Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

SAE TECHNICAL
PAPER SERIES 2002-01-2758

Colorado State University

Clean Snowmobile Challenge 2002
Tim Bauer, Michael Duncan, Nathan Lorenz,
Dan Mastbergen, Steve Nelson, Nick Peterson,
Jessica Rupp, Brian Wedryk and Travis Mathis
Team Members, Colorado State University

Bryan Willson, Patrick Fitzhorn and Steven Schaeffer

Faculty Advisors, Colorado State University

Reprinted From: The Clean Snowmobile Challenge 2002

(SP-1726)

Powertrain & Fluid Systems

Conference & Exhibition
San Diego, California USA
October 21-24, 2002

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.

For permission and licensing requests contact:

SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: permissions@sae.org
Fax: 724-772-4028
Tel: 724-772-4891

For multiple print copies contact:

SAE Customer Service

Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-1615
Email: CustomerService@sae.org

ISSN 0148-7191
Copyright © 2002 Society of Automotive Engineers, Inc.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

2002-01-2758

Colorado State University

Clean Snowmobile Challenge 2002
Tim Bauer, Michael Duncan, Nathan Lorenz, Dan Mastbergen, Steve Nelson,
Nick Peterson, Jessica Rupp, Brian Wedryk and Travis Mathis
Team Members, Colorado State University

Bryan Willson, Patrick Fitzhorn and Steven Schaeffer

Faculty Advisors, Colorado State University

Copyright © 2002 Society of Automotive Engineers, Inc.

reduce noise, the team switched

from the use of three expansion
chambers (i.e. “triple pipes”) on the
stock ZRT600 to a single
expansion chamber. An exhaust
silencer was then fitted to reduce
exhaust noise, and a new intake
airbox was designed to reduce
intake noise. A layer of acoustic
foam was applied to the interior of
the engine compartment to reduce
radiated engine noise. An acoustic
analysis of the engine shows that
the noise level of the snowmobile is
dominated by noise at the engine’s
firing frequency and at twice the
firing frequency. Quarter-wave
reaction tubes were fitted to the
intake and exhaust to reduce
noise. The steps taken reduced
the noise from the unmodified level
ABSTRACT of 83 dBA to 75 dBA, which did not meet the competition
requirements of 74 dBA. Direct injection has improved
A student design team at Colorado State University the fuel efficiency from 13 miles per gallon (MPG) with
(CSU) has developed an innovative snowmobile to the unmodified engine to approximately 20 MPG. The
compete in the Clean Snowmobile Challenge 2002 performance of the engine has been improved. The
competition. The team selected a 600cc two-stroke peak horsepower has been reduced by approximately 4
cycle engine (Arctic Cat ZRT600) due to its favorable horsepower due to the switch from triple-pipes to a
power/weight ratio. In order to reduce emissions, the single pipe exhaust. However, the engine torque in the
team adapted the engine to operate with direct in- mid-range has been increased, and the throttle response
cylinder fuel injection, using the Orbital Combustion is improved across the entire range. These factors
Process (OCP) air-assisted fuel injection system. This combine to give the direct-injected snowmobile the same
conversion required that the team design and cast new elapsed times in a 500-foot acceleration test as the
heads for the engine. The direct-injection system unmodified snowmobile. In summary, the CSU
reduced carbon monoxide (CO) emissions by 70% and snowmobile has reduced emissions by 99%, reduced
total hydrocarbon (THC) emissions by 90%. An fuel consumption by 35%, reduced noise by 8 dBA, and
oxidation catalyst was placed in the engine’s silencer to improved overall performance and ridability. A cost
oxidize the remaining CO and hydrocarbons. The analysis has shown that the team’s technical approach
combination of direct injection and oxidation catalyst would add approximately $497 to the cost of
reduced both CO and THC by over 99%. In order to manufacturing a snowmobile.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

BASE SNOWMOBILE SELECTION

The CSU 2002 team has chosen the Arctic Cat ZRT600
as the base snowmobile (see Figure 1). The stock
engine is a three cylinder 1996 600cc two-stroke utilizing
a “triple-pipe” exhaust with a separate expansion
chamber and exhaust pipe for each cylinder. The team
has redesigned and made modifications to this
snowmobile to enter into the Clean Snowmobile
Challenge 2002.

