Rotational Dynamics

In this review, a systematic methodology is developed to obtain the the moment of inertia of any
system configuration of the PASCO model 9279 apparatus. Final experimental results are obtained
as a weighted average from several independent experiments with corresponding uncertainties. This
is made possible by a set of custom Mathematica code. This methodology is used to explore various
configurations and topics such as conservation of angular momentum and principle axis of rotation
are explored. In addition, frictional torque is well-characterized with a third degree polynomial, and
can be used to achieve higher accuracy of measurements in a future experiment. An unexpected
linear relation between frictional torque and angular velocity was also uncovered.

INTRODUCTION

Rotational dynamics is a crucial component of classical

mechanics. A basic understanding of the subject can be
obtained from exploring the angular acceleration of rigid
objects, calculating their moment of inertia and compare
to theoretical values. Such an experiment is ideal for de-
veloping a systematic approach to experimental physics
as well as practicing uncertainty analysis. During the ex-
periment, a wide variety of data analysis and uncertainty
analysis take place such as linear fitting, evaluate the
quality of fit using ˜2 values, finding the uncertainty in
the fit parameter, combining multiple experimental data FIG. 1: PASCO scientific model 9279 [1]
using weighted average and error propagation. Thanks
to PASCO scientific, we have access to a near-frictionless
is present, air comes out of the metal mounting at two
platform on which metal disks can be spun. Accessories
di↵erent valves - one below the bottom disk and one in
can also be attached to the disk to create more complex
between the two disks. The air flow creates an air cushion
rotational systems. An optical reader records the angular
for the bottom disk to rotate without contacting the sur-
velocity of the disks and the data is easily accessible with
face of the apparatus, dramatically decreasing friction.
a ScienceWorkshop interface. Data acquisition can also
The air flowing out in between the disks separates the
be handled through DataStudio which allows us to per-
top disk from the bottom disk, isolating their motions.
form a wide range of interesting experiments and detailed
Without the valve pin, the air valves in between the disks
data analysis on this subject.
are shut o↵, allowing the two disks to interact through
A methodology for performing experiments with our
friction.
apparatus is developed during the first experiment about
The apparatus comes with two steel disks, one designed
angular acceleration. The same methodology is used
to be the bottom disk and the other the top disk. The
throughout the rest of the experiments and are there-
two are not interchangeable. A lighter aluminum top disk
fore not explained in detail. This review is divided up
was also included. Other accessories will be discussed in
into individual experiments. Within each experiment,
the appropriate experiment section.
the setup, theory and data analysis are presented in or-
In setting up the apparatus, we connect it to a com-
der.
pressed air supply and adjusted the pressure of the air to
9 psi as suggested by the manual. [1] Before any experi-
ment could be conducted, however, the apparatus had to
APPARATUS be leveled, otherwise asymmetry in the mass distribution
of the rotational system (disks and attached accessories)
The PASCO scientific model 9279 rotational dynam- will cause a torque that interferes with experiments. To
ics apparatus(shown in Figure 1) was employed in this do so, a small mass was placed on the edge of the top
experiment. The metal mounting houses two vertically disk with the air supply on. The mass oscillates around
separated disks. The disks are covered in alternating the direction of tilt and eventually rests in that direction.
black and white strips(200 black strips per disk) allow- The three adjustable supporting legs are then adjusted
ing the optical reader to record their angular velocities. according to this information to get the apparatus as level
Air is supplied to both the metal mounting and the air as humanly possible.
bearing pulley to minimize friction. When a valve pin The optical reader records the angular velocities of
 2

FIG. 2: Scienceworkshop 500 interface - PASCO model CI-

6400[2]
FIG. 4: ”both2” indicates that both disks are spinning to-
gether, the 2 signals that this is a second trial.

setup ↵(rad/s2 ) ↵(rad/s2 ) ˜2

top1 2.07 0.01 0.97
top2 2.07 0.01 1.00
top3 2.06 0.01 1.00
both1 1.05 0.01 0.95
both2 1.04 0.01 0.99
both3 1.05 0.01 0.96

