Robert Bosch GmbH_Gasoline-Engine Management Basics Components

The Bosch Yellow Jackets Edition 2001 Technical Instruction Gasoline-engine managementOrder Number 1 987 722 036 AA/PDI-02.01-En

2001

The Program Order Number ISBN

Automotive electrics/Automotive electronicsBatteries 1 987 722 153 3-934584-21-7Alternators 1 987 722 156 3-934584-22-5Starting Systems 1 987 722 170 3-934584-23-3Lighting Technology 1 987 722 176 3-934584-24-1Electrical Symbols and Circuit Diagrams 1 987 722 169 3-934584-20-9Safety, Comfort and Convenience Systems 1 987 722 150 3-934584-25-X

Diesel-Engine ManagementDiesel Fuel-Injection: an Overview 1 987 722 104 3-934584-35-7Electronic Diesel Control EDC 1 987 722 135 3-934584-47-0Diesel Accumulator Fuel-Injection System Common Rail CR 1 987 722 175 3-934584-40-3Diesel Fuel-Injection Systems Unit Injector System/Unit Pump System 1 987 722 179 3-934584-41-1Radial-Piston Distributor Fuel-Injection Pumps Type VR 1 987 722 174 3-934584-39-XDiesel Distributor-Type Fuel-Injection Pumps VE 1 987 722 164 3-934584-38-1Diesel In-Line Fuel-Injection Pumps PE 1 987 722 162 3-934584-36-5Governors for Diesel In-Line Fuel-Injection Pumps 1 987 722 163 3-934584-37-3

Gasoline-Engine ManagementEmission Control (for Gasoline Engines) 1 987 722 102 3-934584-26-8Gasoline Fuel-Injection System K-Jetronic 1 987 722 159 3-934584-27-6Gasoline Fuel-Injection System KE-Jetronic 1 987 722 101 3-934584-28-4Gasoline Fuel-Injection System L-Jetronic 1 987 722 160 3-934584-29-2Gasoline Fuel-Injection System Mono-Jetronic 1 987 722 105 3-934584-30-6Spark Plugs 1 987 722 155 3-934584-32-2Ignition 1 987 722 154 3-934584-31-4M-Motronic Engine Management 1 987 722 161 3-934584-33-0ME-Motronic Engine Management 1 987 722 178 3-934584-34-9Gasoline-Engine Management: Basics and Components 1 987 722 136 3-934584-48-9

Driving and Road-Safety SystemsConventional Braking Systems 1 987 722 157 3-934584-42-XBrake Systems for Passenger Cars 1 987 722 103 3-934584-43-8ESP Electronic Stability Program 1 987 722 177 3-934584-44-6Compressed-Air Systems for Commercial Vehicles (1): Systems and Schematic Diagrams 1 987 722 165 3-934584-45-4Compressed-Air Systems for Commercial Vehicles (2): Equipment 1 987 722 166 3-934584-46-2

Æ

• EGAS electronic throttle control• Gasoline direct injection• NOx accumulator-type catalytic converter

Gasoline-engine managementBasics and components

Tech

nica

l Ins

truc

tion

Gas

olin

e-en

gine

man

agem

ent:

Bas

ics

and

com

pone

nts

The

Bos

ch Y

ello

w J

acke

ts

Automotive Technology

Published by:© Robert Bosch GmbH, 2001Postfach 300220,D-70442 Stuttgart.Automotive Aftermarket Business Sector,Department AA/PDI2Product-marketing, software products,technical publications.

Editor-in-Chief:Dipl.-Ing. (FH) Horst Bauer

Editors:Dipl.-Ing. Karl-Heinz Dietsche, Dipl.-Ing. (BA) Jürgen Crepin.

Authors:Dipl.-Ing. Michael Oder(Basics, gasoline-engine management,gasoline direct injection),Dipl.-Ing. Georg Mallebrein (Systems forcylinder-charge control, variable valve timing),Dipl.-Ing. Oliver Schlesinger (Exhaust-gasrecirculation),Dipl.-Ing. Michael Bäuerle (Supercharging),Dipl.-Ing. (FH) Klaus Joos (Fuel supply,manifold injection),Dipl.-Ing. Albert Gerhard (Electric fuel pumps,pressure regulators, pressure dampers),Dipl.-Betriebsw. Michael Ziegler (Fuel filters),Dipl.-Ing. (FH) Eckhard Bodenhausen (Fuel rail),Dr.-Ing. Dieter Lederer (Evaporative-emissionscontrol system),Dipl.-Ing. (FH) Annette Wittke (Injectors),Dipl.-Ing. (FH) Bernd Kudicke (Types of fuelinjection),Dipl.-Ing. Walter Gollin (Ignition),Dipl.-Ing. Eberhard Schnaibel (Emissions control),in cooperation with the responsible departmentsof Robert Bosch GmbH.

Translation:Peter Girling.

Unless otherwise stated, the above are allemployees of Robert Bosch GmbH, Stuttgart.

Reproduction, duplication, and translation of thispublication, including excerpts therefrom, is onlyto ensue with our previous written consent andwith particulars of source. Illustrations, descrip-tions, schematic diagrams and other data onlyserve for explanatory purposes and for presenta-tion of the text. They cannot be used as thebasis for design, installation, and scope of deliv-ery. Robert Bosch GmbH undertakes no liabilityfor conformity of the contents with national orlocal regulations. All rights reserved.We reserve the right to make changes.

Printed in Germany.Imprimé en Allemagne.

1st Edition, September 2001.English translation of the German edition dated:February 2001.

� Imprint

Robert Bosch GmbH

Gasoline-engine managementBasics and components

Bosch

Robert Bosch GmbH

4 Basics of the gasoline (SI) engine

4 Operating concept7 Torque and output power 8 Engine efficiency

10 Gasoline-engine management10 Technical requirements12 Cylinder-charge control15 A/F-mixture formation18 Ignition

20 Systems for cylinder-charge control

20 Air-charge control22 Variable valve timing25 Exhaust-gas recirculation

(EGR)26 Dynamic supercharging29 Mechanical supercharging30 Exhaust-gas turbocharging33 Intercooling

34 Gasoline fuel injection: An overview

34 External A/F-mixture formation35 Internal A/F-mixture formation

36 Fuel supply37 Fuel supply for manifold

injection39 Low-pressure circuit for

gasoline direct injection41 Evaporative-emissions control

system 42 Electric fuel pump44 Fuel filter45 Rail, fuel-pressure regulator,

fuel-pressure damper, fuel tank, fuel lines

48 Manifold fuel injection49 Operating concept50 Electromagnetic fuel injectors52 Types of fuel injection

54 Gasoline direct injection55 Operating concept56 Rail, high-pressure pump58 Pressure-control valve59 Rail-pressure sensors60 High-pressure injector62 Combustion process63 A/F-mixture formation64 Operating modes

66 Ignition: An overview66 Survey66 Ignition systems development

68 Coil ignition68 Ignition driver stage69 Ignition coil70 High-voltage distribution71 Spark plugs72 Electrical connection and inter-

ference-suppressor devices73 Ignition voltage, ignition energy75 Ignition point

76 Catalytic emissions control76 Oxidation-type catalytic converter77 Three-way catalytic converter80 NOx accumulator-type catalytic

converter82 Lambda control loop84 Catalytic-converter heating

85 Index of technical terms85 Technical terms87 Abbreviations

� Contents

Robert Bosch GmbH

The call for environmentally compatible and economical vehicles, which nevertheless

must still satisfy demands for high performance, necessitates immense efforts to de-

velop innovative engine concepts. The increasingly stringent exhaust-gas legislation

initially caused the main focus of concentration to be directed at reducing the toxic

content of the exhaust gas, and the introduction of the 3-way catalytic converter in the

middle of the eighties was a real milestone in this respect.

Just lately though, the demand for more economical vehicles has come to the fore-

front, and direct-injection gasoline engines promise fuel savings of up to 20%.

This Yellow Jacket technical instruction manual deals with the technical concepts em-

ployed in complying with the demands made upon a modern-day engine, and explains

their operation.

Another Yellow Jacket manual explains the interplay between these concepts and a

modern closed and open-loop control system in the form of the Motronic. This man-

ual is at present in the planning stage.

Robert Bosch GmbH

The gasoline or spark-ignition (SI) internal-combustion engine uses the Otto cycle1)and externally supplied ignition. It burns anair/fuel mixture and in the process convertsthe chemical energy in the fuel into kineticenergy.

For many years, the carburetor was respon-sible for providing an A/F mixture in the in-take manifold which was then drawn intothe cylinder by the downgoing piston.

The breakthrough of gasoline fuel-injection,which permits extremely precise metering ofthe fuel, was the result of the legislation gov-erning exhaust-gas emission limits. Similarto the carburetor process, with manifoldfuel-injection the A/F mixture is formed inthe intake manifold.

Even more advantages resulted from the de-velopment of gasoline direct injection, inparticular with regard to fuel economy andincreases in power output. Direct injectioninjects the fuel directly into the engine cylin-der at exactly the right instant in time.

Operating concept

The combustion of the A/F mixture causesthe piston (Fig. 1, Pos. 8) to perform a recip-rocating movement in the cylinder (9). Thename reciprocating-piston engine, or betterstill reciprocating engine, stems from thisprinciple of functioning.

The conrod (10) converts the piston’s rec-iprocating movement into a crankshaft (11)rotational movement which is maintainedby a flywheel (11) at the end of the crank-shaft. Crankshaft speed is also referred to asengine speed or engine rpm.

Four-stroke principleToday, the majority of the internal-combus-tion engines used as vehicle power plants areof the four-stroke type.

The four-stroke principle employs gas-ex-change valves (5 and 6) to control the ex-haust-and-refill cycle. These valves open andclose the cylinder’s intake and exhaust pas-sages, and in the process control the supplyof fresh A/F mixture and the forcing out ofthe burnt exhaust gases.

1st stroke: InductionReferred to top dead center (TDC), the pis-ton is moving downwards and increases thevolume of the combustion chamber (7) sothat fresh air (gasoline direct injection) orfresh A/F mixture (manifold injection) isdrawn into the combustion chamber pastthe opened intake valve (5).

The combustion chamber reaches maxi-mum volume (Vh+Vc) at bottom dead cen-ter (BDC).

2nd stroke: CompressionThe gas-exchange valves are closed, and thepiston is moving upwards in the cylinder. Indoing so it reduces the combustion-chambervolume and compresses the A/F mixture. Onmanifold-injection engines the A/F mixturehas already entered the combustion cham-ber at the end of the induction stroke. Witha direct-injection engine on the other hand,depending upon the operating mode, thefuel is first injected towards the end of thecompression stroke.

At top dead center (TDC) the combustion-chamber volume is at minimum (compres-sion volume Vc).

4 Basics of the gasoline (SI) engine Operating concept

Basics of the gasoline (SI) engine

1) Named after Nikolaus Otto (1832-1891) who presentedthe first gas engine with compression using the 4-strokeprinciple at the Paris World Fair in 1878.

Robert Bosch GmbH

3rd stroke: Power (or combustion) Before the piston reaches top dead center(TDC), the spark plug (2) initiates the com-bustion of the A/F mixture at a given igni-tion point (ignition angle). This form of ig-nition is known as externally supplied igni-tion. The piston has already passed its TDCpoint before the mixture has combustedcompletely.

The gas-exchange valves remain closedand the combustion heat increases the pres-sure in the cylinder to such an extent thatthe piston is forced downward.

4th stroke: ExhaustThe exhaust valve (6) opens shortly beforebottom dead center (BDC). The hot (ex-haust) gases are under high pressure andleave the cylinder through the exhaust valve.The remaining exhaust gas is forced out bythe upwards-moving piston.

A new operating cycle starts again with theinduction stroke after every two revolutionsof the crankshaft.

Valve timingThe gas-exchange valves are opened andclosed by the cams on the intake and ex-haust camshafts (3 and 1 respectively). Onengines with only 1 camshaft, a lever mecha-nism transfers the cam lift to the gas-ex-change valves.

The valve timing defines the opening andclosing times of the gas-exchange valves.Since it is referred to the crankshaft posi-tion, timing is given in “degrees crankshaft”.Gas flow and gas-column vibration effectsare applied to improve the filling of thecombustion chamber with A/F mixture andto remove the exhaust gases. This is the rea-son for the valve opening and closing timesoverlapping in a given crankshaft angular-position range.

The camshaft is driven from the crank-shaft through a toothed belt (or a chain orgear pair). On 4-stroke engines, a completeworking cycle takes two rotations of thecrankshaft. In other words, the camshaftonly turns at half crankshaft speed.

Basics of the gasoline (SI) engine Operating concept 5

Figure 1a Induction strokeb Compression strokec Power (combustion)

stroked Exhaust stroke1 Exhaust camshaft2 Spark plug3 Intake camshaft4 Injector5 Intake valve6 Exhaust valve7 Combustion

chamber8 Piston9 Cylinder10 Conrod11 CrankshaftM Torque α Crankshaft angles Piston strokeVh Piston displacementVc Compression

volume

α

M

sVh

Vc

UT

a1

b c d

OT

23

4

5

7

8

1011

9

6

Complete working cycle of the 4-stroke spark-ignition (SI) gasoline engine (example shows a manifold-injection engine with separate intake and exhaust camshafts)

1

æU

MM

0011

-1E

Robert Bosch GmbH

CompressionThe compression ratio ε = (Vh+Vc)/Vc iscalculated from the piston displacement Vh

and the compression volume Vc.

