FRONT-WHEEL FRICTION DRIVE

ELECTRIC BICYCLE MOTOR

Oliver Chen

Mireille Ghoussoub

Cherry Zhou

Project Sponsor:

Dr. Andrej Kotlicki

Project 1316

Engineering Physics 459

Engineering Physics Project Laboratory

The University of British Columbia

April 3rd

2013

Executive Summary

Many cyclists are seeking easy solutions to bring electric power to their bikes.

Having an electric motor that can be switched on to get a boost going uphill, to go faster,

or to simply take a break from pedaling can be very useful. With these motivations, we

sought to develop a front-wheel friction-drive motor system that could be easily mounted

and removed from any adult-sized road bike. The objective was to develop an easily

mountable solution containing the motor, controller, battery pack, and throttle that could

be installed within minutes.

Friction-drive motors are a cheaper, lighter alternative to the more common hub

motors that must be permanently attached to the wheel of the bike. Indeed, the motor

mechanism component of our design weighs 1.45 kg, and the total weight of our design

does not exceed 3 kg (depending on the size of battery pack used). The design of our

solution was based on goal of keeping the installation as simple as possible, whilst

ensuring that the mechanism could tolerate the strong rotational forces introduced by the

motor. Our solution mounts at the handlebars by means of two snap-on clamps, as well as

at the fender holes, located at the hub of the front wheel of the bicycle. The design can be

adjusted at two locations: by changing the angle between a horizontal and a vertical

cantilever, as well as by changing the height of the supporting rods. The motor is attached

to a pivot point, such that is has sufficient room to engage and disengage with the front

tire. Blocker pieces are located strategically to ensure that the motor does not swing too

far back and get stuck in the wheel.

The mechanical aspect of our solution has proven to be successful. We were able

to mount the mechanism to five different bicycles, all of which varied greatly in

handlebar shape and tire size. In all these cases, our solution was easily installed in under

two minutes.

Installing a user-friendly throttle proved to be very challenging. The most

successful attempt entailed running the controller by means of a purchased servo tester

circuit; however, our throttle setup was not functioning in time for us to do test ride our

solution. We therefore cannot present meaningful results as the electrical efficiency of our

system.

We recommend three important actions to be taken to ensure that our solution be

safe and fully-functional. First, all electrical parts should be protected in a water-proofed

enclosure. Secondly, a push button throttle should be installed along with the servo tester

circuit. Finally, we recommend that an emergency brake be installed to ensure the safety

of the user in the case where the throttle or controller fails.

