Team 30: Electronic Actuation of a

Motorcycle Clutch
ME 450 F15, Professor Ni
Calvin Chiu, Peter Karkos, Shashwati Haldar, Chi Qiu
12/14/15

TABLE OF CONTENTS
1.0
2.0
3.0
4.0

5.0
6.0
7.0
8.0

9.0
10.0
11.0

12.0
13.0

14.0

15.0

EXECUTIVE SUMMARY ...................................................................... ....................... 4

PROBLEM STATEMENT .............................................................................................. 5
SPONSOR BACKGROUND ........................................................................................... 5
TECHNICAL DETAILS FROM LITERATURE ......................................................... 5
4.1.
Launch Control Theory .......................................................................................... 5
4.2.
Tire Mechanics ...................................................................................................... 6
4.3.
Clutch Mechanics .................................................................................................. 7
4.4.
Powertrain ............................................................................................................. 7
BENCHMARKING .......................................................................................................... 7
CURRENT SOLUTION ................................................................................................... 8
USER REQUIREMENTS AND ENGINEERING SPECIFICATIONS ...................... 9
USER REQUIREMENTS EXTENDED RATIONALE .............................................. 11
8.1.
Lightweight Analysis ............................................................................................ 11
8.2.
Adjustment of Clutch Within .02 Seconds ............................................................ 11
QUALITY FUNCTION DEPLOYMENT (QFD) ........................................................ 12
CONCEPT GENERATION ........................................................................................... 13
CONCEPT SELECTION ............................................................................................... 14
11.1.
Power Source ....................................................................................................... 14
11.2.
Powertrain ........................................................................................................... 14
11.3.
Drivetrain............................................................................................................. 14
11.4.
Sensors ................................................................................................................. 14
11.5.
Receive/Transmit Signal ...................................................................................... 14
11.6.
User Input ............................................................................................................ 14
11.7.
Hand Drawings .................................................................................................... 15
11.8.
Final Concept ...................................................................................................... 15
CONCEPT DESCRIPTION .......................................................................................... 17
KEY DESIGN DRIVERS AND ENGINEERING ANALYSIS .................................. 19
13.1.
Torque Generation ............................................................................................... 19
13.2.
Clutch Position Accuracy .................................................................................... 21
13.3.
ECU/Motor Reaction Speed ................................................................................. 22
13.4.
Reliability/Robustness While Remaining Lightweight ......................................... 23
13.5.
System-Level FEA ................................................................................................ 23
13.6.
Ease of Disassembly ............................................................................................ 27
ELECTRONICS AND CONTROLS ............................................................................ 28
14.1.
Electrical Wiring.................................................................................................. 28
14.2.
Control Architecture ............................................................................................ 29
FMEA/RISK ANALYSIS ............................................................................................... 33
15.1.
FMEA ................................................................................................................... 33
15.2.
Risk Analysis ........................................................................................................ 34
2

16.0
17.0
18.0
19.0

20.0
21.0
22.0

23.0
24.0
25.0
26.0
27.0

MANUFACTURING PLANS AND DRAWINGS ....................................................... 34

ASSEMBLY PLAN ......................................................................................................... 35
VALIDATION PLAN ..................................................................................................... 36
VALIDATION RESULTS ............................................................................................. 36
19.1.
Mechanics ............................................................................................................ 36
19.2.
Electronics ........................................................................................................... 38
19.3.
Controls ............................................................................................................... 38
PROJECT DEVELOPMENTS ..................................................................................... 39
ENGINEERING CHANGE NOTIFICATION ............................................................ 39
DISCUSSION .................................................................................................................. 41
22.1.
Design Critique .................................................................................................... 41
22.2.
Future Work ........................................................................................................ 42
AUTHORS ....................................................................................................................... 42
BIBLIOGRAPHY ........................................................................................................... 43
APPENDIX A: MANUFACTURING ........................................................................... 45
APPENDIX B: CONCEPT GENERATION AND SELECTION .............................. 51
APPENDIX C: VALIDATION PLAN .......................................................................... 54

1.0 EXECUTIVE SUMMARY

The problem that the Michigan Formula SAE Racing Team would like solved is how to reduce tire slip
off the line in the straight-line acceleration event. Solving this problem will allow the car to accelerate
quicker, resulting in the team scoring more points during the competition. Existing solutions include ABS
traction control, engine control with fuel injection cuts, transmission gearing, and launch control through
a Rev Limiter. However, as per the sponsors request, the solution will only be approached through the
implementation of clutch-control, which is a form of launch control using a closed-loop system with
wheel speed feedback. The main user requirements of the project include generating at least 10 ft-lbs of
torque, avoiding bending moment damage with a safety factor of 1.2, disassembling the system from the
vehicle within 20 minutes, and creating a high speed system that can adjust the clutch within .02 seconds.
The overall goal is to beat the fastest acceleration time achieved by the 2015 MRacing vehicle by
approximately 0.2 seconds.
Developing a functional decomposition was the first step taken in creating a final design concept, which
was used to generate concepts for each of the six categories including power source, powertrain,
drivetrain, sensors, controller, and user input. The best design for each category was chosen based on the
score generated by a weighted selection matrix, where the characteristics used to determine a concepts
score was based on the user requirements, engineering specifications and general engineering judgement.
The winning concepts were a 12V battery for power source, a motor for powertrain, gears for drivetrain,
hall effect sensors for sensors, the Pi Innovo for controller, and a switch for user input. Based off these
winning concepts, the AmpFlow A28-150 High Performance motor along with the MAE3 Absolute
magnetic Encoder Kit were chosen for the final design that would be put on the vehicle with a motor
mount system.
The parts that needed to be manufactured to assemble the system included the motor mount, encoder
mount, the gusset, and truss. When finite element analysis (FEA) was completed on the motor mount
system, it was determined that there could be significant deflection, resulting in the manufacturing of
gussets and trusses to reduce this deflection. Manufacturing was completed using milling, turning, water
jetting, and sawing on relatively cheap material including steel and aluminum resulting in the whole
system being subjected to a relatively small amount of tolerance stacks. In addition to manufacturing,
others steps at this time including developing a circuit and control system.
A proof-of-concept electrical circuit that is connected to a small DC motor has been built that allows the
system to drive current from the power source and protect the motor from backwards flowing current by
implementing an opto-isolator. The controls model in Simulink has also been created where mock signals
were run through the code on the computer to determine the programs validity. For torque validation, the
motor was connected to the power source to affirm that the clutch lever can be easily actuated through its
full range of motion, for stiffness validation, the truss was welded onto the mounting frame resulting in
significant decrease of deflection, and for assembly time validation, the disassembly/assembly time was
completed and recorded to be two minutes. The speed of the system could not be validated as there were
issues getting the encoder to work with the Pi Innovo. The next steps for this project would be to create a
more robust circuit involving thicker wiring for AmpFlow motor, determining coefficient for the PID
controller, and testing vehicle lap times with the system integrated into the car.
4

2.0 PROBLEM STATEMENT

The Michigan Formula SAE Racing Team has presented a unique project dealing with controlling a
motorcycle clutch on a Formula SAE (FSAE) car to reduce tire slip. The overall goal is to develop a
control system and install a clutch-actuating mechanism that will allow the car to accelerate quicker,
ultimately scoring more points in the competition. According to the sponsor, wheel slip is the most
limiting factor in the acceleration event, and if a control system were to effectively reduce tire slip then
the team would see a decrease the lap time by up to 7% [1]. MRacing engineers have researched and/or
attempted simpler/alternate solutions to remedy the wheel slip issue, but ultimately came to the
conclusion that clutch-control would be the best means to reduce tire slip. A solution for controlling
wheel slip is through a Rev Limiter, which is a form of open-loop launch control where the engines
RPM is limited at the start of the launch. Another solution is through the cars traction control, where the
Engine Control Unit (ECU) limits the spark and fuel injections on all cylinders. MRacing engineers have
proved the spark/fuel limiting method to be inefficient and worse than a Rev Limiter launch control due
to the engines delayed response time to wheel slip. After documenting and studying MRacings past
research, it was decided that the overall objective will entail executing this clutch-control method.
3.0 SPONSOR BACKGROUND
The MRacing team is a part of the Formula SAE collegiate series, an engineering competition that
designs, builds, tests, and competes an open-wheeled 400cc formula-styled race car each year. Michigan
engineering students are responsible for designing 100% of the car, and are hopeful for a top 3 finish in
each of the competitions they will be competing in during the upcoming year. With the help of this launch
control system, the car will gain approximately 10 more points in the acceleration event.
4.0 TECHNICAL DETAILS FROM LITERATURE
Many of the project requirements translate into problems that need to researched extensively about.
Research sources include SAE journals, books, Google searches, and discussions with MRacing
engineers. The first step will be to investigate system components such as the clutch, powertrain, and tires
of an FSAE vehicle.
4.1 Launch Control Theory
Launch control is a term that is commonly used in the racing industry, which refers to any electrical or
mechanical system that is used to accelerate a car in a straight line with or without driver assistance.
Figure 1 below shows how launch control logic works; absent from it are technical details about car
components. The clutch can be regulated with a PID controller, for which the gains can be calculated
using theoretical methods or through physical testing [2].

Figure 1: Launch control logic utilizes a network of if statements to determine which operation to
perform next. [6]
4.2 Tire Mechanics
According to MRacing Suspension Lead, Jason Ye, The most limiting aspect of the FSAE vehicle is the
tires ability to output enough force through means of friction [1]. During a launch, the engines power
exerts more torque than the tires can physically handle. This is apparent during the acceleration event
when the rear tires spin and slip against the ground. The tire force is controlled through an independent
variable called the slip ratio (SR), which is simply a ratio between the speed of the driving wheels and the
speed of the free rolling wheels [3]. On the car, this is measured through the use of hall effect sensors that
are mounted in order to record the speed of the front and rear wheels [4]. Figure 2 below shows the
relationship between tire friction and SR [5]. This led to the conclusion that the control system must be
designed to adjust the tire slip such that the maximum force observed at the tires can be maintained.

