Incorporation of a High Performance, Four-Cylinder, Four

Incorporation of a High Performance, Four-Cylinder, Four
Incorporation of a High Performance, Four-Cylinder, FourStroke Motorcycle Engine into a Snowmobile Application
Gregory A. Davis, Nick S. Dahlheimer, David A. Meyer,
Aaron S. Messenger, James R. Johnson, Bernhard P. Bettig
Michigan Technological University
Copyright © 2005 SAE International
For the 2003 and 2004 SAE Clean Snowmobile
Challenges, the successful implementation of a clean,
quiet, high-performance four-stroke motorcycle engine
into an existing snowmobile chassis was achieved. For
the 2005 Challenge, a new motor and chassis were
selected to continue the development of a four cylinder,
four stroke powered snowmobile. The snowmobile is as
powerful as today’s production performance models, as
nimble as production touring sleds, easy to start, and
environmentally friendly. This report describes the
conversion process in detail with actual dynamometer,
emissions, noise, and field test data, and also provides
analysis of the development processes and data. The
vehicle meets the proposed 2012 EPA snowmobile
emissions regulations and is significantly quieter than a
stock snowmobile.
In response to the pollution concern of current
snowmobile use in pristine areas, the Clean Snowmobile
Challenge (CSC) was created. The CSC is a national,
collegiate design competition administered by the
Society of Automotive Engineers (SAE). The purpose of
the competition is to challenge universities and their
students to address the rising concern of pollution, both
noise and emissions, from snowmobiles. Michigan
Technological University has diligently strived to
increase team awareness, and as a result, has accepted
nearly 30 new members. The team is composed of
undergraduate and graduate students with majors in
Mechanical Engineering, Mechanical Engineering
Engineering Technology, Biology, Scientific and
Technical Communication, and Business. All of these
students unify as one team to compete and excel in the
2005 Clean Snowmobile Challenge.
The team’s primary goals for the 2004 CSC were to
design and produce a snowmobile with exhaust
emissions below the proposed 2012 EPA snowmobile
regulations and a noise level lower than that of today’s
quietest 4-stroke snowmobiles. These goals had to be
achieved while maintaining reasonable cost, comparable
performance, and expected durability. For the 2005
CSC, the team decided to continue their work with 4cylinder, 4-stroke engines because they felt that a great
deal of improvement could be made over previous
designs. In order to accomplish these improvements,
the team started from “the ground up” by designing and
building a completely new and innovative snowmobile,
utilizing both positive and negative knowledge gained
from previous designs. These goals can be found in
Table 1. A comparison between the 2004 goals and the
2005 goals can also be made to observe the areas in
which the team wanted to improve the snowmobile for
the 2005 CSC.
Table 1: Michigan Tech CSC Goals
2004 Goal
440cc two-stroke
equivalent performance
Emissions passing 2012
EPA Regulations
2005 Goal
600cc two-stroke
equivalent performance
Emissions passing 2012
EPA Regulations as well
as surpassing previous
designs and entrants to
the CSC
Noise output lower than
that of quiet 4-stroke
snowmobile (Arctic Cat
660 touring), 105 dBa
Noise output lower than
that of any production
snowmobile, 105 dBa
Easy maneuvering,
navigating trails as easily
as a manufacturer’s
snowmobile at 64 km/hr
Easy maneuvering, rider
comfort and ergonomics
matching that of
It is observed in Table 1 that the team attained every
one of their goals with the 2004 snowmobile. For 2005,
the team felt that they could take their success to the
next level by designing a snowmobile that would surpass
the expectations of even those from years past in all
areas of the CSC. While there is always room for
improvement in existing designs, the team decided to
start over for 2005. A different motor was selected and a
chassis was chosen to accommodate the new power
plant. The design of this new snowmobile focuses on all
aspects of the CSC, and because of this, the team will
exceed design parameters of years past and achieve
each goal set for 2005.
This report discusses the design of Michigan Tech’s
entry into the 2005 SAE Clean Snowmobile Challenge.
All aspects of the design will be included as well as test
results achieved by the team prior to the 2005 CSC.
The design can be separated into four major categories:
Performance, Emissions Control, Noise Control, and
Consumer Acceptability.
Rockefeller Jr., Memorial Parkway” [4]. The EIS has
been followed by proposed EPA emissions regulations
for off-highway vehicles, including snowmobiles.
The EPA released regulations for snowmobile emissions
in September of 2002 [5]. The three-phase reduction
calls for a 30 percent reduction in emissions by 2006, a
50 percent reduction by 2010, and a 70 percent
reduction by 2012. Table 2 outlines these regulations.
Table 2: EPA Snowmobile Emissions Regulations
Year of
Maximum HC
Maximum CO
Since the late 1960’s, most snowmobile manufacturers
have utilized a two-stroke, spark-ignited engine as the
primary power source. The two-stroke engine provides
a large power output in a compact, lightweight, and cost
effective design. The inherent disadvantage of the twostroke engine is its poor control over the gas exchange
process, as both the exhaust and intake valve are open,
simultaneously allowing intake charge, consisting of air,
fuel, and oil, to pass directly through the combustion
chamber into the exhaust without being ignited. On
average, 20-33 percent of the intake charge is allowed to
pass through the exhaust port without being ignited [1].
Another disadvantage is the fact that oil and gasoline are
mixed into the intake charge and oil is consumed by
combustion. These downfalls lead to high output levels
of hydrocarbon (HC) and carbon monoxide (CO)
emissions [2].
In recent years manufactures have addressed these
emission concerns and have incorporated innovative
technology into 2-stroke engines. Such innovations
include new intake processes, new injector styles, and
the inclusion of Direct Injection. This involves injecting a
precise amount of fuel into the combustion chamber in
contrast to having "approximately" the correct amount of
fuel being drawn in along with the air flow. Also, with
direct injection, the fuel is better atomized than with
standard 2-stroke engines, resulting in a cleaner and
more complete burning of the fuel. On average, the fuel
efficiency of direct-injection 2-stroke engines is 30
percent better than conventional engines [3].
High levels of emissions produced from two-stroke
snowmobiles have caused concern among several key
environmental groups. In 1997, several of these groups
filed suit against the National Park Service, requiring
them to conduct an Environmental Impact Study (EIS).
