Emission potentials of future diesel injection
systems
Dr.-Ing. Hermann Breitbach, Dr.-Ing. Joachim Schommers, Dr.-Ing. Marco
Stotz, Dipl.-Ing. Martin Schnabel
Daimler AG, Stuttgart
Emission potentials of future diesel injection systems
Abstract
In recent years, the diesel engine has enjoyed tremendous success. Driven by its excellent char-
acteristics, the market share went up well above 50% in newly registered cars in Europe. How-
ever, further development is necessary to improve the diesel engine in the next years to keep it
attractive. Emissions and at the same time fuel consumption and noise need to be further re-
duced, while engine power has to go up.
For Mercedes-Benz key steps to reach these goals are reduction of engine displacement, possi-
bly reduction of the number of cylinders for a given output, and the reduction of engine speed
for given vehicle speed. To realise these steps, modern diesel engines will have to further un-
dergo an evolution towards higher boost pressures, together with lower compression ratios,
higher exhaust gas recirculation rates and better EGR cooling. As in the past, the fuel injection
system will also play a key role in the future diesel engine development. It will have to support
combustion systems with new and improved characteristics. In order to operate with signifi-
cantly increased boost pressures, the kinetic energy of the spray will play a more important role
than ever. Also, the ability for robust multiple injection patterns and stable injection quantities
over lifetime will be more important than in the past. High spray energy has to be achieved
especially for small injections and for part load operating points with low pressures. Therefore,
the needle opening and closing velocities are of special importance.
The historical evolution of the diesel engine correlates strongly with fuel injection system de-
velopments. Mercedes-Benz contributed significantly to the recent success of the diesel engine,
being one of the first car manufacturers to introduce a modern common rail diesel engine in the
Mercedes C220 CDI in 1997. Based on this past experience with fuel injection system, Mer-
cedes-Benz has evaluated a new fuel injector concept with regard to the above mentioned char-
acteristics.
The concept analysed is a direct actuated piezo injector, where the needle is driven directly
from an actuator without servo-hydraulic amplification. It was compared to a state of the art
servo-hydraulically driven injector. The hydraulic investigations show an excellent performance
of the direct driven injector in all criteria. The analysis of the concepts on engine confirmed the
superior results with regard to emissions, engine power and noise. A detailed analysis of the
combustion process reveals the physics behind these positive results. With its excellent rate of
injection shape, its robustness and stability, the direct driven injector is an attractive device,
creating excellent conditions to meet the ambitions development targets of future Mercedes-
Benz diesel engines.
3
1. Introduction
Injecting fuel into the diesel engine has already been a challenge in the early days of diesel
engine development. Rudolf Diesel for example used gasoline, being evaporated thermally,
then compressed and intercooled, before it was injected into the combustion chamber. The
large compressor and the dangerous thermal evaporation and compression of the gasoline va-
pour were reasons for the diesel engine initially being restricted to stationary or marine applica-
tions.
The invention of the prechamber diesel engine by Prosper L’Orange and the development of
fuel injection pumps and –nozzles allowed to replace the complex compressors and diesel en-
gines were increasingly used in mobile applications. 1923, the first diesel locomotive was put
into service and also, the first diesel truck with a prechamber engine was presented by
Benz&Cie. In parallel, Bosch started the development of fuel injection pumps, leading to a
serial production of pumps and nozzles in 1927. Improved pumps with higher fuel injection
pressures, combined with centrifugal governors and better nozzles paved the way to the first
application of the diesel engine in a passenger car: the Mercedes-Benz 260D, presented 1936 in
the automobile exposition in Berlin.
Especially in the last 40 years, diesel engine progress was rapid through technologies like tur-
bocharging, introduced in serial production in 1978 by Mercedes-Benz, further increase of
injection pressures, direct injection for passenger cars, or external exhaust gas recirculation.
Gas exchange was enhanced through the introduction of charge air cooling and cooled EGR.
