
The coefficient of thermal expansion (α) tells us how
much the size of an object changes for a given change in
temperature. There are many types of coefficients e.g.
p
Imax = I1 + I2 + 2 I1 I2 (3)
linear, area, volume etc. We will discuss the coefficient
of linear thermal expansion. It is defined, for an object
in the form of a long rod at a given temperature, as the
fractional change in the length of the rod per unit change when,
in temperature. Mathematically,

1 ∆L
α= (1) |δ| = 2nπ, n ∈ 0, 1, 2, ... (4)
L ∆T

where, L is the original length of the object and ∆L is

the change in length when temperature changed by ∆T .
and minima of intensity,
To measure α of an object we need to measure L, ∆L
and ∆T . While measuring ∆T and L is not very dif-
ficult, measuring ∆L poses a lot of problems because p
this quantity is very small for an appropriate change in Imin = I1 + I2 − 2 I1 I2 (5)
temperature. To measure such small lengths we employ
optical methods. Hence, we will use optical instruments
to measure α.
when,

∗ [email protected] |δ| = (2n + 1)π, n ∈ 0, 1, 2, ... . (6)

 2

II. EXPERIMENTAL SETUP called fringes of equal thickness because the air film has
variable thickness and each fringe corresponds to partic-
ular thickness.

FIG. 1. Schematic diagram of the experimental setup. FIG. 2. Schematic diagram of the air-filled trapped between
the glass plate and the glass bean.
The experiment setup for this experiment is called the
Fizeau’s interferometer. Fizeau’s interferometers have
The term λ/2 takes into account the abrupt phase
two reflecting surfaces are placed facing each other. The
change of π radians for the wave reflected from the top of
setup consists of an aluminium stand A about 15 cm in
P 2. From (4) and (6) we know the conditions for max-
height with two projections A1 and A2. A1 holds two
ima and minima of the interference pattern. The phase
glass plates (microscopic slides) P 1 and P 2 where P 2 is
difference δ of the waves depend on the path difference
slightly shorter than P 1. A screw SC passes up through
φ. The conditions for maxima and minima in terms of φ
A2. The copper rod C of length LRT is placed such that
translate as the following. Maxima is obtained when,
it touches P 1. A3 is a metal block screwed on to A2. A3
supports a heater H which is a small cement tube over
which a heating element is wound. A thermocouple (not φ = mλ (8)
shown) is inserted close to the sample. L is the source of
light. The inclined glass plate P 3 reflects the light from and minima is obtained when,
L1 to the interferometer and transmits the light from the
interferometer to the microscope. M is the microscope λ
φ = (2m + 1) (9)
placed over the interferometer to catch the transmitted 2
light from P 3.
where,

A. Measuring α Using Fizeau’s Interferometer |m| = 0, 1, 2, ... . (10)

In this setup when the copper rod is heated, it expands, From (7) the dark fringes would satisfy:
and pushes P 1 upwards. A wedge shaped thin air film
is formed between the surfaces of the two inclined glass 2t = mλ . (11)
plates P 1 and P 2. When the film is illuminated by light
of wavelength λ, coming from L1, at normal incidence,
If the thickness of air film for two consecutive dark fringes
interference occurs between the light rays reflected from
is t1 and t2 , then t2 − t1 = λ/2. The wedge angle (θ)
P 1 and P 2 and an interference pattern is observed when
would then be:
viewed through M . The path difference between the di-
rect and the reflected ray of light is given by: λ
λ tan θ = (12)
φ = 2t + (7) 2β
2
where t is the thickness of the air-film enclosed between where β is the distance between two consecutive minima
the glass plates at the point of interest. These fringes are (also called fringe width).
 3

P 1 properly, fringe pattern was observed through M . A

point of reference was marked. The cursor was made to
coincide with a dark fringe near the point of reference
and the microscope reading was noted. This is the ini-
tial reading. The M cross-wire was moved towards one
side and the readings were noted down for every five dark
fringe up to twenty dark fringes. The power supply was
switched on and C was allowed to heat such that its tem-
perature raised by ten degrees. After C was heated the
procedure of taking readings was repeated. C was heated
to five different temperatures during the experiment. We
marked the point of contact of C and P 1 using a marker
and measure the distance between A and the point of
contact using a scale.
Once we were done taking the readings we computed
θ for different values of T and plotted a graph between θ
and T . The slope of this graph would give us the value
∆θ/∆T necessary to compute α (16).
FIG. 3. Schematic diagram of the change in orientation of
the two reflecting surfaces.
A. Precautions
When C expands on heating it pushes P 1 upwards
which in turns changes the fringe width (12). We can esti- Throughout the course of the experiment necessary
mate the change in length (∆LRT ) of C from this change precautions must be taken to reduce errors in the value
in fringe width. Let θ1 and θ2 be wedge angle at temper- of measured quantities. We should avoid parallax error
ature T1 and T2 = T1 + ∆T1 , respectively. And β1 and while using the vernier caliper and while measuring l.
β2 be fringe width at temperature T1 and T2 = T1 +∆T1 , Backlash error should be avoided while using the vernier
respectively. From (12) we have, caliper as well as the microscope. We must ensure that C
touches P 1 gently at the beginning. When we are heat-
λ ing C we should note down the temperature only after
θ1 = tan−1 , (13) it has stabilised on the digital thermometer. The volt-
2β1
age across the heating filament should be increased very
λ
θ2 = tan−1 . (14) slowly.
2β2

