COMPUTER SIMULATIONS OF A HYDROGEN FUELED

INTERNAL COMBUSTION ENGINE

by

JAAKKO JALMARI HALMARI, B.S.


A THESIS

IN

MECHANICAL ENGINEERING

Submitted to the Graduate Faculty
of Texas Tech University in
Partial Fulfillment of
the Requirements for
the Degree of

MASTER OF SCIENCE

IN

MECHANICAL ENGINEERING

Approved


Timothy Maxwell
Chairperson of the Committee

Atila Ertas

Michael Parten

Accepted


John Borrelli
Dean of the Graduate School

May, 2005
pdfMachine by Broadgun Software - a great PDF writer! - a great PDF creator! - http://www.pdfmachine.com http://www.broadgun.com
ii
ACKNOWLEDGEMENTS

I would like to thank Dr. Maxwell and Dr. Parten for offering me the opportunity
to participate with the FutureTruck competition as well as other projects going on at the
Advanced Vehicle Engineering Laboratory at Reese Center. I would also like to
thank Dr. Ertas for serving in my thesis committee. The support from my family is
greatly appreciated. Pitkään on väännetty ja käännetty mutta tulihan siitä lopulta jotain.


iii
TABLE OF CONTENTS

ABSTRACT .................................................................................................................v
LIST OF TABLES .........................................................................................................vi
LIST OF FIGURES....................................................................................................... vii
CHAPTER
I. INTRODUCTION........................................................................................................1
Production, Distribution and Storage of Hydrogen ...............................................1
FutureTruck.........................................................................................................3
Objective .............................................................................................................3
Gasoline baseline .................................................................................................4
Engine basics .......................................................................................................4
Controlling Fuel Flow and Spark Timing .............................................................6
Speed density.......................................................................................................6
Mass air flow.......................................................................................................7
Other sensors .......................................................................................................8
Actuators ...........................................................................................................10
Hydrogen properties...........................................................................................11
Differences Converting an Engine from Gasoline to Hydrogen ..........................13
Running engine on H2 .......................................................................................17
Hybrid system basics .........................................................................................18
Types of Hybrids ...............................................................................................19
Series hybrid-electric systems ............................................................................19
Parallel hybrid-electric systems..........................................................................20
Pre-transmission parallel ....................................................................................20
Post-transmission parallel...................................................................................21
Road parallel......................................................................................................21
II. ENGINE SIMULATIONS........................................................................................22
Ricardo WAVE background...............................................................................23
iv
Results given......................................................................................................23
Description of how WAVE calculates air flow...................................................23
How models are built .........................................................................................24
Discretization lengths.........................................................................................26
Throttle ..............................................................................................................27
Simulating the intercooler and the catalytic converter.........................................28
Simulating cylinder heads ..................................................................................28
Simulating the camshaft .....................................................................................29
Supercharger modeling ......................................................................................30
Combustion model .............................................................................................32
Calculating the burn duration .............................................................................33
Emission............................................................................................................35
III. SIMULATION RESULTS.......................................................................................37
Spark Timing .....................................................................................................37
Volumetric Efficiency........................................................................................42
The Effect of Volumetric Efficiency on Torque..................................................42
The Effects of Fuel Type, Equivalence Ratio and Supercharging
on Volumetric Efficiency...................................................................................43
Compression Ratio.............................................................................................46
Camshaft Timing ...............................................................................................50
Throttling a Hydrogen Engine............................................................................58
Emissions...........................................................................................................61
Pressure Measurement in the Intake and Exhaust Systems..................................64
Conclusion and Future Research ........................................................................66
BIBLIOGRAPHY.........................................................................................................68


v
ABSTRACT
Ricardo WAVE was used to simulate the hydrogen internal combustion engine
used in Texas Tech University’s FutureTruck Ford Explorer. Initially, a naturally
aspirated gasoline engine was simulated, followed by the supercharged hydrogen engine.
The objective of these simulations was to maximize power of the hydrogen engine, while
minimizing the emissions and fuel consumption. Among the variables which were
changed, were the equivalence ratio, compression ratio, throttle opening, camshaft
timing, and exhaust size. The simulation results studied included the volumetric
efficiency, fuel consumption, as well as NO emissions. Several results were compared to
the gasoline baseline model.

vi
LIST OF TABLES

1.1 Physical properties of hydrogen and gasoline.........................................................13
2.1 Cylinder Flow Data from Two Sources ..................................................................29
2.2 Partial Summary of Camshaft Specifications..........................................................30



vii
LIST OF FIGURES

1.1 Four-Stroke Operating Cycle ...................................................................................5
1.2 Brake thermal efficiency versus equivalence as the compression ratio is varied......16
2.1 Ricardo WAVE model of the supercharged hydrogen engine. ................................25
2.2 A visual representation of a junction ......................................................................26
2.3 View of the input panel for each duct through which air travels .............................27
2.4 Supercharger Efficiency Curves as Given by Rotrex ..............................................31
2.5 Compressor map as entered into WAVE ................................................................31
2.6 Traction gears in the supercharger..........................................................................36
2.7 Oil being used to transmit power in the supercharger .............................................36
3.1 Brake torque v. spark timing for supercharged hydrogen engine ............................37
3.2 Cylinder pressures v. crank angle for different ignition timing ...............................38
3.3 In-cylinder temperatures for three different spark advances....................................39
3.4 Close up of in-cylinder temperatures for different spark advances..........................40
3.5 NO
x
emission levels in the cylinder for different spark timing................................41
3.6 Torque and volumetric efficiency versus engine speed...........................................43
3.7 Volumetric efficiencies versus engine speed for various engine simulations……..…44
3.8 An expanded view of the volumetric efficiencies for the naturally aspirated
gasoline model run rich and stoichiometric…………………………………………46
3.9 Brake thermal efficiency v. equivalence ratio for different compression
ratios for the naturally aspirated hydrogen model………………………………..…47
3.10 Brake thermal efficiency v. compression ratio for different equivalence ratios
for the naturally aspirated hydrogen model……………………………………….. .48
3.11 Brake thermal efficiency v. equivalence ratio for different compression
ratios for the supercharged hydrogen model…………………………………….….49
3.12 Brake thermal efficiency v. compression ratio for different equivalence ratios
for the supercharged hydrogen model……………………………………. …….….50
3.13 Volumetric efficiency v. engine speed as the camshaft timing is changed...............51
3.14 The change in volumetric efficiency.......................................................................52
3.15 NO emissions as the camshaft timing is changed....................................................53
viii
3.16 NO emissions as the camshaft timing are changed for part throttle conditions. .......53
3.17 Changes in NO emissions as the camshaft timing is changed at WOT. ...................54
3.18 Changes in the NO emissions as camshaft timing is changed at part throttle...........54
3.19 BSFC as the camshaft timing is changed at part throttle .........................................55
3.20 Change in the BSFC as the camshaft timing is changed at part throttle...................56
3.21 BSFC as camshaft timing is changed at WOT........................................................57
3.22 Percent change in the BSFC as the camshaft timing is changed at WOT ................57
3.23 Brake power as the throttle opening and equivalence ratio are changed for the
supercharged hydrogen engine……………………………………………….……..58
3.24 Brake specific fuel consumption as the throttle opening and equivalence ratio
are changed for the supercharged hydrogen engine……………………...…………60
3.25 Brake specific fuel consumption as the throttle opening and equivalence ratio
are changed for the supercharged hydrogen engine with throttle mounted
before the supercharger……………………………………………………………..61
3.26 Tailpipe NO emissions as the the throttle opening and equivalence ratio are
changed for the supercharged hydrogen engine…………………………………….62
3.27 NO
x
emissions v. equivalence ratio for Ford test engines running at various
engine speeds, loads, and compression ratios……………………………………….63
3.28 NOx emissions v. equivalence ratio for the supercharged hydrogen model
running at 3000 rpm and WOT……………………………………………………...63
3.29 Mean pressures in the intake and exhaust for naturally aspirated and hydrogen
engine models……………………………………………………………….……....64
3.30 Mass flow rates through supercharged hydrogen engine and the gasoline model…..65

1
CHAPTER I
INTRODUCTION

As the population of the world grows, more people are using vehicles for both
work as well as pleasure. Each one of these vehicles produces harmful emissions and
uses nonrenewable resources. The carmakers in the developed countries have been
forced to reduce the emissions and fuel consumption each year. Fuel cells are viewed as
a potential long-term solution to the emission problem, but they are still in the
developmental stages. Promising technologies for the shorter term are hybrid electric
vehicles and alternative fuels, especially hydrogen.
Hybrid electric vehicles can substantially lower fuel consumption and emissions
when compared to conventional vehicles, no matter what fuel they use. They allow the
engine to be sized smaller, and make up the lost power with an electric motor.
Alternative fuels such as methanol, ethanol, biodiesel, propane, natural gas, and
hydrogen can lower the emissions of vehicles when compared to gasoline engines.
Hydrogen, however, is the only one that can potentially produce no hydrocarbon, carbon
monoxide, or carbon dioxide emissions (Cosworth). All other fuels contain
hydrocarbons, and inherently will produce carbon dioxide. Before the completely
emission free fuel cell vehicles are perfected, hydrogen engines could be used in vehicles
as a stepping stone to the hydrogen economy. The hydrogen engine vehicles could be
used for building the hydrogen infrastructure and solving the on-board storage problems.

Production, Distribution and Storage of Hydrogen
Although hydrogen is the most plentiful gas in the universe and the simplest
element, it does not occur naturally. Thus all hydrogen needs to be separated from some
other compound, most common ones being hydrocarbons and water. Reforming
hydrocarbons is the most economical and common way of producing hydrogen currently.
This however produces CO
2
as a by-product (Karner). Electrolysis of water is a more
2
environmentally friendly approach, but requires large amounts of electricity for the
process to occur, increasing costs. Once the hydrogen is produced, more energy is
needed in order to compress it to very high pressure for vapor storage in vehicles. Liquid
hydrogen tanks have been made and tested, but it also takes considerable amounts of
energy to cool the hydrogen.
Currently, on-board storage of hydrogen in vehicles is done mostly with high-
pressure tanks. The maximum pressure used in the hydrogen tanks has gone up in the
past few years from 3600 psig to 10,000 psig available currently. Most of the refueling
facilities in use can compress hydrogen to 5000 psig. This is the maximum pressure of
the Quantum Technologies tanks used in Texas Tech University’s FutureTruck Explorer.
Even though the hydrogen is compressed to such high pressure, the amount of
hydrogen stored in the tanks is relatively little when compared to liquid fuels. This
means that the tanks have to have a large volume in order to have enough fuel stored on-
board to have a reasonable range for the vehicle. This is still a problem with most
vehicles using vapor hydrogen storage, as much of the trunk space is taken up by the fuel
tanks. Buses have more room, especially on the roof, and do not suffer as much from this
problem.
BMW has researched the use of liquid hydrogen tanks. These tanks store the
hydrogen at a very cold temperature. With the use of cryogenic tanks, the long term
storage can be a problem, as the tanks are not perfect insulators. Some of the liquid
hydrogen would always be evaporating, requiring venting to the atmosphere. In fact,
storing a vehicle with a low fuel level for a week would most likely result in an empty
tank. This is not a big problem for vehicles which get driven daily, but occasionally
driven vehicles would need to be filled every time they are driven. There also is a
potential safety hazard that comes from the vented hydrogen collecting up near the roof
of a garage. Special vents would need to be installed in garage areas in order to minimize
the potential fire hazard from this.
Metal hydride storage might fulfill the shortcomings of the other methods of
hydrogen storage. These store the hydrogen at ambient temperature, at a low pressure
3
(Sapru). The metal hydride can absorb much more hydrogen than the high pressure
storage systems for a given volume. These are a promising way to store hydrogen safely
at a low pressure. Cost is an issue currently preventing the use of these metal hydrides at
a large scale.

FutureTruck
The U.S. Department of Energy, Ford Motor Company and Argonne National Lab
sponsored the Future Truck competition from 2002 through 2004. The goal of the
competition is to convert a stock 2002 Ford Explorer to a hybrid electric vehicle that uses
alternative fuels, has better fuel economy and lower emissions while maintaining the
original performance. The vehicles must be able to operate just like a stock vehicle; plus
be able to go off-road, pull a trailer, and merge into traffic, while lowering the emissions
and improving the fuel economy by 25%.
Fifteen universities have participated in the Future Truck competition. The
approach chosen by Texas Tech University was to use a hydrogen powered internal
combustion engine, with a post-transmission parallel hybrid system. The engine chosen
was a 2.3L 4 cylinder from a 1986 Ford Mustang SVO. This engine was chosen for its
strength and parts availability. Hydrogen was chosen as a fuel in order to stay on the
cutting edge of technology.

Objective
Computer simulations were done for the engine used in Texas Tech University’s
FutureTruck. The objective of these simulations was to maximize the power obtained
from the engine while minimizing harmful emissions. This objective also needed to be
achieved without the expense of drivability. Because the engine designed would go into
a consumer vehicle, the torque output should be reasonably flat. Engines with a very
peaky torque curve are difficult to drive. The power can come on suddenly, forcing the
4
driver to let off the gas quickly. This also means having to shift gears more often than an
engine with a smoother torque curve.

Gasoline baseline
Initially, a naturally aspirated, unmodified, engine was simulated. The
performance of the hydrogen engine is much less familiar, and complicated due to the
addition of the supercharger system. The performance of a gasoline engine is better
known and therefore gives a way to make sure that the model is producing reasonable
results.
The gasoline engine included a throttle body, the stock, cast aluminum, intake
manifold, a cast iron exhaust manifold, and a full exhaust system. Most of the variables
needed to run the model were taken from similar examples provided by Ricardo with
WAVE. The variables included in-cylinder temperatures, such as head temperatures,
cylinder liner temperatures, and piston top temperatures. Also burn duration and timing
data were taken from examples. All duct dimensions were measured from the actual
parts that were used in the engine.

Engine basics
The Otto-cycle, or four-cycle piston engine has been developed for about one
hundred years. These engines are all around us in the modern world, from lawn mowers,
to almost all cars. Numerous improvements have been made to these engines bringing
them from simple single cylinder engines with a single carburetor to multi-cylinder multi-
valve engines with sophisticated electronic fuel injection systems.
The following is a description of the four cycles of a typical spark ignition engine.
1. Intake Stroke - The fuel is mixed into the air stream before it enters the
combustion chamber. At the beginning of the intake stroke, the intake valve starts
to open, letting fresh air-fuel mixture into the cylinder. The piston pulls the air
5
into the cylinder by moving away from the valve, thus expanding the volume in
the cylinder. The air-fuel mixture being at a pressure higher than that in the
cylinder then moves into the cylinder.
2. Compression Stroke - Once the piston has moved all the way down, and begins to
move back up, the intake valve is closed. The cylinder is then essentially sealed,
and the compression of the air-fuel charge is begun. In a typical modern gasoline
engine, compression ratio is about 8:1 to11:1.
3. Expansion or Power Stroke - When the piston nears the top of its travel, a spark
plug is fired, and the air-fuel charge begins to burn, expanding in the process. As
the charge continues to burn and the piston moves down, work is done on the
piston.
4. Exhaust Stroke - Near the bottom of the piston’s travel, the exhaust valve is
opened. This provides a pathway for the burnt gasses to be expelled out of the
cylinder. The piston pushes these gases out of the cylinder as it continues to
move up.
A visual description of the four strokes can be seen in Figure 1.1. This process is
repeated dozens of times each second for high engine speeds.

Figure 1.1 Four-Stroke Operating Cycle (Heywood)
6
Controlling Fuel Flow and Spark Timing
A spark ignition engine needs to have a way to control the fuel flow and spark
timing. Traditionally controlling the fuel flow has been controlled by a carburetor, and
the ignition with a distributor. Both of these used mechanical systems to change the fuel
and spark needs of the engine over a large operating range. These systems have gone
through a lot of improvements, and by the 1980’s were very sophisticated. They still
could not meet the ever stricter emissions and fuel consumption standards, and a more
precise control was needed to replace the carburetor and mechanical distributor.
Computer control and fuel injection met this need.
Although mechanical and analog fuel injection systems have been developed and
used in production vehicles, they still share the problems faced with carburetors. The
main problem with these systems is that they are not able to precisely control the fuel
flow (TCR). This discussion will only concentrate on digital fuel injection.
Modern fuel injection systems fall into two basic types: speed density and mass
air flow, based on the primary sensors used for determining air flow. Both have been
used by the industry, although all new cars sold in the United States now primarily use
the mass air flow system to control fuel flow, with the speed density as a backup system.
The speed density system is generally used to determine the spark requirement of the
engine (Probst). All systems have a variety of sensors, sending information about the
engine operating conditions to an electronic control unit (ECU). The ECU is a computer
into which the fuel and spark requirements are programmed.

Speed density
Speed density systems use engine speed and air density, or pressure, to determine
the amount of air entering an engine. This is an indirect way of measuring airflow based
on testing done at the factory. The airflow is then programmed into the ECU as a
function of the engine speed and pressure in the intake manifold. A three dimensional
7
fuel flow map is developed, with engine speed and manifold pressure on the x- and y-
axis, and amount of fuel flow or spark advance on the z-axis.
With the use of this type of system it is difficult to know if the engine airflow has
changed over time. Common reason for this is because engine parts being changed or
modified by the owner or by regular wear and tear on the engine. Although oxygen
sensors can help this problem at light engine loads, under heavy loads the ECU is unable
to adjust for changes in the engine. As a result of this, these systems are only used as a
backup for determining fuel flow on newer cars.
The primary sensor on this type of system is the manifold absolute pressure, or
MAP, sensor. This sensor has a diaphragm with a fixed pressure on one side. The other
side is connected to a vacuum line on the intake manifold. The diaphragm moves a
potentiometer, which sends a voltage signal to the ECU.
Virtually all aftermarket fuel injection ECU’s use a speed density system for
determining air flow, including the Motec ECU used in Texas Tech University’s
FutureTruck Explorer. Two major advantages of this system are the ease of aftermarket
installation and the lack of restriction created by a mass airflow sensor. Installers of the
aftermarket systems are aware of the setbacks of the systems, but since major tuning
efforts are required in any case initially, performing minor tuning as the engine wears is
not a major problem.

Mass air flow
The mass air flow sensor (MAF), determines the amount of air flowing into the
engine directly. Most common sensors in use now have a heated wire, over which air
flows. The air cools the wire, which is kept at a constant temperature, considerably hotter
than ambient conditions. The power required to keep the wire hot is directly proportional
to the air flowing through the sensor. A two dimensional map is developed with fuel
flow versus airflow on the axes.
MAF sensors are used in newer cars as they can very accurately determine the
actual airflow going to the engine, rather than just approximating it. These sensors are
8
able to compensate for any changes in the airflow as an engine wears, or if modifications
are made.
There are some disadvantages to using MAF sensors. They are expensive to
make and are very delicate. Dirt build up can act like an insulator on the sensor wire,
giving false reading, leading to lean engine operation. These sensors act as a restriction
as all of the incoming air has to travel through the sensor. Aftermarket plumbing of these
can also cause difficulties.

Other sensors
Cold operation of a gasoline engine requires rich operation conditions. This is
because the fuel does not vaporize easily when the engine is cold. Temperature sensors
are used to modify the determined fuel requirements from the MAP or the MAF sensor
based on engine coolant temperature. Air temperature modifications are also needed as
colder air is denser than hot air.
Throttle position sensor measures the opening of the throttle butterfly. The main
reason for using these is for determining when the engine in idling as well as determining
the acceleration enrichment. Since gasoline is much denser than air, when the throttle
plate is opened quickly, the air is accelerated quicker than the gasoline. More gasoline
needs to be injected into the engine during these operating conditions in order to prevent
rough engine operation.
The crankshaft and camshaft positions are also an important piece of information
for the ECU. For most basic systems this information is used to determine the engine
speed, but it is necessary for spark timing also. A lot of the time these sensors are hall-
effects type, placed next to multi-toothed wheels. They generate a signal when a tooth is
near the sensor. Older engines usually use just a signal from the distributor in order to
determine engine speed. This usually is an accurate way of determining average speed,
but it is not accurate enough to tell if an engine has misfired.
The information from the crankshaft and camshaft position sensors can also be
used by the new OBD-II ECU’s to determine if the engine has misfired. OBD-II, or on-
9
board-diagnostics-2, was introduced in all cars sold in the United States after 1996, in
order to make diagnosing engine problems easier. These systems also need to be able to
make sure that all of the emission components and systems are properly working.
Accurate crankshaft position information is very important for many of the OBD-II
functions.
Oxygen sensors are a very important sensor in controlling emissions and lowering
fuel consumption. These sensors measure the amount of oxygen in the exhaust. Narrow
band oxygen sensors can very accurately determine stoichiometric operating conditions.
Operating close to the stoichiometric air fuel ratio minimizes the combined emissions in a
gasoline engine.
During open loop operation, the ECU looks at the sensor inputs, and sets the fuel
requirements based on tables. It is unable to see the outcome. An oxygen sensor allows
the ECU to go into closed loop operation. This means that the ECU first determines how
much fuel to inject based on the sensors and tables mentioned above. The oxygen sensor
determines if the amount of fuel injected is what the engine actually needs. This sensor
input then acts as the ultimate over-ride to the other sensor inputs. This can help to
compensate for changes in the engine, such as wear.
A major disadvantage to using conventional narrow band oxygen sensors is that
they are unable to determine how lean or rich an engine is running. They can only tell
that the engine is running rich, lean, or stoichiometric. In order to prevent engine damage
at high engine loads, a gasoline engine needs to be run rich. Torque is also maximized by
running slightly rich, equivalence ratio being about 1.176. Narrow band oxygen sensors
are unable to determine air fuel rations in this region.
Wide band oxygen sensors have become more popular in the recent years. These
are much more complicated sensors and require a separate controller. However, these are
able to accurately determine the exact equivalence ratio over a wide equivalence ratio
range. Their light off time is usually much quicker than that of a convention narrow band
oxygen sensor.
10
The wideband oxygen sensors have a narrow band sensor and a pump cell. The
pump cell is able to consume oxygen or fuel, depending on the direction of current
applied to it (Gargano). The current in the pump cell is adjusted until the narrow band
reads stoichiometric conditions. Based on the amount of current, and its direction, the
equivalence ratio can be determined. This process is adjusted many times each second.
Since wide band sensors are much more complicated, they cost several times the
price of a narrow band sensor. They also require a controller, and thus their use in
vehicles has not been very widespread. Their use is increasing rapidly there are many
carmakers using these sensors, including Audi, Cadillac, Honda, Porsche, Saturn, Subaru,
Volkswagen and Volvo (Micro-Craft). Most likely in a few years, almost all new cars
will have a wideband oxygen sensor.

Actuators
The fuel injectors have electric solenoids that move a mechanical valve. Because
of this, the injectors only have a on and an off positions, except for a very short transient
conditions. By controlling the duty cycle, or pulse width of the injectors the amount of
fuel flow can be changed.
The inlet side of the fuel injectors is supplied fuel from a fuel rail, at a pressure
few times that of atmospheric conditions. The flow of fuel through an injector is directly
proportional to the pressure differential across it. Fuel flow can be changed by changing
the pressure in the fuel rail. There is a limit to how high the pressure can be raised, since
the solenoid can not open the injector if the inlet pressure is too high. A too low pressure
will also not allow an injector to properly atomize fuel.
Fuel injectors need to atomize liquid fuels, so the fuel is able to travel with the air
flow, as well as vaporize better due to the increased surface area. Obviously gaseous
fuels do not need to vaporize. A larger orifice is also needed to flow a given mass of
vapor than a liquid, if the pressure differential is the same. Because of these differences,
injectors made for a liquid fuel will not work well, if at all, with gaseous fuels. Injectors
11
made for liquid fuels can usually be made to work with almost any liquid fuel. Also,
gaseous injectors will work with most gases. This rule of thumb works as long as the
seals and other materials in the injectors are compatible with the fuel.
There are two basic ways of operating injectors in a multiple injector engine. The
most basic one is batch firing, while a more sophisticated one is sequential. Batch firing
operates all of the injectors simultaneously, usually twice for each engine revolution.
This does not require the ECU to know the position of the crankshaft, and makes the
firmware in the ECU much simpler.
Sequential firing operates each injector separately. The end of the injection is
made to coincide with the time the intake valve is about to close. This requires a separate
injector driver for each injector in the ECU. It also requires crankshaft and camshaft
position signals. This makes the electronic hardware and firmware in the ECU much
more complicated.
Sequential injector timing has virtually no advantage at high engine speeds and
loads. At these operating conditions, the injectors are open most of the time. The time
when the injector is not open makes very little difference when it is open 60-90% of the
time. Sequential injection does improve the operation of the engine at low speeds and
loads, when the injectors are closed most of the time. Idle quality and emissions are
improved, and as a result, virtually all the new vehicles now use sequential injection.