PRINCIPLES OF THE TWO-STROKE ENGINE

The crankcase-scavenged two-stroke cycle engine has

Figure 1. Arctic Cat ZRT600 (base sled used by many performance characteristics that make it the
CSU in the CSC 2000 competition) engine of choice for snowmobiles. Each revolution of a
two-stroke engine produces one power stroke. This
INTRODUCTION
feature allows more power to be produced than a four-
stroke engine of equal trapped displacement, producing
There are currently no requirements limiting the exhaust
a high power-to-weight ratio. In two-stroke engines the
emissions and noise emitted by snowmobile engines.
piston is used to cover and uncover intake and exhaust
Carbon monoxide emissions from a two-stroke
ports. This eliminates the need for a camshaft, intake
snowmobile engine can be twice as high as those from
valves, and exhaust valves, with their accompanying
simple uncontrolled four-stroke snowmobile engines,
cost, complexity, and weight.6 This combination makes
and the hydrocarbon emissions can be 10-40 times as
two-stroke engines inexpensive, powerful, and easy to
high1,2,3. Two-stroke engines used in snowmobiles
maintain.
produce high noise levels, as high as 85 dBA or higher,
based on the SAE J192 test procedure4. The debate
THE EMISSIONS PROBLEM
over snowmobile use is extremely volatile.
Snowmobilers would like access to more national parks,
while other recreationists would like to see snowmobiles Two-stroke engines are currently used for the vast
banned from all national parks. Winter recreational majority of snowmobiles, and the CSU design team has
activities, the economies of surrounding communities, chosen focus on improving the two-stroke engine to
the snowmobile industry, and national parks themselves meet the competition objectives. The emissions from
will all be impacted by this debate. “conventional” two-stroke engines are dominated by high
hydrocarbon emissions produced by a process called
‘short-circuiting’. This process is mostly limited to
THE CLEAN SNOWMOBILE CHALLENGE 2002
‘cross-scavenged’ engines. A cross-scavenged engine
has exhaust ports that lie directly across the cylinder
The “Clean Snowmobile Challenge 2002” is an
from the intake ports. In a cross-scavenged engine, the
intercollegiate design competition focused on innovative
exhaust port opens first and closes last. Typical
solutions to reducing the impact of snowmobiles. Teams
snowmobile engines are carbureted, and use an air/fuel
are to redesign the configuration of current snowmobiles
mixture to “scavenge” the old exhaust products from the
to drastically reduce the emissions and noise while
cylinder. This allows 30%-40% of the fuel to be lost from
maintaining power, handling, safety, and endurance.
the engine during the scavenging process, (up to 70%
under some conditions) resulting in high hydrocarbon
The rules of the CSC 2002 competition have helped to
emissions and high fuel consumption. This process is
shape the team’s approach.5 Carbon monoxide and
illustrated in the top portion of Figure 2. Another source
hydrocarbon emissions must be reduced by a minimum
for two-stroke emissions is the oil premixed within the
of 50 percent each. The noise emitted from the
fuel. The oil is short-circuited along with the fuel,
snowmobile must be no greater than 74dBA, at full
producing further undesired emissions, as well as
throttle, 50 feet from the track.5 The solutions must be
drastically reducing the life of a catalyst.7
cost-effective in order to have the potential for
widespread implementation. The spirit of the
THE NOISE PROBLEM
competition is to generate a win-win situation for
snowmobilers, manufacturers, those concerned about
Noise is produced from many sources on a two-stroke
the environment, and all outdoor recreationists. The
engine. According to Blair,8 noise is primarily created by
development of a cleaner, quieter snowmobile will allow
pressure pulses in the intake and exhaust systems from
everyone to enjoy the wilderness to a greater extent.
the rapid opening and closing of the ports. These pulses
occur at the firing frequency of the engine, typically 250-
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

FUNDAMENTAL DIFFERENCE IN OPERATION BETWEEN CONVENTIONAL

ENGINE W ITH A CARBURETTOR AND THAT W ITH DIRECT FUEL INJECTION
CONVENTIONAL
ENGINE WITH
FUEL & OIL CARBURETTOR
EXHAUST EXHAUST
UP TO 70%
FUEL CHARGE
INTAKE LOST:
POOR FUEL
ECONOMY &
EXH EMISSIONS
CARBURETTOR

DIRECT INJECTED
STRATIFIED FUEL INJECTION
AFTER PISTON COVERS ENGINE
EXHAUST PORT
EXHAUST EXHAUST
ONLY AIR LOST
INTAKE

AIR &
METERED OIL

Figure 2. Scavenging and Fuel Delivery in Carbureted and Direct-Injected Two-Stroke Cycle Engines
Courtesy of Orbital Engine Corporation
400 Hz, and at twice the firing frequency, 500-800 Hz. Engine Corporation DI. This system is the Orbital
The abrupt opening of the exhaust port creates a sharp Combustion Process (OCP), which utilizes pressurized
pressure wave that is made up of many frequencies. air to atomize the fuel for proper mixing. The OCP
Although this sharp pressure wave is effective for system uses separate air and fuel injectors (see Figure
improving the performance of a tuned exhaust system, 4) to optimize the combustion process. The timing and
its broad spectrum is very unpleasant to the human ear. duration of fuel injection and air injection are controlled
The intake system creates noise at both the frequency independently by an electronic control unit (ECU). The
and twice the firing frequency. Additional noise is ECU also controls ignition timing.
caused by the mechanical noise throughout the engine
as well as by the combustion pressure rises transmitted The ECU is used to calibrate the system for optimal
through the walls (creating diaphragm noise). The use performance under different conditions. At light loads,
of reed valves in the intake system creates a “honking” fuel is injected relatively late to create a stratified charge
noise that is associated with the two-stroke engine. with a richer mixture around the spark plug than in the
Figure 3 summarizes the noise generation from a two- rest of the cylinder. At higher loads, the fuel is injected
stroke engine. earlier to allow greater penetration and a more
homogeneous mixture in the cylinder. The team
TECHNICAL APPROACH - EMISSIONS received support from Orbital Engine Corporation (Perth,
Western Australia) and SynerJect (Newport News, Va.),
DIRECT INJECTION a U.S. joint venture between Orbital and Siemens.

In order to reduce the hydrocarbon emissions produced

by scavenging with a fuel/air mixture, the team sought a
solution that allowed separation of the scavenging and
fuel metering functions. This involves scavenging with
air only, not a fuel/air mixture. The air is used for
scavenging and to provide oxygen for the next
combustion cycle. This requires that fuel be introduced
directly into the cylinder. This is done using a Direct
Injection (DI) system. With this system, the timing of the
fuel injection process can be controlled such that short-
circuit fuel losses are greatly reduced. Figure 2
illustrates the benefits of air scavenging with direct
injection.

