Sei sulla pagina 1di 41

7

CHAPTER 2.

BACKGROUND AND LITERATURE REVIEW

Many practical problems in engine operation and performance are controlled by


autoignition chemistry. Classic examples are knock in spark ignition engines and cold
start in diesel engines. The chemistry that controls autoignition in HCCI combustion is
the same as that which leads to knock in SI engines. Studies of autoignition began in the
early 1900s when knock was first realized as a limitation on engine output and fuel
efficiency. Thus, all of the previous research work devoted to knock chemistry in SI
engines over the last 100 years is directly applicable to HCCI combustion, and there is a
wealth of literature that can be used to guide our research.
Instead of reviewing all aspects of hydrocarbon combustion, this chapter provides
background and reviews the past work related to the scope of this research program. First,
a comparison between the CI, SI and HCCI combustion processes and emissions is made.
Second, the history of HCCI engines is reviewed. Then, related research work on fuel
additives and an introduction to their effects on hydrocarbon oxidation is presented.
Finally, previous research on hydrocarbon autoignition and oxidation and on kinetic
mechanism development is discussed.

2.1

Comparison of CI, SI and HCCI Engine Combustion and Emissions


Emissions from the combustion of hydrocarbons in internal combustion engines

are major sources of pollution throughout the world. Regulations introduced by the
Environmental Protection Agency (EPA), California Air Resources Board (CARB), and
international regulatory agencies are requiring vehicles and off-highway powered
equipment to substantially reduce emissions. Significant reductions in carbon monoxide

8
(CO), unburned hydrocarbons (UHC), nitrogen oxides (NOx) and particulate matter (PM)
will be required in almost all classes of engines. Also, fuel efficiency continues to be a
major area of public and policy interest due to its direct relation to carbon dioxide
emissions, which is the pollutant most often associated with global warming.
The simplest way to improve the efficiency of an engine is to increase the
compression ratio.

However, high temperatures and pressures caused by high

compression ratios are normally associated with high NOx. Also, in spark ignition
engines, the high temperature end gases promote autoignition and knock, which limits the
maximum engine compression ratio.

Therefore, the way to improve efficiency by

increasing compression ratio also increases NOx emissions. Also, combustion strategies
that reduce NOx emissions invariably result in increased HC and PM emissions, and
conversely, strategies that reduce HC emissions almost always increase NOx emissions
[Borman, 1980; Turns, 1999; Heywood, 1988].
Generally, due to high temperatures and heterogeneous combustion of the
atomized fuel, Compression Ignition (CI) engines are very efficient, but emit a large
amount of NOx and PM and only small amounts of CO and UHC.

Modern well

controlled catalyst equipped SI engines are modest emitters of CO, UHC and NOx, and
very small emitters of PM, but are less efficient. They also require more refined fuels
than CI engines. A comparison of CI, SI and HCCI engine combustion processes and
emissions is presented in Fig. 2-1.

(a)
Spark Ignition:
spark-ignited
flame propagation
premixed combustion
throttled
port-injection
stoichiometric

(b)
Compression Ignition:
auto-ignition
flame propagation
premixed and diffusive
combustion
unthrottled
direct-injection with swirl
Variable stoichiometry
(lean to rich)

(c)
HCCI:
auto-ignition
no flame propagation
premixed volumetric
combustion
unthrottled
port or direct-injection
lean/ dilute stoichiometry

Figure 21. Comparisons of SI, CI and HCCI combustion


processes (Figures from Ogink, 2004)

In Spark Ignition (SI) engines, the fuel is mixed with air in the intake manifold to
form a premixed charge with equivalence ratio around stoichiometric. When the spark
plug fires, a flame kernel is formed and a flame propagates through the homogenous
charge. As flame propagation occurs, the temperature at the front --- a thin zone of
intense chemical reaction --- is high, and significant NOx formation occurs in the postflame, hot combustion products. Stratified charge SI engines, while attempting to avoid
this high temperature region, still have problems with high emissions [Aoyama et al.,
1996].

10
The thermal efficiency of SI engines depends on the compression ratio.
Unfortunately, the compression ratio is limited by autoignition of the unburned gases.
Severe autoignition leads to knock and limits engine efficiency and thereby increases
emissions. The homogenous premixed combustion in the SI engine contributes to its
very low PM emissions. The NOx formed in the flame front and post flame regime is
primarily NO. The most significant reaction mechanism forming NO is the Zeldovich
[Miller and Bowman, 1989].

CO, a primary intermediate of HC combustion, is

invariably formed and in untreated exhaust CO concentration is the highest of all


emissions.
For all types of engines, hydrocarbon emissions result from the presence of
unburned fuel in the engine exhaust. In SI engines, about 9% of the fuel supplied to an
engine is not burned during the initial flame propagation event. However, most of this
unburned fuel is consumed as a result of post combustion oxidation processes during the
power expansion stroke, including oxidation in the exhaust port during the blow down
process. Ultimately, about 2% of the total fuel flow into the engine will leave with the
exhaust, including partial reaction products, such as acetaldehyde, formaldehyde,
1, 3 butadiene, benzene, etc. [Cheng et al., 1993]. As hydrocarbon emissions represent
lost chemical energy, the UHC emission also represents a decrease in the thermal
efficiency.
There are six primary mechanisms believed to be responsible for hydrocarbon
emissions from SI engines, Table 2-1.

11
Table 2-1. Primary sources for hydrocarbon emissions in SI engines
(Cheng et al., 2003)
% fuel escaping
normal combustion

% contribution to the 2% of
unburned fuel after burnout

Crevices
Oil layers
Deposits
Liquid fuel
Flame quench
Exhaust valve leakage

5.2
1.0
1.0
1.2
0.5
0.1

38
16
16
20
5
5

Total

9.0

100

Source

Crevices these are narrow regions in the combustion chamber into which the
flame cannot propagate because they are smaller than the quenching distance.
Crevices represent about 1 to 2% of the clearance volume.

Oil layers - Since the piston ring is not 100% effective in preventing oil migration
into the cylinder above the piston, an oil layer exists within the combustion
chamber that absorbs fuel.

Deposits Carbon deposits build up on the valves, cylinder and piston crown.
These deposits are porous with pore sizes smaller than the quenching distance so
trapped fuel cannot burn.

Liquid fuel For some fuel injection systems there is a possibility that liquid fuel
is introduced into the cylinder past an open intake valve. The less volatile fuel
constituents may not vaporize (especially during engine warm-up) and be trapped
in the crevices and carbon deposits.

12

Quenching Most of the hydrocarbon contained in the wall quench layer diffuse
into the hot combustion products outside the layer and get consumed during the
post combustion oxidation processed. However, bulk gas quenching can occur
during the decompression and blow down processes when the temperature drops
to a low enough level.

Exhaust valve leakage- Exhaust valves which are normally closed may leak
UHCs directly into the exhaust port.

In CI engines the liquid fuel is injected at high pressure directly into the
combustion chamber near Top Dead Center (TDC). The atomization, vaporization and
mixing of fuel spray with the swirling compressed air in the cylinder occurs in a hightemperature and high-pressure environment. When the in-cylinder temperature is above
the autoignition temperature of the fuel, the mixture will spontaneously ignite following
an ignition delay period.

Subsequently, any vaporized premixed charge with

stoichiometry within the flammability limits will be rapidly consumed.