This study was titled, “Winter Use Plans Final
Environmental Impact Statement for Yellowstone and
Grand Teton National Parks, and the John D.
Table 3 is a list of components and equipment used to
meet the team goals of 600cc two-stroke equivalent
performance, an acceleration run to 500 feet in less than
7 seconds, possession of a specific power of at least
25W/N, 2012 EPA Emissions Regulations, and a noise
level less than 105 dBa utilizing the testing procedure as
outlined in the rules for the 2005 Clean Snowmobile
Challenge [6].
Table 3: Snowmobile Component Specifications
2004 Polaris ProX 600
Honda CBR 954cc RR, inline 4-cylinder,
4-stroke, dual overhead cam,
spark-ignited, liquid cooled
Fuel System
Honda Stock CBR954RR PGM-FI
(Programmed Fuel Injection),
Walbro Inline Fuel Pump
Modified Honda CBR954RR Intake
System, naturally aspirated
Exhaust Headers: Stainless Steel, MTU
Clean Snowmobile Designed and
Fabricated 4-2-1 System,
Thermal Barrier Coating
Catalyst: 500cpsi TS Catalyst
Muffler: MTU Clean Snowmobile
Designed Muffler System
Primary Drive: Micro Belmont Reactor
Four Tower.
Secondary Drive: TEAM Fast Reaction,
Totally Encapsulated Roller Helix
Semi-Direct Drive System incorporating
a FAST Inc. Gearbox, geared reduction
of 2.1:1
Front suspension: Polaris trailing arm
with Fox FLOAT Shocks utilizing air
Rear suspension: Polaris equipped with
Ryde FX shocks having adjustable
121” x 1.25” x 15” Camoplast Ripsaw
Bump Track
The four design areas of performance, emissions
control, noise control, and consumer acceptability will
now be discussed in detail.
In order for the Michigan Tech team to accomplish their
performance strategy goals, team efforts were focused
on three main topics. These include power adaptation
from the engine to a Continuously Variable Transmission
(CVT), incorporation of a semi-direct drive system, and
packaging all components in such a manner that would
allow the center of mass to be as low as possible while
maintaining rider comfort.
In order to accomplish this task, a two part adapter was
designed and fabricated. The first portion of the
adaptation involved a coupler that bolted to the magneto
side of the engine’s crankshaft. The coupler bolted to the
crankshaft in the same way that the original
flywheel/rotor does in the stock system. This allows the
use of the stock starter gear and starter clutch utilization
to start the motor. It also allows for easy attachment of
the entire system. This part was machined from 7075
aluminum. The second portion is a shaft section which
bolts directly to the coupler and protrudes to the outside
of the motor. The shaft portion was machined from 4140
Steel which was heat treated to provide the needed
strength and rigidity for the system. Although the
flywheel was removed, the mass of the adapter portions
accounts for this and is directly used as a flywheel as
well. The end of the shaft has a tapered section that
allows for the attachment of the primary pulley of the
system. See Figures 1, 2, and 3 for models and actual
assembly of this adapter. This design transfers power
directly from the crankshaft of the engine which provides
many advantages. There is no power lost due to
mechanical systems such as gears. Also, the CVT
system operates at speeds comparable to that of
conventional 2-stroke snowmobiles providing for the use
of conventional clutch components, as well as design
When selecting a motor for the 2005 CSC, the team
drew on their previous success by using Honda inline 4cylinder, 4-stroke, spark ignited engines. In order for the
team to achieve its goals for 2005, they selected a
Honda CBR 954RR engine. This engine has a
displacement of 954cc, is spark ignited, and naturally
aspirated. The motor is rated at 114.8 kW (154 Hp) while
maintaining nearly the same external dimensions as the
previously used smaller displacement Honda CBR
600F4i engine. The larger displacement engine was
chosen due to its high power and torque output which
allows the motor to operate in a lower rpm range while
exceeding the performance of the smaller displacement
engine. This choice benefits the design in many ways,
including the advantage of better fuel economy, lower
emissions output, and lower noise output, while still
achieving high performance capabilities.
Figure 1: PTO Adapter Model
In order to fully utilize the motorcycle engine,
modifications had to be made to transmit power from the
engine to the ground. These modifications include the
Power Take Off (PTO) adapter and the semi-direct drive
PTO Adapter
Figure 2: Coupler Portion Installed
Power had to be transferred from the motorcycle engine
to the ground via the CVT system. This required the
design of an output system enabling the mounting of the
primary drive pulley.
previous design assumptions used on the CBR 600
motor adapter which is still in service with no evident
signs of failure. The FEA results may be viewed in
Figure 5.
Figure 5: FEA Results of Adapter
Figure 3: Complete Adapter Assembly
As previously mentioned, the adapter is also used as a
flywheel, thus reducing design complexity so that
precision must be met while balancing the assembly.
This was obtained using a Magna-Matic 7000 Series
balancer which utilizes a free spinning shaft and gravity
to allow for balancing of the part. This tool can be seen
in Figure 4.
Figure 4: Magna-Matic Balancer
The system was designed for infinite life with a load of
maximum belt force constantly applied to the end of the
shaft and the torque applied by the engine.
Pro/Engineer modeling was done for visualization and
interference checking before machining the parts. Both
portions were machined from billet using a ComputerNumerically Controlled (CNC) lathe and mill. The steel
shaft portion was then heat treated to provide added
strength and durability.
The completed analysis for the coupler/shaft assembly
includes estimated life due to a fluctuation moment on
the portion of the shaft extending out of the cover, as
well as a Finite Element Analysis (FEA) of the complete
assembly. For the fatigue calculations, infinite life was
desired and a safety factor of 1.5 was built into the
calculations and the minimum shaft diameter was solved
for. The FEA analysis was done using the solid model
and computer software. Loads assumed were a 400 lbf
on the end of the shaft in the radial direction due to belt
forces seen by the primary pulley, and a pure torsion
force of 136 N-m due to the torque produced by the
motor. These forces were assumed in light of the
The second component involved in the transmission
adaptation is a support cover which incorporates a
bearing that supports the aforementioned shaft. The
cover replaces the existing Honda alternator cover of the
engine. First of all, this particular cover had to be
designed to be load-bearing, primarily in a radial fashion
with respect to the bearing; therefore, a structural
analysis had to be preformed. Secondly, the cover had
to interface with the existing bolt pattern and sealing
surface on the Honda motor. In order to achieve this,
precision measurements were made using a coordinate
measuring machine (CMM). The coordinates were
subsequently used to aid in the design of the part and to
manufacturing. Third, the cover had to allow adequate
inner clearance for the shaft and coupler assembly.