Figure 2: This graph of friction coefficient of the tires versus slip ratio (SR) shows how differing
environmental conditions can have an impact on SR. [5]

4.3 Clutch Mechanics

The way a clutch works is fairly simple. The goal of a clutch is to separate the powertrain (engine) and
drivetrain (transmission) through a series of simple friction plates and springs known as the clutch pack
[6]. When the clutch is disengaged, the springs are compressed and the clutch plates are allowed to
separate, thus resulting in no torque being transmitted from the engine to the transmission. When the
clutch is engaged, the springs are released, allowing 100% of the engine torque to be transferred to the
wheels, not accounting for efficiency losses. Controlling the exact engagement/disengagement of the
clutch will allow for a specific amount of torque to be transmitted [7, 8].
4.4 Powertrain
The clutch-control design has to be integrated into the MR15 vehicle, which houses a Honda CBR600RR
4-cylinder, 4 stroke, naturally aspirated gasoline engine. The transmission of the vehicle, which includes
the clutch, is integrated into the engine block. The basics of an engine, such as components and basic
schematics, has been researched [9]. The inner workings of the engine has been understood by physically
handling and tinkering with the internals of the motorcycle engine while working in the MRacing dyno
room. Scott Trahan, MRacings Powertrain lead, was able to provide a map of Torque and Power vs.
Engine RPM, as shown in Figure 3 [8].

Figure 3: The torque and power curves for the Honda CBR600RR motorcycle engine both show that
there is an optimal rpm at which maximum power will be developed. [8]
5.0 BENCHMARKING
The concept that is being attempted to implement exists in a very narrow market and applies only to a
small number of race car series. Due to secrecy within the racing industry, it is extremely difficult to
obtain an in-depth understanding of a product without actually purchasing or working directly with the
product. However, it has been discovered that Formula 1 has been implementing launch control systems
through clutch-control for almost 15 years [10]. In the automotive industry, products similar to the one
implemented in this project will be of concern only to a limited degree. An exact patent application of this
7

project has not been found, but instead listed below are several related products that deal with clutchcontrol.
Performance enhancing and starting process accelerating method for vehicle with hydrodynamic torque
converter [11]: This patent outlines how a motorized vehicle will begin moving from standstill using
automatic clutch-control in an automatic transmission. This concept is applied to the majority of vehicles
on the road today, apart from anything with a manual transmission. The patent also has a section relating
to how a race car will start, stating, Racing vehicles with automatic transmissions utilize a
hydrodynamic torque converter as a means and method for controlling the start of the vehicle. The
patent explains that this method allows the engine to rev at high speeds, allowing for quick accelerations.
It then goes on to say that this ...function is also referred to as a "launch control" and usually includes a
manually operable switch ("launch button") to enable or disable this Quick Start function of the vehicle
[11]. Due to the use of manual transmission, this project does not infringe upon this patent. However, the
patent gives a general definition for launch control, which is valuable to consider.
Control method and controller for a motor vehicle drive train [12]: This patent relates to torque control
of an automatic transmission by raising engine power and reducing output torque. Its goal is to initiate a
launch with an automatic transmission through an electronic controller. Due to the use of manual
transmission, this project does not infringe upon this patent.
Method and apparatus for operating a clutch in an automated mechanical transmission [13]: This patent
is about an algorithm that calculates the kiss point of the clutch. In response to this mathematical point,
the system will determine the engine power versus clutch position and feed a controller information about
where it should be in order to launch the vehicle. It will continuously compare accelerator position to
clutch position to recalculate the kiss point. Due to the use of manual transmission, this project does not
infringe upon this patent.
6.0 CURRENT SOLUTION
The current solution to prevent tire slip during launch control is by manually holding and dropping the
clutch lever, which is connected to the clutch shaft, to engage/disengage the clutch plates. The problem
with this is that it is extremely difficult for the driver to decide which position he/she should drop the
clutch lever such that the engine will output maximum power without producing wheel slip [14]. Another
solution is through traction control, which is where fuel is withheld from the cylinders. The problem with
this method is that there are only four options for fuel cutting that can reduce the power output of the
engine, thus reducing the acceleration of the wheels/car. Conversely, the clutch-control method has
infinite adjustability [15, 16]. A pneumatic clutch is also one of the existing solutions on the market
designed to remedy wheel slip. A pneumatic clutch can provide a smooth start-up from a standstill and
can also actuate very quickly depending on your engineering needs. This can be done by regulating the air
pressure being applied to the clutch [17]. However, a pneumatic clutch costs around $2000 [18], which
exceeds the budget set by the MRacing team. Also, a pneumatic clutch is difficult to integrate with the
current MRacing car.
8

7.0 USER REQUIREMENTS AND ENGINEERING SPECIFICATIONS

All user requirements and engineering specifications have been determined through discussions with the
sponsor, physical testing on the vehicle and its parts, and internal discussions within the team. Table 1
below summarizes the user requirements and engineering specifications.
Table 1: User requirements and engineering specifications spreadsheet with their relative priority, source
and rationale for each item.
SYSTEMS

USER
ENGINEERING
PRIORITY
REQUIREMENTS SPECIFICATIONS (1 = highest) SOURCE RATIONALE

Mechanical System
General
Large Torque
requirements Generation

Volume Limitation

Light Weight

Mechanical
mounting

Avoid Bending
Moment Damage

Assembly Simplicity

Drivetrain*

High system speed

No backlash

1 Test

Determined via
physical testing of
the clutch torque.

3 Test

Determined via
space
measurements of
car.

< 6 lbs

4 Sponsor

General guideline
given by sponsor,
determined via
discussions with
engineers.

FEA safety factor > 1.2

Team
1 (Internal)

Common practice
within the team
[3]

3 Sponsor

Common practice
for "Pull-off"
systems on the
team [7].

1 Sponsor

Minimum ratio of
1:1 based on
engine spec, max
ratio of 3 to not
exceed volume
limit.

Team
4 (Internal)

Decided to
maintain 1% of
the total torque as
pre-load.

> 10 ft*lb torque

< 100 in^3

< 20 min

Reduction, between 1
and 3

pre-load > 0.5 ft*lb

FEA safety factor > 1.2

Team
2 (Internal)

Common practice
within the team
[8]

Safety

0 current during
working sessions

Team
1 (Internal)

A team voted on
rule for the
project.

Reliability

> 500 launches (1

season)

3 Sponsor

To be costeffective, it must
last 1 season.

kill switch = 1

FSAE
1 Rules

A specific rulenot determined by

us [19].

1 Sponsor

The simplest
method we found
[20].

Team
2 (Internal)

Simplifies
mechanical
design, eliminates
future issues.

3 Test

Specification
given to us from
vehicle data.

2 Sponsor

Maximum
temperature the
outer clutch cover
will achieve,
sometimes it rains.

5 Sponsor

Reliability issues
within the team in
the past have
occurred [7].

5 Sponsor

Ease of integration
into the car is
preferable and
eliminates weight
[7].

Reliable and robust

Electronic/Control System

Wiring

Rules complient
Driver must
activate/deactivate

Must use vehicle

sensors

Powertrain
control

Sensitivity

Enviromental
Independance

Avoid high pressure

Power Source systems

Use car's power

sources

on/off switch = 1

wheel speed sensor = 4

reaction time < .02 s

temperature < 130F

Zero pneumatic or
hydraulic sources

12V battery or 75HP

(mechanical)

10

8.0 USER REQUIREMENTS EXTENDED RATIONALE

Several user requirements need further rationale to be fully explained.
8.1 Light Weight Analysis: The overall concept behind the MRacing vehicles is to be light weight,
considering the fact that weight is the most sensitive parameter on the car. This product, by rules, must
not be removed from MRacing car during competition, unless it is causing failure. This causes an
additional amount of dead-weight on the car during other events, such as the endurance race. To narrow
down the maximum allowable weight, the team was able to compare lap-time simulations and tested track
lap-times. Simulation vehicle weight was iterated until the change in lap time was greater than the
standard deviation of a real life lap time. Standard deviations exist in reality due to driver error. The
results of the simulation and track tests are shown below in Table 2 and Table 3.
Table 2: Simulation lap times using VI-CarRealTime allows us to quantify weight sensitivity
Simulation Parameters and Results
Weight (lbm)

Weight (lbm)

CG_X (in)

CG_Z (in)

Lap Time (s)

Baseline

591

33

11.8

+0.00

+1

592

33

11.8

+.031

+2

593

33

11.8

+.065

+3

594

33

11.8

+.089

+4

595

33

11.8

+0.129

+5

596

33

11.8

+.160

Table 3: 2015 endurance race results give the standard deviation for the lap times.
2015 Competition Endurance Race Parameters and Results
Weight (lbm)

CG_X (in)

CG_Z (in)

591

33

11.8

Avg Lap Time (s)

61.2

Lap Time Stddev (s)

.155

A resulting added mass of 5 lbm will be within the lap time standard deviation of an endurance race, not
including center of gravity (CG) effects.
8.2 Adjustment of Clutch Within .02 Seconds: The rationale behind this is how quickly the wheel speed
sensor can output a new signal. Table 4 below show the results of a test, logging data at 100 Hz (the
maximum allowable frequency by the Daq system used on the vehicle).