Ultimately,

mixing controlled combustion dominates the remainder of the combustion process. The
inhomogeneous mixture and high combustion temperature in CI engines produces NOx
in the oxygen-rich and stoichiometric regions, and particulate in the fuel-rich regions.
NOx is formed in the high temperature regions where both oxygen and nitrogen are
available, and in the post combustion hot gas regions [Miller and Bowman, 1989]. As
temperature is proportional to load in a CI engine, more NOx is formed as the load
increases. Due to the diffusive combustion process and the presence of very rich mixture

13
regions, PM formation is unavoidable. Some of the PM is destroyed in the flame by
oxidation and the unoxidized PM becomes an exhaust emission [Schommers et al., 2000].
It is difficult to reduce both NOx and particulate simultaneously. In CI engines,
the new rules will require electronic engine controls, exhaust gas recirculation (EGR),
and improvements in after-treatment (particulate filter, NOx trap or DeNOx) to reduce
NOx and particulate levels.
The following two factors are believed to be additional sources for UHC emission
in CI engines:

Undermixing of fuel and air - Fuel leaving the injector nozzle at low velocity, at
the end of the injection process, cannot completely mix with air and burn.

Overmixing of fuel and air - During the ignition delay period evaporated fuel
mixes with the air, regions of fuel-air mixture are produced that are too lean to
burn. Some of this fuel makes its way out the exhaust. If ignition delays are
excessively long more fuel becomes overmixed.

Since in-cylinder temperatures are higher in CI engines, UHC emissions are usually
significantly less than in SI engines.
HCCI engines utilize homogeneous charge as in SI engines; however, the charge
is compressed to ignite as in CI engines. This new combustion concept provides the high
fuel efficiency of CI engines and the lower NOx and PM emissions of SI engines. Key to
the application of HCCI is to create a charge that produces a smooth heat release profile
across the entire operating ranges.

This usually requires a dilute, lean charge that

produces maximum temperatures low enough that thermal NOx emissions are

14
dramatically reduced. Due to lean, premixed operation the PM emission is lower too.
High efficiencies are achieved by operating unthrottled with high compression ratios as in
compression ignition engines. There are some other benefits with HCCI engines as well,
such as the capability of using multiple fuels.
However, due to the low combustion temperature, particularly at lower load
conditions, excess CO and UHC emissions are found in HCCI [Dec, 2002; Christensen et
al., 2001; Easley et al., 2001]. A detailed discussion of HCCI engines is given in the
following section.

2.2

Homogenous Charge Compression Ignition (HCCI) Engines


The first HCCI concept was proposed in the late 1970s. This idea has drawn

major attention in the last decade due to the urgency to meet stricter regulations on NOx
and PM emissions. Although tremendous experimental and modeling efforts have been
brought to bear on HCCI phenomena in the past several years, only the recent advent of
electronic sensors and controls has made HCCI engines a potential practical reality
[Epping et al., 2002]. This section provides a brief history of HCCI studies, an overview
of the current state-of-the-art in HCCI technology, and a list of the R&D barriers that
must be overcome before HCCI engines can be considered for commercial application.

2.2.1

Introduction to HCCI
HCCI is an alternative piston-engine combustion process that can provide

efficiencies as high as compression-ignition (CI) engines while producing ultra-low

15
oxides of nitrogen (NOx) and particulate matter (PM) emissions, unlike CI engines.
HCCI engines operate on the principle of having a dilute, premixed charge that reacts and
burns volumetrically throughout the cylinder after compression by the piston. HCCI
incorporates the best features of both spark ignition (SI) and compression ignition (CI).
As in an SI engine, the charge is well mixed, which minimizes particulate emissions, and
as in a CI engine, there are no losses due to inlet throttling, the charge is ignited by the
high ambient pressure and temperature produced by compression, and the load is
determined by the amount of fuel in the charge, which leads to high efficiency. However,
unlike either of these conventional engines, the combustion occurs simultaneously
throughout the volume rather than in a flame region. This important attribute of HCCI
allows combustion to occur below typical flame temperatures, dramatically reducing
NOx emission. The resulting disadvantage of HCCI operation is that the engine may be
hard to start and the combustion process requires new control methods.

These

disadvantages presently restrict the application of HCCI engines. However, the potential
of the HCCI concept has motivated studies designed to understand the ignition and
oxidation chemistry of possible fuels.
The first HCCI operation was reported by Onishi et al. [1979] who measured a
unique combustion behavior they called Active Thermo-Atmosphere Combustion
(ATAC), which was intermediate between SI and CI.

Achieved on a two-stroke

gasoline engine under relatively lean conditions, the ATAC process obtained lower fuel
consumption and low emissions in the region of light and medium loads, with less noise
and vibration. High speed Schlieren photographs showed that ATAC was initiated by a
multipoint autoignition without discernable flame propagation.

16
Later the same year, Noguchi et al. [1979] reported similar self-ignited
combustion in a two-stroke gasoline engine. They named the combustion process TS
(Toyota-Soken) combustion. High levels of HCO, HO2, and O radicals were observed
within the cylinder prior to autoignition, which demonstrated that pre-ignition chemical
reactions had occurred and that these reactions certainly contributed to the autoignition.
In a traditional SI engine, these preignition radical species are primarily associated with
end-gas autoignition, namely knock. After autoignition took place, H, CH, and OH
radicals were detected, which were indicative of high-temperature chemical reactions.
Also, the combustion process seemed to start at lower temperatures and pressure than
those for conventional CI combustion.
Following these two pioneering studies, the operating mode, renamed HCCI, has
been demonstrated on a number of two-stroke engines by several researchers. Lida [1994]
broadened the stable two-stroke ATAC combustion range by using methanol as the fuel.
Later, other alternative fuels such as dimethyl ether, ethanol, and propane were also
tested by Lida [1997] to investigate the fuel sensitivity of HCCI operation on two-stroke
engines. Honda has proven the reliability of the concept for a production two-stroke
engine by placing 5th overall in the Granada-Dakar desert race with a pre-production
motorcycle [Yamaguchi, 1997]. A pre-production two-stroke engine employing HCCI
has also been shown by Duret and Venturi [1996]. In both cases, HCCI was used to
improve combustion stability, reduce HC emissions and improve fuel economy at part
load.
Honda has a 2-stroke cycle, single-cylinder HCCI engine that operates on
gasoline and powers a motorcycle [Ishibashi and Asai, 1996]. This engine operates in

17
HCCI mode at low to moderate loads, and switches to conventional SI operation at high
loads. Even though HCCI is used over only part of the duty cycle, the engine has
demonstrated considerable advantages in fuel economy, which is 27 percent better than a
regular 2-stroke cycle engine under "real-life" riding conditions. Hydrocarbon emissions
are also reduced by 50 percent with respect to a regular 2-stroke cycle engine. However,
without emission controls, hydrocarbon emissions are still very high compared to the
current emissions standards.
While efforts on two-stroke HCCI engines have made significant progress, the
efforts on four-stroke HCCI engines have achieved only marginal success.

