Internal Gearbox for Downhill Mountain Bikes

Final Report

Design Advisor: Professor Kowalski

Design Team
Joshua Filgate, Jesse Kuhn, Morgan Misek,
Jay Seiter, Michael Witonis

December 4th, 2007

Department of Mechanical, Industrial and Manufacturing Engineering

College of Engineering, Northeastern University
Boston, MA 02115
 Design Team
Joshua Filgate, Jesse Kuhn, Morgan Misek
Jay Seiter, Michael Witonis

Design Advisor Sponsor

Gregory Kowalski None

Abstract

In the realm of competitive downhill mountain biking, there is a need for advancement of
the bicycle’s drive train. The conventional sprocket, chain, and derailleur drive train is
not well suited for the high speeds, impacts, and harsh environmental conditions inherrant
to the sport of downhill racing. The gearbox created in this project replaces conventional
drive trains on downhill mountain bikes. The initial design and proof of concept were
designed to provide a comparable gear range and capabilities of a conventional drive
train, while providing increased protection, strength, durability, and performance. The
scope of the project was to provide a working prototype capable of demonstrating the
gearbox’s ability to provide the overall system gear ratio range of 2.1-3.3 when interfaced
with industry standard components, as well as a detailed design of a complete prototype
system.
 Table of Contents:
Acknowledgments............................................................................................................... 5
Copyright ............................................................................................................................ 5
1.0 Problem Statement ........................................................................................................ 7
1.1 Proposed Solution ......................................................................................................... 7
1.2 Background Information............................................................................................... 8 Deleted: 8
1. 3 Modern Drive Train Explained……………………………………………………...10 Deleted: 7
1.4 Market Research ......................................................................................................... 10
2.0 State of the Industry .................................................................................................... 11
2.1 Product Niche.............................................................................................................. 14
3.0 Design Process ............................................................................................................ 16
4.0 Mechanical Design...................................................................................................... 19
4.1 Design of shifting mechanism .................................................................................... 21
4.2 Design of gears ........................................................................................................... 25
5.0 Testing: ....................................................................................................................... 33
6.0 Moving Forward: ........................................................................................................ 33
7.0 Conclusion: ................................................................................................................. 34
Appendix I: G-Con Mounting Standards.......................................................................... 35
Appendix II: Market Research Results ............................................................................. 38
Appendix III: Mechanical Requirements Document ........................................................ 41
Appendix IV: Bill of Materials ......................................................................................... 46
References......................................................................................................................... 52

Figure 1: Conventional downhill bike drive train [12] Deleted: 1

Deleted: 4

3
 Table of Figures

Figure 1- Example of a downhill rider in a technical woods section [3]............................ 8

Figure 2 - Professional racer riding through a typical rock garden [4]............................... 9
Figure 3 - Example of the harsh environment [5]............................................................... 9
Figure 5: Shimano Nexus and Rohloff Speedhub multi-speed rear bicycle hubs [11]..... 12
Figure 6: Early internal gearbox bikes that used modified multi-speed hubs as internal
transmissions. From left to right: Evil, BCD, Brodie [6].......................................... 12
Figure 7: BCD's Racing Coffin used a centralized drive train [6].................................... 12
Figure 8: Derailleur-in-a-box internal drive systems. From Left to Right: Hayes, Phaser,
Pete Speed, Honda [8] [9]......................................................................................... 13
Figure 9: Internal gearbox bikes. From Left to Right: Honda, Nicolai, GT ..................... 13
Figure 10: Photos of Suntour V-boxx Insides taken at Eurobike [10].............................. 14
Figure 11: Centralized mounting of gearbox in frames moves weight from conventional
drive trains to a more central location....................................................................... 15
Figure 12: Model of conventional sequential gearbox concept ........................................ 17
Figure 13: Planetary barrel concept model ....................................................................... 18
Figure 14: Sprocket and chain sequential concept model................................................. 18
Figure 15 - Gearbox Exploded.......................................................................................... 20
Figure 16 - Output shaft shifting assembly....................................................................... 20
Figure 17 - Shifting Bulb Assembly ................................................................................. 21
Figure 18 - Output Shaft ................................................................................................... 22
Figure 19: Output Von Mises stress plot .......................................................................... 23
Figure 20: Pawl model ...................................................................................................... 23
Figure 21 - Pawl FEA ....................................................................................................... 24
Figure 22 - Sun Gear......................................................................................................... 24
Figure 23 - Sun Gear FEA ................................................................................................ 25
Figure 24: Gear set 4 specifications.................................................................................. 27
Figure 25: Case ................................................................................................................. 28
Figure 26: Free Body Diagram of Case ............................................................................ 29
Figure 28 - Finite Element Analysis of aluminum case.................................................... 29
Figure 29 - Shifting Scalar................................................................................................ 30
Figure 30 - 3D Model of Prototype Design ...................................................................... 30
Figure 31 - Cutaway of Prototype Model Showing Three Gear Setup............................. 31
Figure 32 - SLA Case and Cover Plates with Planetary Barrel ........................................ 31
Figure 33 - Planetary system with SLA Sun Gear............................................................ 32
Figure 35 - Gearbox Test bed ........................................................................................... 32
Figure 36: Side View of G-CON standards. Angular dimensions.................................... 35
Figure 37: Side view of G-CON standard. Linear dimensions......................................... 36
Figure 38: Top view of G-CON standards........................................................................ 37

4
Acknowledgments

The internal gearbox design team would like to thank Professor Kowalski for his
advisement throughout the project. Without his support, the team would not have
reached its design goals. We would also like to thank graduate students Brian Weinberg
and Jim Forte for their help during prototype production. Northeastern student Ricky
Defrancisco also deserves thanks for lending an uncommon tool during prototyping. The
design team would like to extend special thanks to Randy Moore. Randy provided gear
design expertise that was unparalleled by any handbook.

Copyright
“We the team members,

Joshua Filgate Jesse Kuhn Morgan Misek Jay Seiter Michael Witonis

Gregory Kowalski

Hereby assign our copyright of this report and of the corresponding Executive Summary
to the Mechanical, and Industrial Engineering (MIE) Department of Northeastern
University.” We also hereby agree that the video of our Oral Presentations ifs the full
property of the MIE Department.

