Direct gasoline injection

Chapter 2: Direct gasoline injection

2.1.- New Standards and how to match them

During the 80´s and 90´s, laws all over the world have been developed more and more restrictive concerning emissions in automobile engines. New standards are being prepared in which admitted levels of carbon dioxide emissions are much lower than the ones that can be achieved today. This tighter emission standards are expected in Europe for the year 2000, along with the new limits on CO2 emissions proposed by the European Parliament. The latter will for the first time force manufacturers to rethink the whole engine concept instead of just working on aftertreatment of exhaust gases like done with the present 3-way catalyst. In order to reduce the CO2 emissions there are only two alternative ways of acting:

  • Decreasing the size (displacement) of the engine and, thus, burning a smaller amount of fuel, which would reduce the CO2 emissions. This way, a smaller engine would be obtained and, consequently, a lower performance.

  • Increasing the engine efficiency.

The reduction of engine performance would directly affect the vehicle safety, comfort and driveability and would not be easily accepted by customers, so a way of increasing engine efficiency must be found. This efficiency is a function of engine type and can only be slightly improved for a given type of engine. As concluded in many SAE papers (refs. 1, 2, 3), the conventional indirect injection spark engine, with the fuel injected at low pressure ahead of the intake ports, will not be able to match expected standards for CO2 emissions without greatly reducing present levels of safety, comfort and driveability. This is particularily true for passenger cars having high engine displacement and performance.

In Table 2.1, Phase III CO2 standards proposed by the European Union Parliament are shown. These standards will be applied on passenger cars. They are expressed in g/km for an ECE + UDC cycle and are compared with todays averaged figures (tolerance: ± 20 g/km). The final CO2 levels achieved by the car result from both engine side and vehicle side contributions and have been obtained by assuming a 15% reduction of the emissions on the vehicle side as estimated for the year 2000.

Displacement (litres)

< 1.4

1.4 ~ 2.0

> 2.0

European Parliament




Today's gasoline cars




Today's IDI Diesel cars




Today's DI Diesel cars




Table 2.1: Phase III CO2 standards proposed by the EU (g/km).

It can be easily seen that even assuming the predicted reduction on the vehicle side, only DI Diesel engines having a displacement bigger than 2 litres will match future laws.

Therefore, a new type of engine should be used in the future. The direct injection gasoline engine, in which fuel is directly injected into the cylinder, seems to be the clearest solution for this problem and will allow to meet both CO2 and emission standards without a loose of performance due to its two basic advantages:

  • Being a stratified charge engine, it is able to run at lean-burn conditions (high air to fuel ratio) with a better combustion optimization and reduction of pumping losses.

  • The direct in-cylinder injection of the fuel allows to increase the air/fuel ratio control during transient running conditions.

2.2.- Developing a DI Gasoline engine

The DI gasoline engine is not a new idea in the thoughts of engine engineers. In the past there were several attempts to manufacture one of this engines, but technological problems have always placed the DI gasoline engine under the condition to chase the evolution of the conventional gasoline engine, either undirectly injected or carburetted. The first example of a DI gasoline engine close to today´s concept was Ford´s PROCO (PROgrammed COmbustion) (ref. 4) in the second half of the seventies. However, the evolution of the conventional IDI engine, along with the massive adoption of 3-way catalyst technology was able to follow the progressive severization of emission standards, so any further effort was not needed. Nowadays, several car manufacturers in Japan, like ISUZU with a stoichiometric DI gasoline engine concept, MAZDA (ref. 5), TOYOTA (ref. 6), MITSUBISHI and NISSAN, as well as in USA, like FORD have research programs running on this direction.

Another interesting approach to a DI gasoline engine is currently being made by ORBITAL for both a two and four stroke cycles. This comnpany proposes air assisted direct injection, in which compressed air is used to assist the atomisation of fuel injected directly into the combustion chamber, in order to fulfil emission standards with a variety of 2 stroke engines with displacements ranging from 800 cm3 to 2000 cm3 (ref 7). This way a very fine control of the A/F ratio around the spark plug is achieved, while DI into the chamber (typically after transfer and exhaust ports have closed) also more effectively eliminates short circuiting than is possible with manifold and port fuel injection systems.

In Europe, car manufacturers ( VOLKSWAGEN, MERCEDES-BENZ, OPEL, RENAULT, FIAT...) too joined a consortium for studying the feasability of a DI gasoline concept. As can be seen on Table 2.2, the total demand of passenger cars in Europe in the year 2010 will be 15.3 million, of which 40% will be equipped with conventional gasoline 3-way catalyst engines, 35 % with Diesel engines and 25% with DI gasoline engines. Thus, the market will demand nearly 4 million of this kind of engines by the year 2010. The data displayed on this table were predicted at the Second Annual Review meeting held in Brussels in December 1994.