TABLE I: Measured angular acceleration, uncertainty (in the

fitting parameter) and the quality of fit are listed above
FIG. 3: Setup for the acceleration experiment[1]

acceleration of the top disk, in the second experiment,

both top and bottom disks and data is handled through we tape the two disks together (using transparent tapes
a PASCO model CI-6400 ScienceWorkshop 500 interface as to not interfere with the optical reader) and measure
(shown in Figure 2). Data are collected using DataStu- their acceleration as a whole. We ran 3 experiments for
dio, and are then analyzed using custom Mathematica each setup.
code. (Appendix A) To start the experiment, we raise the mass to the top of
the table, turn o↵ air supply and make sure the setup is
stable. Then turn on air supply after data acquisition has
ANGULAR ACCELERATION started. Typical collected data are included in Figure 4.
Collected data are analyzed using custom Mathematica
Method and Data code (Appendix A). The code truncates only the acceler-
ating part of the data, for example 3-11s in Figure 3, fits
As a first experiment, we measure the angular acceler- a linear function A+Bt to the data and returns the slope
ation of the steel disks when pulled by a small hanging B, which should be the angular acceleration if friction is
mass and then compare it to the prediction of a minimal negligible. Data are recorded in Table I, the ˜2 values
theoretical model to assess the quality of our setup. The for the linear fits are extremely close to 1, this says that
apparatus is setup as shown in Figure 3 with minor dif- our program is capable of picking out the linear part of
ferences due to lack of accessories. First, we do not have the data also that friction is indeed negligible during the
a holder for multiple masses, so the hanging end of the acceleration process. Data from the three trials are com-
string is fixed to a single mass with m = .024kg. Second, bined using the weighted average method introduced by
there’s no thread holder so the thread is wrapped around Taylor [3, p 175]
the torque pulley and has its end taped to the bottom P
of the pulley. There were two torque pulleys available w i xi
x̄ = P (1)
and we choose the one with a larger radius r1 = .024m wi
to e↵ectively accelerate the heavy steel disks. Without a
where
hollow thumb screw, we can not use the valve pin to bring
the disks into contact, thus in the first experiment we let 1
wi = (2)
the two disks run independently and only measure the 2
i
 3

Mt (kg) Mb (kg) Ma (kg) m(kg)

1.359 1.348 0.469 0.024
Rt1 (cm) Rt2 (cm) Rb1 (cm) Rb2 (cm)
0.55 ± 0.03 6.30 ± 0.03 1.18 ± 0.03 6.30 ± 0.03
Ra1 (cm) Ra2 (cm) r1 (cm) r2 (cm)
0.81 ± 0.03 6.30 ± 0.03 2.40 ± 0.03 1.40 ± 0.03

TABLE II: Mt , Mb , Ma are the masses for the top steel disk,
FIG. 5: Minimal model for acceleration experiment[1] bottom steel disk and aluminum disk respectively. m is the
mass of the pulling object. Rt1 , Rt2 , Rb1 , Rb2 , Ra1 , Ra2 are
inner and outer radii of top steel disk, bottom steel disk and
and the combined standard deviation is aluminum disk. r1 , r2 are the radii of the large and small
pulley respectively
1
= pP (3)
wi