Publication of this report does not constitute approval by Northeastern University, the MIE Department or

its faculty members of the findings or conclusions contained herein. It is published for the exchange and

stimulation of ideas.

5
Nomenclature

RPM - revolutions per minute

σ act - actual calculated stress
σ all - allowable stress
b - face width
m - module
K s - size factor
K m - mounting factor
K 0 - overload factor
K v - velocity factor
J - geometry factor (also 2nd moment of inertia in Torque equation)
K L - life factor
K T - temperature factor
K R - reliability factor
Ft - transmitted tangential load
T – Torque

6
1. 0 Problem Statement

In the realm of competitive downhill mountain biking, there is a need for advancement of
the bicycle’s drive train. Current mountain bike drive train components such as the rear
derailleur have been developed over time from designs intended for road cycling. When
subjected to the abusive environment of downhill mountain biking, these designs perform
unreliably, require constant maintenance, and are easily damaged by a wide variety of
external factors.

1. 1 Proposed Solution

To remedy these problems, it is the group’s intention to design and construct an internal
gearbox transmission. The advantage of a gearbox drive train system is that all shifting is
done in a protected, fully enclosed case. This protects the system from contamination
due to mud or water as well as preventing impact damage from objects in the trail such as
rocks or trees. The design will account for maximum rider inputs of 115 N-m and 90
RPMs, withstand high impacts resulting from crashing at high speeds, seal precision
surfaces from particulates and fluids, and be marketable to racers accustomed to spending
between $4000 and $7000 on a complete bike.

The designed gearbox will also conform to the G-boxx.org G-CON standards. G-
boxx.org is an organization within the bike manufacturing industry working towards
advancing drive train design industry wide. The G-CON standard defines a mounting
scheme for aftermarket bicycle gearboxes. The standard defines specific hardware and a
bolt pattern for the interface between the gearbox and the frame. The bolt pattern also
defines a general shape to the top portion of the case as well as a maximum envelope in
which the box must fit. The standard recognizes the difference between bicycle gearbox
design and gearbox frame design. This allows for designers to specialize in their
respective fields. Please review Appendix I for more details concerning the G-CON
standards.

7
1. 2 Background Information

To fully understand the need for a gearbox drive train on a downhill mountain bike it is
important to understand exactly what the sport of entails. The evolution of downhill
mountain biking began with the introduction of cross country mountain biking. In the
1970s a core group of riders pioneered the sport of mountain biking by retrofitting their
1930s and 40s cruiser style bikes with wider, larger tread tires to handle rougher terrain
of off-road trails. This group of riders, centralized around Marin County, CA, hand built
the first mountain specific bikes. The sport began to grow and the Specialized
Stumpjumper was introduced in 1982 as the first mass produced mountain bike [1].
This initial form of mountain biking is now referred to as Cross Country, also known as
XC. XC riding is generally characterized by a mixture of uphill and downhill with
smooth to moderately aggressive terrain. Downhill mountain biking began with riders
seeking the speeds and challenges from the downhill portions of XC courses. Soon
downhill specific bikes were developed and downhill specific trails were carved.

As technology progressed, the sport of downhill mountain biking progressed along with
it. Rapid frame and suspension technology advancement has allowed riders to conquer
increasingly demanding terrain and at higher speeds. Trail features include large rocks
and boulders, rock gardens, roots, jumps, drop-offs of over 10 vertical feet, and gap
jumps of over 35 feet. Additionally, riding conditions can include dust, snow, rain, and
mud. It is under these circumstances that riders reach speeds up to 50 mph in their runs.

Figure 2- Example of a downhill rider in a technical woods section [3] Deleted: 2

Deleted: 1

8
Figure 3 - Professional racer riding through a typical rock garden [4] Deleted: 3
Deleted: 2

Figure 4 - Example of the harsh environment [5] Deleted: 4

Deleted: 3

9
1. 3 Modern Drive Train Explained
The modern downhill bike drive train is a fairly advanced system made up of six major
components. As seen in Figure 4, these components include the crankset, front chainring,
chain, chain guide, cassette, and rear derailleur. As the rider drives the crankset of the bike
the front chainring drives the chain. The chain runs to the cassette and turns the rear wheel
of the bike. The chain then runs through the rear derailleur as it returns back to the front
chainring. The derailleur moves across the cassette moving the chain from gear to gear
changing the ratio between input and output. It is important to note that majority of these
components are open and exposed.

Figure 5: Conventional downhill bike drive train [12] Deleted: 5

Deleted: 4

1. 4 Market Research

An online study asking riders about their experience with derailleurs was conducted. The
question, “How many derailleurs have you broken this year?” was asked to a total of 192
riders. The poll question was posted on two popular mountain biking forums,
Ridemonkey.com and Bustedspoke.com. The skill level of the questioned riders varied
from new riders to collegiate racers to professionals. Even “new riders” within the sport
of downhill should be considered as skilled bike riders. Many people that are new to
downhill have been involved in some other form of riding for a number of years. It is
important to recognize this because the bicycle gearbox is not intended for an average
bicycle rider; it is a downhill specific product. The results from the survey are shown in
Table 1.

10
Table 1: Study of Broken Derailleurs in 2007

Poll question: How many derailleurs have you broken this year?

Number of derailleurs Number of riders

0 102

1-5 78

5-10 9

10+ 3

Of this study, 47% of riders broke a derailleur in 2007. Many of the riders that indicated
zero said that although they had no derailleur failures, they had broken derailleur hangers.
The derailleur hanger is the structural piece of aluminum that attaches the derailleur to
the frame. As one can imagine, the bicycle gearbox would be a solution to these failures
as well. Appendix II contains personal stories from some of the riders included in this
poll. Their stories are important to understand the failure modes of the external derailleur
system.