11

Table 4: An initial launch test with the vehicle shows the wheel speeds can read at ~50Hz (0.02 s)
FL Wheel Speed
(m/s)

Time (s)

FR Wheel Speed
(m/s)

RL Wheel Speed
(m/s)

RR Wheel Speed
(m/s)

0.01

24.4609

23.9297

48.5

48.4766

0.02

24.8203

24.3047

50.375

48.3359

0.03

25.9609

25.6172

52.5547

54.4844

0.04

26.2109

25.7969

53.6641

54.8516

0.05

28.9063

28.2656

59.8359

58.8281

0.06

29.0729

28.474

60.6901

58.6042

0.07

30.4453

29.9141

61.8828

64.6094

0.08

30.7266

30.1875

62.3438

64.5391

0.09

32.4531

31.9063

64.8516

67.0234

0.1

32.7266

32.1094

64.8438

67.3672

From this data set, it can be seen that the wheel speed sensor outputs a new signal every 0.02 seconds,
faster than the Daq system can log at. The clutch must be moved to its new position, faster than 0.02
seconds, before the next wheel speed signal is read by the controller.
9.0 QUALITY FUNCTION DEPLOYMENT (QFD)
Table 5 shows the QFD that translates the user requirements into engineering parameters that can be
easily understood for design and manufacturing purposes.
Table 5: Quality Function Deployment that translated user requirements into engineering specifications

12

10.0 CONCEPT GENERATION

To accomplish the goal of actuating the clutch of a motorcycle engine, the system was divided into a
series of smaller subsystems through a functional decomposition. The functional decomposition tree,
outlining the process of preventing wheel slip via clutch-control, is shown in Figure 4 below.

Figure 4: Functional Decomposition tree showing the breakdown process for controlling wheel slip.
The functional decomposition was then further divided into categories, which is where the concepts were
organized. The descriptions for each category are listed below.
Power Source: Includes any ideas for energy storage on the car.
Powertrain: Includes any ideas for converting the stored energy into mechanical power.
Drivetrain: Includes any ideas for transmitting the mechanical power to the clutch.
Sensors: Includes mechanisms that will detect a desired car variable in real-time.
Controllers: Includes anything that can process a sensor signal and control the powertrain.
User Input: Includes any ideas for how to start/stop the system.
Major categories include a Power Source, Powertrain system, Drivetrain system, Sensors, Controller, and
User Input mechanism. Each team member then came up with concepts for each category. Table 10 in
Appendix B outlines each generated concept.
11.0 CONCEPT SELECTION
After generating concepts, the team created weighted selection matrices for each category. Every matrix
pitted each concept against a list of requirements. Each requirement was assigned a score on a scale of 110 (in order of increasing importance) based off of our discretion of importance. Then, each concept was
scored on a scale of 0-10 based on how well it matched/met a requirement, as well as how feasible it was
from the perspective of our team members. Some of the requirements that our concepts were compared
against include Reliability, Cost, Weight (mass), and Ease of Use. As a general rule, any concept
that scored a 0 or a 1 for a category would not qualify as a viable concept, regardless of its total score.
Appendix B shows the tables showing the scoring system used and results for each category.

13

11.1 Power Source: The power source category contains ideas for how energy should be stored for the
system. From past experiences, it was determined that the best ideas were the 12V battery, pressurized air,
and gasoline. Pressurized air was ruled out due to a score of 1 for reliability, gasoline was ruled out due to
a score of 1 for ease of control, and propane was ruled out due to a score of 1 for feasibility. Although the
vehicle runs on racing fuel, gasoline scored very low for feasibility because another combustion engine
would be needed to provide the power. Another notable concept was driver effort, but this also failed due
to a score of 1 for strength. The winning concept for power source was the 12 V battery due to its high
score for feasibility and ease of control. One disadvantage of the 12 V battery is that it has a short
lifespan, but this will be accounted for this by swapping batteries whenever needed.
11.12 Powertrain: The powertrain category contains ideas for transmitting energy from our power source
to our drivetrain. The best ideas include the vehicles 75 hp engine, a motor, and the driver. Both the
engine and driver were ruled out immediately due to a score of 1 for ease of control. The winning
powertrain concept was the motor due to its ease of control and its reliability. One downside to the motor
is that it produces relatively low torque in comparison to the other generated concepts.
11.3 Drivetrain: The drivetrain category contains ideas for delivering the energy from our powertrain to
the clutch. Noteable concepts include a chain drive, gears, and a steel rod acting as a moment arm. The
steel rod scored highly overall, but was ruled out because this would necessitate the need for a linear
actuator - a powertrain device that is not feasible according to the teams standards. The top concepts
were a chain drive and gears, but gears were chosen because it is believed that the gear is more feasible
for packaging reasons and meeting the volume constraints. However, it is important to be cautious of how
much is spent on gears since they are relatively costly.
11.4 Sensors: The sensors category contains ideas for how to detect a desired car variable in real-time.
Major ideas include a Hall effect or laser speed sensor for wheel speeds, an accelerometer, and a
potentiometer. The Hall effect sensor scored the highest because it has already been implemented on the
vehicle, thus making it easy to use, cost effective, and efficient with volume. One minor disadvantage to
the Hall effect sensor is its poor accuracy compared to the laser speed sensor. However, laser speed
sensors can cost around $300-$400, which exceeds the budget entirely.
11.5 Receive/Transmit Signal: The receive/transmit category contains a list of controllers that would
bridge the gap between the sensors and the powertrain. Major contenders include a Bosch MS4.4, a Pi
Innovo M220, and an Arduino board. The Bosch MS4.4 was ruled out due to its difficulty of use, despite
the fact that it has already been implemented into the vehicle. The Pi Innovo and the Arduino both ranked
very highly in in ease of use, processing speed, and feasibility. However, the Pi Innovo was chosen
because it is compatible with Simulink, which is something that the team members are familiar with
using.
11.6 User Input: The user input category contains ideas for how a driver may activate and/or deactivate
the system. The simplest, most practical and cost-effective option is a toggle switch mounted into the
cockpit. Since the user can easily distinguish whether the system is on or off with a toggle switch -

14

something that cannot be easily accomplished with a button- it was ranked slightly higher in ease of use,
thus making the toggle switch the chosen concept.
11.7 Hand Drawings
Concept drawings of several ideas were generated to get a better visual representation of how they would
operate within the system. Figure 5 below shows some of the drawings that were produced.

Figure 5: Drawings of concept ideas such as chain-sprocket, solar energy, arduino, and a laser speed
sensor
11.8 Final Concept
The final concept will involve a 12 V battery powering a motor, as well as a gear drivetrain delivering the
power from the motor to the clutch lever arm. The controller will be a Pi Innovo M220, which will read
input signals from several Hall effect sensors mounted to measure the front and rear wheel speeds. To
meet the user requirements listed in the previous section, the whole system will be kept external and
easily accessible on the car. It was initially thought that designing a mount would prove to be a structural
issue. However, when the mock-up was brought up to the car and realistic dimensions were assessed, it
was realized that there would be quite a bit of freedom with system placement. Figure 6 below shows a
basic design scheme while Figure 7 shows the foam mock-up of the final concept. Figure 8 shows a
picture of the Pi Innovo.
15

Figure 6: A free-hand drawing of our final concept.

Figure 7: A foam mock-up of the final concept is shown below.

Figure 8: A mock-up of a Pi Innovo M220 controller was not produced because the sponsor was able to
lend a spare model that they had on hand. Below is a picture of the controller that will be used.

16

12.0 CONCEPT DESCRIPTION

The general design concept that was chosen previously consisted of using a 12 V battery to power a
motor, which would then deliver its power to the clutch lever through a one stage gear reduction. The
motor would also be connected to the Pi Innovo that would in turn be connected to the toggle switch and
the hall effect sensors. Upon further engineering analysis, it was determined that the AmpFlow A28-150
High Performance motor will be able to produce a torque of greater than 10 ft-lbs and actuate the clutch
within 0.02 seconds (both with a 1:3 reduction ratio). It was also determined that the MAE3 Absolute
Magnetic Encoder Kit would allow for an effective resolution of 0.012 (with a 1:3 reduction ratio), thus
providing a very high level of precision. Using this knowledge, a computer aided design (CAD) model
was created for the complete motor mount system as seen below in Figure 9 and Figure 10. A close-up
picture of the motor mount frame is shown below in Figure 11, while Figure 12 shows the spacing
between the motor and the gears.

Figure 9: The CAD model of the motor mount system with all parts labelled. All the part names can be
referred to this figure.

17

Figure 10: The CAD model of the motor mount system with the encoder at the top of the motor and the
gears below the motor. The mount can also be seen with its two bolt slots.

Figure 11: Close-up of motor mount structure

Figure 12: Close-up of the spacing between the motor shaft and its gear.
18

13.0 KEY DESIGN DRIVERS AND ENGINEERING ANALYSIS

In order to effectively accomplish the goal through implementing the final concept design, the following
key design drivers were considered.
13.1 Torque Generation
The overall system must be able to produce enough torque in order to actuate the clutch lever on the
motorcycle engine, which was measured to be approximately 10 ft-lbs at the lever. The torque production
of the motor does not need to be 10 ft-lbs since a drivetrain consisting of a gear reduction designed to
reduce the powertrain torque demand will be used. However, what is more challenging is that the clutch
will need to move at very high speeds, thus requiring more power.
Implementing a motor into the system as a source of torque generation is a very effective means of
delivering enough torque through the drivetrain in order to actuate the clutch. However, it is important
that detailed calculations are performed to verify that the requirements of the Torque Generation design
driver are met.
Being able to produce 10 ft-lbs of torque is key to the success of the project because outputting any less
would result in a failure to actuate the clutch. In order to assure at least 10 ft-lbs of torque can be
produced, a gear reduction will be implemented so that torque delivered through the output shaft can be
increased. Although this may make it seem as if choosing the smallest reduction and overshooting the
amount of torque generated would be the best option, the impact that a small reduction would have on the
system could be very detrimental towards the project. Using a gear reduction decreases the actuation
speed by a factor equivalent to the gear reduction. As a result, it is important to make sure that the
smallest reduction possible is chosen that also meets the requirement of a 0.02 second actuation time. In
order to properly analyze the system, in-depth calculations must be performed to evaluate the rotational
dynamics of the motor to ensure that it fits the projects needs. This theoretical modeling method is the
most appropriate because it is the closest that one can get to evaluating whether a motor will work or fail
prior to actually testing the motor. The layout of a dynamic system identical to the one this project uses
can be seen in Figure 13 below.