TABLE OF CONTENTS ABSTRACT…………………………………..………………………………………..…ii

LIST OF FIGURES…………………………………………….……………………...…iv

LIST OF TABLES………………………………………………………………..…….....v

1.0 INTRODUCTION…………………………………………………………………...7

1.1Background…………………………………………………………………….7

1.2 Objectives……………………………………………………………………..7

1.3Scope and Limitations……………………………………………………….....7

1.4Organiztion…………………………………………………………………….7

2.0 DISCUSSION…………………………………………………………………….…..8

2.1 Theory……………………………………………………................................8

2.2 Methods and Testing Protocol………………………………………………....9

2.3 Mechanical and Electrical Components……………………………………..14

2.4 Results………………………………………………………………………..16

2.5 Discussion of Results………………………………………………………...17

3.0 CONCLUSION………………………………………………………………………18

4.0 PROJECT DELIVERABLES………………………………………………………..18

4.1List of Deliverables…………………………………………………………...18

4.2 Financial Summary…………………………………………………………..18

4.3 Ongoing Commitments by team members…………………………………..19

5.0 RECOMMENDATIONS…………………………………………………………….20

6.0 APPENDICES……………………………………………………………………….21

Appendix A. SolidWorks Drawing of Key Components………………………………...21

Appendix B. Electronics Schematics…………………………………………………….25

Appendix C. Cost of Materials of the Project………………….………………………...27

7.0 REFERENCES………………………………………………………………….......28

LIST OF FIGURES

Figure1. Friction-drive Mechanism

Figure2. Handlebar Clamp

Figure3. Supporting Rod

Figure4. Joint between Horizontal and Vertical Cantilever

Figure5. Slots on the Tip of the Supporting Rod

Figure6. Motor Blocks

Figure7. Servo Tester Device

Figure8. 3D Printed Throttle Attached to Pot

Figure9. Hall Sensor Throttle

Figure10. Jump Cable Circuit

Figure11. Examples of Successful Mounting Cases

Figure12. All-in-one Units in A Bike with Fender

Figure13. Weight Comparison between Hub Motor and Our Design

Figure14. Whole Set of All-in-one Units

Figure15. Solidworks Design of Fully Assembly

Figure16. Solidworks Design of Clamp

Figure17. Solidworks Design of Motor Attachment

Figure18. Solidworks Design of Vertical Cantilever

Figure19. Solidworks Design of Horizontal Cantilever

Figure20. Solidworks Design of Blocker Hole Zoom In

Figure21. Solidworks Design of Supporting Leg

Figure22. Solidworks Design of Fender Hole Attachment

Figure23. 555 Timer

Figure24. Electrical Schematic

LIST OF TABLES

Table1. Financial Summary

Table2. Cost of All-in-one Units

Table3. Industry-standard Benchmarks

1.0 INTRODUCTION

1.1 Background As demonstrated by their popularity, bicycles are one of the most practical, cheap,

and sustainable transportation solutions around the world. However, the physical exertion

required to travel uphill or to go faster poses a disadvantage to many riders. Keeping in

the spirit of sustainability, we built an electric friction-drive motor that can easily attach

to the front wheel of a typical touring bike. The system allows the rider to engage the

motor when they wish to go faster, or take a break from pedaling. The reasons for opting

to build a friction-drive motor, instead of a hub motor, lie in their easier installation,

lower costs, and higher power to weight efficiency. While there currently exists solutions

on the market, most come in the form of incomplete kits, and it falls to the user to

customize the system to their bike. Our design offers an all-inclusive package that can

easily attach to any touring-style bike.

The purpose of this report is to communicate the results of our friction-drive motor

design, as well as to provide recommendations for future improvement of the solution.

1.2 Objectives Our aim is to build a removable, all-in-one unit, including battery pack, motor,

controller, and throttle, which may easily attach and adjust to the front wheel of any

touring-style bike. Specifically, the motor and battery should run for at least 40 minutes,

with the bike moving at an average of 15 km/h. We are not developing a very powerful

electric bike motor, but rather one that may be easily transferred between different

bicycles.

1.3 Scope and Limitations This report primarily addresses the mechanical aspect of our solution. It will

present all features of our mechanical design, the reasons behind our choices, and our

design's ability to mount to different bikes. It does not include information on the riding

experience or on power considerations, as we were unable to obtain sufficient test riding

data. These serious limitations are due to last-minute failures to produce a correct throttle

signal to run the motor, and they will be further addressed in the discussion section of this

report.

1.4 Organization Our recommendation report is comprised of four main sections: Discussion,

Conclusions, Project Deliverables, and Recommendations. The Discussion section covers

the physical theory behind our mechanism, the methods undertaken to construct an

easily-mountable device, the attempts to develop a user-friendly throttle, and the results

of our solution.

2.0 DISCUSSION 2.1 Theory

Theory behind the friction-drive mechanism

In a friction-drive setup, the spinning motor comes into contact with the tire, and causes it

to spin. The motor is attached to the bike such that it can pivot under the impact of its

own angular momentum, thus engaging with the bike wheel. Once it is in contact with the

bike wheel, the high coefficient of friction between the two surfaces prevents it from

disengaging. In order to achieve maximum power transfer from the motor to the wheel,

the motor must engage just enough with the tire to prevent it from slipping without

deforming the tire.

Figure1. Friction-drive Mechanism

2.2 Methods and Testing Protocol

Methods of constructing an easily-mountable device

In order to make solution that could be adjusted to fit different bicycles, we

considered all possible locations to which the motor could be mounted. These options

included the handlebars, the head tube, the hub, and the front-wheel forks. Our selection

was based on which locations would best support the forces introduced by the motor

mechanism, whilst minimizing the number of places where the user must attach the

device. Based from Tao Wang's 479 project recommendation report, we chose not to use

the fork blade curvatures to mount the motor, as they proved to weaken under the force of

the motor. We selected instead to mount the mechanism at the handlebars and the hub of

the front wheel, and this proved to be successful.

Mount Testing Protocol

To test the quality of our solution, we attempted to mount the mechanism to different

bicycles at the UBC Bike Kitchen. These attempts allowed us to identify the weaknesses

in our design, and it took several reiterations of certain parts before our current solution

was achieved. To ensure that the testing provides meaningful insight, we purposely tested

on bikes of varying handlebar width and shape, and of different tire size.