2.0 State of the Industry

One of the earliest production internal gearbox bicycles was a three speed gearbox bike
manufactured by Adler in 1934. However, added weight, design complexity, and
inadequate manufacturing capabilities led to its eventual demise. It was not until the
dawn of downhill mountain biking where gearboxes would emerge as a truly viable
product. In the late 1990s downhill bike manufacturers turned to the internal drive train to
solve issues with durability and performance. [6]

The early downhill bike frame designs with internal drive trains implemented
technologies already familiar to the bike industry: conventional chain drives and
internally shifting hubs. In 1997, BCD and Brodie released frame designs that used the
Shimano Nexus multi-speed hubs as internal transmissions. [6] Following this same
model, many of the internal drive bikes to follow used a modified Rohloff Speedhub in
the same manner. Figure 5 below shows both hubs. Evil’s 2013i and Nicolai’s Nucleon
DH were among the first frames to implement the Speedhub as an internal transmission
(See Figure 6 below). Using the Nexus and the Speedhub enabled frame manufacturers to
turn out designs relatively quickly, but internal hubs have their drawbacks. The Nexus
hubs are designed with the intent of being used for commuter bikes and are not designed
with the rigors of downhill mountain biking. The speedhub, while able to withstand the

11
abuse of downhill is quite costly. The Rohloff hub sells for more than $1200, driving the
overall cost of the gearbox up. [11]

Figure 6: Shimano Nexus and Rohloff Speedhub multi-speed rear bicycle hubs [11] Deleted: 6
Deleted: 5

Figure 7: Early internal gearbox bikes that used modified multi-speed hubs as internal Deleted: 7
transmissions. From left to right: Evil, BCD, Brodie [6] Deleted: 6

Alternatively, BCD’s next attemp at the internal drive train, the Racing Coffin (shown in
Figure 7), was released in 1999. The Racing Coffin’s design was quite simple in that it
used components similar to the conventional chain drive transmission, except it located
the shifting components in a central and protected area of the frame. The concept of
simply moving the conventional drive train to a more central location is now referred to
in the industry as the “derailleur-in-a-box”. The most prominent manufacturers of
derailleur-in-a-box internal gearboxes are Honda, Pete Speed, Hayes, and Phaser. These
products are shown below in Figure 8.

Figure 8: BCD's Racing Coffin used a centralized drive train [6] Deleted: 8
Deleted: 7

12
Figure 9: Derailleur-in-a-box internal drive systems. From Left to Right: Hayes, Phaser, Pete Speed, Deleted: 9
Honda [8] [9] Deleted: 8

Following the Racing Coffin’s entry into the public racing circuits, Honda, GT, and
Nicolai all released frames with internal drives (see figure 9 below). Currently there are
roughly a dozen frame manufacturers working towards internally geared bikes. There is
however, a certain degree of mystery as to what exactly resides in the insides of some
manufacturer’s gearboxes. While most manufacturers rely on familiar technologies such
as derailleurs, chains, and multi-speed hubs, some manufacturers are bringing new
technologies to the bike industry. Photos of Suntour’s V-boxx, taken at Eurobike, shown
in Figure 10, reveal what appears to be a sequential design, similar to what is commonly
used on motorcycles.

Figure 10: Internal gearbox bikes. From Left to Right: Honda, Nicolai, GT Deleted: 10
Deleted: 9
As new gearbox technologies emerge, one of the main limiting factors in the industry is
the division between frame design and internal drive train design. Frame manufacturers
lack the resources to develop their own drive trains and without knowing the specifics of
individual frame designs, drive train designers cannot anticipate the exact layout and
mounting features frame manufacturers require. In order to aid the industry’s advance, G-
boxx.org has clearly outlined both the division and interface between frame and gearbox
design in its G-CON standard. Since its conception, the G-CON standard has enabled
other major frame manufacturers to enter the internal drive market as well as allowing
previous manufacturers to standardize their own designs. Standardized mounting schemes
give frame manufacturers the option to experiment with different internal drives as newer
drive designs reach production. As of 2007, Orange, MSC, Nicolai, Diamondback,
Allutech, and Richi are all designing frames for the G-CON standard.

13
 Figure 11: Photos of Suntour V-boxx Insides taken at Eurobike [10] Deleted: 11
Deleted: 10

2.1 Product Niche

The reason for the industry’s push in the gearbox direction is that the inherent nature of
the gearbox makes it a perfect fit for the sport. Through offering ingress protection,
impact protection, lower maintenance requirements, and an improved center of gravity,
internal gearboxes present the opportunity for technological advancement of downhill
mountain bike design.

Ingress Protection

The fully enclosed nature of the gearbox protects the precision drive train components
from the mud, dust, water, and rocks that are characteristic of a downhill course. As
stated in the Engineering Requirements document in the Appendix III, a minimum IP65
rating will be needed for the enclosure. This will keep the internals free from dust and
slightly pressurized water. Currently, the external derailleur system is completely open to
these elements. Sticks often get stuck in shifting systems and violently rip derailleurs off
frames. Mud and rocks easily get stuck in the small pulley wheels and springs of a
derailleur and cause the system to jam. In addition, using a fully enclosed case facilitates
the use of an open bath lubrication system for all of the precision surfaces on the inside of
the gearbox.

Impact Protection

The fully enclosed nature of the gearbox will also protect the shifting internals from
aggressive impact loads. As previously discussed, downhill courses are littered with

14
rocks, trees, and other immovable objects. Derailleurs are mostly constructed of 1/8”
thick aluminum with some plastic pieces. When these types of materials encounter
boulders at the speeds previously mentioned, they fail catastrophically. This means that
the rider now loses the race and needs to buy new parts for the bike. Since the gearbox
has a case surrounding the internals, the impact is absorbed and distributed throughout
the entire enclosure. Because the design intent is to follow G-CON (see Appendix I)
mounting standards, the case also acts as a structural member to the frame. In this way,
any impact on the drive train is an impact on the frame which can effectively distribute
and handle the forces involved.

Maintenance

Although the gearbox will be designed with serviceability in mind, it should be relatively
maintenance free. Service such as lubrication, alignment, cable tension, and spring
tension are standard short term maintenance events for derailleur systems. Since the
bicycle gearbox will be tuned out of the factory and fully enclosed, these routines will
need to be performed less often.