Fourth, the cover had to incorporate a seal to prevent oil
from escaping the engine. Lastly, the cover had to be
designed in such a fashion that we will be able to be
machined. The new engine cover was modeled in IDEAS using the CMM acquired coordinates. All of the
constraints and requirements stated above were taken
into consideration in the design of the cover. A photo of
the design can be seen below in Figure 6. The cover,
made from 6061-T6 Aluminum, was machined using a
CNC horizontal mill.
Figure 6: Photo of Completed Support Cover
FEA was also used in order to validate the design. The
main concern was assessing the structural integrity of
the cover, so it was decided to use the Finite Element
method, via IDEAS. For the FEA, the radial input load at
the bearing surface was 3 kN (based on worst-possiblescenario loading conditions). This loading condition was
a worst-possible-scenario condition, that is, the existing
bearing just inside the support cover couples the entire
moment caused by the belt tension. It was assumed that
there would be no axial forces at any time during
operation. The bolt holes were fixed (no displacement
allowed), but rotation was permitted. FEA results can be
seen in figure 7.
Figure 7: FEA Results of Support Cover
FEA results for every component can be found below in
Table 4. The values shown are Von Mises stresses and
represent the average stresses on the component.
Table 4: Results of FEA Analysis
4140 +
Semi-Direct Drive System
Due to the orientation of the engine, exhaust facing
rearward, and the position of the crankshaft adapter, the
primary pulley spins the opposite direction of that
required to propel the snowmobile forward. For this
reason a FAST Inc. gearbox was chosen. This gearbox
utilizes a two gear mesh which reverses the direction of
rotation in order to properly drive the snowmobile. The
gearbox also provides a gear reduction of 2.1:1, thus
offering an advantage in torque present to propel the
A semi-direct drive system provided
benefits in several other areas, including overall weight
reduction of the drive system, increased reliability
through the elimination of the chain and sprockets
(common failure components), compact packaging, and
increased safety with the placement of the brake on the
driveshaft. See Figure 9 for a picture of the complete
The primary clutch is an eight inch diameter Micro
Belmont Reactor Four Tower. Common to that of the
Polaris P-85, this clutch offers a wide operating range
while still maintaining excellent tuning characteristics
needed for clutching in an application such as the 2005
design. The Micro Belmont has a peak operating range
of 14,000 revolutions per minute (RPM) which is much
higher then a stock primary of around 9,000 RPM. The
MTU design operating range is 10,000 RPM, thus
justifying the need for the precision machined and
balanced Micro Belmont. The Micro Belmont is tunable
using four variables: weight profile, weight mass, pin
mass, and spring stiffness. These factors will allow
achievement of the peak operating range, cruising RPM,
and the desired engagement RPM.
For the secondary clutch, a TEAM Industries Rapid
Reaction with a 27.3 cm diameter was chosen. The
TEAM secondary utilizes a dual roller mechanism
coupled with progressive angle helixes to offer
exceptional efficiency and quick back shifting
capabilities. The small 30 cm clutch center to center
distance, due to the low engine placement and gearbox
location constraints of the chassis, provide for the use of
a short, but standard, and readily available belt of 116.2
cm inches in length.
Clutching components, including primary springs,
weights, and pins, were used during testing to tune the
drive system. The system was required to operate
between 3,000 and 10,000 RPM. A typical snowmobile
drive system operates between 4,000 and 8,000 RPM.
For this reason, several components were designed and
manufactured in order to obtain the desired engagement
up shift characteristics and full shift RPM. Optimized
weight profiles were designed and utilized to obtain a
well-balanced system. The setup of the Micro Belmont
allowed use of the full RPM range with and engagement
rpm of 3,000 and a peak of 9,500 RPM desired for
competition. The appropriate back shifting was achieved
with the TEAM secondary to allow an optimal cruising
RPM of 5,000. This configuration allows for lower RPM,
which causes improved fuel economy and decreased
noise levels at cruising speeds, while still maintaining
excellent acceleration and top speed.
With the incorporation of the gearbox, a driveshaft was
also designed and fabricated. Splines were cut into the
shaft at the gearbox end to mach the lower gear of the
gearbox, and keyed at the other for the use of the brake
To prevent failure of the gearbox input shaft, a support
arm was designed and fabricated. It was designed to
support both clutches, eliminating the individual motion
of the clutches. This will ensure that the clutches align
under all operating conditions and reduce the moment
caused by the belt force on both the gearbox input shaft
and the engine output shaft. This is a crucial part of the
system due to the high forces seen by both shafts from
the belt force of the high torque motor. The support was
made from 6061-T6 aluminum with a bearing pressed
into each end. See Figures 9 and 10 for pictures of the
support and the FEA analysis performed to ensure that it
would in fact provide the needed support to permanently
affix both clutches. It greatly reduces forces on both
shafts and also keeps the clutches perfectly aligned,
thus greatly increasing efficiency of the overall system.
the brake location also provides greater control when
slowing the sled. This brake location also offers the
assurance of safety, giving the rider the ability to apply
braking force at all times under all conditions, whereas
the traditional chaincase brake does not. If a chain or
gear failure occurs in the chaincase, the brake is
rendered useless and the rider has no control over the
speed of the snowmobile. Applying braking to the
driveshaft eliminates these dangerous scenarios. The
brake used in the 2005 design is a Wilwood caliper and
rotor. The rotor has been machined to a diameter of 18.1
cm to reduce the drop in the belly pan to accommodate
for rotor position. The caliper was then moved closer to
the tunnel to maintain over 90 percent of the stock
braking surface.
Modifications had to take place to make the engine fit
into the chassis in a manner that least effected the
handling of the chassis. These modifications include the
exhaust/seat lift and remote oil filter assembly.