The engine’s compression ratio has a deci-sive effect upon

� The torque generated by the engine,� The engine’s power output,� The engine’s fuel consumption, and the � Toxic emissions.

With the gasoline engine, the compressionratio ε = 7...13, depending upon engine typeand the fuel-injection principle (manifoldinjection or direct injection). The compres-sion ratios (ε = 14...24) which are commonfor the diesel engine cannot be used for thegasoline engine. Gasoline has only very lim-ited antiknock qualities, and the high com-pression pressure and the resulting hightemperatures in the combustion chamberwould for this reason cause automatic, un-controlled ignition of the gasoline. This inturn causes knock which can lead to enginedamage.

Air/fuel (A/F) ratioIn order for the A/F mixture to burn completely 14.7 kg air are needed for 1 kgfuel.

This is the so-called stoichiometric ratio (14.7:1).

The excess-air factor (or air ratio) λ hasbeen chosen to indicate how far the actualA/F mixture deviates from the theoreticaloptimum (14.7:1). λ = 1 indicates that theengine is running with a stoichiometric (in other words, theoretically optimum) A/F ratio.

Enriching the A/F mixture with more fuelleads to λ values of less than 1, and if the A/Fmixture is leaned off (addition of more air)λ is more than 1. Above a given limit (λ > 1.6) the A/F mixture reaches the so-called lean-burn limit and cannot be ignited.

Distribution of the A/F mixture in thecombustion chamber Homogeneous distributionOn manifold-injection engines, the A/Fmixture is distributed homogeneously in thecombustion chamber and has the same λnumber throughout (Fig. 2a). Lean-burnengines which operate in certain ranges withexcess air, also run with homogeneous mix-ture distribution.

Stratified-charge At the ignition point, there is an ignitableA/F-mixture cloud (with λ = 1) in the vicin-ity of the spark plug. The remainder of thecombustion chamber is filled with either avery lean A/F mixture, or with a non-com-bustible gas containing no gasoline at all.The principle in which an ignitable A/F-mixture cloud only fills part of the combus-tion chamber is referred to as stratifiedcharge (Fig. 2b). Referred to the combustionchamber as a whole, the A/F mixture is verylean (up to λ≈10). This form of lean-burnoperation leads to fuel-consumption savings.

In effect, the stratified-charge principle isonly applicable with gasoline direct injec-tion. The stratified charge is the direct resultof the fuel being injected directly into thecombustion chamber only very shortly be-fore the ignition point.

6 Basics of the gasoline (SI) engine Operating concept

Figure 2a Homogeneous A/F-

mixture distributionb Stratified charge

a b

A/F mixture distribution in the combustion chamber

2

æU

MM

0557

Y

Robert Bosch GmbH

Torque and output powerVia the cranks on the crankshaft, the conrodconverts the piston’s reciprocal movementinto crankshaft rotational movement. Theforce with which the expanding A/F mixtureforces the piston downwards is convertedinto torque.

In addition to the force, the lever arm is thedecisive quantity for torque. On the inter-nal-combustion engine, the lever arm is de-fined by the crankshaft throw.

In general, torque is the product of forcetimes lever arm. The lever arm which is ef-fective for the torque is the lever componentvertical to the force. Force and lever arm areparallel to each other at TDC, so that the ef-fective lever arm is in fact zero. At a crank-shaft angle of 90° after TDC, the lever arm isvertical to the generated force, and the leverarm and with it the torque is at a maximumin this setting. It is therefore necessary to se-lect the ignition angle so that the ignition ofthe A/F mixture takes place in the crankshaftangle which is characterized by increasinglever arm. This enables the engine to gener-ate the maximum-possible torque.

The engine’s design (for instance, pistondisplacement, combustion-chamber geome-try) determines the maximum possibletorque M that it can generate. Essentially, thetorque is adapted to the requirements of ac-tual driving by adjusting the quality andquantity of the A/F mixture.

The engine’s power output P climbs alongwith increasing torque M and enginespeed n. The following applies:

P = 2 · π · n · M

Fig. 1 shows the typical torque and power-output curve, against engine rpm, for amanifold-injection gasoline engine. Thesediagrams are often referred to in the test re-ports published in automobile magazines.Along with increasing engine speed, torqueincreases to its maximum Mmax. At higherengine speeds, torque drops again. Today,engine development is aimed at achievingmaximum torque already at low enginespeeds around 2000 min-1, since it is in thisengine-speed range that fuel economy is atits highest. Engines with exhaust-gas tur-bocharging comply with this demand.

Engine power increases along with enginespeed until, at the engine’s nominal speednnom, it reaches a maximum with its nominalrating Pnom.

The power and torque curves of the inter-nal-combustion (IC) engine make it impera-tive that some form of gearbox is installed toadapt the engine to the requirements ofeveryday driving.

Basics of the gasoline (SI) engine Torque and output power 7

Figure 1Mmax Maximum

torquePnenn Nominal powernnenn Nominal engine

speedEngine rpm n

Torq

ue M

100

120

140N.m

Pow

er P

20

40

60

80kW

Engine rpm n nnom

nnom

P

1000

1000

3000 5000

3000 5000

min-1

min-1

Mmax

M

Pnom

Example of the power and torque curves of a manifold-injection gasoline engine

1

æS

MM

0558

E

Robert Bosch GmbH

Engine efficiencyThermal efficiencyThe internal-combustion does not convertall the energy which is chemically availablein the fuel into mechanical work, and someof the added energy is lost. This means thatan engine’s efficiency is less than 100%(Fig. 1). Thermal efficiency is one of the im-portant links in the engine’s efficiency chain.

Pressure-volume diagram (p-V diagram)The p-V diagram is used to display the pres-sure and volume conditions during a com-plete working cycle of the 4-stroke IC en-gine.

The ideal constant-volume cycleFig. 2 (curve A) shows the compression andpower strokes of an ideal process as definedby the laws of Boyle/Mariotte and Gay-Lus-sac. The piston travels from BDC to TDC(point 1 to point 2), and the A/F mixture iscompressed without the addition of heat(Boyle/Mariotte). Subsequently, the mixtureburns accompanied by a pressure rise (point2 to point 3) while volume remains constant(Gay-Lussac).

From TDC (point 3), the piston travelstowards BDC (point 4), and the combus-tion-chamber volume increases. The pres-sure of the burnt gases drops whereby noheat is released (Boyle/Mariotte). Finally, theburnt mixture cools off again with thevolume remaining constant (Gay-Lusac)until the initial status (point 1) is reachedagain.

The area inside the points 1 – 2 – 3 – 4shows the work gained during a completeworking cycle. The exhaust valve opens atpoint 4 and the gas, which is still under pres-sure, escapes from the cylinder. If it werepossible for the gas to expand completely bythe time point 5 is reached, the area de-scribed by 1 – 4 – 5 would represent usableenergy. On an exhaust-gas turbochargedengine, the part above the line (1 bar) canto some extent be utilized (1 – 4 – 5�).

Real p-V diagramSince it is impossible during normal engineoperation to maintain the basic conditionsfor the ideal constant-volume cycle, the ac-tual p-V diagram (Fig. 2, curve B) differsfrom the ideal p-V diagram.

Measures for increasing thermal efficiencyThe thermal efficiency rises along with in-creasing A/F-mixture compression. Thehigher the compression, the higher the pres-sure in the cylinder at the end of the com-pression phase, and the larger is the enclosedarea in the p-V diagram. This area is an indi-cation of the energy generated during thecombustion process. When selecting thecompression ratio, the fuel’s antiknock qual-ities must be taken into account.

Manifold-injection engines inject the fuelinto the intake manifold onto the closed in-take valve, where it is stored until drawn intothe cylinder. During the formation of theA/F mixture, the fine fuel droplets vaporise.The energy needed for this process is in theform of heat and is taken from the air andthe intake-manifold walls. On direct-injec-tion engines the fuel is injected into thecombustion chamber, and the energyneeded for fuel-droplet vaporization is takenfrom the air trapped in the cylinder whichcools off as a result. This means that thecompressed A/F mixture is at a lower tem-perature than is the case with a manifold-in-jection engine, so that a higher compressionratio can be chosen.

Thermal lossesThe heat generated during combustion heatsup the cylinder walls. Part of this thermalenergy is radiated and lost. In the case ofgasoline direct injection, the stratified-charge A/F mixture cloud is surrounded by ajacket of gases which do not participate inthe combustion process. This gas jacket hin-ders the transfer of heat to the cylinder wallsand therefore reduces the thermal losses.

8 Basics of the gasoline (SI) engine Engine efficiency

Robert Bosch GmbH

Further losses stem from the incompletecombustion of the fuel which has condensedonto the cylinder walls. Thanks to the insulating effects of the gas jacket, theselosses are reduced in stratified-charge opera-tion. Further thermal losses result from theresidual heat of the exhaust gases.

Losses at λ =1The efficiency of the constant-volume cycleclimbs along with increasing excess-air fac-tor (λ). Due to the reduced flame-propaga-tion velocity common to lean A/F mixtures,at λ > 1.1 combustion is increasingly slug-gish, a fact which has a negative effect uponthe SI engine’s efficiency curve. In the finalanalysis, efficiency is the highest in the rangeλ = 1.1...1.3. Efficiency is therefore less for ahomogeneous A/F-mixture formation with λ = 1 than it is for an A/F mixture featuringexcess air. When a 3-way catalytic converteris used for efficient emissions control, anA/F mixture with λ = 1 is absolutely impera-tive.

Pumping lossesDuring the exhaust and refill cycle, the en-gine draws in fresh gas during the 1st (in-duction) stroke. The desired quantity of gasis controlled by the throttle-valve opening.A vacuum is generated in the intake mani-fold which opposes engine operation (throttling losses). Since with a gasoline direct-injection engine the throttle valve iswide open at idle and part load, and thetorque is determined by the injected fuelmass, the pumping losses (throttling losses)are lower.

In the 4th stroke, work is also involved inforcing the remaining exhaust gases out ofthe cylinder.

Frictional lossesThe frictional losses are the total of all thefriction between moving parts in the engineitself and in its auxiliary equipment. For in-stance, due to the piston-ring friction at thecylinder walls, the bearing friction, and thefriction of the alternator drive.

Basics of the gasoline (SI) engine Engine efficiency 9

Figure 2A Ideal constant-

volume cycleB Real p-V diagrama Inductionb Compressionc Work (combustion)d ExhaustZZ Ignition pointAÖ Exhaust valve opens

Cyl

inde

r pr

essu

re p

Volume V

Vc Vh

A

2

3

4

15

B

b

c

AÖd 5

a1 bar

ZZ

Sequence of the motive working process in the p-V diagram

2

æU

MM

0559

E

Useful work, drive

Frictional losses, auxiliary equipment

Pumping losses

Losses due to λ =1

Thermal losses in the cylinder, inefficient combustion, and exhaust-gas heat

Thermodynamic losses during the ideal process (thermal efficiency)

13%10%

10%7%

15%

45%

Efficiency chain of an SI engine at λ = 11

æS

MM

0560

E

Robert Bosch GmbH

In modern-day vehicles, closed and open-loop electronic control systems are becom-ing more and more important. Slowly butsurely, they have superseded the purely me-chanical systems (for instance, the ignitionsystem). Without electronics it would beimpossible to comply with the increasinglysevere emissions-control legislation.

Technical requirements

One of the major objectives in the develop-ment of the automotive engine is to generateas high a power output as possible, while atthe same time keeping fuel consumptionand exhaust emissions down to a minimumin order to comply with the legal require-ments of emissions-control legislation.

Fuel consumption can only be reduced byimproving the engine’s efficiency. Particu-larly in the idle and part-load ranges, inwhich the engine operates the majority ofthe time, the conventional manifold-injec-tion SI engine is very inefficient. This is thereason for it being so necessary to improvethe engine’s efficiency at idle and part load

without at the same time having a detrimen-tal effect upon the normal engine’s favorableefficiency in the upper load ranges. Gasolinedirect injection is the solution to this prob-lem.

A further demand made on the engine isthat it develops high torque even at very lowrotational speeds so that the driver has goodacceleration at his disposal. This makestorque the most important quantity in themanagement of the SI engine.

SI-engine torqueThe power P delivered by an SI engine is de-fined by the available clutch torque M andthe engine rpm n. The clutch torque is thetorque developed by the combustion processless friction torque (frictional torque in theengine), pumping losses, and the torqueneeded to drive the auxiliary equipment(Fig. 1).

10 Gasoline-engine management Technical requirements

Gasoline-engine management

Figure 11 Auxiliary equipment

(alternator, A/Ccompressor etc.)