Improved Center of Gravity

The G-CON mounting standard allows the gearbox to be mounted at the junction of the
seat tube, down tube, and chain stay or rear suspension pivot of the bike frame. This
location is where the bottom bracket currently exists. The gearbox will essentially
replace the bottom bracket portion of the frame. This is the lowest possible point on the
bike. By replacing the existing drive train with a gearbox, the center of gravity of the
bike is optomized. A lower center of gravity translates to a more stable and nimble
vehicle. See Figure 11.

Figure 12: Centralized mounting of gearbox in frames moves weight from conventional drive trains Deleted: 12
to a more central location Deleted: 11

Efficiency, Weight, and Cost Considerations

It can be argued that the drawbacks to a gearbox are its efficiency, its added weight over
a traditional drive train, and its cost due to precision parts. While the final efficiency of

15
the gearbox has yet to be determined, it will most likely be less efficient than a
conventional chain drive system which can be up to 93% efficient [2]. The inherent
difference between the designs is that gearboxes have more frictional surfaces in contact
than chains and sprockets. In addition, these surfaces need to be machined precisely and
therefore drive the cost of gearboxes up. Finally, while chains interface with sprockets so
that the tangential load is distributed across nearly half of the sprocket’s teeth, gears mesh
only one set of teeth at the time. In this way, gears have to be wider than sprockets and
therefore require more material, driving the weight higher.

The tradeoffs for higher initial cost, lower efficiency, and higher weight of gearbox drive
trains are durability and reliability. Downhill racing is a sport governed by high speeds
and relatively short runs. Riders require their equipment to function completely
throughout the course of a race run. Failed equipment results in lost races. Additionally,
race courses are only open to practice for limited times at race events. Equipment failures
during practice force riders to spend practice time performing maintenance on their bikes
wasting valuable practice time. Consider that market research showed that some pro
riders replace up to 15 new derailleurs in a year. Apart from the direct results of
decreased practice times and incomplete race runs, derailleurs can cost racers a great deal
of money. A conservative estimate of $70 per derailleur equates to over $1000 in just
derailleurs for one racing season.

3.0 Design Process

The first step in the design process was to brainstorm concepts for both gear
configuration and shifting. Five gear configuration concepts were conceived: the
conventional sequential gearbox, a planetary system, a planetary barrel design
(specialized planetary design), a sprocket and chain sequential system, and a derailleur-
in-a-box concept. Each concept carries with it different inherent advantages and
disadvantages.

Conventional Sequential

Used predominantly in motorcycles, this concept is well proven. Consisting of constantly

meshed gears where face gears engage the desired drive gear ratio, this sequential design
requires some form of a clutch to shift from one gear ratio into another. See Figure 11
below.

16
 Figure 13: Model of conventional sequential gearbox concept Deleted: 13
Deleted: 12
Planetary System

Planetary gearboxes lend themselves to high torque ratios in a small amount of space,
making them a prime candidate for a lightweight and compact design. However, the
linking together of several planetary gear ratios is mechanically challenging and does not
lend itself to ease of packaging or design.

Planetary Barrel

This concept is the team’s innovative solution to the mechanical complications inherent
with using a series of planetary gears. It consists of a series of sun gears that share a
common drive shaft, where the sun gears are free to rotate unless engaged. Each sun gear
and its corresponding planets is a different size, such that the summation of the sun and
planets diameters is equal for each gear set. In this way, it is possible for the entire series
of gears to share a common ring gear, as to act as one single output gear.

17
 Figure 14: Planetary barrel concept model Deleted: 14
Deleted: 13

Sprocket and Chain Sequential

While chain drive systems inherently carry high drive efficiencies, the standard means of
shifting gears does not act as a highly durable or stable platform. Derailleurs require a lot
of tuning and often can become problematic, showing resistance to shifting into the
proper gear or staying in the desired gear. By replacing gears in a conventional sequential
gearbox with chains and sprockets, the efficiency of chain drive systems is mixed with
some of the stability of sequential gearbox design. Still, chains lend themselves to a fair
amount of routine maintenance and a complex shifting mechanism would be required.

Figure 15: Sprocket and chain sequential concept model Deleted: 15

Deleted: 14

18
 Derailleur in a box

The derailleur in a box entails the use of an ordinary bicycle drive train where the shifting
mechanisms and drives would be incased in a protective box. While this design boasts the
efficiency of chain drive systems, it does not appear to be as rugged or easy to maintain
as some of the other concepts. Additionally, the packaging of this system is difficult to fit
within the G-Con standard size restraints.

In order to compare each of the above concepts quantitatively, each concept was put into
a decision matrix and rated against weighted design criteria. The criteria that was
included, in order of emphasis, weight, packaging, durability, interfacing, shifting feel,
and efficiency. Factors such as weight, durability, and efficiency stem from the product
niche and market research while packaging, interfacing, and shifting feel encapsulate the
team’s own design requirements. The concept chosen to advance to the mechanical
design stage was the planetary barrel, based on its winning score in the decision matrix.
See Table 2 below.

Table 2: Concept Decision Matrix

Conventional Planetary Sequential Derailleur in a

Sequential Planetary Barrel Sprockets box
Criteria
Criteria Factor Rating Score Rating Score Rating Score Rating Score Rating Score

Weight 5 2 10 3 15 4 20 4 20 4 20
Packaging 5 4 20 3 15 5 25 3 15 2 10
Durability 5 5 25 4 20 5 25 3 15 2 10
Efficiency 3 3 9 3 9 3 9 5 15 5 15
Interfacing 4 2 8 4 16 3 12 3 12 5 20
Shifting
feel 4 5 20 3 12 3 12 4 16 2 8
Total
Score: 92 87 103 93 83

4.0 Mechanical Design

In order to make the mechanical design process manageable, the group implemented a
from-the-inside-out design strategy. In this way, emphasis was placed more on the
complex mechanisms located in the heart of the system. The process was broken down
into three major design tasks corresponding with major system components. The
breakdown is as follows: shifting assembly, planetary stack assembly, and case. An
exploded view of the finalized design can be viewed below in Figure 15. A detailed
design discussion on each component follows in this section. Each system component
was designed and analyzed through the same basic process.

19
 1) Component conceptualized in Solidworks
2) Initial back of the envelope calculations performed with simplified geometry
3) Solidworks model updated to reflect initial calculations
4) Finite Element Analysis performed to check back of the envelope calculations in
actual geometry.