Figure 13: A motor driving a load through a one stage reduction. [21]
The previous figure depicts a scenario where a motor is driving a motor gear (gear 1), and that gear is
driving a load gear (gear 2), which in turn is driving the load. The governing equation used to calculate
the total inertia experienced by the motor is Jtotal=Jload/N2+Jmotor, where Jload is the inertial load
contribution, Jmotor is the motor inertia, and N is the gear ratio.
19

The first step taken before beginning the analysis was to empirically test the amount of force that was
required to actuate the clutch lever. This resulted in a value of Flever=120 [lbs]. Then the distance between
the axis of rotation of the clutch lever and the point on the lever where the safety wire was attached was
measured to be D=1.2 [in]=0.1 [ft]. The torque was then calculated by T=Flever*D=10 [ft*lbs]=13.6
[N*m]. It was then estimated that the maximum range of motion of the clutch lever would be
=5=0.0873 [rad]. Next,the angular speed was determined by calculating =/t=4.365 [rad/s]; note
that this is a steady state speed and is the best approximation which can be achieved. Similarly, the
angular acceleration was determined by =/t=218.25 [rad/s2]. After all of the preliminary values were
obtained, the next step was determining which values were needed to calculate Jtotal. The inertia of the
motor was determined by Jmotor=T/=0.061 [N*m/s2]. The inertia of the load was calculated by
Jload=(W*r2)/g=1.061 [N*m/s2], where W is the force required to actuate the clutch lever in Newtons and
r is the radius of the load gear in meters. Since a 1:3 reduction ratio will be implemented, the gear ratio is
3:1, thus meaning that N=3. The reason this reduction ratio was chosen was due to the fact that any larger
of a reduction ratio would result in a two stage reduction which is not feasible due to volume constraints.
Finally, Jtotal=Jload/N2+Jmotor=0.180 [N*m/s2].
The value for Jtotal means that the total inertia experienced at the motor (due to the loads of the system) is
0.180 N*m/s2. This is important because, just like how the inertia in a braking car wants to keep one
moving forward, the inertia of the motor keeps the spindle rotating from pole to pole. If the motor inertia
is greater than the total load inertia, then the motor should work just fine for the application. However, if
the motor inertia is less than the total load inertia, then the motor would either have the tendency to bog
down and rotate unsmoothly or the motor just would not work for the application at all.
The motor that was chosen was the AmpFlow A28-150 High Performance motor, as shown below in
Figure 14. This motor fits the engineering specifications exceptionally well for several reasons: a) It can
generate 5.13 ft-lbs of torque (15.39 ft-lbs with the 1:3 reduction), b) It has a maximum speed of 3000
rpm when run at 12 V (1000 rpm with the reduction), which translates into very quick accelerations when
coupled with the high torque output, and c) The system inertia that was just calculated does not seem like
it would be too much for the A28-150 to overcome (the company was contacted to get specifications but
their technicians did not have the information desired). After going over the analysis and ensuring that
none of the technical issues were overlooked, the team is confident that they will not need to perform any
further analysis.

Figure 14: Picture of the motor showing its front mounting plate, spindle, and housing. [23]
20

13.2 Clutch Position Accuracy

The amount of force between the friction plates is determined by the clutch position, which ultimately
determines the amount of power transmitted to the wheels. The accuracy of our system is crucial to a
quicker acceleration time than recorded without an automated clutch system. If the engagement of the
clutch is overshot, then the wheels will slip too much, but if the engagement is undershot then the car will
accelerate too slowly. An analytical model will be developed to determine how much the friction force
will be needed under different circumstances (environmental conditions, quality of the track, etc).
Extensive testing will also need to be conducted on our system to see if the clutch is being moved to the
desired positions. This will be a challenge as the initial parameters values used may not work well when
tests are performed on the track.
It is believed that the resolution of the encoder that is connected to the motor and the accuracy of the hall
effect sensor will affect the accuracy of the clutch position while it is being actuated. The encoder that
was chosen was the US Digital MAE3 Absolute Magnetic Encoder Kit , as shown in Figure 15 below.

Figure 15: The encoder that will be mounted on the rear shaft of the motor to track the position of the
motor. [22]
The positions per revolution (PPR) of the encoder peaks at 4096, which can be converted into degrees by
calculating R=360/PPR=0.0879, where R is the resolution of the encoder. Also, the 3:1 gear ratio makes
the output gear spin 3 times slower than the input gear, resulting in an effective resolution of 0.029. This
means that there are nearly 170 options for the clutch position over the assumed 5 of travel, making this
a very precise operation.
It is very difficult to quantify the impact that the hall-effect sensors have on the accuracy of the clutch
position. For the hall-effect sensor that is used to record the wheel speed, mounted next to the spindle of
the wheels, it is especially challenging to quantify its effect on the accuracy of the clutch position.
However, specification sheets for the sensors on the car show that their accuracy error is 0.001 m/s,
which would not significantly affect the accuracy of the clutch position. This specification comes directly
from the ability of the Bosch MS4.4 controller, which can output signals at a resolution of 0.001 [7].
Something to note for this design driver is that it is extremely difficult to perform any major analyses
prior to obtaining the encoder, assembling the system, and testing it. The only theoretical modeling that
21

needed to be done was to convert the CPR of the encoder into degrees and check to see if the effective
resolution was within the acceptable range. Although not written down, the team internally decided that
an encoder capable of tracking the position of a motor to 100 positions within the desired range of motion
would be chosen. After determining the estimated range of motion to be 5, it was understood that the
encoder resolution would have to be at or exceed (be lower than) 0.05. Therefore, the MAE3 encoder
meets the teams requirements and will be accurate enough to track the motor position according to the
engineering analysis.
13.3 ECU/Motor Reaction Speed
The time that it takes the ECU and powertrain to react to an input wheel speed is vitally crucial to the
success of the project. This is because a delayed reaction time may do nothing to the acceleration time or
even hurt it. This issue can be overcome as long as the sensor and controller selection is adjusted as the
project moves forward.
According to an interview with a design engineer on the MRacing team, as well as online research, the
reaction times for the ECU (Pi-innovo) and the sensor are miniscule compared to the reaction time of a
motor. This results in only considering the reaction time of the motor as the time it takes for the system to
respond to a change in tire slip. However, the reaction time for the motor is quite unpredictable because
the maximum travel range of motion that the motor is going to actuate within is around 5 degrees, which
doesnt allow the motor to accelerate to a steady-state angular velocity. If it is assumed that the motor
reaches the steady-state angular velocity instantly, then the minimum time needed to travel 5 degrees with
the A28-150 motor can be determined by the equation t=5*60(second/min)/ 360(degrees)/speed(rpm).
The speed of the motor can be obtained from Figure 16 below.

Figure 16: With the 10 ft-lbs torque requirement and 1:3 gear reduction, the torque the motor outputs
would be 5.39 ft-lbs where the operating speed found under this torque is around 4500 rpm. [23]
22

Therefore, by using the equation mentioned above, the reaction time for the motor is 0.00019 seconds.
However, this ideal reaction time assumes that the motor will reach this speed instantaneously. In
conclusion, the reaction time of the motor is still an unpredictable factor and the only way to determine its
true reaction time is to test the system.
13.4 Reliability/Robustness While Remaining Lightweight
Creating a reliable, robust, and easy to disassemble system is fairly simple in itself. Unfortunately,
including the need for a lightweight system makes those previous goals much harder to achieve. Mass is
the absolute most sensitive parameter on the car; reducing the mass of the car significantly improves the
cars performance in every single dynamic event. Therefore, it is imperative that the mounting structure is
light yet robust through the use of different structural geometries and materials.
The target range of system weight is set to be between 5 and 6 pounds (lbs) due to a simple weight
sensitivity analysis using simulation tools and track data, with help from the sponsor. Over the course of
an endurance event, driver lap time variability was anywhere from 0.25% to 1%. From these statistics, the
mass should not be increased to offset a lap time by more than .25%. The MRacing Suspension Engineer,
Jason Ye [1], was able to run simple full-vehicle lap time simulations, iterating mass until .25% increased
lap time was achieved, resulting in the target weight of 5-6 pounds (see USER REQUIREMENTS
EXTENDED RATIONALE).
The design must also pass the FEA with a safety factor of 1.2. A safety factor of 1.2 in load-driven
analysis was derived from Doctor David G. Ullmans engineering handbook. The handbook states that a
safety factor of 1.2-1.3 should be used if the nature of the load is defined in an average manner, with
overloads of 20-50% [27]. The loads that the system will experience in this project are understood
relatively well but not fully, therefore a safety factor of 1.2 was chosen.
In order to theoretically analyze this design driver, it became necessary to find the maximum possible
displacement of the gears along with determining the stresses experienced by the motor shaft.
13.5 System-Level FEA
Stress and deflection analysis on the system level was conducted to proof our concept and show how to
improve designs. Applied forces at the gear interface will create stress and deflection distributions on our
shaft and mounts. The calculations for the loads, with a safety factor of 1.2, are as follows [26]:
Wt= 1.2 * 60000 * H/(*d*n) [kN]
Wt = tangential force [kN] =.376 kN = 85.4 lbs
H = maximum power [kW] = .25 kW
d = gear pitch diameter [mm] = 25.4 mm
n = driven speed [RPM] = 600 RPM
Wr= 1.2 * Wt * tan [kN]
Wr = radial force [kN] = .097 kN = 22.1 lbs
= pressure angle of the gears [Deg] = 14.5 deg
23

These two forces were used as inputs for the load simulation done in NX Nastran. Figures 17 and 18
showed how the constraints, tangential force, and radial force were assigned on the model. Figures 19,
20, and 21 show how how the design has been modified based on the FEA to get less gear deflection. The
results from the analysis picture in Figure 21 showed that the final maximum displacement of the gears
would be around 0.0133 in which correlates to a yellowish green color, the color of the gears. This value
is well under the height of the gear teeth, which happens to be 0.1348, therefore meaning that the gear
teeth will not slip. Figures 22 and 23 show the FEA results in terms of stress.