2 Engine3 Clutch 4 Gearbox

Air mass (fresh-gas charge)

Fuel mass

Ignition angle (ignition point)Engine

Exhaust and refill cycle, and friction Auxiliary equipment

Clutch lossesGearbox losses and transmission ratio

Combustion torque

Engine torque

Clutch torque

Drive torque

– –– –– – Clutch Gearbox

1 1 2 3 4

Torque at the drivetrain1

æU

MM

0545

-1E

Robert Bosch GmbH

The combustion torque is generated duringthe power stroke. In manifold-injection en-gines, which represent the majority of to-day’s engines, it is determined by the follow-ing quantities:

� The air mass which is available for com-bustion when the intake valves close,

� The fuel mass which is available at thesame moment, and

� The moment in time when the ignitionspark initiates the combustion of the A/Fmixture.

The proportion of direct-injection SI en-gines will increase in the future. These en-gines run with excess air at certain operatingpoints (lean-burn operation) which meansthat there is air in the cylinder which has noeffect upon the generated torque. Here, it isthe fuel mass which has the most effect.

Engine-management assignments One of the engine management’s jobs is toset the torque that is to be generated by theengine. To do so, in the various subsystems(ETC, A/F-mixture formation, ignition) allquantities that influence torque are con-trolled. It is the objective of this form ofcontrol to provide the torque demanded bythe driver while at the same time complyingwith the severe demands regarding exhaustemissions, fuel consumption, power output,comfort and safety. It is impossible to satisfyall these requirements without the use ofelectronics.

In order that all these stipulations aremaintained in long-term operation, the en-gine management continuously runsthrough a diagnosis program and indicatesto the driver when a fault has been detected.This is one of the most important assign-ments of the engine management, and italso makes a valuable contribution to sim-plifying vehicle servicing in the workshop.

Subsystem: Cylinder-charge controlOn conventional injection systems, the dri-ver directly controls the throttle-valve open-ing through the accelerator pedal. In doingso, he/she defines the amount of fresh airdrawn in by the engine.

Basically speaking, on engine-managementsystems with electronic accelerator pedal forcylinder-charge control (also known asEGAS or ETC/Electronic Throttle Control),the driver inputs a torque requirementthrough the position of the acceleratorpedal, for instance when he/she wants to ac-celerate. Here, the accelerator-pedal sensormeasures the pedal’s setting, and the “ETC”subsystem uses the sensor signal to definethe correct cylinder air charge correspond-ing to the driver’s torque input, and opensthe electronically controlled throttle valveaccordingly.

Subsystem: A/F-mixture formationDuring homogeneous operation and at a de-fined A/F ratio λ, the appropriate fuel massfor the air charge is calculated by the A/F-mixture subsystem, and from it the appro-priate duration of injection and the best in-jection point. During lean-burn operation,and essentially stratified-charge operationcan be classified as such, other conditionsapply in the case of gasoline direct injection.Here, the torque-requirement input fromthe driver determines the injected fuel quan-tity, and not the air mass drawn in by theengine.

Subsystem: IgnitionThe crankshaft angle at which the ignitionspark is to ignite the A/F mixture is calcu-lated in the “ignition” subsystem.

Gasoline-engine management Technical requirements 11

Robert Bosch GmbH

Cylinder-charge control It is the job of the cylinder-charge control tocoordinate all the systems that influence theproportion of gas in the cylinder.

Components of the cylinder chargeThe gas mixture trapped in the combustionchamber when the intake valve closes is re-ferred to as the cylinder charge. This is com-prised of the fresh gas and the residual gas.

The term “relative air charge rl” has beenintroduced in order to have a quantitywhich is independent of the engine’s dis-placement. It is defined as the ratio of theactual air charge to the air charge understandard conditions (p0 = 1013 hPa,T0 = 273 K).

Fresh gasThe freshly introduced gas mixture in thecylinder is comprised of the fresh air drawnin and the fuel entrained with it (Fig. 1). Ona manifold-injection engine, all the fuel hasalready been mixed with the fresh air up-stream of the intake valve. On direct-injec-tion systems, on the other hand, the fuel is in-jected directly into the combustion chamber.

The majority of the fresh air enters thecylinder with the air-mass flow (6, 7) via thethrottle valve (13) and the intake valve (11).Additional fresh gas, comprising fresh airand fuel vapor, can be directed to the cylin-der via the evaporative-emissions controlsystem (3).

For homogeneous operation at λ ≤ 1, the airin the cylinder after the intake valve (11) hasclosed is the decisive quantity for the workat the piston during the combustion strokeand therefore for the engine’s output torque.In this case, the air charge corresponds tothe torque and the engine load. During lean-burn operation (stratified charge) though,the torque (engine load) is a direct productof the injected fuel mass.

During lean-burn operation, the air masscan differ for the same torque. Almost al-ways, measures aimed at increasing the en-gine’s maximum torque and maximum out-put power necessitate an increase in themaximum possible charge. The theoreticalmaximum charge is defined by the displace-ment.

Residual gas The residual-gas share of the cylinder chargecomprises that portion of the cylindercharge which has already taken part in thecombustion process. In principle, one differ-entiates between internal and external resid-ual gas. The internal residual gas is that gaswhich remains in the cylinder’s upper clear-ance volume following combustion, or thatgas which is drawn out of the exhaust pas-sage and back into the intake manifold whenthe intake and exhaust valves open together(that is, during valve overlap).External residual gas are the exhaust gaseswhich enter the intake manifold through theEGR valve.

12 Gasoline-engine management Cylinder-charge control

Figure 11 Air and fuel vapor

(from the evapora-tive-emissions con-trol system)

2 Canister-purge valvewith variable valve-opening cross-section

3 Connection to theevaporative-emis-sions control system

4 Returned exhaustgas

5 EGR valve with vari-able valve-openingcross-section

6 Air-mass flow (ambi-ent pressure pu)

7 Air-mass flow (mani-fold pressure ps)

8 Fresh A/F-mixturecharge (combustion-chamber pressurepB)

9 Residual exhaust-gas charge (com-bustion-chamberpressure pB)

10 Exhaust gas (ex-haust-gas backpressure pA)

11 Intake valve12 Exhaust valve13 Throttle valveα Throttle valve-

angle

1

6 713 108

3

4

11 12

9

5

2

α

Cylinder charge in the gasoline engine1

æU

MM

0544

-3Y

Robert Bosch GmbH

Residual exhaust gas comprises inert gas1)and, during excess-air operation, unburntair. The inert gas in the residual exhaust gasdoes not participate in the combustion dur-ing the next power stroke, although it doeshave an influence on ignition and on thecombustion curve.The selective use of a given share of residualgas can reduce the NOx emissions.

In order to achieve the demanded torque,the fresh-gas charge displaced by the inertgas must be compensated for by a largerthrottle-valve opening. This leads to a reduc-tion in pumping losses which in turn resultsin a reduction in fuel consumption.

Controlling the fresh-gas chargeManifold injectionThe torque developed by a manifold-injec-tion engine is proportional to the fresh-gascharge. The engine’s torque is controlled viathe throttle valve which regulates the flow ofair drawn in by the engine. With the throttlevalve less than fully open, the flow of airdrawn in by the engine is throttled and thetorque drops as a result. This throttling ef-fect is a function of the throttle valve’s set-ting, in other words its opened cross-sec-tion. Maximum torque is developed with thethrottle wide open (Wide Open Throttle =WOT).

Fig. 2 shows the principal correlation be-tween fresh-gas charge and engine speed as afunction of throttle-valve opening.

Direct injectionOn direct-injection (DI) gasoline enginesduring homogeneous operation at λ ≤ 1(that is, not lean-burn operation), the sameconditions apply as with manifold injection.

To reduce the throttling losses, the throttlevalve is also opened wide in the part-loadrange. In the ideal case, there are no throt-tling losses with the throttle wide open (as itis during full-load operation). In order tolimit the torque developed at part load, notall of the air mass entering the cylinder mayparticipate in combustion. In lean-burn ap-plications with excess air (λ > 1), some ofthe air drawn in remains as residual exhaustgas in the cylinder or is forced out duringthe exhaust stroke. In other words, it is notthe air charge trapped in the cylinder whichis decisive for the developed torque, butrather the fuel injected into the combustionchamber.

Gasoline-engine management Cylinder-charge control 13

Fres

h-ga

s ch

arge

rpmmin. max.

WOT

Throttle fully closed

Idle

Throttle characteristic-curve map for an SI engine– – – Intermediate throttle-valve settings

2

æU

MM

0543

-2E

1) Components in the combustion chamber which behaveinertly, that is, do not participate in the combustion process.

Robert Bosch GmbH

Exhaust and refill cycleThe replacement of the used/burnt cylindercharge (= exhaust gas) by a fresh-gas chargetakes place using intake and exhaust valveswhich are opened and closed at precisely de-fined times by the cams on the camshaft(valve timing). These cams also define thevalve-lift characteristic which influences theexhaust and refill cycle and with it the fresh-gas charge which is available for combus-tion.

Valve overlap, that is, the overlap of theopened times of the intake and exhaustvalves, has a decisive influence on the ex-haust-gas mass remaining in the cylinder.This exhaust-gas mass also defines theamount of inert gas in the fresh cylindercharge for the next power cycle. In suchcases, one refers to “internal” EGR.

The inert-gas mass in the cylinder chargecan be increased by “external” EGR. Exhaustpipe and intake manifold are connected byan EGR valve so that the percentage of inertgas in the cylinder charge can be varied as afunction of the operating point.

Volumetric efficiencyFor the air throughput, the total charge dur-ing a complete working cycle is referred tothe theoretical charge as defined by the pis-ton displacement. For the volumetric effi-ciency though, only the exhaust gas actuallyremaining in the cylinder is considered.Fresh gas drawn in during valve overlap,which is not available for the combustionprocess, is not considered.

The volumetric efficiency for naturally aspi-rated engines is 0.6...0.9. It depends uponthe combustion-chamber shape, the openedcross-sections of the gas-exchange valves,and the valve timing.

Supercharging The torque which can be achieved duringhomogenous operation at λ ≤ 1 is propor-tional to the fresh gas charge. This meansthat maximum torque can be increased bycompressing the air before it enters thecylinder (supercharging). This leads to anincrease in volumetric efficiency to valuesabove 1.

Dynamic supercharging Supercharging can be achieved simply bytaking advantage of the dynamic effectsinside the intake manifold. The supercharg-ing level depends on the intake manifold’sdesign and on its operating point (for themost part, on engine speed, but also oncylinder charge). The possibility of changingthe intake-manifold geometry while theengine is running (variable intake-manifold geometry) means that dynamic supercharg-ing can be applied across a wide operatingrange to increase the maximum cylindercharge.

Mechanical supercharging The intake-air density can be further in-creased by compressors which are drivenmechanically from the engine’s crankshaft.The compressed air is forced through theintake manifold and into the engine’s cylin-ders.

Exhaust-gas turbochargingIn contrast to the mechanical supercharger,the exhaust-gas turbocharger is driven by anexhaust-gas turbine located in the exhaust-gas flow, and not by the engine’s crankshaft.This enables recovery of some of the energyin the exhaust gas.

14 Gasoline-engine management Cylinder-charge control

Robert Bosch GmbH

A/F-mixture formationThe A/F-mixture formation system is re-sponsible for calculating the fuel mass appropriate to the amount of air drawn intothe engine. This fuel is metered to the engine’s cylinders through the fuel injectors.

A/F mixtureTo run efficiently, the gasoline engine needsa given air/fuel (A/F) ratio. Ideal, theoreti-cally complete combustion takes place at amass ratio of 14.7:1, which is also referred toas the stoichiometric ratio. In other words,14.7 kg of air are needed to burn 1 kg offuel. Or, expressed in volumes, approx. 9,500liters of air are needed to completely burn 1liter of gasoline.

Excess-air factor λThe excess-air factor λ has been chosen toindicate how far the actual A/F-mixture de-viates from the theoretically ideal mass ratio(14.7:1). λ defines the ratio of the actuallysupplied air mass to the theoretical air massrequired for complete (stoichiometric) com-bustion.

λ = 1: The inducted air mass corresponds tothe theoretically required air mass.

λ < 1: This indicates air deficiency andtherefore a rich A/F mixture. On a cold en-gine, it is necessary to enrich the A/F mix-ture by adding fuel to compensate for thefuel that has condensed on the cold mani-fold walls (manifold-injection engines) andcold cylinder walls and which, as a result, isnot available for combustion.

λ > 1: This indicates excess air and thereforea lean A/F mixture. The maximum value forλ that can be achieved is defined by the so-called lean-misfire limit (LML), and ishighly dependent upon the engine’s designand construction, as well as upon the mix-ture-formation system used. At the lean-misfire limit the A/F mixture is no longercombustible, and this marks the point at

which misfire starts. The engine begins torun very unevenly, fuel consumption in-creases dramatically, and power outputdrops.

Other combustion conditions prevail ondirect-injection (DI) engines, and these arethus able to run with considerably higher λfigures.

Operating modesHomogeneous (λ ≤ 1): On manifold-injec-tion engines, the A/F mixture in the mani-fold is drawn in past the open intake valveduring the induction stroke. This leads to anessentially homogeneous mixture distribu-tion in the combustion chamber.

This operating mode is also possible withDI gasoline engines, the fuel being injectedinto the combustion chamber during the in-duction stroke.

Homogeneous lean (λ > 1): The A/F mixtureis distributed homogeneously in the com-bustion chamber with a defined level of ex-cess air.