Figure 16 - Gearbox Exploded Deleted: 16

Deleted: 15
Shifting Assembly

For ease of packaging, the shifting mechanism was located inside the hollow input shaft
such that each sun gear can be engaged individually. A model of the shifting assembly
can be viewed below in Figure 16.

Figure 17 - Output shaft shifting assembly Deleted: 17

Deleted: 16
As can be delineated from the above image, a shifting bulb is pulled through the shaft via
an industry standard handlebar mounted shifter. As seen in Figure 17, the shifting bulb
has spring loaded balls that depress as the bulb slides through the shaft. The shaft has cut
features where the pawls are located. Each planetary system has an individual set of
pawls that engage the sun gear of that system. As the bulb reaches one of the pawl sets,

20
the balls extend outward and interfere with a surface on the pawl. The springs continue
to push the pawls outward to engage the sun gear.

An added feature designed into the shifting assembly is its ability to freewheel. By
utilizing a fixed gear rear hub the chain is constantly moving when the wheel is moving.
When the rider is not pedaling, the pawls continually ratchet ensuring disengagement
from the sun gear. As soon as the rider pedals, the pawls match the speed of the sun gear
and drive the system. This freewheeling assembly allows the system to be shifted not
only under light pedaling but also when the rider is not pedaling.

4.1 Design of shifting mechanism

The bulb was designed as two separate pieces for ease of maintenance. Springs and balls
may exhibit wear over time so a method of removing the springs and balls was desired.
As seen in Figure 17, the balls and springs can be removed by separating the two halves.
A standard shifting cable will be used to pull the bulb through the shaft. Because the
bulb is located in a constantly spinning shaft, the cable needs to attach to a bearing in
order to not twist. A return spring is attached to the bulb and end of the shaft in order to
maintain a resistive force on the bulb. This spring acts as a return force for a downshift.
The bulb will be made from aluminum for ease of machining since there is no load on the
bulb itself.

Pawl
interference
balls

Cable Bearing

Bulb Body
Extension
springs

Figure 18 - Shifting Bulb Assembly Deleted: 18

Deleted: 17
Shaft Design
The design of the output shaft was very unique. The shaft is required to be hollow to
allow for the shifting bulb, but this shaft is also under load from pedaling forces. As seen
in Figure 18 below, the shaft has many cut features for the pawls which structurally
weaken the component.

21
 Figure 19 - Output Shaft Deleted: 19
Deleted: 18
Back of the envelope calculations were performed on a simplified geometry of the shaft.
An input torque of 119N-m was applied to the outer radius of the shaft for a worst case
scenario. The calculations were as follows:

Τ= Tc/J

T= 119N-m/2 = 58 (half torque due to 2:1 primary drive reduction)

J=1/2(ri4 – ro4)π = 3.33794x10-8 m4
c = .0125

Tshaft= 21.72 MPa

This low stress value is to be expected given the simplified geometry used in the
calculation. It was decided that a finite element analysis was needed to further evaluate
the integrity of the shaft. Figure 19 below shows a rendering from COSMOSWorks. The
shaft was fixed at one end and a force was applied to each pawl pocket surface. The shaft
is designed such that the stress from the pawl is distributed directly to the pocket surface
in the shaft. The pawl pivot rods do not take the majority of the stress. A force of 1900
N was applied to each pocket. This 1900 N was derived from the 58 N divided by 3
pawls at a 10 mm moment arm. Figure 19 shows a max stress of about 370 MPa. The
material selected for the final design was 4130 steel. 4130 is machinable in its normalized
and annealed form and can readily be heat treated to significantly increase its yield
strength.

22
 Figure 20: Output Von Mises stress plot Deleted: 20
Deleted: 19

The pawl is the component that couples the sun gear to the output shaft when the desired
gear ratio is selected. A model of the pawl can be seen in Figure 20.

Engagement faces

Figure 21: Pawl model Deleted: 21

Deleted: 20
The geometry of the pawl was directly based on current free hub mechanisms in the bike
industry. The side feature enables the return of the pawl into the shaft when the particular
gear ratio is not selected. The return is performed via a retaining spring that fits into a
groove in the shaft. The retaining spring also fits into the side feature in the pawl.
Analysis of the pawl is of particular concern due to the high stress over a small surface
area. It is difficult to simplify the geometry of the pawl for back of the envelope
calculations so an FEA analysis was done. A COSMOSWorks rendering of the pawl can
be viewed in Figure 21.

23
 Figure 22 - Pawl FEA Deleted: 22
Deleted: 21
A slight stress concentration can be seen at the corners of the engagement faces. The
maximum stress determined is on the order of 261 MPa. In addition to stress the
ratcheting of the system requires a material that wears well. This drives material
selection towards a moderate steel.

Each sun gear has internal teeth that the pawls interface with upon selection of that gear
ratio. This can be seen in Figure 22. Similar to the pawl, these engagement points have a
low surface area contact point and are of considerable concern. Because of the
complicated geometry, an FEA analysis using COSMOSWorks was performed. The gear
was held in place and three engagement points, 120 degrees apart were loaded with 633N
(previously mentioned 1900 divided by 3 pawls). An FEA rendering can be seen in
Figure 23. The arrow on the scale to the right of the plot shows the yield stress of the
selected alloy steel. The FEA analysis shows that all stress concentration are below this
yield.

Bearing surface

Pawl
Engagement
Points

Figure 23 - Sun Gear Deleted: 23

Deleted: 22

24
 Figure 24 - Sun Gear FEA Deleted: 24
Deleted: 23

4.2 Design of gears

Due to the overall nature of the system, the mechanical design of the planetary gear setup
is quite complex. The first step in gear design was defining the required gearbox ratio
range. As stated in the Mechanical Requirements document (see Appendix III), the
overall system gear range needs to be 2.1-3.3 to match current drive trains on downhill
bikes. Consider however that the gearbox’s range can be scaled both at its input and at its
output. By using chain and sprocket interfaces between the crank arms and the gearbox
input and between the gearbox output and the rear wheel, different scaling factors may be
obtained by changing sprocket ratios. The next consideration for the system was the
maximum diameter of the ring gears. The G-CON standard outlines max dimension for a
front to back length of the gearbox as 165.5 mm. Leaving room for case wall thickness
and clearance between moving parts, the maximum ring outer diameter was set to 112
mm.