Exhaust System
Figure 9: Semi-Direct Drive System
Figure 10: FEA of Clutch Support Arm
For the redesigned drivetrain, the brake system was
moved from its standard location on the jackshaft to the
driveshaft where it is positioned opposite the secondary
clutch. Moving the brake system to the driveshaft places
weight lower in the chassis, improving the vehicle’s
center of gravity. Due to the absence of the reduction
found with a chaincase that affects conventional braking,
The exhaust system used on the CBR954RR engine
incorporated an exhaust valve in the headers that
changed the collector phasing between cylinders. The
utilization of this valve would not work in the 2005 design
since the height of the exhaust needed to be minimized
due to the limited clearance with under the gas tank
exhaust routing. It was required to design a completely
new exhaust to be fabricated. There were several
constraints in the design. The stock length of the
headers in the RPM ranges of 3,000 to 7,000 RPM was
85.09 cm to the first collector and 36.83 cm to the
second collector. This resulted in an overall length of
121.9 cm from the head to the last collector. This put the
last collector less than 30.48 cm away from the end of
the tunnel, not allowing enough distance for overall fit
with 30.48 cm needed for the catalyst and 30.48 cm for
the muffler.
The design of the exhaust utilized Lotus Engine
simulation software. A model of the engine was created,
inputting all of the specific values for the engine such as
bore, stroke, cylinder phase, cam lift, number of valves,
valve area, and intake type. Using this model,
simulations were run to find the optimum length of the
primary and secondary header lengths. The simulation
was setup allowing for a change of the primary and
secondary pipes by 1.27 cm in variations between 1 and
63.5 cm for the primary, and a range of 2.54 to 63.5 cm
for the secondary, keeping the maximum total length to
36 inches. Upon interpreting the simulation results and
noting that the RPM range of 3000 to 7000 is of most
importance in the CSC, the lengths of 53.34 cm for the
primary and 38.1 cm for the secondary headers were
chosen. See Figure 11 for the simulation model used.
To route the exhaust system to the rear of the
snowmobile and still be able to drive the snowmobile in a
comfortable riding position, a new seating area had to be
constructed above the exhaust. A seat frame composed
of C-channel aluminum and aluminum plating was
placed below the seat. A Polaris ProX2 seat was
hollowed out to provide more room for the exhaust. The
Polaris seat provided a comfortable seating position as
well as the ability to remove some of its material. The
gas tank that best suited our design was one from a
Yamaha Rx-1 snowmobile. The Yamaha gas tank has
the bottom of the tank curved upward in the middle to
provide room for the exhaust, which is also routed
beneath the tank/seat on Yamaha’s snowmobile.
Additionally, the gas tank is on a riser made of Cchannel aluminum.
Figure 11: Engine Simulation Model
Stainless steel was the material of choice for the
construction of the exhaust.
Given the high
temperatures due to an optimally running engine and the
closed space that the exhaust would be housed in, along
with the cost, it removed the possibility of using Titanium
for construction.
The welding was done using TiG construction for best
penetration and appearance. Four flex pipes were
installed in the headers to limit the amount of flexing and
possible breaking of the exhaust due to the flexibility of
the snowmobile chassis. After the construction was
completed, the exhaust was thermal barrier coated,
resulting in greater exhaust gas velocity and higher
efficiency. Another reason for the coating was to reduce
the amount of heat radiated from the headers to the gas
tank. The coating was completed using White Lightning
from Swain Tech. The TBC-EX coating is a 3-layer
0.381-0.508 mm thick permanent coating. The coating is
pearl white in color and extremely durable. According to
Swain Tech, it reduces radiant heat by more than 50
percent. Refer to Figure 12 for a photo of the exhaust
Figure 12: Exhaust System
The new seating system allows for the exhaust to be
routed below the rider and places the rider in a more
aggressive riding position. The clearance underneath the
seat also allows for the design of a larger muffler to be
used for noise control.
Another main concern for the exhaust system was heat
generation. Type-S Kaowool ceramic fiber insulation,
one inch in thickness, was used between the exhaust
area and the seat and gas tank. The seat and gas tank
are underlined with a floor and tunnel heat shield which
will reflect heat as well as reducing up to 50 percent of
unwanted noise. The combination of these materials
provides heat insulation needed to protect the
components of the snowmobile along with the rider.
Two electric fans, similar to those used in computers,
are also incorporated into the system. The fans are
positioned at the front of the gas tank blowing in, and in
the rear blowing up and out of the exhaust area. This
allows for better circulation of unwanted hot air under the
gas tank and rider.
Because the airbox is mounted above the head in it's
motorcycle application, the intake runners are more
vertical than would be ideal for a snowmobile. Design
and manufacture of completely new intake runners and
air box was not feasible and therefore the stock
configuration had to be modified. The first modification
made was the redesign of the airbox cover due to its
interference with the steering post. The new design
uses commercially available intake runners that are 72
mm shorter than the original set and are designed to
increase power at high RPM. This change allowed for
the new airbox cover to incorporate a depression that
eliminated the interference issue with the steering post.
Also modified was the way air entered into the air box.
The stock inlets pointed down, and utilizing these inlets
would result in the engine being fed with heated under
hood air, degrading engine performance. Also, the stock
inlets would not fit due to gas tank location. A new air
intake to the air box was made using the same cross
sectional area of the combined dual air intake.
Construction was accomplished utilizing Carbon Fiber
material for its large strength to weight ratio, ease in
forming, and the fact that it would provide less noise
than an aluminum intake that has been attempted by the
team in years past. See Figure 13 for a photo of the
modified airbox.
Figure 13: Modified Airbox
stock snowmobile chassis was modified to incorporate
this engine. 4130 cro-moly tubing with a diameter of
2.54 cm was used for engine mounting due to its
strength, light weight, and ease of manufacture. With all
of the other modifications done to the engine, oil system,
and intake system, the engine was able to be placed as
low and as far back as needed. The motor mounts were
designed to incorporate the clutch side of the gas tank
mount/steering hoop mount and then bolted into the
snowmobile bulkhead and tunnel.