Figure 17: The bracket is constrained by bolt connections on the frame of the car.

Figure 18: Radial and tangential forces are applied to the motor gear.

Figure 19: Gear displacement of the first iteration design set a baseline of ~.056 inches of gear
deflection.
24

Figure 20: Gear displacement of second design iteration improves gear deflection to ~.045 inches.

Figure 21: Gear displacement results of the third design iteration reduces gear deflection to ~.013
inches.
The results from the analysis picture in Figures 22, 23 showed that the motor shaft would experience
maximum stresses of around 8300 psi, which is well under the yield stress of steel (36000 psi). The low
stress that the motor shaft will endure means that the motor shaft will not break under stress. Other areas
of relatively high stress are in various regions of geometry changes and where the bolt connections exist.
However, this value is still very low, around 6000 psi, with respect to the yield stress of 6061-T6
Aluminum (40000 psi) [25].

25

Figure 22: Visual of the motor mount that shows the expected stresses on the motor shaft to be about
8300 psi.

Figure 23: A close-up of FEA in terms of stress around where the max stress occurs.
As the results show that the design is feasible, the team is confident in the analysis. FEA on the opposite
gear/shaft combination was not conducted for reasons such as the gear being three times larger, the shaft
being a Honda-made part, and the moment arm from the gear to bearing (located inside the engine) being
very small, as shown below in Figure 24.

26

Figure 24: Clutch gear (left) is very large and attached to a Honda-made shaft.
13.6 Ease of Disassembly
Since the acceleration event is only one part of the Formula SAE competition, being able to remove the
system from the car very quickly can prevent the team from missing any other events should the system
malfunction and become a nuisance. This goal can be achieved by positioning the clutch-actuating system
on the car such that the fasteners keeping the system constrained to the car are easily accessible.
It was verified that the system could be disassembled quickly by creating a CAD model of the motor
mount. As seen from the CAD model in Figure 25 below, the mechanical portion of our system can be
removed from the car by simply removing two bolts from the mount. This will allow for the system to be
removed from the car in under 20 minutes. One lesson that was learned from creating this CAD model is
that the rigidity of the system could be increased by constraining it through more than the two bolt holes
currently in the design, but doing so would add weight to the system and also make the system more
complex to disassemble, thus making it an option that will not be pursued.
This mode of analysis is appropriate because it produces an approximate time that it will take to
disassemble the system with the existing design - less than 5 minutes. This time of disassembly is well
under the 20 minute user requirement. This design is functional because it includes all the functional
components that were deemed necessary for operation, and its volume approximately 36 in3 which is well
under the maximum volume constraint of 100 in3. This results in high confidence in the analysis.

27

Figure 25: Close-up of the CAD model showing the two points at which the motor-system will be attached
to the FSAE vehicle.
14.0 ELECTRONICS AND CONTROLS
The following sections describe the electronics and controls engineering analysis and design portion of
the project.
14.1 Electrical Wiring
The electrical wiring for this project was not an incredibly difficult design, but due to the high amount of
current that the motor will be drawing (~127 Amps, taken from Figure 11), safety precautions had to be
taken to ensure no components, such as the Pi Innovo M220, or users would be harmed. Figure 26 below
shows the most up-to-date wiring schematic.

Figure 26: Wiring diagram consisting of all components related to the project. Note that the controller
the team will be using is a Pi Innovo M220, not an Arduino.
28

The controller that is in the wiring schematic is not the one that will be implemented in the project. The
program, Fritzing, did not have a Pi Innovo M220 in its database, therefore an Arduino Mega ADK was
used as a placeholder. Also, the components to the left of the battery are already integrated into the car
and therefore were not carefully wired. The components to the right of the purple border are directly
associated with the project.
The components that were used to produce this wiring are as follows: 1) One 12V 4.6 Amp hours battery
to provide the power, 2) One 150 Amp fuse to break in case the system somehow draws above 127 Amps,
3) One optoisolator to prevent high currents from back-flowing and destroying the control unit, 4) One
300 ohm resistor to bring down the voltage being applied to the optoisolator, 5) One 10 kohm and One 1
kohm resistor to act as pull-up resistors, 6) One n-channel metal-oxide-semiconductor field-effect
transistor (MOSFET) to act as a signal delivering field-effect transistor (FET), 7) One n-channel
MOSFET to allow the current from the battery to flow to ground, 8) One DC brushed motor, 9) One
rectifier diode to prevent voltage spikes from damaging the motor, 10) One motor encoder to read
absolute motor positions, 11) One toggle switch to turn the system on/off.
The way in which this circuit functions is actually quite simple. The yellow pole of the motor is
constantly at 12 V, but because there is no input signal into the system, the MOSFET on the right
effectively does not have a voltage difference between its gate and its source and thus the circuit is open.
Once the controller sends a signal to the optoisolator, the isolator produces a voltage difference between
the gate and the source of the MOSFET on the left. Afterwards, a signal-level current flows from the gate
to the source, thus providing a voltage difference between the gate and the source of the MOSFET on the
right. Once this happens, the circuit is grounded and there is a voltage difference between the yellow and
green poles of the motor, thus causing the motor to spin. In the event of a voltage spike, the rectifier diode
will prevent the high voltages from damaging the motor.
14.2 Control Architecture
The fundamental goal behind the control architecture will be to limit output torque from the engine to the
wheels through a closed-loop feedback system. The output torque will be limited by comparing rear
wheel torque to generated engine torque to find an output equilibrium torque through modulation of the
clutch. To simplify the schematics, the architecture has been broken up into three separate categories: tire
slip control, wheel torque control, and clutch force control. The map below shows a simplified breakdown
[24].

Figure 27: Control map displaying how wheel speeds are used to control the clutchs position. [24]
29

The first block represents tire slip control where the inputs are the wheel speeds of the four tires while the
output is torque requested. Essentially, an average speed is taken for the front wheels and the rear wheels.
Then a constant of 1 is subtracted from the ratio of the rear wheel average speed to the front wheel
average speed this is called the slip. This slip ratio is used to determine the torque generated at the
wheels by using a look-up table of slip ratio vs wheel torque. Figure 28 below shows the relationship
between the slip ratio and the force on the wheels; the y-axis of this graph will have to multiplied by 3.11
to translate the units from force to torque. The three curves represent three different normal forces on the
tires.

Figure 28: Team-derived Hoosier LC0 tire map showing the relationship between slip ratio and force on
the wheels.
Once the generated wheel torque is found from the look-up table, it is subtracted from the maximum
possible wheel torque to get the torque requested. As seen from the graph in Figure 28 above, the
maximum torque is at slip ratios of about .155, however, if the slip ratio is higher than these values, the
torque requested will have to be negative. Part of this algorithm determines whether the slip ratio is higher
than .155 and then multiplies the torque requested by negative one if that scenario is satisfied. The block
diagram for tire slip control that has been created in Simulink is below in Figure 29.

30

Figure 29: Block diagram of tire slip control in Simulink.

The second block represents wheel torque control. The input of this block is engine speed and wheel
torque. The engine speed is put into a look-up table of engine speed against engine torque which is shown
in Figure 30 below. Next, the ratio of the requested torque to the generated engine torque is calculated this is called the clutch effort. The Simulink block diagram of the wheel torque control is below in Figure
31.

Figure 30: Engine torque map showing the relationship of engine speed (RPM) to engine torque (ft-lbs).

31

Figure 31: Block diagram of wheel torque control in Simulink.

The third block represents the clutch force control. For this block diagram, a look-up table of clutch effort
against motor position must be created, which will be empirically tested once the system has been
assembled on the car. In order to test clutch effort against motor position, the team will produce a spindle
adapter which will mate with the head of a torque wrench. Then, a motor position will be chosen, held
(stalled), and the team members will use a torque wrench to determine how much torque is required to
cause the clutch plates to slip. The team will perform many tests at different motor positions in order to
obtain a graph of clutch effort versus motor position.
For this third block diagram, as shown in Figure 32 below, an input of actuator position is used to
determine an initial clutch effort using the clutch effort graph. This initial clutch effort is summed with
the clutch effort output from the second block diagram and again inputted into the clutch effort graph to
obtain the new actuator position. This new actuator position is inputted into the PID controller that will
determine the voltage required using a pulse width modulation signal. Pulse width modulation will be
used to adjust the motor to a specific position. The constants for the PID controller will need to be
obtained through testing (KP, KI, KD). First KI and KD will be set to zero to determine the optimal Kp that
will be large enough without any overshoot. Next KD will be determined to minimize oscillation and
finally KI will be determined to reduce the steady state error.

Figure 32: Block diagram of clutch force control in Simulink.

