You are on page 1of 6

Speed of Light Laboratory

Nicholas Prezioso
University of Rhode Island, 45 Upper College Rd, Kingston RI,02881
nprezioso@y!"ri!ed"
September 30
th
, 2014
Abstract: In this experiment, the speed of light is measured using the methods developed by Foucault in the 1
th
century!
"he speed of light is recogni#ed as an important value due to its constant nature as described by Special $elativity! "hus,
accuracy and precision are of the utmost priority %hen it comes to measuring this value! "he measurement of c is thus
not the sole focus of this experiment! "%o methods are used to calculate the value&curve fitting of data and direct
computation&and their 'ualities in terms of precision and accuracy are discussed!
Key#ords$ speed of light, opti%s
INTRODUCTION
The velocity of light has been a fascination of
scientists for some time. For awhile, it was believed
that light travelled instantaneosly! that is to say,
it had infinite s"eed. #ventally this was shown to
not be tre. Frthermore, the actal vale for the
s"eed of light has been determined with greater
and greater "recision over time. $ trly remar%able
discovery was made by $lbert #instein in his
&"ecial Theory of Relativity that noted that the
s"eed of light is consant regardless of the reference
frame in which it is measred. 'hether the s"eed
of light is measred from a lantern over a hill, or
from a distant star hrtling throgh the niverse,
when yo measre the s"eed of light yo derive the
same vale.
&everal attem"ts have been made throghot
history to find this vale. The first attem"ts at
determining the s"eed of light were done by
(alileo. )is methods were relatively "rimitive* )e
and an assistant wold stand on hills some
distance from each other holding lanterns. One
wold ncover their lantern and the other wold,
"on seeing the light, sbse+ently ncover their
own. The goal was to detect some time lag between
the lanterns being ncovered. Unfortnately, this
method was not ade+ate in "rodcing a
measreable reslt.
Fifty years later Danish astronomer Ole
Roemer observed that ,"iter-s moon, Io, wold
occasionally be ot of "lace in its "redicted orbit.
.oreover, it was noticed that Io wold be behind in
its orbit when the #arth was relatively far from
,"iter, and ahead in its orbit when #arth was
closer. Roemer "osited that this time delay was de
to the light having to travel an e/tra distance* The
diameter of #arth-s orbit arond the sn. Romer
estimated, throgh carefl observation, that this
time delay was a""ro/imately 00 mintes. 1y
sim"ly dividing the diameter of #arth-s orbit by
this time delay, the s"eed of light cold be
determined.
Fig. $* Diagram of Romer-s method. 2oint $ and 1 re"resent
the #arth and ,"iter, res"ectively, when they are closest.
2oint C and D re"resent ,"iter and #arth, res"ectively, when
they are farthest. Romer "ostlated that the time delay in
observing Io-s "redicted ecli"se is de to the light from ,"iter
having to travel the e/tra distance from $ to D 3a""ro/imately4.
$while later the mathematician Christian
)ygens actally did the calclations and
determined the s"eed of light to be 565,777 miles
"er second. The vale is %nown today to be 589,777
miles "er second. Romer-s observation, althogh
ingenis, was marred by a lac% of "recise
%nowledge of both the time delay and the diameter
of the #arth-s orbit. Regardless, his wor% is the first
observed evidence that the s"eed of light was not
infinite. Over the ne/t several hndred years, a
variety of methods were develo"ed to "rodce
accrate measrements of the s"eed of light.
One sch method, and the one sed in this
e/"eriment, was develo"ed by Focalt in 5890. $
beam of light is sent throgh a series of lenses and
then reflected off of a rotating mirror. The beam
then stri%es a s"herical meter fi/ed some distance
away from the rotating mirror, where it is then
reflected bac% towards a measring a""arats. If
the mirror is rotating sfficiently fast, there is a
noticeable dis"lacement between the beam-s
original "ath and that of its retrn tri". Using this
dis"lacement, the s"eed of the mirror, and the
distance between the rotating mirror and the fi/ed
s"herical mirror, the s"eed of light cold be
determined.
T)#OR:
The theory behind the e/"eriment de"ends "on a
finite s"eed of light! shold light "ro"agate
instantaneosly, there wold be no dis"lacement
3as mentioned above4 between the beam-s original
"ath and the reflected "ath. Figre 5 illstrates the
"ath of a single ;"lse; of light generated by the
laser a""arats. First, it travels throgh the first
lens 3<54 where it is focsed to a "oint s and then
contines throgh the second lens 3<04 and is then
reflected off of the rotating mirror 3.R4. The "lse
then stri%es the fi/ed mirror 3.F4 and retrns
down its original "ath, where it is refocsed by <0
to the "oint s- after "assing throgh a beam
s"litter.