Also incorporated in the design were the shock towers
where all of the front suspension forces enter into the
chassis. The mounts tied into the engine, utilizing 6 of
the stock mounting locations, and were designed by
taking into account removal of the engine. The mounts
bolt to the engine outside of the snowmobile, and then
the assembly is attached. The unit can be removed with
the unhooking of limited wiring and only two coolant
lines. This should ease any possible engine work,
although none is anticipated.
Remote Oil Filter and Cooler
Heavy modification took place with the development of
the oil system. On a stock motorcycle, the oil cooler and
oil filter are mounted in the front of the engine behind the
headers. With the team’s adaptation of putting the
motorcycle engine in the snowmobile, the system
needed to be relocated, allowing the engine to be moved
back towards the center of the snowmobile. This was
accomplished by making oil distribution blocks. The
length of the cooler is 7.62 cm and the oil filter has a
length of 12.7 cm. The design was made with a total
length of 3.81 cm from the back of the engine. This is
3.81 cm shorter than the design used in previous
competition snowmobile. The oil blocks were made of
aluminum because of the light weight and ease of
machining needed. It was decided that the location of
the oil cooler and filter would be at the front of the engine
mounted to the shock tower cross brace of the engine
Modification also took place with the oil pan. The stock
version had the sump located directly where the main
support for the chassis was located with the final engine
placement. The oil pan was modified to allow for
clearing of the chassis structure. With the oil pan
modified for clearance and mounting of a new oil drain
plug, the oil sump pickup then needed to be modified.
The sump was moved out 5.08 cm and back towards the
rear of the sled by 3.81 cm.
These overall oil system changes allowed the engine to
be moved an additional 16.51 cm rearward and 7.62 cm
downward. This move resulted in a much lower center of
gravity and helped to reduce any negative effects with
the larger and heavier 4 cylinder engine.
Engine Mounting
One of the main problems with installing an inline 4
cylinder motorcycle engine is the sheer size of it. The
Incorporated into the motor mounts is a mounting
location for the oil distribution block with the oil cooler
and filter, as well as a mount designed to hold the
bottom of the steering shaft. Issues have arisen in past
years with not properly supporting the u-joint on both
sides. The 2005 mount allows for no flex in the u-joint
and removes the need for a separate support. This
saves weight and makes for a tighter and more compact
package. See Figure 14 for the completed engine
mounts and all associated parts.
Figure 14: Engine Mounting Frame Installed
To improve performance, the power to weight ratio can
be increased by either increasing power or decreasing
weight. Reducing weight is a popular choice. Utilizing a
four-stroke engine for a power-plant is a disadvantage
because it has a greater mass than a two-stroke engine
of the same displacement. To counteract this obstacle,
the team reduced weight in two main areas: the engine
and the drivetrain.
Engine Mass Reduction
The conversion from a two-stroke engine to a four-stroke
engine inherently results in an overall engine weight
increase. This is due to the fact that a four-stroke engine
produces power every two revolutions of the crankshaft,
where a two-stroke engine produces power every
revolution of the crankshaft. Typically, to achieve
equivalent power outputs between two-strokes and fourstrokes, the later is twice the displacement of the twostroke engine.
To reduce the mass of the engine, the transmission
components were removed. The removal of these
components resulted in a reduction of 62.7N of rotating
Drivetrain Mass Reduction
To reduce drivetrain weight, the team implemented a
gearbox which eliminated the jackshaft and placed the
gearbox lower in the snowmobile chassis providing for
reduced and lowered mass.
In testing last year’s cooling system, it was found that
the stock motorcycle radiator would not provide enough
cooling capacity given the high load characteristics of a
snowmobile and the limited airflow through the radiator
due to poor ducting of fresh air and limited speeds of a
snowmobile. In choosing a radiator, the stock motorcycle
radiator was deemed inadequate and a radiator of larger
size was needed. The stock radiator had an area of 987
square cm with a length of 45.72 cm and a height of
21.59 cm. The thickness of the radiator was 1.9 cm. The
chosen radiator has an area of 1239 square cm, with a
length of 16 inches and a height of 30.48 cm. The
thickness of the radiator is 5.08 cm. When comparing
the two radiators, it was found that the radiator chosen
covered 25 percent more area and had a volume 200
percent larger than the stock motorcycle radiator. A
22.86 cm DC electric fan, controlled by the Engine
Control Unit (ECU), provides airflow to the radiator at low
speeds. The fan turns on at a water temperature of
approximately 101 degrees Celsius, and turns off at 99
degrees Celsius. This total system provides more than
adequate cooling for the 2005 high performance engine.
Electrical System
Total Vehicle Mass
Table 5 compares the weight of the 2005 Michigan Tech
four-stroke snowmobile and the 2004 Michigan Tech
four-stroke snowmobile to an average 600cc two-stroke
snowmobile. As the table shows, the 2005 four-stroke
design is superior in the area of specific power, even
when compared to a production two-stroke snowmobile,
and far surpasses designs from years past.
Table 5: Weight, Power and Specific Power:
Comparison of Three Snowmobiles
2005 MTU 954cc
2004 MTU 600cc
600cc two-stroke
Engine Cooling System
The cooling system for the 2005 design uses an electric
water pump, electronic pump controller, and a radiator.
In past years, cooling was accomplished by utilizing the
stock snowmobile cooling system of front, tunnel, and
rear heat exchangers. Due to the exhaust placement
and chassis modification, the snowmobile cooling
system was discarded and a radiator was implemented.
Due to the CVT adaptation design, the stock
flywheel/stator assembly was removed in order to
provide a mounting location for the adapter shaft. This
required a new means of generating electrical power. It
was decided that the power would be generated from an
alternator driven by a belt with a pulley attached to the
primary clutch. The flywheel/stator assembly is rated at
50 amps at 5,000 RPM. Due to the added current draw
caused by the mounting of multiple fans, electronic water
pump, and hand warmers, the stock rating was used as
the minimum rating acceptable for the system. The new
design incorporated a 60 amp mini alternator that
features a one wire hookup due to its internal regulation.
The alternator is designed to start charging around 2,500
RPM and its upper limit is 10,000 RPM. The alternator
will not tolerate an RPM higher than 10,000. A ratio of
1:1 would be appropriate given that the expected clutch
engagement would be above 3,000 RPM and the
maximum engine RPM for the design is 10,000.