2.3 Mechanical and Electrical Components Handlebar mounts

The mechanism attaches at the handlebars by two polyethylene clamps. The

pieces were waterjet-cut, and are flexible enough to accommodate handlebars of 2 to 3

cm in diameter. A bolt screws through each clamp to ensure rigidity. The clamps snap

onto the handlebars from the under, rather than over (see Figure 2), and this setup avoids

interference from gear cables, which often lie directly in front of the handlebars.

Figure2. Handlebar Clamp

Hub mounts

It quickly became apparent that the mounting at the handlebars alone could not

withstand the rotational forces of the motor mechanism, and therefore made it necessary

to include an attachment at the hub of the front wheel. We initially designed two rods that

mounted to the hub; however, the difference between quick release and nut-type wheels

made it difficult to find a solution that would fit all. From surveying the bikes available at

the UBC Bike Kitchen, we found that roughly 50% of the bikes were of the quick-

release-type, and 50% of the nut-type. Fortunately, during the survey we observed over

90% of the bikes have fender holes at the end of fork. Thus we decided to choose fender

holes as the supporting point of the hub mount

Fender Hole mounts

As fender holes are designed to attach fender to the bikes, they have standard size

of diameter, which fit to either M5 or M6 metric bolts. Using this advantage, we design

our supporting rod as shown below: two holes are used to connect to the fork, and the rest

for the fender if needed.

Figure3. Supporting Rod

Adjustable designs

There are several designs in the mechanism to fit different bikes; they together

ensure a rigid mount and perfect distance between motor and the tire.

Joint between horizontal and vertical cantilever

The joint has one pivot point and a circular slot so that the angle between two

pieces can be adjusted continuously from 0° to 90°.

Figure4. Joint between Horizontal and Vertical Cantilever

Slots on the tip of the supporting rod

These slots provide multiple level of height the motor can be mounted so that the

motor is close enough to engage when running but not too close that it hits the tire

when not running. The distance between the top slot and the bottom slot is 9 cm.

Figure5. Slots on the Tip of the Supporting Rod

Motor blockers

Motor blockers are used to block the motor from over engaging, in which case the

motor get stuck with the tire. Multiple holes were included should the user wish to

relocate the blockers fit the specific blocking angle required for their bike.

Figure6. Motor Blocks

Throttle

The controller can run from a purchased servo tester circuit (see Figure 7). The

device can operate in three modes, one of which allows the user to manually turn a

potentiometer-knob to change the signal's duty cycle, and thus control the speed of the

motor.

Figure7: Servo Tester Device

Attempt 1: 3D Printed Thumb Throttle

Although the knob allows for smooth control of the motor speed, it is awkward to

use whilst cycling. For this reason, we designed and 3D-printed a thumb throttle that

mounts to the potentiometer knob (see Figure 8).

Figure8. 3D Printed Throttle Attached to Potentiometer knob

Attempt 2: Hall Sensor Throttle

Still, we were not satisfied with this solution as we were unable to implement a

spring that would make the throttle bounce back once the user ceases to press. The

reliability and robustness of the throttle is critical to the safety of the user, and we

therefore decided to use a purchased thumb throttle (see Figure 9). This throttle was

actually a hall sensor throttle, which produces a signal varying linearly from 0 to 5 V with

the angle of rotation. We then tried to control the output of the servo tester circuit by

attaching the hall sensor throttle to its input. Following the advice of hobbyists online, we

removed the potentiometer and jumped a cable on the servo tester circuit; however, this

failed to control the duty cycle of the signal.

Figure9. Hall Sensor Throttle

Figure10. 555 Timer Circuit

Attempt 3: On/Off Switch

We tried to implement a simple On/Off switch button onto the servo tester circuit.

However, the changes we made to the circuit in our previous attempt seemed to have

caused some damage, and output signal gave too small a duty cycle to have the motor

speed be high enough to engage with the wheel.

Attempt 4: 555 Timer Circuit

In another attempt, we built a circuit to generate a square wave pulse that we

could control by an on/off switch. The circuit uses a 555 Timer, and the duty cycle and

frequency of the square wave output can be determined by the choice of resistor and

capacitor values. Since this circuit can only produce square waves of duty cycles from

50% to 99%, we added an inverter to the output to obtain our desired duty cycle of 12%.

The circuit generated the correct output signal when powered from a voltage supply.