The combined data is listed and compared to theoretical Z tZ Z

2⇡ R2
predictions in Table III.
Idisk = ⇢r2 rdrd✓dz (8)
0 0 R1

Uncertainty where ⇢ = M/⇡(R22 R12 )t. Evaluating the integral,

⇡ 4 1
Given the set of measured angular velocity yi , linear re- Idisk = (R R14 )t ⇤ M/⇡(R22 R12 )t =
M (R12 + R22 )
gression is performed to fit angular velocity against time, 2 2 2
(9)
A + Bxi , the slope B is the desired angular acceleration. To obtain the theoretical value of the angular acceler-
The uncertainty in the acceleration are calculated accord- ation, we perform a standard mechanics analysis of the
ing to Taylor[3, p 184-188]. setup shown in Figure 5 (b). We denote the mass of the
r pulling object by m, the radius of the big pulley by r1 ,
N
B = y (4) tension in the string is T and the gravitational constant
is taken to be g = 9.8m/s2 . Tension in the string causes
where the disk to accelerate
v
u N ⌧ = Isys ↵ (10)
u 1 X
y =t (yi A Bx2i )2 (5)
N 2 1 where ⌧ is the torque exerted on the disk-pulley system,
Isys is the moment of inertia of the entire system and ↵ is
and the angular acceleration. As a minimum model we do not
X X take into account the moment of inertia of the pulley and
=N x2 ( x)2 (6) the air bearing as well as the additional torque exerted
by air flow. Thus Equations (10) simplifies to
The quality of the fit is accessed using
N T r1 = Idisk ↵ (11)
X yi f (xi )
2 2
˜ = ( ) /d (7)
i The linear acceleration of the mass is given by Newton’s
1
2nd law
where yi is the expected value of the experimental data
f (xi ) with input xi , and d is the degree of freedom. i a=g T /m (12)
are approximated by y [3, p 268]
Substitute in Equation (11) and relate linear and angular
acceleration by ↵ = a/r1 we arrive at
Theory and Analysis mgr1
↵= (13)
mr12 + Idisk
In a minimum theoretical model, we treat the steel disk
as a cylindrical shell with inner radius R1 , outer radius The necessary dimensions and weights are measured and
R2 , thickness t, mass M with constant density, as shown listed in Table II. A comparison of theoretically and ex-
in Figure 5 (a). The angular momentum of the disk is perimentally obtained accelerations are compared in Ta-
obtained through a standard mechanics calculation ble III. The uncertainties in the theoretical predictions
 4

e↵top (rad/s2 ) t↵top (rad/s2 ) e↵both (rad/s2 ) t↵both (rad/s2 )

2.07 ± 0.01 2.07 ± 0.04 1.04 ± 0.01 1.026 ± .014

TABLE III: Theoretical and experimental accelerations for

top and both disks

FIG. 8: Experimental set up for the collision experiment [1]

trial b!t (deg/s) b!b (deg/s) a!(deg/s)

1 876 ± 5 944 ± 7 33 ± 4
2 895 ± 7 654 ± 6 120 ± 4
3 1033 ± 8 1062 ± 7 8±2
FIG. 6: When only top disk is spinning 4 94 ± 5 894 ± 7 397 ± 4
5 88 ± 4 927 ± 7 420 ± 4
6 83 ± 5 929 ± 6 420 ± 5
are calculated using prorogation of uncertainty on Equa- 7 166 ± 6 972 ± 5 565 ± 5
tion (13), with the constants from Table II. A more 8 176 ± 6 1175 ± 8 672 ± 4
9 81 ± 2 1258 ± 7 665 ± 6
visual comparison is included in Figures 6,7, from these
figures we conclude that our theoretical and experimen-
tal angular accelerations of the system in question agree TABLE IV: Angular velocity of top and bottom disk
within the uncertainties of the experiment. This implies before(b!t ,b!b ) and after(a!) the collision
that our set up is reasonable and friction can be ignored
on short time scales (10-60s). It also means that the anal-
ysis code we developed indeed do return fairly accurate 1. Top disk spin fast counter-clockwise, bottom disk
results, therefore those scripts will be used throughout spin slow clock-wise
the rest of the experiment.
2. Top disk spin slow counter-clockwise, bottom disk
spin fast clock-wise
COLLISION
3. Top disk spin slow counter-clockwise, bottom disk
spin fast counter-clockwise
Method and Data
With the air supply on, we spin the disks to desired
In this experiment we aim to verify the conservation of rates with the valve pin plugged in (disks are isolated),
angular momentum during collision. The setup is shown start data acquisition wait 3 seconds and then let the
in Figure 8. Only the steel disks and the valve pin are disks collide, after collision, we continue to take data for 3
used in this experiment. Three di↵erent scenarios are seconds. At the default acquisition rate of 10Hz, we have
explored, they vary in the direction of spin and the spin 30 data points before the collision and after the collision.
rates. The scenarios are The mean and standard deviation of these data points
are taken to be the angular velocity and its corresponding
uncertainty. Data are recorded in Table IV.