32

15.0 FMEA/RISK ANALYSIS

The following sections describe the analysis of the FMEA and Risk Analysis.
15.1 FMEA
Table 6 below shows all the failure modes in an FMEA chart.
Table 6: The FMEA

The failure mode in the FMEA with highest risk for the projects design was Shaft Yield/Fracture under
the Motor category. Since the shaft that is connected to the clutch pack was designed by Honda and has
already been tested on the motorcycle, the shaft that causes concern is the one on the motor. Although the
shaft is designed for the motor and definitely wont fail under the rotational forces generated by the motor
itself, it was important to be cautious since implementing a one stage gear reduction to transmit torque
33

(this would introduce bending stresses. The likelihood of this failure was determined to be moderate,
however, contrary to the prior section discussing the stresses experienced by the motor shaft, the true
stresses experienced by the motor shaft is relatively unpredictable since it is not known quickly the motor
will accelerate when the system draws maximum current. If this mode fails, then another motor shaft
would need to be manufactured that would need to be heat treated for increased strength and rigidity.
15.2 Risk Analysis
Table 7 below shows the analysis for all the possible risks.
Table 7: Risk analysis

In the risk analysis, the Broken Parts category would be of the highest risk. Although it doesnt seem as
though it has high probability of occurring, if it did happen, then broken parts from gears, shafts, and any
other broken components would fly out and may cause severe injuries. Also, once any part breaks, it will
be necessary to re-design and re-manufacture a new set to replace the broken components. The ways to
minimize the potential for hazards is to design the parts to meet the strength requirements by properly
running FEA and ensure that they have a safety factor of 1.2. Also, when the system is tested, everybody
in the vicinity will be required to wear long sleeve clothing, long pants, cover-toe shoes and safety
glasses.
16.0 MANUFACTURING PLANS AND DRAWINGS
The manufacturing plans have been moved to Appendix A. Please reference the Bill of Materials and
Manufacturing Plans in Appendix A.
16.1 Manufacturing Pros and Cons
The prototype contains a relatively small number of parts to be manufactured, as well as a very low
number of manufacturing method needed to produce the parts: milling, water jetting, and sawing. This is
a huge Pro for the project, because it means the system will be subjected to a relatively small amount of
34

error and stackup tolerances, which could greatly affect the performance of the system in a negative
manner. Another Pro is that the material that will be used is fairly cheap- steel and aluminum. Exotic,
expensive materials, such as Titanium and carbon fiber, are not required for this system. A concern that
has been come across in the prototyping is the manufacturing error and tolerance stackups on the
frame/engine combination. The position of the engine in the overall CAD is approximated because the
engine model was produced by scans from a coordinate measuring machine (CMM), and could be off
by .05-.10 in any direction. This issue is completely uncontrollable, which is why a large amount of
adjustability was incorporated into the design of the mount. The combination of the upper and lower slots
give the motor two degrees of (limited) freedom.
17.0 ASSEMBLY PLAN
Assembly drawings have been created to make sure our system is going to be built correctly. Figure 33
shows an overall assembly view with steps. All of the steps will be explained on the next page.

Figure 33: Shows the overall assembly view with step numbers and the instruction for each step is show
below.
Step 1:
Screw encoder mount on the top of the motor and press fit the encoder adaptor onto the rear shaft. Then
press fit the encoder on the other side of the encoder adaptor and screw them on to the encoder mount.
Step 2:
Screw the motor onto motor mount bracket with washers in between.
35

Step 3:
Weld the quarter gear and gear adaptor to the clutch shaft.
Step 4:
Adjust the position of the bolts relative to the four slots until two gears mesh perfectly with each other,
then tighten the upper two screws that connect the motor mount frame with the motor mount bracket.
Step 5:
Take out the two bottom bolts that fix the motor mount frame on the chassis and then weld the gusset
plate between the motor mount frame and the motor mount bracket.
Step 6:
Adjust the bottom two bolts until the gears mesh perfectly with each other and then tighten them.
18.0 VALIDATION PLAN
The validation segment can be broken up into three categories before the entire system is tested on the
vehicle. Individually, the mechanics, electronics, and code (controls) must all pass validation testing in
order to minimize full-vehicle testing and other potential failures. This section describes each individual
plan, followed by a full-vehicle validation plan. The full validation plan has been moved to Appendix C.
19.0 VALIDATION RESULTS
Progress and current results are explained in the following sections.
19.1Mechanics
The mechanical validation results of the weight, torque generation, stiffness, and failure (stress) have
been complete with passing grades. Figure 34 below shows the final mechanical product mounted on the
vehicle.

Figure 34: The final mechanical product has been mounted and tested on the vehicle.
36

Torque, stiffness, and failure were tested by running the motor, slowly, over its maximum range of
motion (~10 degrees). A power box was used to control the current flow to prevent a quick and
unwarranted lash. The results are shown in Table 8 below.
Table 8: Incrementing the current shows that the static clutch torque is 11.7 ft-lb.

Current (Amp)

Voltage (V)

Full-Range
Movement
(Y/N)

25

12

30

12

2.51

7.53

35

12

3.02

9.06

40

12

3.53

10.59

45

12

3.9

11.7

Motor Torque (ftlb)

Clutch Torque (ft-lb)

The clutch moved under 11.7 ft-lbs of torque, 1.7 ft-lbs over the user requirement. This can be accounted
for in motor and gear efficiency losses. Significant deflection was observed by the engineers, instigating
the welding of a stiffening truss onto the mount. A second test at 12 V, 45 A proved that deflection was
noticeably reduced. This validated our FEA model, as well as the fact that failure did not occur. The final
system weight measured in at 5.09 lbm, as shown below.

Figure 35: The weight of the system is below 6 lb, meeting the user requirement.
The original target of 5 lb was only missed by +0.09 lb.
The speed validation (under .02 seconds) has not yet been validated for safety concerns on the controls
end. Encoder/controller issues delayed testing and raised concerns within the team, causing a decision to
be made to prove the system on a test bench first. Any sudden lash or unwarranted motion from the motor
37

could cause major failure to our entire project or the MR15 Honda CBR 600 engine, which could be
catastrophic for the entire MRacing team.
19.2 Electronics
The team has proven that the designed circuitry is fully functional through a proof-of-concept model that
is hooked up to a dummy motor (very cheap motor that draws low current), as shown in Figure 36,
below. The next step is to remake the circuitry with thicker wires and with heat sinks for the MOSFETs
without the use of a breadboard. Afterwards, the team will implement the wiring into the wiring system of
the vehicle.

Figure 36: A proof-of-concept circuitry makes it much easier to troubleshoot wiring issues prior to
vehicule implementation.
This concept was validated by running power through the circuit and verifying that the optoisolator
completed the circuit to the motor when a voltage (5 V) was applied.
19.3 Controls
Basic proof of concept trials were ran, with the results shown in Table 9 below. The reason only a couple
data sets were taken was due to the fact that the code at this moment still does not account for incredibly
high slip ratios where the torque requested would essentially be infinite. To solve this problem, the code
will have to have a default torque requested for when the slip ratio is too high.
38

Table 9: Input wheel speeds and engine speeds create a simulated launch, resulting in a Clutch Effort
Left Front
Wheel
Speed (mph)

Right Front
Wheel
Speed (mph)

Left Rear
Wheel
Speed (mph)

Right Rear
Wheel Speed
(mph)

Torque
Requested
(Nm)

Engine
Speed
(RPM)

Clutch
Effort
(Nm)

33.18997

33.02487

46.64166

46.81645

-33.5

8434

-785

35.85019

36.65604

43.05908

44.58337

-9.64

8969

-784

38.33729

39.27419

42.13026

42.14641

32.3

9112

-727

A basic test bench, using the small motor provided in Figure 36, above, to prove the controller outputs
could be translated physically through the PWM. Using the same fake signals from above, the results
were positive, and the motor responded in the same manner. The only difference between this motor and
the actual motor are the PID coefficients.
20.0 PROJECT DEVELOPMENTS
The team constructed a steel frame on which some of the project components were mounted. For the
project components that could not be mounted, the team decided to use placeholder parts in order to
represent the parts not on the model. Figure 37 below shows the steel frame that was built by team 30.

Figure 37: Steel frame consisting of a representative frame/motor/gear/clutch model off of the vehicle.

39

The photo above shows the visualization model that was built by the team. The model consists of a frame
bar to represent the chassis bar on the vehicle that the motor and mount are mounted to. The oilite bushing
on which the big gear is mounted to is meant to represent the clutch lever that is partially encased inside
of the engine. This component rotates and engages/disengages the clutch pack by pulling on a set of 5
springs, which are represented by the 1 spring in the visualization model.
21.0 ENGINEERING CHANGE NOTIFICATION
An engineering change occurred during the encoder assembly process, which lead to the recognition of a
missed variable in our encoder mount. The encoder specification sheet indicates an axial tolerance of
+ .025, which was not met by the original mount. The new mount was engineered and manufactured
with much higher precision. This design change has not affected the team in any way. The drawings in
Figure 38 and Figure 39 below show the design change.

Figure 38: Engineering change on the encoder mount allows higher precision from the encoder.

Figure 39: Cross-section view of the new encoder mount shows a better representation of the system.
40

This new encoder mount has been manufactured and assembled onto the motor without any issues. Figure
40 below shows the manufactured product.