The

inherently high Exhaust Gas Recirculation (EGR) rate of two-stroke engines helps to
control the rate of heat release, and thus the knock intensity of the engine. For a fourstroke engine, controlling the rate of heat release with little or no EGR while maintaining
the engine performance is an obstacle to achieving HCCI operation.
The first success in applying HCCI combustion to a four-stroke engine was
achieved by Najt and Foster [1983]. They successfully conducted HCCI experiments
with blends of paraffinic and aromatic fuels over a range of engine speeds and dilution
levels in a four-stroke CFR test engine with a variable compression ratio. The intake air
was heated to a high level to achieve HCCI operation and mimic the benefit of high
internal residuals present in two-stroke engines. Ignition and smooth energy release were
obtained by varying the engine operating parameters, such as equivalence ratio, inlet
temperature, and EGR rate. They used global autoignition chemistry and kinetics to
analyze the experimental results. It was concluded that HCCI ignition is controlled by
low temperature (below 950 K) hydrocarbon oxidation (and they recommended the use of

18
a skeletal reaction model proposed by Shell-Thornton Research Labs), and that the
energy release process is controlled by the high temperature (above 1000 K) hydrocarbon
oxidation kinetics as characterized by Dryer and Glassman [1978].

An empirical

equation was also developed based on Dryer and Glassmans global kinetics, and
successfully predicted the average rate of energy release.
An early effort to determine the permissible operating parameters of a four-stroke
HCCI engine was conducted at Southwest Research Institute by Thring [1989] using
gasoline as fuel. Using a Labeco CLR engine, Thring mapped the HCCI operating range
by varying equivalence ratio, EGR rate, engine speed, and inlet temperature. In this work,
HCCI combustion could only achieve stable operations at conditions of low speed and
low load in a four-stroke engine, and the overall HCCI operating range was very narrow.
Diesel engine like fuel economy was achieved under selected conditions (ISFC in the
range of 180 to 200 g/kWh). High EGR rates (in the range of 13 to 33 percent) and high
intake temperatures were necessary for HCCI operation.
It is widely accepted that HCCI combustion is dominated by chemical kinetic
reaction rates [Najt and Foster, 1983], with no requirement for flame propagation. This
notion has been supported by numerous studies, which indicated that the order of radical
formation in HCCI combustion corresponds to that of self-ignition instead of flame
propagation [Noguchi et al., 1979; Oguma et al. 1997]. Experimental [e.g., Shimazaki et
al., 1999] and modeling [e.g., Aceves et al., 1999, 2000, 2001a and b] efforts have also
supported this idea.
Recent chemical kinetics modeling of HCCI combustion has concluded that HCCI
ignition is controlled by hydrogen peroxide (H2O2) decomposition. Hydrogen peroxide is

19
formed as a result of low temperature chemical reactions in the engine charge and at a
high enough temperature it decomposes into two OH radicals, which are very efficient at
attacking the fuel and releasing energy. Hydrogen peroxide decomposition occurs over a
temperature range of 1050-1100 K at the elevated in-cylinder pressures after compression.
This fundamental chemistry of HCCI autoignition and combustion is identical to the
chemistry of knock in spark-ignition engines. With high-octane fuels, little heat is
released prior to this main ignition event at 1050-1100 K; however, with low-octane fuels
(e.g., diesel fuel) significant heat-producing reactions begin at temperatures of about 800
K [Kelly-Zion and Dec, 2000]. Although the amount of energy liberated is too small to
be considered ignition, these low-temperature reactions quickly drive the mixture up to
the 1050-1100 K temperature necessary for H2O2 decomposition and main ignition. It is
this effect that requires HCCI operating parameters to be adjusted with changes in fuel
type [Kelly-Zion and Dec, 2000]. Active radicals (i.e., reactive chemical compounds,
such as H, OH; HO2) present in the exhaust gases do not survive the exhaust and intake
strokes and play a very minor role in starting HCCI combustion; however, partial
oxidation products formed during fuel decomposition can be carried over and, under
proper conditions, sensitize the incoming charge and initiate early pre-ignition reactions.
While the HCCI process has been studied intensively over the past several years,
the chemical mechanisms that control the combustion process are still far from being
completely understood. This statement is based on the observation that no universal and
reliable model has been developed for HCCI prediction, in spite of the tremendous efforts
that have been made towards these objectives. This is not to say that progress has not
been made. Multi-zone detailed chemical kinetic models coupled with CFD codes [Kraft

20
et al., 2000; Aceves et al., 2000] have shown progressively better ability to predict the
heat release rate and the onset of HCCI ignition in engines. While these studies are very
encouraging, they are limited with respect to experimental conditions and fuels, because
the chemical mechanism used in these models were developed under conditions not
directly applicable to HCCI conditions.

The development of chemical mechanism

information for hydrocarbon oxidation under the highly dilute, large percentage of
Exhaust Gas Recirculation (EGR), and pre-heated inlet charge conditions expected in
HCCI engines should improve these chemical kinetic models.

2.2.2

Advantages of HCCI
Relative to SI gasoline engines, HCCI engines are more efficient, approaching the

efficiency of a CI engine due to the following three factors: (1) the elimination of the
throttling losses, (2) the use of high compression ratios (similar to a CI engine), and (3) a
shorter combustion duration (since it is not necessary for a flame to propagate across the
cylinder). HCCI engines also have lower engine-out NOx than SI engines. Although
three-way catalysts are adequate for removing NOx from current-technology SI engine
exhaust, low NOx is an important advantage relative to direct-injection, spark-ignition
(DISI) technology, which is being considered for future SI engines.
Relative to CI engines, HCCI engines have substantially lower emissions of PM
and NOx. Emissions of PM and NOx are the major challenges for CI engines to meet
future emissions standards, and hence controlling these emissions is the focus of
extensive current research. The low emissions of PM and NOx in HCCI engines are a
result of the dilute homogeneous air and fuel mixture in addition to low combustion

21
temperatures. The charge in an HCCI engine may be made dilute by being very lean, by
using Exhaust Gas Recirculation (EGR), or by some combination of the two. Because
flame propagation is not required, dilution level can be much higher than the levels
tolerated by either SI of CI engines. Combustion is induced throughout the charge
volume by compression heating due to the piston motion, and it will occur in almost any
fuel/air/exhaust-gas mixture once the 800 to 1100 K ignition temperature (depending on
the type of fuel) is reached. As combustion occurs, the temperature will rise above the
ignition temperature, but complete combustion can be achieved at temperatures below
those at which significant NOx is produced. In contrast, in CI engines, minimum flame
temperatures are 1900 to 2100 K, high enough to make unacceptable levels of NOx
[Flynn et al., 2000]. Additionally, the combustion duration in HCCI engines is much
shorter than in CI engines since it is not limited by the rate of fuel/air mixing. This
shorter combustion duration gives the HCCI engine an efficiency advantage.
Another advantage of HCCI combustion is its fuel-flexibility. HCCI operation
has been demonstrated for a wide range of fuels [Oguma et al., 1997; Christensen et al.,
1997; Gray and Ryan, 1997]. HCCI engines can operate on gasoline, diesel fuel, and
most alternative fuels, such as methanol, ethanol, LPG and natural gas etc. However,
gasoline is particularly well suited for HCCI operation. High efficiency CI engines, on
the other hand, cannot run on gasoline due to its low cetane number.
HCCI is potentially applicable to virtually every size-class of transportation
engines from small motorcycles to large ship engines which certainly encompasses
automobiles and trucks. In fact, the smallest commercially available engines, those for
model airplanes, are actually HCCI engines [Heywood and Sher, 1999]. HCCI is also

22
applicable to reciprocating engines used outside the transportation sector such as those
used for electrical power generation and pipeline pumping.
If we assume that vehicles with HCCI engines would be 25 percent more efficient
than their non-HCCI counterparts, large reductions in the demand for petroleum are
possible [Epping et al., 2002]. (The 25 percent difference seems reasonable given that
current diesel versions of vehicles use 40 percent less fuel than their gasoline
counterparts).