Theory and Analysis

By conservation of angular momentum

It b!t + Ib b!b = (It + Ib )a! (14)

Thus
It Ib
FIG. 7: When both disks are spinning a! = b!t + b!b (15)
It + Ib It + Ib
 5

trial ta!t (deg/s) ea!(deg/s)

1 43 ± 11 33 ± 4
2 113 ± 10 120 ± 4
3 24 ± 13 8±2
4 405 ± 7 397 ± 4
5 424 ± 7 420 ± 4
6 428 ± 7 420 ± 5
7 573 ± 6 565 ± 5
8 680 ± 7 672 ± 4
9 675 ± 7 665 ± 6

TABLE V: Theoretical and experimental results for collision

experiment

FIG. 10: Setup for linear and angular momentum

experiment[1]

FIG. 9: Theoretically predicted and experimentally measured FIG. 11: Top view of the linear and angular momentum ex-
angular velocity after collision, the closer the points are to the periment setup[1]
diagonal line, the more theory and experiment agree

Using theoretical values obtained from Equation (9), pre- Speed of the metal ball right before collision is esti-
dictions for a! can be obtained. The results are com- mated using the setup shown in Figure 12 where h is
pared with experimental data in Table V. Uncertainties measured to be 89.0 ± 0.1cm. To determine L, we coat
are obtained using standard uncertainty propagation. [3, the metal ball with dry erase marker and place a piece
p 75] All data agree within 2 standard deviations, a more of plain paper close to the landing spot. After 5 tri-
visual comparison is provided in Figure 9. als the ball landed 62.5, 63.2, 63.2, 63.2, 64.5cm away from
the table, whose mean and standard deviation are used
for L, namely L = 63.3 ± 0.9cm. Thus the velocity of the
LINEAR MOMENTUM AND ANGULAR
MOMENTUM

Knowing that angular momentum is conserved, we’d

like to show that linear momentum can be viewed as
angular momentum in a naı̈ve manner. That is a point
mass m traveling with velocity ~v perpendicular to the axis
of rotation will have angular momentum
L = I! = mr2 v/r = mvr (16)
The setup for the experiment is shown in Figure 10. The
lighter aluminum top disk is used in this experiment to
amplify the e↵ect of the collision for better measurability.
The experiment procedure can be more clearly seen from
a top view(Figure 11). A metal ball is release from the
top of the ramp, it is then caught in the ball-catcher, FIG. 12: Setup for estimating of the velocity of the ball after
which causes the top disk to spin. accelerating down the ramp[1]
 6

D(cm) Li (g · m2 rad/s) !(rad/s) Lf (g · m2 rad/s)

8.0 3.40 ± 0.18 1.69 ± 0.11 1.36 ± 0.14
7.5 3.19 ± 0.17 1.43 ± 0.13 1.12 ± 0.16
7.0 2.98 ± 0.16 1.45 ± 0.15 1.10 ± 0.18
6.5 2.77 ± 0.15 1.41 ± 0.11 1.05 ± 0.13
6.0 2.55 ± 0.14 1.30 ± 0.09 0.94 ± 0.11
5.5 2.34 ± 0.13 1.17 ± 0.12 0.83 ± 0.14
5.0 2.13 ± 0.11 0.93 ± 0.04 0.64 ± 0.05
4.5 1.92 ± 0.10 0.92 ± 0.08 0.63 ± 0.09
4.0 1.70 ± 0.09 0.82 ± 0.10 0.55 ± 0.12
3.5 1.49 ± 0.08 0.64 ± 0.03 0.42 ± 0.04
3.0 1.28 ± 0.07 0.80 ± 0.06 0.52 ± 0.06 FIG. 13: Setup for the variable radius mass experiment

TABLE VI: D is the distance from the ball to the axis of VARIABLE RADIUS MASS
rotation, Li is calculated by Equation (18),! is the angular
velocity of the system after collision and Lf is calculated using
Experiment
Equation (19)