After analyzing different DI options, the CSU team

selected a fuel injection system developed by the Orbital Figure 3. Noise from a Two-Stroke Engine
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

FUEL
PASSAGE FUEL
INJECTOR
AIR
PASSAGE
AIR
INJECTOR

Figure 4. Orbital OCP DI System, Showing the Fuel

Injector (top) and Air Injector (bottom) Figure 5. Platinum / Palladium Oxidation Catalyst

ENGINE MODIFICATIONS

CATALYST Heads

Although the use of DI fuel injection was expected to In order to adapt the OCP system to a stock 600ZRT
dramatically reduce CO and THC emissions, oxidation engine, new cylinder heads were required. These heads
catalysts were installed in the exhaust silencer used to were designed and manufactured by the design team.
oxidize the remaining emissions. The catalysts selected Among the challenges the team faced were tradeoffs
(see Figure 5) consist of a stainless steel substrate between cooling capability, combustion chamber
coated with a layer of gamma-alumina with a washcoat geometry, compression ratio, location of the injector and
of platinum and palladium. The catalysts were five- spark plug, and manufacturability. The team chose to
inches in diameter and two-inches thick and placed switch from the individual heads used on the stock
within the exhaust silencer, for aid in silencing. Two ZRT600 to a single piece head. This simplified the
catalysts were placed in series to increase the surface cooling, prevented potential issues with side loading of
area. The cell density of each catalyst is 300 cells per the fuel injectors, and provided the necessary rigidity for
square inch. Carbon monoxide begins to react at the o-rings that seal the injectors. The ZRT600 engine
around 250° Celsius (500° Fahrenheit), whereas, is water-cooled. This is desirable for DI since these
hydrocarbons react more readily at 400° Celsius - 550° engines typically produce higher exhaust temperatures
Celsius (750° Fahrenheit - 1000° Fahrenheit), than carbureted engines.
depending on the fuel composition9. Since the
temperatures of the exhaust are around 650° Celsius The heads were fabricated in two pieces. A lower plate,
(1200° Fahrenheit), the catalyst efficiency was expected which contains the combustion chamber, was machined
to be high. At these temperatures, Miratech Corporation from aluminum plate. This plate contains the mounting
has rated their catalysts to reduce hydrocarbons by 95% surfaces for the fuel injector, spark plug, and combustion
(on methane – oxidation of higher hydrocarbons was pressure transducer. The combustion pressure
expected to be higher) and CO by 99%. transducer was used for development, and would not be
included in a production engine. An upper cap for the
LUBRICATION head was cast of aluminum and functions as the top
cover of the head. Figure 6 shows the shape of the
In a conventional two-stroke engine, the oil is dissolved combustion chamber. Figure 7 shows the inner surface
in the fuel, either by using oil injection or a “premix” of of the lower plate. Figures 8 and 9 show the
fuel and oil. The fuel dilutes the oil, reducing its viscosity manufacturing of the lower plate. Figure 10 shows the
and allowing effective distribution in the crankcase. In steps in casting of the upper head cap. Finally, Figure
the modified engine, oil is injected into the intake air to 11 shows the initial solid model and final head assembly.
provide crankcase lubrication. It is believed that there is
less oil lost to scavenging than would normally be lost
through fuel scavenging, although this has not been
confirmed. One negative feature of eliminating oil
dilution is that the oil is more viscous during cold start,
although the engine fires at a lower speed with direct
injection. Cold start performance is good.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Figure 6. Cross Section of Combustion Chamber

Figure 7. Lower Head Plate

Figure 8. Machining of Lower Head Plate from

Billet Aluminum
Figure 9. CNC Machining of Combustion
Chamber in the Lower Head Plate

Figure 10. Casting of Head Cap Figure 11. Final Head Assembly
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Figure 12. Fuel Injector (left) and

Air Injector (right) Figure 13. Prototype Air Pump

The compression ratio was raised from 6.4:1 to 6.9:1 to The second major component of the head design was
provide a slight benefit in efficiency. Due to the high the head cover. This component had to seal the water
elevation of the competition (7,000 – 9,000 feet), the jacket, incorporate the coolant outlet, and provide a solid
compression ratio could have been increased to as high mounting pad for the fuel rail. This head cover was
as 8.0:1 to maintain the same compression pressure essentially a 3/8” flat plate, with four 1” bosses for
that would be produced at sea level. However, the mounting of the fuel rail. If this part were machined,
compression ratio selected represents a balance over two-thirds of the material would ultimately be
between the variables considered. The combustion removed, which would take a lot of time and material for
chamber must maintain a critical geometry between the two copies of this part. Since this piece was only sealing
spark plug and injector tip. The chamber must also the water pressure, a lower strength alloy could be used.
maintain certain flow characteristics. It was felt that These two considerations made casting a superior
further increase in compression ratio might compromise option to machining for this piece. For casting, a single
the flow characteristics of the cylinder. Time constraints pattern was made out of wood. Figure 10 shows the
did not allow the fabrication and evaluation of different solid model, mold, and finished casting. Approximately 5
head designs and compression ratios, so a relatively hours of finish machining was required after casting.
conservative combustion chamber design and
compression ratio were selected. Figure 11 shows the solid model of the head. The
design work was done using the SolidWorks CAD
The combustion chamber utilizes an offset dome. This program. The final head assembly, including the lower
is visible in Figure 6. The offset dome improves the plate and upper cap with the fuel rail mounted, are also
combustion process in multiple ways. Since the injector shown in Figure 11. It should be emphasized that the
is offset to the intake side, the fuel spray is further from production of the prototype head was much more
the exhaust port, which helps to reduce short-circuiting. involved than it would be on a production engine. In
production, the head would be produced as a single
Head Fabrication casting, with a cost similar to other single-piece heads.