The next aspect considered was ensuring that the gears chosen would mesh. Where
planetary systems have a number of planets meshing into both the sun and the ring gears,
the actual feasibility of gear mesh needs to be calculated. This calculation is given by
Equation 1 below. If the feasibility value is an integer the system will mesh. Otherwise,
the system will not mesh evenly and is not a possibility.

Equation 1: Planetary gear mesh feasibility

(#of teeth in the ring gear + #of teeth in the sun gear) = Feasibility Value
#of planets

The next consideration was the gear stress calculated via the AGMA method. According
to AGMA, the maximum bending stress is a function of gear geometry and the tangential
force. Shown below is the AGMA equation utilized to design the planetary system.

25
 Equation 2: Bending Stress on Gear Teeth

1. 0 K s K m
σ = Ft K 0 K v
bm J

Due to the relatively low RPM, the cyclic nature of human pedaling, the quality available
in precision gears, and predicted number of teeth in the system, the size, mounting,
overload, velocity, and geometry constants were set to the following:

K s - Size Factor = 1
K m - Mounting Factor = 1.2
K 0 - Overload Factor = 1.25
K v - Velocity Factor=1
J - Geometry Factor = .35

After all of the constraints were defined, the next step in the process was to optimize gear
selection based on the max ring dimensions, planetary feasibility values, the overall
desired gear ratio range, input and output scaling factors, and gear geometries. A
spreadsheet was developed to carry out this process and later modified to include wear
loads and material properties. The AGMA stress values for the final design ranged from
173Mpa to 277Mpa.

Wear analysis was carried out using the Buckingham Equation, where the wear load is a
function of the gear and pinion sizes, tooth width, pressure angle, and material
properties. For this particular system, wear loads were relatively low ranging up to 200
MPa. This is due to the thin profile of each gear used to satisfy geometric constraints of
the system.

Equation 3: Buckingham Gear Wear Load

Fw = K Q b Dp
The fatigue life was calculated for a three year lifetime of the product. The estimated time
for pedaling during that time frame was found to be 343 hours. This time accounts for 40
runs a week every week of the year, at 6 minutes each when the rider pedals at an average
of 60 RPM. A factor of safety of 2 was used to ensure a conservative analysis. The final
material selection was achieved by building a matrix with available gear materials and

26
comparing their physical properties to the AGMA stress and fatigue analysis. Materials
that satisfied both conditions were then compared using the Buckingham wear load
calculations. The finalized material selection is depicted in Table 3. See Figure 24 for an
example of optimized sizing and material selection of one of the planetary gear sets.

Table 3: Material Optimization

2024 Aluminum 303 Stainless Steel 416 Stainless Steel 17-4 PH Stainless Steel Selected Material
AGMA Fatigue AGMA Fatigue AGMA Fatigue AGMA Fatigue
Planet N N N N N N Y Y 17-4 PH Stainless Steel
Sun N N N N N N Y Y 17-4 PH Stainless Steel
t1

Ring N N N N N N Y Y 17-4 PH Stainless Steel

Se

Planet Y N Y N Y N Y Y 17-4 PH Stainless Steel

Sun Y N N N N N Y Y 17-4 PH Stainless Steel
2t

Ring Y Y N N N N Y Y 2024 Aluminum

Se

Planet Y N Y N Y Y Y Y 416 Stainless Steel

Sun Y N Y N Y Y Y Y 416 Stainless Steel
3t

Ring Y Y Y N Y Y Y Y 2024 Aluminum

Se

Planet Y N Y N Y Y Y Y 416 Stainless Steel

Sun Y N Y N Y Y Y Y 416 Stainless Steel
t4

Ring Y Y Y N Y Y Y Y 2024 Aluminum

Se

Planet Y N Y Y Y Y Y Y 416 Stainless Steel

Sun Y N Y Y Y Y Y Y 303 Stainless Steel
5t

Ring Y Y Y Y Y Y Y Y 2024 Aluminum

Se

Planet Y N Y Y Y Y Y Y 303 Stainless Steel

Sun Y N Y Y Y Y Y Y 303 Stainless Steel
6t

Ring Y Y Y Y Y Y Y Y 2024 Aluminum

Se

Planet Y N Y Y Y Y Y Y 303 Stainless Steel

Sun Y N Y Y Y Y Y Y 303 Stainless Steel
t7

Ring Y Y Y Y Y Y Y Y 2024 Aluminum

Se

Output Y Y N N Y N Y Y 2024 Aluminum

Figure 25: Gear set 4 specifications Deleted: 25

Deleted: 24

27
Case Design

The starting point for the case was the geometric layout and mounting pattern defined by
the G-CON 2.0 standards defining bottom bracket location, upper profile, and interfacing
bolt pattern are defined. With the overall general layout defined, the next step was to
design the support features for the rotating shafts in the planetary barrel design. For ease
of assembly and maintenance, the case was designed to be three pieces: a main body and
two sides. Given the nested nature of the planetary barrel system, the mounting features
for the case needed to be placed at the two sides. Aluminum 6061 was chosen as a
material due to its weight, strength, and machinability properties. See figure 25 below.

Figure 26: Case Deleted: 26

Deleted: 25
4.3 Design of case

Given the complexity of the case, exclusively finite element analysis was performed for
analysis. Shown below in Figure 26, a free body diagram of the case demonstrates that
the only loading on the case when the bike is in its horizontal orientation is generated
from the rider’s legs pushing down on the pedals. Given the capabilities of the human
leg, this force was estimated to be 250 lbs for each leg. FEA analysis demonstrates a
maximum stress of 7Mpa for the case. While this value falls significantly below the yield
stress of 6061 Aluminum, the seemingly conservative design of the case should enable
the case to withstand direct impacts that would occur in collisions.