To explore how the mass distribution of a rotational

ball is determined by the relation system a↵ects its angular momentum, we set up the
experiment such that 2 masses (each having a mass of
r 0.201kg) mounted on a metal support can move in the
g
v=L (17) radial direction as shown in Figure 13. The basic setup is
2h
the same as that in the angular acceleration experiment.
But here the aluminum plate is in use and the holes on
numerically v = 1.49 ± 0.02m/s.
top of the plate are covered to provide enough lift for the
The catcher can catch the ball about anywhere from heavy masses. The two masses should always be kept the
3cm to 7cm away from the axis of spin. In the experiment same distance away from the axis of rotation, otherwise
we shoot the ball at di↵erent distances from the axis and the top plate will tilt in one direction and come in con-
record the angular velocity of the top disk after the colli- tact with the bottom plate, which causes a lot of friction
sion - the top disk stays stationary before collision. The and interferes with the experiment.
catcher has a ruler glued on top of it whose readings are The distance between the axis of rotation and the cen-
distances from the axis, and we record the distance of ter of the variable radius mass d are measured and the
the farther end of the ramp before each experiment. The angular acceleration of the system is obtained in the same
distance from the center of the ball to the axis can then way as in the angular acceleration experiment. To iso-
be calculated by subtracting o↵ the radius of the ball. late the inertia contribution from the masses, we conduct
The angular momentum of the system before collision if an additional experiment with the two masses taken o↵.
the ball collides at a distance D from the axis is The moment of inertia measured in this case is that of the
base. This moment of inertia was subtracted from that
Li = mb vD (18) of the system to give the moment of inertia of the two
masses by themselves. Best estimates and corresponding
and the angular momentum after collision is given by uncertainties are obtained using our custom Mathemat-
ica code (Appendix A). All the ˜2 values are extremely
Lf = Isys ! (19) close to 1 (mean 0.995) if not less (minimum 0.97), and
we have already seen that a linear fit to the angular ve-
where locity data on this time scale is completely justifiable,
therefore ˜2 values are not presented in detail. All other
Isys = Ibase + mb D2 (20) data are listed in Table VII.

Data are recorded in Table VI. Our data is mysteriously

o↵ by roughly a factor of 2. We remeasured the angular Theory
momentum of the base (aluminum disk and ball catcher),
the velocity of the ball and the angular acceleration of the We treat the masses as cylinders with uniform density,
system after collision at multiple distances from the axis, the height and radius of the cylinder are determined to
the results are robust - but systematically o↵. Unfortu- be h = 3.20±.01cm and R = 0.016±0.01cm respectively.
nately we were not able to track back to the cause of this Consider one such mass spinning around the x axis as
systematic error. shown in Figure 14 The moment of inertia of the mass
 7

FIG. 14: Uniform cylinder model for variable radius mass FIG. 15: Comparison of theoretically calculated and exper-
imentally obtained moment of inertia of the variable radius
mass system
d(cm) eI(gm2 ) tI(gm2 )
3.6 ± 0.1 1.05 ± 0.16 0.55 ± 0.15
4.6 ± 0.1 1.22 ± 0.26 0.88 ± 0.14
5.6 ± 0.1 1.52 ± 0.19 1.29 ± 0.12
6.6 ± 0.1 2.07 ± 0.23 1.78 ± 0.11
7.6 ± 0.1 2.14 ± 0.34 2.35 ± 0.09
8.6 ± 0.1 2.74 ± 0.28 3.00 ± 0.07
9.6 ± 0.1 3.49 ± 0.33 3.73 ± 0.06

TABLE VII: eI is the experimentally measured moment of in-

ertia contribution of the two masses and tI is the theoretically
predicted moment of inertia contribution of the two masses

can be determined through basic mechanics as

ZZZ
FIG. 16: Triangular mass [1]
I= ⇢(y 2 + z 2 )dxdydz (21)
cylinder

of the small pulley has a large fractional uncertainty and

where ⇢ = ⇡rm2 h , parametrize in cylindrical coordinates
the large theoretical uncertainty came from the axis-mass
and put in the integration limits we have
distance d, whose measurement is limited by our ability
Zh/2 Z2⇡ZR to distinguish where the center of mass of the masses are
m and where exactly the center of the apparatus is located.
I= (r2 sin2 ✓ + z 2 )rdrd✓dz (22)
⇡R2 h
h/2 0 0