Even if HCCI engines were to achieve only a 25 percent market

penetration, the savings in oil consumption would be significant. Additional savings may
accrue from reduced refining requirements for fuels for HCCI engines relative to gasoline
for conventional SI technology.
HCCI is a potential low emission alternative to CI engines in light-, medium- and
heavy-duty applications. Even with the advent of effective exhaust emission control
devices, CI engines are currently seriously challenged to meet the future emission
standards.

Although the actual cost and fuel-consumption penalties of CI emission

controls are uncertain, the use of HCCI engines or engines operating in HCCI mode for a
significant portion of the driving cycle could significantly reduce the overall cost of
operation, thus saving fuel and reducing the economic burden of lowering emissions.
SI engines for automotive applications also require intensive design efforts to
improve overall vehicle fuel efficiency. It appears that SI engines will require advanced
NOx emission control devices similar to those being developed for CI engines.
While HCCI engines have several inherent benefits as replacements for SI and CI
engines in vehicles with conventional powertrains, they are particularly well suited for
use in internal combustion engine/electric series hybrid vehicles.

In these hybrids,

23
engines can be optimized for operation over a fairly limited range of speeds and loads,
thus eliminating many of the control issues normally associated with HCCI, creating a
highly fuel-efficient vehicle. In addition to the on-highway applications discussed above,
it should be noted that the benefits of HCCI engines could be realized in most other
internal combustion engine applications such as off-road vehicles, marine applications,
and stationary power applications. The resulting benefits would be similar to those
discussed previously.

2.2.3

The Importance of HCCI Research


Although stable HCCI operation and its substantial benefits have been

demonstrated at selected steady-state conditions, several technical barriers must be


overcome before HCCI engines can be widely used. The main disadvantages of HCCI
and efforts to overcome these technical barriers are briefly listed below:

Hard to control ignition timing and combustion rate


HCCI ignition is determined by the charge mixture composition and its

temperature history (and to a lesser extent, its pressure history). Changing the power
output of an HCCI engine requires a change in the fueling rate and, hence, the charge
mixture.

As a result, the temperature history must be adjusted to maintain proper

combustion timing. Similarly, changing the engine speed changes the amount of time for
the autoignition chemistry to occur relative to the piston motion. Again, the temperature

24
history of the mixture must be adjusted to compensate. These control issues become
particularly challenging during rapid transients.
Several potential control methods have been proposed to adjust operational
parameters for changes in speed and load. Some of the most promising include varying
the amount of hot exhaust gas recirculation (EGR) left in the cylinder after combustion,
using a fuel additive to enhance ignition [Flowers et al., 2000; Olsson et al., 2001; Flynn
et al., 1999], using a Variable Compression Ratio (VCR) mechanism to alter TDC
temperatures [Christensen et al., 1997, 1999; Flynn et al., 1999; Sharke, 2000], and using
Variable Valve Timing (VVT) to change the effective compression ratio and/or the
amount of hot residual retained in the cylinder [Theobald and Henry, 1994; Kaahaaina et
al., 2001; ]. VCR, VVT, and fuel additives are particularly attractive because their time
response can be made sufficiently fast to handle rapid transients.

Although these

techniques have shown strong potential, they are not yet fully proven, and cost and
reliability issues must be addressed.

The possibility also exists to control HCCI

combustion by controlling the temperature, pressure, and composition of the mixture at


the beginning of the compression stroke. In this methodology, thermal energy from
exhaust gas recirculation (EGR) or compression of the inlet charge is used to vary charge
inlet (and subsequent in-cylinder) conditions [Martinez-Frias et al., 2000]. The main
advantage of this method is its simplicity. The disadvantage of this method is that it may
be too slow to react to the rapidly changing conditions that typically exist in
transportation applications. A full transient response analysis of this type of system has
yet to be performed and would depend on the specific system used.

25

High CO and UHC emissions, particularly at lower load conditions


HCCI engines have inherently low emissions of NOx and PM, but relatively high

emissions of hydrocarbons (HC) and carbon monoxide (CO). Some potential exists to
mitigate these emissions at light load by using direct in-cylinder fuel injection to achieve
appropriate partial-charge stratification. However, in most cases, controlling HC and CO
emissions from HCCI engines will require exhaust emission control devices. Catalyst
technology for HC and CO removal is well understood and has been standard equipment
on automobiles for many years. However, the cooler exhaust temperatures of HCCI
engines may increase catalyst light-off time and decrease average effectiveness. As a
result, meeting future emission standards for HC and CO will likely require further
development of oxidation catalysts for low-temperature exhaust streams. However, HC
and CO emission control devices are simpler, more durable, and less dependent on scarce,
expensive precious metals than are NOx and PM emission control devices.

Thus,

simultaneous chemical oxidation of HC and CO (in an HCCI engine) is much easier than
simultaneous chemical reduction of NOx and oxidation of PM (in a CI engine).

Relatively narrow operating range


Although HCCI engines have been demonstrated to operate well at low-to-

medium loads, difficulties have been encountered at high-loads. Combustion can become
very rapid and intense, causing unacceptable noise, potential engine damage, and
eventually unacceptable levels of NOx emissions. Expanding the controlled operation of
an HCCI engine over a wide range of speeds and loads is a big challenge for HCCI.
HCCI starts having NOx problems as load increases ( = 0.5 to 0.6), and will likely

26
require transitioning to conventional operation at high load. Thus, the biggest problem
for HCCI may be the control of the transitions into and out of HCCI.
Preliminary research indicates the operating range of HCCI can be extended
significantly by producing a broad temperature distribution inside the cylinder and/or by
partially stratifying the charge mixture (i.e., SCCI combustion) at high loads to stretch
out the heat-release event. Several potential mechanisms exist for achieving this partial
charge stratification, including varying in-cylinder fuel injection, injecting water, varying
the intake and in-cylinder mixing processes to obtain non-uniform fuel/air/residual
mixtures, and altering cylinder flows to vary heat transfer. The extent to which these
techniques can extend the operating range is currently unknown, and R&D will be
required. Because of the difficulty of high-load operation, most initial concepts involve
switching to traditional SI or CI combustion for operating conditions where HCCI
operation is more difficult. This dual mode operation provides the benefits of HCCI over
a significant portion of the driving cycle but adds to the complexity by switching the
engine between operating modes.

Difficulty with cold start and light load


At cold start, the compressed-gas temperature in an HCCI engine will be reduced

because the charge receives no preheating from the intake manifold and the compressed
charge is rapidly cooled by heat transfer to the cold combustion chamber walls. Without
some compensating mechanism, the low compressed-charge temperatures could prevent
an HCCI engine from firing. Various mechanisms for cold-starting in HCCI mode have
been proposed, such as using glow plugs, using a different fuel or fuel additive, and

27
increasing the compression ratio using VCR or VVT.

Perhaps the most practical

approach would be to start the engine in spark-ignition mode and transition to HCCI
mode after warm-up. For engines equipped with VVT, it may be possible to make this
warm-up period as short as a few fired cycles, since high levels of hot residual gases
could be retained from previous spark-ignited cycles to induce HCCI combustion.
Although solutions appear feasible, significant R&D will be required to advance these
concepts and prepare them for production engines.

