Fuel injection

Malcolm Latarche, in Pounder's Marine Diesel Engines and Gas Turbines (Tenth
Edition), 2021

Developments and trends

Fuel injection systems have a significant influence on the combustion process and
hence a key role to play in improving engine fuel consumption and reducing noxious
exhaust emissions. The following characteristics of an injection system are desirable in
achieving these goals:

•
Injection pressures during the whole process should be above 1000–1200 bar for a
good spray formation and air–fuel mixture; a tendency in practice to 1600–1800 bar and
higher is noted.

•
The total nozzle area should be as small as possible in relation to cylinder diameter for
good combustion, particularly at part load.

•
Total injection duration should be 20° of crank angle or less for achieving a minimum
burning time in order to exploit retarded combustion for reduced NOx emissions without
loss in efficiency. A high compression ratio is desirable.

•
High pressures at the beginning of injection promote reduced ignition delay, while
increased mass flow can result in overcompensation and increased pressure gradients.
Consequently, rate shaping is necessary in some cases, particularly with high-speed
engines.

•
High-aromaticity fuels cause increased ignition delay in some cases. Preinjection with
high injection pressures is necessary and can achieve nonsensitivity to fuel quality.
•
Electronically controlled adjustment of injection timing should be applied for optimized
NOx emissions at all loads, speeds, and other parameters.

•
The load from the torque of the injection equipment on the camshaft and/or the gear
train should be as low as possible in order to prevent unwanted additional stresses and
noise.

•
For safety reasons, even a total breakdown of electrical and other power supplies
should not result in the engine stopping.

To understand fully the complex hydraulic events during the fuel injection process, the
Swiss-based specialist Duap says that it is essential to appreciate the function of all the
elements (pump, pressure pipe, fuel valve, and nozzle) forming the injection system.

It is commonly believed by nonspecialists that the plunger within the pump pushes the
fuel upwards like a pillar, thus affecting the immediate injection of fuel into the
combustion chamber. The reality, however, is that the plunger moves with such a high
speed that the fuel likely to be conducted is highly compressed locally (elasticity of fuel).
The compressed fuel now generates a pressure wave, which runs through the pipe and
valve, causing the nozzle to open and inject. The pressure wave of the fuel is forced
through the system at a speed of around 1300 m/s. In the same way that sound in the
air is reflected by houses and hills, the wave is reflected between the fuel valve and the
pump. It may therefore easily run back and forth between those components several
times before the nozzle is actually forced to open and inject.

Duap warns that this process underlines the importance of maintaining the injection
system in an excellent condition and ensuring that the fuel is properly treated and free
of dirt. If, for example, the nozzle spray holes are partially blocked by extraneous
elements or carbon particles, the pressure wave may not be sufficiently reduced within
the system. This eventually results in the destruction of the fuel pump cam or other vital
parts of the injection system upon the next stroke.
The task of the injection system is to feed fuel consecutively to each cylinder within a
very short period of time (0.004–0.010 s, depending on the engine revolutions per
minute). It is also essential that the same amount of fuel is delivered with each stroke:
deviations in the quantity supplied to different cylinders will adversely affect the
performance of the engine and may result in crankshaft damage due to resonance.

All of the key elements of the system must be manufactured to very small tolerances;
the clearance between plunger and barrel, for example, is not larger than 4–16 μm
(depending on the plunger diameter). Such a high-precision clearance also dictates an
adequate surface quality (within 0.2–0.5 μm), a finish which can be achieved only by
careful grinding and lapping.

Before the fuel is pushed into the pressure pipe (linking the pump and fuel valve), it has
to pass the pressure valve, which has several tasks. The first task is to separate the
pump hydraulically from the pressure pipe after the fuel has passed the valve; the
second is to smooth the pressure wave running back and forth within the pipe: this
calming down is necessary to secure the proper closing of the fuel valve without having
additional and uncontrolled injection, and the third is to maintain a certain pressure
within the pressure pipe for the next injection stroke (rest pressure). The fulfilment of
these tasks can be guaranteed only by using pressure valves manufactured to very tight
tolerances (bearing in mind that the valves play an important role in ensuring that
identical amounts of fuel are delivered to each engine cylinder).