Previously, the servo tester circuit was powered by 5 V from the controller; however, we

were unable to power our 555 Timer circuit this way.

2.4 Results

Results of Test Mounting

The motor unit can be successfully mounted to different bikes with different kinds

of handlebar designs. Three adjustable designs are discussed above to ensure the

flexibility of the mounting.

Figure11. Examples of successful mounting cases

We even managed to mount our all-in-one units to a bike with a fender as shown

in Figure 12.

Figure12. The unit attached to a bike with a front-wheel fender

In testing our solution, we measured the time taken to mount the motor unit to

each bicycle. In every case, the time was under 2 minutes. An Allan key is the only tool

required by the user to mount the mechanism.

The weight of our friction drive motor unit is 1.45 kg (not including controller, battery,

and throttle), which is significantly lighter than a typical hub motor (see Figure 13).

Figure13. Weight Comparison between Hub Motor (5.32 kg) and Our Design (1.45 kg)

2.5 Discussion of results

In mount testing on a total of 7 bikes, 5 of them work perfectly, 1 of them could

work after lowering down the handlebar, and the other one would not work due to fender

hole is located on only one side of the fork. More testing can be performed to find other

limitation and constraints of our design.

Speed testing has not been performed yet due to the problems we encounter with

the electrical design. Test rides would be performed after electrical design is fixed.

Figure14. Motor unit

3.0 CONCLUSION Based on the results, our solution has succeeded in being easily adjusted to fit a

variety of road bikes. Its ability to fit onto a variety of handlebar shapes and sizes, and to

accommodate different types of fender holes, indicates that it is a practical solution for

people seeking a non-permanent electrically-powered motor solution for their bicycle.

The two cases in which the unit failed to fit to the bike are exceptions that do not

undermine the success of our solution. In the instance where the bike offered only one

fender hole was considered to be rare by bike mechanics at the UBC bike kitchen. In the

case where the distance from the handlebars to the front-wheel tire was too long, the

handlebar stem was adjustable and could have been lowered to accommodate the motor

mechanism.

Given the circumstances of our throttle circuit, we were unable to perform test

rides, and therefore cannot provide information as to the electrical efficiency of our

solution. This task will remain an ongoing commitment of all our team members.

4.0 PROJECT DELIVERABLES

4.1 List of Deliverables The primary deliverable is the motor mechanism, comprising of the motor unit,

and the supporting rods, and a controller and servo tester are provided for speed control.

4.1.1 Motor Unit

The motor unit is comprised of the cantilever mechanism, as well as the two supporting

rods. As mentioned previously, the unit will attach at the handlebars by means of clamps.

The unit has the ability to adjust to various heights along the mount in order to achieve

the optimal location for any bike. The batteries, controller, and cables are velcroed

compactly to the unit.

4.1.2 Throttle

A servo tester device serves as the throttle and should be attached to the handlebar mount

where user can easily reach. Our current solution does not include a fully-functional

throttle; however, our team is committed to ensuring that this deliverable will be ready by

April 19th.

4.2 Financial Summary

# Description Quantit

y

Cost Purchased

By:

Funded By:

1 Brushless DC Motor 1 $75.77 Bernhard Project Lab

2 Programming Box 1 $9.91 Bernhard Project Lab

3 Controller 1 $73.58 Bernhard Project Lab

4 Battery 1 $45.00 Bernhard Project Lab

5 Hall Sensor Throttle 1 $15.00 Bernhard Project Lab

6 Servo Tester Device 1 $4.15 Bernhard Project Lab

7 Water jet cut pieces 10 $20.00 Bernhard Project Lab

Total Cost $243.41

Table1. Final Summary

4.3 Ongoing commitments by team members

Our team remains committed to ensuring that our bike motor is fully functional,

and wish to include a working throttle as part of our final design. Our target date for this

goal is, and we will turn in all our project deliverables to our sponsor, Andrzej Kotlicki,

by Friday, April 19th.

5.0 RECOMMENDATIONS Despite the success of our solutions' ability to mount to different bikes, it still

lacks some important features that prevent it from being fully functional. The following is

a list of recommended actions that should ensure that our final deliverables meet the

original project objectives:

1. Waterproofing

We recommend that the controller and battery sit in a water-proof enclosure,

underneath the metal frame.