2.2.4

Results Using Different Fuels


One of the advantages of HCCI combustion is its intrinsic fuel flexibility. The

literature shows that HCCI can be achieved with a range of hydrocarbons [Oguma et al.,
1997; Christensen et al., 1997, Gray and Ryan, 1997], including gasoline, diesel fuel,
propane, natural gas, and neat or binary mixtures of the SI engine primary reference fuels
(PRF), iso-octane and n-heptane.

HCCI combustion has little sensitivity to fuel

characteristics such as lubricity and laminar flame speed. Fuels with any octane or cetane
number can be burned, although the operating conditions must be adjusted to
accommodate different fuels, which can impact efficiency, as discussed below. An HCCI
engine, in principle, can operate on any hydrocarbon or alcohol liquid fuel, as long as the
fuel is vaporized and mixed with the air before ignition.
The applicability of typical fuels to HCCI engines is discussed below. Other fuels
(methanol, ethanol, and acetone) have also been tried in experiments, but with
inconclusive results.

28
Gasoline:
Gasoline has several advantages as an HCCI fuel, one being a high Octane
Number (ON). ON is used to indicate the resistance of a motor fuel to knock, and it is
based on a scale in which isooctane is 100 ON and n-heptane is 0 ON. ONs are typically
in the range of 87 to 92 in the U.S. and up to 98 in Europe, which allows the use of
reasonably high compression ratios in HCCI engines. Actual compression ratios for
gasoline-fueled HCCI engine data vary from 12:1 to 21:1 depending on the fuel octane
number, intake air temperature, and the specific engine design (which may affect the
amount of hot residual naturally retained). This compression-ratio range allows gasolinefueled HCCI engines to achieve relatively high thermal efficiencies (in the range of
diesel-fueled CI engine efficiencies). A potential drawback to higher compression ratios
is that the engine design must accommodate the relatively high cylinder pressures that
can result, particularly with high engine loads. Additional advantages of gasoline include
easy evaporation, simple mixture preparation, and a ubiquitous refueling infrastructure.
Gasoline is a complex mixture of hundreds of hydrocarbons. While the majority
of research engine tests utilize full-boiling range fuels, often it is desirable to limit the
chemical and/or physical complexity of the fuel to generate insight and understanding
into the underlying fundamental processes.

This parallels the problem with

computational chemistry models of the combustion processes.

The need exists for

models of the chemistry of real fuels; unfortunately, it is currently not possible to


represent the chemistry of all these complex hydrocarbon mixtures with detailed
chemical kinetic models. Consequently, it is advisable to develop computational models

29
for simpler mixtures (validated against experimental data) before moving to the
complexity of real fuels.
The simplest surrogate fuels for gasoline consist of single components, e.g., the
use of iso-octane as a gasoline surrogate. Binary blends of n-heptane and iso-octane, the
octane rating scale primary reference fuels (PRFs), also find wide-spread use as
convenient surrogates for variable RON/MON fuels. Mixtures of these two PRFs are
used to define the octane number (ON) scale, specifically by the volumetric percentage of
iso-octane in the mixture.
Therefore, the present work concentrated on the SI PRFs and their mixtures. nHeptane, which is also used as a representative diesel fuel component, and iso-octane
have quite different oxidation chemistries. Studies show that n-heptane autoignition
occurs in two stages, while iso-octane autoignition happens in a single stage at higher
temperature [Epping et al., 2002]. Further experiments show that HCCI combustion of
PRFs and PRF blends in engines is usually characterized by a two-stage heat release
process due to the separate contributions of low temperature reactions (LTR) and high
temperature reactions (HTR) [Rao et al., 2004].

Research also shows that HCCI

operation with pure n-heptane requires a compression ratio of about 11:1 to phase
autoignition at TDC without inlet air preheating, while iso-octane and high octane
gasoline (RON 98) require compression ratios of 21.5:1 and 22.5:1, respectively
[Christensen et al., 1999].

30
Diesel Fuel:
The HCCI combustion of diesel type fuels can be more easily achieved than with
gasoline type fuels because of diesel fuels lower autoignition temperature. However,
overly advanced combustion phasing can cause low thermal efficiency. In addition,
mixture preparation is a critical issue. There is a problem getting diesel fuel to vaporize
and premix with the air due to its low volatility [Christensen et al., 1999; Peng et al.,
2003]. Therefore, to obtain premixed HCCI combustion using diesel fuel, the air-fuel
mixture must be heated considerably to evaporate the fuel, and the compression ratio of
the engine must be very low (8:1 or lower) to obtain satisfactory combustion, which
results in a low engine efficiency. Alternatively, the fuel can be injected in the intake
port or in-cylinder but, without air preheating, temperatures are not sufficiently high for
diesel-fuel vaporization until well into the compression stroke. This strategy often results
in incomplete fuel vaporization and poor mixture preparation, which can lead to PM and
NOx emissions. However, one concept for direct injection of diesel fuel, involving late
injection (after TDC) with high swirl, has been successful at thoroughly vaporizing and
mixing the fuel before ignition at light to moderate loads. Using this method, diesel-like
compression ratios of 15:1 to 16:1 can be used resulting in high efficiency. This mode of
operation is used in the Nissan MK engine [Kimura et al., 1999 and 2001]. Like gasoline,
diesel fuel has an extensive refueling infrastructure.
The HCCI operation using diesel fuel was extensively tested at Southwest
Research Institute (SwRI) by Ryan and Callahan [1996] and Gray and Ryan. [1997]. For
the first time, Knock Intensity (KI) was used to trace knock and determine the acceptable
HCCI operating range. According to the definition, knock that is just marginally audible

31
is used to define a KI of 5 on a scale from zero to ten. The rate of pressure rise is
measured and used to yield a KI. A KI of 4 was used to identify permissible HCCI
operations. The HCCI operating range was tested by varying EGR rate, compression
ratio, and inlet temperature. They found that management of EGR rate and equivalence
ratio was critical to achieving HCCI. Under 50 percent EGR rate and stoichiometric
fresh charge condition, the engine would produce acceptable power output with near total
elimination of smoke.
A simple empirical model was also proposed by SwRI to predict HCCI ignition
delay time:

td = 0.021*(O2)-0.53*(Fuel)0.05*()0.13*exp(5914/T)

Where td is the ignition delay time (ms), O2 is the oxygen molar density (moles/m3), Fuel
is the fuel molar density (moles/m3), is the density (kg/m3), and T is the air temperature
(K). However, the compression ratio had to be lowered from 16:1 to 8:1 to achieve
HCCI operation, and the unburned hydrocarbons were very high. Also, they found that it
makes little difference whether the dilute mixture was achieved by going very lean (e.g.,
below the equivalence ratio in which a flame can propagate, ~ 0.6 or 0.7) or by adding
exhaust gas recirculation.

32
Propane:
High efficiencies can be achieved with propane-fueled HCCI engines because
propane has a high octane number (105). In addition, because propane is used as a gas, it
can be easily mixed with air. Some infrastructure also exists for propane and it has a high
energy density during storage, as it is a liquid at moderate pressures.

Natural Gas:
Because natural gas has an extremely high octane rating (about 110), natural gas
HCCI engines can be operated at very high compression ratios (15:1 to 21:1), resulting in
high efficiency.

However, similar to gasoline or propane, the engine design must

accommodate the relatively high cylinder pressures that can result. Natural gas is widely
available throughout the U.S.