Stratified charge: This operating mode andthose given below are only possible with di-rect-injection gasoline engines. Fuel is in-jected only shortly before the ignition point,and an A/F-mixture cloud forms in thevicinity of the spark plug.

Homogenous stratified charge: In addition tothe stratified charge, there is a homogeneouslean A/F mixture throughout the completecombustion chamber. Dual injection is ap-plied to achieve this form of A/F-mixturedistribution.

Homogeneous anti-knock: Here, dual injec-tion is also used to achieve an A/F-mixturedistribution which to a great extent preventscombustion knock.

Stratified-charge/catalyst heating: Retarded(late) injection leads to the rapid warm-upof the catalytic converter.

Gasoline-engine management A/F-mixture formation 15

Robert Bosch GmbH

Specific fuel consumption,power and exhaust emissionsManifold injectionManifold-injection gasoline engines developtheir maximum power output at 5...15 % airdeficiency (λ = 0.95...0.85), and their lowestfuel consumption at 10...20 % excess air(λ = 1.1...1.2). Figs. 1 and 2 indicate the ex-tent to which power output, fuel consump-tion, and exhaust emissions are all a func-tion of the excess-air factor λ. It is immedi-ately apparent that there is no excess-airfactor at which all factors are at their “opti-mum”. Best-possible fuel consumptiontogether with best-possible power outputare achieved with excess-air factors ofλ = 0.9...1.1.

When a 3-way catalytic converter is used forthe treatment of the exhaust gases, it is ab-solutely imperative that λ = 1 is maintainedprecisely when the engine has warmed-up.In order to comply with these requirements,the mass of the intake air must be measuredexactly and a precisely metered fuel quantityinjected.

An optimal combustion process thoughnot only demands precision fuel injection,but also a homogeneous A/F mixture, whichin turn necessitates efficient atomization ofthe fuel. If the fuel is not perfectly atomized,large fuel droplets are deposited on the walls

of the manifold and/or combustion cham-ber. Since these fuel droplets cannot burncompletely, they lead to increased HC emis-sions.

Gasoline direct injectionFor gasoline direct injection, during homo-geneous operation at λ ≤ 1, the same condi-tions apply as with manifold injection. Withstratified-charge operation though, a practi-cally stoichiometric A/F mixture is only pre-sent in the stratified-charge mixture cloudnear the spark plug. Outside this area, thecombustion chamber is filled with fresh airand inert gas. Regarding the combustionchamber as a whole, the A/F mixture ratio isvery high (λ > 1).

Since the complete combustion chamberis not filled with a combustible A/F mixturein this operating mode, torque output andpower output both drop. Similar to mani-fold injection, maximum power can only bedeveloped when the complete combustionchamber is filled with a homogeneous A/Fmixture.

Depending upon the combustion process,and the A/F-mixture distribution in thecombustion chamber, NOx emissions aregenerated in the lean-burn mode whichcannot be reduced by the 3-way catalyticconverter. Here, for emissions control, it is

16 Gasoline-engine management A/F-mixture formation

Figure 1a Rich A/F mixture

(air deficiency)b Lean A/F mixture

(excess air)

0.8 1.0 1.2

a b

P

be

Pow

er P

,sp

ecifi

c fu

el c

onsu

mpt

ion

be

Excess-air factor λ

Influence of the excess-air factor λ on the power Pand on the specific fuel consumption be under con-ditions of homogeneous A/F-mixture distribution

1

æU

MK

0033

-1E

0.6 1.0 1.4

Rel

ativ

e qu

antit

ies

of

CO

; H

C;

NO

X

Excess-air factor λ0.8 1.2

COHC NOX

Effect of the excess-air factor λ on the pollutantcomposition of untreated exhaust gas under con-ditions of homogeneous A/F-mixture distribution

2

æU

MK

0032

-1E

Robert Bosch GmbH

necessary to take additional measures whichcall for a NOx accumulator-type catalyticconverter.

Engine operating modesIn some engine operating modes, the fuel re-quirement differs considerably from thesteady-state requirements with the engine atoperating temperature. This makes it neces-sary to take corrective measures in the A/F-mixture formation.

Start and warm-upWhen starting with the engine cold, the in-ducted A/F-mixture leans-off. This is thedue not only to inadequate mixing of the in-take air with the fuel, but also to the fuelhaving less tendency to evaporate at lowtemperatures, and the pronounced wall wet-ting (condensation of the fuel) on the still-cold intake manifold (only on manifold-in-jection engines) and on the cylinder walls.To compensate for these negative effects, andto facilitate engine start, additional fuel mustbe provided during the cranking process.

Even after the engine has started, addi-tional fuel must continue to be injected untilit reaches operating temperature. This alsoapplies to the gasoline direct-injection en-gine. Depending upon the engine’s designand the combustion process, stratified-charge lean-burn operation is only possiblewith the engine at operating temperature.

Idle and part loadOnce they have reached their operating tem-perature, conventional manifold-injectionengines all run on a stoichiometric A/F mix-ture at idle and part load. On direct-injec-tion gasoline engines though, the objective isto run the engine as often as possible with astratified-charge. This is feasible at idle andat part load, the two operating modes withthe highest potential for saving fuel, wherefuel savings of as much as 40 % can beachieved with lean-burn operation.

Full loadEssentially, the conditions for manifold in-jection and gasoline direct injection arepretty much the same at full load. At WOT,it may be necessary to enrich the A/F mix-ture. As can be seen from Fig. 1, this permitsthe generation of maximum-possible torqueand power.

Acceleration and decelerationWith manifold injection, the fuel’s tendencyto evaporate depends to a large extent uponthe manifold pressure. This leads to the de-velopment of a fuel film (wall film) on theintake manifold in the vicinity of the intakevalves. Rapid changes in manifold pressure,as occur when the throttle-valve openingchanges suddenly, lead to changes in thiswall film. Heavy acceleration causes the in-take-manifold pressure to increase so thatthe fuel’s evaporation tendency deteriorates,and the wall film thickens as a result. Beingas a portion of the fuel has been depositedto form the wall film, the A/F mixture leans-off temporarily until the wall film has stabi-lized. Similarly, sudden deceleration leans toenrichment of the A/F mixture since thedrop in manifold pressure causes a reduc-tion in the wall film and the fuel from thewall film is drawn into the cylinder. A tem-perature-dependent correction function(transitional compensation) is used to cor-rect the A/F mixture so as to ensure not onlythe best possible driveability, but also theconstant A/F ratio as needed for the catalyticconverter.

Wall-film effects are also encountered atthe cylinder walls. With the engine at oper-ating temperature though, they can be ig-nored on direct-injection gasoline engines.

OverrunAt overrun (trailing throttle), the fuel supplyis interrupted (overrun fuel cutoff). Apartfrom saving fuel on downhill gradients, thisprotects the catalytic converter against over-heating which could result from inefficientand incomplete combustion.

Gasoline-engine management A/F-mixture formation 17

Robert Bosch GmbH

IgnitionIt is the job of the ignition to ignite the com-pressed A/F-mixture at exactly the right mo-ment in time and thus initiate its combustion.

Ignition systemIn the gasoline (SI) engine, the A/F mixtureis ignited by a spark between the electrodesof the spark plug. The inductive-type igni-tion systems used predominantly on gaso-line engines store the electrical energyneeded for the ignition spark in the ignitioncoil. This energy determines how long(dwell angle) the current must flow throughthe ignition coil to recharge it. The interrup-tion of the coil current at a defined crank-shaft angle (ignition angle) leads to the igni-tion spark and the A/F-mixture combustion.

In today’s ignition systems, the processesbehind the ignition of the A/F mixture areelectronically controlled.

Ignition pointChanging the ignition point (ignition timing)Following ignition, about 2 milliseconds areneeded for the A/F mixture to burn com-pletely. The ignition point must be selectedso that main combustion, and the accompa-nying pressure peaks in the cylinder, takes

place shortly after TDC. Along with increas-ing engine speed, therefore, the ignition an-gle must be shifted in the advance direction.

The cylinder charge (or fill) also has an ef-fect upon the combustion curve. The lowerthe cylinder charge the slower is the flamefront’s propagation. For this reason, with alow cylinder charge, the ignition angle mustalso be advanced.

Influence of the ignition angleThe ignition angle has a decisive influenceon engine operation. It determines

� The delivered torque,� The exhaust-gas emissions, and� The fuel consumption.

The ignition angle is chosen so that all re-quirements are complied with as well as pos-sible, whereby care must be taken that con-tinued engine knock is avoided.

Ignition angle: Basic adaptationOn electronically controlled ignition sys-tems, the ignition map (Fig. 1) takes into ac-count the influence of engine speed andcylinder charge on the ignition angle. Thismap is stored in the engine-managementdata storage, and represents the basic adap-tation of the ignition angle.

The x and y axes represent the enginespeed and the relative air charge. The map’sdata points are formed by a given number ofvalues, typically 16. A certain ignition angleis allocated to each pair of variates so thatthe map has 256 (16x16) adjustable igni-tion-angle values. By applying linear inter-polation between two data points, the num-ber of ignition-angle values is increased to4096.

Using the ignition-map principle for theelectronic control of the ignition anglemeans that for every engine operating pointit is possible to select the best-possible igni-tion angle.

These ignition maps are generated by runningthe engine on the engine dynamometer.

18 Gasoline-engine management Ignition

Engine rpmRelative air charge

Ignition angle

Ignition map based on engine rpm n and relativeair charge rl

1

æU

MZ

0030

-1E

Robert Bosch GmbH

Additive ignition-angle adjustmentsA lean A/F mixture is more difficult to ig-nite. This means that more time is neededbefore the main combustion point isreached. A lean A/F mixture must thereforebe ignited sooner. The A/F ratio λ thus hasan influence on the ignition angle.

The coolant temperature is a further vari-able which affects the choice of the ignitionangle. Temperature-dependent ignition-an-gle corrections are therefore also necessary.Such corrections are stored in the data stor-age in the form of fixed values or character-istic curves (e.g. temperature-dependentcorrection). They shift the basic ignition an-gle by the stipulated amount in either theadvance or retard direction.

Special ignition angleThere are certain operating modes, such asidle and overrun, which demand an ignitionangle which deviates from those defined bythe ignition map. In such cases, access ismade to special ignition-angle curves storedin the data storage.

Knock controlKnock is a phenomenon which occurs whenignition takes place too early. Here, once reg-ular combustion has started, the rapid pres-sure increase in the combustion chamberleads to the auto-ignition of the unburntresidual mixture which has not been reachedby the flame front. The resulting abruptcombustion of the residual mixture leads toa considerable local pressure increase. Thisgenerates a pressure wave which propagatesthrough the combustion chamber until ithits the cylinder wall. At low engine speedsand when the engine is not making toomuch noise, it is then audible as combustionknock. At high speeds, the engine noisesblanket the combustion knock.

If knock continues over a longer period oftime, the engine can be damaged by thepressure waves and the excessive thermalloading. To prevent knock on today’s high-compression engines, no matter whether ofthe manifold-injection or direct-injectiontype, knock control belongs to the standardscope of the engine-management system.With this system, knock sensors detect thestart of knock and the ignition angle is re-tarded at the cylinder concerned. To obtainthe best-possible engine efficiency, therefore,the basic adaptation of the ignition angle(ignition map) can be located directly at theknock limit.

On direct-injection gasoline engines, com-bustion knock only takes place in homoge-neous operation. There is no tendency forthe engine to knock in the stratified-chargemode since there is no combustible mixturein the stratified charge at the combustionchamber’s peripheral zones.

Dwell angleThe energy stored in the ignition coil is afunction of the length of time current flowsthrough the coil (energisation time). In or-der not to thermally overload the coil, thetime required to generate the required igni-tion energy in the coil must be rigidly ad-hered to. The dwell angle refers to the crank-shaft and is therefore speed-dependent.

The ignition-coil current is a function of thebattery voltage, and for this reason the bat-tery voltage must be taken into accountwhen calculating the dwell angle.

The dwell-angle values are stored in a map,the x and y axes of which represent rpm andbattery voltage.

Gasoline-engine management Ignition 19

Robert Bosch GmbH

On a gasoline engine running with a homo-geneous A/F mixture, the intake air is thedecisive quantity for the output torque andtherefore for engine power. This means thatnot only is the fuel-metering system of spe-cial importance but also the systems whichinfluence the cylinder charge. Some of thesesystems are able to influence the percentageof inert gas in the cylinder charge and thusalso the exhaust emissions.

Air-charge control

For it to burn, fuel needs oxygen which theengine takes from the intake air. On engineswith external A/F-mixture formation (mani-fold injection), as well as on direct-injectionengines operating on a homogeneous A/Fmixture with λ = 1, the output torque is di-rectly dependent upon the intake-air mass.The throttle valve located in the inductiontract controls the air flow drawn in by theengine and thus also the cylinder charge.

Conventional systemsConventional systems (Fig. 1) feature a me-chanically operated throttle valve (3). Theaccelerator-pedal (1) movement is trans-ferred to the throttle valve by a linkage (2)or by a Bowden cable. The throttle valve’svariable opening angle alters the openingcross-section of the intake passage (4) andin doing so regulates the air flow (5) drawnin by the engine, and with it the torque out-put.

