Dynamometer Project Writeup

Presented To: Dr. Milne Andrew

Project Name: Dynamometer Project
Jingkai Shao, Jintai Li, Jiya Patel, Joanie Hui, & Jennifer Xia
7 October 2025

Design Decisions:
We chose the design shown in [1] for the following reasons:
1.​ Experiment Accuracy
a.​ While recording data for the graph of RPM vs Torque exerted on the TT motor,
this design ensures that the rope of the hanging mass is clear from the edge of
the table or any other obstruction that can potentially create additional force on
the system, which reduces systematic error.
b.​ Design of a truss-structured base [1.C] will ensure that the structure remains
static while a higher magnitude of torque done by weight is applied onto the
system during the testing procedure.
c.​ While inserting the metal rod into the cup [1.B] in order to increase the magnitude
of the torque done by weight onto the TT-motor, the rods were inserted directed
under the table without touching the top section of the dynamometer, which
reduces the risk of someone accidentally tipping the mass and systematic error
during the active experiment process.
2.​ Experiment Efficiency
a.​ Inserting the circuit board onto the dynamometer [1.C] will reduce the time
needed to start the motor.
3.​ Easy to Transport
a.​ The handle on the tip of the dynamometer [1.C] will make the system easier to
transport from the locker to the table.

Test Procedure:
First, the initial length, L0, of the rubber band was measured when the cup was empty, as shown
in [1]. Weights of known mass were added to the cup. After each weight increment, ΔW, the
stretch of the rubber band was measured and recorded as L1, L2, L3, … Ln, with Ln as the length
of the rubber band when the motor stalls. The rotational speeds were measured by recording
the rotational motion using a smartphone for every increment in weight. The video was slowed
down to count the number of rotations per time increment recorded. Since the recordings were
an average of 9 seconds long, the ratios needed to be multiplied by 60 to get the number of
rotations per one full minute. The elastic constant was measured by hanging the string,
connected to the empty cup at and the elastic band at opposite ends, on a vertically suspended
hook. A ruler was used to measure the initial length, L0, of the rubber band when the cup was
empty. Weights were added gradually, and the stretch of the rubber band was measured after
each increase in weight, ΔW, and recorded as LA, LB, LC, … Ln, where n is the final test number.

Figure 1: [A] Front view, [B] top view, [C] left view, and [D] load cell calibration of the
dynamometer setup, showing the TT motor, elastic band, hanging mass, string, and dimensions
of the dynamometer.
Appropriateness of Using an Elastic Band as a Load Cell:

According to Hooke's law, 𝐹𝑒𝑙𝑎𝑠𝑡𝑖𝑐 = 𝑘∆𝑥, the restoring force or elastic force of an elastic material

(𝐹𝑒𝑙𝑎𝑠𝑡𝑖𝑐) is directly proportional to the displacement from the elastic material’s equilibrium state (

∆𝑥). The elastic constant (𝑘) is a measure of the stiffness of the elastic material, which in this
case is an elastic band. The use of an elastic band as the load cell in the dynamometer is
appropriate for several reasons, including its basic function, accuracy, and accessibility. It fulfills
its purpose of measuring the elastic force by evaluating the band’s elongation from known loads,
calibrated separately from the system, as outlined in the procedure. By plotting the elastic force
as the dependent variable and the respective change in length of the rubber band (elongation)
as the independent variable in [2], a proportional relationship between force and elongation is
observed, consistent with Hooke’s Law. The slope of the best-fit line from the elastic force vs
elongation graph [2] gives the elastic constant (k) in N/m, which is 58.002 N/m. In addition, the
elastic band is lightweight, inexpensive, and readily available, thereby making it straightforward
to implement in the system as a load cell. As such, these factors make an elastic band a
reasonable choice for the dynamometer’s load cell that can reliably be used in the calculation of
torque for the TT motor.