Hydrogen properties
Hydrogen as a fuel has some properties which vary greatly from gasoline or other
fuels. Table 1.1 has most of the important properties listed for hydrogen and gasoline.
The ones which are important to the operation of an internal combustion engine will be
listed and discussed shortly below.
1. Boiling Point – The extremely low boiling point of hydrogen means that it will
always be a vapor once it enters the engine, although onboard storage of it can be
in liquid form. Gasoline on the other hand is injected as a liquid, which partially
12
will evaporate in the engine before combustion. Vapor fuels take up space in the
intake manifold as well as the combustion chamber, lowering the volumetric
efficiency.
2. Density – The low density of hydrogen means that it will take up more space in
the intake manifold and the combustion chamber, even when compared to other
vapor fuels, such as natural gas and propane.
3. Octane Rating – The high octane rating of hydrogen allows the possibility of
increased compression ratio, which increases thermal efficiency.
4. Quenching Gap – This is the largest gap that still prevents flame front propagation
(Alcock). Since this is smaller than that for most other fuels, the flame front can
escape the combustion chamber through the intake valve more easily, resulting in
a back flash. It might also be possible to get the flame front to go past the piston
rings.
5. Limits of Flammability – Hydrogen has a very large difference between the upper
and lower limits of flammability. This allows hydrogen to burn over a much
wider range of mixtures, from extreme lean to rich. Gasoline and most other fuels
will not burn if the mixture is not near stoichiometric. This property allows a
hydrogen engine to run extremely lean, but also makes hydrogen more dangerous
as almost any mixtures of hydrogen and air will burn.
6. Minimum Ignition Energy – The low ignition energy of hydrogen can be very
problematic in an engine as it can ignite from even the smallest hot spots in the
combustion chamber.
7. Auto-ignition Temperature – Although this temperature is higher for hydrogen
than gasoline, the low ignition energy of hydrogen makes it more likely to pre-
ignite, or even back flash.
8. Burning Velocity – Hydrogen burns extremely quickly when compared to other
fuels. This means that the combustion process takes up only a few crankshaft
degrees, and the pressure changes are almost instantaneous. This requires either
very little spark advance or even retard timing. This property also means that any
13
pre-ignition that happens is much more violent, and can damage the engine faster
than that occurring with gasoline.

Table 1.1 Physical Properties of Hydrogen and Gasoline (Dempsey, Turns)
Property Hydrogen Gasoline
Specific Gravity at STP relative to air 0.07 ~ 4.0
Normal Boiling Point (K) 20.3 310-478
Critical Pressure (atm) 12.8 24.5-27
Density of Liquid at STP (kg/L) 0.0708 ~ 0.70
Density of Gas at STP (kg/m
3
) 0.838 ~ 4.40
Density Ratio, STP 845 ~ 150
Octane Rating 130+ 86-110
Thermal Diffusivity in STP air (cm
2
/s) 0.61 ~0.05
Diffusion Velocity in STP air (cm/s) ~ 2 ~ 0.34
Quenching Gap in STP air (mm) 0.64 2
Limits of Flammability in air Vol. (%) 4-75 1-7.6
Limits of Detonation in air Vol. (%) 18.3-59 1.1-3.3
Minimum Energy for Ignition in air (mJ) 0.02 0.24
Auto-ignition Temperature (K) 858 501-744
Maximum Burning Velocity in STP air 278 37-43
Flame Temperature in air (K) 2318 2470
There are major differences in the properties of hydrogen and gasoline. Hydrogen
is a fuel which ignites very easily at a wide range of concentrations and burns very
quickly. Gasoline on the other hand is harder to ignite, burns relatively slowly and
requires a near stoichiometric mixture. Gasoline is also much denser, and will puddle on
the ground for a long time if spilled, when hydrogen will disperse in the air very quickly.

Differences Converting an Engine from Gasoline to Hydrogen
Biggest problem with running an engine on hydrogen is pre-ignition and back-
flash. Pre-ignition occurs when the air-fuel charge ignites before the spark plug is fired.
When this occurs during the middle of the compression cycle, severe engine damage can
result. Back-flash is pre-ignition that happens when the intake valve is still open. This
14
explosion will travel through the intake manifold to the air intake. They are much louder,
but less destructive to the engine, as the high pressure is not contained in the cylinder.
A few possible sources of pre-ignition have been suggested. They include hot
spots from carbon deposits or sharp edges in the combustion chamber. Hot exhaust
gasses could also start pre-ignition. A way around the pre-ignition problem is running
the engine leaner, but this decreases power. Injecting the fuel when the intake valve is
open helps cool any hot spots by not introducing hydrogen into the combustion chamber
immediately after the intake valve opens.
When converting a gasoline engine to run on hydrogen, several changes should be
made, many to avoid pre-ignition. The most important and obvious one is changing the
fuel delivery and control systems. In fuel injected gasoline engines, the liquid fuel is
pumped from a fuel tank at a constant volume to the fuel injectors by a positive
displacement pump. Usually a bypass regulator is installed after the fuel injectors, letting
unused fuel back to the fuel tank.
High pressure tanks generally store hydrogen onboard of vehicles. This makes a
fuel pump unnecessary; however regulators are needed to bring the fuel pressure to the
levels required by the injectors. Often, two regulators are needed in order to avoid
variations in fuel pressure. Most pressure regulators will increase the outlet pressure as
the inlet pressure decreases a great deal, which is the case when high pressure tanks are
used. The first regulator brings the pressure down substantially. The second one is able
to keep the pressure constant, even if there is some variation in the pressure output of the
first regulator.
Fuel lines need to be changed. Stainless steel tubing is usually used with
hydrogen. It is possible to use aluminum, plastic, and steel with gasoline. The hydrogen
would most likely leak out of these tubes and their junctions over time.
Most fuel injectors are roughly the same size externally, allowing the same intake
manifold to be used. Even the fuel rail should fit the hydrogen injector. However, the
fuel rails for liquid fuels are very small in diameter (roughly a centimeter). A large
diameter fuel rail is needed in order to dampen pressure pulses and to serve as a reservoir
15
for the injectors. The fuel rail used in Texas Tech’s FutureTruck Explorer has an inside
diameter of one inch. This size was recommended by hydrogen engine engineers at Ford
Motor Company.
Although the stock ECU’s can operate hydrogen injectors, they are not
programmed for use with hydrogen. Engine controllers which are programmed for use
with hydrogen are not available commercially. The high performance aftermarket does
come into rescue, with user programmable ECU’s. Several manufacturers make engine
controllers which can be used with hydrogen. Being able to control injectors sequentially
is a feature that roughly half of the aftermarket engine controllers have. A production
hydrogen engine could use an ECU made for gasoline engine with just a change of
programming.
A Motec M4 ECU was chosen for Texas Tech’s FutureTruck. This is a very
sophisticated aftermarket controller, being able to control 4 injectors sequentially, control
multiple coil packs, and use a large variety of sensors including a wideband oxygen
sensor. The M4 also has 512kb internal data logger, which can record a large amount of
sensor data (Motec). This is useful when tuning the engine as the ECU can be set to
always log data. The data can then be used to change the fuel and timing maps when the
engine is not running. This can save time and fuel while tuning.
The material used in most older cylinder heads and engine block is cast iron. In
the past twenty years, the use of aluminum has increased dramatically due to the weight
savings and increased heat conduction. Since aluminum conducts heat much better, it
can reduce the likelihood of pre-ignition. This is a very desirable property when using
hydrogen as a fuel in an engine.
Because hydrogen has much higher octane rating than gasoline, higher
compression ratio can be used. A higher compression ratio increases engine efficiency
by increasing cylinder pressure during the power stroke. In a study done by Heffel et al.,
the thermal efficiency was compared while changing compression ratio. All testing was
done at 3000 rpm and WOT. The best brake thermal efficiency was attained at 14.5:1
16
compression ratio. At 12.5:1 and 15.3:1, the efficiency dropped. The efficiencies can be
seen in Figure 1.2.
Theoretically, higher compression ratio should give a higher efficiency
(Heywood). The drop in the efficiency of the test engine as the compression ratio was
increased to 15.3:1 could be a result of increased work that needs to be done to compress
the air-fuel mixture before the combustion. Since hydrogen burns very fast, it is possible
that there is very little increase in the work done on the piston during the expansion
stroke as the compression ratio is increased. Another possibility is that the increased
cylinder temperatures before ignition promote pre-ignition, requiring retarded ignition
timing, which lowers the power output of the engine.

Figure 1.2 Brake thermal efficiency versus equivalence ratio as the compression ratio is
varied. Data is from dynamometer testing done by Heffel et al., for a hydrogen engine
running at 3000 rpm at WOT.
Changing the compression ratio can be done by decreasing the combustion
chamber volume or increasing the engine displacement. Increasing the displacement is
17
possible by boring the cylinder, but this is only marginal, since cylinder walls need to
remain thick enough to withstand combustion pressures. Increasing the stroke can
increase the compression ratio, but usually requires a new crankshaft, connecting rods
and pistons.
Decreasing the combustion chamber volume is usually accomplished easier by
milling the cylinder head. Changing the pistons to one’s that protrude past the deck
height also increases the compression ratio. The cylinder head used in Texas Tech’s
FutureTruck was changed to an aftermarket version made of aluminum. This head was
also milled to produce a compression ratio of about 12.6:1.
Sharp edges can stay very hot during engine operation, and thus lead do pre-
ignition. Therefore the elimination of these might be desirable. This can be
accomplished prior to cylinder head assembly by smoothing out all internal surfaces.
Polishing the combustion chamber surfaces might reduce pre-ignition further.
Cooler cylinder temperatures will reduce likelihood of pre-ignition. This can be
accomplished by installing a lower temperature thermostat. A downside to this is
increased engine wear and efficiency. Engines today use thermostats which keep the
coolant temperature at about 100 C. Thermostats opening at about 80 to 85 C are
commonly used in modified, high performance engines. This type of thermostat was
installed in Texas Tech’s FutureTruck.

Running an engine on H2
Considerable difficulties were encountered trying to initially start the test engine
used in the FutureTruck. The main reason for the difficulties was the lack of base fuel
and ignition timing maps for the ECU. Another part of the problem was inability to
accurately check ignition timing on the engine. This was caused because the engine
chosen did not originally have crankshaft and camshaft sensors. The addition of the
crankshaft sensor required removing the timing indicator.
18
Once the initial difficulties were solved, and the engine started, more difficulties
were caused by back-flash and detonation. Reducing the injected fuel amount generally
will reduce this, but power suffers dramatically. As a general rule, a conventional, port
injected hydrogen engine will not run well at equivalence ratios above 0.75 at high loads.
It is possible to run the engine at these equivalence ratios, and possible a little above, at
very low loads, such as idling.
The simulation done in this study cover a broad range of equivalence ratios from
very lean up to stoichiometric. The stoichiometric simulations are done as a reference,
although running a conventional engine so rich is not feasible.

Hybrid system basics
Conventional internal combustion engines have traditionally been used in
vehicles. They posses good power to weight and volume, and have long operating
ranges. They do have some disadvantages such as having poorly matched fuel efficiency
characteristics with operation of a normal driving, leading to poor fuel economy; being
unable to harness the vehicle’s kinetic energy during braking, and a low efficiency of
vehicle transmissions, especially in stop-and-go traffic. The internal combustion engines
also produce higher emissions since they have to operate over a broad speed and load
ranges. Electric vehicle overcome the tailpipe emissions, but batteries have a poor
energy density, reducing the operating range (Ehsani).
Hybrid vehicles can take the good characteristics of both internal combustion
engines and electric drive trains, being able to lower the emissions and fuel consumption
considerably. The way these systems work is by using a smaller displacement engine in
conjunction with an electric motor. The smaller engine makes enough power to
comfortably allow the vehicle to cruise, but does not necessarily have the power to
accelerate the vehicle fast enough while merging and going up hills. The total power of
the vehicle is then brought up with the use of the electric motor. This idea is not new, in
19
fact, a patent was filed in 1905 for a vehicle that used an electric motor to augment the
internal combustion engine by H. Piper (Wouk).
A newer use of the hybrid system is regeneration. A large amount of the kinetic
energy of a vehicle can be recovered by using the electric motor as a generator to
recharge the batteries. In a conventional vehicle this energy is lost through heat in the
brakes. It is not possible to regain all of the kinetic energy. Conventional, hydraulic
brakes are still needed to bring the vehicle to a complete stop as it is difficult to control
an electric motor at low speeds. Also some of the energy is lost to friction and
aerodynamic drag.
Batteries play an important role in hybrid-electric vehicles. They store and
release energy as needed. This allows the engine to work at a lower, more constant
output, making less emissions and lowering fuel consumption. The batteries are charged
during cruising and decelerating, when the engine is making more power than needed by
the propulsion of the vehicle.

Types of Hybrids
Traditionally, there have been two basic types of hybrids: parallel and series. In
the resent years, more ways of incorporating the engine and electric motor to the drive
train have been invented, which fit into neither one of the basic categories. This has
given rise to the series-parallel hybrid, which can run in either one of the basic modes,
and the complex hybrid, which incorporate more than one electric motor. Each one of
them has its advantages and disadvantages. This discussion will focus only on the more
conventional series and parallel hybrid drive trains.
Series hybrid-electric systems
In a series hybrid-electric vehicle the energy flows in a series from one
component to another. The engine drives an electric generator which provides electric
power to the drive motors and charges batteries. These vehicles are set up very similarly
20
to an all-electric vehicle, as there is no way to transfer mechanical power directly to the
wheels from the engine.
This system has the advantage that the vehicle and engine speeds are not coupled
to each other. It is possible to operate the engine at its peak efficiency and tune it for best
emissions at that operating point. The engine provides a constant power output, while
power from the batteries is used to fill in the power needs of acceleration and recapture
the energy during braking. A major drawback to this system is that there are significant
energy losses while converting energy from mechanical to electric and back to
mechanical. Another drawback is that a high power, and thus heavy, electric motor is
needed to accelerate the vehicle since it is the sole source of propulsion.

Parallel hybrid-electric systems
In this system power can flow from either the electric motor or the engine to the
wheels. An electric motor is placed in parallel to the drive train. This kind of system is
easier to incorporate into an existing design since the basic engine and transmission
system are left in place. There are three basic parallel systems: pre-transmission, post-
transmission and road parallel.

Pre-transmission parallel
In this hybrid-electric type the electric motor is placed in the drive-train before the
transmission. It is possible to have the motor directly coupled to the engine, requiring
them to spin at the same speeds, or through a system of clutches so that the motor and/or
the engine can be uncoupled from the transmission. Advantages of this system are that
the torque of the electric motor is multiplied through the transmission and it is possible to
charge batteries while the vehicle is not moving. This does mean that there are power
losses through the transmission. Also this design can make regenerating while
decelerating difficult as some transmissions do not efficiently transfer power backwards.
21

Post-transmission parallel
This set up gets rid of any regeneration problems that might occur in the pre-
transmission systems. The electric motor is placed after the transmission, and will
always turn at a speed relative to the vehicle speed. Two disadvantages of this are that
there is no torque multiplication through the transmission and it is not possible to charge
batteries while the vehicle is not moving. The torque multiplication can be attained by
coupling the motor through a step down gear ratio. There is no simple solution to
charging the batteries while the vehicle is not in motion.
An advantage to this system is the ease of incorporating it, especially into an
existing rear wheel drive design. The engine and transmission can be left intact, while all
of the hardware is added to the space in between the transmission and the rear end. Texas
Tech’s FutureTruck is this type of design. The electric motor is a Solectia AC 55, with a
modified shaft with splines on both ends. This allows a drive shaft to connect the
transmission to the electric motor, and another one to connect the motor to the rear end.
In essence, a piece of the drive shaft was replaced with the electric motor.

Road parallel
The road parallel system incorporates two separate, unconnected, propulsion
systems driving the front and rear wheels. Usually the engine and transmission are in a
front wheel drive configuration, while the electric motor drives the rear wheels. This
system has the advantage that the room in between the front and rear wheels is not taken
up or split by the drive shaft. This can leave the space for use by batteries, electric
motors, controllers, fuel tank, etc. This system does share most of the disadvantages of
the post-transmission parallel system including not having the capability to charge
batteries while the vehicle is not in motion.
22
CHAPTER II
ENGINE SIMULATIONS

Ever since the first engines were built, designers have wanted to simulate their
engines prior to building them. It was realized early that several different analyses need
to be done in order to accurately model an engine. These include performing fluid-
dynamics for the air and fuel flow to and from the cylinder, thermodynamics for the
combustion process, heat-transfer for the energy losses to the cylinder walls, and kinetics
for the frictional losses (Atherton).
During the late 1800’s and early 1900’s, only rough combustion analysis could be
done. As knowledge of fluid flow increased, it became possible to do steady state models
of simple ducts of an engine. The ducts in a multi-cylinder engine can be very
complicated, and thus the early designers could not easily model them. Also, air flow
through these ducts is far from being a steady state process as valves open and close
several times each second even for low engine speeds, such as idling.
With the advent of computers, it became possible to perform thousands of
calculations in a very short time. Calculations for many small pieces of the ducts could
be done. Each piece of ducting is looked at separately, and as the results are calculated
for one section, those conditions are used as the boundary conditions for the next section.
The calculations for each section are done for small increments of the engine position.
This can accurately model an engine, even if the model is one-dimensional. Ricardo
WAVE, which was used in this study, is this kind of a program.
In the Texas Tech FutureTruck project, two basic models were simulated. One
was a naturally aspirated, unmodified, gasoline engine. This was done as a baseline, and
as a way to verify that the simulation results were reasonable. This engine model was
modified, much like the real life test engine used in the FutureTruck Explorer. The
modifications included the following: adding a supercharger and an intercooler,
increasing the compression ratio, changing the cylinder head, camshaft timing, and
exhaust manifold.
23
Ricardo WAVE background
Ricardo is a large, multinational corporation that does research for several major
auto manufacturers. Ricardo has developed the engine simulation package WAVE to
analyze the dynamics of pressure waves, mass flows, and energy losses in ducts,
plenums, and manifolds of various systems and machines. This program provides fully
integrated time-dependent fluid dynamic and thermodynamic calculations using a one-
dimensional formulation. WAVE uses a general treatment of working fluids including
air, air-hydrocarbon mixtures, combustion products, and liquid fuels. (Ricardo)

Results given
WAVE is capable of giving many different results, and is not limited to just
calculating torque curves. It can also give emission data, as well as fuel consumption
results. This study concentrated on the power, torque, volumetric efficiency, and
emission results.
Once a model is run, the results should be correlated with dynamometer data.
This assures that changes made on the model will be very accurate, and do not
necessarily need to be checked with the dynamometer. Because of time and equipment
limitations, the initial correlation was not done for this study. This should not make the
simulation results unusable; general trends shown by the simulations should be correct.

Description of how WAVE calculates air flow
WAVE uses a quasi-one dimensional compressible model to calculate the airflow
using conservation of mass, momentum, and energy. The equations are:
Mass
¿
= m
dt
dm
&
Energy
¿
+ = sources h m
dt
dme
&
24
Momentum
¿
÷ + ÷ = losses u m dx
dx
dp
A
dt
dmu
&

How models are built
The engine models in WAVE are built using the dimensions of each runner that
the air goes through from the air intake to the tip of the exhaust. Every volume is
assumed to be either a cut-off cone or a sphere. Bends are entered separately, and the
model takes the energy losses into account, although they might not be shown on the
visualization of each runner.
Figure 2.1 shows the model used for the supercharged hydrogen engine. Each
block on the model represents a junction between ducts, while the lines are the ducts.
The model starts at the top-right corner, on an ambient junction which connects to
the supercharger. The intercooler is located on the top, four ducts after the supercharger.
More ducts connect the intercooler to the intake manifold and the plenum, from which
four separate ducts direct the air to the intake manifold. The fuel injectors can be seen on
the intake manifold. The intake ports connect to the cylinder, with similar exhaust ports
on the right side of the cylinders. The exhaust manifold is 4-into-2-into-1 type. This is
connected to the catalytic converter. The muffler is modeled by the dead ended duct,
almost at the end of the model, next to another ambient junction.
Large volumes and junctions are modeled using either simple- or complex-
junctions. The simple-junctions are the round ones on the model, and are modeled as
spheres. The complex-junctions are the square junctions in the model and represent more
duct-like volumes. They are recommended for use with junctions containing perforated
walls (such as mufflers), or for volumes with many parallel passages (such as intercoolers
and catalytic converter inlets and outlets) (Ricardo).
Figure 2.2 shows the input panel for the junctions. The junctions set the positions
at which the ducts come into and out of the junction in three dimensions. This is done by
setting the angles from each axis to the center of the duct. The junctions assume that all
25
ducts originate from the center of the junction. In the example, all of the ducts are in a
single plane since the angle from y-axis is 90 degrees for all of them. They do not have
to be necessarily in a single plane. The angles used to define the direction have to agree
with each other, i.e. it is not possible to say that the angles are 10 degrees from every
axis, as the directions would not coincide.

Figure 2.3 Ricardo WAVE model of the supercharged hydrogen engine.
The basic shapes make it somewhat difficult to model complex shapes such as
intake and exhaust runners. The actual dimensions of some parts are very difficult to
accurately measure without making molds of the runner shapes or cutting the part open.
These setbacks can mostly be overcome as actual flow data can be entered for any runner
in the model, yielding very accurate models.
26


Figure 2.4 A visual representation of a junction. This is the junction of two exhaust
manifold passages coming to form one. All of the tubes coming to the junction can be set
anywhere in three dimensions, although all of these are in a single plane.

Discretization lengths
Figure 2.3 shows the input panel for each duct. The main points to note on it are
the end diameters, overall length, and bend angle. The discretization length is also an
important part of the model. It sets the number of different sub-volumes that WAVE uses
to calculate the airflow. The equation for conservation of momentum is solved at the
boundary of each volume. The shorter this is set, the more calculations need to be done,
and the model will be slower. This cannot be set too long, as then the accuracy of the
model is compromised. Ricardo recommends using the following for discretization
lengths:
27
Intake side: dx = 0.45 * Bore
Exhaust side: dx = 0.55 * Bore
On the models used, these recommendations were usually used, or rounded down.

Figure 2.5 View of the input panel for each duct through which air travels. The diameters
at each end, the bend angle, and lengths are entered. The dotted lines dividing the duct
split it into several sub-volumes. Note the injector placed toward the right end of the
duct, one of the intake runners in this case.
Throttle
A butterfly valve is used as a throttle on most engines. Since this is a relatively
complicated piece to model, a simpler approach is used. The throttle on WAVE models
is made by placing a junction which is smaller than the tubes attached to it. For these
simulations, the tubes attached to the throttle were 2.0 inches in diameter. The butterfly
valve does restrict the airflow even when it is completely open, thus at WOT conditions,
a 1.8 inch diameter hole was used in the junction. In order to throttle the engine, the
diameter of the junction could be made smaller. At idle conditions, the hole needs to be
very small.

28
Simulating the intercooler and the catalytic converter
The intercooler and the catalytic converter are two pieces that cannot be modeled
like regular ducts. This is because they have multiple small ducts. The way WAVE
recommends modeling them is by using regular spheres for the open ends of them and
attaching many small ducts to these. The small ducts only need to be entered as a single
duct, and specifying that there is a number of these.
For the intercooler the number of small ducts was counted to be 17. Their size
could be measured with reasonable accuracy. The catalytic converter was welded into
the exhaust, and thus it was not possible to count the ducts or measure their exact size.
Based on the examples in WAVE, and by using the cross sectional area of the catalytic
converter, the number of ducts was estimated to be 1500.

Simulating cylinder heads
Flow data for the cylinder head on the engine were entered into WAVE. The data
were measured by two separate sources on identical Esslinger aluminum heads
(mercurencyclopedia, turboford). The data were virtually identical, and therefore should
be a good representation of what the head used on the experimental engine flowed. The
data can be seen in Table 2.1. The data used in the simulation were from
mercurencyclopedia as that had data for higher valve lifts. The other source was used
only to check the validity of the data.


29
Table 2.2 Cylinder Flow Data from Two Sources (cfm)

Valve
Lift
(inches) 0.1 0.2 0.3 0.4 0.5 0.6 0.7
mercurencyclopedia.com Intake 65 117 162 201 228 241 247
turboford.org Intake 65 116 162 201 228 241

mercurencyclopedia.com Exhaust 52 101 135 154 172 183 192
turboford.org Exhaust 51 101 134 154 172 183

The table above shows how much air will flow through a single port in the
cylinder head. The measurements were taken at the standard pressure difference of 28
inches of water. At the top, the valve lift is indicated. The maximum valve lift for the
model is approximately 0.4 inches. Camshaft lifts greater than this can be used in
engines, but very stiff valve springs are required if high valve lifts are used. Stiff valve
springs can also lead to premature camshaft wear.

Simulating the camshaft
The camshaft profile was measured using Cam Doctor software. This software
gives all of the camshaft specifications, including lift, duration, lobe center separation,
etc. The information received from Cam Doctor was used to modify a generic camshaft
profile that comes with WAVE. Using a generic profile does introduce some error;
however, this should not be significant. Most modern camshafts try to open the valves as
quickly as possible, and therefore there should be minimal differences in the profiles.
Table 2.2 shows the overview of the camshaft profile.







30
Table 2.3 Partial Summary of Camshaft Specifications
Lobe center separation 121.7 Cam deg
Valve overlap -53.7 Crank deg

Intake Valve
Duration (at 0.050 in) 202.5 Crank deg
Max cam lift 0.23713 Inches

Exhaust Valve
Duration (at 0.050 in) 205.3 Crank deg
Max cam lift 0.23791 Inches

Supercharger modeling
A Rotrex centrifugal, belt driven supercharger was used in the hydrogen model.
The total ratio of the turbine speed to engine speed is 22.752:1. This includes an internal
ratio of 9.48:1 inside the supercharger, and a ratio of the belt pulleys of 2.4:1. Rotrex
provided the efficiency data for the supercharger, which compared the mass flow of air,
pressure ratio of inlet and outlets, turbine speed, and efficiency. Seventy data points were
entered from the Rotrex data into WAVE. WAVE has a function to create the continuous
four-dimensional compressor map. Both of the efficiency curves are shown below in
Figures 2.4 and 2.5 to indicate that they are the same.
31

Figure 2.6 Supercharger Efficiency Curves as Given by Rotrex

Figure 2.7 Compressor map as entered into WAVE
The compressor maps above show the supercharger efficiency. The horizontal
axis shows the mass flow of air through the supercharger. The vertical axes show the
pressure ratio, or the ratio of the absolute outlet pressure to the absolute inlet pressure.
Lines which run approximately horizontal in Figure 2.4 and continue almost vertically
32
down in Figure 2.5 indicate the turbine speed. Curves toward the bottom of the figure are
slower than the ones above. The islands, which are colored in Figure 2.5, are constant
turbine efficiencies. Ideally, an engine will operate as close to the efficiencies in the
middle of the curve, as these are the highest.
WAVE automatically subtracts the power used to drive the supercharger from the
engine output. The power to drive the supercharger is taken form this efficiency map.

Combustion model
The combustion model used by WAVE is Wiebe based (Ricardo). This model
calculates the rate of mass that is burned. The formula used by WAVE is:
( )
|
|
.
|

\
|
|
.
|

\
| A
÷ ÷ =
+1
exp 1
WEXP
BDUR
AWI W
u

where: W = cumulative mass fraction burned
u A = crank degrees past start of combustion
BDUR = user-entered 10-90% burn duration in crank degrees
WEXP = user-entered Wiebe exponent
AWI = internally calculated parameter to allow BDUR to cover the range of
10-90%
The model requires inputs for the duration, the Wiebe exponent, how much of the
fuel gets burned, as well as the timing when the combustion happens. The duration is the
number of crank degrees for 80% of the combustion to occur. This does not include the
first or the last 10% of the combustion. The timing is entered as crank degrees for half of
the combustion and heat release to have occurred (50% burn point).
The Wiebe exponent controls how fast the combustion happens relative to the
50% burn point. A smaller (less than 2) Wiebe exponent burns the fuel quickly initially
and slowly after the 50% burn point. A larger exponent starts the combustion process
very slowly initially, and quickly later on. WAVE recommends using 2 as the Wiebe
exponent, and thus that was used in all simulations.
33
The amount of fuel burned needs to be set also. For the hydrogen simulations, the
fraction of fuel burned was always set to 1.0, because of the extremely lean mixtures that
the hydrogen engines run. All of the available fuel should burn in the engine. This
fraction needs to be set to less than one for engines running under rich mixtures, such as a
gasoline engine running under a high load.

Calculating the burn duration
In order to get accurate results, cylinder pressure data would be needed from an
engine running on a dynamometer. These pressure sensors are expensive and difficult to
obtain, so these data were not available. The duration inputs from the example gasoline
engines were used, along with multipliers to correct them to the hydrogen engine. The
multipliers were the ratios of the calculated, theoretical flame speeds.
The theoretical flame speed is a function of temperature and pressure. Average
figures for these were taken from the engine simulations, for the 80% burn duration.
Initially, a 1 to 10 fraction of indoline to hydrogen flame speeds was set. The resulting
average pressure and temperatures were used to calculate new flame speeds for the
hydrogen model, and the burn duration was adjusted accordingly. This iteration process
was used until the burn duration for the hydrogen was within 0.5 degree of the previous
iteration for the hydrogen model.
The following is an example of the burn duration for the simulations running at
3000 rpm. The same procedure was used for 1500 and 4500 rpm. Linear curve fitting
between these points was used to obtain the burn duration for the engine speed around
these points.
The flame speed for the V8 example indolene engine, running at 12:1 air fuel
ratio (Ö=1.125) was calculated from the following (Turns). This engine was chosen as it
had the closest bore and stroke, as well as bore/store ratio to the engine being simulated.
The reference flame speed, at pressure of 1 atmosphere, and 298K temperature, is:

34
( )
2
2 m m ref
B B S u ÷ u + =
=26.87cm/s.
The flame speed in the combustion chamber should be:
| ¸
|
|
.
|

\
|
|
|
.
|

\
|
=
ref ref
ref
P
P
T
T
S S = 699cm/s.
Where: ( ) 1 8 , 0 18 . 2 ÷ u ÷ = ¸ = 2
( ) 1 22 . 0 16 . 0 ÷ u + ÷ = | = -0.1105
P=45 atm, this is an average pressure in the combustion chamber during
combustion, taken from P-V diagrams after running WAVE for the example model.
T=1876 K average temperature for the example engine in the combustion
chamber during combustion.
The flame speed calculation for the hydrogen simulation running at 55:1 air fuel
ration (Ö=0.62) uses the same formula as above, with the exception of the following
coefficients:
( ) 1 8 , 0 18 . 2 ÷ u ÷ = ¸ = 2.4855
( ) 1 22 . 0 16 . 0 ÷ u + ÷ = | = -0.244
P=80 atm
T=1846 K
The reference flame speed, at 1 atm, 298 K and stoichiometric mixture, is 210
cm/s (Turns).
The pressure and temperature figures are the ones that were calculated at the end
of the iteration process. They resulted in the flame speed of 6712cm/s. The average
temperature and pressure were taken for the burn timing that resulted in maximum
torque.
The flame speed ration is then 6712:699, or 9.6:1. This means that the burn
duration in the indoline engine of 28.2 degrees corresponds to 2.9 degrees in the
hydrogen engine. This burn duration was entered back into WAVE and simulations with
multiple spark timings were run. The one which made maximum torque was picked to
35
supply new temperature and pressure data, and the flame speed calculation process was
repeated until the same results were obtained as the previous run.

Emission
Emissions have played an increasingly significant role in automotive development
for the past several decades. Today emissions cannot be ignored, as the norms faced by
the vehicle manufacturers are getting increasingly more stringent. The main emissions
measured today are hydrocarbons (HC), carbon-monoxide (CO), and nitrous-oxides
(NO
x
).
The HC and CO emissions are extremely low in an engine using hydrogen as a
fuel. This is due to the fact that the fuel has no hydrocarbons. The HC emissions are
mainly made by unburned fuels. The lack of hydrocarbons also keeps the CO emissions
very low. Ideally all of the carbon in the hydrocarbons would be converted to CO
2
.
When the chemical reactions are not completed during combustion, some CO is made.
Although at first it seem unlikely to make any CO and HC emissions in a
hydrogen engine, some amounts can still be made. This is due to burning the lubricating
oil in the engine. The oil is not intended to be burned, and steps have been taken to
minimize getting it to the combustion chamber. The oil can make its way there past the
piston rings, leakage through the intake valve guide, or through the crankcase ventilation
system. Some solutions to minimize these are: run a tighter piston ring gap, producing
higher piston ring to cylinder wall pressure; improving the valve guide seals to prevent
oil from getting past the guides; and improving the oil trap for the crankcase ventilation
system. This simulation did not take into account any of the HC and CO emissions, and
only concentrated on the NO
x
emissions.
Using synthetic oils has been said to decrease the emissions. These oils are not
made with hydrocarbons, and therefore do not contribute to those emissions. They also
are slicker, decreasing the internal friction in the engine. In some cases it is not possible
to run synthetic oils. This is the case of the Texas Tech’s FutureTruck Explorer. The
36
supercharger in it uses the engine oil for its lubrication. The supercharger uses planetary
roller traction gears, shown in Figure 2.6. The power comes in through the three smaller
rollers inside the larger ones. The larger rollers then transmit the power to the output
shaft placed in the middle. These gears use a thin film of oil to transmit power from one
smooth gear to another. This is shown in the close up view in Figure 2.7. The synthetic
fluids are so slick that they would let the gears slip more than the designed amount, thus
transmitting the power poorly to the compressor wheel.


Figure 2.8 Traction gears in the supercharger (Rotrex).

Figure 2.9 Oil being used to transmit power in the supercharger (Rotrex).

37
CHAPTER III

SIMULATION RESULTS

Spark Timing
The spark timing is set indirectly in WAVE. The crank angle when half of the
fuel has burned is specified, and the burn duration then sets the spark timing. In order to
determine the optimal timing, for wide open throttle runs, a sweep of different timing
settings were ran. The optimal timing for WOT was taken to be the one that maximized
torque, although this increases emissions. This can be seen in Figure 3.1.
80
90
100
110
120
130
140
-20 -10 0 10 20 30 40
Spark timing (degrees before TDC)
T
o
r
q
u
e

(
N
-
m
)

Figure 3.1 Brake torque v. spark timing for supercharged hydrogen engine running at
3000 rpm, 55:1 AFR and WOT.
It can clearly be seen from the graph that having too much spark advance reduces
the torque as the in-cylinder pressure increases before the piston reaches TDC, doing
negative work on the crankshaft. As the fuel mixture is lighted too late, less work is done
on the piston since it has already begun to move down.
The surprising effect of the timing is that maximum torque is produced, not at
slightly advanced, or no advance, but at a retarded timing. The maximum torque for the
38
sample case occurred at retarded spark timing around 7 degrees. Possible reason for this
is that the cylinder pressure is very high when the piston is at TDC increasing blow-by
past the piston rings. At the same time, the work done near TDC is a lot less than when
the piston has moved away from it. Thus the net effect is that the maximum torque
occurs at slightly retarded spark timing.
The spark timing directly affects the cylinder temperatures and pressures.
Because of this, it changes the emissions, especially the NO
x
. The in-cylinder pressure
for different crank angle can be seen in Figure 3.2. The beginning of the combustion can
be clearly seen in the graphs. Another note worth mentioning is that even slightly
advanced timing produces very high pressures that fight the piston moving up to TDC.
0
400
800
1200
1600
2000
-45 -30 -15 0 15 30 45
Crank Angle (degrees)
C
y
l
i
n
d
e
r

P
r
e
s
s
u
r
e

(
K
P
a
)
-14deg
-7 deg
0 deg
7 deg
14 deg
21 deg

Figure 3.2 Cylinder pressures v. crank angle for different ignition timing at 3000 rpm and
55:1 afr.
Beginning of the combustion process can also be seen in the in-cylinder
temperature graphs. Figures 3.3 and 3.4 also show the effect of ignition timing in
cylinder temperatures. The fast flame speed of hydrogen can be seen in the graphs as the
temperature increases almost the same amount, for different timing settings. Only small
differences in the maximum temperature can be seen due to the cooling effect of the
cylinder volume increasing for the retarded spark timing. The advanced timing
39
temperatures do go slightly higher than the retarded due to the cylinder volume
decreasing after ignition.
The in-cylinder pressures have very small differences after the combustion is
completed. The temperatures however take a longer time to fall, and thus are different
until after the intake valve opens. This is because the heat transfer takes time to occur.
Only after fresh air enters the cylinder, do the temperatures return to the same level. This
can be seen in Figure 3.4.
0
1000
2000
3000
4000
5000
-90 0 90 180 270 360 450 540
Crank Angle (Degrees)
T
e
m
p
e
r
a
t
u
r
e

(
K
)
-14 deg
0 deg
14 deg

Figure 3.3 In-cylinder temperatures for three different spark advances on the
supercharged hydrogen model running at 3000 rpm, 55:1 AFR and WOT.

40
0
1000
2000
3000
4000
5000
-30 -15 0 15 30 45 60
Crank Angle (degrees)
T
e
m
p
e
r
a
t
u
r
e

(
K
) -14 deg
-7 deg
0 deg
7 deg
14 deg
21 deg

Figure 3.4 Close up of in-cylinder temperatures for different spark advance at 3000 rpm,
55:1 AFR and WOT.
In-cylinder NO
x
levels can be seen in Fig 3.5. In all of the lines, the NO
x
levels
start low, and increase very rapidly during the combustion process. The level then stays
constant until the intake valve is opened and fresh air starts entering the combustion
chamber, diluting the mixture.
Maximum NO
x
is produced at about 7 degrees of spark advance. This produces
the maximum combined pressure and temperature. Advancing the timing to 14 degrees
before TDC lowers the pressure during combustion, and thus reduces the NO
x
levels.
Retarding the timing lower the NO
x
levels considerably, as the peak pressure and
temperature is decreased. Having a highly retarded timing lowers the NO
x
levels
tremendously, but torque output is decreased.
41
0
500
1000
1500
2000
2500
3000
3500
-90 0 90 180 270 360 450 540
Crank Position (degrees)
I
n
-
C
y
l
i
n
d
e
r

N
O
x

c
o
n
c
e
n
t
r
a
t
i
o
n

(
p
p
m
)
-7 deg
-14 deg
0 deg
7 deg
14 deg
21 deg

Figure 3.5 NO
x
emission levels in the cylinder for different spark timing at 3000 rpm,
55:1 afr at WOT. Negative timing refers to spark occurring BTDC. Note that maximum
NO
x
levels are not attained at earliest timing, but at slightly advanced timing.
Setting the timing to produce maximum torque produces about half of the
maximum NO
x
, attained at the slightly advanced setting. Retarding the timing a little
more, from 7 degrees to 14 degrees retarded, decreases the NO
x
emissions by about a
third, with only about 2% decrease in torque. Retarding the timing by another 7 degrees
to 21 degrees retarded, decreases the NO
x
by another third, but decreases the torque by
about 6% from the maximum. Retarding the timing any more, decreases the torque
significantly. The timing setting can be set to produce the maximum power, or minimize
emission. If getting the minimum emissions at WOT is most important, a little power can
be sacrificed. If getting every last bit of power out of the engine is wanted, running less
retard will make the power and emissions go up. Running too much ignition advance
will decrease power, increase emissions, and most likely increase likelihood of pre-
ignition or detonation. This can lead to severe engine damage.
42
Volumetric Efficiency
The volumetric efficiency is an important measure of a four-stroke engine’s
ability to pump air. This is closely related to the torque output of the engine. It is
defined as the volume flow rate of air into the cylinder divided by the rate at which
volume is displaced by the piston.
N V
m
d i a
a
v
,
2
µ
q
&
=
where
a
m& is the mass flow rate of the air entering the cylinder,
i a,
µ is the air density,
d
V is
the displaced volume, and N is the engine speed.
The volumetric efficiency is generally measured at wide open throttle, as this
produces the maximum efficiencies. The efficiency can be lowered by closing the
throttle.

The Effect of Volumetric Efficiency on Torque
Since the volumetric efficiency is a measure of how much air enters the cylinder,
it also measures how much fuel enters the cylinder at a given equivalence ratio. The
amount of air-fuel mixture burned in the cylinder is directly related to the mean effective
pressure. The mean effective pressure is very closely related to torque output of the
engine.
Figure 3.6 shows the torque output and volumetric efficiency for the naturally
aspirated, gasoline engine. The relation between volumetric efficiency and torque can be
clearly seen, as the curves follow each other.

43
100
120
140
160
180
200
220
1000 2000 3000 4000 5000 6000
Engine Speed (rpm)
T
o
r
q
u
e

(
N

m
)

a
n
d
V
o
l
u
m
e
t
r
i
c

E
f
f
i
c
i
e
n
c
y
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Torque
Volumetric
Efficiency

Figure 3.6 Torque (left axis) and volumetric efficiency (right axis) versus engine speed
for a naturally aspirated gasoline model.

The Effects of Fuel Type, Equivalence Ratio and Supercharging
on Volumetric Efficiency
The volumetric efficiency is affected by the fuel being burned in the engine.
Liquid fuel takes up very little space in the intake ports and the combustion chamber.
Fuel vapor can take considerably more of this space, leaving less room for the air being
pumped into the cylinder. As a vapor fuel is burned richer, the fuel takes more space,
decreasing the volumetric efficiency.
Figure 3.7 shows volumetric efficiencies for several models varying the
equivalence ratio. The naturally aspirated gasoline model was modeled twice. The first
model was ran at a stoichiometric mixture (14.7:1 air-fuel-ratio), while the second one
was ran at a rich mixture (12.5:1 air-fuel-ratio). Both models used a 10% fuel vapor, and
90% liquid being injected from the fuel injectors. The two volumetric efficiencies are
almost identical, with less than one percent difference between each data point. Since the
two resulting lines are almost identical, an expanded view of them is shown in Figure 3.8.
44
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
1000 2000 3000 4000 5000 6000
Engine Speed (rpm)
V
o
l
u
m
e
t
r
i
c

E
f
f
i
c
i
e
n
c
y
S/C H2 eq 0.4
S/C H2 eq 0.62
S/C H2 eq 1.0
N/A Gasoline
eq 1.176
N/A Gasoline
eq 1.0
N/A H2 eq 0.62

Figure 3.7 Volumetric efficiencies for a supercharged (S/C) hydrogen fueled engine
varying air-fuel ratio, a naturally aspirated (N/A) gasoline engine for stoichiometric and
rich mixtures, and a naturally aspirated hydrogen engine.
The same, naturally aspirated engine was also modeled using hydrogen as the
fuel, with no other changes. The air-fuel-ratio used was 55:1, or 0.62 equivalence ratio.
This line follows the shape of the gasoline models, but at a much lower level. Just
because of the change in fuel, the volumetric efficiency dropped by approximately 20-
25%.
The supercharged models show the decrease in volumetric efficiency very
dramatically. The three efficiency curves are almost parallel. The leanest air-fuel-ratio is
the most efficient, while the richest is the least efficient.
There are two curves on the figure which show the effect of supercharging an
engine. This is for the 55:1 air-fuel-ratio for hydrogen. Although at first it might appear
that the supercharged model, with its bends and long pipes, is less efficient at low engine
speeds, it is not true. The 55:1 AFR supercharged model is the middle one of the
supercharged models. Even at low speeds the efficiency of the naturally aspirated 55:1
AFR model is lower, being very close to the curve of the stoichiometric, supercharged
model.
45
Comparing the curves of the naturally aspirated and the supercharged model, it
appears that the supercharged model is more sensitive to the engine speed. This is true,
but the reason for this is the increasing outlet pressure from the supercharger. The curves
are smoother than the naturally aspirated ones.
The volumetric efficiency curves for the naturally aspirated engines are much
more sensitive to tuning. Tuning is adjusting the intake and exhaust duct diameters and
lengths, as well as valve timing. As air is flowing through the ducts, there are pressure
waves traveling in them. The pressure waves occurring in the intake and exhaust ducts
are more favorable at the engine speeds where the volumetric efficiency is high. If the
pressure in the intake port is high when the intake valve opens, more air will flow to the
cylinder. On the other hand, if this pressure is low, less air will flow into the cylinder,
decreasing the volumetric efficiency.
The supercharged model could also be tuned for some engine speed; however, the
effects of the tuning are considerably smaller than those caused by the increasing outlet
pressure from the supercharger. Since the basic intake and exhaust manifolds were used
in all models, the same tuning effects should be seen on all models. The supercharged
model barely shows the increased volumetric efficiency near 2500 rpm, which is very
prominent in the naturally aspirated model. The drop-off in the efficiency at higher
engine speeds can be seen a little better in the supercharged model by decreasing slope of
the curve. It is possible that the supercharger itself is not as efficient at the higher speeds,
thus adding to this effect.
Changing the air-fuel-ratio for the naturally aspirated gasoline model does not
change the volumetric efficiency significantly. This can be seen in Figures 3.7 and 3.8.
In fact, the difference in volumetric efficiency between the stoichiometric and rich
mixtures is less than one percent. Surprisingly, the volumetric efficiency for engine
speeds less then 4500 rpm is higher for the rich mixture. As some of the fuel is vaporized
when it is injected, similar results for the hydrogen would be expected, but at a smaller
scale. This does happen for the high engine speed, when the volumetric efficiency for the
rich mixture dips below the one for the stoichiometric.
46
0.6
0.65
0.7
0.75
0.8
0.85
0.9
1000 2000 3000 4000 5000 6000
Engine Speed (rpm)
V
o
l
u
m
e
t
r
i
c

E
f
f
i
c
i
e
n
c
y
Stoichiometric
Rich

Figure 3.8 An expanded view of the volumetric efficiencies for the naturally aspirated
gasoline model run at a stoichiometric (14.7:1 air-fuel-ratio) and rich (12.5:1 air-fuel-
ratio). The same curves are shown in Figure 2.1.
The seemingly backwards results for the low engine speeds could possibly be an
error in the inputs for the simulation. The same burn duration and timing were used for
both models. It is possible that the rich mixture would need a different timing advance,
and would burn at a different rate. Changing these would most likely change the pressure
and temperature in the combustion chamber when the valves are open, which directly
affects the volumetric efficiency. It is also possible that the simulation result is correct,
and the engine burns the air more efficiently for the rich mixture.

Compression Ratio
Since compression ratio can affect the engine performance, varying it was
simulated. Dynamometer data were available for hydrogen engines from Ford. The
brake thermal efficiency was plotted as a function of equivalence ratios. Figure 3.9
shows the brake specific fuel consumption the naturally aspirated hydrogen engine
47
running at 3000 rpm at WOT. These are the same operating conditions used for the data
in Figure 1.1. It can be seen that the simulation results do match trends of the
dynamometer data.
The simulation results do not fall down as quickly as the dynamometer data as the
fuel mixture is richened. Most likely the reason for this is that the simulations do not
have a knock model. As a real engine is ran richer than about 0.75 equivalence ratio,
knock starts becoming a major problem, requiring highly retarded ignition timing, which
lowers the power output considerably.
0.26
0.27
0.28
0.29
0.3
0.31
0.4 0.5 0.6 0.7 0.8 0.9 1
Equivalence Ratio
B
r
a
k
e

T
h
e
r
m
a
l

E
f
f
i
c
i
e
n
c
y
15.3
14.5
17.5
12.55
20
10

Figure 3.9 Brake thermal efficiency as the equivalence ratio is changed for the naturally
aspirated hydrogen model for different compression ratios.
The dropping in the efficiency of the engine as the compression ratio is raised
could be as a result of a few things. One potential reason is the increased knock. Also,
since hydrogen burns so quickly, it is possible that there is no significant increases in the
work done by the burned mixture. As the compression ratio is increased, there is a
significant increase in the work that must be done on the mixture to compress it. It is
possible that these two do not grow equally. Another possibility is that increased blow-
by past the rings causes some of the fuel to be not converted into usable work. This also
will increase the pressure in the crankcase, which works against the work done by the
pistons.
48
Figure 3.10 shows the same date as shown in Figure 3.9. The difference is that in
Figure 3.10 the thermal efficiency is plotted as a function of compression ratio, for
different equivalence ratios. This makes it easier to see the ideal compression ratio for
engine, as the lines are very close to each other in the Figure 3.9. For moderate
equivalence ratios, such as 0.62, the peak in the efficiency occurs for a compression ratio
of about 14.5:1 to 15:1. For leaner equivalence ratios, like 0.48, this is compression ratio
is still along the peak values of efficiency.
0.27
0.28
0.29
0.3
0.31
10 12 14 16 18 20
Compression Ratio
B
r
a
k
e

T
h
e
r
m
a
l

E
f
f
i
c
i
e
n
c
y
0.96
0.75
0.62
0.48

Figure 3.10 Brake thermal efficiency as the compression ratio is changed for the naturally
aspirated hydrogen model for different equivalence ratios.
The same simulations were also done for the supercharged engine. The results
from it are not as clean as the ones for the naturally aspirated engine. Figure 3.11 shows
the simulation results as the equivalence ratio is changed. The curves are virtually the
same shape as the ones for the naturally aspirated engine, but shifted higher.
Figure 3.12 shows is again a better way of representing the results. A very
interesting point to note on that figure is the drop in the efficiency in the middle of two
peaks. The two peaks are virtually the same efficiencies. The engine would be just as
efficient running with approximately a 13:1 or a 17.5:1 compression ratio.
49
The reason for this odd behavior is that the supercharger efficiency changed as the
compression ratio is changed. A combination of the engine and supercharger efficiency
is shown in Figure 3.12. The lower compression ratio would be a better choice to use in
an engine. This reduces the tendency to knock and is generally easier on most engine
parts.
0.29
0.295
0.3
0.305
0.31
0.315
0.32
0.325
0.33
0.4 0.6 0.8 1
Equivalence Ratio
B
r
a
k
e

T
h
e
r
m
a
l

E
f
f
i
c
i
e
n
c
y
17.5
12.55
14.5
15.3
20
10

Figure 3.11 Brake thermal efficiency as the equivalence ratio is changed for different
compression ratios on the supercharged hydrogen engine running at 3000 rpm and WOT.

50
0.3
0.305
0.31
0.315
0.32
0.325
0.33
10 12 14 16 18 20
Compression Ratio
B
r
a
k
e

T
h
e
r
m
a
l

E
f
f
i
c
i
e
n
c
y
0.96
0.75
0.62
0.48

Figure 3.12 Brake thermal efficiency as the compression ratio is changed for the
supercharged hydrogen engine for different equivalence ratios at 3000 RPM and WOT.

Camshaft Timing
Camshaft specifications affect engine performance dramatically. There are a
number of variables that can be changed in the camshaft including lift, duration, lobe
center separation angle, timing and the lift profile. In an engine with a single camshaft
changing the timing is the only variable that can be changed without grinding a totally
new camshaft, provided that the camshaft has an adjustable pulley. In an engine with
separate intake and exhaust camshafts it is possible to change the lobe separation angle
since it is possible to adjust the timing of each camshaft separately. Factory pulleys are
usually not adjustable, as adjustment is not needed on a stock engine and the adjustable
pulleys need to be made from at least two parts, costing more to manufacture. The
aftermarket does make adjustable pulleys for most of the popular engines, including the
2.3L Ford used in this study.
51
Since the engine being studied has a single camshaft, only camshaft timing was
altered. Substantial changes in the volumetric efficiency can be attained by this, as can
be seen by Figure 3.13. Retarding the camshaft increases the volumetric efficiency at
higher engine speeds, while lowering it at low engine speeds. Advancing does the
opposite.
Figure 3.14 shows the change in the volumetric efficiency as the camshaft timing
is changed. The curves are the differences between the volumetric efficiencies in Figure
3.13 and the stock curve on that figure. There is no change from the stock curve, and
thus that lies along the x-axis. Obviously, the highest volumetric efficiency at all engine
speeds is desired. Since maximizing it at one speed hurts it at other engine speeds, the
curve with the maximum area under the most often used engine speeds is the desired
curve.
0.6
0.8
1
1.2
1.4
1.6
1000 2000 3000 4000 5000
Engine Speed (rpm)
V
o
l
u
m
e
t
r
i
c

E
f
f
i
c
i
e
n
c
y
10 deg retard
5 deg retard
Stock
5 deg Advance
10 deg advance

Figure 3.13 Volumetric efficiency versus engine speed as the camshaft timing is changed
at 55:1 AFR.

52
-0.1
-0.075
-0.05
-0.025
0
0.025
0.05
1000 2000 3000 4000 5000
Engine Speed (rpm)
C
h
a
n
g
e

i
n

V
o
l
u
m
e
t
r
i
c

E
f
f
i
c
i
e
n
c
y
10 deg retard
5 deg retard
Stock
5 deg Advance
10 deg advance

Figure 3.14. The change in volumetric efficiency from the values attained with the stock
timing at WOT at 55:1 AFR.
Normal, street driven engines typically operate under 3500 rpm and part throttle
most of the time. Retarding the camshaft hurts the volumetric efficiency at those speeds,
while helping it above that speed. It is at these speeds that the supercharger is increasing
the volumetric efficiency dramatically as compared to the naturally aspirated engine.
When a vehicle is cruising, the engine operates at a much reduced volumetric
efficiency than wide open throttle accelerations. Since part throttle operation is more
expressive of the engine operation, some simulations were done at a throttle opening of
0.9 inches, instead of the 1.8 inches that were used for the WOT.
NO Emissions were also looked at while changing the camshaft timing. The
sensor was placed in the exhaust manifold collector. Figure 3.15 shows the NO output at
WOT in ppm, and Figure 3.16 for the part throttle case. Since the changes are very
small, Figures 3.17 and 3.18 show the change in the NO output when compared to the
stock timing. It can be seem that camshaft timing does alter the NO output, but not
significantly. Thus the maximizing volumetric efficiency is the deciding factor in
choosing the camshaft timing.
53
0
5
10
15
20
25
30
35
40
45
1000 2000 3000 4000 5000
Engine Speed (rpm)
N
O

e
m
i
s
s
i
o
n
s

(
p
p
m
)
10 deg advance
5 deg advance
stock
5 deg retard
10 deg retard

Figure 3.15 NO emissions as the camshaft timing is changed. AFR 55:1, WOT.
0
5
10
15
20
25
30
1000 2000 3000 4000 5000
Engine Speed (rpm)
N
O

e
m
i
s
s
i
o
n
s

(
p
p
m
)
10 deg retard
5 deg retard
stock
5 deg advance
10 deg advance

Figure 3.16 NO emissions as the camshaft timing are changed for part throttle conditions
at 55:1 AFR.
54
-1
-0.5
0
0.5
1
1.5
1000 2000 3000 4000 5000
10 deg advance
5 deg advance
stock
5 deg retard
10 deg retard

Figure 3.17 Changes in NO emissions (in ppm) as the camshaft timing is changed for the
WOT case at 55:1 AFR.
-1
-0.5
0
0.5
1
1000 2000 3000 4000 5000
Engine Speed (rpm)
C
h
a
n
g
e

i
n

N
O

E
m
i
s
s
i
o
n
s

(
p
p
m
)
10 deg retard
5 deg retard
stock
5 deg advance
10 deg advance

Figure 3.18 Changes in the NO emissions as camshaft timing is changed for the part
throttle case at 55:1 AFR.

55
As part throttle fuel economy is an important part of engine design, it was also
looked at during the part throttle operation. The brake specific fuel consumption (BSFC)
is shown in Figure 3.19. Since the data lines are fairly close to each other, the change in
the BSFC was also graphed in Figure 3.20.
Figure 3.19 shows the BSFC for the WOT case, and 3.20 shows the changes in it.
The same basic results are shown for the WOT case as were for the part throttle case.
The best camshaft timing from a fuel economy point of view is 5 degrees retarded. This
is a different from getting the maximum power, which would suggest either stock or
slightly advanced camshaft timing.
0.09
0.1
0.11
0.12
0.13
0.14
0.15
1000 2000 3000 4000 5000
Engine Speed (rpm)
B
S
F
C

(
k
g
/
k
W

h
r
)
10 deg advanced
5 deg advanced
stock
10 deg retarded
5 deg retarded

Figure 3.19 BSFC as the camshaft timing is changed for the part throttle case at 55:1
AFR.
Figure 3.20 clearly shows that advancing the camshaft increases the fuel
consumption at part throttle. Retarding the camshaft five degrees will increase the fuel
economy, but retarding it ten degrees is already too much. At ten degrees of retard, the
fuel consumption starts coming back up to the stock levels.
As a side note, comparing Figures 3.19 and 3.21, it is possible to see the increased
pumping losses as the throttle is closed. When the throttle plate is completely open, it is
56
easy for the pistons to pull in the air into the cylinder. As it is closed, more power goes
into pumping the air into the cylinder, even though less air gets ingested. This increases
the amount of fuel that gets burned for a given power output, or BSFC.
-4
-2
0
2
4
6
8
1000 2000 3000 4000 5000
Engine Speed (rpm)
P
e
r
c
e
n
t

C
h
a
n
g
e

i
n

B
S
F
C
10 deg advanced
5 deg advanced
stock
10 deg retarded
5 deg retarded

Figure 3.20 Change in the BSFC as the camshaft timing is changed for the part throttle
case at 55:1 AFR.
57
0.08
0.09
0.1
0.11
0.12
0.13
1000 2000 3000 4000 5000
Engine Speed
B
S
F
C

(
k
g
/
k
W

h
r
)
10 deg advance
5 deg advance
stock
10 deg retard
5 deg retard

Figure 3.21 The BSFC as camshaft timing is changed for the WOT case at 55:1 AFR.
-2
-1
0
1
2
3
4
5
6
7
8
1000 2000 3000 4000 5000
Engine Speed (rpm)
P
e
r
c
e
n
t

C
h
a
n
g
e

i
n

B
S
F
C

10 deg advance
5 deg advance
stock
10 deg retard
5 deg retard

Figure 3.22 Percent change in the BSFC as the camshaft timing is changed for the WOT
case at 55:1 AFR.

58
Throttling a Hydrogen Engine
The power output of most spark ignition engines is cut back to the levels needed
by the vehicle operation by restricting the airflow into the engine. An engine running on
most fuels, especially liquid fuels, needs to operate very close to the stoichiometric air
fuel ratio. This is not the case with hydrogen since it can ignite at very lean equivalence
ratios.
By lowering the equivalence ratio, and keeping the airflow constant, less fuel goes
into the engine, thus decreasing the power output. This gives a second way to cut back
the power output of the engine. It is possible to get the needed power while lowering the
fuel consumption by opening the throttle blade more thus decreasing the pumping losses.
Constant power output curves are plotted in Figure 3.23, as the throttle opening and
equivalence ratio are changed, for 3000 rpm engine speed.

Figure 3.23 Brake power (kW) as the throttle opening and equivalence ratio are changed
for the supercharged hydrogen engine running at 3000 rpm.
The curve in the bottom, left hand corner is for no power output (all the power is
used to run the engine). Below this curve, the engine produces negative torque, or
requires power to run. Each of the curves above the first curve represents an increase of
5 kW, up to 60 kW for the short, almost horizontal curve in the top, right hand corner.
59
From the curves it is possible to see that there are an unlimited number of
different combinations of throttle openings and equivalence ratios to get a specific power
output. For large throttle openings, the curves are almost horizontal. This means that the
engine power is almost completely dependent on the equivalence ratio at the large throttle
openings. The equivalence is the larger restriction at these operating parameters.
For the smaller throttle openings, the curves start pointing upward. This means
that the equivalence ratio is not as large a factor in determining power output, and the
importance of the throttle opening increases. For medium throttle openings (around 0.75
inches), the importance of throttle opening and equivalence ratio are equally important.
A different fuel consumption is obtained for all different operating points for a
given power output. The brake specific fuel consumption (BSFC) is shown in Figure
3.24 as the throttle opening and equivalence ratio are changed. The lowest BSFC is at
WOT and approximately 0.8 equivalence ratio. The curve is very flat for the blue colors;
there is very little change in the BSFC in that area. The figure does not demonstrate that
the fuel consumption rises extremely quickly in the red area. In fact, the BSFC is infinite
for the bottom left corner, where the engine makes either negative or no torque.
The necessary power for a given vehicle speed changes with driving conditions.
If the transmission does not change gears, the torque output from the engine must change
with the driving conditions in order to maintain constant vehicle speed. Comparing both
Figures 3.23 and 3.24., it is possible to come up with ideal operating points for the
engine. These are points which consume the least amount of fuel for a given power
output. It turn out that the least fuel is consumed at WOT. Thus it might be possible to
operate a hydrogen engine without a conventional throttle, using equivalence ratio as the
only way to regulate torque output.
60

Figure 3.24 Brake specific fuel consumption (kg/kW Hr) as the throttle opening and
equivalence ratio are changed for the supercharger hydrogen engine running at 3000 rpm.
Since Rotrex recommended that the supercharger be mounted after the throttle,
this setup was also simulated. The BSFC for this case is shown in Figure 3.25. The
reason for mounting the supercharger after the throttle is that it is under much lower load
while the vehicle is cruising, thus lowering the fuel consumption. When the supercharger
is mounted before the throttle, and the throttle is open very little, the supercharger is still
making boost that the throttle is preventing from reaching the intake manifold.
Rotrex recommends at least using a re-circulation valve if the throttle is placed
after the supercharger. This valve is operated by manifold vacuum, which opens it to let
air from the high pressure side to the low pressure side of the supercharger. When the
supercharger is placed after the throttle, a recirculation valve is not as critical for
maximum efficiency the compressor spins in a vacuum. However, most original
equipment superchargers do have these, even though virtually all of them mount the
throttle before the supercharger.

61

Figure 3.25 The BSFC for the supercharged hydrogen engine with the throttle mounted
before the supercharger as the throttle and the equivalence ratio are changed at 3000 rpm.
From Figures 3.24 and 3.25 it appears that it would be better to mount the throttle
after the throttle. The simulation results are most likely not completely accurate for the
supercharger efficiency. Most likely the airflow or the pressure ratio that WAVE
calculates across the supercharger is smaller than in the actual engine, and it is better to
mount the supercharger after the throttle.
Emissions
Since hydrogen has no carbon atoms, the main pollutant from a hydrogen engine
is NO
x
. WAVE is unable to simulating any other NO
x
than NO, so it is the only one that
was studied and compared to available data. The emissions were considered when
throttling the engine and the NO emissions were compared to data published by Ford
Motor Company.
Figure 3.26 shows the NO emissions in the tailpipe in ppm, as the throttle opening
and equivalence ratio are changed. These simulation results are for the case where the
throttle is mounted after the supercharger. It can be seen that the NO emissions are
almost completely a function of the equivalence ratio. Thus, from the emissions
perspective it is also better to run the engine leaner, with the throttle opened further.
62
The Ford Motor company has studied the NO
x
emissions in a hydrogen engine
(Stockhausen). They found that the NO
x
emissions were very small (less than 10ppm) at
equivalence ratios less than 0.38, and less than 100ppm at equivalence ratios less than
0.5. Their study shows that concentration increases dramatically when the engine is run
richer.
The results from Ford are shown in Figure 3.27. The graph includes data from
their test engine ran at three different compression ratios (12.5:1, 14.5:1, and 15.3:1).
The points were taken at various engine loads and speeds. This graph also supports the
results shown in Figure 3.26, as the NOx emissions are strongly a function of equivalence
ratio.

Figure 3.26 NO emissions (ppm) in the tailpipe as the throttle opening and equivalence
ratio are changed for the supercharger hydrogen engine running at 3000 rpm. The bottom
line is 4 ppm, and increases by 4 ppm for each line above it.
Similar trends were found with the simulations. The NOx emissions for the
supercharged model are shown in Figure 3.28. There is very little NOx made below 0.5
equivalence ratio, but it as the mixture is richened the NOx increases. In order to
minimize NOx Emissions, the engine should not be ran richer than about 0.6 equivalence
ratio. Running a real engine at that equivalence ratio is possible, and as the air fuel
mixture is richened from this at higher loads, pre-ignition problems increase dramatically.
63

Figure 3.27 NOx emissions as a function of equivalence ratio for Ford test engines
running at various engine speeds, loads and compression ratios.
Nox Emissions v. Equivalence Ratio
0
2000
4000
6000
8000
10000
12000
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Equivalence Ratio
N
O
x

E
m
i
s
s
i
o
n
s

(
p
p
m
)

Figure 3.28 NO
x
emissions for the supercharged hydrogen engine at 3000 RPM and
WOT, as the equivalence ratio is changed.
64
Pressure Measurement in the Intake and Exhaust Systems
Pressure sensors were placed in the intake manifold plenum, exhaust manifold
collector, and just behind the catalytic converter on the supercharged hydrogen model.
The mean pressures are shown in Figure 3.29. This figure also shows the pressure in the
intake plenum for the naturally aspirated gasoline model.
For the gasoline model, the pressure in the intake plenum, at low engine speeds is
very close to the ambient pressure (1 Bar). It can be seen that the pressure decreases as
engine speed increases. This decrease is very small, but it does show that as the airflow
is increased, pressure losses through the open throttle plate and intake plumbing are
increased.
On the supercharged model, the pressure in the intake increases as the engine
speed increases. This is because of the centrifugal supercharger having an increased
outlet pressure with higher speeds. Significant increases in the exhaust pressures can also
be observed.
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
1000 2000 3000 4000 5000
Engine Speed (rpm)
A
b
s
o
l
u
t
e

P
r
e
s
s
u
r
e

(
B
a
r
)
S/C Intake
S/C Exhaust
S/C Catalytic Conv.
N/A Exhaust
N/A Catalytic Conv.
N/A Intake

Figure 3.29 Mean pressures in intake manifold plenum, exhaust manifold, and after the
catalytic converter brick, for the supercharged hydrogen model running at 55:1 AFR and
the stoichiometric, naturally aspirated, gasoline model. Ambient pressure is 1 Bar.

65
The mass flow through the exhaust system was also noted for the models. They
are shown in Figure 3.30. The mass flows were measured near the end of the tailpipe,
where only vapor passed through the meter. The mass flow rate through the engine
should be mostly a function of engine speed and volumetric efficiency.
0
0.05
0.1
0.15
0.2
1000 2000 3000 4000 5000
Engine Speed (rpm)
M
a
s
s

F
l
o
w

t
h
r
o
u
g
h

E
n
g
i
n
e

(
k
g
/
s
)
S/C Hydrogen
N/A Gasoline

Figure 3.30 Mass flow through the engine for the supercharged, hydrogen engine (55:1
AFR) and naturally aspirated, stoichiometric, gasoline engine.
The mass flow rates along with the volumetric efficiencies explain why the
pressure curves in Figure 3.29 level off for the gasoline model, but not for the hydrogen
model. For the naturally aspirated gasoline model, the volumetric efficiency starts falling
after about 4000 rpm. It is because of this that the mass flow rates level off for the higher
engine speeds. For the supercharged, hydrogen model, the mass flow rates just keep
climbing as the volumetric efficiency and engine speed increase.
The significant back pressure in the exhaust system for the hydrogen engine
means that the engine needs a bigger exhaust system. This was done by increasing the
exhaust pipe diameter from 1.875 inches to 2.5 inches, for a cross-sectional area increase
of almost 80%. The catalytic brick size was doubled by increasing the number of
passages from 1500 to 3000.
66
Increasing the exhaust size decreased the back pressure 10-20% above 3500 rpm.
This did increase the volumetric efficiency, but only by 1-3%. Under 3500 rpm, where
the engine operates most of the time, there were no significant changes in the back
pressure or volumetric efficiency. A down side to increasing the size of the exhaust
tubing is increased noise. Thus, the smaller exhaust would be a better choice for the
engine.

Conclusion and Future Research
This study was done in order to show general trends as variables are changed for a
hydrogen engine, in order to recommend changes in the engine. The effects of camshaft
timing, compression ratio, equivalence ratio were shown in the volumetric efficiency,
power, and emissions. Recommendations for improving the power, fuel efficiency, and
emissions for Texas Tech’s FutureTruck explorer are discussed below.
Probably the biggest, and most difficult, improvement would be to chance the
throttle plate to a fly-by-wire throttle. This would allow the throttle to be opened more
while cruising in order to lower the fuel consumption. Controlling this type of a throttle
would not be simple. These systems are relatively rare in stock vehicles, but are
becoming more common, much like the wideband oxygen sensors. A stock controller
would most likely not be programmable, and would not allow leaning out the equivalence
ratio as the throttle plate is opened. Thus, a new engine controller would need to be built
from scratch.
Leaving out the throttle plate all together might be possible with the use of
hydrogen. The power output could be controlled by regulating the equivalence ratio. It
is unclear if an engine would run at low speeds at very lean equivalence ratios.
Controlling this would not be difficult, and the current ECU could be used.
The compression ratio of 12.6:1 seems to give approximately the highest fuel
efficiency, and thus would not need to be changed. If the engine did not have a
supercharger or a turbo, a higher compression ratio should be used. The highest fuel
67
efficiency is given by approximately 14.5:1 to 15:1 compression ratio for a naturally
aspirated model, and approximately 12.5:1 to 13:1 for the supercharged model.
The camshaft advance of 8 degrees is a little bit too much. While this does
increase power below 2000 rpm, it decreases power above 2500 rpm. If the engine is
operated above 3500 rpm, the stock timing is more suitable. Since the engine would be
used at about 1500 to 2500 most of the time, camshaft advance should be approximately
5 degrees.
If maximum fuel economy is wanted, the camshaft should be advanced as little as
possible. Five degrees of camshaft retard provided the best fuel economy.
As the supercharger makes very little boost at low engine speeds, the ratio of the
pulleys could be changed. Currently the pulley sizes are set to give maximum rotational
speeds of the supercharger at engine speed of 5000 rpm. By decreasing the maximum
engine speed to 4500 rpm, more torque could be made at the low engine speeds.
Changing to a positive displacement supercharger would help the low boost levels
at low engine speeds. These superchargers in general are much more difficult to mount
to stock intake manifolds, but in a production vehicle this should not be an issue.
These simulations have shown that the specifications of the hydrogen engine used
in Texas Tech’s FutureTruck does not need any major changes in order to optimize its
operation. Most likely the recommended changes will not be made to this engine, as the
FutureTruck competition ended in 2004. However, these results can be use in the
FutureTruck replacement Challenge X.
Challenge X is going to use a hydrogen-ethanol engine, which will be have very
similar low engine load characteristics as a pure hydrogen engine. The engine will be ran
an either mostly or completely on hydrogen at light loads, in order to minimize
emissions. At higher loads and engine speeds, when hydrogen is unable to produce
power, the engine will use mostly ethanol. Controlling and tuning this engine will
present new challenges, but the experience gained from the FutureTruck will help
overcome these obstacles.
68
BIBLIOGRAPHY

Alcock, J.L., “Compilation of Existing Safety Data on Hydrogen and Comparable Fuels.”
http://www.eihp.org/public/documents/CompilationExistingSafetyData_on_H2_a
nd_ComparativeFuels_S..pdf Jan. 16, 2005.

Autery, Clay. turboford.org
http://www.turboford.org/faq/ta.shtml, Sept. 24, 2004.

Atherton, Larry. http://www.motionsoftware.com/simtech.htm, Jan. 14, 2005.

Cosworth Technology. “Alternative Fuels.” http://www.cosworth-
technology.co.uk/100_engineering/106_alt_fuels.htm, Jan. 20, 2005.

Dempsey, J. “Module 2: Hydrogen Use.” Energy technology Training Center College of
the Desert, 2001.

Ehsani, Mehrdad; Yimin Gao, Sebastien E. Gay, Ali Emadi, Modern Electric, Hybrid
Electric, and Fuel Cell Vehicles; Fundamentals, Theory, and Design. Boca
Raton: CRC Press, 2005.

Endres, Christopher P. Chevy LS1/LS6 Performance. HPBooks. New York, New York.
2003.

Gargano, Peter. “NTK L1H1 Sensor Information.”
http://techedge.com.au/vehicle/wbo2/wbntk.htm, Feb. 8, 2005.

Heffel, J.; Kabat, D.; Natkin, R.; Stockhausen, W.; Tang, X. Ford P2000 Hydrogen
Engine Dynamometer Development. SAE, 2002

Heywood, John B. Internal Combustion Engine Fundamentals. New York: McGraw-Hill.
1988.

Hunkins, Johnny. “The Lowly Oxygen Sensor, More power and clean living through
science – and Bosch.” Popular Hotrodding.
http://www.popularhotrodding.com/tech/0407phr_bosch/, Feb. 8, 2005.

Karner, Don; James Francfort. Arizone Public Service – Alternative Fuel (Hydrogen)
Pilot Plant Design Report. U.S. Department of Energy, FreedomCAR & Vehicle
Technologies Program, Advanced Vehicle Testing Activity. INEEL/EXT-03-
00976. Dec. 2003.

69
Micro-Craft. “Planar wide band O2/Lambda sensor.”
http://www.airfuelmeter.com/english/lsu4_en_sensor.htm, Feb. 8, 2005.

Motec. “MoTeC M4 Specifications.” http://www.motec.com/products/ecu/m4ecu.htm,
Feb 16, 2005.

Probst, Charles O. Corvette Fuel Injection & Electronic Engine Management 1982
Through 2001: How to Understand, Service and Modify. Cambridge, MA:
Bentley Publishers. 2001.

Ricardo Software, “WAVE Knowledge Center. A Comprehensive Guide to WAVE
Practices & Techniques.” 2004 Ricardo Software.

Rotrex. http://www.rotrex.com/, Nov. 29, 2004.

Sapru, K., et al. Hydrogen Internal Combustion Engine Two Wheeler with on-board
Metal Hydride Storage. Proceeding of the 2002 U.S. DOE Hydrogen Program
Review. NREL/CP-610-32405

Slocum, Allen. “Esslinger Aluminum Head.”
http://www.merkurencyclopedia.com/Motor/ESSLINGER_Head.html, Sept. 24,
2004

Stockhausen, William F. et al. Ford P2000 Hydrogen Engine Design and Vehicle
Development Program. SAE 2002-01-0240

TCR Automotive and Performance. “Basic EFI Theory.”
http://toyotaperformance.com/basic.htm, Feb. 8, 2005.

Turns, Stephen R., An Introduction to Combustion, Concepts and Applications. 2
nd

Edition. Boston: McGraw-Hill. 2000.

Wouk, Victor, “Hybrid Electric Vehicles” Scientific American. Oct. 1997.

70
PERMISSION TO COPY



In presenting this thesis in partial fulfillment of the requirements for a master’s
degree at Texas Tech University or Texas Tech University Health Sciences Center, I
agree that the Library and my major department shall make it freely available for research
purposes. Permission to copy this thesis for scholarly purposes may be granted by the
Director of the Library or my major professor. It is understood that any copying or
publication of this thesis for financial gain shall not be allowed without my further
written permission and that any user may be liable for copyright infringement.

Agree (Permission is granted.)
Jaakko Halmari April 17, 2005
________________________________________________ ________________
Student Signature Date



Disagree (Permission is not granted.)



_______________________________________________ _________________
Student Signature Date

ACKNOWLEDGEMENTS

I would like to thank Dr. Maxwell and Dr. Parten for offering me the opportunity to participate with the FutureTruck competition as well as other projects going on at the Advanced Vehicle Engineering Laboratory at Reese Center. I would also like to thank Dr. Ertas for serving in my thesis committee. The support from my family is greatly appreciated. Pitkään on väännetty ja käännetty mutta tulihan siitä lopulta jotain.

ii

TABLE OF CONTENTS

ABSTRACT .................................................................................................................v LIST OF TABLES .........................................................................................................vi LIST OF FIGURES.......................................................................................................vii CHAPTER I. INTRODUCTION........................................................................................................1 Production, Distribution and Storage of Hydrogen ...............................................1 FutureTruck .........................................................................................................3 Objective .............................................................................................................3 Gasoline baseline .................................................................................................4 Engine basics .......................................................................................................4 Controlling Fuel Flow and Spark Timing .............................................................6 Speed density .......................................................................................................6 Mass air flow .......................................................................................................7 Other sensors .......................................................................................................8 Actuators ...........................................................................................................10 Hydrogen properties...........................................................................................11 Differences Converting an Engine from Gasoline to Hydrogen ..........................13 Running engine on H2 .......................................................................................17 Hybrid system basics .........................................................................................18 Types of Hybrids ...............................................................................................19 Series hybrid-electric systems ............................................................................19 Parallel hybrid-electric systems ..........................................................................20 Pre-transmission parallel ....................................................................................20 Post-transmission parallel...................................................................................21 Road parallel......................................................................................................21 II. ENGINE SIMULATIONS ........................................................................................22 Ricardo WAVE background...............................................................................23

iii

..................................................................................................................................................................................................................................37 Volumetric Efficiency .....................................................61 Pressure Measurement in the Intake and Exhaust Systems............................................................................................................66 BIBLIOGRAPHY ...........................................................................................................................................................................................................................................................................................................................................................................................................68 iv .................................................................................33 Emission .........................................................32 Calculating the burn duration .........................................28 Simulating cylinder heads ................................ Equivalence Ratio and Supercharging on Volumetric Efficiency ..........................37 Spark Timing .......................................................................23 Description of how WAVE calculates air flow ........................................... SIMULATION RESULTS.....................46 Camshaft Timing .............................................................................................................................................................Results given.........................50 Throttling a Hydrogen Engine ............................................................................35 III...........27 Simulating the intercooler and the catalytic converter......................................................................................23 How models are built ................................................................43 Compression Ratio..................................................................42 The Effects of Fuel Type........................................28 Simulating the camshaft ......................................26 Throttle ...............58 Emissions.......................................................64 Conclusion and Future Research ......29 Supercharger modeling ..........................................................42 The Effect of Volumetric Efficiency on Torque.....................................................................24 Discretization lengths............................................30 Combustion model ...........................................

The simulation results studied included the volumetric efficiency. were the equivalence ratio. a naturally aspirated gasoline engine was simulated. throttle opening. Among the variables which were changed. followed by the supercharged hydrogen engine.ABSTRACT Ricardo WAVE was used to simulate the hydrogen internal combustion engine used in Texas Tech University’s FutureTruck Ford Explorer. v . camshaft timing. fuel consumption. compression ratio. while minimizing the emissions and fuel consumption. Initially. and exhaust size. as well as NO emissions. Several results were compared to the gasoline baseline model. The objective of these simulations was to maximize power of the hydrogen engine.

...................29 2.....2 Partial Summary of Camshaft Specifications..13 2.......................30 vi ...............................................LIST OF TABLES 1...........................................1 Physical properties of hydrogen and gasoline ...................................1 Cylinder Flow Data from Two Sources ...........

.......3 View of the input panel for each duct through which air travels ......................................... ..............................37 3..........….............................8 An expanded view of the volumetric efficiencies for the naturally aspirated gasoline model run rich and stoichiometric…………………………………………46 3...............39 3. …….43 3......................50 3................... equivalence ratio for different compression ratios for the supercharged hydrogen model……………………………………..53 vii .... compression ratio for different equivalence ratios for the supercharged hydrogen model……………………………………..3 In-cylinder temperatures for three different spark advances..…..40 3.....................51 3......31 2............2 Brake thermal efficiency versus equivalence as the compression ratio is varied.31 2.6 Traction gears in the supercharger...............................4 Supercharger Efficiency Curves as Given by Rotrex ....2 Cylinder pressures v....................5 Compressor map as entered into WAVE ...7 Volumetric efficiencies versus engine speed for various engine simulations……...........................48 3............25 2.................1 Brake torque v................... compression ratio for different equivalence ratios for the naturally aspirated hydrogen model………………………………………...............14 The change in volumetric efficiency..........49 3..................................15 NO emissions as the camshaft timing is changed..............13 Volumetric efficiency v.....................36 2..........27 2........11 Brake thermal efficiency v...........................................1 Four-Stroke Operating Cycle .......................10 Brake thermal efficiency v.... crank angle for different ignition timing .........52 3........ spark timing for supercharged hydrogen engine ..............…47 3......1 Ricardo WAVE model of the supercharged hydrogen engine...... equivalence ratio for different compression ratios for the naturally aspirated hydrogen model………………………………..............................12 Brake thermal efficiency v...4 Close up of in-cylinder temperatures for different spark advances.......................36 3........26 2...6 Torque and volumetric efficiency versus engine speed ...5 1..…44 3..............9 Brake thermal efficiency v..7 Oil being used to transmit power in the supercharger .................................. ................38 3...............2 A visual representation of a junction ....................41 3...................16 2..5 NOx emission levels in the cylinder for different spark timing.....LIST OF FIGURES 1................... engine speed as the camshaft timing is changed........................

..18 Changes in the NO emissions as camshaft timing is changed at part throttle.27 NOx emissions v........21 BSFC as camshaft timing is changed at WOT ............63 3............62 3.56 3................24 Brake specific fuel consumption as the throttle opening and equivalence ratio are changed for the supercharged hydrogen engine……………………......…………60 3... ..19 BSFC as the camshaft timing is changed at part throttle .....30 Mass flow rates through supercharged hydrogen engine and the gasoline model….......55 3.......54 3...22 Percent change in the BSFC as the camshaft timing is changed at WOT . equivalence ratio for the supercharged hydrogen model running at 3000 rpm and WOT……………………………………………………........17 Changes in NO emissions as the camshaft timing is changed at WOT.58 3. and compression ratios……………………………………….........29 Mean pressures in the intake and exhaust for naturally aspirated and hydrogen engine models……………………………………………………………….........................16 NO emissions as the camshaft timing are changed for part throttle conditions.……........53 3.........23 Brake power as the throttle opening and equivalence ratio are changed for the supercharged hydrogen engine………………………………………………....64 3...61 3.....20 Change in the BSFC as the camshaft timing is changed at part throttle.3......25 Brake specific fuel consumption as the throttle opening and equivalence ratio are changed for the supercharged hydrogen engine with throttle mounted before the supercharger…………………………………………………………….......54 3..26 Tailpipe NO emissions as the the throttle opening and equivalence ratio are changed for the supercharged hydrogen engine…………………………………….......57 3. equivalence ratio for Ford test engines running at various engine speeds....63 3. loads............…….65 viii .28 NOx emissions v.. .....57 3.

Before the completely emission free fuel cell vehicles are perfected. Promising technologies for the shorter term are hybrid electric vehicles and alternative fuels. ethanol. Distribution and Storage of Hydrogen Although hydrogen is the most plentiful gas in the universe and the simplest element. most common ones being hydrocarbons and water. Production. hydrogen engines could be used in vehicles as a stepping stone to the hydrogen economy. Thus all hydrogen needs to be separated from some other compound. The carmakers in the developed countries have been forced to reduce the emissions and fuel consumption each year. Hydrogen. however. propane. Each one of these vehicles produces harmful emissions and uses nonrenewable resources. Reforming hydrocarbons is the most economical and common way of producing hydrogen currently. more people are using vehicles for both work as well as pleasure. but they are still in the developmental stages. and hydrogen can lower the emissions of vehicles when compared to gasoline engines. They allow the engine to be sized smaller.CHAPTER I INTRODUCTION As the population of the world grows. is the only one that can potentially produce no hydrocarbon. Alternative fuels such as methanol. especially hydrogen. natural gas. The hydrogen engine vehicles could be used for building the hydrogen infrastructure and solving the on-board storage problems. Fuel cells are viewed as a potential long-term solution to the emission problem. Hybrid electric vehicles can substantially lower fuel consumption and emissions when compared to conventional vehicles. or carbon dioxide emissions (Cosworth). no matter what fuel they use. biodiesel. and make up the lost power with an electric motor. This however produces CO2 as a by-product (Karner). it does not occur naturally. Electrolysis of water is a more 1 . and inherently will produce carbon dioxide. All other fuels contain hydrocarbons. carbon monoxide.

requiring venting to the atmosphere. as much of the trunk space is taken up by the fuel tanks. Liquid hydrogen tanks have been made and tested. storing a vehicle with a low fuel level for a week would most likely result in an empty tank. more energy is needed in order to compress it to very high pressure for vapor storage in vehicles. These tanks store the hydrogen at a very cold temperature. Even though the hydrogen is compressed to such high pressure. Currently. This is not a big problem for vehicles which get driven daily. This is still a problem with most vehicles using vapor hydrogen storage. on-board storage of hydrogen in vehicles is done mostly with highpressure tanks. Some of the liquid hydrogen would always be evaporating. This is the maximum pressure of the Quantum Technologies tanks used in Texas Tech University’s FutureTruck Explorer. In fact.000 psig available currently. This means that the tanks have to have a large volume in order to have enough fuel stored onboard to have a reasonable range for the vehicle. BMW has researched the use of liquid hydrogen tanks. and do not suffer as much from this problem. at a low pressure 2 .environmentally friendly approach. but it also takes considerable amounts of energy to cool the hydrogen. but occasionally driven vehicles would need to be filled every time they are driven. Special vents would need to be installed in garage areas in order to minimize the potential fire hazard from this. but requires large amounts of electricity for the process to occur. With the use of cryogenic tanks. These store the hydrogen at ambient temperature. The maximum pressure used in the hydrogen tanks has gone up in the past few years from 3600 psig to 10. the amount of hydrogen stored in the tanks is relatively little when compared to liquid fuels. as the tanks are not perfect insulators. There also is a potential safety hazard that comes from the vented hydrogen collecting up near the roof of a garage. Most of the refueling facilities in use can compress hydrogen to 5000 psig. especially on the roof. increasing costs. the long term storage can be a problem. Buses have more room. Metal hydride storage might fulfill the shortcomings of the other methods of hydrogen storage. Once the hydrogen is produced.

The objective of these simulations was to maximize the power obtained from the engine while minimizing harmful emissions. plus be able to go off-road. This engine was chosen for its strength and parts availability. Hydrogen was chosen as a fuel in order to stay on the cutting edge of technology. has better fuel economy and lower emissions while maintaining the original performance. Cost is an issue currently preventing the use of these metal hydrides at a large scale. These are a promising way to store hydrogen safely at a low pressure. The metal hydride can absorb much more hydrogen than the high pressure storage systems for a given volume.S. the torque output should be reasonably flat. while lowering the emissions and improving the fuel economy by 25%. Objective Computer simulations were done for the engine used in Texas Tech University’s FutureTruck. and merge into traffic. The goal of the competition is to convert a stock 2002 Ford Explorer to a hybrid electric vehicle that uses alternative fuels. Ford Motor Company and Argonne National Lab sponsored the Future Truck competition from 2002 through 2004. Fifteen universities have participated in the Future Truck competition. FutureTruck The U. pull a trailer. The power can come on suddenly. forcing the 3 . The vehicles must be able to operate just like a stock vehicle. with a post-transmission parallel hybrid system. Engines with a very peaky torque curve are difficult to drive. Because the engine designed would go into a consumer vehicle.(Sapru).3L 4 cylinder from a 1986 Ford Mustang SVO. The approach chosen by Texas Tech University was to use a hydrogen powered internal combustion engine. The engine chosen was a 2. This objective also needed to be achieved without the expense of drivability. Department of Energy.

to almost all cars. and complicated due to the addition of the supercharger system. The performance of the hydrogen engine is much less familiar. the intake valve starts to open.driver to let off the gas quickly.The fuel is mixed into the air stream before it enters the combustion chamber. All duct dimensions were measured from the actual parts that were used in the engine. Engine basics The Otto-cycle. Most of the variables needed to run the model were taken from similar examples provided by Ricardo with WAVE. a cast iron exhaust manifold. The piston pulls the air 4 . cylinder liner temperatures. such as head temperatures. This also means having to shift gears more often than an engine with a smoother torque curve. Also burn duration and timing data were taken from examples. Numerous improvements have been made to these engines bringing them from simple single cylinder engines with a single carburetor to multi-cylinder multivalve engines with sophisticated electronic fuel injection systems. The performance of a gasoline engine is better known and therefore gives a way to make sure that the model is producing reasonable results. The variables included in-cylinder temperatures. The following is a description of the four cycles of a typical spark ignition engine. and a full exhaust system. a naturally aspirated. intake manifold. These engines are all around us in the modern world. 1. At the beginning of the intake stroke. Intake Stroke . unmodified. cast aluminum. or four-cycle piston engine has been developed for about one hundred years. and piston top temperatures. Gasoline baseline Initially. The gasoline engine included a throttle body. engine was simulated. from lawn mowers. the stock. letting fresh air-fuel mixture into the cylinder.

Exhaust Stroke . and begins to move back up.1. and the compression of the air-fuel charge is begun.1 Four-Stroke Operating Cycle (Heywood) 5 . work is done on the piston. The cylinder is then essentially sealed. compression ratio is about 8:1 to11:1. and the air-fuel charge begins to burn. A visual description of the four strokes can be seen in Figure 1. The air-fuel mixture being at a pressure higher than that in the cylinder then moves into the cylinder. expanding in the process. In a typical modern gasoline engine. 2. 4.Once the piston has moved all the way down. the intake valve is closed. 3. a spark plug is fired. Expansion or Power Stroke . thus expanding the volume in the cylinder. the exhaust valve is opened. Figure 1. This process is repeated dozens of times each second for high engine speeds.Near the bottom of the piston’s travel. This provides a pathway for the burnt gasses to be expelled out of the cylinder.into the cylinder by moving away from the valve. The piston pushes these gases out of the cylinder as it continues to move up.When the piston nears the top of its travel. Compression Stroke . As the charge continues to burn and the piston moves down.

A three dimensional 6 . Speed density Speed density systems use engine speed and air density. to determine the amount of air entering an engine. All systems have a variety of sensors. Modern fuel injection systems fall into two basic types: speed density and mass air flow. or pressure. The ECU is a computer into which the fuel and spark requirements are programmed. they still share the problems faced with carburetors. and the ignition with a distributor. although all new cars sold in the United States now primarily use the mass air flow system to control fuel flow. The speed density system is generally used to determine the spark requirement of the engine (Probst). Both have been used by the industry. This is an indirect way of measuring airflow based on testing done at the factory. with the speed density as a backup system.Controlling Fuel Flow and Spark Timing A spark ignition engine needs to have a way to control the fuel flow and spark timing. Both of these used mechanical systems to change the fuel and spark needs of the engine over a large operating range. and by the 1980’s were very sophisticated. based on the primary sensors used for determining air flow. The main problem with these systems is that they are not able to precisely control the fuel flow (TCR). Computer control and fuel injection met this need. The airflow is then programmed into the ECU as a function of the engine speed and pressure in the intake manifold. They still could not meet the ever stricter emissions and fuel consumption standards. and a more precise control was needed to replace the carburetor and mechanical distributor. Traditionally controlling the fuel flow has been controlled by a carburetor. These systems have gone through a lot of improvements. sending information about the engine operating conditions to an electronic control unit (ECU). Although mechanical and analog fuel injection systems have been developed and used in production vehicles. This discussion will only concentrate on digital fuel injection.

which is kept at a constant temperature. considerably hotter than ambient conditions. rather than just approximating it. With the use of this type of system it is difficult to know if the engine airflow has changed over time. with engine speed and manifold pressure on the x. sensor. Virtually all aftermarket fuel injection ECU’s use a speed density system for determining air flow. and amount of fuel flow or spark advance on the z-axis. or MAP.and yaxis. The diaphragm moves a potentiometer. including the Motec ECU used in Texas Tech University’s FutureTruck Explorer. Most common sensors in use now have a heated wire. Two major advantages of this system are the ease of aftermarket installation and the lack of restriction created by a mass airflow sensor. Although oxygen sensors can help this problem at light engine loads. these systems are only used as a backup for determining fuel flow on newer cars. Mass air flow The mass air flow sensor (MAF). A two dimensional map is developed with fuel flow versus airflow on the axes. which sends a voltage signal to the ECU. The power required to keep the wire hot is directly proportional to the air flowing through the sensor. These sensors are 7 . The primary sensor on this type of system is the manifold absolute pressure. The air cools the wire. This sensor has a diaphragm with a fixed pressure on one side. under heavy loads the ECU is unable to adjust for changes in the engine. determines the amount of air flowing into the engine directly. performing minor tuning as the engine wears is not a major problem. Common reason for this is because engine parts being changed or modified by the owner or by regular wear and tear on the engine.fuel flow map is developed. over which air flows. The other side is connected to a vacuum line on the intake manifold. MAF sensors are used in newer cars as they can very accurately determine the actual airflow going to the engine. As a result of this. Installers of the aftermarket systems are aware of the setbacks of the systems. but since major tuning efforts are required in any case initially.

This usually is an accurate way of determining average speed. placed next to multi-toothed wheels. A lot of the time these sensors are halleffects type. Throttle position sensor measures the opening of the throttle butterfly. but it is not accurate enough to tell if an engine has misfired. Older engines usually use just a signal from the distributor in order to determine engine speed. This is because the fuel does not vaporize easily when the engine is cold. but it is necessary for spark timing also. They generate a signal when a tooth is near the sensor. leading to lean engine operation. More gasoline needs to be injected into the engine during these operating conditions in order to prevent rough engine operation. The main reason for using these is for determining when the engine in idling as well as determining the acceleration enrichment. OBD-II. There are some disadvantages to using MAF sensors. Dirt build up can act like an insulator on the sensor wire. Temperature sensors are used to modify the determined fuel requirements from the MAP or the MAF sensor based on engine coolant temperature. the air is accelerated quicker than the gasoline.able to compensate for any changes in the airflow as an engine wears. when the throttle plate is opened quickly. The crankshaft and camshaft positions are also an important piece of information for the ECU. or on8 . These sensors act as a restriction as all of the incoming air has to travel through the sensor. Other sensors Cold operation of a gasoline engine requires rich operation conditions. Since gasoline is much denser than air. Air temperature modifications are also needed as colder air is denser than hot air. Aftermarket plumbing of these can also cause difficulties. They are expensive to make and are very delicate. or if modifications are made. The information from the crankshaft and camshaft position sensors can also be used by the new OBD-II ECU’s to determine if the engine has misfired. For most basic systems this information is used to determine the engine speed. giving false reading.

176. lean. The oxygen sensor determines if the amount of fuel injected is what the engine actually needs. in order to make diagnosing engine problems easier. These sensors measure the amount of oxygen in the exhaust. or stoichiometric. and sets the fuel requirements based on tables.board-diagnostics-2. was introduced in all cars sold in the United States after 1996. Accurate crankshaft position information is very important for many of the OBD-II functions. They can only tell that the engine is running rich. Narrow band oxygen sensors can very accurately determine stoichiometric operating conditions. A major disadvantage to using conventional narrow band oxygen sensors is that they are unable to determine how lean or rich an engine is running. these are able to accurately determine the exact equivalence ratio over a wide equivalence ratio range. Their light off time is usually much quicker than that of a convention narrow band oxygen sensor. These systems also need to be able to make sure that all of the emission components and systems are properly working. In order to prevent engine damage at high engine loads. However. such as wear. 9 . Torque is also maximized by running slightly rich. An oxygen sensor allows the ECU to go into closed loop operation. It is unable to see the outcome. This can help to compensate for changes in the engine. This sensor input then acts as the ultimate over-ride to the other sensor inputs. These are much more complicated sensors and require a separate controller. This means that the ECU first determines how much fuel to inject based on the sensors and tables mentioned above. Operating close to the stoichiometric air fuel ratio minimizes the combined emissions in a gasoline engine. Narrow band oxygen sensors are unable to determine air fuel rations in this region. During open loop operation. Wide band oxygen sensors have become more popular in the recent years. a gasoline engine needs to be run rich. equivalence ratio being about 1. Oxygen sensors are a very important sensor in controlling emissions and lowering fuel consumption. the ECU looks at the sensor inputs.

the equivalence ratio can be determined. There is a limit to how high the pressure can be raised. Obviously gaseous fuels do not need to vaporize. except for a very short transient conditions. Because of this. They also require a controller. Since wide band sensors are much more complicated. Saturn. the injectors only have a on and an off positions. injectors made for a liquid fuel will not work well. Most likely in a few years. and thus their use in vehicles has not been very widespread. or pulse width of the injectors the amount of fuel flow can be changed. Because of these differences. Actuators The fuel injectors have electric solenoids that move a mechanical valve. if the pressure differential is the same. A too low pressure will also not allow an injector to properly atomize fuel. if at all. they cost several times the price of a narrow band sensor. A larger orifice is also needed to flow a given mass of vapor than a liquid. Porsche. depending on the direction of current applied to it (Gargano). almost all new cars will have a wideband oxygen sensor. Volkswagen and Volvo (Micro-Craft). The pump cell is able to consume oxygen or fuel. This process is adjusted many times each second. with gaseous fuels. The current in the pump cell is adjusted until the narrow band reads stoichiometric conditions. since the solenoid can not open the injector if the inlet pressure is too high. Injectors 10 . at a pressure few times that of atmospheric conditions.The wideband oxygen sensors have a narrow band sensor and a pump cell. Their use is increasing rapidly there are many carmakers using these sensors. The inlet side of the fuel injectors is supplied fuel from a fuel rail. and its direction. Fuel flow can be changed by changing the pressure in the fuel rail. Cadillac. so the fuel is able to travel with the air flow. Subaru. Fuel injectors need to atomize liquid fuels. including Audi. By controlling the duty cycle. Honda. The flow of fuel through an injector is directly proportional to the pressure differential across it. Based on the amount of current. as well as vaporize better due to the increased surface area.