To compensate for the higher levels of fric-tion, the cold engine requires a larger airmass and extra fuel. And when, for instance,the A/C compressor is switched on more airis needed to compensate for the torque loss.This information is inputted to the ECU (8)in the form of an electrical signal (9), andthe extra air is supplied by the air bypass ac-tuator (7) directing the required extra air (6)around the throttle valve. Another methoduses a throttle-valve actuator to adjust thethrottle valve’s minimum stop. In both casesthough, it is only possible to electronicallyinfluence the air flow needed by the engineto a limited extent, for instance for idle-speed control.

20 Systems for cylinder-charge control Air-charge control

Systems for cylinder-charge control

Figure 11 Accelerator pedal2 Bowden cable or

linkage3 Throttle valve4 Induction passage5 Intake air flow6 Bypass air flow7 Idle-speed actuator

(air bypass actuator)8 ECU9 Input variables (elec-

trical signals)

78

1 2

9

34

6

5

Principle of the air control in a conventional systemusing a mechanically adjustable throttle valve andan air bypass actuator

1

æU

MK

1677

-1Y

Robert Bosch GmbH

ETC systemsWith ETC (Electronic Throttle Control, alsoknown as EGAS), an ECU (Fig. 2, Pos. 2) isresponsible for controlling the throttle valve(5). The DC-motor throttle-valve drive (4)and the throttle-valve-angle sensor (3) arecombined with the throttle valve to form aunit, the so-called throttle device. To triggerthe throttle device, the accelerator-pedal po-sition, in other words the driver input, isregistered by two potentiometers (accelera-tor-pedal sensor, 1). Taking into account theengine’s actual operating status (enginespeed, engine temperature, etc.) the engineECU then calculates the throttle-valve open-ing which corresponds to the driver inputand converts it into a triggering signal forthe throttle-valve drive.

Using the feedback information from thethrottle-valve-angle sensor regarding thecurrent position of the throttle valve, it thenbecomes possible to precisely adjust thethrottle valve to the required setting.Two potentiometers on the accelerator-pedaland two on the throttle unit are a compo-nent part of the ETC monitoring concept.

The potentiometers are duplicated for re-dundancy reasons. In case malfunctions aredetected in that part of the system which isdecisive for the engine’s power output, thethrottle valve is immediately shifted to a pre-determined position (emergency or limp-home operation).

In the latest engine-management systems,the ETC control is integrated in the engineECU which is also responsible for control-ling ignition, fuel injection, and the auxiliaryfunctions. There is no longer a separate ETCcontrol unit.

The demands of emissions-control legisla-tion are getting sharper from year to year.They can be complied with though thanksto ETC with its possibilities of further im-proving the A/F-mixture composition.

ETC is indispensable when complyingwith the demands made by gasoline directinjection on the overall vehicle system.

Systems for cylinder-charge control Air-charge control 21

Figure 21 Accelerator-pedal

sensor2 Engine ECU3 Throttle-valve-angle

sensor4 Throttle-valve drive

(DC motor)5 Throttle valve

Accelerator-pedal module

Engine ECU

CAN

Sensors1 Actuators

Monitoring modul

C

Throttle device

M

5432

The ETC system (Electronic Throttle Control or EGAS)2

æU

MK

1627

-1E

Robert Bosch GmbH

Variable valve timingApart from using the throttle-valve to throt-tle the flow of incoming fresh gas drawn inby the engine, there are several other possi-bilities for influencing the cylinder charge.The proportion of fresh gas and of residualgas can also be influenced by applying vari-able valve timing.

Of great importance for valve timing is thefact that the behaviour of the gas columnsflowing into and out of the cylinders variesconsiderably as a function of engine speedor throttle-valve opening. With invariablevalve timing, therefore, this means that theexhaust and refill cycle can only be ideal forone single engine operating range. Variablevalve timing, on the other hand, permitsadaptation to a variety of different enginespeeds and cylinder charges. This has thefollowing advantages:

� Higher engine outputs,� Favorable torque curve throughout a wide

engine-speed range,� Reduction of toxic emissions,� Reduced fuel consumption,� Reduction of engine noise.

Camshaft phase adjustmentIn conventional IC engines, camshaft andcrankshaft are mechanically coupled to eachother through toothed belt or chain. Thiscoupling is invariable.

On engines with camshaft adjustment, atleast the intake camshaft, but to an increas-ing degree the exhaust camshaft as well, canbe rotated referred to the crankshaft so thatvalve overlap changes. The valve openingperiod and lift are not affected by camshaftphase adjustment, which means that “intakeopens” and “intake closes” remain invariablycoupled with each other.

The camshaft is adjusted by means ofelectrical or electro-hydraulic actuators. Onless sophisticated systems provision is onlymade for two camshaft settings. Variablecamshaft adjustment on the other hand per-mits, within a given range, infinitely variableadjustment of the camshaft referred to thecrankshaft.

Fig. 1 shows how the “position”, or lift, ofthe open intake-valve changes (referred toTDC) when the intake camshaft is adjusted.

Retard adjustment of the intake camshaft Retarding the intake camshaft leads to theintake valve opening later so that valve over-lap is reduced, or there is no valve overlap atall. At low engine speeds (<2000 min–1), thisresults in only very little burnt exhaust gasflowing past the intake valve and into theintake manifold. At low engine speeds, thelow residual exhaust-gas content in the in-take of A/F mixture which then follows leadsto a more efficient combustion process and asmoother idle. This means that the idlespeed can be reduced, a step which is partic-ularly favorable with respect to fuel con-sumption.

22 Systems for cylinder-charge control Variable valve timing

Figure 11 Camshaft retarded 2 Camshaft normal3 Camshaft advancedA Valve overlap

300° 420° 480° 600°0

Intake (variable)

Exhaust (invariable)

A

1

2

3

Val

ve li

ft s

Crankshaft angleBDCTDC

540°360°

Camshaft adjustment1

æU

MM

0534

-1E

Robert Bosch GmbH

The camshaft is also retarded at higher en-gine speeds (>5,000 min–1). Late closing ofthe intake valve, long after BDC, leads to ahigher cylinder charge. This boost effect re-sults from the high flow speed of the freshgas through the intake valve which contin-ues even after the piston has reversed its di-rection of travel and is moving upwards tocompress the mixture. For this reason, theintake valve closes long after BDC.

Advance adjustment of the intake camshaftIn the medium speed range, the flow of freshgas through the intake passage is muchslower, and of course there is no high-speedboost effect.At medium engine speeds, closing the intakevalve earlier, only shortly after BDC, pre-vents the ascending piston forcing thefreshly drawn-in gas out past the intakevalve again and back into the manifold. Atsuch speeds, advancing the intake camshaftresults in better cylinder charge and there-fore a good torque curve.

At medium speeds, advanced opening of theintake camshaft leads to increased valveoverlap. Opening the intake valve earlymeans that shortly before TDC, the residualexhaust gas which has not already left thecylinder is forced out past the open intakevalve and into the intake manifold by the as-cending piston. These exhaust gases are thendrawn into the cylinder again and serve toincrease the residual-gas content of thecylinder charge. The increased residual gascontent in the freshly drawn in A/F mixturecaused by advancing the intake camshaft, af-fects the combustion process. The resultinglower peak temperatures lead to a reductionin NOx.

The higher inert-gas content in the cylindercharge makes it necessary to open the throttle valve further, which in turn leads toa reduction of the throttling losses. Thismeans that valve overlap can be applied toreduce fuel consumption.

Adjusting the exhaust camshaftOn systems which can also adjust the ex-haust camshaft, not only the intake camshaftis used to vary the residual-gas content, butalso the exhaust camshaft. Here, the totalcylinder charge (defined by “intake closes”)and the residual-gas content (influenced by“intake opens” and “exhaust closes”) can becontrolled independently of each other.

Camshaft changeoverCamshaft changeover (Fig. 2) involvesswitching the camshaft between two differ-ent cam contours. This changes both thevalve lift and the valve timing (cam-contourchangeover). The first cam defines the opti-mum timing and the valve lift for the intakeand exhaust valves in the lower and mediumspeed ranges. The second cam controls theincreased valve lift and longer valve-opentimes needed at higher speeds.At low and medium engine speeds, mini-mum valve lifts together with the associated

Systems for cylinder-charge control Variable valve timing 23

Figure 21 Standard cam2 Supplementary cam

Crankshaft angle

240° 480° 600°

BDC

0

Val

ve li

ft s

TDCBDC

120°

1

1

22

360°

Intake (variable)

Exhaust (variable)

Camshaft changeover2

æU

MM

0535

-1E

Robert Bosch GmbH

small valve-opening cross-sections lead to ahigh inflow velocity and therefore to highlevels of turbulence in the cylinder for thefresh air (gasoline direct injection) or for thefresh A/F mixture (manifold injection). Thisensures excellent A/F mixture formation atpart load. The high engine outputs requiredat higher engine speeds and torque demand(WOT) necessitate maximum cylindercharge. Here, the maximum valve lift isselected.

There are a variety of methods in use forswitching-over between the different camcontours. One method, for instance, relieson a free-moving drag lever which engageswith the standard rocking lever as a functionof rotational speed. Another method useschangeover cup tappets.

Fully variable valve timing and valve liftusing the camshaftValve control which incorporates both vari-able valve timing and variable valve lift is re-ferred to as being fully variable. Even morefreedom in engine operation is permitted by3D cam contours and longitudinal-shiftcamshafts (Fig. 3). With this form ofcamshaft control, not only the valve lift(only on the intake side) and thus the open-ing angle of the valves can be infinitely var-ied, but also the phase position betweencamshaft and crankshaft.

Since the intake valve can be closed earlywith this fully variable camshaft control, thispermits so-called charge control in whichthe intake-manifold throttling is consider-ably reduced. This enables fuel consumptionto be slightly lowered in comparison withthe simple camshaft phase adjustment.

Fully variable valve timing and valve liftwithout using the camshaftFor valve timing, maximum design freedomand maximum development potential areafforded by systems featuring valve-timingcontrol which is independent of thecamshaft. With this form of timing, thevalves are opened and closed, for instance,by electromagnetic actuators. A supplemen-tary ECU is responsible for triggering. Thisform of fully variable valve timing withoutcamshaft aims at extensive reduction of theintake-manifold throttling, coupled withvery low pumping losses. Further fuel sav-ings can be achieved by incorporating cylin-der and valve shutoff.

These fully variable valve-timing conceptsnot only permit the best-possible cylindercharge and with it a maximum of torque,but they also ensure improved A/F-mixtureformation which results in lower toxic emis-sions in the exhaust gas.

24 Systems for cylinder-charge control Variable valve timing

Figure 3a Minimum liftb Maximum lift

a b

Example of a system with fully variable ad-justment of valve timing and of valve lift

3

æU

MM

0536

-1Y

Robert Bosch GmbH

Exhaust-gas recirculation(EGR)

The mass of the residual gas remaining inthe cylinder, and with it the inert-gas con-tent of the cylinder charge, can be influ-enced by varying the valve timing. In thiscase, one refers to “internal” EGR. The inert-gas content can be influenced far more byapplying “external” EGR with which part ofthe exhaust gas which has already left thecylinder is directed back into the intakemanifold through a special line (Fig. 1,Pos. 3). EGR leads to a reduction of the NOx

emissions and to a slightly lower fuel-con-sumption figure.

Limiting the NOx emissionsSince they are highly dependent upon tem-perature, EGR is highly effective in reducingNOx emissions. When peak combustiontemperature is lowered by introducing burntexhaust gas to the A/F mixture, NOx emis-sions drop accordingly.

Lowering fuel consumptionWhen EGR is applied, the overall cylindercharge increases while the charge of fresh airremains constant. This means that the throt-tle valve (2) must reduce the engine throt-tling if a given torque is to be achieved. Fuelconsumption drops as a result.

EGR: Operating conceptDepending upon the engine’s operatingpoint, the engine ECU (4) triggers the EGRvalve (5) and defines its opened cross-sec-tion. Part of the exhaust-gas (6) is divertedvia this opened cross-section (3) and mixedwith the incoming fresh air. This defines theexhaust-gas content of the cylinder charge.

EGR with gasoline direct injectionEGR is also used on gasoline direct-injectionengines to reduce NOx emissions and fuelconsumption. In fact, it is absolutely essen-tial since with it NOx emissions can belowered to such an extent in lean-burn oper-ations that other emissions-reduction mea-sures can be reduced accordingly (forinstance, rich homogeneous operation forNOx “Removal” from the NOx accumulator-type catalytic converter). EGR also has afavorable effect on fuel consumption.

There must be a pressure gradient be-tween the intake manifold and the exhaust-gas tract in order that exhaust gas can bedrawn in via the EGR valve. At part loadthough, direct-injection engines are oper-ated practically unthrottled. Furthermore aconsiderable amount of oxygen is drawninto the intake manifold via EGR duringlean-burn operation.

Non-throttled operation and the intro-duction of oxygen into the intake manifoldvia the EGR therefore necessitate a controlstrategy which coordinates throttle valve andEGR valve. This results in severe demandsbeing made on the EGR system with regardto precision and reliability, and it must berobust enough to withstand the depositswhich accumulate in the exhaust-gas com-ponents as a result of the low exhaust-gastemperatures.