Another highly underestimated component of the fuel injection system is the pressure
pipe. Despite its small size and apparent simplicity, Duap notes that the pipe has to
endure pressure waves of up to 1800 bar; even the smallest mark may thus lead to
fracture.

The last link in the injection system chain is the fuel valve (nozzle holder assembly and
nozzle), which is designed to inject and atomize the fuel into the combustion chamber
when the pressure wave has reached a predetermined strength.

Spray stability from VCO and a new Diesel nozzle design concept
N. Mitroglou, ... D. Arcoumanis, in Fuel Systems for IC Engines, 2012
INTRODUCTION
Fuel injection systems for modern direct-injection Diesel engines play a distinct and
paramount role that is well recognised as a means of controlling their performance and
meeting the ever more stringent emission regulations. A variety of nozzle designs and
engine optimisation injection strategies are employed by electronic common-rail
injection systems that cover the wide range of operating conditions of modern Diesel
engines. Increasing injection pressure, piezo-controlled needle valves and multiple
injections are among the methods explored and known to improve combustion and
engine performance. However, under such operating conditions, cavitation phenomena
become the dominant and frequently uncontrolled flow characteristics that affect the
durability of the fuel system and the properties of the near-nozzle spray. Investigations
over the last few years have demonstrated that Diesel injector nozzles cavitate [1–3]
under typical operating conditions, a fact that complicates further the already complex
design of high-pressure Diesel injection systems. As demonstrated in [4–6], two distinct
forms of cavitation have been identified inside injection nozzles: geometric-induced and
vortex or string cavitation. Geometric-induced cavitation is the most common form of
cavitation in fuel injector nozzles and it has become gradually better understood. It
initiates at the hole inlet corners due to the abrupt acceleration of the fuel flow as it
enters the nozzle holes. On the other hand, string or vortex cavitation structures have
been observed in the bulk of the liquid inside sac, mini-sac and VCO nozzles [7–9].
Vortex cavitation is also commonly found in propellers, hydraulic turbines and hydrofoils
as explained in [10–13]. Recent studies have also confirmed similar behaviour in multi-
hole nozzles for direct-injection gasoline engines and low-speed two-stroke Diesel
engines [7,14–17].

Cavitation can be linked to undesirable effects such as sharp reduction in engine

performance, increase in noise and vibrations, as well as surface erosion [15]. In most
cases of practical interest, cavitation bubbles survive until the nozzle hole exit [4].
Therefore, it is generally accepted that cavitation promotes fuel atomisation, which is
generally believed to be desirable for enhanced air fuel mixing. However, some studies
[18] have indicated that cavitation may also be associated with hole-to-hole and cycle-
to-cycle variations. Appropriate design of the inlet hole curvature and non-cylindrical
shape of injection holes as shown in [19,20], alter cavitation inception and development
characteristics; at the same time, careful system optimisation is found to achieve
reduction of engine exhaust emissions while maintaining similar performance standards
[7,15]. Until recently, these cavitation regimes have been addressed independently.
However, there is evidence [5] that geometric-induced and vortex cavitation possibly
interact in various ways. Additionally, it has been concluded that inception of string
cavitation is dramatically enhanced and owes its existence to sources of vapour already
present inside the nozzle volume [8] even at locations where the static pressure of the
fluid is well above its vapour pressure.
The present paper represents a continuation of the work presented recently in [8,17,21].
The investigation reported here focus on the effect of string cavitation on the near
nozzle spray dispersion angle both, in large-scale and real-size transparent nozzle tips.
Acquired data cover a number of fully transparent nozzle replicas, as well as real-size
nozzle tips that provide unobstructed optical access of the nozzle sac volume, injection
holes and the emerging sprays plumes. Investigated designs include 6-hole nozzle
configurations of a VCO and a new prototype nozzle design that promises to eliminate
string cavitation and produce steady spray shapes. Then the various results obtained
are presented while the most important findings are summarised at the end.