Once the head design was finalized, the next challenge FUEL INJECTION HARDWARE
became how to produce two prototypes in a very short
amount of time. The first part produced was the lower The OCP fuel injection system, as applied on this
head plate. This component required a lightweight project, utilizes an electronic control unit (ECU), fuel
material with high yield strength to support combustion injectors, an engine-driven air pump (for the atomization
pressure. The material selected was 7050 T7451 air), crank encoders, a medium-pressure fuel source, a
aluminum, which is a solution heat-treated alloy with a throttle position sensor and a coolant temperature
yield strength of 68 ksi. Due to complex 3-D and 2-D sensor.
surfaces, CNC machining was used throughout the
manufacturing process (see Figures 8 and 9). Figure 9
shows pictures of the tool path, validation of the tool
path using foam prototyping, and final machining of the
combustion chambers. Two copies of the lower plate
were fabricated. Each required approximately 30 hours
of CNC machining.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Figure 15. Throttle Body and

Throttle Position Sensor

Figure 14. Dual Engine Encoders

Injectors Throttle Body / TPS

Two independent injectors are required for each A throttle body from a port-injected engine was used.
cylinder. The first injector meters the fuel, and the The injector ports were initially plugged, but are now
second atomizes the air-fuel mixture directly into the used to mount quarter-wave tubes to reduce intake
combustion chamber. The individual injectors are noise. A throttle position sensor is mounted to the
shown in Figure 12. The fuel injector is positioned on throttle shaft and is the primary load input used by the
top of the air injector as shown in Figure 4. ECU. At idle, the ECU switches to an idle fuel strategy,
which incorporates closed loop control of idle speed.
Air Pump
Engine Control Unit
Based on other Orbital applications, a 40cc air pump
was selected for this 600cc engine. This pump is shown The engine controller used in this system is a “previous
in Figure 13. Modifications were made to accommodate generation” Orbital ECU. It uses inputs such as throttle
lubrication. The pump was mounted to the engine with a position, speed, and jacket water temperature to
cast aluminum housing manufactured by the team. The determine the timing and duration of fuel injection, timing
pump has adequate flow capacity at 550 KPa, the and duration of air injection, and timing of ignition. The
pressure setpoint of the air rail. ECU uses an Electrically Erasable Programmable Read
Only Memory (EEPROM) to allow real-time calibration
Encoders during engine operation. This would not be the case on
a production unit.
The Orbital direct injection system, as applied in this
project, utilizes two separate crank encoders (see Figure ECU Calibration
14). This 2-rotor system allows the ECU to determine
crank position within 3-4 degrees of rotation after Orbital provided the team with calibration software and
commencement of a pull-start. This is particularly training during the system shakedown and early
important during cold starting. The system also provides calibration stages. Initial calibration was performed on
rapid and accurate position resolution, which is an engine dynamometer at CSU’s Engines and Energy
important during rapid engine transients. Conversion Laboratory. Field calibration was performed
at sites in Colorado, in southern Wyoming, and at the
Fuel Pump & Fuel Rail competition site (the Flagg Ranch, adjacent to the South
Entrance to Yellowstone National Park; near Jackson,
A fuel pump is needed which can produce adequate flow WY).
at a pressure approximately 70 kPa higher than that of
the air rail. Thus, the fuel pump must produce at least Development work included calibration of the following:
620 kPA, the air rail being set at 550 kPA. The start-up parameters, a closed-loop speed control
maximum flow rate of fuel through the fuel injectors is strategy at idle, a stratified charge strategy when running
3.7 g/s at the differential pressure setpoint of 70 kPa. at light loads, and a homogeneous charge strategy when
An aftermarket automotive electric fuel pump was operating at higher speeds and loads. The closed-loop
selected to provide the required fuel pressure and flow. speed control strategy for idle has been discussed.
The fuel rail is equipped with two regulators, one for air Combustion at idle is highly stratified to enhance stability
pressure and one for differential fuel. Return flow from and minimize short-circuiting of fuel. The stratified
the fuel rail is returned to the fuel tank. Excess air from charge strategy at light load operation utilizes relatively
the air rail is exhausted into the airbox. The fuel rail was late injection timing to keep the fuel cloud close to the
manufactured by SynerJect to our specifications. The spark plug. This produces a combustible mixture near
fuel rail assembly is shown mounted on the engine head the spark plug, while maintaining a much leaner mixture
in Figure 11. throughout the rest of the cylinder. At higher loads and
speeds, a homogeneous air/fuel mixture is desirable to
attain maximum power; this is achieved by using more
advanced injection timing to enhance fuel atomization.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Figure 17. Cast Aluminum Housing for Mounting

of Magnetic Pickup and Air Pump

Figure 16. Representative Traces of Combustion

Pressure During Calibration

During calibration, combustion analysis, subjective THE NOISE SOLUTION

ridability testing, and emissions measurements were
used to guide the calibration process. Combustion The CSU team has adopted a comprehensive approach
pressure transducers were mounted in each cylinder to to noise reduction.
monitor cyclic combustion variations. Representative
combustion pressure traces are shown in Figure 16. Exhaust Silencer
Ridability and throttle response were assessed
subjectively during field testing. Calibration values were The major focus was to reduce the noise at the
adjusted to give crisp throttle response. Initially, it was fundamental firing frequency with the use of an exhaust
anticipated that additional calibration would be required silencer. As the first step, the triple-pipe exhaust on the
after emissions testing. However, after calibrating for ZRT600 was removed and replaced with a modified
combustion stability and ridability, the emissions of CO single-pipe exhaust and silencer from a 3-cylinder
and THC were reduced by over 99%, so no further Polaris engine. The catalysts are mounted in the
calibration was required. silencer, so a single-pipe system was required. The
silencer was cut open, and a flanged extension was
Magnetic Pick-up Bracket added. The assembled catalyst/silencer system is
shown in Figure 18. The exhaust was wrapped with
In order to mount the crank encoders, the air pump drive insulation and a heat shield was fitted. Although the
pulley, and the pull start pulley, modifications to the insulation was added to control temperature, it also
flywheel side of the crank were made. An extension reduces noise radiated from the exhaust system. After
shaft was added to the crank to allow mounting of the initial testing, noise at the firing frequency (≈ 400 hz) is
components mentioned. still excessive so a set of quarter-wave silencers and a
perforated exhaust extension have been added.
The use of a crank extension required the fabrication of
a housing to mount the encoder pickups, the air pump, Intake Silencer
and the pull-start housing. A rigid bracket (shown in
Figure 17) was manufactured by sand casting an Noise from the intake system has also been reduced by
aluminum/magnesium alloy. The magnesium provides the use of an intake silencer, which is a combination
good vibration damping to help prevent propagating baffle/absorption silencer. Intake air is drawn in through
frequencies while the aluminum provides a strong and air filters and enters the intake box through three pipes
lightweight structure. This bracket was required that are each capped at one end and perforated around
because this is a retrofit application. On a purpose-built the outer edges. This allows noise to diffuse into the air
engine all of the mounting needs could be met through box. The box is lined with a barrier/absorber foam
minor modification of the original casting. composite to block air-born noise and absorb noise due
to vibrations. Air flowing through the air box is diverted
and reflected by baffles lined with the composite so that
air has no straight path to escape and is more likely to
be attenuated by the foam. After initial testing, noise at
twice the firing frequency (≈ 800 hz) was still excessive,
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Acoustic Foam Sound Transmission Loss as