Fig. 5* $bstract =iew of the Focalt .ethod.
It shold be noted that Figre 5 shows an
abstract view of the method* The actal set " sed
is discssed more thoroghly in the #/"erimental
&et" section. Now, let s consider Figre 0. Figre
0a shows the "lse of light stri%ing .R while it is
at an angle >. Then, since the "lse stri%es .R at >
from thee normal, and the angle of incidence e+als
the angle of reflection, the angle between the
incident and reflected rays is 0> . 'e label the "oint
that the "lse stri%es .F at as &.
Fig. 0 a, b* Incident and reflected light rays at varios angles >.
Now consider 0b where > is dis"laced by an
angle ?>. The angle between rays then becomes
3>@?>4, and, labelling the new "oint on .F as &5,
we develo" a relationshi" between & and &5*
& A &5 B DC03>@?>4 A 0>D B 0D3?>4 #E5
where D is the distance between .F and .R. <et s
now consider Figre 6, which sim"lifies the geometry of
the method by ta%ing a closer loo% at the virtal image
formed by the beam "ath.
Fig. 6* =irtal image of Focalt setF". D is as described
above! 1 is the distance between <0 and .R! and $ is the
distance between <5 and <0 mins the focal length of <5.
The image that is observed is at ?s-, and it is
the same dis"lacement that wold be observed at ?s.
Using the law of similar triangles, we can obtain a
relationshi" between ?s- and ?&*
?s- G $ B ?& G 3D @ 14
BH ?s- B C$ G 3D @ 14D ?& #E0
&bstitting or derived vale for ?& from #E5 into
#E0 we obtain*
?s- B C$ G 3D @ 14DIC0D3?>4D #E6
Now, we can determine the de"endence of ?> by noting
that the distance the "lse of light travels is 0D.
Dividing by the s"eed of light gives the time that the
"lse ta%es to reflect off of and retrn to .R.
.lti"lying this time by the anglar fre+ency J yields*
?> B 0DJ G c #EK
2lgging #EK into #E6 yields*
?s- B CK3DL04$JD G Cc 3D @ 14D
BH c B CK3DL04$JD G C3?s-43D @ 14D #EM.
It is im"ortant at this Nnctre to note that the
s"eed of light, c, does not de"end on the "ro"erties of the
fi/ed mirror. That is to say, there are no de"endencies
"on the focal "oint or the center of crvatre of the .F.
&o, "rovided the "lses of light are reflected bac% along
their incident "aths, the s"herical mirror cold be
re"laced by a lens and flat mirror. Figre K illstrates
the geometry in this case.
Fig. K* $ diagram of the Focalt method sing a lens and flat
mirror in lie of a s"herical mirror. The sorce of light s
bonces rays off of the rotating mirror .R, where they are
then focsed by lens < onto flat mirror .F. The rays are then
reflected bac% towards the lens symmetrically, where they are
reFfocsed bac% towards .R and, ths, the sorce "oint.
In the case of a flat mirror, it is of tmost im"ortant that
the mirror be "laced at the focal length of the lens. For
this to wor%, we se the lens e+ation. <et D be the
distance from the lens to the fi/ed mirror, and let D- be
the distance from the lens to the sorceGobservation
"oint. Then, %nowing the focal length of the lens, we can
obtain or critical distances by the formla*
5 G f B 35 G D4 @ 35 G D-4
#O2#RI.#NT$< &#TU2
First, the O&FP576 O"tics 1ench is set " along with the
&#FP69Q <aser with O&FP5Q0 $lignment 1ench. The
beam is then aligned sch that it stri%es the center of
the rotating mirror sing two alignment Nigs. 'ith the
beam aligned, "lace the K8 mm focal length lens 3<54 on
the bench at the P6 cm mar% and the 0M0 mm focal
length lens 3<04 at the 90.0 cm mar%. The measring
microsco"e with beam s"litter is then "laced sch that
its left edge is aligned with the 80.7 cm mar%. Ne/t, a
series of relaying mirrors are set " to increase the
effective vale of D. Finally, the fi/ed mirror is set "
sch that it reflects bac% on the final relaying mirror.
$n intermediate ste" at this "oint is
determining D. This is done by measring the distance
between each relay mirror sing a ta"e measre. The
vale for D, then, is the sm of the distances between
the relay mirrors. The distances $ and 1 can be
determined directly by reading the "ositions on the
o"tics bench, which has centimeter mar%ers along its
side.