28
 Figure 27: Free Body Diagram of Case Deleted: 27
Deleted: 26

Figure 28 - Finite Element Analysis of aluminum case Deleted: 28

Deleted: 27
4.4 Interfacing
The case was designed with an industry standard threaded bottom bracket shell to
facilitate the integration with industry standard crank sets and bottom brackets. In order
to integrate with industry standard trigger shifters, a shifting scalar was conceptualized
and prototyped. The shifting scalar was necessary because a typical bicycle shifter only
pulls 3.45mm of cable in single throw. Because of the spacing between each planetary
system, the shifting bulb needs to move 10mm. Therefore; in order to use a standard
shifter, a scaling device is necessary. Figure 29 shows the shifting scalar design. As seen
from the figure, the cable from the shifter rotates the smaller diameter wheel and in turn
rotates the larger wheel of the pulley. The cable attached to the larger wheel pulls the
shifting bulb through the shaft.

29
 Figure 29 - Shifting Scalar Deleted: 29
Deleted: 28

5.0 Prototyping
The limitations of the overall Capstone budget caused the group to focus its finances on a
proof of concept prototype. The proof of concept was designed around three gear ratios
instead of the finalized seven. The group took advantage of campus laboratory SLA rapid
prototyping for the more complex components, such as the case, sun gears, and hollow
drive shaft in order to shorten the prototyping timeline and to reduce cost. The majority
of the prototype cost can be attributed to high precision thin section bearings needed to fit
the design into the overall geometric specifications as well as stock purchased for the
machining of components. See Figures 29and 30 below for a detailed model of the proof
of concept prototype.

Figure 30 - 3D Model of Prototype Design Deleted: 30

Deleted: 29

30
 Figure 31 - Cutaway of Prototype Model Showing Three Gear Setup Deleted: 31
Deleted: 30

Due to its material cost and complexity, the case was a prime candidate for SLA
production. Since the proof of concept was not designed with the intent of being
subjected to any significant loading or impact conditions, the material properties of the
SLA material were not an issue. Similarly, the sun gears, drive shafts, and pawls were all
produced via SLA due to their complexity. The added benefit of rapid prototyping these
parts was that they could be reproduced if any necessary modifications needed to be
made. This flexibility was preferable due to the uncertainly surrounding the shifting
mechanism and planetary barrel configuration. Figure 31 shows the planetary stack and
case of the prototype.

Figure 32 - SLA Case and Cover Plates with Planetary Barrel Deleted: 32
Deleted: 31
Given the simplicity of the necessary machining operations needed, the ring and planet
gears were purchased from Action Automation. The gears were stock from their

31
catalogue and the post machining was completed in the Capstone machine shop. One of
the complete planetary gear set ups is shown below in Figure 33.

Figure 33 - Planetary system with SLA Sun Gear Deleted: 33

Deleted: 32

In order to test the proof of concept prototype, a test stand was constructed from 80/20
aluminum extrusions and mounting hardware. The sliding nature of 80/20 hardware
allowed for the adjustment of the distance from a mounted rear wheel to the gear box
output and the overall versatility means that the stand can be used for future testing.

Figure 34 - Gearbox Test bed Deleted: 34

Deleted: 33

32
5.0 Testing:
In order to demonstrate the feasibility of the overall design, the proof of concept was
tested methodically. Each sub system was tested individually and then placed into a
complete assembly. Issues arose with each of the sub assemblies but were overcome. The
lack of precision machining in the support oriented components caused the planetary
barrel configuration to exhibit a larger level of resistance than was expected, and the lack
of precision machining of the shaft and bulb caused for a finicky shifting mechanism.
Before the full system was assembled, the clearance was opened up between the bulb and
the shaft as to allow to the bulb to slide easier. Upon full system integration, the system
exhibited the abilities to provide three distinct gear ratios and to switch between them as
well as driving a wheel in the test fixture.

6.0 Moving Forward:

While the current design exemplifies the group’s initial design objectives, it could be
advanced to a beta prototype by redesigning specifically for weight reduction,
manufacturability, cost, ease of maintenance, and achieving an extended lifetime. The
current gear material was selected due to its low cost and availability. The group is
considering different aluminum and heat treated steels for future testing in order to
reduce weight, cost, and extend the lifetime of the gears. In addition, a lubrication system
should be added to the system. The proposed system is an open bath system where any
particulates in the system would be washed from the precision surfaces and collect in a
basin at the bottom of the case.

While weight was not a main concern for the proof of concept or initial design, it is an
issue that should be addressed in future iterations. Many of the components in the system
are machined from stock and have more material than is actually required from a
mechanical load standpoint. Additional machining operations could easily remove
material from the ring gears, the case, backing plates, and gears. One other main area for
reduction of weight is the case structure as whole. The use of a composite skin with an
aluminum support frame would eliminate weight currently attributed to the precision
machined aluminum case while maintaining the desired strength. It is estimated that
optimization of strength and weight for the system would eliminate 30% of the weight.

Another key improvement for the system would be to manufacture the barrel output shaft
out of one piece of material, opposed to assembling it from seven separate pieces. This
would replace a series of post machining operations, eliminate assembly procedures,
remove material required to support fastening, and decrease the tolerance stack up
incurred by fastening seven different gears together.

One of the main areas for cost reduction would be to eliminate the need for the thin
section precision bearings used to support the sun and ring gears. This could be achieved
in two ways. The first would be to redesign the geometry of the gears and the support

33
structures with cheaper bearings in mind. The second would be to replace the cartridge
bearing assemblies with machined bearing surfaces and use integrated ball bearings.

The final area for improvement in the current system would be to redesign with the intent
of ease of assembly. This process entails the reduction in the quantity of fasteners, using
more common fasteners, and eliminating nested assemblies and would reduce assembly
time, repair time, and tools required for any required assembly or repair operations.

7.0 Conclusion:

The gearbox designed and prototyped in this project is a viable replacement for
conventional drive trains on downhill mountain bikes. It provides impact and ingress
protection while adding minimal weight to the system. The proof of concept
demonstrates the feasibility of the design of the sub systems as well as the overall
mechanical integration and the project is ready to move to the alpha prototyping phase.