evaluate the integral PRINCIPLE AXIS

1
I= m(3R2 + h2 ) (23) The setup is the same as the acceleration experiment
12
with some changes. Firstly the thumb screw is replaced
by the parallel axis theorem, when the center of the cylin- with by a triangular mass that can be rotated around a
der is distance d away from the x axis, the moment of horizontal axis, see Figure 16. Secondly, the steel disk
inertia is is replaced by the lighter aluminum disk, and the tiny
1 1 wholes in the aluminum disk are covered by tapes to force
I= md2 + m(3R2 + h2 ) (24) air down as to provide more lift to the disk. Thirdly,
2 12
instead of the big pulley we use the smaller pulley since
the theoretically calculated values of the moment of in- the lighter aluminum disk is in use. Recall the calculation
ertia of the two masses and the experimentally measured for angular acceleration in the first experiment Equation
values are listed for comparison in Table VII. A more (13). It can be adopted to this system by simply changing
visual comparison is shown in Figure 15 Large experi- Idisk into Isys , also we’re using the small pulley instead
mental uncertainties arise from the fact that the radius of the big one this time so r1 needs to be substituted by
 8

FIG. 18: Theoretical and experimentally obtained angular

acceleration for the rotating triangle depending on the orien-
tation of the axis of rotation
FIG. 17: Triangular mass placed in a coordinate system(it’s
also centered at the x-axis but not shown)
the theoretical angular acceleration can then be calcu-
lated and compared with experimentally obtained val-
r2 . By solving for Isys
ues in Figure 18. Notice there’s an axis that orientation
mgr2 that gives the minimum angular acceleration, this is the
Isys = mr22 (25)
↵ principle axis. An interesting observation is that if the
triangle is originally o↵set from the principle axis and
In the experiment, we first find the acceleration of the the mounting screw is not tight - that is the triangle can
”base”, that is the system with everything but the tri- spin around its mounting axis with relative easy - after
angle attached, in the same way we did the acceleration the triangle reaches a high enough angular velocity it ori-
experiment. The acceleration is then converted into an- entates itself back to the principle axis. Our explanation
gular momentum using Equation (25). In a similar man- for this phenomenon is that the principle axis is the axis
ner, the moment of inertia of the system is measured for of no ”wobbling” and is therefore the stationary configu-
many orientations of the triangle. ration. It is also the configuration of least angular accel-
eration, which is also the configuration that minimized
energy lost, since friction is proportional to angular ve-
Theory and Analysis
locity which is proportional to angular acceleration as we
will discuss later.
To analyze the moment of inertia of the triangle, we
place it in a coordinate system as shown in Figure 17.
The inertia tensor of this triangular mass is calculated
FRICTION
using standard mechanics
ZZZ
Iij = ⇢( ij rij xi xj )dV (26) Friction is arguably the most difficult element to take
into account in this series of experiment. It is not a
simple linear acceleration for objects in contact, rather
where density ⇢ is assumed to be a constant, ij is the a velocity dependent toque caused by the airflow under
Kronecker delta. We also assumed that the hole in the the disks. We have no theoretical model for the cause
middle of the triangle has negligible a↵ect on its moment of this frictional torque, but it can be characterized very
of inertia. Thus ⇢ = mt /( 12 abt). Using a custom Math- well experimentally.
ematica program, the inertia tensor was determined to
The apparatus is setup the same way as in the collision
be
experiment(Figure 8). We perform the experiment in 2
0 1
2.13 0 0 di↵erent configurations
I = @ 0 1.27 0.3 A g · m2 (27)
0 0.3 0.86 1. top disk spinning, bottom disk taped to the base
of the apparatus, the two disks spins independently
The angular momentum around axes in the form ~a = with the presence of the valve pin
(0, cos ✓, sin ✓) are evaluated as
2. both disks are spinning together without the pres-
I~a = ~aT I~a (28) ence of the valve pin
 9

FIG. 21: Angular acceleration caused by friction is linearly

FIG. 19: Friction data for top disk spinning fitted to 3.28 ⇥ related to angular velocity
10 12 t3 + 2.95 ⇥ 10 6 t2 0.13t + 1462