'ith the lenses, microsco"e, and mirrors in
"osition, the a""arats can then be calibrated.
2olariRers are "t in "lace and adNsted so that the
beam can be viewed safely. The beam is then traced
from mirror to mirror sch that it stri%es each one in its
center. &mall adNsting %nobs on the bac% of each mirror
are sed to adNst the vertical and horiRontal reflection
of the beam. This is done for each relaying mirror ntil
the beam is reflected onto the s"herical fi/ed mirror. .F
also contains these %nobs, and is adNsted as well so that
it reflects the beam bac% onto the final relaying mirror
and therefore bac% onto the rotating mirror.
'ith the beam reflecting bac% into the
microsco"e, the image is fine tned by adNsting <0. The
goal being to obtain a centered, fine "oint on the
microsco"e. The beam s"litter-s angle can also be
adNsted to center the viewed image. Frthermore, the
microsco"e itself can slide " and down to obtain a more
resolte image.
'hen the beam is aligned and "ro"erly resolte
for viewing, the trials can begin. The rotating mirror is
allowed to warm " for several mintes. $ trial is
"erformed by first setting the rotating mirror to a
desired s"eed and direction 3cloc%wise or conterF
cloc%wise4. 'hen an image is obtained, it is then
centered sing the micrometer %nob on the microsco"e.
$ reading is obtained and the direction is then reversed
3%ee"ing the same s"eed4. To clarify, for a given
rotational s"eed, two trials are collected* $ cloc%wise
trial and a conterFcloc%wise trial. This is re"eated for
increasingly faster rotational s"eeds " to the ma/imm
s"eed of 5M77 revsGsec. $fter all trials are com"leted, the
data is "t into a table, with a corres"onding direct
com"tation of c 3sing #EM4. The data are also sch
that the slo"e of a linear a""ro/imation yields the vale
of c as well 3discssed below4.
R#&U<T& S #RROR $N$<:&I&
Uncertainties are assigned to all direct measrements
done 3e.g. $, 1, D, etc.4 "ro"ortional to the smallest
significant digit of each measrement. This acconts for
random error in re"eated measrements of these
vales. In the case of D, it is noted that five
measrements are needed 3de to the relay mirrors4,
and ths the ncertainty reflects this. $ table of
e/"erimental constants is "rovided below*
Table 5. #/"erimental Constants
$ 3mm4 K87 T 5
1 3mm4 09Q T 5
D 3mm4 57077 T M
U 3m
0
4 56.7Q T .7Q
The constant U is derived by isolating c, J , and
?s- and noting that all other terms are constant
throghot the e/"eriment 3This is highlighted below in
#E9, all constant terms are in boldface4. $s discssed
above, two trials are rn for each vale of J* $ cloc%wise
trial and a conterFcloc%wise trial. This is done to
effectively doble the "recision of a given trial.
Throghot the trials, thogh, the rotational s"eeds
between cloc%wise and conterFcloc%wise trials are held
e+al. $lso, to %ee" the nits consistent, #EM is
mlti"lied by 0V to convert to radians "er second. #EM
then becomes*
c B CK3DL04$3Jcw @ JccwD G C3s-cw F s-ccw43D @ 14D
BH c B C59V3DL04$JD G C3s-cw F s-ccw43D @ 14D #E9.
BH c B UJ G 3s-cw A s-ccw4 #E9b.
Table 0 3ne/t "age4 shows the reslts of the
e/"eriment sing direct calclation. C is calclated
directly by first ta%ing a high vale 3based on ""er
limits of ncertainty4 and a low vale 3based on the
lower limits4. C is then ta%en as the mid"oint of the two
vales, with its ncertainty being the variance between
them. The relative error is also shown, which gives
insight into how the "recision of the e/"eriment
increases for larger vales of J.
Figre M 3ne/t "age4 shows a "lot of the
dis"lacement of the beam "ath verss the anglar
fre+ency. $ line of best fit is a""lied, as the data is
nearly linear. This is consistent with #E9, which im"lies
a the dis"lacement is a linear fnction of fre+ency with
slo"e m*
m B U G c
Ths deriving c from this line of best fit is sim"le.'ith a
slo"e of 7.7K00 3in microns G rev I sec
F5
4, we get*
c B 6.7P / 57
8
3T 5.9 / 57
9
4 mGs
BH ?c G c B 7.M W
CONC<U&ION $ND <I.IT$TION&
'hen we com"are the relative errors of the linear fit
method to the direct calclation method, we see that the
linear fit is star%ly more "recise even for the highest
vales of omega. This is de to the fact that a linear fit
yields a slo"e vale inde"endent of the rather large
ncertainties we get from re"eated measrement of or
e/"erimental constants. In "articlar, becase of the
modified method sedXThat is, the relaying mirrorsX
we obtain a significant sorce of ncertainty when we
measre or vale for D. &ince the slo"e of the line can
be determined withot these ncertainties, and the
gra"h shows a strong correlation between the "lotted
data and the fitted crve, we get a vale for c with a
small relative error.