Air Pollution from Small Two-Stroke Engines and Technologies to Control It

Yuji Ikeda, ... Eran Sher, in Handbook of Air Pollution From Internal Combustion
Engines, 1998

13.3.4 Air-Assisted Fuel Injection

Fuel injection systems are indispensable to the improvement of two-stroke engines to
enhance their advantages in automotive engine applications. There are numerous
reports of injector developments [35–42] but very few contain sufficient information
related to the detailed spray droplet characterization. Spray and injection systems have
been thoroughly researched, especially in diesel engines. The two-stroke engine
involves complex processes, such as the scavenging process, cyclic variation, and
misfiring, which are closely related to pressure wave propagation and reflection.
Although the scavenging process has been the key feature in the development of two-
stroke engines [20,22–24,43–46], very little experimental evidence is available to
explain the relationships between gasoline spray vaporization, mixture formation, and
the scavenging process [47–54].

For small two-stroke engines, direct fuel injection has been considered as a way to
solve the problems of incomplete combustion and excessive HC concentration in
exhaust gases. In particular, air-assisted fuel injection was developed as a powerful tool
in the creation of a more combustible fuel-air mixture at lean combustion conditions. Air-
assisted injection uses compressed air to atomize fuel in the injector and enhance the
penetration of small droplets. Many different kinds of injector mechanisms have evolved
around the world. The spray formation of an air-assisted injector is dominated by the
assisting airflow, so the droplet dispersion process and its atomization should be
understood, together with droplet dynamics.
Laser diagnostic tools like laser sheet [55], exciplex [56], and LDV [14] can provide
information regarding spray angle, spray shape, penetration, vapor region, and so on,
but detailed spray information, such as the droplet diameter distribution and its velocity
in a two-dimensional plane, has not yet been obtained. A visualization technique can
supply sufficient spatial, but very poor temporal, information of spray characteristics. A
phase Doppler anemometer (PDA) can measure the diameter of a droplet and its
velocity with very high spatial and temporal resolution, but this is a single-point
measurement method. An alternative technique is required to determine a two-
dimensional spray image with detailed droplet characteristics.

In this section, the usefulness of Sauter mean diameter (SMD) [57,58] in a periodic
injector is proven and droplet size classes are implemented in order to better
understand the momentum transfer between liquid and gas phases.

The air-assisted injector used in this experiment was a commercial injector for a two-
stroke marine engine with over 22 kW (30 ps) per cylinder, as shown in Figure 13.21.
Fuel is first injected into a cavity and the air injector is operated by opening a poppet
valve. The air-fuel ratio can be controlled by changing the opening period of the valve
when the pressure difference between the air and the fuel is set at a certain level. The
injector has a 36-mm-long, straight pipe before the valve, where preatomization is
carried out. Air-assisted fuel is injected from the 5-mm diameter poppet valve.

Sign in to download full-size image

Fig. 13.21. Air-assisted injector.

(reprinted with permission from SAE)

A dry solvent with a refractive index of 1.427 was used as fuel instead of gasoline. The
specific density of the dry solvent is 0.77 g/cm3, which is very similar to that of gasoline
(0.7–0.8 g/cm3). The scattering angle of 68° was determined by the first-order refraction
angle [59]. For vector measurements, the one-component LDV was used by changing
the incident beam angles of ± 45°.

Direct photographs of the injected spray are shown [60] in Figure 13.22. It is clear that a
mushroom vortex is induced by shear stress at the spray shell. The spray tip velocity
calculated from these pictures is about 64 m/s. A YAG laser sheet was used to take a
two-dimensional spray image, as shown in the same figure. These shots are direct
pictures at a certain cycle. It is well known that there is cycle-to-cycle variation in this
kind of air-assisted injector. The figure also shows two pictures at different cycles at the
same timing. These photographs indicate the importance and need to analyze the spray
by a two-dimensional image having high temporal resolution, as laser sheet
visualization cannot provide time variation information and diameter information. Single-
point measurements do not reveal the cycle-to-cycle variations and spatial structure
variations. However, using single-point measurement with ensemble averaged data, the
two-dimensional image of the spray can be demonstrated with its spatial structure as
shown [61] in Figure 13.23. The Sauter mean diameter (SMD) and the relevant velocity
vectors are also shown.

Sign in to download full-size image

Fig. 13.22. Images of injected spray structure.