Steering Ability
The steering system that was used in the 2004
competition proved to work sufficiently. For the 2005
competition, the team used the same principle, an over
the engine steering system. The adaptation of the over
the engine steering system is a little different from the
one used previously in the Arctic Cat chassis, being that
there wasn’t as much room in the engine compartment
as that of the new Polaris chassis. The steering post is
routed over the engine and is linked with a universal joint
near the front of the snowmobile. This system
accommodated for the use of the stock steering rack and
tie rods. The approach angle to the rider positions the
handle bars in a comfortable location, allowing for a
smooth handling snowmobile. See Figure 15 for a photo
of the complete steering system as mounted in the
in the following table. The testing was done with a
maximum engine RPM set at 9000rpm. For each of the
3 modes, exhaust gas emissions readings were taken
using a 5-gas analyzer testing for HC, CO, and NOx
emissions. Readings were taken two minutes after
desired rpm and throttles settings were reached to allow
for any settling or delay of the analyzer readings. See
Table 6 for definitions of the modes used.
Table 6: Mode Definitions
Mode 5
Mode 4
Mode 3
Percentage of Maximum
Figure 15: Installed Steering System
With the introduction of the subjective handling and
human exposure to whole body vibration events for the
2005 competition, the sled is designed to handle and
drive as similar to a production snowmobile as possible.
This design was implemented using specific seating and
handlebar locations, along with the incorporation of Fox
FLOAT shocks. While providing a wide range of settings,
the Fox shocks also reduce vehicle mass by
incorporating an air spring and eliminating the standard
steel coil spring.
In order to reduce the emissions of the engine, the
addition of a catalyst and fuel injection tuning were
utilized in meeting our goals for the 2005 competition
sled. The catalyst chosen is a three way catalyst, which
will help in the reduction of hydrocarbons (HC), carbon
monoxide (CO), and nitrogen oxides (NOx). In order for
the catalyst to function properly the engine needs to be
properly tuned. The three way catalyst requires the
engine to operate around the stoichiometric value of
gasoline, which is 14.7 mass parts fuel to one part air.
This value represents the ideal air to fuel ratio in which
the catalyst will operate most efficiently. Fuel Injection
tuning was accomplished using a DynoJet Power
Commander. This unit allows for the changing of fuel
and ignition maps easily using a laptop computer, while
keeping the robust integrity of the stock Honda PGM-FI
system. In order to test steady state conditions under
different load conditions a Land- and-Sea dynamometer
was used. This system was setup utilizing an electronic
servo load valve in order to accurately set specific rpm
and load conditions.
In doing the emissions testing, a system was setup using
3 different RPM and load conditions which can be seen
Three different engine setups were tested to be used in
comparing the effects upon emissions related
components. The first test was done with the only
included modification of the engine being the exhaust
and intake. The stock motorcycle muffler was used in
order to limit any backpressure changes and therefore
stage of tuning of the stock system. The second test
setup included the addition of the three way catalyst,
again utilizing the stock motorcycle muffler. The third
test setup incorporated the addition of the custom
designed muffler and tuning of the fuel injection system.
The results of the emissions testing can be seen in
Figures 16, 17, and 18.
Hydrocarbon Emissions
Suspension/Ride Quality
Stock Honda System
Addition of Catalyst
Catalyist w ith Fuel
Injection Tuning
Figure 16: Hydrocarbon Emissions; Stock Honda
system data for Mode 5 is off chart, value is 3300.
Carbon M onoxide Em is s ions
Stock Honda System
Addition of Catalyst
Catalyist w ith Fuel
Injection Tuning
The three main noise sources from a snowmobile are
the engine intake, the engine exhaust, and the track and
suspension. By analyzing each source and treating each
component separately in noise reduction strategy, the
team felt that the highest level of success would be
M ode
Figure 17: Carbon Monoxide Emissions
NOx Em is s ions
Stock Honda System
A ddition of Catalyst
Catalyist w ith Fuel
In the 2004 CSC, the team goal was to win the noise
event and achieve a score of 105 dBA. While the team
did succeed in winning the “2004 Quietest Snowmobile
Award,” the goal of 105 dBA was not reached. To
achieve this goal for 2005, a new strategy was
developed for the new engine and chassis package to
focus on noise control throughout the design and
construction of the snowmobile.
Exhaust Noise Reduction
The use of resonators in the exhaust system is effective
in removing dominant frequencies of noise produced by
the exhaust of the engine. Devices such as Helmholtz
Resonators are incorporated into the exhaust system to
actively cancel out problematic exhaust frequencies.
Figure 18: NOx Emissions; note Stock Honda
system NOx data for Mode 3 and 4 is off chart,
values are 560 and 922 respectively.
In comparing the results of the data for the three types of
emissions, an interesting note to mention is the fact that
when the catalyst was added, emissions increased for
CO. This is probably due to the fact that the catalyst is a
large restriction in the system and does not allow the
engine to operate efficiently using the stock Honda fuel
mapping. The last configuration shows a noticeable
decrease in emissions largely due to the fact that the
fuel injection system was tuned to make the catalyst
operate in its most efficient state of an Air/ Fuel ratio.
Refer to Table 7 for percent reductions in emissions.
Table 7: Percent Reductions in Emissions
Percentage Reduction in Emissions
Catalyst and Fuel
Injection Tuning
To maximize noise cancellation, mufflers were designed
around the frequency output of the engine exhaust. A
primary muffler acts as a low-cut filter by removing low
frequencies, and a secondary muffler acts as a high-cut
filter by removing higher frequencies. The secondary
muffler is simply a tube insulated with fiberglass around
the outside. The primary muffler is a chamber type
muffler, which uses a series of tuned chambers and
baffles to cancel out certain exhaust frequencies.
The first step in designing the muffler was to analyze the
noise of the engine exhaust. Using a Land & Sea water
brake dynamometer for engine loading, the conditions
that the engine will see during the noise event were
simulated. A 01dB brand microphone and Symphony
data acquisition software were used to analyze the
sound output of the engine. The exhaust noise was
analyzed for frequency content at an engine speed of
5000 RPM, which is the speed while the snowmobile is
moving at 40 miles per hour that of the fuel economy or
“cruising” speed as supplied in the rules for the 2005
CSC [6]. This speed also falls within the range of
speeds for the noise event test speeds of 35 to 55 miles
per hour. The results of the frequency analysis can be
seen in Figure 19.