a Function of Frequency

70

Sound Transmission Loss

60

50

40

(dB)
30

20

10

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500

Frequency (Hz)

Figure 18. Exhaust System, Including Tuned Pipe

(right) and Silencer (left) which Houses the Figure 19. Absorption Characteristics of
Oxidation Catalysts Acoustic Foam Used on Hood

so a quarter-wave resonator was mounted in each Laboratory (EECL) at CSU. The lab has given the CSC
throttle body to reduce this frequency component. team its support and the use of such equipment for the
2002 project. The Rosemount five-gas rack measures
Structural Damping dry emissions of “criteria pollutants” such as
hydrocarbons, CO, and NOx, and also measures CO2,
Structural damping was added to the engine cowling to and O2. The FTIR uses spectral analysis of infrared
attenuate mechanical noise radiated underhood. The absorption of different gaseous compounds, and was
foam chosen was Tufcote by EAR Specialty primarily used to look for the presence of non-regulated
Composites. This foam has aluminized polyester facing “hazardous air pollutants” (HAPs), such as formaldehyde
to resist moisture, including grease, oil, and engine and acetaldehyde. The portable analyzer was used to
fluids, as well as deflect heat. Factory specifications allow emissions measurement during field testing. It is
from EAR Specialty Composites of sound transmission an infrared emissions analyzer with a similar detection
loss as a function of frequency of this acoustical foam mechanism and similar accuracy to the remote sensing
are shown in Figure 19. measurement equipment to be used in the competition.
It was used to simulate the competition measurements.
A summary schematic of the overall noise strategy for
the CSU design is shown in Figure 20. ENGINE TEST STAND

TESTING The team adapted a SuperFlow SF-1 water brake

dynamometer to mount the ZR600 engine. The SF-1
EMISSIONS MEASUREMENT was used for high-load calibration, but could not be
unloaded sufficiently to allow it to be used to simulate
The emissions data provided in this report were attained the Mode 3 (75% speed, 33% load) and Mode 4 (65%
using a Rosemount NGA-2000 five-gas emissions speed, 19% load) low-load points required to perform
analyzer, a Nicolet “Magna” Fourier Transform Infrared the 5-mode snowmobile test cycle.10 For the 5-mode
Spectrometer (FTIR), and a Vetronix PXA-1100 portable testing, a smaller Land & Sea “Dynomite” water brake
emissions analyzer. This emissions equipment is dynamometer was used. Unfortunately, the Land & Sea
housed in the Engines and Energy Conversion dynamometer did not have a feedback load control
system, and manual attempts at dynamometer load
control were too unstable to allow final FTIR emissions
testing at the full-load “mode 1” point.

SOUND TESTING

A Quest 2400 sound meter was used for sound testing.

The analog output from the Quest 2400 was then input
into a Panasonic CF-71 laptop, which recorded the
sound as a .WAV file. The .WAV files were then post
processed with various sound analysis programs to
assess the frequency content of the sound.

Figure 20. Summary of Noise Solution

 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

RESULTS

EMISSIONS UNDER COMPETITION CONDITIONS Total Hydrocarbon (THC)

Emissions were tested under simulated competition Unmodified THC 19,343 ppm (as hexane)
conditions during a 4% uphill run at 20 mph. Emissions emissions
were measured with the PXA-1100 portable gas THC emissions using DI 2,050 ppm (as hexane)
analyzer in order to allow comparison with results from 89.4% reduction of
the CSC 2000 competition, and to predict results in the unmodified THC emissions
CSC 2002 competition. The data was analyzed using THC emissions using DI & 64 ppm (as hexane)
the FEAT equations provide by Dr. Gary Bishop11 to catalyst 96.9% reduction of DI
account for dilution effects. The results summarized in emissions
Table 1 show that DI alone reduced the THC emissions
99.7% overall THC reduction
by approximately 90% and CO emissions by 70%.
When the oxidation catalyst was added, the emissions of Carbon Monoxide (CO)
both compounds were reduced by over 99%. This same Unmodified CO emissions 108,150 ppm
information is shown in a graphical form in Figures 21
and 22. Particulate emissions were not measured, but CO emissions using DI 32,973
the addition of the catalyst appears to completely 69.5% reduction of
unmodified CO emissions
eliminate visible smoke once the catalyst begins to warm
up. CO emissions using DI & 65 ppm
catalyst 99.8% reduction of DI
emissions
Uncontrolled HC Emission of 19,343 ppm 99.9% overall CO reduction
Table 1. Summarized Results of Emissions Testing
17,293 ppm removed with DI Note: The results above were obtained under competition conditions
traveling uphill on a 4% grade at 20 mph. Emissions were measured
using a Vetronix PXA-1100 portable emissions analyzer. The results
1986 ppm were “post-processed” using the University of Denver FEAT CO2 ratio
removed with technique to eliminate the effects of dilution. This will allow the results
to be compared with the results obtained using the remote sensing
catalyst
technique, which will be used during the competition.