2.3

Fuel Additives
Fuel additives can be grouped into different categories based on functions, such as

engine performance, fuel stability, and fuel handing and contaminant control. Engine

performance additives discussed here are a class of additives that can improve engine
performance usually by changing autoignition characteristics.
Historically, the study of fuel ignition-enhancing (or suppressing) additives was
motivated by the need for a cetane (or octane) number improver. Cetane Number (CN) is
a rating scale used for diesel engines to indicate the tendency of a fuel to autoignite. The

33
rating compares a fuels performance in a standard engine with that of a mixture of
cetane (CN = 100) and alpha-methyl-napthalene (CN = 0).

While the major source of

diesel fuel has been straight-run distillates separated from crude oil; however, the
increase in market demand for diesel fuel has led to the position where oil refiners are
incorporating more cracked distillates into diesel fuels.

Diesel fuels derived from

cracked distillates generally have a relatively low cetane rating (i.e., poor ignition
quality), as the cracking processes result in higher proportions of aromatic molecules in
the product. Such fuels normally require higher temperatures to ignite than their straightrun counterparts. In diesel combustion, this results in extended ignition delay periods and
faster initial burn and rate of pressure rise, with the consequent effects of greater noise
output and rough running.
In the 1940s and 1950s, a number of investigations into the potential of additives
for diesel fuel ignition quality improvement were carried out [Bogen and Wilson, 1944;
Robbins et al., 1951; Anderson and Wilson, 1952; Brien, 1956; Hurn and Hughes, 1956].
During the period of plentiful petroleum supplies (approximately 1950-1970), there
seemed to be little need for fuels research and the literature reflects this with relatively
few publications. Some of the studies that were performed were engine based [McGreath,
1971; Kamel, 1984] while others were of a more fundamental nature [Salooja, 1962;
Dunskus and Westwater, 1961; Satcunanathan and EI-Nesr, 1972; Kirsch et al., 1981].
However, with the oil crisis of the seventies and the then growing use of cracked distillate
fuels, there was renewed interest in developing suitable ignition promoting additives [Li
and Simmons, 1986; Pishinger et al., 1988; Inomata et al., 1990; Clothier et al., 1990].

34
The effects of additives on knock in SI engine were also studied by Downs et al.
[1951]. Alkyl peroxides, aldehydes and hydrogen peroxide were investigated. The
results demonstrated the key role of alkyl peroxides in the knock process. Formaldehyde,
acetaldehyde, propionaldehyde, and butylaldehyde were added in molar concentrations of
5% or more to a full boiling gasoline. Interestingly, the formaldehyde acted as an antiknock and the other aldehydes were only slightly pro-knock.
As noted, ideally, autoignition in HCCI should occur at the point where the piston
reaches top dead center to provide optimum power and efficiency; therefore the timing of
the autoignition is critical. Without the help of an external triggering event, HCCI has a
problem in controlling the ignition timing. One option for ignition control is to use small
amounts of ignition-enhancing additives to alter the ignition properties slightly.
Although a large number of ignition additives have been shown to be effective in
enhancing the ignition quality of the parent fuel to which they are added, their precise
role in promoting ignition remains uncertain. One school of thought is that they modify
the physical processes contributing to the delay period.

For example it has been

suggested that ignition additives may act in much the same way as water, when the latter
is introduced into diesel fuels in the form of an emulsion. Being at supercritical pressure,
additives may evaporate instantaneously; thereby shattering fuel droplets and assisting
atomization [Valdmanis and Wulfhorst, 1970]. Others claim that additives act as heatflux improvers, considerably increasing the heat transfer rate in nucleate boiling and so
reducing evaporation time [Dunskus and Westwater, 1961; Satcunanathan and EI-Nesr,
1972]. Another school of thought maintains that the main effect of additives is that of
accelerating the autoignition chemistry. Most effective additives are thermally unstable

35
and their thermal decomposition could be expected to yield free radicals. It has been
suggested that these are effective in enhancing the chain branching reactions leading to
ignition [Hurn and Hughes, 1956; Salooja, 1962]. It has also been claimed that the local
temperature rise, caused by the heat release from the thermal decomposition of the
additive may be of equal importance in stimulating the autoignition of the fuel [Inomata
et al., 1990].
Di-tertiary Butyl Peroxide (DTBP), (CH3)3COOC(CH3)3, is one such additive and
it has been suggested as a commercial cetane number improver in diesel engines. DTBP
was selected in this study because (1) it is readily available; (2) it has been used by a
number of previous workers, permitting cross reference to other tests; and (3) its structure
and decomposition mechanism is reasonably well understood.

Like most effective

additives, DTBP is thermally unstable and its thermal decomposition liberates heat and
yields free radicals.
Experiments show that DTBP can reduce ignition delay both in rapid compression
machines for SI PRFs and PRF blends [Inomata et al., 1990; Tanaka et al., 2003] and in
engines for diesel fuels [Ai-Rubaie et al., 1991]. However, it was unclear whether the
effect of DTBP is thermal or chemical. The present work is aimed at determining its the
mode of action and its effects on HCCI operation.
DTBP was adopted as the reference additive and the performance of other
additives and mixtures were compared to it by Ai-Rubaie et al. [1991]. Experiments
show that DTBP can reduce ignition delay both in rapid compression machines for SI
PRFs and PRF blends [Inomata et al., 1990; Tanaka et al., 2003] and in engines for diesel
fuels [Ai-Rubaie et al., 1991]. In rapid compression studies, the oxidation of 1% DTBP

36
alone in air is capable of raising the compressed gas temperature by 35C [Inomata et al.,
1990], and with 2% addition of DTBP to PRF90, the ignition delay time was cut in half
[Tanaka et al., 2003]. Addition of 1% by volume of DTBP to a diesel fuel with cetane
number 40 at an injection temperature of 880 K in an engine caused a 14% reduction of
the ignition delay [Ai-Rubaie et al., 1991].
The experimental investigations of DTBP described above were conducted at
relatively high temperatures and varying conditions in engines, shock tubes and rapid
compression machines. In spite of this work, it remains unclear whether the effect of
DTBP on the lower temperature autoignition processes was thermal or chemical. Thus,
part of the present work was aimed at elucidating DTBPs mode of action. The work was
carried out primarily in a Pressurized Flow Reactor (PFR) which can effectively simulate
the conditions occurring during the critical preignition regime of hydrocarbon oxidation,
and involved measurement of CO concentration, serving as a reactivity marker, at 8 atm
and 650 K < T < 900 K. Measurements were made for seven PRF fuel blends, with and
without DTBP addition. Additional tests were carried out in a CFR engine. The effects
of DTBP on primary reference fuels in the PRF and in an engine are reported in Chapters
4 and 5, respectively.

2.4

Models of Hydrocarbon Oxidation Mechanisms


Although hydrocarbon combustion properties have been studied for over a

century, numerical combustion modeling work did not become an essential part of
combustion research and development programs until the 1980s [Dryer, 1991] with the

37
development and advance of computers.

There have been two distinctly different

approaches to the numerical modeling of hydrocarbon oxidation. One involves explicitly


accounting for all possible chemistry detail and the other elects to account for a minimal
set of features.

Generally mechanisms are classified as detailed, lumped, reduced,

skeletal, or global [Zheng et al., 2004].

Detailed models try to include all of the

important elementary reactions and individual species using the best available rate
parameters and thermochemical data. The other four model types are all driven by the
desire to minimize the model size. Their general characteristics are shown in Table 2-2.