(reprinted with permission from SAE)

Sign in to download full-size image

Fig. 13.23. Droplet velocity vectors and SMD.

(reprinted with permission from SAE)

Spatial droplet dispersion is best explained using planar information sources such as a
photograph or a laser sheet image. The PDA technique provides single-point
information, but a phase-locked ensemble averaged method can demonstrate a two-
dimensional image, as shown in Figure 13.23. The spray axisymmetry was checked by
measuring at opposite locations up to r = –3 mm. This figure shows the SMD variation
and its spatial structure versus time. The vector length was calculated as droplet
trajectory length within 0.25 ms, and the color represents the SMD. The maximum SMD
was set to be 130 micron.

At 1.6 ms after the injection signal, which was used as the assisted air drive signal, the
first droplet was observed on the axis. After 0.25 ms, the spray tip velocity reached
approximately 65 m/s and droplet dispersion in a radial direction was observed. The
spray tip velocity of 65 m/s was almost the same as that calculated from the direct spray
picture. The SMD was about 25 micron at the spray tip. On the center axis, droplet
direction was parallel to the axis, while droplet direction at the spray shell region was
over 45 degrees in a radial direction.

At 2.3 ms, the spray tip velocity on the axis increased and the next droplet from the
nozzle formed a larger-sized droplet group. The region penetrated by the droplets
resembled an umbrella. The small and fast droplets existed until 2.8 ms. After 2.8 ms,
the spray tip velocity decreased and the SMD enlarged near the center axis. The larger
droplets caught up and collided with the smaller droplets and, hence, the diameter
began to increase. The droplets of spray at the outer region had lower velocity due to
the strong shear flows, and the direction of the droplets then showed a wavy spray
structure. The very large droplet near the nozzle, which is red in color, was formed at
2.875 ms, when the spray tip droplet size was 30 micron.

Furthermore, spray droplets influenced by turbulent air tended to follow air motion but
large droplets with a high momentum penetrated the highly turbulent flow regions, such
as the recirculation flow areas. These droplet dynamics could not then be demonstrated
by Sauter mean diameter alone but require other advanced methods such as size-
classified analysis.

Four size-classified droplet velocity vectors are shown frozen at 2.875 ms in Figure
13.24. It is clear that there is a shear flow-induced mushroom vortex formation in the
small droplet regions. At the spray tip, small droplets show larger velocity gradients than
larger droplets. The larger droplet vectors have straighter and narrower injection angles.
There are no droplets larger than 30 μm in the spray shell region.

Sign in to download full-size image

Fig. 13.24. Size-classified droplet dynamics at 2.875 ms.

(reprinted with permission from SAE)

The spray angle of each size class and the momentum decay should be quantified in
order to understand evaporation processes and mixture formation. The air motion and
turbulent energy intensity profiles are shown in Figure 13.25. The large turbulent energy
area revealed by the dark area in the figure indicates the presence of a strong shear
flow region. At an early injection period, the larger spot is located at the center of the
axis. The dark area appears at the spray shell region in the next stage. The slip velocity
vector shows the large vector angle at the strong shear region.

Sign in to download full-size image

Fig. 13.25. Airflow motion, turbulent kinetic energy, and slip velocity of small droplet.

(reprinted with permission from SAE)

The spray characteristics of an air-assisted gasoline injector were investigated by phase
Doppler measurements. A summary of the foregoing results follows.

A two-dimensional planar image of size-classified droplets was used to demonstrate the

spatial structure of spray formation. It was found that the Sauter mean diameter is not
the best representation value at the acceleration region, and that the size-classification
technique is very useful for understanding detailed spray characteristics. The slip
velocity and relative Reynolds number were implemented to show the momentum
transfer region due to strong drag force. A mushroom vortex was formed by the strong
shear flow at the spray shell and consisted of small droplets of 10 to 20 μm. A double
spray tip structure was found near the nozzle, which diminished rapidly with distance.
Droplets larger than 30 μm penetrated almost straight through to downstream. It was
found that this spray animation can be a most powerful tool in understanding
momentum transfer processes.