EMISSIONS UNDER 5-MODE TESTING

Remaining HC The 5-mode emissions test allows the evaluation of a

emissions of 64 ppm snowmobile’s performance across a much wider
operating range than will be tested in the CSC 2002
competition. The CSU snowmobile was evaluated under
Figure 21. Reduction of Hydrocarbon Emissions the same speed and load conditions that would be used
in the 5-mode test. The results are summarized in Table
Uncontrolled CO Level of 108,150 ppm A.1, at the end of the report. The results show that THC
and CO levels are higher than those measured with the
portable infrared analyzer. This was expected since the
infrared analyzer cannot measure the aromatic
75177 ppm 32,908 ppm compounds in gasoline, and has lower sensitivity to the
removed with DI removed with ethanol in the fuel. The data also shows that CO and
catalyst HC levels rise slightly at higher load conditions (Mode
2). NOx emissions rise with increasing load.

FUEL ECONOMY

65 ppm remaining By using the direct injection system, fuel efficiency is

greatly increased. Fuel is no longer short-circuited
Figure 22. Reduction of Carbon Monoxide through the exhaust port without being burned. Fuel
Emissions efficiency is increased from 12-15 miles per gallon to 18-
22 miles to the gallon. These values are based on 35
mph operation. Fuel consumption for the DI engine was
calculated from the fuel injector duty cycle and the
injector flow characteristics.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Figure 23. Noise Spectrum, Indicating the Need

to Further Reduce Noise at Firing Frequency
(400 hz) and 2X Firing Frequency (800 hz)
Figure 24. CSU Snowmobile “Grabbing Air”
During Performance Testing of the Direct
Injected Engine in mid-February
NOISE

Noise Equipment increased. This is partly attributable to the DI system,

and partly to the favorable mid-range operation of the
Noise measurements were made with a Quest Model single-pipe exhaust. The modified sled is able to run the
2400 sound level meter using the SAE J19212 acceleration event in 7.8 seconds, which is comparable
snowmobile noise test procedure. The snowmobile had to the 7.6 seconds time for the carbureted engine with
a base noise measurement of 83dBA. With the initial the triple-pipe exhaust. The precise control and the
modifications (intake silencer, exhaust silencer, minimal transportation time of the fuel to the combustion
acoustical foam in the hood), the noise at the halfway chamber on the DI system provide crisp throttle
point of a 500 foot full-acceleration run (the competition response.
event where noise is evaluated) was 75dBa. The noise
spectrum in Figure 23 shows that noise peaks are HANDLING
occurring at about 400 Hz and 800 Hz. 400 Hz is the
firing frequency of the engine when the engine is The Orbital DI system allows the engine to retain power
operating at its normal “high-load” speed of 8000 rpm. comparable to the stock sled. This provides the
The remaining 400 Hz noise is believed to be primarily experienced, aggressive rider with enough power to “find
due to unattenuated exhaust noise. The 800 Hz noise is the limitations” of the stock suspension. In order to
believed to be due to intake noise and to the first improve the handling of the sled in aggressive jumps,
harmonic of the firing frequency. the rear suspension was replaced with a long-travel A.
D. Boiven “Expert” suspension. To improve stability, the
Additional noise mitigation features have been added. front A-arms were replaced with longer A-arms to allow
Quarter-wave reaction tubes tuned to 400 Hz and 800 increased front suspension travel. Figure 24 shows the
Hz have been added to the exhaust. Quarter wave direct-injected snowmobile “in action” during calibration
tubes tuned to twice the firing frequency have been fitted in February 2002. This picture suggests the excellent
to the intake throttle-bodies. A more dissipative tailpipe power / weight ratio of the snowmobile and its
is being added to produce more rapid noise dispersion. exceptional handling characteristics.
Finally, hood sealing has been significantly improved.
ERGONOMIC CONSIDERATIONS
PERFORMANCE
Improved rider comfort was considered in the CSU
As discussed earlier in this report, the power-to-weight design. Changes to the CSU sled include a seat
ratio of a two-stroke engine is superior to that of a four redesign, a handlebar riser, and a higher lifting bar.
stroke. The stock ZRT600 engine produces These modifications were based on the results of an
approximately 75 hp on an engine dynamometer at CSU ergonomic study conducted by an earlier student design
(note: The CSU Engines Laboratory is at an altitude of team. The seat was raised to allow a more comfortable
5,000 feet, so this is significantly lower than the engine’s operation when the driver is seated. The handlebars
sea level rating). The engine with direct injection were raised to improve the shoulder joint angles while
produces 70-75 hp, so the power is comparable. Some riding. This was accomplished with the addition of a new
upper end power was lost due to the use of single tuned 2-inch riser block (6061-T6 solid block with four 4-inch
pipe, although the DI system appears to “gain back” bolts) connecting the steer tube and bar mounts, and
some of the tuned-pipe loss. Torque in the mid-range is new 4-inch rise chromoly bars. The combination of the
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