And ∆LRT can be calculated as follows: IV. OBSERVATIONS

∆LRT = l∆θ = l(θ2 − θ1 ) (15)

1. l = 3.71 cm.
where l is the length of P 1. 2. Least count of vernier calliper = 0.01 cm.
From (1), we finally have,
3. Least count of travelling microscope = 0.01 mm.
 
l ∆θ
α= . (16) 4. Wavelength of light from sodium lamp = 583.9 nm.
LRT ∆T

Length of Copper Rod (LRT )

S. No. Mean LRT (cm)
III. METHODOLOGY M.S.R. (cm) V.S.R. Total (cm)
1 2 8 2.08
2 2 7 2.07 2.07
First of all, we measured the room temperature. The 3 2 8 2.08
length of C was measured using vernier caliper. Three
readings were taken. TABLE I. Observation table for measurement of LRT .
C was placed in the interferometer and sodium lamp
was switched ON. P 3 was adjusted so that it reflects
light towards P 2. Once it was ensured that C touches

TABLE II. Obsevation table for measure of θ for different T .

 4

Temperature (T) No. of Microscope Readings Width of Five Fringe Average Fringe Wedge Angle
S. No.
(in degree Celcius) Fringes M.S.R. (mm) V.S.R. Total (mm) Fringes (mm) Width (mm) Width (mm) (in radians)
0 5 1 5.01 0 0
5 6 20 6.2 1.19 0.238
1 25 10 7 39 7.39 1.19 0.238 0.2375 0.001240630942
15 8.5 5 8.55 1.16 0.232
20 9.5 26 9.76 1.21 0.242
0 5 4 5.04 0 0
5 6 6 6.06 1.02 0.204
2 35 10 7 7 7.07 1.01 0.202 0.2025 0.001455060702
15 8 5 8.05 0.98 0.196
20 9 9 9.09 1.04 0.208
0 5 6 5.06 0 0
5 5.5 42 5.92 0.86 0.172
3 46 10 6.5 27 6.77 0.85 0.17 0.169 0.001743489358
15 7.5 7 7.57 0.8 0.16
20 8 44 8.44 0.87 0.174
0 5 10 5.1 0 0
5 5.5 35 5.85 0.75 0.15
4 55 10 6.5 10 6.6 0.75 0.15 0.146 0.002018147945
15 7.5 27 7.77 1.17 0.234
20 8 2 8.02 0.25 0.05
0 5 2 5.02 0 0
5 5.5 14 5.64 0.62 0.124
5 65 10 6 26 6.26 0.62 0.124 0.127 0.002320074577
15 6.5 45 6.95 0.69 0.138
20 7.5 6 7.56 0.61 0.122
0 4.5 50 5 0 0
5 5.5 7 5.57 0.57 0.114
6 74 10 6 15 6.15 0.58 0.116 0.113 0.002607516214
15 6.5 16 6.66 0.51 0.102
20 7 26 7.26 0.6 0.12

V. DATA ANALYSIS A. Computing α

From (16),
In figure 4, slope of red line,  
3.71
α= (2.8180 × 10−5 ) K −1
2.07
∆θ = 5.0506 × 10−5 K −1 .
= s = (2.8180 ± 0.0948) × 10−5 K −1 . (17)
∆T

B. Errors

We have from [1],

s 2  2  2
δα δLRT δl δs
Wedge Angle (θ in radians)

0.0026 Recorded Data = + + . (18)

Fitted Line α LRT l s
0.0024

0.0022 Here,
0.002

0.0018
δLRT = 0.005 cm , (19)
0.0016 δl = 0 , (20)
0.0014 δs = 0.0948 × 10−5 K −1 . (21)
0.0012

30 40 50 60 70
Temperature (T in degree Celcius)
Substituting the values of α, LRT , l, s, δLRT , δl and δs
in (18), we get:
FIG. 4. Plot of θ vs T . δα = 0.1703 × 10−5 K −1 . (22)
 5

VI. RESULTS AND CONCLUSION pattern consists of hyperbolic lines with equal spacing.

CODE AND DATA

−6 −1
The value of α is (50.506 ± 1.703) × 10 K . The
value of α obtained in the experiment is more than the The graph has been generated with pyROOT and
value of α present in the literature [2]. The literature line fitting has been done with ROOT’s inbuilt linear
value (16.6×10−6 K −1 ) is more than 19σα away from the fit function. The slope of the graph comes from χ2
obtained value. Here, σα is the standard deviation in the fit. All the code used to generate this document is
value of α. This large deviation from the literature value hosted at: https://github.com/SlowGraviton/P241/
should be attributed to systematic errors. The fringe tree/main/Lab%201.

[1] J. R. Taylor, An Introduction to Error Analysis, 2nd ed. Research Bhubaneswar (2021).
(University Science Books, 1997). [3] M. Born and E. Wolf, Principles of Optics, 7th ed. (Cam-
[2] Coefficient of Thermal Expansion of Metals by Fizeau’s In- bridge University Press, 2019).
terferometer, National Institute of Science Education and