Figure 2.1: Expected output and consumption for a DI gasoline engine in comparison to other types of engines (Source: MISUBISHI Motors Co.)







Passenger cars


15.30 (+13%)



Light duty trucks





Heavy duty trucks





Table 2.2: Trend of vehicle population (million) in Europe according to the Second Annual Review meeting (Brussels, December 1994, ref. 10).

Apart from the availability of DENOx catalyst, three main problems are found concerning the realisation of a DI gasoline engine:

1.-Optimized injection system that is designed also taking into consideration future production constraints. This was found to be the major obstacle to the past attempts concerning a DI gasoline engine. In fact, up to the eighties, the gasoline injection systems were always derived from diesel fuel injector systems; the excessive control rigidity and the spray characteristics of these systems have proved unsuitable for application on gasoline engines and specially on stratified charge engines. Once this problem was found, several research programs addressed to the development of specific injection systems started, like the japanese NICE (refs. 8 and 9). Nowadays, several systems specially developed for DI gasoline engines can be found at prototype level. However, they will have to be rethinked in such a way they are able to meet, at the same time, performance and future production/service requirements. The main objective of the present “Diplomarbeit“ is analysing the overall behaviour of one of these prototype injection systems developed by a major injection system supplier. During the tests and measures performed with them, the manufacturer continuously received information referring to their product´s behaviour and they introduced modifications in their design. The measurements concentrated on the analysis of spray formation and evaporation behaviour. The influence of turbulence was as well analysed as the ignition behaviour.

2.-Engine System. Associated to direct injection in gasoline engines is the present concept of stratified charge (ref. 10) that is really the technological step forward for achieving reduction in consumption (along with CO2 emission reductions that are not possible to achieve with conventional internal combustion systems). Stratification means a combustion system able to supply, at all engine running conditions, the proper motion, generated by the intake manifold and combustion chamber, to the injected fuel. In order to realize its fuel economy potential at partial loads, the direct injection spark ignition engine should be operated unthrottled in a extremely lean condition by distinctively stratifying the charge and by preparing a rich air-fuel mixture around the spark plug. For that purpose, many of the direct injection systems proposed so far have adopted the basic configuration of locating the spark plug gap near the fuel spray cone. Some manufacturers (like HONDA, TOYOTA and CHRYSLER) tried to develop a lean burn indirect injection spark ignition engine in which mixture was led to the spark plug via an extra small valve, but DI seems to be a better sollution for stratified charge burning. It has been confirmed stable combustion can be realized by such configurations. Ignition fouling caused by the inpingement of the liquid fuel on the spark plug, however, prevented such a concept from being developed into mass production. Although higher energy ignition systems have sometimes shown the effects of reduced ignition fouling, problems associated with the deterioration of the spark plug durability due to the higher ignition energy have not been solved.

A second part of the present “Diplomarbeit“ will deal with ignition, therefore, a spark plug will be adapted to the high pressure chamber. The ignition behaviour of actual injectors will be analysed.

In order to achieve its higher performance potential at high loads, the DI gasoline engine should be operated under stoichiometric or slightly rich conditions. When charge is stratified, soot is generated in the rich zone, so sufficient excess of air should be provided around that zone in order to burn the soot generated. Therefore, when the average mixture strength is stoichiometric or slightly rich, that is, when the equivalence ratio is larger than unity, the mixture should be homogenous so as to supress formation of soot. With a DI system, homogenuous mixture can be obtained by an early injection.

In other words, at partial loads, the goal is to prepare a rich gaseous mixture around the spark plug and at high loads it is to prepare a homogenuos mixture. To achieve this goals, the combustion chamber geometry has to be optimized as a function of the given injection system and the in-cylinder air motion has to be controlled ,for example, by using variable valve actuation or port throttling. Furthermore, the exhaust aftertreatment has to be fully integrated in the engine system, even a new type of catalyst has to be developed that is able to work at lean conditions.

It is obvious that the whole engine concept is changed when introducing DI and the development of every system should be carried out along with that of the rest of the engine.

3.- Control Strategy. A strategy able to manage engine control parameters, such as air and fuel, in an integrated way and during transient engine operations has to be developed. The control strategy will harmonize the different requirements of low fuel consumption, satisfactory engine performance and vehicle driveability, and high conversion efficiency of the aftertreatment system. It will also manage technical innovations such as variable valve actuation.