With Muffler
Sound Pressure (dB)
10 0
12 0
16 0
20 0
Frequency (Hz)
Figure19: Exhaust Frequency Content at 5000
Engine RPM without Muffler
Figure 20: Exhaust Frequency Content at 5000
Engine RPM with Muffler
From Figure 19, the “peaks” in the exhaust frequencies,
or the problem frequencies can be seen. The main peak
in Figure 22 is at 315 Hz. This band makes up part of a
wider peak that ranges from 315 Hz to 1000 Hz. The
secondary peak occurs at the 160 Hz to 200 Hz band
level. These two ranges are where the muffler is
designed to reduce sound pressure levels the most.
Using the temperature of the exhaust, the speed of
sound, c, for that medium was found to be 1200 m/s.
The wavelength corresponding to the dominant
frequencies can be found using equation 4.
Since the exhaust layout is under the seat, the muffler
needs to be compact so it doesn’t interfere with the seat
lift or with the header and catalytic converter system. For
this reason, the team chose to use a combination type
muffler, or one in which the primary or reactive muffler,
and the secondary or absorptive muffler, are both
housed within the same component.
Wavelenth, λ =
Each baffle is a quarter-wavelength tube so as the
sound wave destructively interferes with itself after
reflection. Therefore the baffle length is calculated by
dividing the dominant wavelength by four. Photographs
of the baffles used for the muffler can be viewed in
Figure 21.
To increase ease of packaging decided to use a
combination type muffler, which was designed around
these frequencies.
Both reactive and absorptive
mufflers are combined in this design. The absorptive
muffler portion is a perforated tube with fiberglass
packing surrounding it. The reactive muffler involves a
series of baffles and chambers. For this application, a
combination of three chambers and three baffles were
chosen. The baffle length is tuned to eliminate the
dominant frequencies in the noise. The test results with
muffler installed are shown in Figure 20. The results
indicate that the problematic frequencies were
attenuated as planned, as well was an overall
attenuation across the spectrum
Figure 21: Muffler Baffles
Once the exhaust exits the muffler it is directed through
the tunnel and towards the track of the snowmobile.
This leads to the absorption of any noise exiting the
muffler by the track, snow, and ground.
Intake Noise Reduction
To gain the most intake noise reduction as possible,
resonator chambers were once again implemented, and
sound-dampening material was utilized under the hood.
capable of driving the track in a controlled manner due to
its high torque rating. This test stand can be viewed in
Figure 22.
The Honda intake system utilizes resonator chambers
both in the air box as well as on the air ducts. Both of
these systems were retained in the design of the 2005
The sound dampening material used in the engine
compartment is SoundProof brand acoustical foam. It is
a dense, egg-carton type foam insulation. This foam
allows for maximum absorption of sound waves, while
still being fire resistant. To aid in the foam’s sound
absorbing ability, as many vents as possible in the hood
and belly pan were closed off. This creates an anechoic
environment in the engine bay and prevents noise from
escaping the engine compartment.
Figure 22: Redesigned Suspension/Track Test Stand
Track and Track Interface Noise Reduction
One of the dominant noise sources of a snowmobile in
motion comes from the track area. However, the exact
cause of the noise is somewhat unknown. Possibilities
include the idler wheel contact on the track, the metal
clips contacting the rear suspension components, or the
track rotation itself.
Variables such as track clips, compounds, lug height,
windows, and tension all impact the noise output of the
track. Test data provided the necessary information to
begin to understand what modifications could be done to
reduce chassis noise levels. Track tension was found to
have a direct relation to track noise, which means that as
tension increases, so does track noise. Therefore, the
track is run with as little tension applied to it as possible.
Two other areas that needed to be analyzed in depth
were the affect of the track clips and track compound on
noise production, as they have the greatest influence on
chassis noise production. The following information will
allow the future design and implementation of quieter
chassis components.
To explore the discussed track possibilities, a test was
designed in 2003 to compare systems with varying
aspects in these areas. First, a stand was constructed
that supported a snowmobile tunnel. An electric motor
was mounted to this tunnel in a way that it could provide
power to the driveshaft. With this setup, suspensions
and tracks and the combination of the two could be
tested for variations in noise levels. Sound pressure
levels were taken from 1 meter in areas around the
Two different Camoplast tracks were tested with the new
test stand. The tracks each had similar characteristics
including clip pattern, compound, lug style, and basic
features. Both tracks had constant lug height. Table 8
compares the three different tracks.
Table 8: Tracks Tested for Noise
Lug Height
Lug Type
Clip Interval
3 windows
Bump Track
3 windows
Using a Sound Level Meter (SLM) in a quiet, controlled
setting, the tracks were all run at 235 RPM. Measuring
the sound level at three locations around the stand, and
averaging the results after three trials, the results found
in Figure 23 were obtained.
2005 Track Test
Bump Ripsaw
Standard Ripsaw
For 2005, the test stand was modified to work with a
variety of rear suspension designs. This new stand was
more stable, more rigid, and provided the ability to
interchange suspensions and tracks easily. A Leeson,
one horsepower, ball-bearing DC electric motor was
used to drive the track on the test stand. This motor was
Track Number
Figure 23: Track Noise Output Results
As shown in Figure 25, the two tracks provided identical
noise output. For the 2005 design, the Bump Track was
chosen based on the fact that it was designed to reduce
noise by preventing the idler wheels from contacting the
belts inside the track.
When the results of the dynamic properties test were
completed, the hardest urethane was chosen to prevent
excess engine movement under load. This was done to
prevent misalignment between the primary and
secondary clutches.
Further research on noise included the design of engine
mount bushings to reduce the transmittance of vibration
from the engine to the chassis. The goal of this project
was to reduce the amount of vibration transmitted from
the engine to the chassis without sacrificing the rigidity of
the engine mounts and risking clutch misalignment. The
mount bushings are incorporated into the motor mounts.
This prevents the motor from shifting when the soft
mounts are used.
While designing, fabricating, and refining the
snowmobile for the 2005 Clean Snowmobile
Competition, the Michigan Tech team kept the consumer
in mind. The team wanted to produce a vehicle that was
designed for snowmobile rental agencies and personal
consumers alike. Cost, durability, fuel economy, comfort,
ride, and cold engine starting were the most important
characteristics for this market.