Medium and heavy duty diesel fuel injection system requirements to meet future
emissions legislation
S. Daum, ... H. Theissl, in Fuel Systems for IC Engines, 2012

SUMMARY
Modern fuel injection systems are developing in parallel with aftertreatment systems
and the choice of aftertreatment technology is affecting the demands placed on the fuel
injection system. For Euro V, engine solutions varied from the use of SCR alone to
engines using cooled EGR without particulate filters. This meant that on one engine the
peak injection pressure on engines using SCR aftertreatment could be as low as 1600
bar, while on engines using cooled EGR, the injection pressure requirement was as
high as 2200 bar. Predictions were made for injection pressures to increase in the
direction of 3000 bar.

For Euro VI, EPA10, JNPLTR and Stage IVb/Tier 4 emissions, the development of more
efficient SCR systems and an effective combination of different aftertreatment systems
is relaxing the demand for high injection pressures. Based on the available
aftertreatment technologies and their combination, the peak injection pressures required
to meet future emissions regulations are not likely to exceed 2500 bar in the near future.

As future emissions legislation is likely to concentrate on maintaining current levels

while reducing CO2 and fuel consumption, measures to reduce parasitic losses and
improve combustion efficiency will gain in emphasis.

Fuel systems will continue to have to be flexible, especially with regard to multiple
injection capabilities. As injection quantities in the smaller injections reduce, so the
accuracy of the injection, with regard to speed and repeatability of needle opening and
closing will increase to ensure that injection to injection and injector to injector variability
is minimised. This is likely to require systems which control the injection nozzle needle
more directly than some heavy duty systems currently do.

Transient control of injection pressure and/or rate will also be increasingly important for
future generations of fuel injection systems as they offer potential benefits in engine out
emissions and noise.

Fuel quality is unlikely to improve in the future and lubricity could decrease. Currently a
wear scar diameter of 460 μm is given in EN590, but increases to 520 μm are likely.
This will also influence the design of fuel injection systems as they will have to cope with
these as well as fuels sourced from renewable sources.

Control systems functionality and power will also increase, not only as the functionality
of the injection system increases, but also the requirements for controlling the total
engine (including aftertreatment as well as on board diagnostics) also grow.

The automobile
Tariq Muneer, Irene Illescas García, in Electric Vehicles: Prospects and Challenges,
2017

Direct injection
Port fuel-injection systems long ago replaced carburettors in cars because of their
efficiency and lower maintenance requirements. But now, automakers are moving
towards the even more efficient direct injection. With port fuel injection, gasoline is
sprayed into the intake manifold, where it mixes with air, and then is sucked down into
the cylinders. Direct injection places an injector on each cylinder, spraying gasoline into
the cylinder itself. Direct injection leads to more efficient engines, as the gasoline-air
mixture burns more completely. The individual injectors also ensure that each cylinder
gets the same amount of fuel, and the spray can be more precisely timed.

Doxford Low-Speed Engines

Doug Woodyard, in Pounder's Marine Diesel Engines and Gas Turbines (Ninth Edition),
2009

Fuel Injection
The fuel injection system operates on the common rail principle in which timing valves,
operated by cams on the camshaft, control the injection of fuel from a high-pressure
manifold through spring-loaded injectors to the cylinders (Figure 14.12). Fuel is
delivered to the high-pressure main by the multi-plunger pump fitted at the after end of
the engine (Figure 14.13), the pressure being maintained at the desired value by means
of a pneumatically operated spill valve.

Sign in to download full-size image

Figure 14.12. Common rail fuel system

Sign in to download full-size image

Figure 14.13. High-pressure fuel pump
Two fuel injectors fitted to each cylinder open when the timing valves are operated by
the cams on the camshaft. The duration of opening of the timing valves and hence the
period of injection of fuel into the cylinders is controlled by the governor. Either a
centrifugal governor with a hydraulically operated output or an electronic governor with
a pneumatic actuator could be fitted. In the former case the governor input is set
pneumatically from the control station; in the latter case it is set electrically. In this way
the speed of the engine can be adjusted as required. Means of adjusting the setting of
the timing valves by direct mechanical linkage are also provided. The timing of injection
is determined by the positions of the timing valve cams. Only one cam is required for
each cylinder for both ahead and astern running.