Complex mechanical and later electrical components allowed higher flexibility in fuel injection
timing and quantity. However, engine speed and fuel injection pressures were still closely cou-
pled, since both radial and axial piston pump and also unit injector piston speed and thus pres-
sure were related to engine speed. Multiple injections were possible, but needed complex me-
chanical devices and were limited in number and flexibility.
In the late 80’s, Mercedes-Benz started as one of the first vehicle manufacturers development
of engines with fuel injection systems, that decoupled the components for injection quantity and
timing control from the components generating pressure: common rail systems. They allowed to
choose injection pressures to a certain extent independent from engine speed and to adjust the
injection timing over a wide range and very flexible, since injection pressure was constantly
available. Fast solenoid valves in the fuel injectors enabled accurate and fast metering depend-
ing on engine operating condition, and multiple injection over a wide timing range.
The introduction of the first production common rail system in the Mercedes C220 CDI in 1997
was an almost revolutionary step in the development of the diesel engine [3]. Common rail
systems allowed to increase power and torque of diesel engines, while at the same time, drive-
ability and NVH were significantly improved and emissions decreased [5].
However, despite this significant progress, challenges remain, as for example the reduction of
the diesel engine NOx-emissions. Here, Mercedes-Benz is again playing a key role with its
BLUETEC-technology. Besides the after treatment, a key part of this technology is the NOx-
reduction of engine out emissions [1]. With both NOx-storage catalyst and selective catalytic
reduction technologies, lower engine out emissions allow a more effective NOx-after treatment,
Emission potentials of future diesel injection systems
for example through reduction of the number or regenerations of the NOx-storage catalyst, or
reduction of the AdBlue additive consumption.
Also, despite the usage of particulate filters, particulate emissions from the engine remain im-
portant, to avoid high loading of the particulate filter and subsequently a high number of filter
regenerations.
In view of the recent CO2-discussion, the future diesel engine development will have to focus
even more on fuel consumption. While with regard to CO2, the diesel engine is still superior to
most other powertrain concepts, gasoline or gasoline hybrid powertrains are going to catch up.
Thus, the diesel engine has to further improve to keep its market share. Also, power and torque
output need to increase, and NVH behaviour needs further refinement. As in the past, the fuel
injection system will again play a key role in all these challenges. Thus, it is a key question,
how future fuel injection systems need to look like, to foster diesel engine development with
regard to fuel consumption, emissions, reliability and NVH. Based on its long experience with
common rail systems [4, 9], Mercedes-Benz investigated this question in basic research, fuel
injection test benches and single and multi-cylinder studies [10, 11], comparing a state of the
art injector with a new fuel injector concept.
2. Fuel injection system requirements for modern diesel
engine combustion
To identify key requirements for the fuel injection system, a detailed analysis of modern diesel
engines and their combustion systems is necessary. Most remarkable is the strong increase in
torque and power of recent diesel engines. Fig. 2.1 shows on a time scale from the beginning of
the last century up to today, how specific power output is going up to values of around 70 kW/l.
Also, key enablers for such progress are identified. The rapid increase of power in the last years
is, besides fuel injection system improvements, mainly due to the development of turbo charg-
ing and intercooling. Both lead to a significant increase of charge air through boost pressure
and air density. That allows - in conjunction with improved mixture formation and thus lower
fuel-air ratios - a significant increase of injection quantities and thus power and torque. This
trend was supported by measures in engine mechanical structure, allowing higher peak pres-
sures and temperatures, and the enhancement of engine thermodynamics, like lowering the
compression ratio, motivated by power, emission and NVH-considerations [7].
5
Fig. 2.1: Development of diesel engine power output over time with key enabling technolo-
gies
Fig. 2.2: Emission benefits for low compression ratio
Fig. 2.2 shows, how engine emissions are influenced by a decrease in compression ratio. At
constant fuel consumption and combustion noise, NOx and particulate emissions decrease sig-
Emission potentials of future diesel injection systems
nificantly, however, hydrocarbon and carbon monoxide emissions go up slightly, since peak
combustion temperatures go down with lower compression ratio.
Principally, the increase in power and torque can serve different purposes: besides using it for
vehicle performance, it can also be used to downsize the engine in order to improve fuel econ-
omy. For the diesel engine, this mainly works through lower engine mechanical friction, while
increased engine load hardly has an influence, since for diesel engines, part load efficiency does
not significantly improve with load. Therefore, combining downsizing of displacement with a
lower number of cylinders is especially effective, since friction can be decreased substantially.
The increase in torque can be used to lower engine speed for a given vehicle speed by changing
overall engine to wheel transmission ratio (downspeeding). Fuel economy benefits again result
from lower friction, since for a given vehicle driving profile, the engine speed profile on aver-
age is lower. Also, lower engine speeds are positive for NVH.
Both downsizing and downspeeding create the same new boundary conditions for the fuel in-
jection system: engine load and subsequently charge density go up, while, as described above,
piston bowl size increases due to lower compression ratio. Under these conditions, the kinetic
energy of the spray has to significantly increase in order to achieve good fuel air mixture forma-
tion, leading to both good spray penetration into the piston bowl and excellent atomisation of
the spray.
To create high kinetic spray energy, it needs two main parameters: high rail pressure and low
pressure loss between rail and nozzle. For given high pressure pump capacity and injection
quantity, the achievable rail pressure is determined by the injector return flow. The higher the
return flow, e.g. due to fuel quantities to operate the servo hydraulic circuit or due to concept
related injector leakage, the lower the maximum pressure.
Two aspects have to be considered with regard to the pressure loss between rail and injector.
The first one is the steady state pressure loss due to restrictions inherent to the injector concept
or due to filters, small bores or tube diameters. For example, some injectors need a flow orifice
to reduce pressure around the needle seat when the injector is injecting, in order to facilitate the
closing of the injector.
The second aspect is the pressure loss due to transient effects, especially during opening and
closing of the injector [2]. While opening and closing, pressure loss is a combination of flow
restriction through injection hole and needle seat. With the needle opening, the main pressure
loss occurs over the needle seat and hardly any over the nozzle holes. Only at significant stroke,
the pressure loss over the nozzle hole becomes predominant. Finally, at full stroke, the needle
seat pressure loss is typically a few percent of overall pressure loss in the nozzle holes, and the
fluid pressure is almost fully transformed into kinetic spray energy.
The overall kinetic energy in the spray is determined by the integral of the pressure loss during
opening and closing and the steady state flow restrictions while fully open. Only considering
the transient pressure losses, the kinetic energy in the spray could be optimised, if opening and
closing speed of the injector were infinite. Then, only steady state pressure losses would have
an effect. Fig. 2.3 shows, with opening and closing speed as a parameter, how the actual kinetic
energy from an injector relates to the maximum possible kinetic energy with infinite needle
speed. Two examples for fuel quantities of 30 mm3 and 7 mm3 have been simulated. As can be
seen, the influence of opening and closing speed is lower for larger fuel quantities, since the
steady state flow phase is longer, the influence for small injected quantities is significant, be-
7
cause the entire injection is essentially only a brief needle opening followed by immediate nee-
dle closing. Full lift is not even achieved.
For the smaller injected quantity, the slowest injector does only reach about 65% of the theo-
retically possible energy, while the fastest injector achieves around 85%. The slowest injector
would need 20% higher rail pressure to generate a similar spray kinetic energy. With this, it
becomes clear, that opening and closing speed have a similar effect as pressure has, and fast
needle speed becomes a key requirement for the combustion concepts discussed above.
Fig. 2.3: Effective kinetic spray energy in relation to maximum theoretical kinetic spray en-
ergy for different injector opening and closing speeds
Minimum injected quantity is another requirement, which has been discussed extensively in
recent years. However, now, with compression ratios and pressures, and thus gas temperatures
going down, the conditions for pilot quantity ignition become less favourable. To compensate,
the pilot injection quantity has to go up. Also, with the piston bowl growing in size, the fuel-air
mixing of the main injection is enhanced leading to a higher amount of premixed fuel-air mix-
ture. Also, the ignition conditions worsen due to high amounts of EGR. Both effects lead to a
higher demand in pilot quantity, in order to enhance ignition conditions for the main injection
by warming up the gas in front of the main injection.