Urethane was used due to its damping and spring
properties, as well as for it’s predominate use and
reputation in the automotive industry. Three different
durometer urethane materials from Sunray Inc. were
selected for testing. The finished bushings can be
viewed in Figure 24.
With rental operators having to cover the cost of a rental
snowmobile up front before any kind of profit can be
made, the initial cost of the snowmobile must be
relatively low. Also, in order for a consumer to buy a
snowmobile, the machine must be economical while still
providing all comforts and accessories to the
The cost of the Michigan Tech snowmobile over a
conventional snowmobile, as determined from the CSC
2005 Technology Implementation Cost Assessment
(TICA) form, is $1314.25. This additional cost compared
to current expenditures could be recouped by the rental
businesses and customers from reduced maintenance
costs, reduced oil consumption, higher durability and
lower fuel consumption that are inherent in a four-stroke
Figure 24: Urethane Bushings
After using a turning operation to obtain the correct size,
the three different materials were dynamically tested at a
displacement of 1 mm to determine their dynamic
properties. A photograph of this testing can be observed
in Figure 25.
Honda motor products are well known for their durability
and reliability. The CBR954RR is no exception. This
engine undergoes rigorous durability tests by the
manufacturer and has also been extensively tested by
the team. Many hours of dynamometer testing as well
as over 200 miles of actual riding has been done to test
the engine and the overall design of the snowmobile.
Electronic Fuel Injection continuously optimizes the
amount fuel delivered to the engine, thus maintaining a
consistent air to fuel ratio and increasing fuel economy.
Figure 25: Dynamic Testing of Urethane Bushing
By optimizing the air/fuel ratio throughout the fuel map
and avoiding rich conditions, a minimal amount of fuel is
used during combustion. This condition benefits the
consumer, as the rider will be spending less money on
fuel each time he/she rides the snowmobile. It also
makes the machine a very clean burning and
environmentally friendly machine.
The team placed great emphasis on the overall ride
quality of the snowmobile. In designs of years past, ride
quality was not emphasized nearly as much. The 2005
snowmobile will treat every rider to the gentlest of rides
while still maintaining a high level of performance. This
was accomplished through the incorporation of Fox
FLOAT shocks into the tried and true Polaris front
suspension. This allows the rider to adjust the damping
of the suspension with a quick air pressure adjustment,
making it easy to tune to varying riders and trail
conditions. The rear suspension utilizes shocks capable
of adjusting damping as well. These factors combined
with design parameters that include low center of mass,
and comfortable seating and steering position give the
snowmobile an outstanding ride quality.
The snowmobile designed for the 2005 CSC is not only
clean and quiet, but also very performance oriented.
Four-stroke snowmobiles would be better accepted by
the snowmobile community if they possessed equal or
better performance qualities to the two-stroke machines
that made snowmobiling the popular sport that it is. The
2005 Michigan Tech entry in the CSC is able to do this.
Using an engine capable of producing 114.8 kW, the
snowmobile can be an exciting machine even to the
most veteran riders. Combined with simple and fast
suspension tuning, the snowmobile can easily adapt to
various riding conditions.
Starting a cold snowmobile can also prove to be a
challenge to riders. When a cold start test was
conducted, the snowmobile started in less than 3
seconds when exposed to an environment with
temperatures averaging -5 degrees Fahrenheit for a
duration of 7 hours.
snowmobiling enthusiasts have come to demand. This
machine represents the future and longevity of a sport
that continues to grow in numbers each and every year,
guaranteeing that snowmobiling will be enjoyed by
generations to come.
1. Haines, Howard, “The Snowmobile Dilemma or Who
Spilled What in the Refrigerator vs. Who's Going to
Clean It Up?”, 1999, Montana Department of
Environmental Quality
2. Heywood, J.B., “Internal Combustion Engine
Fundamentals”, McGraw-Hill, New York, 1988.
3. PWIA, “Outboard/PWC Engine Technologies &
Water Quality”.
4. National Park Service, U.S. Department of the
Interior. Winter Use Plans: Final Environmental
Impact Statement for the Yellowstone and Grand
Teton National Parks and John D. Rockefeller, Jr.,
Memorial Parkway. October 2000.
5. United States Environmental Protection Agency
“Emission Standards for New Non-road Engines”
6. “The SAE Clean Snowmobile Challenge 2005
Rules.” 2004.
7. Barr, B., Cramer, R., Kayser, M., Luskin, L.,
Messenger, A., Pitcher, A., Bettig, B., “Design and
Development of a 4-Cylinder, 4-Stroke Powered
Snowmobile,” 2004
8. Barr,
Seidenstucker, J., Kallio, D., Hoffman, D., Bettig, B.,
Motorcycle Engine to Continuously Variable
Transmission for Snowmobile Application” SAE
9. Miers, S., Anderson, C., Hayes, R., Ballmer, J.,
Wegleitner, J., “Design and Testing of a Four-Stroke,
EFI Snowmobile With Catalytic Exhaust Treatment,”
SAE 2001-01-3657.
The 2005 Michigan Tech clean snowmobile is a reliable,
efficient, quiet, and excellent riding vehicle. It is a
product of an intense level of research and development.
The 2003 and 2004 Michigan Tech entries into the CSC
performed well, but at the same time, left a great deal of
potential unexplored. The 2005 entry is of the same
concept but is completely redesigned from the ground up
to defeat the faults of the previous versions. It continues
to improve on the incorporation of a high performance,
four-cylinder, four-stroke motorcycle engine into a
snowmobile application. The result is a vehicle that
defines the scope of research, development, and sheer
determination, a product of experience, education, and
dedication. The vehicle is substantially cleaner and
quieter than a stock two or four-stroke snowmobile, yet
it still maintains the performance characteristics that
Dr. Bernhard Bettig is an Assistant Professor in the
Department of Mechanical Engineering at Michigan
Technological University and the faculty advisor for the
MTU Clean Snowmobile Team.
ME-EM Department
Michigan Technological University
1400 Townsend Drive
Houghton, MI 49931
Phone: (906) 478-1897
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF