Study of Low-Temperature-Combustion Diesel Engines
As an On-Board Reformer for Intermediate Temperature
Solid Oxide Fuel Cell Vehicles
by
Tairin Hahn
Bachelor of Science in Mechanical and Aerospace Engineering
Seoul National University, Republic of Korea, 2003
Submitted to the Department of Mechanical Engineering
in Partial Fulfillment of the Requirements for the Degree of
Master of Science in Mechanical Engineering
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
JN
Ae L 2006
0 2006 Massachusetts Institute of Technology. All Rights Reserved
Signature of Author:
Department of Mechanical Engineering
May 02, 2006
Certified by:
Wai K. Cheng
AQ
Professor of Mechanical Engineering
Thesis Supervisor
Accepted by:
Lallit Anand
MASSACHUSES INS
E
Chairman, Department Committee on Graduate Students
OF TECHNOLOGY
JUL 1 4 2006
LIBRARIES
BARKER
2
To
Z
3
g
atherand
mother
4
Study of Low-Temperature-Combustion Diesel Engines
As an On-Board Reformer for Intermediate Temperature
Solid Oxide Fuel Cell Vehicles
by
Tairin Hahn
Submitted to the Department of Mechanical Engineering
on MAY 02, 2006 in Partial Fulfillment of the Requirements
for the Degree of Master of Science in Mechanical Engineering
ABSTRACT
Fuel cells have been recognized as a feasible alternative to current IC engines. A significant
technical problem yet to be resolved is the on bound fuel supply before fuel cells can be
practically used for vehicles.
Use of an on-board fuel reformer can mitigate the fuel supply issue. In particular, using a
diesel engine as on-board fuel reformer, combined with a fuel cell, is a strong candidate for
the next generation power plant for vehicles. This study investigates feasibility of using a
diesel engine as a fuel reformer. To supply the hydrogen and carbon monoxide for Solid
Oxide Fuel Cell (SOFC), a primary Low-Temperature-Combustion (LTC) technology diesel
engine using will burn fuel-rich mixture to provide the high temperature combustion products necessary for SOFC. This study models the ignition properties of the fuel-rich mixture,
and then applies the model to estimate operation map of a LTC diesel engine. This research
provides useful design for the combined cycle prove plant.
Thesis Supervisor: Wai K. Cheng
Title: Professor of Mechanical Engineering
5
6
ACKNOWLEDGEMENTS
First of all, I would like to thank Professor Wai K. Cheng for the best guidance that any
graduate student would expect from an advisor. During the numerous short meetings, he
has not only taught me about researches but also showed me a way to think when problems
came on me. I am looking forward to continuing my researches and experiments with him
over the next few years.
I am grateful to many other people in the Sloan Automotive Laboratory for their supports
over the past one year. However, there have been numerous of people who deserve my special thanks. When things went wrong, I used to run to Morgan as well as Tony, our alumni.
Thanks for Morgan's patient and help. For me, because the Sloan Lab has been always a
friendly, amusing, and cozy place, I could enjoy my research more. I appreciate my dear
office mates: Steve, Francois, Yong, and another Yong (Jeff). Generous tips of Michael and
Dong-kun on researching and surviving in Sloan Lab are appreciated, too.
Beside the Sloan Lab, I would like to express my deep gratitude to Leslie for her immense
practical help. I also would like to thank Dr.Hu of Eaton Co.; my project sponsor, and John
of Numerica Technology. I would like to thank my S&P girls who made MIT as an awesome place to live and work. Many thanks to KGSA members, especially the KGSAME
members.
Last but not least, I would like to give my gratitude to my praiseworthy father, Gi-Chuk
Han, who has been always there for me, and my mother, So-Youn Park, the best and the
one in the world to me. Also, I appreciate my precious sister, Serin, and bother in law, Seungwon. My gratitude to my family is beyond description. Without them, I couldn't be here
to follow my dreams.
Tairin Hahn
Cambridge, MA, May, 2006
7
8
TABLE OF CONTENTS
A B ST RA C T ...........................................................................................................
5
ACKNOWLEDGEMENTS...................................................................................7
TABLE OF CONTENTS.....................................................................................9
LIST OF FIGURES AND TABLES.........................................................................11
NOMENCLENTURE.......................................................................................15
CHATER 1: INTRIDUCTION..............................................................................17
1.1 Motivation for the Present Research...........................................17
1.2 Backgrounds of the Research......................................................19
21
1.3 Objectives of the Research........................................................
23
CHATER 2: RICH COMBUSTION AT LOW TEMPERATURES ...........
23
2.1 Introduction.................................................................................
2.2 Analysis of Chemical Kinetic Mechanisms.................................26
2.3 Modification of Chemical Kinetic Mechanisms..........................30
CHATER 3: IGNITION DELAY STUDY...............................................................
3.1 Introduction.................................................................................
3.2 Simulation Model........................................................................
3.3 Stationary Ignition Delay Results...............................................
39
39
42
44
CHATER 4: ENGINE OPERATION WINDOW STUDY.................51
51
4 .1 Introduction ...................................................................................
52
4.2 Simulation Model Setup for LTC Engines..................................
59
4.3 Engine Operation Maps and Results...........................................
9
CHATER 5: SUMMERY AND CONCLUSIONS..................................................67
5.1 Overview ......................................................................................
5.2 Fuel-Rich Combustion..................................................................
5.3 Ignition D elay in Engines..............................................................
5.4 Conclusions..................................................................................
67
67
69
70
REFEREN CES.....................................................................................................
71
10
LIST OF FIGURES AND TABLES
FIGURES
CHATER 1
Figure 1.1 Schematic of power generating reformer and ITSOFC powerplant.............19
CHATER 2
Figure 2.1 Main hydrocarbon oxidation pathways.....................................................26
Figure 2.2 Difference between experimental data and predictions from the original LLNL
32
kinetic m odel for n-butane.....................................................................
Figure 2.3 Schematic of the applied modification....................................................35
Figure 2.4 Agreement after modification in our present study....................................36
CHATER 3
Figure 3.1 Definition of ignition delay with NTC behavior............................................40
Figure 3.2 Agreements of simulation results from the different batch reactor model according to the various pressure where T=900K, Xr=O.15, and phi= 2.5..............43
Figure 3.3 Effect of equivalence ratio and pressure on ignition delay under T= 700K and
X r = 0.3........................................................................................................---44
Figure 3.4 Effect of pressure and initial temperature on ignition delay under phi=1.5 and
45
Xr = 0 .3 .................................................................................................................
Figure 3.5 Changing the NTC behavior region related to a fuel-equivalence ratio as a function of initial temperature at P=40atm and Xr =0.3......................................45
Figure 3.6 Effect of residual gas fraction on ignition delay as a function of initial tempera46
ture at P=40atm and phi=1.5 .........................................................................
Figure 3.7 Effect of residual gas fraction on ignition delay at P=40atm........................47
Figure 3.8 Effect of residual gas fraction on ignition delay at phi = 2.0........................48
11
CHATER 4
Figure 4.1 Schematic of the system configuration.........................................................
Figure 4.2 Convergence of solution after cycle-by-cycle variation...............................
53
54
Figure 4.3 Convergence of solutions in terms of ignition delay [ms] and ignition timing
[d eg] ..................................................................................................................
55
Figure 4.4 Convergence of solutions in terms of pressure, temperature, and gas compositions in engines............................................................................................
55
Figure 4.5 Sensitivity of ignition delay with bum duration ...........................................
57
Figure 4.6 Heat release rate profile of the engine.........................................................
58
Figure 4.7 Engine operation window: data points with 10 % EGR..............................
59
Figure 4.8 Ignition points as function of EGR at different phi.....................................
Figure 4.9 Effect of equivalence ratio on ignition timing..............................................
60
Figure 4.10 Temperature Profile in the Cylinder............................................................
Figure 4.11 Engine operation map..................................................................................62
61
61
Figure 4.12 Indicated mean effective pressure, indicated specific fuel consumption, and
indicated thermal efficiency of an engine as a function of fuel equivalence ratio ......................................................................................................................
63
Figure 4.13 Speed effects on the break thermal efficiency.............................................
64
Figure 4.14 Exhaust gas temperatures under 1500rpm ..................................................
65
12
TABLE
CHATER 2
Table 2.1 Characters of rich hydrocarbon chemical reaction, where RH, A, and B is hydrocarbon species, X and X' is oxidizer, and M is 3 rd body.................................24
13
14
NOMENCLENTURE
ACRONYMS
ATDC
After Top Dead Center
CAD
Crank Angle Degree
CI
Compression Ignition
EGR
External Gas Recirculation
HCCI
Homogenous Charge Compression Ignition
IC
Internal Combustion
ID
Ignition Delay
IMEP
Indicated Specific Fuel Consumption
ISFC
Indicated Mean Effective Pressure
ITSOFC
Intermediate Temperature Solid Oxide Fuel Cell
IVC
Intake Valve Closing
LTC
Low Temperature Combustion
NTC
Negative Temperature Coefficient
PM
Particulate Matters
RPM
Revolutions Per Minute
SI
Spark Ignition
Xr
Residual Gas Fraction
SYMBOLS
Indicated thermal efficiency
(D
Fuel air equivalence ratio, phi
15
16
Chapter 1
INTRIDUCTION
1.1 Motivation for the Present Research
For more than a hundred years, fossil fuels have been the most significant energy
source for transportation. However, fossil fuels have two major problems. One is the fossil
fuel shortage, which we will encounter in a few decades. The other is emissions from fossil
fuels such as NOx, CO, C0 2 , and Particulate Matters (PM). Because of fuel economy and
low emission of C0 2 , diesel engines have been considered as the most efficient internal
combustion engines in common use [1]. Even though there have been remarkable improvements in the conventional internal combustion (IC) engine, new power generators and
energy sources are required upon the fossil fuel shortage problem and the environmental
issue.
Fuel cells have been recognized as the most feasible alternative to current IC engines.
They produce electrical energy directly from chemical reactions. If hydrogen is used for
fuel cells, water is the only by-product from the overall power generation process. Current
17
thermodynamic efficiency of fuel cells is about 60%, based on the heating value of the hydrogen supplied [2]. Many researchers believe that IC engines will eventually be replaced
with fuel cells in a few decades.
Nevertheless, there are still various technical problems in fuel cells applications. For
example, a power to weight (or volume) ratio of fuel cells is not as high as that of the IC
engines. The substantially high price of the hydrogen gas and lack of hydrogen-stations also
limit their early adaptation. For the above reasons, on-board fuel reformer using liquid fuels
is attractive. First, the fuel reformer does not need high pressurized storages of gaseous fuel,
hence the energy needed for compression is avoided [3]. Much work is necessary to put hydrogen, which is the lightest gas, into a pressurized tank, and the fuel storage means causes
a low power to weight ratio of the overall fuel cell system. As a result, the overall system
efficiency increases if automobiles use a liquid fuel such as gasoline, methanol, or methane
and convert these fuels into the feed stream appropriate for the fuel cell. Second, the fuel
reformers allow the use of conventional automotive fuels in a vehicle [3]. Therefore, the
logistic problem of supplying the fuel is alleviated.
18
Fuel flow
F
Exhaust "
enthalpy F
flow
Gearbox
and motor
F
L
Figure 1.2 Schematic of power generating reformer and ITSOFC powerplant from 141
1.2 Background of the Research
The power plants combining an intermediate temperature solid oxide fuel cell
(ITSOFC) with an IC engine is a strongly candidate of the next generation power plant for
vehicles. Figure 1.2 describes the schematic of this system. The IC engine acts as a power
generating reformer, and the mechanical output of the system is provided both by the engine and by an electric motor driven by the fuel cell.
The proposed configuration will substantially improved the efficiency of the fuel reforming process. Since the bestial oxidation of the fuel is CO and H2, and the main gas shift
reaction of converting CO and H20 to CO 2 and H2 are both exothermic, the released energy
19
is usually wasted. The engine as a reformer would recover this energy as useful work.
Moreover, the ITSOFC has higher energy conversion efficiency (45% to 65%) than a conventional IC engine (33% to 37%). Therefore, the overall powertrain energy conversion efficiency of this system will be higher than that of an IC engine.
To maximize the overall efficiency of this system and to minimize NOx and PM emissions, which adversely affect the latter fuel cell operation, the engine will operate in the
fuel-rich LTC mode using compression ignition of the diesel fuel. Because the high fuel
conversion efficiency of a diesel engine originates from the essential characters of its fuellean operable cycles [1], many diesel engine researches have been focused on only fuellean operating regions. However, many essential concepts have not been adequately addressed under the fuel rich condition. Therefore, evaluating basic combustion behavior and
operable condition of the LTC diesel engine as a power generating reformer is necessary.
20
1.3 Objectives of the Research
This project aims to theoretically assess two main points: the ignitability of a fuel rich,
heavily diluted mixture; and the operating parameters for a power-generating reformer using a LTC diesel engine. The research steps can be summarized as follows:
0
Using chemical kinetics calculations of n-heptane chemistry [5], the ignition delay of the fuel-air mixture is tabulated as a function of a fuel-air equivalence ratio(CD) under the rich condition, temperature (T), pressure (P), and residual gas fraction (Xr).
*
The tabulated ignition delay is incorporated into a cycle simulation software,
WAVE [6], to define the LTC operating region.
The scope of the study is limited to steady state operation at each load speed point;
dynamic issues in the load/speed trajectory of the driving cycle are left for future study.
Furthermore, it is assumed that the powerplant is a stand-alone unit without any energy
storage device, and the overall driving cycle optimization via energy storages is not considered. Despite significant uncertainty in the real LTC fuel rich diesel engine's operating
characteristics, it is still useful to establish the operating characteristics for the LTC engine /
21
ITSOFC combined powerplant based on the theoretical assessment of the LTC engine.
Therefore, the output of this research will be used as guidance for further development of the concept in terms of testing hardware and obtaining real data for product optimization.
22
Chapter 2
RICH COMBUSTION
AT LOW TEMPERATURES (650-1000K)
2.1 Introduction
We consider the combustion of a homogenous mixture at a rich-fuel equivalence ratio
(between 1.5 and 3), at low temperatures (between 650 K and 1000 K), and at high pressures (between 20 atm and 60 atm). Ignition delay is used to access the ignitability of the
charge. However, the hydrocarbon ignition chemistry has not been well established for fuelrich condition. Even the commonly used detailed n-heptane mechanism developed by Lawrence Livermore National Laboratory (LLNL) is established only within an equivalence
ratio between 0.3 and 1.5 [5]. Thus, we need to adjust this mechanism to use it for our study.
To modify this mechanism, we must understand the basic chemical reaction mechanism for
hydrocarbon ignition. In particular, because of negative temperature coefficient (NTC) behavior, ignition delay at low temperatures (T < 900 K) must be carefully defined [8].
Briefly, the reaction mechanism depends on temperature and pressure due to the differ-
23
ent path of oxidation. Oxidation generally occurs through three steps; initiation, propagation, and termination. Table 2.1 shows a simple expression for hydrocarbon fuel reaction
mechanism. Two main different pathways are selected by initial temperature of the mixture.
At high T, the reaction rate of hydrocarbon with H radical is much faster than reaction rate
of H- + 02->O- + OH. Reactions that produce H radical are most important reaction because these reactions dominate the over all reaction rate involving chain reactions. The
presence of hydrocarbon provides an effective competitor to oxygen for available H atoms,
reducing chain branching rate as well as slowing the overall combustion rate. Therefore,
At high Temperature (>1000K)
initiation
RH -+ A+B
At low Temperature (r 10001K)
RH + X -+ R + X'
with elevated high P ( >20 atm)
main
reaction
H- + 0 2 -O+
Reaction rate of
A
relation
-+ OH
H- +0
2 +M--*HO 2+
M
Followed by
of B+ H- - R +H20 2
H0,+H
much faster
inhibit the iydrogen oxidation
RH +HO2
2 -+H
H202
-
+
2+
02
R + H202
M - OH + OH + M
Richer mixture reacts slower. Richer mixture reacts faster.
with cD
Table 2.1 Characters of Rich Hydrocarbon Chemical Reaction, where RH, A, and B is hydrocarbon
species, X and X' is oxidizer, and M is 3rd body.
24
little increase hydrocarbon amount can inhibit the hydrogen oxidation. In particular, because the fuel does not react as rapidly or completely under fuel rich condition, fuel consumption reaction plays a larger role than for lean or stoichiometric. Therefore, we need to
carefully modify mechanism under rich region.
25
2.2 Analysis of Chemical Kinetic Mechanisms
Hydrocarbon ignition is governed mainly by chain initiation, propagation, reaction, and
degenerated branching which leads to rapid multiplication of radicals. Like other chemical
reactions, these processes generally show strong dependence on temperature. The overall
hydrocarbon oxidation pathways, having more than three carbon atoms, can be described
schematically as shown in Figure 2.1 [9].
The hydrocarbon oxidation rate is dominated by the following reaction, regardless of
temperature:
H + 02 = 0+
OH
Still, there are locally governing reactions when we consider a specific temperature. At
decomposition
I
oxygenated intermediate + OH
f -scission
at high T
02
RH-> R
OH,
H02
02
02
'1-
ROO0
QOOH
Ir
p-scission + H02
02
OOQOOH
O=R'OOH+OH
Branching
alkene + H02
at low T
intermediate molecular products + OH
Figure 2.1 The main hydrocarbon oxidation pathways
26
high temperatures (above 1100 K at 1 atm), the reaction rate is governed mainly by
P-
scission decomposition of alkyl radicals (R). At intermediate temperatures (between 850 K
and 1100 K), the chain branching of alkanes (RH) with OH or HO 2 begins to act as a major
oxidation pathway [10]. Lastly, at low temperatures (below 850 K), the chain branching of
ketohydropreoxide (ROO) dominates the reaction rate [11].
The hydrocarbon oxidation at low temperatures must be studied more carefully due to
its complexities. Below 800 K, ROO cannot be decomposed into R. Therefore, instead of
decomposition, QOOH isomerization occurs. Since QOOH isomerization leads to complex
chain propagation involving OH/HOO, there is a range of temperature in which the reaction rate decrease rather then increase with temperature; such a response is called NTC behavior.
As a result of such different chemical pathways at high temperatures and low temperatures, the reaction rate of the whole system also depends on the fuel-air equivalence ratio
[12, 13]. Due to the import relationship of H radical pool with reaction rate, any reaction
which consumes H radicals will decelerate the reaction rate. On the other hand, any reaction which produces H radicals will accelerate the whole reaction rate. Particularly for fuel
rich combustion, because hydrocarbon does not ignite rapidly or completely, the initial
27
amount of fuel plays a large role than lean or stoichiometric combustion.
At relatively high temperatures (T>1000 K), the main reaction of H+02 = O+OH initiates all the oxidization reactions. Therefore, as the fuel-air equivalence ratio decreases,
which means the system has more oxygen, the system ignites faster due to the higher energy levels. Moreover, because the reaction rate of hydrocarbon and H atoms is much
faster then that of hydrogen and oxygen reaction, the hydrocarbon is good competitor of
oxygen to hydrogen atoms. Therefore, the small increasing amount of hydrocarbon is able
to inhibit the whole system reactivity.
In contrast, at low temperatures (T <1000K) with elevated pressures (P>20 atm), because the reaction of Fuel +HOO = R + H20
2
dominates the reactivity of the whole system,
the increasing fuel-air equivalence ratio accelerates the overall reaction of the system. Because this is not strong exothermic reaction, temperature could not have comprehensive
effect on reaction rate. Therefore, though temperature is increased in this region, the reac-
tion rate will remain same or even slower, which is NTC behavior.
However, if the fuel-
air equivalence ratio increases at lower pressures (P<20atm), H-atom abstraction provides
inhibiting effects by competing with the main reaction (H+0 2 = 0+0H) [14]. Therefore, the
equivalence ratio does not have a large effect on the rate of reaction under the low pressure,
28
even though the fuel-air equivalence ratio is high.
The dominant reactions in the operating region of interest can be summarized as follows:
At low temperatures: ROO = QOOH (QOOH isomerization)
Under fuel-rich conditions: RH + HOO = R + H 2 0
29
2
(H-atom abstraction)
2.3 Modification of Chemical Kinetic Mechanisms
A detailed n-heptane chemical kinetic mechanism developed by LLNL describes 2450
elementary reactions involving 550 different species [7]. Our planned operation window is
beyond the established range of this mechanism: pressures between 1 and 42 atm, temperatures between 550 and 1700 K, and an equivalence ratio between 0.3 and 1.5. It is much
broader operation conditions than where the Curran mechanism addressed on.
Considerable research has been conducted in order to achieve better agreement in extended application regions for the butane mechanism proposed by LLNL [7, 15].
Fur-
thermore, since one of our objectives for this study is to assess the guidelines for the reasonable operation range for the LTC fuel-rich combustion region, the actual experimental
data for n-heptane combustion is insufficient. As butane combustion behavior shows generally good agreement with n-heptane, and the butane mechanism developed by LLNL was
built on the same algorithm as n-heptane, it means the butane mechanism has comprehensive analogy with heptane. So, we modified the butane mechanism based on the experimental data of Kitsopanidis. Therefore, because the initial process of H atom abstraction from
the fuel does not strongly depend on the fuel structure larger than butane, firstly, our modification is performed based on butane experimental data. Then, we applied the same modi-
30
fication to n-heptane.
The LLNL inhibition mechanism is based on the reaction rate equation:
- Ea
k= Ax T"xe
RT
where k is a reaction rate constant, A-factor is a constant, Ea is the activation energy, R is
the universal gas constant, T is the temperature, and n is the stoichiometric coefficient,
which is determined by the fuel. In the LLNL mechanism, the reaction rate is decided by
estimated values of n, A-factor and Ea, based on the previous studies and their correlation
with the actual experimental data [14, 16-19]
Many empirical reactions rate equations are represented based on Arrhenius rate expressed as k = AT" exp(-Ea/ R T)[Fue]a[Oxidizer]b, with the assumption that the overall
reaction is generally the first order as regards both fuel and oxidizer (a=l, b =1). Therefore,
reaction rate naturally depends on the reaction rate constant. However, this reaction constant, k = AT" exp(-Ea/ R T), is not a real constant. It is the function of temperature, activation energy, A factor, and n. R and T are given by the system, but n, A factor, or Ea is decided based on empirical data. However, as the reaction rate strongly depends on the initial
condition, the different initial condition affect a lot to the decision. Therefore, these coeffi-
31
cients could not grantee the generality [5, 20]. We modified the mechanism reaction rate by
changing these coefficients in the particular activation energy.
Figure 2.2 describes the difference between the LLNL mechanism prediction and the
experimental data of Kitsopanidis for butane ignition delay under fuel-rich condition. As
shown in Figure 2.2, the predicted ignition is generally earlier than the real ignition regardless of temperature, and especially the prediction of the delay in the NTC region does not
agree with the experimental data. This discrepancy in the n-butane mechanism is mainly
I
*
-
U
Experimental Data
Original LLN L model
U
P =14.6 atm .4=3.0
U
10
-
C
1 IflO
I. I
1.05
I
1.10
1.15
I
120
125
1.30
1.35
1.40
'
I
1.45
I
1.50
1000OIT [1 IK]
Figure 2.2
Difference between experimental data and predictions from the original LLNL kinetic
model for n-butane.
32
due to the fuel-rich combustion condition. Moreover, this LLNL mechanism is only addressed between 0.3 and 1.5. Therefore, it could not predict the ignition delay in this region
In the previous study of Kitsopanidis, a better agreement is achieved by a model that
1
modifies two reaction groups (H-atom abstraction from RH by HOO: R2 and OOQOOH
isomerization: R23). In later version of LLNL model, it assumed that the activation energy
of R23 set same as that of R12. Even it does not seem physically sound, the computational
results showed better agreement than original results.
In our study, we made three adjustments over three reaction groups. First, to make the
better prediction regardless of temperature, we modified the main initiate reaction pathway
that is the root of the whole reaction in the fuel-rich region (RH + HOO = R + H2 0 2 ).
Second, to improve agreement in the NTC region, we selected a low temperature submechanism, which initiates the NTC behaviors. As we can see from the schematic of hydrocarbon oxidation pathways in Figure 2.1, two main pathways exist relatively independent of other restraints except for temperature: QOOH isomerization (R12), and OOQOOH
isomerization with release of OH radicals (R23). As the reaction equations clearly show, Hatom transfer occurs within these reactions, except that OOQOOH isomerization produces
'Reaction group notation such as R2, RI 2, and R23 followed by Curran, the author of original model.
33
OH radicals at the end of isomerization. Therefore, even though the activation energy of
OOQOOH will be lower than that of QOOH, the activation energies of R12 are strongly
analogous to those of R23, because the both activation energies caused by the H atoms abstraction, which are bound to the carbon atoms.
The original LLNL kinetic mechanism assumed the difference in the activation energy
between these two reactions to be 3 kcal/mol. However, in a later version of a reduced nbutane mechanism, the activation energy is given the same value for these two reaction
groups. Even though it seems physically unreasonable, this modified mechanism shows
better agreement with the experimental data in the fuel-rich regions. Similarly, in
Kitsopanidis' mechanism, the activation energy difference between the two reaction groups
is assumed to be 1.5 kcal/mol. These modifications change only the activation energy of
OOQOOH isomerization (R23).
However, when the estimated activation energy of QOOH isomerization (R12) is lower
than the actual activation energy of R12, the computation rate of autoignition is always less
than the observed rate of that [5]. Therefore, we also reduced the activation energy of R12.
In our study, we achieved better agreement by adjusting not only the activation energy of
OOQOOH isomerization but also that of QOOH isomerization. Our schematic mechanism
34
for the applied modification is described in Figure 2.3. To retard the ignition at high temperatures within the fuel-rich condition, the activation energy of R2 was increased to 1
kcal/mol. Since H-atom attraction has small impact at low temperatures, the difference between the activation energy of R12 and that of R23 set to 2 kcal/mol instead of 3 kcal/mol
used in the original mechanism. Furthermore, to retard feedback on the reaction rate from
temperature, the activation energy of R12 was increased by 1 kcal/mol.
As shown in Figure 2.4, the prediction of our modified mechanism for butane and the
experimental data showed better agreement than previous studied mechanisms. When we
consider the heat transfer effect at low temperatures due to longer ignition delay, the faster
Ea
k
Ea'
+1
Ea.R1
1+4
2
3
E8.R2
Ea.R23
Original model
Original model
Modified model
Figure 2.3 Schematic of the applied modification.
35
::
+
Modified model
ignition predictions at low temperatures and slower ignition predictions at high temperatures are understandable and meet our expectation well.
Because the initial process of H atom abstraction from the fuel does not strongly depend
on the fuel structure larger than butane, the butane kinetic mechanism has comprehensive
analogy with the n-heptane kinetic mechanism [5, 7, 21]. Therefore, we applied the same
modification to three reaction groups of the LLNL n-heptane mechanism. This validated
modified mechanism is used for development of the operating window of the engine, which
*
--
I-,
0
~1o-
Experlmental Data
use d mech anism in pros ant study
Khopanidis's modfied mechanism
...modifed LLNL meohanism
P = 145 atm. =3.0
a
*15
.2
C
1-
0.95
1.00
10
1.10
1.15 1.20
1.25
1.30
Figure 2.4 Agreement after modification in our present study.
36
135
1.40
1.45
1.0
is presented in the next chapter.
37
38
Chapter 3
IGNITION DELAY STUDY
3.1 Introduction
In our study, the ignition occurred through the reaction of a mixture of fuel and oxidizer without any ignition source after some residence time. The ignition delay time can be
defined as the period between the creation of a combustible mixture by some means (such
as a step change in temperature from a very low temperature), and the onset of the rapid
reaction phase leading to the heat release. Ignition delay times can also be indicated by either a fixed temperature increase or the evolution of a certain chemical species. In our study,
ignition delay was defined as the time duration from the creation of the combustible mixture up to the point at 10% of total heat released. As shown in the Figure 3.1, the ignition
occurred with two distinguishable stages, which were demonstrated by NTC behavior
dominated by not strong exothermic reaction.
39
1.1 -.
-
/
160
II
1.0
140
C
0.90.8-
120
0.1
100
04
60
C,,
(D
0.3-
75 220.
0.2-
12-1.,26
E
3
U.1I
--
- -
127
12
1.
20
0.0
0.0
12.4
12.5
I'd
3
-
- -
0
I
TIK
12.6
12.7
12
129
131)
time [ms]
Figure 3.1 Definition of ignition delay with NTC behavior
We need to apply the ignition delay data, which is based on a fixed initial temperature
and pressure, to the ignition problem in engine. Because the combustion environments in
diesel engines change continuously due to the temperature change produced by the piston
motion during the ignition delay, to estimate the ignition delay is more complicate than stationary condition. Generally, the ignition delay is defined as a function of mixture pressure,
temperature, fuel equivalence ratio, and fuel properties [1]. The following empirical integral equation used to reflect the changing effect of pressure and temperature for the engine:
40
f
t
0
where
t
is stationary ignition delay of the reactor under the same conditions as the engine at
time t, and ti. is the ignition delay of the engine. Therefore, to estimate the ignition delay of
a diesel engine, we first calculated stationary delay.
41
3.2 Simulation Model
To estimate stationary ignition delay, we built a batch reactor for the modified nheptane mechanism using OpenChem Pro and Jacobian. Jacobian is a dynamic modeling
and optimization software program that can solve problems by synchronizing with
OpenChem Pro, which can build and apply chemical kinetic mechanisms. Our batch reactor
could be run at constant volume or constant pressure. It is a closed system that requires initial temperature, pressure, a fuel equivalence ratio and a residual gas fraction. We assumed
that the residual gas would have a burned gas equilibrium composition consistent with an
equivalence ratio at 1740K, which would contain only hydrogen, water vapor, carbon monoxide, and carbon dioxide.
Since there is very little heat release in the pre-ignition chemistry [1, 5], the ignition
delays found by the constant pressure or the constant volume reactor would be the same. In
our simulations, this fact is confined in Figure 3.2; the ignition delay does not depend on
the combustor type.
42
W
a
a
U-O
E
0
a)
E
4
4
0
20
60
40
80
100
Percentage of total released heat (%)
with constant pressure reactor
with constant pressure reactor
with constant pressure reactor
with constant volume reactor
P=50atm, with constant volume reactor
P=60atm, with constant volume reactor
where T= 900K, Xr-0.15, 4k = 2.5 , *
P=40atm,
P=50atm,
P=60atm,
P=40atm,
e
e>
o
o
A
Figure 3.2 Agreement of Simulation results from the different batch reactor model according to the
various pressure where T=900K, Xr=0.15, and Phi= 2.5
43
3.3 Stationary Ignition Delay Results
In the present work, the full ignition simulation was run for 900 points on T(600K1000K), P (20 atm- 60atm), Xr(0.0-0.6), and Phi(l.0-2.5). The fuel-rich combustion in
those tests showed the same tendency in regard to pressures and fuel-equivalence ratio as
the fuel-lean combustion did. As shown Figure 3.3, a richer air-fuel mixture extends ignition delay. Also, as shown Figure 3.4, an increase of pressure reduces the ignition delay.
However, the Figure 3.5 clearly shows NTC behavior. The temperature region of the NTC
behavior depends on a fuel-equivalence ratio. While the fuel-air mixture is richer, the temperature range representing NTC behavior becomes wider and lower.
0
0
A
0
10
M
I.
1.0
1.2
I
I
1.4
1.6
P=30atm
*
P=40atm
A
P=50atm
I
I
1.6
2.0
.
Fuel Equivalence Ratio,
m
cD
Figure 3.3 Effect of Equivalence Ratio and Pressure on Ignition Delay at T= 700K, Xr =0.3
44
100-
'A
E 10
A.
.1
U
U
A
A.
U
>1
A
0
4~
C
U
C
A
.2
C
1
0)
SP=30atm
*
P=40atm
A P=50atm
1.0
1.1
1.2
1.3
1.4
1.6
1.5
1000fT
Figure 3.4 Effect of Pressure and Initial Temperature on Ignition Delay at phi=1.5 and Xr =0.3
S
100-
.4
E
U
0
I
10
U
U
I
2.
1.0
o=1.5
*
*
U
(D=2.0
o=2.5
A
1.1
1.2
1.3
1.4
1.5
1.6
1000
Figure 3.5 Changing the NTC behavior region related to a fuel-equivalence ratio as function of initial
temperature at P=40atm and Xr =0.3
45
Figure 3.6 shows that the ignition delay strongly depends on residual gas fraction particularly at low initial temperature combustion. Generally, as we expect, the larger residual
gas fraction produces the longer ignition delay. However, under the high temperature condition, the residual gas fraction does not have significant effects on the ignition delay because
the initial temperature is maintained in our model regardless residual gas fraction. In the
real engine, because increasing residual gas fraction also increases the initial temperature at
the start of a compression stroke, the combustion characteristic is more sensitive to the residual gas fraction.
0
100U
U
*
A
10
I--i
a
0
is
A
A
SXr=0.6
1
SXr=0.3
Xr=O
&
I
I
1.0
1.1
I
I
1.3
1.2
1.4
1.5
1.6
1000/T
Figure 3.6 Effect of Residual Gas Fraction on Ignition Delay as function of Initial Temperature at
P=40atm and phi=1.5
46
Figure 3.7 shows the dependency of temperature regardless an equivalence ratio. However, under high temperature, the ignition delay does not increase even though residual gas
fraction is increased due to the different ignition initiation pathway between high temperature and low temperature as mentioned in chapter 2. In particular, before NTC region, the
ignition delay depends on the equivalence ratio. On the other hand, in NTC region, the
equivalence ratio is not in charge of the ignition delay.
P
0
U)
E 10-
.0
1) -
<t=1.5
0
0
>=2.0
c+=2.5 where T=850 K
:6
0.0
0.1
0.2
0.3
<D
0.5
0
0
0
0.6
0
0
0
E
0.4
0
0
40
30
o <=1.5
o 4=2.0
2-
4>=2.5
0.0
0.1
0.2
0.3
0.4
Residual Gas Fraction, Xr
Figure 3.7 Effect of Residual gas fraction on Ignition delay at P=40atm
47
0.5
where T=950 K
0.6
Figure 3.8 clearly shows that the same dependency is reflected by pressure. Because
the NTC behavior region complexly depends on temperature and equivalence ratio, the ignition delay around NTC region is hard to be predicted. Thus, these stationary results are
used for interpolation of each ignition delay in given engine conditions and integration to
find out the actual engine ignition delay in the next chapter.
I
U
40
30
A
20
10
0
4
3
2
(0
C
U
0
I
0
*
&
0.1
7
0.0
0.3
0.2
0.1
P=30atm
P=40atm
P=50atm
where T=850 K
0.4
M
U
N
.A.
C
C
0
*~2
SP=30atm
* P=40atm
P=50atm
A
0.0
0.1
0.3
0.2
0.4
R esidual Gas Fraction, Xr
Figure 3.8 Effect of Residual gas fraction on Ignition delay at Phi = 2.0
48
0.5
where T=950 K
0.6
Consequently, due to these complicated dependencies on reaction conditions, such as
pressure, temperature, and gas compositions, the ignition delay in the engine is hard to be
predicted in particular with the fuel-rich mixtures. Therefore, we interpolated previous results and estimated an instant ignition delay during the compression stroke, and then find a
ignition delay point in an engine in stead of inferring a general equation of ignition delay in
the engine from these results in this chapter.
49
50
Chapter 4
ENGINE OPERATION WINDOW STUDY
4.1 Introduction
The purpose of the study is to define the operating boundary for a LTC engine operating in the fuel rich mode. The approach was to incorporate the ignition delay would into a
cycle simulation program. A simple Wiebe function was applied to calculating a cumulative
burned mass fraction. Then the engine cycle was complicated to find the residual gas fraction and temperature to define the change for the next cycle. The pressure was unpacted to
see if the cycles converged to a stable LTC operating point.
51
4.2 Simulation Model Setup for LTC Engines
In the engine combustion modeling, we made a dynamic link library between Fortran
and WAVE, which is commercial engine simulation package. The WAVE calculates thermal
environments in an engine and gives the internal engine conditions to user combustion
model at each time step, 0.5 crank angles, such as temperature, pressure, and chemical
compositions in the engine.
The user combustion model will estimate the starting point of the ignition by using a
following equation:
9ign
1
f
-
dO
- N=1;
Olvc T(TP,) 360N
where, N is speed of the engine,
0
N=Rev/sec
ign is the starting point of the ignition, and this equation
starts to be calculated the right after intake valve closed, Oj,. After starting ignition, the
cumulative burned mass fraction is represented as a function of crank angles (CA) using the
Wiebe correlation:
W
I - EXP
AWOI
M-
B
WEXP
~BDUR)
where, AO is CA past start of combustion, BDUR is the 10-90% burn duration in CA,
WEXP is the Wiebe exponent(=2.0), and AWI(= 1.67) is a internally calculated parameter.
52
Because the burn duration of HCCI is shorter than that of SI, the burn duration of HCCI are
generally assumed as 10 CA [28]. This Wiebe function is generally used for the combustion
modeling of the SI engine or the heat release modeling of the typical HCCI engine. However, because we addressed only the exhaust temperature and enveloped the operating conditions, we simplified the combustion period.
Figure 4.1 shows the schematic of our engine configuration. This engine inhaled and
exhaled ambient temperature air, and an injector located right after an intake port to make
homogeneity mixture. The cylinder is a CI combustor. It means this engine has the same
configuration with a HCCI engine. A simple EGR system is used with a valve actuator and
a PI controller to control the residual gas fraction. The external EGR works well due to its
&A
Figure 4.1 Schematic of the system configuration
53
Oiuct 1i
tvtf4
du .da
simplicity, but its thermal effect is limited due to heat loss in the EGR system and slow response during the transient time [22]. As a result of this simulation, we can take exhaust gas
temperatures, peak pressures, indicated mean effective pressures, and thermal heat efficiencies.
As we can see Figure 4.2, the start of ignition timing is various during the first 15 cycles.
However, after
15 th
cycle, cycle variation is negligible and the solutions are converged.
Therefore, to decide the start of ignition regardless cycle variation, we assumed this system
is always stabilized after 100 cycles. If the system has one more misfire cycle before 15th
cycle, we consider it as a misfired case even if the system is stable after 100 cycles. If the
20 Is -
-10 -
=2.0. Xr =15%
-
=1.6, X r = 16%
Xr = 16%
---15----4=1.2,
S
10
a
20
~
30
~ i ~a ~
to
~ = 7| 1r i
s30
Cycle
Figure 4.2 Convergence
of solution after cycle-by-cycle variation
54
70
so
90
-r--1ie
1GO
30
14
13
28
-
I--Ignition delay
-0-Ig n itio n t im in g
4- = 1.5, 1500 rpm. WOT
12
11
26
24
10
9
E
22
w
(3
a
in
'0
C
-o0
20
E
'-*-----k-
*-k-k-*-*-*
--
-*-0-0-k-k-k--
-
-
18
6
5
16
3
0
0 2---
.
14
-o-a3--o- -a - -c-a -- - -a- -a -a -a -o-o- -a -o -oo
12
- - -
,
,
.
I
15
10
5
.
20
-
25
30
cycle
Figure 4.3 Convergence of solutions in terms of Ignition delay [ms] and Ignition timing[deg]
820
I
-
0 -0 - 0 - -0
to
I
I
I
-0 -0 -0-0-0 -0 -0-0 -0 -0-0 -0-0-0 -0-0 0-0-0 -0
800
38
180
36
160
34 a)
a)
740
32
C:
120
30
(D
C)
E
E
-0p
0
-Mu u-
- ---
n-- --
--
--
- - ---
Co
u- -0u-0
On'
0
03
Co
OD2
0.03
-0-
02
012
Co
Ono
Pressure
----
Temperature
-----
Risidual gas fraction
.=15,
1500 rpm. WO T
(C
-
0
-
-
5
10
15
20
25
30
Cycle
Figure 4.4 Convergence of solutions in terms of Pressure, Temperature, and Engine compositions
55
engine has misfire cycles, it exerts a bad influence on engine reliability by increasing thermal gradients. Figure 4.3 and Figure 4.4 also represent system stabilization of system properties such as ignition delay, temperature, pressure, and residual gas fraction.
As we discussed earlier, Wiebe function is used to determine combustion profile. Other
constant properties in Wiebe function are easily defined for HCCI combustion except burn
duration. Burn duration is defined by number of CAD, which takes the engine to complete
10-90% of heat release. Burn duration can be used as a basis to determine the speed of
combustion. Longer burn duration provides a smoother and quieter operation. For HCCI
combustion, the burn duration is shorter than SI combustion and it is suggested generally
between 10 and 20 CAD [23, 24]. Also, it should be noted that burn duration is high dependent on an EGR ratio. Because higher EGR ratio make inlet temperature increase and
outlet temperature decrease, the EGR ratio increase the burn duration even the burn duration does not change much with different inlet temperature. As a result, burn duration can
be increased by as much as 12 CAD with 40% of EGR [22].
56
Figure 4.5, however, depicts that the ignition delay does not strongly depend on the
burn duration. Thus, we can choose burn duration as 20 CAD regardless engine operating
conditions. Still, this averaging of the combustion duration over cycles is one of the potential sources of discrepancies with experiments [22, 25-27]. Figure 4.6 shows a sample engine combustion simulation result.
28
-
-
24
20
E
16-
C
12-
0
a
I
6
a
I
10
I2 .1
12
14
.
1I
16
15
Burn d uratio n [CA]
Figure 4.5 Sensitivity of Ignition delay with Burn duration
57
20
22
24
26
0025-
*
burned ftl In cylinder
Heat release rate~
**Cum.
0.020-
018
7-
0.015-
E
0
0.010-
0.2
0
z
41
01105
-
0.0004
I-
I1
-30 -20
-10
1 1
0
10
1 1
1
20
I I,-
1Ir
30
40
ATDC(CAJ
Figure 4.6 Heat release rate profile of the engine
58
50
60
I
70
nn
IIu~.u
'
80
90
100
4.3 Engine Operation Maps
In present work, the integrated full cycle model was run for 3240 points corresponding
to each specific fuel equivalence ratio, EGR, and speed. As depicted in Figure 4.7, the operation window is bounded by three regions: misfired region, unstable combustion region,
and stable combustion region. Misfire occurs when the engine cannot sustain stable operation or when substantial number of the engine cycles fails to ignite. Unstable combustion
occurs when the engine has the boundary condition between misfire and stable combustion,
causing less than 1% of the engine cycles failing to ignite after the initial period.
17 11
'1
V
-
Missfired region
-..
22-
* * .... .............
Y.
*y
2.0
-
I.y
Unstable
combustion
region
1B
: Stable combustion region 1
1.2
Y
~
Wr
Sao
1000
1500
speed [rpm]
Figure 4.7 Engine operation window: Data points with 10 % EGR
59
2000
2500
Figure 4.8 depicts the calculated ignition starting points as a function of EGR at different equivalence ratio. The EGR effect gets bigger as the fuel is richer. Also, Figure 4.9
shows that the NTC behavior is stronger when the fuel is richer and the EGR ratio is higher.
Such NTC behavior is caused by the temperature in the engine during the compression
stroke as shown in Figure 4.10. The inlet temperature, which strongly depends on the EGR
ratio, is a convenient and efficient way to control the ignition timing and combustion phasing [28, 29]. The high intake temperature speeds up the onset of low-temperature chemistry
with NTC behavior, in turns reducing the main stage ignition delay. Therefore, the ignition
1.5
16
14
12
10
-l 4
E$OR(%)
Figure 4.8 Ignition points as function
of EGR at different <D
60
20
-
1510< 5 C
0
0-5 -5
-10 -
EGR
-15.
1
1
*'
'
1
I
. "
I
'
*
I
10%
*
2.2
2.0
1.8
1.6
1.4
1.2
1.0
=
Figure 4.9 Effect of Equivalence Ratio on Ignition Timing
n
lb 0A^
TEMPERATURE v& CRANK ANGLE
eq.rafo:1.0 eq.ralo:1.8
900
850
Bo
2500-
10
700
2000
450
No
w
860
00
1500-
00
400
00
-80-70
-60
-80
-to
-30
-20
-10
Crang angle [ATDC]
1000-
.2
500-
7VC
(I
iei
COMPRESSION
IVO
EI
II
VD
T4
EXPANSION
CRANK ANGLE
Figure 4.10 Temperature Profile in the Cylinder
61
E~C
NO .
I
EXHAUST
[deg]
INTAKE
0
to
timing will be advanced or maintained though the EGR ratio is increased as shown in Figure 4.8.
Figure 4.11 well represents the ignition points with regard to various fuel-air equivalence ratio and EGR ratio. This Figure 4.11 also provides an insight into misfire phenomenon and a possible way that could be accounted for while modeling and experimenting with
an engine. The misfire was denoted by considering a multiple cycle simulation where the
intake charge failed to bum during some of the cycles. Around the boundary near the misfire conditions, the ignition timing is later than suggested ignition timing for the engine
Start of Ignition [ATDC]
25
20
W0 -
-
80
4 0
0
12
0
-4.
0
1.0
1.1
1.2
1.3
1-4
1.5
Figure 4.11 Engine operation map
62
1.6
1.7
1.8
1.9
2.0
2.1
which is less than about Ims ( 9 CAD with 1500 rpm) [1]. Therefore, it is not recommendable to operate the engine in this region. Also, Figure 4.12 depicts that this boundary region
was denoted by a low Indicated Mean Effective Pressure (IMEP) and a high Indicated Specific Fuel Consumption (ISFC). As a result, it has a very low indicated thermal efficiency
(ru). Indicated thermal efficiency and indicated specific fuel consumption are generally
used to express the efficiency of an engine. Therefore, Figure 4.11 represents the effective
operation windows in terms of an overall engine efficiency as well. When a engine system
30
9W0
.0
3
1100-
-
-200
150
E24
E24
?60
90022
7000
850,
20
400
$300-$
300-
260-
2S0_
_ -
"00
In
12
1.4
1A
1
350
16
350
20
500
V
In
20
12
1.6
1.4
4
EGRuO%
12
I2 100
I
y
1.8
1.8
2.0
22
EGRa10%
Figure 4.12 Indicated Mean Effective Pressure, Indicated Specific Fuel Consumption, and Indicated
Thermal Efficiency of an engine as a function of Fuel Equivalence ratio
63
2400
30
2200
2000
27
1800
C-
1600
25
2 19
1400
23
2
0)
Ca
1200
1000
800
600
Break thermal efficiency
with EGR = 10 %
14
'I.
1.0
1.2
1.4
1.6
1.8
2.0
Figure 4.13 Speed effects on the break thermal efficiency
is operated under an appropriate EGR ratio, the engine can be utilized with richer fuel mixture without an efficiency penalty. Figure 4.13 also depicts the operating potential of the
engine with richer mixture without a penalty of efficiency under appropriate speed.
In summery, the ignition points and the engine parameters are predicted by the model
over the entire rich operating range based on calculated stationary delay data with the modified n-heptane mechanism in this study. In consequence, for developing an operating window of an engine, the Ricardo-WAVE based engine simulator offers a time and cost effec-
64
tive tool. Also, as shown in Figure 4.14, the exhaust gas temperature of this system can supply enough thermal energy to operate and maintain the latter intermediate temperature solid
oxide fuel cell.
Exhaust gas temperature [K]
0
100 207
79
725
70
70
00
0
Figure 4.14 Exhaust Gas Temperature under 1500rpm
65
66
Chapter 5
SUMMERY AND CONCLUSIONS
5.1 Overview
The purpose of this research was to investigate stable engine operation windows and
exhaust gas temperature of Low-Temperature-Combustion Diesel Engines with fuel rich
mixtures to build the guidance for future studies of LTC Diesel Engines as on-board fuel
reformers for Intermediate Temperature Solid Oxide Fuel Cell Vehicles. This study will assist testing hardware and obtaining experimental data for real products optimization. In this
chapter, the findings of the research will be summarized, and conclusions will be drawn.
5.2 Fuel-rich combustion
This research investigated the oxidation characteristics of hydrocarbon fuels under
high pressure and fuel-rich conditions using detailed chemical kinetic mechanism model.
Because of the discrepancies with experimental data and predicted data, which are obtained
from conventional chemical kinetic mechanisms, especially under fuel-rich condition, we
67
developed the modified chemical kinetic mechanisms, which provide more accurate predictions of ignition delays. In this study, the main oxidation pathways of fuel-rich mixtures,
which are mainly adjusted as shown in chapter 2, are selected by the comprehensive analysis of chemical oxidation. The stationary ignition delays, which are used for interpolations
to find the ignition starting points in engines, are calculated based on this modified chemical mechanism of n-heptane. The findings are summarized as followings:
.
Richer mixture can react faster in particular under high pressure, and low temperature. Moreover, because the combustion cannot be completed under fuelrich condition, the rate of reaction under fuel-rich condition depends on the fuelair equivalence ratio heavier than that under fuel-lean conditions,.
.
The temperature range of interest coincides with the negative temperature coefficient region of the ignition delay curve. Thus, the ignition delay is not a strong
function of temperature.
68
5.3 Ignition Delay in Engines
In engines, because pressure and temperature are changing, method needs to be devised
to connect the ignition calculations done at fixed condition to the engine environment.
The method need was by integrating the inverse of the stationary ignition delay value. To
utilize CO and H2 in exhaust gas, the engine should be run under fuel-rich condition, and
to minimize NOx and PM, which adverse effect on latter fuel cell operation, the engine
will be better to use under Low Temperature Combustion. To achieve the operation stability of the engine under these conditions: fuel-rich and LTC, which extend ignition delays, the engine will be operated with external gas recirculation system that affect burned
gas temperature as well the ignition delays. Because thermal energy in the burned gas
will use a source of maintaining operation temperature for ITSOFC, the burned gas temperature should be examined into details. The findings are summarized as following:
.
According to operation maps in chapter 4, engines can be run as rich as a 2.0
fuel-air equivalence ratio under 10% EGR.
.
Generally, exhaust gas temperatures are between 700 C and 900 C so that the
energy in the exhaust gas are enough to heat up fuel cell system during cold start
periods and can maintain the appropriate operation temperature for ITSOFC.
69
*
Indicated thermal efficiencies are between 20-30%. By using EGR system, the
engine can be run under richer condition with maintaining same efficiency with
leaner condition.
5.4 Conclusions
From this study, the fuel-rich LTC engines were established as on-board reformers for
ITSOFC. By using this configuration, not only the thermal energy for maintaining and operating the fuel reforming process, which is exceedingly required by conventional fuelreformers, are reduced, but also there is no start up dead time, which is necessary for warming up the conventional SOFC system. Besides, under certain conditions from chapter 4,
engines can be operated without the penalty of thermal efficiency with the fuel-richer mixture.
Therefore, the implementation of a LTC diesel engine as an on-board fuel reformer for
ITSOFC vehicle is expected a promising alternative power plant for future vehicles and the
results from the study will be used as guideline for future developments of real products.
70
References
1.
Heywood, J.B., Internal Combustion Engine Fundamentals,Mc Grqw-Hill, New
York, 1988.
2.
Ghoniem, A. F. Fundamentalsof advanced energy conversion, Lecture note, Cam-
3.
bridge, MA, 2005.
Pettersson, L. J. & Westerholm, R. State of the art of multi-fuel reformers for fuel
cell vehicles: problem identification and research needs. Int. J.Hydrogen Energy 26,
243-264, 2001.
4.
Cheng, W.K., Assessment of Low Temperature Combustion Diesel Engine as an
On-Board Reformer for Intermediate Temperature Solid Oxide Fuel Cell powered
Vehicle, Proposalto Eaton Corporation,2005.
5.
Curran, H.J., Gaffuri, P., Pitz, W.J., and Westbrook, C.K., "A Comprehensive Modeling Study of n-Heptane Oxidation," Combustion and Flame, 114:149-177, 1998.
6.
WAVE, Ricardo Consulting Co.
Curran, H. J., Gaffuri, P., Pitz, W. J., and Westbrook, C. K. "A Comprehensive
7.
Modeling Study of iso-Octane Oxidation" Combustion and Flame 129:253-280,
2002.
8.
Chevalier, C., Pitz, W.J., Warantz, J., Westbrook, C.K. and Melenk,H., "Hydrocarbon Ignition : Automatic Generation of Reaction Mechanism and Applications to
Modeling of Engine Knock," 24th proc. the combustion institute,p. 93-101, 1992.
9.
Griffiths, J., MacNamara, J.P., Mohamed, C.,WhitakerB.J., and Shepperd, G.W.,
"Temperature fields during the development of autoignition in a rapid compression
machine," FaradayDiscuss., 119:287-303 ,2001.
Westbrook, C.K., Warnatz, J. and Pitz, W.J., "Simulation of Auto-Ignition over a
10.
Large Temperature Range," 22th proc. the combustion institute,p. 893-901, 1989.
11.
12.
Marinov, N. M., Pitz, W. J., Westbrook, C. K., Vincitore, A. M., Castaldi, M. J.,
Senkan, S. M. "Aromatic and Polycyclic Aromatic Hydrocarbon Formation in a
Laminar Premixed n-Butane Flame," Combustion and Flame, 114:192-213, 1998.
Dagaut, P., Reuillon, M., and Cathonnet, M., "High Pressure Oxidation of Liquid
Fuels from Low to High Temperature," Combust. Sci. and Tech., 95:233-260, 1994.
71
13.
14.
Minetti, R., Carlier, M., Ribaucour, M., Therssen. E., and Sochet, L. R., Combust.
Flame, 102:298-309, 1995.
Westbrook, C. K. and Pitz, W. J., Combust. Sci. Tech., 37:117-125,1984.
15.
Kitsopanidis, I., " Experimental and Computational Study of Soot Formation Under Diesel Engines Conditions", Ph D Thesis, Massachusetts Institute of Technology,
Cambridge, MA, 2004.
16.
Allara, D. L. and Shaw, R., J. Phys. Chem. Ref Data, Vol. 9, No. 3, p. 523-559,
1980.
17.
18.
Droege, A. T. and Dly, F. P., J.Phys. Chem., 90:1949-1965, 1986.
Michael, J. V., Keil, D. G., Klemm, R. B., Int. J.Chem. Kinet., 15:705-732, 1983.
Baker, R. R., Baldwin, R. R., and Walker, R. W., 1kans. Famday Sot., 66:28122829, 1970.
19.
20.
Westbrook, C. K. and Dryer, F.L., "Chemical kinetic modeling of hydrocarbon
combustion," Prog.Energy Combust. Sci., 10:1-57, 1984.
21.
Dagaut, P., Reuillon, M., and Cathonnet, M., "Experimental Study of the Oxidation
of n-Heptane in a Jet Stirred Reactor from Low to High Temperature and Pressures
22.
up to 40 atm, " Combustion and Flame, 10 1:132-140, 1995.
Zhao, F., Asmus, T. W., Assanis, D. N., Dec, J. E., Eng, J. A. and Najt, P. M. "Homogeneous Charge Compression Ignition (HCCI) Engines: Key Research and Development Issues," 1st edition, SAE Inc., Canada, pp. 1-658, 2003.
23.
Bhave, A., Kraft, M., Oakley, A., Zhao, H., and Mauss, F., "Analyzing the limits of
a stable Homogeneous Charge Compression Ignition (HCCI) operation," CCCE,
5 th
Aug., 2004.
24.
Au, M., Girard, J. and Dibble, R. "1.9-Liter Four-Cylinder HCCI Engine Operation with Exhaust Gas Recirculation", SAE Paper 2001-01-1894.
25.
Christensen, M., Johansson, B., "Influence of Mixture Quality on Homogeneous
Charge Compression Ignition," SAE Paper 982454.
26.
Morimoto, S., Kawabata, Y, Sakurai, T., Amano, T., "Operating Characteristics of
a Natural Gas fired HCCI Engine", SAE Paper 2001-01-1034.
27.
Oakley, A., Zhao, H., Ladommatos, N., Ma, T., "Experimental Studies on CAI
Combustion of Gasoline in a 4-Stroke Engine", SAE Paper 2001-01-1030.
28.
Thring, R. H. (1989). Homogeneous charge compression ignition (hcci) engines,
SAE Paper 892068.
72
29.
Najt, P. and Foster, D. E. (1983). Compression ignited homogeneous charge com-
bustion, SAE Paper 830264.
73
74