Systems for cylinder-charge control Exhaust-gas recirculation (EGR) 25

Figure 11 Fresh-air intake 2 Throttle valve3 Recirculated

exhaust gas4 Engine ECU5 EGR valve 6 Exhaust gasn Engine rpmrl Relative air charge

33

216

5

4nrl

Exhaust-gas recirculation (EGR)1

æU

MK

0913

-2Y

Robert Bosch GmbH

Dynamic superchargingApproximately speaking, the achievable en-gine torque is proportional to the fresh-gascontent in the cylinder charge. This meansthat the maximum torque can be increasedto a certain extent by compressing the airbefore it enters the cylinder.

The exhaust-and-refill processes are notonly influenced by the valve timing, but alsoby the intake and exhaust lines. The piston’sinduction work causes the open intake valveto trigger a return pressure wave. At theopen end of the intake manifold, the pres-sure wave encounters the quiescent ambientair from which it is reflected back again sothat it returns in the direction of the intakevalve. The resulting pressure fluctuations atthe intake valve can be utilized to increasethe fresh-gas charge and thus achieve thehighest-possible torque.

This supercharging effect thus depends onutilization of the incoming air’s dynamic re-sponse. In the intake manifold, the dynamiceffects depend upon the geometrical rela-tionships in the intake manifold and on theengine speed.

For the even distribution of the A/F mixture,the intake manifolds for carburetor enginesand single-point injection (TBI) must haveshort pipes which as far as possible must beof the same length for all cylinders. In thecase of multipoint injection (MPI), the fuelis either injected into the intake manifoldonto the intake valve (manifold injection),or it is injected directly into the combustionchamber (gasoline direct injection). WithMPI, since the intake manifolds transportmainly air and practically no fuel can de-posit on the manifold walls, this provideswide-ranging possibilities for intake-mani-fold design. This is the reason for there be-ing no problems with multipoint injectionsystems regarding the even distribution offuel.

Ram-tube superchargingThe intake manifolds for multipoint injec-tion systems are composed of the individualtubes or runners and the manifold chamber.

In the case of ram-tube supercharging(Fig. 1), each cylinder is allocated its owntube (2) of specific length which is usuallyattached to the manifold chamber (3). Thepressure waves are able to propagate in theindividual tubes independently.

The supercharging effect depends uponthe intake-manifold geometry and the en-gine speed. For this reason, the length anddiameter of the individual tubes is matchedto the valve timing so that in the requiredspeed range a pressure wave reflected at theend of the tube is able to enter the cylinderthrough the open intake valve (1) and im-prove the cylinder charge. Long, narrowtubes result in a marked supercharging ef-fect at low engine speeds. On the otherhand, short, large-diameter tubes have apositive effect on the torque curve at higherengine speeds.

26 Systems for cylinder-charge control Dynamic supercharging

Figure 11 Cylinder2 Individual tube3 Manifold chamber4 Throttle valve

4

1

2

3

Principle of ram-tube supercharging1

æU

MM

0587

Y

Robert Bosch GmbH

Tuned-intake-tube chargingAt a given engine speed, the periodic pistonmovement causes the intake-manifold gas-column to vibrate at resonant frequency.This results in a further increase of pressureand leads to an additional supercharging ef-fect. On the tuned intake-tube system (Fig.2), groups of cylinders (1) with identical an-gular ignition spacing are each connected toa resonance chamber (3) through shorttubes (2). The chambers, in turn, are con-nected through tuned intake tubes (4) witheither the atmosphere or with the manifoldchamber (5) and function as Helmholtz res-onators.

The subdivision into two groups of cylin-ders each with its own tuned intake tubeprevents the overlapping of the flowprocesses of two neighboring cylinderswhich are adjacent to each other in the firingsequence.

The length of the tuned intake tubes andthe size of the resonance chamber are afunction of the speed range in which the su-percharging effect due to resonance is re-quired to be at maximum. Due to the accu-mulator effect of the considerable chambervolumes which are sometimes needed, dy-namic-response errors can occur in somecases when the load is changed abruptly.

Variable-geometry intake manifoldThe supplementary cylinder charge resultingfrom dynamic supercharging depends uponthe engine’s working point. The two systemsjust dealt with increase the achievable maxi-mum charge (volumetric efficiency), aboveall in the low engine-speed range (Fig. 3).

Practically ideal torque characteristics canbe achieved with variable-geometry intakemanifolds in which, as a function of the en-gine operating point, flaps are used to im-plement a variety of different adjustmentssuch as:

� Adjustment of the intake-tube length,� Switch over between different intake-tube

lengths or different tube diameters,� Selected switchoff of one of the cylinder’s

intake tubes on multiple-tube systems,� Switchover to different chamber volumes.

Electrical or electropneumatically actuatedflaps are used for change-over operations inthese variable-geometry systems.

Systems for cylinder-charge control Dynamic supercharging 27

Figure 21 Cylinder2 Short tube3 Resonance chamber4 Tuned intake tube5 Manifold chamber6 Throttle valveA Cylinder group AB Cylinder group B

Figure 31 System with tuned-

intake-tube charging2 System with conven-

tional intake mani-fold

Engine speed

14

nnnom.

Vol

umet

ric e

ffici

ency

12

34 1

1

2

Increasing the maximum-possible cylinder aircharge (volumetric efficiency) by means ofdynamic supercharging

3

æU

MM

0589

E

1

23

5

6

4

A B

Principle of tuned-intake-tube charging2

æU

MM

0588

Y

Robert Bosch GmbH

Ram-tube systemsThe manifold system shown in Fig. 4 canswitch between two different ram tubes. Inthe lower speed range, the changeover flap(1) is closed and the intake air flows to thecylinders through the long ram tube (3). Athigher speeds and with the changeover flapopen, the intake air flows through the short,

wide diameter ram tube (4), and thus con-tibutes to improved cylinder charge at highengine revs.

Tuned-intake-tube systemOpening the resonance flap switches in asecond tuned intake tube. The changedgeometry of this configuration has an effectupon the resonant frequency of the intakesystem. Cylinder charge in the lower speedrange is improved by the higher effectivevolume resulting from the second tuned in-take pipe.

Combined tuned-intake-tube and ram-tubesystemWhen design permits the open changeoverflap (Fig. 5, Pos. 7) to combine both the res-onance chambers (3) to form a single vol-ume, one speaks of a combined tuned-in-take-tube and ram-tube system. A single in-take-air chamber with a high resonantfrequency is then formed for the short ramtubes (2).

At low and medium engine revs, thechangeover flap is closed and the systemfunctions as a tuned-intake-tube system.The low resonant frequency is then definedby the long tuned intake tube (4).

28 Systems for cylinder-charge control Dynamic supercharging

Figure 4a Manifold geometry

with changeover flapclosed

b Manifold geometrywith changeover flapopen

1 Changeover flap2 Manifold chamber3 Changeover flap

closed: Long, nar-row-diameter ramtube

4 Changeover flapopened: Short,wide-diameter ramtube

Figure 51 Cylinder2 Ram tube

(short intake tube)3 Resonance chamber4 Tuned intake tube5 Manifold chamber6 Throttle valve7 Changeover flapA Cylinder group AB Cylinder group B

a Intake-manifold con-ditions withchangeover flapclosed

b Intake-manifold con-ditions withchangeover flapopen

1

5

6

7

4

Aa B b

23

Combined tuned-intake-tube and ram-tube system

5

æU

MM

0591

Ya

b 2 4 1

2

3

1

Ram-tube system4

æU

MM

0590

Y

Robert Bosch GmbH

Mechanical superchargingDesign and operating conceptThe application of supercharging units leadsto increased cylinder charge and therefore toincreased torque. Mechanical supercharginguses a compressor which is driven directlyby the IC engine. Mechanically driven com-pressors are either positive-displacement su-perchargers with different types of construc-tion (e.g. Roots supercharger, sliding-vanesupercharger, spiral-type supercharger,screw-type supercharger), or they are cen-trifugal turbo-compressors (e.g. radial-flowcompressor). Fig. 1 shows the principle offunctioning of the rotary-screw super-charger with the two counter-rotating screwelements. As a rule, engine and compressorspeeds are directly coupled to one anotherthrough a belt drive.

Boost-pressure controlOn the mechanical supercharger, a bypasscan be applied to control the boost pressure.A portion of the compressd air is directedinto the cylinder and the remainder is re-turned to the supercharger input via the by-pass. The engine management is responsiblefor controlling the bypass valve.

Advantages and disadvantagesOn the mechanical supercharger, the directcoupling between compressor and enginecrankshft means that when engine speed in-creases there is no delay in supercharger ac-celeration. This means therefore, that com-pared to exhaust-gas turbocharging enginetorque is higher and dynamic response isbetter.

Since the power required to drive the com-pressor is not available as effective enginepower, the above advantage is counteractedby a slightly higher fuel-consumption figurecompared to the exhaust-gas turbocharger.This disadvantage though is somewhat al-leviated when the engine management isable to switch off the compressor via aclutch at low engine loading.

Systems for cylinder-charge control Mechanical supercharging 29

Figure 11 Intake air2 Compressed air

2

1

Rotary-screw supercharger: Principle of functioning1

æU

MM

0592

Y

Robert Bosch GmbH

Exhaust-gas turbochargingOf all the possible methods for supercharg-ing the IC engine, exhaust-gas turbocharg-ing is the most widely used. Even on engineswith low swept volumes, exhaust-gas super-charging leads to high torques and poweroutputs together with high levels of engineefficiency.

Whereas, in the past, exhaust-gas tur-bocharging was applied in the quest for in-creased power-weight ratio, it is todaymostly used in order to increase the maxi-mum torque at low and medium enginespeeds. This holds true particularly in com-bination with electronic boost-pressure con-trol.

Design and operating conceptThe main components of the exhaust-gasturbocharger (Fig. 1) are the exhaust-gasturbine (3) and the compressor (1). Thecompressor impeller and the turbine rotorare mounted on a common shaft (2).

The energy needed to drive the exhaust-gasturbine is for the most part taken from thehot, pressurized exhaust gas. On the otherhand, energy must be also used to “dam” theexhaust gas when it leaves the engine so as togenerate the required compressor power.

The hot gases (Fig. 2, Pos. 7) are applied ra-dially to the exhaust-gas turbine (4) andcause this to rotate at very high speed. Theturbine-rotor blades are inclined towardsthe center and thus direct the gas to the in-side from where it then exits axially.

30 Systems for cylinder-charge control Exhaust-gas turbocharging

Figure 11 Compressor

impeller2 Shaft3 Exhaust-gas turbine4 Intake for exhaust-

gas mass flow5 Outlet for com-

pressed air 1 2 3

5

4

Passenger-car exhaust-gas turbocharger (Shown: 3K-Warner, type K14)1

æS

MM

0593

Y

Robert Bosch GmbH

The compressor (3) also turns along withthe turbine, but here the flow conditions arereversed. The fresh incoming gas (5) entersaxially at the center of the compressor and isforced radially to the outside by the bladesand compressed in the process.

Since the exhaust-gas turbocharger is lo-cated directly in the flow of hot exhaust gasit must be built of highly temperature-resis-tant materials.

Exhaust-gas turbochargers: DesignsWastegate superchargerThe objective is for IC engines to develophigh torques at low engine speeds. The tur-bine casing has therefore been designed for alow level of exhaust-gas mass flow, for in-stance WOT at ≤ 2000 min–1. With high ex-haust-gas mass flows in this range, part ofthe flow must be diverted around the tur-bine and into the exhaust system in orderthat the turbocharger is prevented fromovercharging the engine. Diversion is via abypass valve, the so-called wastegate (Fig. 2,Pos. 8). This flap-type bypass valve is usuallyintegrated into the turbine casing.

The wastegate is actuated by the boost-pres-sure control valve (6). This valve is con-nected pneumatically to the pulse valve (1)through a control line (2). The pulse valvechanges the boost pressure upon being trig-gered by an electrical signal from the engineECU. This electrical signal is a function ofthe current boost pressure, information onwhich is provided by the boost-pressure sen-sor (BPS).

If the boost pressure is too low, the pulsevalve is triggered so that a somewhat lowerpressure prevails in the control line. Theboost-pressure control valve then closes thewastegate and the proportion of the ex-haust-gas mass flow used to power the tur-bine is increased.

If, on the other hand, the boost pressure isexcessive, the pulse valve is triggered so thata somewhat higher pressure is built up inthe control line. The boost-pressure control

valve then opens the wastegate and the pro-portion of the exhaust-gas mass flow used topower the turbine is reduced.

VTG turbochargerThe VTG (Variable Turbine Geometry) isanother method which can be applied tolimit the exhaust-gas mass flow at higher en-gine speeds (Fig. 3, next page). The VTG su-percharger is state-of-the-art on diesel en-gines, but has not yet become successful ongasoline engines due to the high thermalstressing resulting from the far hotter ex-haust gases.

By varying the geometry, the adjustableguide vanes (3) adapt the flow cross-section,and with it the gas pressure at the turbine, tothe required boost pressure. At low speeds,they open up a small cross-section so thatthe exhaust-gas mass flow in the turbinereaches a high speed and in doing so alsobrings the exhaust-gas turbine up to highspeed (Fig. 3a).