Table 2-2. Categories of chemical kinetic models


(Zheng et al., 2004)
Category

Description

Species

Reactions

Detailed

the latest comprehensive reaction set

100s

1000s

Lumped

uses a lumped description for larger species

100s

1000s

Reduced

a subset of the detailed model

10s

10s- 100s

Skeletal

employs class chemistry and lumping concepts

10s

10s

Global

utilizes global reactions to minimize reaction set

<10

<10

Regardless of the type of mechanism, each reaction requires the associated


reaction rate coefficients and species thermodynamic properties. Accurate estimation of
heats of formation for all radicals and stable species are needed to identify possible
pathways and to estimate activation energies and the rates of reversible reactions. As the
fuels of interest increase in size and complexity, the estimation becomes more difficult.
A detailed description for each category is given below.

38
Detailed Mechanism
A detailed mechanism, as its name implies, includes almost all of the important
elementary reactions and individual species with available rate parameters and
thermochemical data.

As the level of understanding and the size of the molecules

increase, detailed mechanisms become extremely large. A detailed C7 hydrocarbon


mechanism may contain thousands of reactions and hundreds of species [Curran et al.,
1998, 2002]. Coupled with CFD and if used in an engine simulation, such a model would
require tremendous computational resources.
In 1984, Westbrook and Pitz [1984] introduced a comprehensive chemical kinetic
mechanism for the oxidation and pyrolysis of propane and propene. This model was later
extended to lower temperatures [Smith et al., 1985] and to much more complex fuels
[Westbrook and Pitz, 1987]. In a later work by Westbrook et al. [1991], they included
low temperature reaction paths involving alkylperoxy radical isomerization in the
program and examined the chemical kinetic process leading to knocking in spark-ignition
internal combustion engines. Since then there have been efforts to develop detailed
models for butane [Green et al., 1987a, b], pentane [Wang et al., 1999] and heavier
hydrocarbons such as n-heptane and isooctane [Curran et al., 1998 and 2002].

Lumped Mechanism
Lumped mechanisms typically classify the primary propagation reactions of the
parent fuel with a limited set of reference kinetic parameters and group the primary
intermediate isomers into a limited number of lumped components [e.g., Violi et al.,

39
2002; Agosta et al., 2004]. The smaller species are typically treated in a detailed manner
nearly identical to the detailed mechanism.
The size of a lumped mechanism can vary significantly, but for large hydrocarbon
applications they can still encompass thousands of reactions among hundreds of species
[e.g., Granata et al., 2003; Nehse and Warnatz, 1996].

Reduced Mechanism
A reduced mechanism begins with either a detailed or lumped mechanism. Then
the critical reactions and species are selected by one of several reduction methods. The
more useful techniques for automatic reduction are: Quasi-Steady-State Approximation
(QSSA) [Peters and Rogg, 1993], Intrinsic Low-Dimensional Manifolds (ILDM) [Maas
and Pope, 1992], Computational Singular Perturbations (CSP) [Lam and Goussis, 1988],
Directed Relation Graphs [Lu and Law, 2004], and others.

Reduced mechanisms

nominally have tens to hundreds of reactions among tens of species.

Skeletal Mechanism
Skeletal mechanisms are developed from the opposite perspective from the
mechanism categories just described. Instead of starting with an all inclusive detailed
model, they are assembled with just enough of the chemical skeleton to model the
parameters of interest. Normally these mechanisms consist of tens of reactions and tens
of species [e.g., Zheng et al., 2001; 2002a, b].

In skeletal mechanisms, the rate

parameters and thermochemistry are based on classes of reactions.

40
A review of such kinetic models and their applications has been made by Griffiths
[1995]. The earliest skeletal kinetic model, based on degenerate-branched-chain and
class chemistry concepts, was developed at the Shell Thornton Research Center by
Halstead et al. [1975, 1977]. This model consisted of 8 generalized reactions and 5
species with the primary interest being to match the ignition delay behavior, while the
phenomenological complexity of hydrocarbon oxidation, such as cool flames, two stage
ignition and NTC behavior were considered to be of secondary importance. This work
formed the basis for later developments and the model was widely used in engine
applications.
Cox and Cole [1985] developed a skeletal chemical kinetic model consisting of 15
reactions and 10 active species. The model was tested against the ignition data using isooctane and PRF90 in a rapid compression machine.

Hu and Keck [1987] further

developed a skeletal chemical kinetic model of 18 reactions and 13 active species.


Keeping a better representation of the chemical reactions similar to Cox and Cole model,
Hu and Keck model treated exothermicity as enthalpy change in each step reaction. The
rate parameters were calibrated using measured explosion limits in a combustion bomb.
The fuels studied were C4-C8 straight chain paraffins and iso-octane. The effects of fuel
structure are reflected in the rate parameter of the RO2 isomerization reaction.

The

model was applied to predict selected data of ignition delay measured in a rapid
compression machine. Cowart et al. [1990] reproduced the overall trend for ignition
delay of specific hydrocarbons of interest with this model. However, significant physical
features such as preignition fuel consumption, cumulative heat release, and species
concentrations are not included in the model [Li et al., 1996]. These deficiencies were

41
addressed by Li et al. [1992, 1995, 1996] and the basic model was further refined by
Zheng et al., [e.g., 2001, 2002a and b] and in the present work.

Global Mechanism
A global mechanism describes the chemistry in terms of a few principal reactants
and products in a small number of functional relations. Typically, global mechanisms
have fewer than ten reactions among fewer than ten species. These types of mechanisms
are extremely attractive for CFD and other heavy computational applications where
large mechanisms are computationally expensive.
Global models were first developed to describe high temperature chemistry
[Dryer, 1991]. Later, a 4-reaction model [Mller et al., 1992] and a 5-reaction model
[Schreiber et al. 1994] were developed to describe the full temperature regime. However,
neither of these global models can reflect hydrocarbon oxidation behavior in the Negative
Temperature Coefficient (NTC) regime, since NTC behavior inherently involves
intermediate species (for example, HOOH) that provide branching at 900-1100 K. Hence
these two global models were only used to predict ignition delays, and they are not
suitable for prediction of the full HCCI behavior that occurs with PRF fuels. Bourdon et
al. [2004] proposed an optimized 5-step model for HCCI applications. Zheng et al. [2004]
at Drexel University developed a 7 step model to successfully predict temperature,
pressure, ignition delay, combustion duration, and heat release for PRF20 in an engine
operating in HCCI mode. The model includes five reactions that represent degenerate
chain branching in the low temperature region, including chain propagation, termination
and branching reactions and the reaction of HOOH at the second stage ignition. Two

42
reactions govern the high temperature oxidation, to allow formation and prediction of CO,
CO2, and H2O.

2.5

Low and Intermediate Temperature Regime Fuel Oxidation


Generally, the combustion environment, such as temperature, pressure, and

equivalence ratio effects the location of the boundaries between each regime. At one
atmosphere, the hydrocarbon oxidation process can be divided along the following
approximate boundaries:

(1) low temperature, < 650 K


(2) intermediate temperature, 650-1000 K
(3) high temperature, > 1000 K

In engines, fuel spends a relatively long time in the low and intermediate
temperature regime (<1000 K), where it decomposes significantly and generates many
intermediate species prior to autoignition [Smith et al., 1985; Cernansky et al., 1986;
Green et al., 1987 a and b; Leppard, 1987, 1988; Henig et al., 1989]. As illustrated in the
in-cylinder end gas temperature - pressure trajectories presented in Figure 2-2, engine
autoignition, which is associated with knocking, cold start and misfire, is a low and
intermediate temperature phenomenon.