Due to these physical effects, the minimum injected quantity goes up, the requirement for very
small quantities is less important. Smallest quantities are now expected to be around 1.5 to 2
mm3 for medium to high injection pressures.
On the other hand, the requirement for multiple injections becomes more important. Combus-
tion systems with up to two pre-, a split main- and two post injections are thermodynamically
advantageous. The post injections are especially important for operating modes like particulate
filter or NOx-storage catalyst regeneration. With five and more injections, the requirement for
Emission potentials of future diesel injection systems
stable injection quantities with regard to both stability over lifetime and stability for different
injection patterns becomes obvious [6].
Figure 2.4 shows, how stability for varying injection patterns can differ depending on injector
concepts. For identical pulse width, the quantity of the main injection varies with the pilot to
main injection time gap. The black dashed and the blue dotted curves show the main quantity
for a state of the art injector before and after model based correction respectively. The full red
curve finally shows the quantity for a hydraulically optimised injector. It is obvious, that using
an optimised injector can provide a more stable environment for engine calibration than even
sophisticated corrections for a less stable injector.
Fig. 2.4: Deviation of main injection quantity with variation of pilot injection timing
The following analysis compares two injector concepts to each other: a state of the art servo-
hydraulic injector and a newly developed concept with direct needle actuation. The injectors
are first analysed hydraulically with regard to the above criteria, then evaluated on engine. The
concepts are shown in Fig. 2.5. Figure 2.5.1 shows an injector, where the actuator is controlling
a pressure reservoir directly above the needle. This servo-hydraulic actuator is located in the
injector body, close to the pressure controlled volume. Since the injector diameter is usually
restricted to 19 mm in the cylinder head, the actuator has to be small, such that it can be inte-
grated into the injector body. On the market, this type of injector concept is available from
various suppliers with both solenoid and piezo actuators.
The concept in Fig. 2.5.2 shows an injector with the needle being directly actuated through a
piezo actuator [8]. This concept moves the needle with significantly higher speed, because it
does not have time delays through a servo-hydraulic circuit, slowing the other concept down
especially at low rail pressures. However, compared to the forces required for the servo valve,
this actuator has to deal with significantly higher forces directly from the needle that can only
be dealt with by a piezo actuator.
9
Fig. 2.5: Injector operating concepts
3. Hydraulic characterization of injector concepts
The basic behaviour of the injector concepts is compared in Fig. 3.1, showing the injected
quantity vs. the electrical pulse width supplied to the injector.
Fig. 3.1: Injection quantity as function of energizing time
Emission potentials of future diesel injection systems
The direct actuated injector does open significantly faster for all pressures, however, with a
steep quantity vs. pulse width gradient for small quantities. The servo-hydraulic concept shows
a lower gradient for the small quantities. Usually, for servo-hydraulic injectors, this is prefer-
able, because it leads to more accuracy and stability of pilot injections. For a direct actuated
injector, however, there is a paradigm shift. Despite the steep gradient in pilot quantity over
pulse width, small quantities can still be injected with high accuracy and stability, because the
operating principle does not rely on the servo-hydraulic circuit any more and the robustness of
quantity to external influences is significantly decreased.
As discussed above, the mixture formation is heavily influenced by the rate of injection and
subsequently, the kinetic spray energy. Fig. 3.2 compares the rate of injection for both injector
concepts. The direct actuated injector shows a much steeper rate of injection gradient due to its
high needle opening speed. As discussed above, this leads to a much lower pressure loss of the
spray in the opening and closing phase, thus higher spray velocities and kinetic energy.
Fig. 3.2: Rate of injection for the injector concepts
Inherent to the design are significant differences of the injector concepts with regard to the
overall fuel volume. The servo-hydraulic actuator has a long, small dia