Systems for cylinder-charge control Exhaust-gas turbocharging 31

1

3

4

2

89

6

7

5

pD

VWG

VT

p2

Design and construction of an exhaust-gas turbo-charger using a wastegate turbocharger as an example

2

æU

MK

1320

-1Y

Figure 21 Pulse valve2 Pneumatic control

line3 Compressor4 Exhaust-gas turbine5 Fresh incoming air6 Boost-pressure

control valve7 Exhaust gas8 Wastegate9 Bypass duct

Triggering signal forpulse valve

VT Volume flowthrough the turbine

VWG Volume flowthrough thewastegate

p2 Boost pressurepD Pressure on the

valve diaphragm

Robert Bosch GmbH

At high engine speeds, the adjustable guidevanes (3) open up a larger cross-section sothat more exhaust gas can enter without ac-celerating the exhaust-gas turbine to exces-sive speeds (Fig. 3b). This limits the boostpressure.

It is an easy matter to adjust the guide-vaneangle by rotating the adjusting ring (2).Here, the guide vanes are adjusted to the de-sired angle either directly through individualadjusting levers (4) attached to the guidevanes, or by adjusting cam. The adjustingring is rotated pneumatically via a baromet-ric adjustment cell (5) using either vacuumor overpressure. This adjustment mecha-nism is triggered by the engine managementso that the boost pressure can be set to thebest-possible level in accordance with theengine’s operating mode.

VST superchargerOn the VST (Variable Sleeve Turbine) super-charger, the “turbine size” is adapted bymeans of successively opening two flow pas-sages (Fig. 4, Pos. 2 and 3) using a specialcontrol sleeve (4).

Initially, only one flow passage is opened,and the small opening cross-section resultsin high exhaust-gas flow speed and high tur-bine speeds (1). As soon as the permissibleboost pressure is reached, the control sleevesuccessively opens the second flow passage,the exhaust-gas flow speed reduces accord-ingly, and with it the boost pressure.

Using the bypass channel (5) incorpo-rated in the turbine casing, it is also possibleto divert part of the exhaust-gas mass flowpast the exhaust-gas turbine.

The control sleeve is adjusted by the en-gine management via a barometric cell.

32 Systems for cylinder-charge control Exhaust-gas turbocharging

Figure 3a Guide-vane setting

for high boost pres-sure

b Guide-vane settingfor low boost pres-sure

1 Exhaust-gas turbine2 Adjusting ring3 Guide vanes4 Adjusting lever5 Barometric cell6 Exhaust-gas flow�– High flow speed�– Low flow speed

Figure 4a Only 1 flow passage

openb Both flow passages

open1 Exhaust-gas turbine2 1st flow passage3 2nd flow passage4 Special control

sleeve5 Bypass duct6 Adjustment fork

b

a 1 2 3 54

6

Variable Turbine Geometry of the VTG supercharger3

æU

MM

0594

Y

21 3 4 5 6a

b

Turbine geometry of the VST supercharger4

æU

MM

0552

-1Y

Robert Bosch GmbH

Exhaust-gas turbocharging: Advantages and disadvantagesCompared with a naturally-aspirated IC en-gine with the same output power, the majoradvantages are to be found in the tur-bocharged engine’s lower weight and smallersize (“downsizing”). The turbocharged en-gine’s torque characteristic is betterthroughout the usable speed range (Fig. 5,curve 4 compared to curve 3). All in all, at agiven speed, this results in a higher output(A � B).

Due to its more favorable torque charac-teristic at WOT, the turbocharged enginegenerates the required power as shown inFig. 5 (B or C) at lower engine speeds thanthe naturally aspirated engine. At part load,the throttle valve must be opened further,and the working point is shifted to an areawith reduced frictional and throttling losses(C � B). This results in lower fuel-con-sumption figures even though turbochargedengines in fact feature less favorable effi-ciency figures due to their lower compres-sion ratio.

The low torque that is available at very lowengine speeds is a disadvantage of the tur-bocharger. In such speed ranges, there is notenough energy in the exhaust gas to drivethe exhaust-gas turbine. In transient opera-tion, even in the medium-speed range, thetorque curve is less favorable than that of thenatually aspirated engine (curve 5). This isdue to the delay in building up the exhaust-gas mass flow. When accelerating from lowengine speeds, this is evinced by the turboflat spot.

The effects of this flat spot can be min-imised by making full use of dynamiccharge. This supports the supercharger’srunning-up characteristic. There are a num-ber of other versions available, including aturbocharger with electric motor, or with anextra compressor driven by an electric mo-tor. Independent of the exhaust-gas massflow, these accelerate the compressor im-peller and/or the air-mass flow, and in doingso avoid the turbo flat spot.

IntercoolingThe air warms up in the compressor duringthe compression process, but since warm airhas a lower density than cold air, this tem-perature rise has a negative effect uponcylinder charge. The compressed, warmedair must therefore be cooled off again by theintercooler. Compared to supercharged en-gines with this facility, intercooling results inan increase in the cylinder charge so that it ispossible to further increase torque and out-put power.

The drop in the combustion-air tempera-ture also leads to a reduction in the temper-ature of the cylinder charge compressedduring the compression cycle. This has thefollowing advantages:

� Reduced tendency to knock,� Improved thermal efficiency resulting in

lower fuel-consumption figures,� Reduced thermal loading of the pistons,� Lower NOx emissions.

Systems for cylinder-charge control Exhaust-gas turbocharging, intercooling 33

Figure 51, 3 Naturally aspirated

engine in steady-state operation

2, 4 Supercharged en-gine in steady-stateoperation

5 Torque curve of thesupercharged en-gine in transient(dynamic) operation

Engine speed

Torq

ue M

Pow

er o

utpu

t P

1/4 1/2 3/4 1

1

CB

A

2

43

5

Extra power

Same power output at lower engine speed

Iden

tical

en

gine

spe

ed

n/nnom

Power and torque characteristics of an exhaust-gas-turbocharged engine compared with thoseof a naturally aspirated engine

5

æS

MM

0595

-E

Robert Bosch GmbH

It is the job of the fuel-injection system, orcarburetor, to meter to the engine the best-possible air/fuel mixture for the actualoperating conditions.

Fuel-injection systems, particularly whenthey are electronically controlled, are farsuperior to carburetors in complying withthe tight limits imposed on A/F-mixturecomposition. In addition, they are betterfrom the point of view of fuel consumption,driveability, and power output. In the auto-motive sector, the demands imposed byincreasingly severe emission-control legisla-tion have led to the carburetor beingcompletely superseded by electronic fuelinjection.

At present, on the majority of these injectionsystems the A/F mixture is formed externallyoutside the combustion chamber (manifoldinjection). Systems based on internal A/F-mixture formation, that is with the fuel in-jected directly into the cylinder (gasoline di-rect injection), are coming more and moreto the forefront though, since they haveproved to be particularly suitable in thenever-ending endeavours to reduce fuel con-sumption.

OverviewExternal A/F-mixture formation On gasoline injection systems with externalA/F-mixture formation, the mixture isformed outside the combustion chamber,that is, in the intake manifold. Developmentof such systems was forced ahead to enablethem to comply with increasingly severedemands. Today, only the electronicallycontrolled multipoint injection systems areof any importance in this sector.

Multipoint fuel-injection systemsOn a multipoint injection system, everycylinder is allocated its own injector whichsprays the fuel directly onto the cylinder’sintake valve (Fig. 1). Such injection systemsare ideal for complying with the demandsmade on the A/F-mixture formation system.

Mechanical fuel-injection systemThe K-Jetronic injection system operateswithout any form of drive from the engine,and injects fuel continuously. The injectedfuel mass is not defined by the injector butby the system’s fuel distributor.

Combined mechanical-electronic fuel-injection systemThe KE-Jetronic is based on the basicmechanical system used for the K-Jetronic.Thanks to additional operational-dataacquisition, this system features electroni-cally controlled supplementary functionswhich permit the injected fuel quantity to beeven more accurately adapted to changingengine operating conditions.

34 Gasoline fuel injection: An overview

Gasoline fuel injection: An overview

Figure 11 Fuel2 Air3 Throttle valve4 Intake manifold5 Injector6 Engine

3

2

5

6

4

1

Multipoint fuel-injection system1

æU

MK

0662

-2Y

Robert Bosch GmbH

Electronic fuel-injection systemsElectronically controlled fuel-injectionsystems inject the fuel intermittentlythrough electromagnetically operated injec-tors. The injected fuel quantity is defined bythe injector opening time (for a givenpressure drop across the injector).

Examples: L-Jetronic, LH-Jetronic, andMotronic in the form of an integratedengine-management system (M andME-Motronic).

Single-point injectionSingle-point injection (also known as throt-tle-body injection or TBI) features an elec-tromagnetically operated injector located ata central point directly above the throttlevalve. This injection system intermittentlyinjects fuel into the intake manifold (Fig. 2).The Bosch single-point injection systems aredesignated Mono-Jetronic and Mono-Motronic.

Internal A/F-mixture formation On direct-injection (DI) systems, the fuel isinjected directly into the combustion cham-ber through electromagnetic injectors, oneof which has been allocated to each cylinder(Fig. 3).A/F-mixture formation takes place insidethe combustion chamber.

A/F-mixture formation inside the com-bustion chamber permits two completelydifferent operating modes: In homogeneousoperation, similar to external A/F-mixtureformation, a homogeneous A/F mixture ispresent throughout the combustion cham-ber, and all the fresh air in the combustionchamber participates in the combustionprocess. This operating mode is thereforeapplied when high levels of torque are calledfor. In stratified-charge operation on theother hand, it is only necessary to have anignitable A/F mixture around the sparkplug. The remainder of the combustionchamber only contains fresh gas and resid-ual gas without any unburnt-fuel content.This results in an extremely lean mixture atidle and part-load, with a correspondingdrop in fuel consumption.

The MED-Motronic is used for the man-agement of gasoline direct-injection en-gines.

Gasoline fuel injection: An overview 35

Figure 21 Fuel2 Air3 Throttle valve4 Intake manifold5 Injector6 Engine

Figure 31 Fuel2 Air3 Throttle valve (ETC)4 Intake manifold5 Injectors6 Engine

2

5

3

6

4

1

Single-point injection (TBI) system2

æU

MK

0663

-2Y

2

4

3

5

6

1

Direct-injection (DI) system3

æU

MK

1687

-3Y

Robert Bosch GmbH

The injectors (injection valves) of a gasolineinjection system inject the fuel into theintake manifold (manifold injection), ordirectly into the combustion chamber(direct injection). In both cases, the fuelmust be supplied to the injectors at adefined pressure.

This chapter describes the componentswhich are involved in the supply of fuelfrom the fuel tank to the injectors or, in thecase of gasoline direct injection, from thefuel tank to the high-pressure pump.

Overview

Basically speaking, the following compo-nents are mainly concerned with the supplyof fuel as defined above (Fig. 1):

� Fuel tank (1),� Electric fuel pump (2),� Fuel filter (3),� Fuel-pressure regulator (4), and� Fuel lines (6 and 7).

With manifold injection, the fuel pumpforces the fuel to the injector (8) via the fuelrail (5). On gasoline direct-injection en-gines, the fuel is forced into the high-pres-sure circuit by the high-pressure pump.

On older systems, the electric fuel pump islocated outside the fuel tank in the fuel lineitself (so-called “in-line” pump). On morerecent systems, the fuel pump is inside thefuel tank (“in-tank” pump). It can also becombined with other components (e.g. pre-liminary filter, fuel-level sensor) in the tankin an in-tank unit.

The electric fuel pump delivers fuel continu-ously from the fuel tank and through the fil-ter to the engine. The fuel-pressure regula-tor maintains a defined pressure in the fuelcircuit, depending on the type of fuel-injec-tion system.

In order that the required fuel pressurecan be maintained under all operating con-ditions, the fuel pump delivers more fuelthan is actually required by the engine. Ex-cess fuel is returned to the tank.

So that the required fuel pressure is availablefor starting the engine, the electric fuelpump comes into operation immediately theignition/starting switch is turned. If theengine is not started, it stops again afterabout 1 second.

To a great extent, the pressure generated bythe fuel pump serves to prevent the forma-tion of vapor bubbles in the fuel. The fuelsystem is provided with an integral non-re-turn valve which decouples it from the fueltank by preventing fuel returning to thetank. After the fuel pump has been switchedoff, the non-return valve maintains thesystem pressure for a certain period. Thisprevents the formation of vapor bubbles inthe fuel system when the fuel heats up afterthe engine has been switched off.

36 Fuel supply An overview

Fuel supply

Robert Bosch GmbH

Fuel supply for manifoldinjectionHere, there are two systems for fuel supplywhich differ according to the type of fuel re-turn.

Fuel-supply system with fuel returnExcess fuel is that fuel which the injectordoes not inject (Fig 1, Pos. 8 and Fig. 2, nextpage, Pos. 8). It is returned to the fuel tank(1) via the fuel-pressure regulator (4) whichis usually located on the fuel rail (5).

The intake-manifold pressure is applied asthe reference for system-pressure control.Since the fuel-pressure regulator is situatedvery close to the manifold, it is possible hereto locate the reference connection directlyon the manifold. Using this reference pres-sure results in a constant difference betweenthe fuel-system pressure and the intake-manifold pressure.