Overview of gasoline direct injection engines

H. Zhao, in Advanced Direct Injection Combustion Engine Technologies and
Development: Gasoline and Gas Engines, 2010

1.4 High-pressure fuel injection system

The fuel injection system is a key component of the DI gasoline engine. It should have
the capability of providing both late injection for stratified charge combustion at part load
and early injection during the intake stroke for homogeneous charge combustion at high
load operations. The homogeneous charge operation requires a well-atomised and
evenly dispersed fuel spray with early injection at low in-cylinder pressure. For stratified
charge operation, a well-atomised but compact and repeatable spray pattern is
desirable to achieve rapid mixture formation and controlled stratification.

The key technology enabler for the modern DI gasoline engine is the development of
electronically controlled fuel injection systems. Until the 1990s, mechanical pump–line–
nozzle fuel injection systems had been used with fixed injection timings and single
injection. Initially developed for two-stroke direct injection gasoline engines, the
solenoid-actuated electronic high-pressure fuel injector became available in the late
1980s and was soon adopted for the development of four-stroke direct injection
gasoline engines. As shown in Fig. 1.2, the high-pressure fuel system for direct injection
gasoline engines comprises a high-pressure pump driven directly by one of the
camshafts, which supplies the pressurised fuel to a common rail mounted in the cylinder
head, and high-pressure electronic fuel injectors.

The first generation of modern DI gasoline engines are designed with a wall-guided
combustion system. The high-pressure solenoid injectors are mostly of swirl-type
design, as shown in Fig. 1.7, which features an inwardly opening pintle and a single exit
orifice (e.g. Hentschel et al., 1999). The liquid emerges from the single discharge orifice
as an annular sheet that spreads radially outwards to form a hollow-cone spray.
However, the spray pattern from such an inward-opening swirl injector undergoes
significant changes with the injection pressure, the ambient pressure or density, and the
injector operating temperature. At the designed injection pressure (between 50 and 100
bar) and elevated ambient density during the late injection of the stratified charge
operation, collapse of the hollow-cone spray will occur, forming a narrow spray
envelope with an increased spray penetration. As a result, the structure of spray from
the swirl injector changes substantially over the operational range of the in-cylinder
density and fuel rail pressure, leading to significant difficulty in the optimisation of
stratified charge operations over a wide range of part load conditions.

Sign in to download full-size image

1.7. Direct injection gasoline injectors in production.

One of the major limitations of first-generation DI gasoline engines with wall-guided

combustion systems is the requirement for strong in-cylinder charge motion, such as
tumble or swirl. Either a high tumble intake port or a helical port is required to produce
the charge motion, which often leads to reduced volumetric efficiency and hence lower
full-torque performance. In order to reduce the reliance on a variable flow-control
system, a high-pressure slit-type injector was used in the Toyota second-generation DI
gasoline engine (Keanda et al., 2000). The slit-type injector has a single rectangular
orifice and its slit is positioned to produce a fan-shaped spray either on-axis or off-axis.
The ratio of the length and width of the rectangular slit can be adjusted to generate a
range of nominal fan included angles. Employment of the slit-type injector was reported
to provide both an improved torque curve and a wider range of stratified charge
operations.

In order to achieve a spray-guided stratified charged combustion system at part load

and improve the full load performance of the direct injection gasoline engine, the
solenoid-actuated multi-hole nozzle injector and piezo-electrically actuated injector have
been developed and put into mass production vehicles. A major advantage of the multi-
hole injectors is that any spatial fuel distribution pattern can be produced in principle by
the number of holes, including the angle or angles of the spray pattern, on or offset from
the injector axis. A portfolio of multi-hole injectors can therefore be designed to suit the
combustion system optimisation needed, as shown in Fig. 1.7. However, as pressure
atomisation is the only mechanism to generate the fuel droplets, a relatively higher
injection pressure (c. 150 Mpa) is normally employed to produce good-quality
atomisation. In addition, the small nozzle hole diameter and the higher charge
temperature of spark ignition combustion increase the tendency for the injector hole to
become blocked by soot deposits. The injector should therefore be placed in a region
where the injector could be well cooled to below 130 °C to prevent the formation of a
soot deposit.