This rule of thumb works as long as the seals and other materials in the injectors are compatible with the fuel. Sequential injector timing has virtually no advantage at high engine speeds and loads. There are two basic ways of operating injectors in a multiple injector engine. the injectors are open most of the time. Boiling Point – The extremely low boiling point of hydrogen means that it will always be a vapor once it enters the engine. and as a result. gaseous injectors will work with most gases. Gasoline on the other hand is injected as a liquid. The most basic one is batch firing. This makes the electronic hardware and firmware in the ECU much more complicated. although onboard storage of it can be in liquid form. The ones which are important to the operation of an internal combustion engine will be listed and discussed shortly below. Hydrogen properties Hydrogen as a fuel has some properties which vary greatly from gasoline or other fuels. Sequential firing operates each injector separately.made for liquid fuels can usually be made to work with almost any liquid fuel. Sequential injection does improve the operation of the engine at low speeds and loads. Idle quality and emissions are improved.1 has most of the important properties listed for hydrogen and gasoline. This does not require the ECU to know the position of the crankshaft. It also requires crankshaft and camshaft position signals. Batch firing operates all of the injectors simultaneously. The time when the injector is not open makes very little difference when it is open 60-90% of the time. when the injectors are closed most of the time. and makes the firmware in the ECU much simpler. Table 1. usually twice for each engine revolution. Also. The end of the injection is made to coincide with the time the intake valve is about to close. which partially 11 . virtually all the new vehicles now use sequential injection. This requires a separate injector driver for each injector in the ECU. 1. At these operating conditions. while a more sophisticated one is sequential.

resulting in a back flash. Limits of Flammability – Hydrogen has a very large difference between the upper and lower limits of flammability. 6. Octane Rating – The high octane rating of hydrogen allows the possibility of increased compression ratio. Gasoline and most other fuels will not burn if the mixture is not near stoichiometric. This requires either very little spark advance or even retard timing.will evaporate in the engine before combustion. even when compared to other vapor fuels. such as natural gas and propane. 2. Vapor fuels take up space in the intake manifold as well as the combustion chamber. This allows hydrogen to burn over a much wider range of mixtures. lowering the volumetric efficiency. which increases thermal efficiency. the flame front can escape the combustion chamber through the intake valve more easily. This means that the combustion process takes up only a few crankshaft degrees. Since this is smaller than that for most other fuels. or even back flash. Burning Velocity – Hydrogen burns extremely quickly when compared to other fuels. Quenching Gap – This is the largest gap that still prevents flame front propagation (Alcock). 8. 3. the low ignition energy of hydrogen makes it more likely to preignite. from extreme lean to rich. and the pressure changes are almost instantaneous. but also makes hydrogen more dangerous as almost any mixtures of hydrogen and air will burn. 5. This property allows a hydrogen engine to run extremely lean. It might also be possible to get the flame front to go past the piston rings. Density – The low density of hydrogen means that it will take up more space in the intake manifold and the combustion chamber. 4. Minimum Ignition Energy – The low ignition energy of hydrogen can be very problematic in an engine as it can ignite from even the smallest hot spots in the combustion chamber. Auto-ignition Temperature – Although this temperature is higher for hydrogen than gasoline. 7. This property also means that any 12 .

(%) 4-75 1-7.24 Auto-ignition Temperature (K) 858 501-744 Maximum Burning Velocity in STP air 278 37-43 Flame Temperature in air (K) 2318 2470 There are major differences in the properties of hydrogen and gasoline. When this occurs during the middle of the compression cycle. Table 1. Gasoline on the other hand is harder to ignite.40 Density Ratio.5-27 Density of Liquid at STP (kg/L) 0. Hydrogen is a fuel which ignites very easily at a wide range of concentrations and burns very quickly.64 2 Limits of Flammability in air Vol.02 0. severe engine damage can result.0 Normal Boiling Point (K) 20. (%) 18. STP 845 ~ 150 Octane Rating 130+ 86-110 Thermal Diffusivity in STP air (cm2/s) 0. Pre-ignition occurs when the air-fuel charge ignites before the spark plug is fired. burns relatively slowly and requires a near stoichiometric mixture.3 310-478 Critical Pressure (atm) 12.70 3 Density of Gas at STP (kg/m ) 0.1-3. Differences Converting an Engine from Gasoline to Hydrogen Biggest problem with running an engine on hydrogen is pre-ignition and backflash.0708 ~ 0. and will puddle on the ground for a long time if spilled.6 Limits of Detonation in air Vol.07 ~ 4.pre-ignition that happens is much more violent.1 Physical Properties of Hydrogen and Gasoline (Dempsey.8 24. Turns) Property Hydrogen Gasoline Specific Gravity at STP relative to air 0.61 ~0.05 Diffusion Velocity in STP air (cm/s) ~2 ~ 0.34 Quenching Gap in STP air (mm) 0. Gasoline is also much denser.3-59 1.3 Minimum Energy for Ignition in air (mJ) 0. This 13 .838 ~ 4. when hydrogen will disperse in the air very quickly. and can damage the engine faster than that occurring with gasoline. Back-flash is pre-ignition that happens when the intake valve is still open.

Injecting the fuel when the intake valve is open helps cool any hot spots by not introducing hydrogen into the combustion chamber immediately after the intake valve opens. Hot exhaust gasses could also start pre-ignition. letting unused fuel back to the fuel tank. The most important and obvious one is changing the fuel delivery and control systems. They are much louder. Most pressure regulators will increase the outlet pressure as the inlet pressure decreases a great deal. However. but less destructive to the engine. Usually a bypass regulator is installed after the fuel injectors. many to avoid pre-ignition. In fuel injected gasoline engines. two regulators are needed in order to avoid variations in fuel pressure.explosion will travel through the intake manifold to the air intake. plastic. the fuel rails for liquid fuels are very small in diameter (roughly a centimeter). The first regulator brings the pressure down substantially. and steel with gasoline. This makes a fuel pump unnecessary. Often. allowing the same intake manifold to be used. They include hot spots from carbon deposits or sharp edges in the combustion chamber. several changes should be made. Stainless steel tubing is usually used with hydrogen. The second one is able to keep the pressure constant. A way around the pre-ignition problem is running the engine leaner. A few possible sources of pre-ignition have been suggested. It is possible to use aluminum. Fuel lines need to be changed. A large diameter fuel rail is needed in order to dampen pressure pulses and to serve as a reservoir 14 . High pressure tanks generally store hydrogen onboard of vehicles. Even the fuel rail should fit the hydrogen injector. but this decreases power. When converting a gasoline engine to run on hydrogen. which is the case when high pressure tanks are used. Most fuel injectors are roughly the same size externally. however regulators are needed to bring the fuel pressure to the levels required by the injectors. even if there is some variation in the pressure output of the first regulator. as the high pressure is not contained in the cylinder. The hydrogen would most likely leak out of these tubes and their junctions over time. the liquid fuel is pumped from a fuel tank at a constant volume to the fuel injectors by a positive displacement pump.

A Motec M4 ECU was chosen for Texas Tech’s FutureTruck. the use of aluminum has increased dramatically due to the weight savings and increased heat conduction. This is useful when tuning the engine as the ECU can be set to always log data. it can reduce the likelihood of pre-ignition.. being able to control 4 injectors sequentially. All testing was done at 3000 rpm and WOT. The data can then be used to change the fuel and timing maps when the engine is not running. they are not programmed for use with hydrogen. The material used in most older cylinder heads and engine block is cast iron. Although the stock ECU’s can operate hydrogen injectors. The high performance aftermarket does come into rescue. Several manufacturers make engine controllers which can be used with hydrogen. A higher compression ratio increases engine efficiency by increasing cylinder pressure during the power stroke. with user programmable ECU’s. higher compression ratio can be used. In a study done by Heffel et al. This is a very sophisticated aftermarket controller. control multiple coil packs. In the past twenty years.5:1 15 . This is a very desirable property when using hydrogen as a fuel in an engine. This size was recommended by hydrogen engine engineers at Ford Motor Company. This can save time and fuel while tuning.for the injectors. The fuel rail used in Texas Tech’s FutureTruck Explorer has an inside diameter of one inch. Because hydrogen has much higher octane rating than gasoline. the thermal efficiency was compared while changing compression ratio. The best brake thermal efficiency was attained at 14. Being able to control injectors sequentially is a feature that roughly half of the aftermarket engine controllers have. The M4 also has 512kb internal data logger. Engine controllers which are programmed for use with hydrogen are not available commercially. and use a large variety of sensors including a wideband oxygen sensor. which can record a large amount of sensor data (Motec). Since aluminum conducts heat much better. A production hydrogen engine could use an ECU made for gasoline engine with just a change of programming.

for a hydrogen engine running at 3000 rpm at WOT. At 12. higher compression ratio should give a higher efficiency (Heywood). it is possible that there is very little increase in the work done on the piston during the expansion stroke as the compression ratio is increased. which lowers the power output of the engine. Figure 1..2. Changing the compression ratio can be done by decreasing the combustion chamber volume or increasing the engine displacement. Data is from dynamometer testing done by Heffel et al. Another possibility is that the increased cylinder temperatures before ignition promote pre-ignition. Since hydrogen burns very fast. The efficiencies can be seen in Figure 1.2 Brake thermal efficiency versus equivalence ratio as the compression ratio is varied. the efficiency dropped. Theoretically.3:1.5:1 and 15. The drop in the efficiency of the test engine as the compression ratio was increased to 15. Increasing the displacement is 16 . requiring retarded ignition timing.3:1 could be a result of increased work that needs to be done to compress the air-fuel mixture before the combustion.compression ratio.

The addition of the crankshaft sensor required removing the timing indicator. but this is only marginal. This was caused because the engine chosen did not originally have crankshaft and camshaft sensors. since cylinder walls need to remain thick enough to withstand combustion pressures. This can be accomplished prior to cylinder head assembly by smoothing out all internal surfaces. 17 . and thus lead do preignition. Polishing the combustion chamber surfaces might reduce pre-ignition further. but usually requires a new crankshaft. Another part of the problem was inability to accurately check ignition timing on the engine. Therefore the elimination of these might be desirable. high performance engines. connecting rods and pistons. Cooler cylinder temperatures will reduce likelihood of pre-ignition. Sharp edges can stay very hot during engine operation. Engines today use thermostats which keep the coolant temperature at about 100 C. Thermostats opening at about 80 to 85 C are commonly used in modified. A downside to this is increased engine wear and efficiency. Running an engine on H2 Considerable difficulties were encountered trying to initially start the test engine used in the FutureTruck.possible by boring the cylinder. This head was also milled to produce a compression ratio of about 12. The main reason for the difficulties was the lack of base fuel and ignition timing maps for the ECU. Decreasing the combustion chamber volume is usually accomplished easier by milling the cylinder head. The cylinder head used in Texas Tech’s FutureTruck was changed to an aftermarket version made of aluminum. This can be accomplished by installing a lower temperature thermostat. Increasing the stroke can increase the compression ratio. Changing the pistons to one’s that protrude past the deck height also increases the compression ratio. This type of thermostat was installed in Texas Tech’s FutureTruck.6:1.

in 18 . Reducing the injected fuel amount generally will reduce this. and have long operating ranges. As a general rule. The stoichiometric simulations are done as a reference. They do have some disadvantages such as having poorly matched fuel efficiency characteristics with operation of a normal driving. The simulation done in this study cover a broad range of equivalence ratios from very lean up to stoichiometric. being unable to harness the vehicle’s kinetic energy during braking. and a low efficiency of vehicle transmissions. at very low loads. leading to poor fuel economy. especially in stop-and-go traffic. and possible a little above. Hybrid vehicles can take the good characteristics of both internal combustion engines and electric drive trains. They posses good power to weight and volume. The internal combustion engines also produce higher emissions since they have to operate over a broad speed and load ranges. The smaller engine makes enough power to comfortably allow the vehicle to cruise. a conventional. more difficulties were caused by back-flash and detonation. It is possible to run the engine at these equivalence ratios. The total power of the vehicle is then brought up with the use of the electric motor. reducing the operating range (Ehsani). and the engine started. The way these systems work is by using a smaller displacement engine in conjunction with an electric motor.Once the initial difficulties were solved. Electric vehicle overcome the tailpipe emissions. such as idling.75 at high loads. being able to lower the emissions and fuel consumption considerably. but power suffers dramatically. Hybrid system basics Conventional internal combustion engines have traditionally been used in vehicles. although running a conventional engine so rich is not feasible. This idea is not new. but does not necessarily have the power to accelerate the vehicle fast enough while merging and going up hills. but batteries have a poor energy density. port injected hydrogen engine will not run well at equivalence ratios above 0.

Each one of them has its advantages and disadvantages. This allows the engine to work at a lower. A large amount of the kinetic energy of a vehicle can be recovered by using the electric motor as a generator to recharge the batteries. Series hybrid-electric systems In a series hybrid-electric vehicle the energy flows in a series from one component to another. Types of Hybrids Traditionally. when the engine is making more power than needed by the propulsion of the vehicle. there have been two basic types of hybrids: parallel and series. In the resent years. which fit into neither one of the basic categories. They store and release energy as needed. A newer use of the hybrid system is regeneration. which incorporate more than one electric motor. and the complex hybrid. Batteries play an important role in hybrid-electric vehicles. Piper (Wouk). This has given rise to the series-parallel hybrid. which can run in either one of the basic modes. a patent was filed in 1905 for a vehicle that used an electric motor to augment the internal combustion engine by H. more constant output. Also some of the energy is lost to friction and aerodynamic drag. In a conventional vehicle this energy is lost through heat in the brakes. These vehicles are set up very similarly 19 .fact. The batteries are charged during cruising and decelerating. making less emissions and lowering fuel consumption. Conventional. hydraulic brakes are still needed to bring the vehicle to a complete stop as it is difficult to control an electric motor at low speeds. The engine drives an electric generator which provides electric power to the drive motors and charges batteries. This discussion will focus only on the more conventional series and parallel hybrid drive trains. It is not possible to regain all of the kinetic energy. more ways of incorporating the engine and electric motor to the drive train have been invented.

There are three basic parallel systems: pre-transmission. and thus heavy. as there is no way to transfer mechanical power directly to the wheels from the engine. Pre-transmission parallel In this hybrid-electric type the electric motor is placed in the drive-train before the transmission. posttransmission and road parallel. Also this design can make regenerating while decelerating difficult as some transmissions do not efficiently transfer power backwards. Another drawback is that a high power. requiring them to spin at the same speeds. This system has the advantage that the vehicle and engine speeds are not coupled to each other. Parallel hybrid-electric systems In this system power can flow from either the electric motor or the engine to the wheels. This kind of system is easier to incorporate into an existing design since the basic engine and transmission system are left in place. Advantages of this system are that the torque of the electric motor is multiplied through the transmission and it is possible to charge batteries while the vehicle is not moving. electric motor is needed to accelerate the vehicle since it is the sole source of propulsion. An electric motor is placed in parallel to the drive train. This does mean that there are power losses through the transmission. or through a system of clutches so that the motor and/or the engine can be uncoupled from the transmission.to an all-electric vehicle. A major drawback to this system is that there are significant energy losses while converting energy from mechanical to electric and back to mechanical. while power from the batteries is used to fill in the power needs of acceleration and recapture the energy during braking. The engine provides a constant power output. 20 . It is possible to have the motor directly coupled to the engine. It is possible to operate the engine at its peak efficiency and tune it for best emissions at that operating point.

especially into an existing rear wheel drive design. electric motors.Post-transmission parallel This set up gets rid of any regeneration problems that might occur in the pretransmission systems. and another one to connect the motor to the rear end. etc. This system has the advantage that the room in between the front and rear wheels is not taken up or split by the drive shaft. This can leave the space for use by batteries. The electric motor is a Solectia AC 55. This system does share most of the disadvantages of the post-transmission parallel system including not having the capability to charge batteries while the vehicle is not in motion. with a modified shaft with splines on both ends. Two disadvantages of this are that there is no torque multiplication through the transmission and it is not possible to charge batteries while the vehicle is not moving. An advantage to this system is the ease of incorporating it. In essence. The electric motor is placed after the transmission. unconnected. Usually the engine and transmission are in a front wheel drive configuration. The torque multiplication can be attained by coupling the motor through a step down gear ratio. a piece of the drive shaft was replaced with the electric motor. 21 . Road parallel The road parallel system incorporates two separate. This allows a drive shaft to connect the transmission to the electric motor. and will always turn at a speed relative to the vehicle speed. fuel tank. There is no simple solution to charging the batteries while the vehicle is not in motion. propulsion systems driving the front and rear wheels. Texas Tech’s FutureTruck is this type of design. The engine and transmission can be left intact. controllers. while the electric motor drives the rear wheels. while all of the hardware is added to the space in between the transmission and the rear end.

22 . such as idling. those conditions are used as the boundary conditions for the next section. and exhaust manifold. air flow through these ducts is far from being a steady state process as valves open and close several times each second even for low engine speeds. which was used in this study. Calculations for many small pieces of the ducts could be done. Also. heat-transfer for the energy losses to the cylinder walls. During the late 1800’s and early 1900’s. and as the results are calculated for one section. and kinetics for the frictional losses (Atherton). This engine model was modified. increasing the compression ratio. two basic models were simulated. The ducts in a multi-cylinder engine can be very complicated. much like the real life test engine used in the FutureTruck Explorer. it became possible to perform thousands of calculations in a very short time. designers have wanted to simulate their engines prior to building them. The modifications included the following: adding a supercharger and an intercooler. is this kind of a program. Each piece of ducting is looked at separately. changing the cylinder head. As knowledge of fluid flow increased. and thus the early designers could not easily model them. even if the model is one-dimensional. gasoline engine.CHAPTER II ENGINE SIMULATIONS Ever since the first engines were built. This was done as a baseline. One was a naturally aspirated. camshaft timing. With the advent of computers. and as a way to verify that the simulation results were reasonable. This can accurately model an engine. In the Texas Tech FutureTruck project. it became possible to do steady state models of simple ducts of an engine. only rough combustion analysis could be done. unmodified. The calculations for each section are done for small increments of the engine position. thermodynamics for the combustion process. Ricardo WAVE. These include performing fluiddynamics for the air and fuel flow to and from the cylinder. It was realized early that several different analyses need to be done in order to accurately model an engine.

and liquid fuels. and is not limited to just calculating torque curves. the results should be correlated with dynamometer data. Because of time and equipment limitations. mass flows. This program provides fully integrated time-dependent fluid dynamic and thermodynamic calculations using a onedimensional formulation. momentum. Once a model is run. and emission results. This should not make the simulation results unusable. torque. The equations are: Mass Energy dm   m dt dme    mh  sources dt 23 . plenums. and energy. the initial correlation was not done for this study. multinational corporation that does research for several major auto manufacturers. Description of how WAVE calculates air flow WAVE uses a quasi-one dimensional compressible model to calculate the airflow using conservation of mass. and manifolds of various systems and machines. volumetric efficiency.Ricardo WAVE background Ricardo is a large. WAVE uses a general treatment of working fluids including air. air-hydrocarbon mixtures. (Ricardo) Results given WAVE is capable of giving many different results. general trends shown by the simulations should be correct. as well as fuel consumption results. combustion products. It can also give emission data. and energy losses in ducts. and do not necessarily need to be checked with the dynamometer. This study concentrated on the power. Ricardo has developed the engine simulation package WAVE to analyze the dynamics of pressure waves. This assures that changes made on the model will be very accurate.

They are recommended for use with junctions containing perforated walls (such as mufflers). The junctions assume that all 24 . The exhaust manifold is 4-into-2-into-1 type.Momentum dmu dp    A dx   mu  losses dt dx How models are built The engine models in WAVE are built using the dimensions of each runner that the air goes through from the air intake to the tip of the exhaust.1 shows the model used for the supercharged hydrogen engine. This is done by setting the angles from each axis to the center of the duct.or complexjunctions. with similar exhaust ports on the right side of the cylinders. The complex-junctions are the square junctions in the model and represent more duct-like volumes. The model starts at the top-right corner. Each block on the model represents a junction between ducts. or for volumes with many parallel passages (such as intercoolers and catalytic converter inlets and outlets) (Ricardo). The intercooler is located on the top. and the model takes the energy losses into account. The fuel injectors can be seen on the intake manifold. on an ambient junction which connects to the supercharger. while the lines are the ducts. Every volume is assumed to be either a cut-off cone or a sphere. Large volumes and junctions are modeled using either simple. from which four separate ducts direct the air to the intake manifold. Bends are entered separately. four ducts after the supercharger. Figure 2. The junctions set the positions at which the ducts come into and out of the junction in three dimensions. More ducts connect the intercooler to the intake manifold and the plenum. This is connected to the catalytic converter. although they might not be shown on the visualization of each runner. almost at the end of the model. The intake ports connect to the cylinder. Figure 2. and are modeled as spheres. The simple-junctions are the round ones on the model.2 shows the input panel for the junctions. The muffler is modeled by the dead ended duct. next to another ambient junction.

Figure 2. it is not possible to say that the angles are 10 degrees from every axis. i. They do not have to be necessarily in a single plane. The basic shapes make it somewhat difficult to model complex shapes such as intake and exhaust runners. all of the ducts are in a single plane since the angle from y-axis is 90 degrees for all of them. The angles used to define the direction have to agree with each other. In the example. These setbacks can mostly be overcome as actual flow data can be entered for any runner in the model. 25 .e. The actual dimensions of some parts are very difficult to accurately measure without making molds of the runner shapes or cutting the part open. as the directions would not coincide.3 Ricardo WAVE model of the supercharged hydrogen engine.ducts originate from the center of the junction. yielding very accurate models.

3 shows the input panel for each duct. This cannot be set too long.Figure 2. although all of these are in a single plane. The main points to note on it are the end diameters.4 A visual representation of a junction. This is the junction of two exhaust manifold passages coming to form one. The equation for conservation of momentum is solved at the boundary of each volume. It sets the number of different sub-volumes that WAVE uses to calculate the airflow. Discretization lengths Figure 2. Ricardo recommends using the following for discretization lengths: 26 . All of the tubes coming to the junction can be set anywhere in three dimensions. and bend angle. as then the accuracy of the model is compromised. The shorter this is set. the more calculations need to be done. overall length. The discretization length is also an important part of the model. and the model will be slower.