Figure 40: New encoder mount is installed properly onto the motor.
22.0 DISCUSSION
This section discusses the design critiques and work that needs to be completed post-ME450.
22.1 Design Critique
Several engineering critiques can be made after completion of this project. Some include spending more
time on the controls sub-system, verifying the CAD model before manufacturing, and working with the Pi
Innovo M220 more closely.
Controls: The methodology behind the controls system was loosely understood and became a time issue
once the mechanics were complete. Implementation of the model also became difficult once the
methodology was fully understood. For the future, the team recommends having a suspension engineer
and electrical engineer work together, where the suspension engineer handles the methodology and
electrical engineer implements the code.
CAD: It was understood from the MRacing team that the frame/engine CAD model was not exact.
Adjustability was built into our design for this reason, but the magnitude of CAD error on the engine
model was much higher than expected. For the future, the team recommends redesigning the bracket for a
higher range of adjustability.
Pi Innovo M220: Using the built-in MATLAB software on the Pi was difficult and not complying with
our original code. For the future, the team recommends building all of the code directly inside the
software.
41

22.2 Future Work

The next steps for this project are creating a circuit based on the proof-of-concept circuit for the prototype
system that is able to withstand a much higher current and finding PID coefficients and incorporating
them into the Simulink model to control the movement of the motor effectively. After this is completed,
the system can be hooked up onto the car and testings can be conducted to assess the performance of the
system and validate how helpful the system is to the acceleration performance of the car. The team
recommends taking high precautions as these steps are carried out- any error in the circuit or code has
potential for heavy damage. In particular, the electronic circuit must have thick enough wire to carry the
current, and the components (MOSFETS, diode, resistors) must be mechanically attached to a very good
heat sink. The code must also be able to comply with any error signals from sensors, or odd scenarios that
could occur on track.
23.0 AUTHORS
Peter Karkos is a Mechanical Engineering student in his 4th year here at Michigan. He
is currently the Technical Director of the MRacing Formula SAE Team, specializing in
vehicle dynamics and suspension design. In his spare time, he enjoys playing hockey.
He is essentially a permanent resident at the Wilson Student Team Project Center due
to the amount of hours spent there.
Calvin Chiu is a Mechanical Engineering student working on his Victory Lap 5th
year. He has experience in the field of automotive engineering and will be attending
graduate school next year at Michigan for Biomedical Engineering. In his spare time,
he enjoys hitting the gym, dancing, and eating food.

transferred
next year at
playing

Chi Qiu is a Mechanical Engineering student in his 4th year who

from Purdue after his first year. He will be attending graduate school
Michigan for Mechanical Engineering. In his spare time, he enjoys
basketball.

Shashwati Haldar is a Mechanical Engineering student in her 4th year at Michigan.

She is pursuing a minor in Electrical Engineering which has contributed to her interest
in the field of controls. In her spare time, she likes to go shopping for clothes and
shoes.

42

24.0 BIBLIOGRAPHY
[1] Ye, J., 2015, Suspension Lead, MRacing , private communication
[2] Malmgren, D., 2006, Automotive Electronics and their implementation in a race car, Lulea
University of Technology, Lulea, Sweden
[3] Miliken, D. L., Miliken, W. F., 1994, Race Car Vehicle Dynamics, SAE International, Warrendale,
PA, pp. 25-71
[4] Mellet, D.S., Plessis, M. du, 2014, A Novel CMOS Hall Effect Sensor, Sensors and Actuators A:
Physical, Vol. 211, pp. 60-66
[5] Ming, Z., Hong, N., Xiao-Hui, W., Xiaomei, Q., and Enzhi, Z., 2009, Modeling and simulation of
aircraft anti skid braking and steering using co simulation method, The international journal for
computation and mathematics in electrical and electronic engineering, pp. 14711488
[6] Delagrammatikas, G. J., Fedullo, T., 2011, The Traction Control System of the 2011 Cooper Union
FSAE Vehicle, SAE International, New York, NY
[7] Keller, A., 2015, Electronics Lead, MRacing, private communication
[8] Trahan, S., 2015, Powertrain Lead, MRacing, private communication
[9] W. Harris, How Motorcycles Work, HowStuffWorks. [Online]. Available at:
http://auto.howstuffworks.com/motorcycle1.htm. [Accessed: 2015].
[10] Wikipedia Contributors, 2015, Launch Control (automotive), Wikipedia Foundation,
https://en.wikipedia.org/wiki/Launch_control_(automotive)
[11] Bek, M., Popp, C., Schiele P., Schmidt, T., and Schwemer, C., 2005, Performance enhancing and
starting process accelerating method for vehicle with hydrodynamic torque converter, Germany,
DE10356194
[12] Murray, S., 2009, Control method and controller for a motor vehicle drive train, US, US7563194
[13] Baer, K., Patel, A., Wheeler, J., Method and apparatus for operating a clutch in an automated
mechanical transmission, US, US6309325
[14] Crane, J., MRacing Driver, MRacing, private communication
[15] Sobotka, J., 2010, Traction Control System for the formula CTU CarTech, Czech Tehnical
University in Prague, Prague, Czech Republic
[16] Udomkesmalee, S., Utama, D., and Nicolls, M., 2004, Evolution and Design of the 2003 Cornell
University Engine Control Module for FSAE Racecar, SAE International, Ithaca, NY
[17] Leland, M., Pneumatic clutch mechanism, US, US2512360
[18] 2014, Wichita Clutch, Altra Industrial Motion Corp., http://www.wichitaclutch.com/Default.asp
[19] FSAE Rules Committee, 2015, 2015 Formula SAE Rules, International SAE, pp 96-98
[20] Hughes, A., Drury, B., 2013, Electric Motors and Drives: Fundamentals, Types, and Applications,
Fourth, Newnes, Oxford
[21] "Inverters." Further Information of Frequency Inverters Technical Guide for Frequency Inverters.
Web. 22 Oct. 2015
[22] "MAE3 Absolute Magnetic Encoder Kit." US Digital. Web. 7 Nov. 2015.
[23] " ." Three-Inch High Performance Motor. Web. 23 Oct. 2015.
[24] Lovell, J., 2015, Systems Engineer, Pi Innovo, private communication
[25] Aerospace Specificaiton Metals Inc., Aluminum 6061-T6; 6061-T651, ASM,
http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA6061t6
43

[26] Umbriac, Michael., 2015, LEO Lecturer IV, Gears- Force Analysis ME350 Lecture 11. University
of Michigan.
[27] Ullman, David. "The Factor Of Safety As A Design Variable." The Mechanical Design Process. 2nd
ed. New York City: McGraw-Hill Companies. 316. Print.

44

25.0 APPENDIX A: MANUFACTURING

Bill of Materials

Material

Type

Width,
Length Radius Height
Part
Purpose (in.)
(in.)
(in.)
Quantity Manufacturer Number Cost

Aluminum
6061-T6 Sheet

Motor
mount
Frame

0.375

Aluminum
6061-T6 Sheet

Motor
mount

0.25

18-8
Stainless
Steel

Bar

Key
Stock

0.125

0.125

Steel

Gear

Clutch
Gear

0.25

1-

$50.6
1 McMaster-Carr 6325K21
1

Steel

Gear

Motor
Gear

0.25

3-

1 McMaster-Carr 6867K41

316
Stainless

Washe Motor
r
Washer

Motor

AmpFl
ow
A28150
Motor

0.038

0.438

5.6

92530A1
1 McMaster-Carr 00
$3.18

$46.0
5

90107A0
100 McMaster-Carr 11
$4.80

1 AmpFlow

A28-150 $309

Manufacturing Plans
Part Number: ME450-Team30-001
Part name: Motor Mount Frame
Raw Material: 6061 Aluminum Plate, 3/8" x
7" x 4"
Step #

Process Description

Machine
45

Fixtures

Tool(s)

Speed
(RPM)

Save the model as dxf file with the

necessary post-processing
1 modifications.
PC
2 Save the dxf file to the waterjet.

designated PC
for water jet

Follow computer station

instructions of waterjet to
appropriately compile the file
3 to .ord file

designated PC
for water jet

Start waterjet to cut the component

4 profile.
waterjet

Part Number: ME450-Team30-002

Part name: Motor Mount Bracket
Raw Material: 6061 Aluminum Plate, 1/4" x
4" x 4"
Step #

Process Description

Machine

46

Fixtures

Tool(s)

Speed
(RPM)

Save the model as dxf file with the

necessary post-processing
1 modifications.

PC

2 Save the dxf file to the waterjet.

designated PC
for water jet

Follow computer station instructions

of waterjet to appropriately compile designated PC
3 the file to .ord file
for water jet
Start waterjet to cut the component
4 profile.

waterjet

Install drill chuck and hold part in

5 vise.

Mill

6 Find datum lines for X and Y.

Mill

vise

vise

edge
finder,
drill chuck

900

1000
1000

7 Centerdrill both holes

Mill

vise

Center
drill,
19/64"
drill bit,
drill chuck

8 Drill both holes

Mill

vise

#21 Drill
bit

9 Tap both holes

Mill

vise

10-32 Tap

47

Part Number: ME450-Team30-003

Part name: Encoder Mount
Raw Material: 6061 Aluminum Plate, 1/4" x
4" x 4"
Step #

Process Description

Machine

Save the model as dxf file with the

necessary post-processing
1 modifications.

PC

2 Save the dxf file to the waterjet.

designated PC
for water jet

Fixtures

Tool(s)

Speed
(RPM)

Follow computer station instructions

of waterjet to appropriately compile designated PC
3 the file to .ord file
for water jet
Start waterjet to cut the component
4 profile.

waterjet

Install drill chuck and hold part in

5 vise.

Mill

6 Find datum lines for X and Y.

Mill
48

vise
vise

edge
finder,

900

drill chuck
Write down the location of all the
7 holes

pencil

Centerdrill both holes on each

8 mount

vise

center
drill,
19/64"
drill bit,
drill chuck

1000

Mill

vise

#43 Drill
bit

1000

Mill

vise

#32 Drill
bit

1000

Mill

vise

4-40 Tap

8 Drill all the tap holes

Drill all the through holes
9 Tap all the tap holes

Mill

Part Number: ME450-Team30-004

Part name: Clutch Gear-Weight Reduction
Raw Material: 3" Steel Gear

49

Step #

Process Description

Machine

Fixtures

Tool(s)

Use protractor and marker to

estimate a 45 degree pie-slice on the
1 gear

Protractor

Mark center of gear with hole

2 punch

Hole punch

3 Cut along marked lines

Bandsaw

Speed
(RPM)

150
Deburring
tool

4 Deburr and/or file edges

Part Number: ME450-Team30-005
Part name: Motor Mount Strip
Raw Material: 1/16" Steel Plate
Step #

Process Description

Machine

Save the model as dxf file with the

necessary post-processing
1 modifications.