The accracy of the e/"eriment, thogh, is on
the order of Y6W error for the linear fit, and Y5.QW error
for the direct calclation! however, this is only when we
consider the direct calclation vale with the highest
"recision. $ big limitation for this e/"eriment is the siRe
of the laboratory. &ince a larger D will allow each "lse
of light more time to "ro"agate, we can e/"ect easier
measrement for the dis"lacement of the "aths. This is
alleviated somewhat by the relaying mirrors, which
effectively increase D, bt as discssed above we
introdce random error in measrement by having to
ta%e more direct measrements to derive a vale for D.
.oreover, the sensitivity of the entire a""arats
can also yield n%nown sorces of error. In "articlar,
the difficlty in alignment can be a hindrance, and
obtaining a resolte image in the microsco"e involves a
lot of very "recise mani"lations of the microsco"e, the
mirrors, and the lenses.
)owever, des"ite these isses, the e/"eriment
still yields a relatively "recise and accrate vale. The
difficlty in measring c is e/em"lified by the strggle of
scientists who soght it for hndreds of years. #ven
today, there is no limit to the need for "recision in
determining this vale. .oreso now that so mch
de"ends "on it 3for e/am"le, the reFdefining of the
meter in terms of c4, and its im"lications are far
reaching in all as"ects of "hysics.
Table 0 3above4* Data and derived reslts of e/"eriment. Fig. M*
&catter "lot of dis"lacement verss anglar fre+ency with
linear fit.
Tabl e 0. Data and Resl ts
577 0.P0#@778 0.69#@778 0.9K#@778 0.Q8#@77Q 57.M
077 6.7P#@778 0.QK#@778 0.P5#@778 5.QQ#@77Q 9.5
677 6.5M#@778 0.8P#@778 6.70#@778 5.60#@77Q K.K
K77 6.5P#@778 0.PQ#@778 6.78#@778 5.7Q#@77Q 6.M
M77 6.07#@778 6.70#@778 6.55#@778 P.78#@779 0.P
977 6.7P#@778 0.PK#@778 6.70#@778 Q.K0#@779 0.M
Q77 6.70#@778 0.8P#@778 0.PM#@778 9.6K#@779 0.5
877 6.06#@778 6.57#@778 6.5Q#@778 9.M7#@779 0.5
P77 6.0K#@778 6.50#@778 6.58#@778 9.77#@779 5.P
5775 6.7P#@778 0.PP#@778 6.7K#@778 M.5Q#@779 5.Q
5KPM 6.7P#@778 6.75#@778 6.7M#@778 K.70#@779 5.6
J
3revs G second4
&-
cw
3Zm4
&-
ccw
3Zm4
?s-T7.M
3Zm4
C
@
3mGs4
C
FF
3mGs4
C
3mGs4
?C
3mGs4
Relative #rror
3?C G C4
5095 T 7.M 50M9 T 7.M M
5096 T 7.M 50MK T 7.M P
509M T 7.M 50M0 T 7.M 56
509Q T 7.M 50M7 T 7.M 5Q
509P T 7.M 50K8 T 7.M 05
50Q5 T 7.M 50KM T 7.M 09
50QK T 7.M 50K6 T 7.M 65
50QM T 7.M 50K0 T 7.M 66
50QQ T 7.M 50K7 T 7.M 6Q
5085 T 7.M 5068 T 7.M K6
50P5 T 7.M 500Q T 7.M 9K
R#F#R#NC#&
[&"eed of <ight Demonstration 1y Focalt .ethod\,
Uevin .acFarland,
Internet* htt"*GGwww."as.rochester.edGY"avoneG"articleF
wwwGteachersGdemonstrationsGFocaltDemonstration.htm
[FiReaFFocalt $""arats\
'i%i"edia Contribtors
Internet* htt"*GGen.wi%i"edia.orgGwi%iGFiRea
W#0W87WP6Focalt]a""arats
[Ole Roemer and the &"eed of <ight\
#dited 1y* Neil De(rasse Tyson and &teven &oter
Internet*
htt"*GGwww.amnh.orgGedcationGresorcesGrflGwebGessayb
oo%sGcosmicG"]roemer.html

You might also like