two results in significantly improved perfect shoulder using the same technique. The TICA score for the stock
joint and elbow for most riders. The rear lift bar of a snowmobile was $505, producing a differential price of
snowmobile is used extensively in deep snow. The $497. The cost differences are summarized in Table 2.
stock height of the bumper/lift bar, which is
approximately mid-shin, results in a “hunched over” The cost analysis provided is based on the cost to
posture during lifting. The lift bar uses welded 6061-T6 produce a production run of 5,000 snowmobiles with the
aluminum footplates and tubular 6061-T6 aluminum design features of the CSU snowmobile. Many of the
bolted to the tunnel at an angle that allows for an more complex features to prototype (i.e. heads) could be
optimum lifting height. produced by “normal” production techniques at little
additional cost. Two of the purchased items used on the
DURABILITY AND MAINTENANCE CSU snowmobile, the catalyst and the air pump, cost
significantly more in prototype quantities than in modest
The CSU design which utilizes an Arctic Cat engine and production quantities. In order to provide a realistic cost
the Orbital OCP fuel system is a durable and robust analysis for a production snowmobile, the team has
design. The current engine has logged over 50 hours secured quotes for comparable components to provide a
with only minor problems – none related to the new cost analysis that demonstrates the cost benefits of
heads or fuel injection system. There has been no higher volume manufacturing.
abnormal wear on the heads and cylinders. Inspection
of the heads shows they are holding up well. The
durability of the Orbital OCP system has been
demonstrated by small sample tests13 and by the
System Cost
successful conclusion of a large fleet trial. In this trial,
Differential
100 Ford Festiva vehicles were equipped with two-
stroke Orbital engines. These vehicles then
accumulated over five million kilometers of operation Fuel System: Costs include: $154
with excellent reliability and durability.14 The Orbital injectors, fuel rail, fuel pump, air
OCP system is now in use in two-stroke engines sold by compressor. Credits include
Mercury Marine (the “Optimax” engine), Bombardier- carburetors and low-pressure
Rotax, Tohatsu, Aprilia, and Piaggio. Other two-stroke fuel pump.
and direct-injection four-stroke applications are in
various stages of commercialization. Exhaust System: Costs include $181
catalyst and muffler. Assumes
If this system were to be pursued for production, the that an expansion chamber
lubrication system should receive attention. The lower- would be used for either engine.
end was designed to be lubricated by a gas / oil mixture.
Noise Reduction: Costs include $29
The current system is different in that the fuel is silencing foam and quarter-wave
removed from the mixture (as described by the resonators.
Lubrication section of this report). For production, a
lubrication system similar to that used in the fleet trial Electronics: Includes the Electronic $133
engines or Optimax engines should be considered. This Control Unit (ECU), and
system incorporates an electronic lubing system that additional sensors. Credits
eliminates the dependency of air to transport the oil to include the stock ignition module
the various areas within the crankcase areas, i.e. the and crank sensor.
main bearings, connecting rod bearings, wrist pins, and
cylinder walls. The cost increment in production would Total Additional Cost $497
be negligible. A snowmobile must be able to start
reliably in very cold weather. This requires an oil with Table 2: Cost Differential of the CSU Direct Injected
low cold-weather viscosity but excellent hot operation Snowmobile
protection. The current oil (a Yamaha low-ash
SAFETY
watercraft oil, chosen to prevent catalyst fouling) is
almost certainly non-optimal. In spite of the comments
on the lubrication system, the current lubrication system No modifications were made which would degrade the
appears to be functioning very well. structural integrity of the engine or chassis, so no impact
on safety is expected. Again, extensive field testing and
commercial use of the Orbital OCP system has
COST OF IMPLEMENTATION
demonstrated a strong safety record.
In assessing the cost, the team used the standardized
CSC Technical Implementation Cost Assessment
(TICA). The modified snowmobile totaled $1002 using
the TICA method. The stock snowmobile was evaluated
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

COMPETITION RESULTS Category Results

The Colorado State University team has Emissions 99.7% reduction in hydrocarbon
demonstrated the effectiveness of direct injection emissions
technology. The team won the Lowest Emissions
Award, showing decisively that a direct-injected two- 99.9% reduction in carbon
stroke with a catalytic converter can meet or beat the monoxide emissions
emissions of a sophisticated four-stroke engine. There Noise Reduced from 83 dBA to 75 dBA
were only two two-stroke engines that even passed the
emissions requirements. The results from the Fuel economy Increased by 35%, from 13 mpg to
competition mode-2 testing (uphill at a steady 20 MPH) 20 mpg
was comparable to the results mentioned previously in
this report. As time constraints did not allow for Fuel savings of $40 for each 1,000
thorough transient calibration, the emissions taken miles of operation
during acceleration in mode-1 testing of the competition
were not optimal. The averaged emissions reductions Performance Acceleration: equal to stock, ≈7.8
over the two-mode testing were 99.4% for carbon seconds in 500 foot acceleration
monoxide emissions and 88.6% for hydrocarbon Mid-range torque: increased over
emissions. stock
A few mishaps during the competition cost CSU Throttle response: improved over
significant points. During the endurance event, a fitting stock
on the air pump worked its way loose. Without adequate
air pressure, atomization of the fuel was degraded. This Handling Improved longer travel suspension
resulted in longer combustion duration and significantly
Ergonomics • Improved seat height
hotter exhaust temperatures. The heat was so hot that
the belly pan caught on fire, burning a hole in the side of • Improved handlebar height
the sled. This forced the CSU team out of the
endurance/fuel economy event. Estimations at the time • Improved lift bar height
of the breakdown indicated aforementioned fuel
economy levels. This breakdown also affected the noise Cost $447 increase in manufacturing
of the snowmobile. Pre-competition testing showed that cost
the snowmobile was within the range allowed for noise Table 3. Summary of Results
emissions. The repair to the belly pan was not as tight
as the original unit, allowing louder under-hood noise.
This was one of the factors impacting the noise
savings in overall operating cost. In fact, the DI
performance. In the design of the quarter-wave
snowmobile would save $40 in fuel cost for every 1,000
attenuators, the team decided on their lengths based on
miles of operation, based on $1.50 gasoline. The
warm engine conditions. Since the engines were shut
system will pay for its manufacturing cost differential
down until right before run, the tubes were never up to
through fuel savings in less than 10,000 miles of
their design temperature. These two factors caused the
operation.
noise levels to be 75dBA, not enough to pass the
rigorous requirements.
The CSU snowmobile has significantly reduced
snowmobile noise, to 75 dBA. The rider as well as those
The CSU team won all of the design events, including
in the vicinity of the snowmobile will appreciate the
oral design presentation, written design report, and
benefits of a cleaner, quieter snowmobile. The rider
design display. However, since the snowmobile did not
comfort has been enhanced through improved
meet the sound requirements, the team did not receive
ergonomics.
the Best Design Award.
The CSU entry is exciting to ride while being economical
CONCLUSION and quiet. Furthermore, the emissions reduction is
remarkable. The use of a direct-injected engine and
The CSU team was guided by the philosophy that oxidation catalyst have reduced emissions of carbon
environmentally friendly snowmobiles will not be monoxide and hydrocarbons by over 99%, and visible
adopted voluntarily unless they improve performance smoke is eliminated.
and reduce the cost of operation. The direct injected
two-stroke snowmobile engine developed by CSU team The results of the development effort are summarized in
does both. The CSU snowmobile maintains equivalent Table 3.
peak power while increasing mid-range torque and
improving throttle response. The direct-injected engine
reduces fuel consumption by 35%, producing significant
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