is developed to efficiently analyze data obtained us-

ing DataStudio. GetAcl[filename,acqfreq] automatically
truncates the data to only the linear part and returns the
angular acceleration. GetDAcl[filename,acqfreq] does
the same linear fit but returns the uncertainty in the
angular acceleration using Taylor’s estimate for uncer-
tainty in linear fitting. It also returns ˜2 for the lin-
ear fit using Taylor’s method with i estimated using
y (Equation 5). GetFreq[filename] takes in angular ve-
locity of a collision scenario where the disk start from
rest and returns the angular acceleration of the disk af-
ter collision. Acl2Mom[↵,m,r], takes in the angular ac-
celeration, pulling mass and pulley radius of the angular
FIG. 20: Friction data for both disks spinning fitted to
3.74 ⇥ 10 10 t3 + 1.58 ⇥ 10 5 t2 0.26t + 1634
acceleration setup (Figure 3) and calculates the angular
momentum ignoring friction and the moment of inertia
of the pulley. For an experiment to determine the an-
To start the experiment, we spin the desired disk(s) up gular acceleration of a system configuration, we run the
to a fairly high angular velocity (⇠1500 Hz), start data angular acceleration experiment 3 times and average the
acquisition and see the disk(s) spin down. The obtained obtained acceleration and uncertainty then combined the
data can be fitted extremely well with a third order poly- results using weighted average introduced by Taylor.
nomial as shown in Figure 19,20 Angular acceleration experiment employs a cylindrical
What’s more interesting is that if we break the data shell model for the rotating disks, whose angular mo-
into 100s chunks and calculate the angular acceleration of mentum is given by I = 21 M (R12 + R22 ) where M is the
each chunk using Mathematica code in Appendix A and mass of the disk, and R1 , R2 are the inner and outer ra-
compare them to the angular velocity data, we found that dius of the cylindrical shell respectively. Experimentally
the two are linearly related. See Figure 21. measured angular accelerations agree very well with the
We’re not sure of the origin of this result, but it can theoretical values. Confirming the validity of our model.
be used as a correction factor to improve previous re- Collision experiment confirms the law of conservation of
sults. We realized this at the end of our experiment and angular momentum. Linear and angular momentum ex-
therefore did not pursue the question further, but it is periment failed to confirm the relation between angular
definitely one of great interest for anyone who would like and linear momentum, with an unexplained systematic
to explore the subject in the future. error of 160%. Potential sources of error were investi-
gated. Unfortunately the error was not tracked down.
Variable radius mass experiment shows a that a cylinder
CONCLUSION model for the masses is valid and the moment of inertia
of the system increase quadratically with the separation
A methodology for performing rotational dynamics ex- of the masses. In the principal axis experiment, we found
periments with PASCO model 9272 is developed. For that the moment of inertia of a rigid mass di↵ers depend-
a single data acquisition, a set of Mathematica code ing on the axis of rotation. However, there’s a special axis
 10

where the moment of inertia of the system is maximum periment. The author would also like to thank Dr. Daniel
and the mass doesn’t wobble during the rotation. Gibson who provided the equipment for this lab and of-
Last but not least, an analysis of the frictional torque fered helpful comments throughout the experiment.
on the disks reveals that the angular acceleration caused
by friction can is linearly related to angular velocity, this
can be used to improve experimental accuracy for this
series of experiment, and this relation itself is interesting
enough to be pursuit further in a future exploration of [1] Ed Pitkin. Rotational Dynamics Apparatus. PASCO sci-
entific, 1999.
the subject. [2] PASCO scientific. ScienceWorkshop 500 Interface, 2001.
[3] John R. Taylor. An Introduction to Error Analysis. Uni-
versity Science Book, 1997.
ACKNOWLEDGMENT

The author would like to acknowledge his lab partner

for setting up and running the ex-

APPENDIX A MATHEMATICA ANALYSIS CODE

In the first segment of our custom code, best estimates of the angular acceleration and moment of inertia of
the rotational system is calculated. The GetAcl function uses FindStart and FindEnd to pick out the linearly
increasing portion of the input data and calculate the slope of the most appropriate fitted line. This slope is processed
to have the appropriate units and returned as the angular acceleration of the system. Acl2Mom further converts
this information to the angular momentum of the system in question.

FIG. 22: Code for extracting the moment of inertia of the rotational system

The second segment of our code calculates the uncertainty in the slope as well as the ˜2 value of the fit.
The last segment of our code extracts the angular velocity of a collision experiment if the disks are originally at
 11

FIG. 23: Code for calculating the uncertainties in the angular acceleration as well as ˜2 value of the fit performed in GetAcl

FIG. 24: Code for extracting angular velocity after collision from rest

rest.