Figure 2: A scatter plot graph made in Excel showing the relationship between elastic force
and elongation of the elastic band, with the elastic constant as the slope.
Note: In Figure 2, the relation between elongation and Felastic is a 2nd degree polynomial
trendline with equation
2
𝐹(𝐸) =− 418. 36𝐸 + 86. 266𝐸 + 0. 0173, 𝑤ℎ𝑒𝑟𝑒 𝐹 (𝐹𝑒𝑙𝑎𝑠𝑡𝑖𝑐) 𝑖𝑠 𝑎 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑜𝑓 𝐸 (𝐸𝑙𝑜𝑛𝑔𝑎𝑡𝑖𝑜𝑛).
In order to get the elastic constant 𝑘, we want to find the instantaneous rate of change of the
𝑑𝐹
elastic force at a certain Elongation. The value of 𝑑𝐸
was evaluated at 𝐸 = 0. 018𝑚. Calculation

is shown below:
2
𝐹(𝐸) =− 418. 36𝐸 + 86. 266𝐸 + 0. 0173
𝑑𝐹 43133
​ ​ ​ 𝑑𝐸
= 500
− 836𝐸​ ​ ​
𝑑𝐹
𝐴𝑡 𝐸 = 0. 018𝑚, 𝑑𝐸
= 71. 218 = 𝑘 (𝑉𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑒𝑙𝑎𝑠𝑡𝑖𝑐 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑢𝑠𝑒𝑑 𝑖𝑛 𝑓𝑜𝑟𝑚𝑢𝑙𝑎 𝑜𝑓 𝑇𝑜𝑟𝑞𝑢𝑒)

Physics and Sample Calculation for Torque of the Motor:

The net torque acting on the drum is the torque of the motor (τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛) + torque caused by the

frictional force between the string and the drum (τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛). τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛 is defined positive as its

direction is counterclockwise, and τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛 is defined negatively as its direction is clockwise.

Since the drum is always rotating at constant speed, its rotational acceleration is 0. The
relationship can be modeled as:
τ𝑁𝑒𝑡 = τ𝑀𝑜𝑡𝑜𝑟 − τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛 = 𝐼α = 0, 𝑏𝑒𝑐𝑎𝑢𝑠𝑒 α = 0 ⇒ ω𝑓 = ω𝑖 𝑎𝑡 𝑒𝑎𝑐ℎ 𝑝𝑜𝑖𝑛𝑡 𝑡ℎ𝑎𝑡 𝑤𝑒 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 𝑡𝑜𝑟𝑞𝑢𝑒

We can then derive the relationship between the torque of the motor and the torque caused by
the friction force to be equal:
τ𝑀𝑜𝑡𝑜𝑟 = τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛

There are two tension forces (𝑇𝑊𝑒𝑖𝑔ℎ𝑡 & 𝑇𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒) acting on the string and therefore the drum

to cause the friction force between the string and the drum. Torque caused by the 𝑇𝑊𝑒𝑖𝑔ℎ𝑡 is

positive as it is counterclockwise, and the torque caused by the 𝑇𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒 is negative as it is

clockwise. The torque caused by friction can then be written as the following:
τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛 = τ𝑤𝑒𝑖𝑔ℎ𝑡 − τ𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒 = (𝑇𝑊𝑒𝑖𝑔ℎ𝑡 − 𝑇𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒) * 𝑅𝑑𝑟𝑢𝑚

Knowing 𝑇𝑊𝑒𝑖𝑔ℎ𝑡 and 𝑇𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒 is has the same magnitude and direction as the Weight and

Elastic force, τ𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛 can then be substituted in the previous relationship to get:

τ𝑀𝑜𝑡𝑜𝑟 = (𝑇𝑊𝑒𝑖𝑔ℎ𝑡 − 𝑇𝐸𝑙𝑎𝑠𝑡𝑖𝑐 𝐹𝑜𝑟𝑐𝑒) * 𝑅𝑑𝑟𝑢𝑚

Along with constants like the acceleration due to gravity (g), the radius of the drum (𝑅𝑑𝑟𝑢𝑚), and

the calibrated elastic constant of the load cell [2], the torque of the motor can be calculated
depending on the mass of the hanging weights (𝑚) and the measured elongation (𝑥):
τ𝑀𝑜𝑡𝑜𝑟 = (𝑚𝑔 − 𝑘Δ𝑥) * 𝑅𝑑𝑟𝑢𝑚

Therefore, the torque-speed curve [3] is obtained by plotting the measured rotation speed of the
motor with respect to the torque of the motor.

Figure 3: A scatter plot graph made in Excel using the experimental data, showing the
torque-speed curve of the dynamometer [2].

Explanation of the Level of Quality of Torque Speed Curve:

Calculation of Percent Error for this experiment:
According to Figure 4, our group's experimental data for the stall torque and the numerical
simulation data for the stall torque have a percent error of
​ ​
𝐸𝑥𝑝𝑒𝑟𝑖𝑚𝑒𝑛𝑡𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 − 𝑇ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 |
% 𝑒𝑟𝑟𝑜𝑟 𝑜𝑓 𝑠𝑡𝑎𝑙𝑙 𝑡𝑜𝑟𝑞𝑢𝑒 = || 𝑇ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 | × 100%
0.072−0.08
​ ​ % 𝑒𝑟𝑟𝑜𝑟 𝑜𝑓 𝑠𝑡𝑎𝑙𝑙 𝑡𝑜𝑟𝑞𝑢𝑒 = || 0.08 || × 100% = 10%
Analysis of the Quality of Torque vs Speed Curve Data:
The torque-speed curve shows a clear negative relationship between torque and rotational
speed. y=-0.0003x+0.741 closely follows the theoretical data. There are mainly three types of
experimental errors:
1.​ Random errors
a.​ According to Figure #4, the points is distributed along the trendline. So, we are
able to calculate the standard deviation (σ) to quantify the random error in this
experiment.
𝑛
2
∑ (𝑥𝑖−𝑥)
𝑖=1
σ= 𝑛−1
, 𝑤ℎ𝑒𝑟𝑒 𝑛 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑎𝑡𝑎 𝑝𝑜𝑖𝑛𝑡, 𝑥𝑖 = 𝑒𝑎𝑐ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑑𝑎𝑡𝑎, 𝑥 = 𝑚𝑒𝑎𝑛 𝑜𝑓 𝑥𝑖

2.​ Systematic errors

a.​ Potential Vibration of the table during the testing procedure will cause the mass
of the bottom to swing, which can provide imprecise results.
b.​ Each stretch of the rubber band may vary slightly (due to rubber aging, different
stretching rates, and temperature effects)
c.​ Assumption: the force is distributed evenly throughout the string
i.​ If the force is not distributed evenly throughout the rope, we will have to
split the entire rope into an infinite amount of small chucks of rope and
then look at the tension of each small chuck of rope. At the end, using
definite integration (meaning limit as the number of small chucks of rope
goes to ∞) and the function of tension, we are able to know the tension
function at every point along the rope.
𝑦𝑓

𝑇(𝑦) = ∫ 𝑓(𝑦)𝑑𝑦, 𝑤ℎ𝑒𝑟𝑒 𝑦 𝑖𝑠 𝑡ℎ𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑡𝑟𝑖𝑛𝑔 𝑡ℎ𝑎𝑡 𝑦𝑜𝑢 𝑤𝑎𝑛𝑡 𝑡ℎ𝑒 𝑡𝑒𝑛𝑠𝑖𝑜𝑛 𝑎𝑡
𝑦1

𝑇(𝑦) 𝑖𝑠 𝑡ℎ𝑒 𝑎𝑛𝑡𝑖𝑑𝑒𝑟𝑖𝑣𝑎𝑡𝑖𝑣𝑒 𝑜𝑓 𝑓(𝑦)

3.​ Reading errors

a.​ Resulted from occasional reading or recording mistakes
i.​ The limitations of human reaction while tracking the RPM in the recording
ii.​ The limitations or estimations of the ruler’s measurement while measuring
the length of elongation
iii.​ Rounding differences during any calculation performed during the
experiment.

Overall, the biggest error that appears in the experiment is human reaction time of tracking the
rpm and the potential deformation of the rubber band material. However, the percent error is
relatively small, meaning this dynamometer has achieved the expected relationship between
Torque and RPM.
Figure 4: Scatter plot graph showing Torque vs RPM (Experimental Data vs Python Simulation
Data using data of the torque done by TT motor from Google)