4
In conclusion, the CSU two-stroke cycle snowmobile Control of Snowmobile Noise. Volume 1 – Technology
utilizing the Orbital OCP direct injection system and an and Cost Information. EPA Document 550/9-74-003-A.
oxidation catalyst has shown the ability to dramatically 1974.
5
reduce emissions and noise while improving SAE Clean Snowmobile Challenge 2002 rules,
performance and reducing operating cost. available online:
http://www.engr.colostate.edu/eecl/Rules.pdf
6
ACKNOWLEDGMENTS Alan T Kirkpatrick , C R Ferguson. Internal Combustion
Engines - Applied Thermosciences, Second Edition.
The CSU team gratefully acknowledges the assistance John Wiley & Sons. 2001.
7
of Orbital Engine Corporation and SynerJect. Special A. McDowell, et al. “Catalyst Deactivation on a Two-
thanks to Stephen Malss, Mark Norman, Gavin Dickson, Stroke Engine”. SAE paper #982015.
8
Paul Gulvin, Fred Morin, and Harlow Rutledge. We Gordon P Blair. Design and Simulation of Two-Stroke
appreciate the support of our major team sponsors: Bob Engines. SAE International. 1996.
9
and Bonnie Walker, Enginuity Inc., Colorado J B Heywood. Internal Combustion Engine
Snowmobile Association, and Vail Associates. Fundamentals. McGraw-Hill. 1988. p. 652.
10
C W Wright, J J White. “Development and Validation
REFERENCES of a Snowmobile Engine Emission Test Procedure.”
SAE #982017. Later modified in 2001-01-1832/4235.
11
Gary Bishop. “FEAT Equations for CO, HC, and NO.”
1
S W Coates, G G Lassanske, “Measurement and Private communication. February 2002.
12
Analysis of Gaseous Exhaust Emissions from SAE standard J192. “Exterior Sound Level for
Recreational and Small Commercial Marine Craft.” SAE Snowmobiles.” March 1985.
13
#901597. K Eisenhauer. “Durability Development of an
2
Gordon P Blair. Design and Simulation of Two-Stroke Automotive Two-Stroke Engine.” SAE paper #956006.
Engines. SAE International. 1996. 1995.
3 1414
J B Heywood, Eran Sher. The Two-Stroke Cycle David Shawcross, C Pumphrey, D Arnall. “A Five-
Engine: Its Development, Operation, and Design. SAE / Million Kilometre, 100-Vehicle Fleet Trial of an Air-Assist
Taylor & Francis, 1999. Direct Fuel Injected, Automotive 2-Stroke Engine.” SAE
paper #2000-01-0898.
 Downloaded from SAE International by University of British Columbia, Thursday, August 02, 2018

Emissions Emissions
Mode Speed (RPM) Load (ft*lb)
(ppm) (FEAT) (ppm)
Start up
O2 = 17.1%
FEAT → O2 = 0%
CO2 = 3.08%
Idle, no load CO2 = 13.73%
HC = 7,206
HC = 32,116
1,300 0 CO = 1,124
Measured at CO = 5,009
NOx = 92.7
startup, before NOx = 413
CH2O = 96
catalyst
CH3CHO = 145
“lightoff”
5 O2 = 15.04%
CO2 = 4.42% FEAT → O2 = 0%
Idle, no load HC = non-detect CO2 = 15.05%
1,300 0 CO = 27.7 HC = non-detect
Measured after NOx = 168.81 CO = 94
catalyst CH2O = too low NOx = 574
“lightoff” CH3CHO = too low
O2 = 7.03%
CO2 = 9.76% FEAT → O2 = 0%
4
HC = 94.69 CO2 = 15.04%
5,350 7.7 CO = 50.8 HC = 145
65% speed
NOx = 98.9 CO = 78
19% torque
CH2O = non-detect NOx = 152
CH3CHO = non-detect
O2 = 8.69%
CO2 = 8.51% FEAT → O2 = 0%
3
HC = 80.5 CO2 = 15.05%
6,000 12.3 CO = 29.7 HC = 142
75% speed
NOx = 306 CO = 53
33% torque
CH2O = non-detect NOx = 541
CH3CHO = non-detect
O2 = 9.00%
CO2 = 8.17% FEAT → O2 = 0%
2
HC = 195 CO2 = 15.03%
6,800 19.5 CO = 68.2 HC = 359
85% speed
NOx = 987 CO = 125
51% torque
CH2O = non-detect NOx = 1,816
CH3CHO = non-detect
1
Dynamometer instability prevented control
8,000 38.2
100% speed of load, emissions not measured.
100% torque
Notes:
• Emissions reported in the far right column were post-processed using the FEAT equations
used by the University of Denver’s “CO2 Ratio” technique. This technique is used to factor
out the effects of dilution, and therefore assumes that the O2 level is 0%.
• HC levels at point 5 (idle, with catalyst “lit”) were below the range of accuracy of the FTIR
• Formaldehyde and acetaldehyde emissions were below the detection limits except in the
startup case without catalyst lightoff.
Table A.1 – Emissions from 5-Mode Snowmobile Cycle