In order to throttle the engine. the diameter of the junction could be made smaller.8 inch diameter hole was used in the junction. 27 . one of the intake runners in this case. the tubes attached to the throttle were 2. the bend angle.55 * Bore On the models used. Throttle A butterfly valve is used as a throttle on most engines. At idle conditions. The throttle on WAVE models is made by placing a junction which is smaller than the tubes attached to it. or rounded down. Since this is a relatively complicated piece to model.Intake side: dx = 0. The dotted lines dividing the duct split it into several sub-volumes. The diameters at each end. Note the injector placed toward the right end of the duct.0 inches in diameter.5 View of the input panel for each duct through which air travels. For these simulations. a simpler approach is used.45 * Bore Exhaust side: dx = 0. the hole needs to be very small. and lengths are entered. thus at WOT conditions. Figure 2. The butterfly valve does restrict the airflow even when it is completely open. these recommendations were usually used. a 1.

and specifying that there is a number of these. 28 . For the intercooler the number of small ducts was counted to be 17. and by using the cross sectional area of the catalytic converter. The catalytic converter was welded into the exhaust. The way WAVE recommends modeling them is by using regular spheres for the open ends of them and attaching many small ducts to these. This is because they have multiple small ducts. The small ducts only need to be entered as a single duct. Based on the examples in WAVE. The data were measured by two separate sources on identical Esslinger aluminum heads (mercurencyclopedia. the number of ducts was estimated to be 1500. Simulating cylinder heads Flow data for the cylinder head on the engine were entered into WAVE. Their size could be measured with reasonable accuracy. The data used in the simulation were from mercurencyclopedia as that had data for higher valve lifts. and therefore should be a good representation of what the head used on the experimental engine flowed. and thus it was not possible to count the ducts or measure their exact size.1. turboford). The other source was used only to check the validity of the data. The data can be seen in Table 2. The data were virtually identical.Simulating the intercooler and the catalytic converter The intercooler and the catalytic converter are two pieces that cannot be modeled like regular ducts.

org mercurencyclopedia. The measurements were taken at the standard pressure difference of 28 inches of water. including lift.3 0. Stiff valve springs can also lead to premature camshaft wear. etc. Most modern camshafts try to open the valves as quickly as possible. 29 .com turboford.org Intake Intake 0. this should not be significant. Using a generic profile does introduce some error. and therefore there should be minimal differences in the profiles.2 shows the overview of the camshaft profile.2 0.5 0. however. Table 2. duration. Camshaft lifts greater than this can be used in engines. The information received from Cam Doctor was used to modify a generic camshaft profile that comes with WAVE. the valve lift is indicated. Simulating the camshaft The camshaft profile was measured using Cam Doctor software.com turboford.6 0.4 inches. but very stiff valve springs are required if high valve lifts are used.1 65 65 0. The maximum valve lift for the model is approximately 0.7 117 162 201 116 162 201 101 135 154 101 134 154 228 241 247 228 241 172 183 192 172 183 Exhaust 52 Exhaust 51 The table above shows how much air will flow through a single port in the cylinder head. At the top.4 0.Table 2. This software gives all of the camshaft specifications.2 Cylinder Flow Data from Two Sources (cfm) Valve Lift (inches) mercurencyclopedia. lobe center separation.

30 .4:1.5 0. WAVE has a function to create the continuous four-dimensional compressor map. Seventy data points were entered from the Rotrex data into WAVE. and a ratio of the belt pulleys of 2. belt driven supercharger was used in the hydrogen model. This includes an internal ratio of 9.4 and 2.3 0.3 Partial Summary of Camshaft Specifications Lobe center separation 121.48:1 inside the supercharger.752:1.7 Crank deg Intake Valve Duration (at 0.050 in) Max cam lift 202. Rotrex provided the efficiency data for the supercharger.Table 2.5 to indicate that they are the same. The total ratio of the turbine speed to engine speed is 22. Both of the efficiency curves are shown below in Figures 2. turbine speed.23713 Crank deg Inches Exhaust Valve Duration (at 0. and efficiency.23791 Crank deg Inches Supercharger modeling A Rotrex centrifugal.7 Cam deg Valve overlap -53. pressure ratio of inlet and outlets.050 in) Max cam lift 205. which compared the mass flow of air.

Figure 2. The horizontal axis shows the mass flow of air through the supercharger. The vertical axes show the pressure ratio. or the ratio of the absolute outlet pressure to the absolute inlet pressure.7 Compressor map as entered into WAVE The compressor maps above show the supercharger efficiency. Lines which run approximately horizontal in Figure 2.4 and continue almost vertically 31 .6 Supercharger Efficiency Curves as Given by Rotrex Figure 2.

5. which are colored in Figure 2. The formula used by WAVE is: WEXP 1     AWI     W  1  exp     BDUR     where: W  = cumulative mass fraction burned = crank degrees past start of combustion BDUR = user-entered 10-90% burn duration in crank degrees WEXP = user-entered Wiebe exponent AWI = internally calculated parameter to allow BDUR to cover the range of 10-90% The model requires inputs for the duration. A larger exponent starts the combustion process very slowly initially. This model calculates the rate of mass that is burned. the Wiebe exponent. and quickly later on. The duration is the number of crank degrees for 80% of the combustion to occur. This does not include the first or the last 10% of the combustion. The Wiebe exponent controls how fast the combustion happens relative to the 50% burn point. Ideally. The islands. A smaller (less than 2) Wiebe exponent burns the fuel quickly initially and slowly after the 50% burn point.down in Figure 2. Curves toward the bottom of the figure are slower than the ones above. as well as the timing when the combustion happens. how much of the fuel gets burned. WAVE automatically subtracts the power used to drive the supercharger from the engine output.5 indicate the turbine speed. WAVE recommends using 2 as the Wiebe exponent. 32 . are constant turbine efficiencies. The timing is entered as crank degrees for half of the combustion and heat release to have occurred (50% burn point). and thus that was used in all simulations. Combustion model The combustion model used by WAVE is Wiebe based (Ricardo). an engine will operate as close to the efficiencies in the middle of the curve. as these are the highest. The power to drive the supercharger is taken form this efficiency map.

such as a gasoline engine running under a high load. This iteration process was used until the burn duration for the hydrogen was within 0. These pressure sensors are expensive and difficult to obtain. a 1 to 10 fraction of indoline to hydrogen flame speeds was set. All of the available fuel should burn in the engine. theoretical flame speeds. The same procedure was used for 1500 and 4500 rpm. The flame speed for the V8 example indolene engine. for the 80% burn duration. The following is an example of the burn duration for the simulations running at 3000 rpm. so these data were not available. running at 12:1 air fuel ratio (Ö=1. Initially. because of the extremely lean mixtures that the hydrogen engines run. For the hydrogen simulations. Calculating the burn duration In order to get accurate results.0. at pressure of 1 atmosphere. Average figures for these were taken from the engine simulations. The resulting average pressure and temperatures were used to calculate new flame speeds for the hydrogen model. the fraction of fuel burned was always set to 1. as well as bore/store ratio to the engine being simulated. The duration inputs from the example gasoline engines were used. and 298K temperature. This fraction needs to be set to less than one for engines running under rich mixtures.5 degree of the previous iteration for the hydrogen model. is: 33 . The multipliers were the ratios of the calculated. cylinder pressure data would be needed from an engine running on a dynamometer.The amount of fuel burned needs to be set also. This engine was chosen as it had the closest bore and stroke. The theoretical flame speed is a function of temperature and pressure. Linear curve fitting between these points was used to obtain the burn duration for the engine speed around these points. The reference flame speed.125) was calculated from the following (Turns). along with multipliers to correct them to the hydrogen engine. and the burn duration was adjusted accordingly.

244 P=80 atm T=1846 K The reference flame speed.22  1 = -0.87cm/s.9 degrees in the hydrogen engine.8  1 = 2  P  P  ref   = 699cm/s.      0.4855   0.1105 P=45 atm.S ref  Bm  B2    m  2 =26.62) uses the same formula as above. This means that the burn duration in the indoline engine of 28.22  1 = -0. taken from P-V diagrams after running WAVE for the example model. at 1 atm. The flame speed calculation for the hydrogen simulation running at 55:1 air fuel ration (Ö=0.18  0. The pressure and temperature figures are the ones that were calculated at the end of the iteration process.6:1. The average temperature and pressure were taken for the burn timing that resulted in maximum torque. The flame speed ration is then 6712:699.16  0.18  0.8  1 = 2.2 degrees corresponds to 2. They resulted in the flame speed of 6712cm/s. is 210 cm/s (Turns). The one which made maximum torque was picked to 34 . The flame speed in the combustion chamber should be:  T S  S ref  T  ref      Where:   2. this is an average pressure in the combustion chamber during combustion. or 9. T=1876 K average temperature for the example engine in the combustion chamber during combustion. This burn duration was entered back into WAVE and simulations with multiple spark timings were run. 298 K and stoichiometric mixture. with the exception of the following coefficients:   2.16  0.

or through the crankcase ventilation system. The oil is not intended to be burned. The HC emissions are mainly made by unburned fuels. Using synthetic oils has been said to decrease the emissions. In some cases it is not possible to run synthetic oils. and improving the oil trap for the crankcase ventilation system. and only concentrated on the NOx emissions. producing higher piston ring to cylinder wall pressure. improving the valve guide seals to prevent oil from getting past the guides. some CO is made. Today emissions cannot be ignored. and the flame speed calculation process was repeated until the same results were obtained as the previous run. and therefore do not contribute to those emissions. some amounts can still be made. leakage through the intake valve guide. and nitrous-oxides (NOx). The oil can make its way there past the piston rings.supply new temperature and pressure data. The main emissions measured today are hydrocarbons (HC). carbon-monoxide (CO). as the norms faced by the vehicle manufacturers are getting increasingly more stringent. This is due to burning the lubricating oil in the engine. Emission Emissions have played an increasingly significant role in automotive development for the past several decades. This is due to the fact that the fuel has no hydrocarbons. The 35 . When the chemical reactions are not completed during combustion. Some solutions to minimize these are: run a tighter piston ring gap. This is the case of the Texas Tech’s FutureTruck Explorer. The lack of hydrocarbons also keeps the CO emissions very low. They also are slicker. These oils are not made with hydrocarbons. Ideally all of the carbon in the hydrocarbons would be converted to CO2. decreasing the internal friction in the engine. This simulation did not take into account any of the HC and CO emissions. and steps have been taken to minimize getting it to the combustion chamber. Although at first it seem unlikely to make any CO and HC emissions in a hydrogen engine. The HC and CO emissions are extremely low in an engine using hydrogen as a fuel.

supercharger in it uses the engine oil for its lubrication. This is shown in the close up view in Figure 2. shown in Figure 2.7. 36 . The power comes in through the three smaller rollers inside the larger ones. thus transmitting the power poorly to the compressor wheel. These gears use a thin film of oil to transmit power from one smooth gear to another. The larger rollers then transmit the power to the output shaft placed in the middle.8 Traction gears in the supercharger (Rotrex).6. The supercharger uses planetary roller traction gears. Figure 2. Figure 2. The synthetic fluids are so slick that they would let the gears slip more than the designed amount.9 Oil being used to transmit power in the supercharger (Rotrex).

although this increases emissions. The surprising effect of the timing is that maximum torque is produced.1.1 Brake torque v. In order to determine the optimal timing. It can clearly be seen from the graph that having too much spark advance reduces the torque as the in-cylinder pressure increases before the piston reaches TDC. spark timing for supercharged hydrogen engine running at 3000 rpm. less work is done on the piston since it has already begun to move down. The maximum torque for the 37 . 55:1 AFR and WOT. 140 130 Torque (N-m) 120 110 100 90 80 -20 -10 0 10 20 30 40 Spark timing (degrees before TDC) Figure 3. but at a retarded timing. The crank angle when half of the fuel has burned is specified. The optimal timing for WOT was taken to be the one that maximized torque. doing negative work on the crankshaft. As the fuel mixture is lighted too late. not at slightly advanced. and the burn duration then sets the spark timing. a sweep of different timing settings were ran. This can be seen in Figure 3. or no advance. for wide open throttle runs.CHAPTER III SIMULATION RESULTS Spark Timing The spark timing is set indirectly in WAVE.

it changes the emissions. The spark timing directly affects the cylinder temperatures and pressures. especially the NOx. The advanced timing 38 . Possible reason for this is that the cylinder pressure is very high when the piston is at TDC increasing blow-by past the piston rings. Because of this. Thus the net effect is that the maximum torque occurs at slightly retarded spark timing.3 and 3. The in-cylinder pressure for different crank angle can be seen in Figure 3. At the same time. 2000 Cylinder Pressure (KPa) 1600 -14deg 1200 800 400 0 -45 -30 -15 0 15 30 45 Crank Angle (degrees) -7 deg 0 deg 7 deg 14 deg 21 deg Figure 3. The fast flame speed of hydrogen can be seen in the graphs as the temperature increases almost the same amount. the work done near TDC is a lot less than when the piston has moved away from it. Beginning of the combustion process can also be seen in the in-cylinder temperature graphs.2 Cylinder pressures v. Figures 3.sample case occurred at retarded spark timing around 7 degrees.2. Another note worth mentioning is that even slightly advanced timing produces very high pressures that fight the piston moving up to TDC. crank angle for different ignition timing at 3000 rpm and 55:1 afr.4 also show the effect of ignition timing in cylinder temperatures. Only small differences in the maximum temperature can be seen due to the cooling effect of the cylinder volume increasing for the retarded spark timing. for different timing settings. The beginning of the combustion can be clearly seen in the graphs.

3 In-cylinder temperatures for three different spark advances on the supercharged hydrogen model running at 3000 rpm. The temperatures however take a longer time to fall.temperatures do go slightly higher than the retarded due to the cylinder volume decreasing after ignition. 39 . This can be seen in Figure 3. This is because the heat transfer takes time to occur. Only after fresh air enters the cylinder. and thus are different until after the intake valve opens. The in-cylinder pressures have very small differences after the combustion is completed.4. 55:1 AFR and WOT. 5000 4000 Temperature (K) 3000 -14 deg 0 deg 14 deg 2000 1000 0 -90 0 90 180 270 360 450 540 Crank Angle (Degree s) Figure 3. do the temperatures return to the same level.

Maximum NOx is produced at about 7 degrees of spark advance. and increase very rapidly during the combustion process. This produces the maximum combined pressure and temperature. but torque output is decreased. diluting the mixture. Advancing the timing to 14 degrees before TDC lowers the pressure during combustion. the NOx levels start low. Retarding the timing lower the NOx levels considerably.4 Close up of in-cylinder temperatures for different spark advance at 3000 rpm. In-cylinder NOx levels can be seen in Fig 3. 55:1 AFR and WOT. Having a highly retarded timing lowers the NOx levels tremendously.5. as the peak pressure and temperature is decreased.5000 4000 Temperature (K) -14 deg 3000 -7 deg 0 deg 7 deg 2000 14 deg 21 deg 1000 0 -30 -15 0 15 30 45 60 Crank Angle (degrees) Figure 3. and thus reduces the NOx levels. 40 . The level then stays constant until the intake valve is opened and fresh air starts entering the combustion chamber. In all of the lines.

The timing setting can be set to produce the maximum power. from 7 degrees to 14 degrees retarded.5 NOx emission levels in the cylinder for different spark timing at 3000 rpm. This can lead to severe engine damage. decreases the torque significantly. Running too much ignition advance will decrease power. or minimize emission. Negative timing refers to spark occurring BTDC. If getting the minimum emissions at WOT is most important. increase emissions. Retarding the timing a little more. decreases the NOx by another third. 55:1 afr at WOT. and most likely increase likelihood of preignition or detonation. with only about 2% decrease in torque. Retarding the timing any more. but at slightly advanced timing. a little power can be sacrificed. 41 . Retarding the timing by another 7 degrees to 21 degrees retarded. running less retard will make the power and emissions go up. attained at the slightly advanced setting. Setting the timing to produce maximum torque produces about half of the maximum NOx.3500 3000 In-Cylinder NOx concentration (ppm) 2500 2000 1500 1000 500 0 -90 0 90 180 270 360 450 540 Crank Position (degrees) -7 deg -14 deg 0 deg 7 deg 14 deg 21 deg Figure 3. Note that maximum NOx levels are not attained at earliest timing. If getting every last bit of power out of the engine is wanted. but decreases the torque by about 6% from the maximum. decreases the NOx emissions by about a third.

as this produces the maximum efficiencies. The mean effective pressure is very closely related to torque output of the engine. 42 . The Effect of Volumetric Efficiency on Torque Since the volumetric efficiency is a measure of how much air enters the cylinder. as the curves follow each other.Volumetric Efficiency The volumetric efficiency is an important measure of a four-stroke engine’s ability to pump air. The relation between volumetric efficiency and torque can be clearly seen.  a.iVd N  where ma is the mass flow rate of the air entering the cylinder.i is the air density. This is closely related to the torque output of the engine. it also measures how much fuel enters the cylinder at a given equivalence ratio. gasoline engine. The efficiency can be lowered by closing the throttle. It is defined as the volume flow rate of air into the cylinder divided by the rate at which volume is displaced by the piston. and N is the engine speed. Figure 3. The amount of air-fuel mixture burned in the cylinder is directly related to the mean effective pressure.6 shows the torque output and volumetric efficiency for the naturally aspirated. v   2m a  a. Vd is the displaced volume. The volumetric efficiency is generally measured at wide open throttle.

The two volumetric efficiencies are almost identical. 43 .220 1.10 0.70 0. leaving less room for the air being pumped into the cylinder.90 0. Fuel vapor can take considerably more of this space. The first model was ran at a stoichiometric mixture (14.40 0.6 Torque (left axis) and volumetric efficiency (right axis) versus engine speed for a naturally aspirated gasoline model.60 Torque Volumetric Efficiency 0.5:1 air-fuel-ratio).80 0. decreasing the volumetric efficiency. Since the two resulting lines are almost identical.20 0. The Effects of Fuel Type. and 90% liquid being injected from the fuel injectors. with less than one percent difference between each data point. The naturally aspirated gasoline model was modeled twice.7:1 air-fuel-ratio). Equivalence Ratio and Supercharging on Volumetric Efficiency The volumetric efficiency is affected by the fuel being burned in the engine.00 0.00 6000 Torque (N m) and Volumetric Efficiency 200 180 160 140 120 100 1000 2000 3000 4000 5000 Engine Speed (rpm) Figure 3. the fuel takes more space.50 0. Liquid fuel takes up very little space in the intake ports and the combustion chamber. while the second one was ran at a rich mixture (12. As a vapor fuel is burned richer.30 0. Figure 3.7 shows volumetric efficiencies for several models varying the equivalence ratio.8. Both models used a 10% fuel vapor. an expanded view of them is shown in Figure 3.

62 1. a naturally aspirated (N/A) gasoline engine for stoichiometric and rich mixtures. There are two curves on the figure which show the effect of supercharging an engine. and a naturally aspirated hydrogen engine. This line follows the shape of the gasoline models.62 equivalence ratio. The supercharged models show the decrease in volumetric efficiency very dramatically. The same.6 0.0 N/A Gasoline eq 1. 44 . while the richest is the least efficient.6 Volumetric Efficiency S/C H2 eq 0. is less efficient at low engine speeds. The air-fuel-ratio used was 55:1. The 55:1 AFR supercharged model is the middle one of the supercharged models. the volumetric efficiency dropped by approximately 2025%. Just because of the change in fuel. This is for the 55:1 air-fuel-ratio for hydrogen. The leanest air-fuel-ratio is the most efficient.176 N/A Gasoline eq 1. supercharged model. with its bends and long pipes. it is not true. or 0.62 S/C H2 eq 1. Although at first it might appear that the supercharged model. naturally aspirated engine was also modeled using hydrogen as the fuel.8 1. The three efficiency curves are almost parallel.8 0.7 Volumetric efficiencies for a supercharged (S/C) hydrogen fueled engine varying air-fuel ratio. being very close to the curve of the stoichiometric.0 N/A H2 eq 0.2 0 1000 2000 3000 4000 5000 Engine Speed (rpm) 6000 Figure 3.1. with no other changes. but at a much lower level.2 1 0.4 0. Even at low speeds the efficiency of the naturally aspirated 55:1 AFR model is lower.4 1.4 S/C H2 eq 0.

If the pressure in the intake port is high when the intake valve opens. the volumetric efficiency for engine speeds less then 4500 rpm is higher for the rich mixture. similar results for the hydrogen would be expected.8. if this pressure is low. less air will flow into the cylinder. The pressure waves occurring in the intake and exhaust ducts are more favorable at the engine speeds where the volumetric efficiency is high. In fact. The supercharged model barely shows the increased volumetric efficiency near 2500 rpm. it appears that the supercharged model is more sensitive to the engine speed. As some of the fuel is vaporized when it is injected. the effects of the tuning are considerably smaller than those caused by the increasing outlet pressure from the supercharger. Changing the air-fuel-ratio for the naturally aspirated gasoline model does not change the volumetric efficiency significantly. however. the same tuning effects should be seen on all models. thus adding to this effect. Tuning is adjusting the intake and exhaust duct diameters and lengths. Surprisingly. which is very prominent in the naturally aspirated model. This does happen for the high engine speed. As air is flowing through the ducts. more air will flow to the cylinder. The supercharged model could also be tuned for some engine speed. the difference in volumetric efficiency between the stoichiometric and rich mixtures is less than one percent.7 and 3. The volumetric efficiency curves for the naturally aspirated engines are much more sensitive to tuning. This is true. there are pressure waves traveling in them. Since the basic intake and exhaust manifolds were used in all models. decreasing the volumetric efficiency. as well as valve timing. but the reason for this is the increasing outlet pressure from the supercharger. when the volumetric efficiency for the rich mixture dips below the one for the stoichiometric. On the other hand. but at a smaller scale. The curves are smoother than the naturally aspirated ones. The drop-off in the efficiency at higher engine speeds can be seen a little better in the supercharged model by decreasing slope of the curve.Comparing the curves of the naturally aspirated and the supercharged model. This can be seen in Figures 3. 45 . It is possible that the supercharger itself is not as efficient at the higher speeds.

8 An expanded view of the volumetric efficiencies for the naturally aspirated gasoline model run at a stoichiometric (14. It is also possible that the simulation result is correct.9 Volumetric Efficiency 0.9 shows the brake specific fuel consumption the naturally aspirated hydrogen engine 46 .1. The same burn duration and timing were used for both models.7:1 air-fuel-ratio) and rich (12. Figure 3. The same curves are shown in Figure 2.5:1 air-fuelratio).6 1000 Stoichiometric Rich 2000 3000 4000 5000 6000 Engine Speed (rpm) Figure 3. and would burn at a different rate. The brake thermal efficiency was plotted as a function of equivalence ratios. Dynamometer data were available for hydrogen engines from Ford. Changing these would most likely change the pressure and temperature in the combustion chamber when the valves are open. It is possible that the rich mixture would need a different timing advance.85 0. The seemingly backwards results for the low engine speeds could possibly be an error in the inputs for the simulation.8 0.0. and the engine burns the air more efficiently for the rich mixture. Compression Ratio Since compression ratio can affect the engine performance.7 0. varying it was simulated. which directly affects the volumetric efficiency.65 0.75 0.

3 0.5 17. The dropping in the efficiency of the engine as the compression ratio is raised could be as a result of a few things. Most likely the reason for this is that the simulations do not have a knock model.27 0. 0.3 15. which lowers the power output considerably. One potential reason is the increased knock.31 Brake Thermal Efficiency 0. there is a significant increase in the work that must be done on the mixture to compress it.28 20 10 0.8 0.5 12. It is possible that these two do not grow equally. which works against the work done by the pistons.75 equivalence ratio.9 Brake thermal efficiency as the equivalence ratio is changed for the naturally aspirated hydrogen model for different compression ratios.26 0. Another possibility is that increased blowby past the rings causes some of the fuel to be not converted into usable work. These are the same operating conditions used for the data in Figure 1. requiring highly retarded ignition timing.55 0. As a real engine is ran richer than about 0.5 0.1.29 14.9 1 Equiv ale nce Ratio Figure 3.6 0. This also will increase the pressure in the crankcase.7 0. As the compression ratio is increased. Also. It can be seen that the simulation results do match trends of the dynamometer data. since hydrogen burns so quickly. 47 .running at 3000 rpm at WOT. The simulation results do not fall down as quickly as the dynamometer data as the fuel mixture is richened. knock starts becoming a major problem. it is possible that there is no significant increases in the work done by the burned mixture.4 0.

this is compression ratio is still along the peak values of efficiency.75 0. The same simulations were also done for the supercharged engine.3 0.12 shows is again a better way of representing the results. A very interesting point to note on that figure is the drop in the efficiency in the middle of two peaks.5:1 to 15:1.48. The engine would be just as efficient running with approximately a 13:1 or a 17.48 0.11 shows the simulation results as the equivalence ratio is changed.Figure 3.9. as the lines are very close to each other in the Figure 3.10 shows the same date as shown in Figure 3. The two peaks are virtually the same efficiencies. For moderate equivalence ratios.29 0.28 0. 0.10 Brake thermal efficiency as the compression ratio is changed for the naturally aspirated hydrogen model for different equivalence ratios. such as 0.96 0. Figure 3. for different equivalence ratios. like 0. For leaner equivalence ratios.5:1 compression ratio. The results from it are not as clean as the ones for the naturally aspirated engine.62.10 the thermal efficiency is plotted as a function of compression ratio. The curves are virtually the same shape as the ones for the naturally aspirated engine.62 0. Figure 3. but shifted higher. the peak in the efficiency occurs for a compression ratio of about 14. The difference is that in Figure 3.9. 48 . This makes it easier to see the ideal compression ratio for engine.31 Brake Thermal Efficiency 0.27 10 12 14 16 18 20 Compression Ratio Figure 3.