PC

2 Save the .dxf file to the waterjet.

designated PC
for water jet

Follow computer station

instructions of waterjet to
appropriately compile the file
3 to .ord file

designated PC
for water jet

Fixtures

Tool(s)

Speed
(RPM)

Fixtures

Tool(s)

Speed
(RPM)

Start waterjet to cut the component

4 profile.
waterjet
Part Number: ME450-Team30-006
Part name: Motor Mount Gusset Plate
Raw Material: 6061 Aluminum Plate, 1/4" x
4" x 4"
Step #

Process Description
Save the model as dxf file with the
necessary post-processing
1 modifications.

Machine

PC

50

2 Save the dxf file to the waterjet.

designated PC
for water jet

Follow computer station

instructions of waterjet to
3 appropriately set up the software.

designated PC
for water jet

Start waterjet to cut the component

4 profile.
waterjet

26.0 APPENDIX B: CONCEPT GENERATION AND SELECTION

Table 10: Concept generation matrix
Power Source

Powertrain

Drivetrain

Sensors

Controllers

User
Input

Driver Effort

Vehicles 75 hp
engine

Chain

Hall effect

Bosch MS4.4

Button

12 V Battery

Motor

Belt

Laser speed
sensors

Pi Innovo

Switch

Pressurized Air

Pneumatic
actuator

Human eye

Arduino

Motion
sensor

Pressurized Liquid

Hydraulic actuator Spool

Human touch

Human brain

Heat
sensor

Load cell (or

torque cell)

Beagleboard

Gear

Solar energy

Linear motor

CVT

Biofuel energy

Driver's arm

Very strong
magnets
Microphone

Radio
communication

Carbon
fiber rod

XBox

Coal energy

51

Temperature
sensor

Hydroelectric energy
via random waterfall

Steel Rod

Nuclear energy

Safety wire Accelerometer

Burning propane (or

any -ane gas)

Titanium
rod

Potentiometer

Raspberry Pi
MSP430

Pressure sensor Nanode

Potato battery

Strain gauge

Pinguino

Lemon battery

Variable
reluctance
sensor

STM32

Exhaust energy
(wind
energy)

Throttle
position sensor Teensy 2.0

Gasoline

MEMS
magnetic field
sensor

Intel Galileo

Yaw rate sensor

Power Source Weighted Selection Matrix

Strength
Reliability
Volume
Ease of control
Feasibility
Cost

Driver
Weight Effort
1
8
10
8
10
7
5
7
0
10
6

Weighted Total

Strength
Reliability
Volume
Ease of control
Feasibility
Cost
Weighted Total

Weight
8
8
7
7
10
6

12 V
Battery

Pressurized
Air

Pressurized
Liquid

Solar Biofuel Coal

Energy Energy Energy

5
5
5
8
10

10
1
5
6
6

10
4
3
6
6

1
1
0
5
0

2
1
2
6
0

7
6
0
6
0

10

288

325

267

2533

51

86

Hydroelectric
Energy
3
6
0
0
0
0
72

158

Nuclear Burning Potato Lemon Exhaust Gasoline

Energy Propane Battery Battery Energy
10
10
1
1
1
10
9
10
1
1
7
7
8
8
3
3
9
10
4
5
8
8
2
1
0
1
0
0
2
2
0
6
10
10
5
10
236
320
184
184
208
320

Powertrain Weighted Selection Matrix

Weight

Vehicles 75 hp
Engine
Motor

Pneumatic
Actuator

52

Hydraulic
Actuator

Linear
Motor

Driver's
Arm

Ease of Control
Reliability
Strength (Power)
Cost
Weight
Weighted Total

1
10
10
10
10
260

10
10
3
5
7

10
10
4
7
7
296

5
1
10
5
5
150

6
1
10
1
1
112

8
9
1
1
8
234

1
10
5
10
10
245

Drivetrain Weighted Selection Matrix

Strong
Carbon Steel Safety Titanium
Wire
Rod
Weight Chain Belt Gear Spool CVT Magnets Fiber Rod Rod
8
6
10
8
8
10
10
10
6
10
5
10
7
10
7
5
10
4
10
4
10
9
9
7
10
7
8
1
8
10
2
10
7
8
6
10
6
9
10
7
9
6
9
10
7
8
5
10
1
6
1
10
10
1
5
7
8
6
7
0
5
10
3
9
8
7
9
9
10
9
0
2
5
5
6
5
10
447 388 477
401 236
332
337
421
313
411

Efficiency
Reliability
Rigidity
Accuracy
Cost
Weight
Feasibility
Weighted Total

Sensors Weighted Selection Matrix

Accuracy
Ease of use
Reliability
Cost
Volume
Weighted Total

Accuracy
Ease of use
Reliability
Cost
Volume
Weighted Total

Weight
10
7
9
5
6

Hall
effect
5
10
8
10
9
296

Laser
Speed
Sensor

Human
Eye

8
6
8
3
9
263

1
10
2
10
10
208

Load Cell
Yaw
Human or Torque
Temperature Rate
Touch
Cell
Microphone
Sensor
Sensor
1
5
4
7
7
10
7
6
7
6
2
5
6
5
6
10
8
7
8
6
10
9
5
9
9
208
238
201
258
250

Pressure Strain
Weight Potentiometer Accelerometer Sensor Gauge
7
7
5
5
10
7
7
7
6
7
6
6
6
6
9
8
8
7
6
5
7
7
8
9
6
255
255
236
230

MEMS
Variable
Throttle Magnetic
Reluctance Position
Field
Sensor
Sensor
Sensor
6
7
7
5
6
4
5
5
5
3
8
1
7
9
8
197
251
196

Receive/Transmit Weighted Selection Matrix

Ease of use
Processing Speed
Cost
Feasibility

Pi
Bosch Innovo
Human
Radio
Beagleboard communication XBox
Weight MS4.4 M220 Arduino brain
1
10
8
10
3
5
8
7
10
7
8
10
3
10
8
8
10
10
8
10
7
1
1
5
10
9
10
0
9
3
3
10

53

Weighted Total

237

Ease of use
Processing Speed
Cost
Feasibility
Weighted Total

266

260

200

80

Weight Raspberry Pi MSP430 Nanode Pinguino STM32

10
3
3
5
6
7
0
5
8
6
8
8
8
1
2
8
4
5
0
3
3
2
3
10
110
96
125
143
156

70

155

Teensy
2.0 Intel Galileo
3
5
6
10
4
1
2
0
109
120

User Input Selection Matrix

Ease of Use
Cost
Practicality
Weighted Total

Weight
7
5
10

Button
8
10
10
206

Toggle
Switch
10
10
10
220

Motion
Sensor

Heat
Sensor
2
2
2
44

1
2
1
27

27.0 APPENDIX C: VALIDATION PLAN

Mechanics
The validation for bracket stiffness can be verified by running the FEA for von mises stress and
deflection. This has already been done in previous design process and has been shown in previous section
of the report (Page 23-26). According to the results of the FEA, the maximum stress that occurs in our
design was far below the yield strength of the material we are using. However for the deflection, we
couldnt determine a certain value as our target, so we decided to do the validation by running our basic
design (without any additional truss and gusset) and evaluate its performance based on our engineering
judgement. If the performance is considered bad, truss and gusset will be welded to our mounting
structure.
To validate the 10 ft*lbs of torque that needs to be generated from the motor, we are going to observe the
system running and see if the motor can move the clutch for a full range of motion which determines if
the motor can provide enough torque to move against the full-loaded springs inside of the clutch pack.
In order to validate the 0.02-second motor reaction speed requirement, which is the hardest one among all
the user requirement we have, data from the encoder will be used to determine how long it takes for the
motor to travel for the full range of motion.
Electronics
54

The electronics portion of this project can be validated through empirical testing. The validation can be
done in two steps: 1) Hook up the completed circuitry to a power source and use a multimeter to test each
component and each node to ensure that the voltage readings make sense, 2) Implement the circuitry into
the electrical system of the car and test it to see if the motor functions properly.
If the voltage readings that are acquired in step 1 demonstrate that the circuitry is in working order, then it
is okay to proceed to step 2 (jumping straight to step 2 is dangerous and can be very time consuming
should the system not function properly).
Controls
Validation for each control block (tire, engine, and clutch) will be tested by running mock-signals into
each sensor port via Simulink, and comparing the outputs to manual hand-calculations. This is a relatively
simple concept, but helps eliminate problems experienced during odd scenarios. Below is a list of
scenarios that must pass, virtually, before implementing it physically on the motor.
Scenario 1: Extremely high wheel slip off of the line. In this case, the slip ratio goes to infinite,
because the front wheels will be at zero velocity.
Scenario 2: Extremely low wheel slip off of the line. In this case, slip ratio goes to zero or
negative.
Scenario 3: Large difference between front wheels (left/right) and/or rear wheels. In this case, a
sensor error, or unknown driving situation, causes the difference between front left and front right
to be large (or rear left/rear right).
Scenario 4: Grip is fully gained on the vehicle, and the motor is no longer needed. In this case,
the motor must be shut off completely
Running this code in real-life will follow the completion of the scenarios listed above.

Full-Vehicle
Full-vehicle testing will, theoretically, be the most simple step, considering the fact that the individual
sub-system validations have all passed. This will include running low-speed trials through first gear on
different surfaces, such as dry asphalt, wet asphalt, and concrete. Proof of functionality will be
determined on whether or not wheel slip can be controlled appropriately. Wheel speed data will be logged
to determine this.

55