2.3.- ADIGA Program

With the aim of solving these problems, a new Consortium was created in which first order car and components manufacturers from several countries take part. This Consortium, based on the experience taken out of a former project called GADI, has started several programs to research in the different topics needed for this new concept of engine. The main objective of the project is the realization of research prototype demonstrators of the DI gasoline engine which, tested at vehicle level on the European ECE+UDC driving cycle, will be able to show:

  • more than 15% CO2 reduction with respect to the today´s IDI gasoline engine (i.e. 3% lower than the proposed EU Parliament limits shown in Table 2.1).

  • Pollutant reduction down to 1.5 g/km of CO and 0.2 g/km of HC + NOx. Nowadays limits in the European Union for Spark ignition engines are 2.2 g/km for CO and 0.5 g/km for HC + NOx. In fact, these predicted emission figures would match even the strongest enviromental laws that are applied today in California (2 g/km for CO and 0.47 g/km for HC + NOx).

Another objective of the project is to supply information to the European Commission as a prenormative action for delivering Phase III standards concerning the emission level that could be achieved with a Direct Injection spark ignition engine.

This consortium has launched six different projects that intend to develop the proper technology in order to stablish the potential of DI gasoline and Diesel engines to achieve very low CO2, CO, HC, NOx, PM and noise emissions. The mentioned six programes are the following:

1/A.- ADvanced Valve Control system (ADVACO)

2/A.- Integrated electronics for dynamic emission control (ELSEC)

3/A.- Advanced Direct Injection GAsoline engine (ADIGA)

1/B.- DI Diesel NOISE and vibration control technologies (DINOISE)

2/B.- Integrated afterTREATment of DI Diesel engines for cars (DITREAT)

2/B.- Advanced DI Diesel with high pressure injection control (ADDI)

The first three projects form a cluster addressed to the developement of an advanced DI gasoline engine with ultrahigh fuel economy, while the last three form another cluster addressed to the development of an ultralow pollutant and noise emitting Diesel engine to be mounted on passenger cars. Even if the projects of the two clusters are structured in a modular manner so that they can be run independently of each other, a deep interchange exists among deliverables of the different projects in order to meet the final result consisting in DI engine prototype demonstrators to be tested at vehicle level. This way, a comparison can be made between the two different types of internal combustion engines in terms of their overall CO2, CO, HC, NOx, PM and noise potential.

Within the ADIGA program, the injection and combustion systems development will be performed. The first generation of injector prototypes was delivered the manufacturer to the LAT for testing in the second semester of 1996. This system followed the guidelines for the engine design given by the project coordinator. This guidelines were defined according





CO2 reduction (%)


> 15

With respect to the mean of the best IDI gasoline engines

HC + NOx emissions (g / km)

< 0,5

< 0,2

Mass producibility (engine / day)



Compatibility with the most advanced mass prodution techniques

Durability (km)

> 150.000 (1)

> 150.000 (2)

(1) with injector maintenance.

(2) without injector maintenance

Reliability (M.T.B.F)

= IDI engines

> IDI engines

M.T.B.F. (Mean Time Between Failures)

Table 2.3: Main objectives for a DI gasoline engine. Emissions in terms of carbon dioxide, unburned hydrocarbons, nitrogen oxides and carbon monoxide are referred to the European ECE + EUDC test cycle.

to in-house research programs already made by car manufacturers partners of the project, information from the previous GADI project and the most recent technological objectives shown on Table 2.3. Even higher CO2 reduction (down to 22% instead of 15%) could be achieved if cylinder deactivation should be found convinient. However, an increase in NOx emissions should be expected in this case.

In Table 2.4 it is shown how every one of the major tasks involved in the project will affect emissions. RELATIVE CO2 and NOx reduction figures are referred to each previous step, while TOTAL CO2 and NOx figures consider base engine levels of CO2 = 220 g / km and NOx = 3.5 g / km, which are the achievable levels nowadays for a 1.4 to 2.0 litres engine. As far as CO2 is concerned, an equivalent reduction is expected from the vehicle improvement in order to approach the 150 g / km proposed by the European Union Parliament.

2.4.- Engine design Guidelines

With all this information, the basic line for researching on injection and combustion systems will be developed. Requirements for the injection system will be defined with reference to a common rail high pressure liquid fuel injection system characterized by the following figures:

  • Conical spray pattern: hollow cone with an angle of 30° - 75°

  • Fuel flow of 30 mg / ms

  • Linearity between time and injected fuel in the fuel delivery range of 8 - 60 mg / (cycle

  • Descargar
    Enviado por:Fernando Lopez Velasco
    Idioma: inglés
    País: España

    Te va a interesar