55 14.5 12.5 15.32 0.12.11 Brake thermal efficiency as the equivalence ratio is changed for different compression ratios on the supercharged hydrogen engine running at 3000 rpm and WOT. The lower compression ratio would be a better choice to use in an engine.4 0.3 20 10 Figure 3.29 0.3 0. 49 .315 0. A combination of the engine and supercharger efficiency is shown in Figure 3.8 1 Equivalence Ratio 17.33 0.325 Brake Thermal Efficiency 0.31 0.305 0.6 0. This reduces the tendency to knock and is generally easier on most engine parts.295 0. 0.The reason for this odd behavior is that the supercharger efficiency changed as the compression ratio is changed.

The aftermarket does make adjustable pulleys for most of the popular engines.33 0. as adjustment is not needed on a stock engine and the adjustable pulleys need to be made from at least two parts.75 0.0. Factory pulleys are usually not adjustable.48 Figure 3.96 0.12 Brake thermal efficiency as the compression ratio is changed for the supercharged hydrogen engine for different equivalence ratios at 3000 RPM and WOT. duration. costing more to manufacture. provided that the camshaft has an adjustable pulley. In an engine with a single camshaft changing the timing is the only variable that can be changed without grinding a totally new camshaft. There are a number of variables that can be changed in the camshaft including lift. 50 . lobe center separation angle.305 0.32 0.62 0. timing and the lift profile.3L Ford used in this study.315 0. Camshaft Timing Camshaft specifications affect engine performance dramatically.325 Brake Thermal Efficiency 0.31 0. including the 2.3 10 12 14 16 18 20 Compression Ratio 0. In an engine with separate intake and exhaust camshafts it is possible to change the lobe separation angle since it is possible to adjust the timing of each camshaft separately.

while lowering it at low engine speeds. Figure 3. Obviously.Since the engine being studied has a single camshaft. Advancing does the opposite. the highest volumetric efficiency at all engine speeds is desired.13. The curves are the differences between the volumetric efficiencies in Figure 3. Substantial changes in the volumetric efficiency can be attained by this. Since maximizing it at one speed hurts it at other engine speeds. and thus that lies along the x-axis. 1.2 1 0. as can be seen by Figure 3.13 Volumetric efficiency versus engine speed as the camshaft timing is changed at 55:1 AFR.6 1. the curve with the maximum area under the most often used engine speeds is the desired curve. There is no change from the stock curve. only camshaft timing was altered. Retarding the camshaft increases the volumetric efficiency at higher engine speeds.13 and the stock curve on that figure.8 0.14 shows the change in the volumetric efficiency as the camshaft timing is changed. 51 .6 1000 10 deg retard 5 deg retard Stock 5 deg Advance 10 deg advance Volumetric Efficiency 2000 3000 4000 5000 Engine Speed (rpm) Figure 3.4 1.

15 shows the NO output at WOT in ppm. some simulations were done at a throttle opening of 0. instead of the 1. NO Emissions were also looked at while changing the camshaft timing. It is at these speeds that the supercharger is increasing the volumetric efficiency dramatically as compared to the naturally aspirated engine.05 -0. while helping it above that speed. street driven engines typically operate under 3500 rpm and part throttle most of the time. Since the changes are very small.16 for the part throttle case. the engine operates at a much reduced volumetric efficiency than wide open throttle accelerations. Thus the maximizing volumetric efficiency is the deciding factor in choosing the camshaft timing.8 inches that were used for the WOT.14. It can be seem that camshaft timing does alter the NO output.9 inches. and Figure 3.18 show the change in the NO output when compared to the stock timing.17 and 3.0. but not significantly. Figure 3. When a vehicle is cruising. Normal. 52 .025 2000 3000 4000 5000 10 deg retard 5 deg retard Stock 5 deg Advance 10 deg advance -0.075 -0. The sensor was placed in the exhaust manifold collector.05 Change in Volumetric Efficiency 0. Retarding the camshaft hurts the volumetric efficiency at those speeds. Since part throttle operation is more expressive of the engine operation. The change in volumetric efficiency from the values attained with the stock timing at WOT at 55:1 AFR. Figures 3.025 0 1000 -0.1 Engine Speed (rpm) Figure 3.

15 NO emissions as the camshaft timing is changed.16 NO emissions as the camshaft timing are changed for part throttle conditions at 55:1 AFR. 53 . 30 NO emissions (ppm) 25 10 deg retard 20 15 10 5 0 1000 5 deg retard stock 5 deg advance 10 deg advance 2000 3000 4000 5000 Engine Speed (rpm) Figure 3. WOT.45 40 NO emissions (ppm) 35 30 25 20 15 10 5 0 1000 10 deg advance 5 deg advance stock 5 deg retard 10 deg retard 2000 3000 4000 5000 Engine Speed (rpm) Figure 3. AFR 55:1.

1. Change in NO E m issions (ppm ) 1 0.5 -1 Figure 3.18 Changes in the NO emissions as camshaft timing is changed for the part throttle case at 55:1 AFR.5 0 1000 -0. 54 .5 -1 Engine Speed (rpm) Figure 3.17 Changes in NO emissions (in ppm) as the camshaft timing is changed for the WOT case at 55:1 AFR.5 10 deg retard 5 deg retard stock 2000 3000 4000 5000 5 deg advance 10 deg advance 0 1000 -0.5 1 10 deg advance 5 deg advance stock 5 deg retard 2000 3000 4000 5000 10 deg retard 0.

Retarding the camshaft five degrees will increase the fuel economy.15 0. When the throttle plate is completely open.21. but retarding it ten degrees is already too much.19 BSFC as the camshaft timing is changed for the part throttle case at 55:1 AFR. Since the data lines are fairly close to each other.20 shows the changes in it. 0. it was also looked at during the part throttle operation. The brake specific fuel consumption (BSFC) is shown in Figure 3.13 10 deg advanced 5 deg advanced stock 10 deg retarded 5 deg retarded 0.19 shows the BSFC for the WOT case. it is possible to see the increased pumping losses as the throttle is closed.20.20 clearly shows that advancing the camshaft increases the fuel consumption at part throttle.12 0. This is a different from getting the maximum power. which would suggest either stock or slightly advanced camshaft timing. and 3. The best camshaft timing from a fuel economy point of view is 5 degrees retarded. the change in the BSFC was also graphed in Figure 3. Figure 3.09 1000 2000 3000 4000 5000 Engine Speed (rpm) Figure 3. At ten degrees of retard.1 0.19 and 3. Figure 3.19.11 0.14 BSFC (kg/kW hr) 0. the fuel consumption starts coming back up to the stock levels. comparing Figures 3. As a side note.As part throttle fuel economy is an important part of engine design. The same basic results are shown for the WOT case as were for the part throttle case. it is 55 .

or BSFC. 56 . 8 6 Percent Change in BSFC 4 10 deg advanced 5 deg advanced stock 10 deg retarded 5 deg retarded 2000 3000 4000 5000 2 0 1000 -2 -4 Engine Speed (rpm) Figure 3.easy for the pistons to pull in the air into the cylinder. even though less air gets ingested. As it is closed. more power goes into pumping the air into the cylinder.20 Change in the BSFC as the camshaft timing is changed for the part throttle case at 55:1 AFR. This increases the amount of fuel that gets burned for a given power output.

21 The BSFC as camshaft timing is changed for the WOT case at 55:1 AFR.13 0.22 Percent change in the BSFC as the camshaft timing is changed for the WOT case at 55:1 AFR.0. 8 7 Percent Change in BSFC 6 5 4 3 2 1 0 1000 -1 -2 Engine Speed (rpm) 10 deg advance 5 deg advance stock 10 deg retard 5 deg retard 2000 3000 4000 5000 Figure 3.09 0.11 10 deg advance 5 deg advance stock 10 deg retard 0.08 1000 2000 3000 4000 5000 Engine Speed Figure 3. 57 .12 BSFC (kg/kW hr) 0.1 5 deg retard 0.

This gives a second way to cut back the power output of the engine. the engine produces negative torque. thus decreasing the power output. less fuel goes into the engine. especially liquid fuels. Constant power output curves are plotted in Figure 3. up to 60 kW for the short. The curve in the bottom. left hand corner is for no power output (all the power is used to run the engine). for 3000 rpm engine speed.Throttling a Hydrogen Engine The power output of most spark ignition engines is cut back to the levels needed by the vehicle operation by restricting the airflow into the engine. as the throttle opening and equivalence ratio are changed. It is possible to get the needed power while lowering the fuel consumption by opening the throttle blade more thus decreasing the pumping losses. This is not the case with hydrogen since it can ignite at very lean equivalence ratios.23. or requires power to run. right hand corner. Below this curve. almost horizontal curve in the top. Figure 3. needs to operate very close to the stoichiometric air fuel ratio. Each of the curves above the first curve represents an increase of 5 kW. 58 . By lowering the equivalence ratio.23 Brake power (kW) as the throttle opening and equivalence ratio are changed for the supercharged hydrogen engine running at 3000 rpm. An engine running on most fuels. and keeping the airflow constant.

8 equivalence ratio. The equivalence is the larger restriction at these operating parameters.24 as the throttle opening and equivalence ratio are changed. The necessary power for a given vehicle speed changes with driving conditions.23 and 3. These are points which consume the least amount of fuel for a given power output. A different fuel consumption is obtained for all different operating points for a given power output. and the importance of the throttle opening increases. The brake specific fuel consumption (BSFC) is shown in Figure 3. This means that the equivalence ratio is not as large a factor in determining power output. The lowest BSFC is at WOT and approximately 0. The figure does not demonstrate that the fuel consumption rises extremely quickly in the red area.. the curves start pointing upward. 59 .24. it is possible to come up with ideal operating points for the engine. Thus it might be possible to operate a hydrogen engine without a conventional throttle. the BSFC is infinite for the bottom left corner. This means that the engine power is almost completely dependent on the equivalence ratio at the large throttle openings. Comparing both Figures 3. where the engine makes either negative or no torque. In fact. the curves are almost horizontal. If the transmission does not change gears. For the smaller throttle openings. there is very little change in the BSFC in that area. the importance of throttle opening and equivalence ratio are equally important. the torque output from the engine must change with the driving conditions in order to maintain constant vehicle speed. For medium throttle openings (around 0. The curve is very flat for the blue colors. using equivalence ratio as the only way to regulate torque output. For large throttle openings. It turn out that the least fuel is consumed at WOT.From the curves it is possible to see that there are an unlimited number of different combinations of throttle openings and equivalence ratios to get a specific power output.75 inches).

The reason for mounting the supercharger after the throttle is that it is under much lower load while the vehicle is cruising. When the supercharger is placed after the throttle. 60 . a recirculation valve is not as critical for maximum efficiency the compressor spins in a vacuum. However. Rotrex recommends at least using a re-circulation valve if the throttle is placed after the supercharger. and the throttle is open very little. even though virtually all of them mount the throttle before the supercharger. the supercharger is still making boost that the throttle is preventing from reaching the intake manifold. This valve is operated by manifold vacuum. The BSFC for this case is shown in Figure 3.24 Brake specific fuel consumption (kg/kW Hr) as the throttle opening and equivalence ratio are changed for the supercharger hydrogen engine running at 3000 rpm. thus lowering the fuel consumption.Figure 3. which opens it to let air from the high pressure side to the low pressure side of the supercharger. Since Rotrex recommended that the supercharger be mounted after the throttle. When the supercharger is mounted before the throttle. most original equipment superchargers do have these.25. this setup was also simulated.

The simulation results are most likely not completely accurate for the supercharger efficiency.26 shows the NO emissions in the tailpipe in ppm.25 The BSFC for the supercharged hydrogen engine with the throttle mounted before the supercharger as the throttle and the equivalence ratio are changed at 3000 rpm. From Figures 3. It can be seen that the NO emissions are almost completely a function of the equivalence ratio. and it is better to mount the supercharger after the throttle. Thus. Figure 3. with the throttle opened further. from the emissions perspective it is also better to run the engine leaner. 61 . The emissions were considered when throttling the engine and the NO emissions were compared to data published by Ford Motor Company. These simulation results are for the case where the throttle is mounted after the supercharger. Emissions Since hydrogen has no carbon atoms. WAVE is unable to simulating any other NOx than NO.24 and 3. so it is the only one that was studied and compared to available data. as the throttle opening and equivalence ratio are changed.25 it appears that it would be better to mount the throttle after the throttle.Figure 3. Most likely the airflow or the pressure ratio that WAVE calculates across the supercharger is smaller than in the actual engine. the main pollutant from a hydrogen engine is NOx.

5. The bottom line is 4 ppm.26.5:1.28.The Ford Motor company has studied the NOx emissions in a hydrogen engine (Stockhausen). 14.38. Their study shows that concentration increases dramatically when the engine is run richer. There is very little NOx made below 0.5 equivalence ratio. Running a real engine at that equivalence ratio is possible.27. as the NOx emissions are strongly a function of equivalence ratio. pre-ignition problems increase dramatically. This graph also supports the results shown in Figure 3. Similar trends were found with the simulations.5:1. They found that the NOx emissions were very small (less than 10ppm) at equivalence ratios less than 0. and less than 100ppm at equivalence ratios less than 0.3:1). but it as the mixture is richened the NOx increases. Figure 3. In order to minimize NOx Emissions. The graph includes data from their test engine ran at three different compression ratios (12. and as the air fuel mixture is richened from this at higher loads. the engine should not be ran richer than about 0. and 15. The results from Ford are shown in Figure 3. and increases by 4 ppm for each line above it.6 equivalence ratio. The points were taken at various engine loads and speeds. The NOx emissions for the supercharged model are shown in Figure 3. 62 .26 NO emissions (ppm) in the tailpipe as the throttle opening and equivalence ratio are changed for the supercharger hydrogen engine running at 3000 rpm.

27 NOx emissions as a function of equivalence ratio for Ford test engines running at various engine speeds.3 0.Figure 3.5 0.9 1 Equivalence Ratio Figure 3.4 0.7 0.8 0. as the equivalence ratio is changed. 63 .28 NOx emissions for the supercharged hydrogen engine at 3000 RPM and WOT.2 0. Equivalence Ratio 12000 NOx Emissions (ppm) 10000 8000 6000 4000 2000 0 0. loads and compression ratios.6 0. Nox Emissions v.

64 . exhaust manifold.75 2. the pressure in the intake plenum.5 1. gasoline model.29 Mean pressures in intake manifold plenum. pressure losses through the open throttle plate and intake plumbing are increased. This decrease is very small. Ambient pressure is 1 Bar.25 1 0. for the supercharged hydrogen model running at 55:1 AFR and the stoichiometric. The mean pressures are shown in Figure 3. Significant increases in the exhaust pressures can also be observed. This figure also shows the pressure in the intake plenum for the naturally aspirated gasoline model. the pressure in the intake increases as the engine speed increases. For the gasoline model. at low engine speeds is very close to the ambient pressure (1 Bar). This is because of the centrifugal supercharger having an increased outlet pressure with higher speeds. naturally aspirated. N/A Exhaust N/A Catalytic Conv.75 1.29.Pressure Measurement in the Intake and Exhaust Systems Pressure sensors were placed in the intake manifold plenum. N/A Intake Figure 3. exhaust manifold collector. and just behind the catalytic converter on the supercharged hydrogen model.25 2 1.5 Absolute Pressure (Bar) 2. and after the catalytic converter brick.75 1000 2000 3000 4000 5000 Engine Speed (rpm) S/C Intake S/C Exhaust S/C Catalytic Conv. but it does show that as the airflow is increased. 2. On the supercharged model. It can be seen that the pressure decreases as engine speed increases.

The mass flow through the exhaust system was also noted for the models. They are shown in Figure 3.30. The mass flows were measured near the end of the tailpipe, where only vapor passed through the meter. The mass flow rate through the engine should be mostly a function of engine speed and volumetric efficiency.
0.2

Mass Flow through Engine (kg/s)

0.15

0.1

S/C Hydrogen N/A Gasoline

0.05

0 1000

2000

3000

4000

5000

Engine Speed (rpm)

Figure 3.30 Mass flow through the engine for the supercharged, hydrogen engine (55:1 AFR) and naturally aspirated, stoichiometric, gasoline engine. The mass flow rates along with the volumetric efficiencies explain why the pressure curves in Figure 3.29 level off for the gasoline model, but not for the hydrogen model. For the naturally aspirated gasoline model, the volumetric efficiency starts falling after about 4000 rpm. It is because of this that the mass flow rates level off for the higher engine speeds. For the supercharged, hydrogen model, the mass flow rates just keep climbing as the volumetric efficiency and engine speed increase. The significant back pressure in the exhaust system for the hydrogen engine means that the engine needs a bigger exhaust system. This was done by increasing the exhaust pipe diameter from 1.875 inches to 2.5 inches, for a cross-sectional area increase of almost 80%. The catalytic brick size was doubled by increasing the number of passages from 1500 to 3000. 65

Increasing the exhaust size decreased the back pressure 10-20% above 3500 rpm. This did increase the volumetric efficiency, but only by 1-3%. Under 3500 rpm, where the engine operates most of the time, there were no significant changes in the back pressure or volumetric efficiency. A down side to increasing the size of the exhaust tubing is increased noise. Thus, the smaller exhaust would be a better choice for the engine.

Conclusion and Future Research This study was done in order to show general trends as variables are changed for a hydrogen engine, in order to recommend changes in the engine. The effects of camshaft timing, compression ratio, equivalence ratio were shown in the volumetric efficiency, power, and emissions. Recommendations for improving the power, fuel efficiency, and emissions for Texas Tech’s FutureTruck explorer are discussed below. Probably the biggest, and most difficult, improvement would be to chance the throttle plate to a fly-by-wire throttle. This would allow the throttle to be opened more while cruising in order to lower the fuel consumption. Controlling this type of a throttle would not be simple. These systems are relatively rare in stock vehicles, but are becoming more common, much like the wideband oxygen sensors. A stock controller would most likely not be programmable, and would not allow leaning out the equivalence ratio as the throttle plate is opened. Thus, a new engine controller would need to be built from scratch. Leaving out the throttle plate all together might be possible with the use of hydrogen. The power output could be controlled by regulating the equivalence ratio. It is unclear if an engine would run at low speeds at very lean equivalence ratios. Controlling this would not be difficult, and the current ECU could be used. The compression ratio of 12.6:1 seems to give approximately the highest fuel efficiency, and thus would not need to be changed. If the engine did not have a supercharger or a turbo, a higher compression ratio should be used. The highest fuel

66

efficiency is given by approximately 14.5:1 to 15:1 compression ratio for a naturally aspirated model, and approximately 12.5:1 to 13:1 for the supercharged model. The camshaft advance of 8 degrees is a little bit too much. While this does increase power below 2000 rpm, it decreases power above 2500 rpm. If the engine is operated above 3500 rpm, the stock timing is more suitable. Since the engine would be used at about 1500 to 2500 most of the time, camshaft advance should be approximately 5 degrees. If maximum fuel economy is wanted, the camshaft should be advanced as little as possible. Five degrees of camshaft retard provided the best fuel economy. As the supercharger makes very little boost at low engine speeds, the ratio of the pulleys could be changed. Currently the pulley sizes are set to give maximum rotational speeds of the supercharger at engine speed of 5000 rpm. By decreasing the maximum engine speed to 4500 rpm, more torque could be made at the low engine speeds. Changing to a positive displacement supercharger would help the low boost levels at low engine speeds. These superchargers in general are much more difficult to mount to stock intake manifolds, but in a production vehicle this should not be an issue. These simulations have shown that the specifications of the hydrogen engine used in Texas Tech’s FutureTruck does not need any major changes in order to optimize its operation. Most likely the recommended changes will not be made to this engine, as the FutureTruck competition ended in 2004. However, these results can be use in the FutureTruck replacement Challenge X. Challenge X is going to use a hydrogen-ethanol engine, which will be have very similar low engine load characteristics as a pure hydrogen engine. The engine will be ran an either mostly or completely on hydrogen at light loads, in order to minimize emissions. At higher loads and engine speeds, when hydrogen is unable to produce power, the engine will use mostly ethanol. Controlling and tuning this engine will present new challenges, but the experience gained from the FutureTruck will help overcome these obstacles.

67

New York.popularhotrodding. 24. Feb.turboford.” Energy technology Training Center College of the Desert.org/faq/ta.com/tech/0407phr_bosch/. D. 2002 Hydrogen Heywood.au/vehicle/wbo2/wbntk. Advanced Vehicle Testing Activity.cosworthtechnology. “Compilation of Existing Safety Data on Hydrogen and Comparable Fuels. Yimin Gao. FreedomCAR & Vehicle Technologies Program. 2001. R.. Hybrid Electric. “Module 2: Hydrogen Use.” http://www. 2003. Stockhausen. 8.org/public/documents/CompilationExistingSafetyData_on_H2_a nd_ComparativeFuels_S. http://www. Arizone Public Service – Alternative Fuel (Hydrogen) Pilot Plant Design Report. 68 . turboford. 2005.” http://techedge. 14. Kabat. Endres. Hunkins. Clay. J. Karner. X. Gay. Gargano. Dempsey. Jan. Larry. Heffel.eihp. Peter..htm. New York.. U. Ford P2000 Engine Dynamometer Development..uk/100_engineering/106_alt_fuels.” http://www.co. 2005. Department of Energy. Cosworth Technology. Ehsani. 2003. “NTK L1H1 Sensor Information.” Popular Hotrodding. James Francfort.com. Ali Emadi. Dec. W. J.com/simtech. 8. J.shtml. Modern Electric.motionsoftware. 2005.BIBLIOGRAPHY Alcock. Jan. and Fuel Cell Vehicles. HPBooks. Atherton. Feb. Internal Combustion Engine Fundamentals. 1988. John B. Mehrdad. “The Lowly Oxygen Sensor. Tang. Don. Johnny. http://www. 2005. 2005.org http://www. Sept.. Christopher P.pdf Jan. Theory..htm.htm. 20. Autery. Chevy LS1/LS6 Performance. Natkin. INEEL/EXT-0300976. 16. Sebastien E. “Alternative Fuels. Fundamentals. and Design. New York: McGraw-Hill. 2004. SAE. 2005.S.L. More power and clean living through science – and Bosch. Boca Raton: CRC Press.

com/. 69 . Nov. Stephen R. Ricardo Software. Sapru. Rotrex. “MoTeC M4 Specifications. William F. “Planar wide band O2/Lambda sensor. A Comprehensive Guide to WAVE Practices & Techniques.” http://toyotaperformance.com/products/ecu/m4ecu. Boston: McGraw-Hill. “Basic EFI Theory.merkurencyclopedia.htm.html.” 2004 Ricardo Software. 29. Corvette Fuel Injection & Electronic Engine Management 1982 Through 2001: How to Understand. 2000. Service and Modify. Turns. Sept. 2005. 24. DOE Hydrogen Program Review. 2004.S. et al.” http://www.airfuelmeter. K. 2004 Stockhausen.” http://www. Feb 16. “Esslinger Aluminum Head.. Proceeding of the 2002 U.Micro-Craft. 2nd Edition. MA: Bentley Publishers. “Hybrid Electric Vehicles” Scientific American.com/english/lsu4_en_sensor. 8. Probst. Concepts and Applications.com/basic. Ford P2000 Hydrogen Engine Design and Vehicle Development Program.” http://www. Victor. Oct.com/Motor/ESSLINGER_Head.rotrex.motec. Hydrogen Internal Combustion Engine Two Wheeler with on-board Metal Hydride Storage. Motec. Wouk. Allen. 8. 2005. Feb.. NREL/CP-610-32405 Slocum. SAE 2002-01-0240 TCR Automotive and Performance. “WAVE Knowledge Center. http://www. An Introduction to Combustion. Feb. Cambridge.htm.htm. et al. 1997. Charles O. 2001. 2005.

Agree (Permission is granted. Permission to copy this thesis for scholarly purposes may be granted by the Director of the Library or my major professor. I agree that the Library and my major department shall make it freely available for research purposes.) _______________________________________________ Student Signature _________________ Date 70 . 2005 ________________ Date Disagree (Permission is not granted.) Jaakko Halmari ________________________________________________ Student Signature April 17. It is understood that any copying or publication of this thesis for financial gain shall not be allowed without my further written permission and that any user may be liable for copyright infringement.PERMISSION TO COPY In presenting this thesis in partial fulfillment of the requirements for a master’s degree at Texas Tech University or Texas Tech University Health Sciences Center.

Sign up to vote on this title
UsefulNot useful