The outward-opening injector, as shown in Fig. 1.7, in comparison can effectively

eliminate blocking of the injector nozzle by soot through its outward opening pintle.
Moreover, the initial liquid sheet thickness of the spray is directly controlled by the pintle
stroke. As a result, the outwardly opening injector allows the spray angle, penetration
and droplet size to be controlled.

Piezo-electric actuation utilises the rapid dimensional change in certain ceramics when
subject to an electric field. The rapid opening and closing times allow a significant
reduction in the minimum opening period and more fuel injection taking place at full
pintle lift. The variation in the opening characteristics from actuation to actuation is also
superior with the piezo actuator. The ability to operate with much shorter injection
duration with repeatable actuation dynamics and fuel quantity leads to a substantial
improvement in the dynamic range and the working flow rate of the injector. The
extended dynamic range and greater flow rate are pre-requisite for the development of
boosted DI gasoline engines and engines that can operate with both alcohol and
gasoline fuels. Moreover, the fast piezo-electric actuated injector allows multiple
injections per cycle to be utilised.

Table 1.1 summarises the main characteristics of the three main types of injectors as
discussed above. Both solenoid-actuated multi-hole injectors and piezo-electric
actuated outward-opening injectors are now in series production with high-pressure
pumps supplying the fuel pressure up to 20 Mpa (Stach et al., 2007; Achleitner et al.,
2007).

Table 1.1. Comparison of three types of DI gasoline injectors

Characteristics Multi-hole injector Piezo outward opening injector Swirl/inward

opening injector
Flexibility of spray pattern + + + +
Inclined spray axis + – +
Atomisation quality at 10 MPa – + +
Spray shape dependence on back-pressure ++ ++ –
Homogeneous mixture preparation with early injection ++ +
Flow rate and dynamic range – ++ –
Multiple injection + ++ +
Robustness against fouling – ++ +
Low cost electronic fuel injection for 2 and 3 wheeled motorcycles
J. Allen, ... G. Farmer, in Innovations in Fuel Economy and Sustainable Road Transport,
2011

2 Application of PCI on 125 cc 4-Stroke Motorcycle

As discussed in [ref 4,5,6] there are a number of Pulse Width Modulated (PWM)
pressurised fuel injection systems currently being developed for use on small engines.
Typically these systems are adapted or derived from the technology used in automotive
systems. This means the systems comprise a high pressure fuel pump and pressure
regulator to control the fuel pressure at an accurate preset value and an injector
(housed in the throttle body) which controls the flow of fuel into the engine by a variable
duration (width) pulse.

Shown on the left-hand side in figure 3 is the fuel handling system of a conventional
PWM system, with its complex and hence expensive multi component layout. In
comparison the Pulse Count Injection (PCI) system shown on the right-hand side has a
simple feed and vent line directly from the tank to the PCI injector (housed in the
throttlebody) with no other parts required, keeping the overall system cost to an
absolute minimum.

Sign in to download full-size image

Figure 3. Comparison of fuel components between PWM and PCI systems.

Figures 4 and 5 show the installation of the PCI engine management system on
demonstration vehicles including a 125 cc motorcycle, a 250 cc 3 wheeled utility vehicle
and a 50 cc scooter. In all cases the injector is located within a fuel chamber integrated
directly into the throttle body. This arrangement allows the free flow of fuel direct from
the fuel tank to immerse the injector and gives a free return of fuel vapour back to the
tank. This free flow of fuel and vapour ensures the injector is well supplied with liquid
fuel even under extreme heat conditions such as hot soak conditions without the need
for expensive fuel pumps and without the need for high fuel pressures. Figure 5 also
shows the air bypass and fuel mixing system used to ensure good fuel atomisation.

Sign in to download full-size image

Figure 4. Photo of the PCI engine management system fitted to 2 and 3 wheeled
vehicles.

Sign in to download full-size image

Figure 5. Injectors and main components in position on the PCB and Sectional view of
integrated throttle body and PCI engine management system.