43

1400
CI Engine Inlet
CI Engine Cold Start Autoignition
SI Engine Inlet
SI Engine Autoignition
Adiabatic Engine Trajectories

Temperature [K]

1200
1000

H+

O2

e
nov
T ur

High

800
600
400

Intermediate R + O2 2

Turno

ver

Low

200
10-1

101
100
Pressure [atm]

102

Figure 22. Typical SI engine envelope of end gas temperature and pressure
histories leading up to the point of autoignition (Wang, 1999)

Motored engine experiments are good for studying autoignition, because they can
provide the low and intermediate temperature and higher pressure environment
conditions, in which autoignition takes place. Green et al. [1987a and b] studied the
chemical aspects of autoignition of iso-butane and n-butane using a skip-fired
technique in a single cylinder research engine. They concluded that low temperature
chemistry plays an important role in end-gas autoignition. Leppard [1988] also studied
the oxidation of iso-butane using a motored engine and developed a reaction mechanism.
Later, Leppard [1989] studied more fuels using the same technique and reported that

44
olefins do not exhibit negative temperature coefficient behavior. (Note: a study at Drexel
University found that large olefins do exhhibit NTC behavior [Prahbu et al., 1996]).
At Drexel University, initial engine experiments were conducted by Henig et al.
[1989] using n-butane, iso-butane and blends using the same skip-fired strategy as Green
et al. [1987 a and b] to investigate the effects of fuel structure on autoignition. Products
sampled from the end gas in fired cycles confirmed the importance of low and
intermediate temperature chemistry prior to autoignition and examined the interaction
between n- and iso-butane. The heat release and chemical species in the second motored
cycles were examined in a later investigation by Addagarla et al. [1989a]. Chemical
pathways were discussed based on the species data. Wilk et al. [1990] modeled the
species data using a detailed chemical kinetic model.
Addagarla et al. [1989b] measured the critical inlet fuel/air conditions of
temperature and pressure which induce autoignition for n-pentane, n-hexane, and the
primary reference fuels under motored engine conditions.

Then, based on gas

composition measurements in the engine prior to ignition, Addagarla et al. [1991] studied
the n-pentane mechanism. Filipe et al. [1992a and b] examined the preignition reactivity
and autoignition behavior of several PRF blends under motored conditions.

Time

resolved concentration profiles of fuels and light intermediate species (C4) were
measured. The experimental results indicated that significant amount (up to 40-50%) of
both n-heptane and iso-octane reacted during the cycle.
Li et al. [1994, 1995] conducted experiments in a motored research engine fueled
with neat PRFs, an 87 octane blend of PRFs (PRF87), and PRF87 blended with methyl
tert-butyl ether (MTBE), ethyl tert-butyl ether (ETBE), methyl tert-amyl ether (TAME),

45
diisopropyl ether (DIPE), methanol and ethanol. Detailed evolution profiles of reactants,
molecular intermediates, and products were measured prior to autoignition via in-cylinder
sampling combined with gas chromatographic analysis. The results showed that all of the
ethers and alcohols were effective in reducing preignition reactivity and retarding
autoignition, and mechanistic explanations for the behavior were proposed.
Yang [2002] measured species evolution profiles of PRF20 at different
equivalence ratio, additives, such as 1-pentene and toluene, and major EGR components,
such as CO2 and NO. The results were used to elucidate the chemical kinetics controlling
HCCI operation. As noted previously, Zheng et al. [2001, 2002, 2004] has conducted
experimental and computational studies on skeletal mechanisms for HCCI operation.
Generally, in motored engine experiments, chemical species are typically sampled
and analyzed at selected crank angles. When coupling with air flow and fuel mixing, it is
difficult to get detailed data on fuel oxidation under desired temperature and pressure
conditions. Therefore, it is difficult to deduce and develop reaction mechanisms solely
based on experimental data from motored engines.
To obtain detailed speciation and species evolution data, flow reactors are often
used, especially for those conditions where reactions are very fast. The advantages of
flow reactors can be summarized as follows: (1) the reaction temperature and pressure
can be well controlled; (2) the reaction time can be controlled over a broad range (tens of
milliseconds to a few seconds); and (3) gas samples at different locations along the
reactor, which correspond to different reaction times, are easy to withdraw for analysis.
Therefore, more detailed reaction information can be obtained from flow reactors.

46
In the past 20 years, flow reactors were utilized extensively by many researchers
and their results have significantly contributed to our understanding of combustion
chemistry. Dryer and Glassman [1973] applied the flow reactor and studied the CO and
CH4 oxidation at high temperature. Cohen [1977] studied the mechanism of ethane
oxidation at high temperature. Later, Hautman et al. [1981] used a range of flow reactor
data and proposed a multiple-step overall kinetic mechanism for the oxidation of
hydrocarbons. Callahan et al. [1996] performed experiments to study the oxidation of
primary reference fuels over an initial reactor temperature range of 550 - 850 K and with
a constant pressure of 12.5 atm in the Princeton variable pressure flow reactor. Other
experimental efforts to provide experimental data using flow reactors include work done
by Vermeersh et al. [1991] and Bales-Gueret et al. [1992].
At Drexel University, a pressurized flow reactor has been utilized extensively in
investigating hydrocarbon oxidation chemistry. Koert [1990] designed the flow reactor
system and employed it to examine the effect of pressure on the oxidation of propane,
and then he collaborated on a modeling effort to develop a pressure-dependent kinetic
mechanism for propane based on this experimental data [Koert et al., 1996]. In the same
facility, Wood [1994] studied the oxidation of n-pentane and 1-pentene in the low and
intermediate temperature region.

McCormick [1994] studied the C4 hydrocarbon

oxidation and developed an FTIR technique to analyze the samples taken from the reactor.
Later, Prabhu et al. [1996] investigated 1-pentene oxidation and its interaction with nitric
oxide. Wang et al. [1999] employed the pressurized flow reactor to obtain species
information of neopentane to develop a detailed model.

47
The oxidation and ignition characteristics of pure alkanes (n-dodecane and
isocetane),

naphthenes

(methylcyclohexane

and

decalin),

and

aromatics

(-methylnaphthalene and hexylbenzene) and of their mixtures have been experimentally


studied by Agosta [2002]. The analysis of the interactions controlling the ignition of
binary, ternary and larger mixtures of the compounds listed above has been applied to the
synthesis of a multi-component surrogate for the military aviation fuel JP-8, which is
very similar to the commercial aviation fuel Jet-A. The surrogate has been tailored to
closely match the hydrocarbon distribution in JP-8: a mixture containing 26% n-dodecane,
36% isocetane, 18% -methylnaphthalene, 14% methylcyclohexane, and 6% decalin,
was shown to accurately reproduce the chemical behavior of JP-8 over different
experimental conditions.
Oxidation of samples of JP-8 and Jet-A were experimentally studied by Lenhert
[2004b]. In his study, a 4-component surrogate of JP-8, with mixture of 43% n-dodecane,
27% iso-cetane, 15% methyl-cyclohexane, and 15% -methyl-naphthalene was
developed to match the average JP-8.

Neat, binary mixtures of the components, and

the full surrogate were oxidized in the PFR and stable intermediate and product species
were identified and quantified using permanent gas analyzers, and gas chromatography
with mass spectrometry (GC/MS).

These detailed studies provided kinetic and

mechanistic information in the low and intermediate temperature ranges (600 1000 K)
and at elevated pressures.

Potrebbero piacerti anche