2. Throttle Installation

Given that the only success in running the motor came from using the servo tester

circuit, we recommend using this, along with a an ON/OFF push button. The servo tester

circuit should operate in manual mode, with the potentiometer knob set to a position that

ensures a duty cycle of at least 12%, to ensure that the motor speed is high enough to

engage with the tire. The COM port of the push button should connect to the input of the

servo tester. The NC port should connect to the controllers ground signal, and the NO

port to the controller's 5V signal. This setup should allow the user to activate the motor

by pushing the putton, and deactivates it when the button is released. The user still has

the option of adjusting the motor speed by changing the potentiometer knob on the servo

tester circuit, as one speed may not be optimal for all bikes.

3. Emergency Brake

To ensure the safety of the solution, an emergency brake should be installed in

case the throttle or the controller fails. The brake should therefore break the circuit

between the battery and the controller, and the brake throttle should be mounted at an

easily-reachable location.

4. Bicycle Basket

While our final recommendation is more a matter of personal preference, we

would like to recommend that a basket be fit around our current mechanism. Aside from

improving our design aesthetically, a basket would provide space to the controller and

battery.

As mentioned earlier, our team members are commitment to the ongoing

improvement of our solution, and we hope to have carried out the recommendations

(notably 1 and 2) by the time the project deliverables are handed to our project sponsor.

Appendix A SolidWorks Drawing of Key Components

Figure15. Full motor mechanism assembly

Figure16. Handlebar clamp

Figure17. Motor attachment piece

Figure18. Vertical Cantilever

Figure19. Horizontal Cantilever

Figure20. Blocker Hole Zoom In

Figure21. Supporting Rod

Figure22. Fender Hole Attachment

Appendix B Electronics Schematics

Figure23. 555 Timer Circuit

The frequency and the duty cycle can be adjusted by selecting appropriate resistor and

capacitor values for R1, R2, and C1.

THigh Time for which the signal is high, s

TLow Time for which the signal is low, s

F Frequency, Hz

THigh = 0.693*(R1+R2)*C

TLow = 0.693*R2*C

F = 1.44/[(R1+R2)*C]

At its maximum duty cycle, the PWM signal generated from the servo tester circuit has

THigh = 2.10 ms, TLow = 14.3 ms, and F = 61.5 Hz. For the purpose of calculating R1 and

R2, we assign the value of THigh to TLow and vice versa, since the output of the 555 timer

is then inverted.

Resulting component values:

C = 47 μF

R1 = 360 Ω

R2 = 62 Ω

Figure24. Electrical Schematic

Appendix C Cost of Materials for the Project

Removable all-in-one Units of Friction Drive Electric Bike Motor

Material Price Weight/time Cost

Aluminum $8-10 /lb 174 g $3.84

Polyethylene $3-9 /kg 45g $0.45

Water-jet $1/min 5min $5

Friction Motor

SK3-6374-149KV

$75.77 N/A $75.77

Controller $73.58 N/A $73.58

Battery $35 N/A $35

Other Cost $20 N/A $20

TOTAL COST $213.67

Table2. Cost of All-in-one Units

The information of the Industry- Standard Benchmarks as shown in following table

Table3. Industry-standard Benchmarks

REFERENCE

Analytic Cycling (2012). Forces on Rider. Retrieved December 10, 2013. From

http://www.analyticcycling.com/ForcesPower_Page.html

Bicycle Design (2013) Retrieved January 10, 2013. From

http://bicycledesign.net/wp-content/uploads/2012/09/Footloose-assembly-pair-49

8x343.jpg

BikeFanatic (2010). Eboost-kepler friction drive Review and Testing.

Retrieved March 20, 2013. From

http://endless-sphere.com/forums/viewtopic.php?f=4&t=22026

Commuter Booster (2013) Retrieved January 9, 2013. From

https://sites.google.com/site/commuterbooster/photo-album

Eboost Power Assist (2013) Retrieved February 15, 2013. From

http://www.eboo.st/index.php?main_page=product_info&cPath=9&products_id1

&zenid=hv5n2hrgclcklqji52ejmlud91

MIT (2012) The Copenhagen Wheel. Retrieved September 29, 2012. From

http://senseable.mit.edu/copenhagenwheel/gallery.html

MIT (2013). Understanding D.C. Motor Characteristic. Retrieved November 19,

2013. From http://lancet.mit.edu/motors/motors3.html

Rainer Pivit (2/1990). Drag Forces in Formulas. Pp.44-46. Retrieved November 18,

2013. From http://sheldonbrown.com/rinard/aero/formulas.htm