Table 1 shows a comparison of Euro 3 emission test cycles run at a UK emission test
facility. The PCI and PWM systems are direct comparisons with the same motorbike
and catalyst being used with the two electronic fuel systems. The table shows that both
electronic fuel injection systems are well capable of passing Euro 3 emission limits, with
the PCI system emitting less than 23% of the Euro 3 limits on all 3 measured pollutants.
Both electronic systems deliver quick starting and smooth riding characteristics as you
would expect from a well calibrated electronic engine management system, with virtually
identical fuel consumption figures being achieved in real world UK urban driving (2.798
l/100 km for the PWM system and 2.747 l/100 km for the PCI system [ref 8]). Although
these numbers are real world results with some degree of inherent variability they are
both generated by the same driver on the same drive route, being driven in a typical UK
driving manner. As a comparison data from [ref 9] indicates that the fuel injected 125 cc
vehicle is slightly better than the carburetted vehicle when compared on the Indian drive
cycle (1.41 l/100 km EFI compared to 1.42 l/100 km carburetted). These numbers also
indicate the very different figures achieved using different driving cycles.

Table 1. Comparison of Euro 3 emission results between PCI, PWM.

Standard Honda Limits of Euro 3 % Limits

CO 0.6122 2 30.61
NOx 0.04797 0.15 31.38
HC 0.10012 0.8 12.515
Scion Sprays SS2 Limits of Euro 3 % Limits
CO 0.46097 2 23.0485
NOx 0.03413 0.15 22.75333
HC 0.11887 0.8 14.85875
2.1 ECU and ignition control
The PCI fuel injection system also has full ignition control integrated into the ECU, as is
expected on full engine management systems. As well as basic optimised ignition
mapping additional spark control can be manipulated under certain conditions to enable
for example; idle speed control, easy start, or fast catalyst light-off, to further improve
the emissions, fuel economy and engine smoothness. The control software used in the
PCI controller implements a software based crank decoder and ignition output. This
allows a lower cost, non-automotive, ECU to be selected [ref 10]. To achieve the
required performance the system currently uses a 32-bit Atmel ARM7 processor. As
application requirements change there is further potential to optimise processor
selection and cost by moving up and down the processor range. The microprocessor
software is MISRA-C compliant in-house software with K-Line diagnostics functionality.
The interface software, which is also in-house developed, gives a MS Windows
compatible system with very intuitive and easy to use calibration functionality.

The effect of closely coupled pilot injections on diesel engine emissions

A. Weigand, ... A. Schwarte, in Internal Combustion Engines: Improving Performance,
Fuel Economy and Emissions: IMechE, London, 29–30 November 2011, 2011

ABSTRACT
Modern common rail fuel injection systems allow for very sophisticated injection
strategies to facilitate the compromise between low emissions, fuel consumption and
combustion noise. One such injection strategy is the reduction of the dwell time
between injections to extremely low values.

In general the Diesel mixture formation is governed by the geometry of injection nozzle
and combustion bowl. Common rail injection systems additionally offer the possibility to
inject the fuel in a train of discrete injection pulses (multiple injection). The spacing
between the pulses and the individual fuel mass per pulse allow for a “radial
stratification” of the injected fuel, and this can be used to favourably control engine
emissions and combustion noise.

The current work focuses on optimum injection strategies as a function of engine speed
and load within the NEDC. For this, two operating points were assessed, one
representing the city driving and the other one standing for the acceleration to 120 kph
motorway speed. For these points especially the “minimum dwell”, MD, injection
strategies were investigated for a number of operating conditions. The low engine load
and speed regime was explored at fully warmed up engine as well as for cold start and
warm-up, in particular with respect to the reduction of unburnt hydro carbons, HC, and
carbon monoxyde, CO. The high engine load and speed “acceleration regime” to 120
kph was studies for two operating provisos. One was that of purely engine internal
reduction of nitrous oxides, NOx, via high EGR rates and retarded centre of combustion,
HR50, the other one that of NOx reduction mainly from exhaust gas aftertreatment
(DeNOx), which allows for an advance of the centre of combustion. This improves the
thermodynamic efficiency to its best, but poses the challenge of high combustion noise,
CN. The reduction of combustion noise can be achieved by suitable injection strategies
and good results were found with MD.

The MD injection strategy yielded good results throughout the engine map, compared to
conventional dwell injection patterns, and it was attempted to elucidate the mechanisms
causal for this advantage.