34
Appendix I: G-Con Mounting Standards

Figure 35: Side View of G-CON standards. Angular dimensions. Deleted: 35

Deleted: 34

35
Figure 36: Side view of G-CON standard. Linear dimensions Deleted: 36
Deleted: 35

36
Figure 37: Top view of G-CON standards Deleted: 37
Deleted: 36

37
Appendix II: Market Research Results
These are personal accounts of riders included in the market research poll. These riders
were asked the question, “How many derailleurs have you broken this year?” Not all 192
voters posted stories and some of the stories were not relevant to the topic. Note that
“XT”, “105”, “Dura ace”, “ultegra”, “shadow”, “saint”, and “X9” are different models of
SRAM and Shimano derailleurs.

“One retired due to being really old.

One of my buddies went through 6 XT's last year.

He keeps hitting the cable clamp bolt and pushing into the [parallelogram], an XT
shadow is in his future”

-from Zark

“I have both SRAM and shimano and I must say I like the shimano more.

I've gone [through] 2-3 this year. An XT and an ultegra. XT was bent, and tweaked the
ultegra a little, then the tube on my chain stay became unwrapped and it got pulled into
the derailleur and I pedaled and cracked it. 105 works well at the moment.”

-from jcook90

“[I've] done a Dura ace in this year and the 105 that replaced it is looking a bit worse for
wear. [I've] also done 2 [mechanical] hangers. The 105 is being replaced with a shadow
for this weekend and it will become my spare. I have used both SRAM and shimano and
work as a bike mechanic so set the stuff up all day. I [personally] think SRAM is not
worth the extra money it costs (here in the UK anyway).
The only problem with shimano now is that the shadow doesn't come in SS. I don’t need
a cage that big when I only have a 12-25 cassette”

-from Pow wow

“I just changed my derailleur, it was the second this year.”

-from Vena

“I [blew] up two this year. Then I decided to switch to shortcage road stuff.”

-from Dan wask

38
“Took out two xt's this year, one short cage, one medium rapid rise, moved up to Sram x9
now, havent had enough riding on it yet to see if its the brand, or if its just trail abuse. I
will know next time I need to replace it.”

-from DirtyMike

“I don't trash derailleurs, just lots of hangers.”

-from greenhood

“I know people who go through upwards of 25 sram derailluers a year, yet I have ruined
2 in my lifetime, simply because of sticks getting stuck in my spokes. It has a lot to do
with riding style.”

-from Red Bull (edited for relevancy)

“I've bent 5 so far at the Parallelogram. I also haven't really ridden in the last 2 months,
so I'm sure I would have gone through at least 1, maybe 2 more. I have a tendency to
bend more because of the 24" wheel that my old [Bighit] frame had to have. That and I
live in AZ, and there is a TON of rocks that are just aiming for it.”

-from DhDork

“2 this year. I broke a DuraAce in half (R.I.P.), and ripped a Saint apart. Was able to
repair the Saint with parts, it's 100% again.”

-from zahgurim

“One SRAM X9. On a super-fast, super-rocky descent, the [derailleur] sheared off and
ripped out half of the drive-side spokes. Managed to fix the wheel as best a possible and
complete the remainder of the ride. This was on a loop ride tour of the DH tracks here in
Portes du Soleil. Had loads of chairlifts to get back home.”

-from DamoDh

“I broke 1 or 2 XT's. The one I remember bent across the top side of my cassette after a
hip jump gone wrong (sideways landing). I also bent my chain.”

-from mtnbrider

39
“None this year, but twig season is just beginning!!

[I’m] hoping to [demolish] my XT soon so I can replace it with a SRAM. I just can't get
the XT adjusted right...have straightened hanger and replaced every other part of the
drive train. Different people have attempted to fix the problem with no success. Of
course, this is the reason I haven't [demolished] it yet.”

-from miatagal96

“Just one. I seem to destroy one [derailleur] per season... that’s enough for me.

EDIT- O ya and I did put a pretty nice tweak/bend on my brand new short cage x9 at the
last mt snow race... but its still working. Good thing for derailleur hangers.”

-from manitouman28

“[derailleur's] still [working] fine, but i've toasted 4 [derailleur] hangers. Much props to
my x9. And by toasted hangers, i mean bent [beyond] [straightening] back or smashing so
hard the threads got messed. 1 of those 4 was a OEM specialized one that's wicked
flimsy, the other 3 were the nice solid wheels manufacturing ones from QBP.”

-from yeti187

“last season i killed about 15. this year i have yet to kill one, but i'm sure it will now fall
off on my next ride for saying that.”

-from ford

“1 saint down. it actually still works but the cage is somewhat toast. just threw a new one
on the other day. Before going to saint I was toasting hangers and [derailleurs] at a
[ridiculous] rate.”

-from Doro

40
Appendix III: Mechanical Requirements Document

41
42
43
44
45
Appendix IV: Bill of Materials

46
47
48
49
50
51
References
[1] Bradbury, Doug; Farrel, Al; Ready, Steve.
http://www.mtnbikehalloffame.com/home.cfm. 16 September 2007

[2] Wilson DG. 2004. Bicycling Science 3rd Edition. Cambridge (MA): MIT Press.

[3] Littermag. Online magazine.

http://littermag.com/features/vermont07/photos/2.jpg. 14 September 2007

[4] Littermag. Online magazine.

http://www.littermag.com/features/msa07/day4/5.jpg. 14 September 2007

[5] Littermag. Online magazine.

http://www.littermag.com/features/deervalley07/day1/6.jpg. 14 September 2007

[6] Universal Transmission GmbH. http://www.g-boxx.org

14 September 2007

[7] Dirtragmag. Online magazine.

www.dirtragmag.com/web/interbike2004/page4.php

[8] Sicklines. Online magazine. www.sicklines.com/2007/03/19/phaser-gearbox/

[9] Dirty-Pages. Online magazine. http://www.dirty-pages.net/nieuws/petespeed.html

[10] Google Picsaweb. Online photo album.

http://picasaweb.google.de/TFJ777/SuntourVBoxxInside1/photo#5111621402612997
362

[11] Sheldon Brown. http://www.sheldonbrown.com/harris/rohloff.html

21 October 2007

[12] Ironhorse Bicycle Company. http://www.ironhorsebikes.com/bikes/bikes-

2008.php 21 October 2007

52