This has the advantage that the injected fuelquantity is a function of the injection time.It is independent of intake-manifold pres-sure and therefore also of cylinder charge.

VersionsThere are a variety of different versions ofthe fuel-supply system with return. Thestandard version with fuel flowing throughthe rail is shown in Fig. 2a. There are alsoversions on the market in which the fuel line(6) in connected to the same end of the railas the fuel-pressure regulator, so that there isno direct flow through the rail.

System pressureOn present-day systems with fuel return, thesystem pressure is approx. 0.3 MPa (3 bar).

Fuel supply Fuel supply for manifold injection 37

Figure 11 Fuel tank2 Electric fuel pump

(here integrated inthe fuel tank),

3 Fuel filter,4 Fuel-pressure

regulator5 Fuel rail6 Fuel line7 Fuel-return line8 Injector

1 2

45

3

6

87

Fuel supply system for a manifold-injection engine (version with fuel return) 1æ

UM

K17

02-1

Y

Robert Bosch GmbH

Returnless Fuel SystemThe fuel-pressure regulator (Fig. 2b, Pos. 4b)for the Returnless Fuel System (RLFS) isusually installed inside the fuel tank or in itsvicinity. It can also be installed as a compo-nent part of the in-tank unit. On such sys-tems, the fuel-return line from the fuel railto the fuel tank can be dispensed with. Theexcess fuel delivered by the pump is re-turned directly to the tank via a short returnline from the pressure regulator. Only thefuel actually injected by the injectors is de-livered to the fuel rail.

This system has two advantages: Firstlylower costs, and secondly the fact that thefuel in the tank does not heat up since nohot fuel is returned from the engine com-partment. This leads to a reduction in theHC emissions at the fuel tank, and thereforeto reduced loading of the evaporative-emis-sions control system.

VersionsThere are a number of different returnlessfuel systems available:� Fuel filter and pressure regulator outside

the fuel tank,� Fuel filter outside, pressure regulator in-

side the fuel tank,� Fuel filter and pressure regulator both

integrated in the in-tank unit (fuel-supplymodule).

System pressureSince it would be too far away from the in-take manifold it is practically impossible toprovide an manifold reference connection atthe fuel-pressure regulator. The fuel-pres-sure regulator therefore regulates the systempressure to a constant pressure differentialreferred to the surrounding/ambient pres-sure. This means that the injected fuel quan-tity is a function of the manifold pressure.This fact is taken into account when calcu-lating the injection duration.

On returnless fuel systems the pressure isapprox. 0.35...0.4 MPa (3.5...4 bar).

38 Fuel supply Fuel supply for manifold injection

Figure 2a With fuel returnb Without fuel return1 Fuel tank2 Electric fuel pump3 Fuel filter4a Fuel-pressure

regulator (intake-manifold pressureused as reference)

4b Fuel-pressure regu-lator (surroundingpressure used asreference)

5 Fuel rail6 Fuel line7 Fuel-return line 8 Injectors

a

3 7 8

1

b

2

54a6

7

4b

8

1

2

5

6

3

Fuel-supply system for a manifold-injection engine (examples)2

æU

MK

1252

-1Y

Robert Bosch GmbH

Low-pressure circuit forgasoline direct injectionOn the gasoline direct-injection system, thefuel-supply system can be divided into the

� Low-pressure circuit, and� High-pressure circuit.

The high-pressure circuit is described in theChapter “Gasoline Direct Injection”.

Depending upon the vehicle manufacturer’srequirements, the low-pressure circuits forsuch injection systems can differ consider-ably in design. Similar to the manifold injec-tion system, there are also variants here

� With fuel return, and� Without fuel return (RLFS).

Example of an installationFig. 1 shows a fuel system featuring both fuelreturn and primary-pressure changeover.Here, the pressure in the low-pressure (pri-mary pressure) circuit can be switched be-tween two different levels.

Higher primary pressureWhen the fuel is hot, measures must betaken to prevent the formation of vaporbubbles in the high-pressure pump (7) dur-ing the starting phase and the subsequenthot-idle phase. Increasing the primary pres-sure is a suitable step. Here, the shutoff valve(3) remains closed so that the pressure lim-iter integrated in the electric fuel pump (2)comes into operation and adjusts the pri-mary pressure to 0.5 MPa (5 bar).

When located in the fuel tank, the pres-sure limiter not only protects the compo-nents against excess pressure, but also as-sumes responsibility for pressure-controlfunctions.

Lower primary pressureAfter 30...60 seconds, the high-pressurepump has been thoroughly flushed andcooled off far enough so that there is nolonger any danger of vapor-bubble forma-tion. The shutoff valve opens, and the pres-sure regulator (4) takes over the pressure-control function and adjusts the primarypressure to 0.3 MPa (3 bar).

In this case, the pressure regulator is lo-cated in the engine compartment. This is afuel system with return.

Fuel supply Low-pressure circuit for gasoline direct injection 39

Figure 1Low-pressure (primary)circuit with1 Fuel tank2 Electric fuel pump

with integral pres-sure limiter and fuelfilter

3 Shutoff valve4 Pressure regulator5 Fuel line6 Fuel return lineHigh-pressure circuitwith7 High-pressure pump8 Rail 9 High-pressure

injectors10 Pressure-control

valve11 Fuel-pressure

sensor

35

2

4 6

7 8

9

11

10

1

Fuel supply for a gasoline direct-injection system (example with fuel return and primary-pressure changeover)1

æU

MK

1775

Y

Robert Bosch GmbH

40 Fuel supply Integration in the vehicle: In-tank unit

1 Fuel filter2 Electric fuel pump3 Jet pump (closed-

loop controlled)4 Fuel-pressure

regulator5 Fuel-level sensor6 Preliminary filter

4

5

63

2

1

Integration in the vehicle: In-tank unit�

�

æU

MK

1439

-1Y

In-tank unit: The complete unit for a returnless fuelsystem (RLFS)

In the early years of electronically controlledgasoline injection, the electric fuel pump wasalways installed in the fuel line (“in-line”) out-side the fuel tank. Today, on the other hand,the majority of electric fuel pumps are of the“in-tank” type and, as the name implies, arepart of an “in-tank unit”, the so-called fuel-sup-ply module. This contains an increasing num-ber of other components, for instance:

� A preliminary filter,� A fuel-level sensor,� Electric and hydraulic connections, and� A special fuel reservoir for maintaining the

fuel supply when cornering or in sharpbends.

Usually, a jet pump or a separate stage in theelectric fuel pump keep this reservoir full.

On RLFS systems, the fuel-pressure regu-lator (4), is usually integrated in the in-tank unitwhere it is responsible for the fuel return. Thepressure-side fine fuel filter can also be lo-cated in the in-tank unit.

In future, the fuel-supply module will incorpo-rate further functions, for instance diagnosisdevices for detecting tank leaks, or the timingmodule for triggering the electric fuel pump.

Robert Bosch GmbH

Evaporative-emissions control systemIn order to comply with the legal limits forevaporative hydrocarbon emissions, vehiclesare being equipped with evaporative-emis-sions control systems. This system preventsfuel vapor escaping to the atmosphere fromthe fuel tank.

Fuel-vapor generationMore fuel vapor escapes from the fuel tankunder the following circumstances:

� When the fuel in the fuel tank warms up,due either to high surrrounding tempera-tures, or to the return to tank of excessfuel which has heated up in the enginecompartment, and

� When the surrounding pressure drops, forinstance when driving up a hill in themountains.

Design and operating concept The evaporative-emissions control system(Fig. 1) comprises the carbon canister (3),into which is led the venting line (2) fromthe fuel tank (1), together with the so-calledcanister-purge valve (5) which is connectedto both the carbon canister and the intakemanifold (8).

The activated carbon in the carbon canis-ter absorbs the fuel contained in the fuel va-por and thus permits only air to escape intothe atmosphere. As soon as the canister-purge valve opens the line (6) between thecarbon canister and the intake manifold, thevacuum in the manifold causes fresh air to bedrawn through the activated carbon. The ab-sorbed fuel is then entrained with the freshair (purging or regeneration of the activatedcarbon) and burnt in the normal combustionprocess. The system control reduces the in-jected fuel quantity by the amount returnedthrough canister-purge valve. Regeneration isa closed-loop control process, whereby thefuel concentration in the canister-purge gasflow is continuously calculated based on thechanges it causes in the excess-air factor λ.

The canister-purge gas quantity is controlledas a function of the working point and canbe very finely metered using the canister-purge valve. In order to ensure that the car-bon canister is always able to absorb fuel va-por, the activated carbon must be regener-ated at regular intervals.

Gasoline direction injection: Special featuresDuring stratified-charge operation on gaso-line direct-injection engines, the possibilityof regenerating the carbon canister’s con-tents is limited due to the low level of vac-uum in the intake manifold (caused by prac-tically 100 % “unthrottled” operation) andthe incomplete combustion of the homoge-neously distributed canister-purge gas. Thisresults in reduced canister-purge gas flowcompared to homogeneous operation.For instance, if the canister-purge gas flow isinadequate for coping with high levels ofgasoline evaporation, the engine must beoperated in the homogeneous mode untilthe high concentrations of gasoline in thecanister-purge gas flow have dropped farenough.

Fuel supply Evaporative-emissions control system 41

Figure 11 Fuel tank2 Fuel-tank venting

line3 Carbon canister4 Fresh air5 Canister-purge valve6 Line to the intake

manifold7 Throttle valve8 Intake manifold

7 3

1

6

6

8

5

4

2

Evaporative-emissions control system1

æU

MK

1706

-1Y

Robert Bosch GmbH

Electric fuel pumpAssignmentThe electric fuel pump (EKP) must at alltimes deliver enough fuel to the engine at ahigh enough pressure to permit efficient fuelinjection. The most important performancedemands made on the pump are:

� Delivery quantity between 60 and 200 l/hat rated voltage,

� Pressure in the fuel system between 300and 450 kPa (3...4.5 bar),

� System-pressure buildup even down to aslow as between 50 and 60 % of rated volt-age.

Apart from this, the EKP is increasingly be-ing used as the pre-supply pump for themodern direct-injection systems used ondiesel and gasoline engines.

On gasoline direct-injection systems forinstance, pressures of up to 700 kPa aresometimes required during hot-delivery op-erations.

Design and constructionThe electric fuel pump is comprised of:

� End plate (Fig. 1, A), incorporating spark-suppression elements if required,

� Electric motor (B), and� Pump element (C), designed as either

positive-displacement or turbine pump(for description, see Section “Types”below).

TypesPositive-displacement pumpsIn this type of pump, the fuel is drawn in,compressed in a closed chamber by rotationof the pump element, and transported to thehigh-pressure side. For the EKP, internal-gear pumps or roller-cell pumps (Figs. 2a,2b) are used. When high system pressuresare needed (400 kPa and above), positive-displacement pumps are particularly suit-able. These feature a good low-voltage char-acteristic, that is, they have a relatively flatdelivery-rate characteristic as a function ofthe operating voltage. Efficiency can be ashigh as 25 %.

Pressure pulsations, which are unavoid-able, can cause audible noise dependingupon the particular design details and in-stallation conditions. The fact that the deliv-ery rate can drop when the fuel is hot is another disadvantage which can occur in exceptional cases. This is due to vapor bub-bles being pumped instead of fuel, and forthis reason conventional positive-displace-ment pumps are equipped with peripheralpreliminary stages for degassing purposes.

Whereas in electronic gasoline-injection sys-tems the positive-displacement pump has toa great extent been superseded by the tur-bine pump for the classical fuel-pump requirements, it has captured a new field ofapplication as the presupply pump on direct-injection systems wihich operate withfar higher fuel-pressures.

42 Fuel supply Electric fuel pump

Figure 11 Electric connections2 Hydraulic connec-

tions (fuel outlet)3 Non-return valve4 Carbon brushes5 Permanent-magnet

motor armature6 Turbine-pump

impeller ring7 Hydraulic connec-

tion (fuel inlet)

12

3

4

5

A

B

C6

7

Electric fuel pump: Design and construction using a turbine pump as an example

1

æU

MK

1280

-3Y

Robert Bosch GmbH

Turbine pumpsThis type of pump comprises an impellerring with numerous blades inserted in slotsaround its periphery (Fig. 2c, Pos. 6). Theimpeller ring with blades rotates in a cham-ber formed from two fixed housing sections,each of which has a passage (7) adjacent tothe blades which starts at the level of the in-take port (A) and terminates where the fuelis forced out of the pump at system pressurethrough the fuel outlet (B). The “Stopper”between start and end of the passage pre-vents internal leakage.

At a given angle and distance from the in-take opening a small degassing bore hasbeen provided which provides for the exit ofany gas bubbles which may be in the fuel.This, although improving the hot-deliverycharacteristics, is at the cost of very slight in-ternal leakage. The degassing bore is notneeded with diesel applications.

Pressure builds up along the passage (7) as aresult of the exchange of pulses between thering blades and the liquid particles. Thisleads to spiral-shaped rotation of the liquidvolume trapped in the impeller ring and inthe passages. In the case of the peripheralpump (Fig. 2c), the ring blades around theperiphery of the ring are surrounded com-pletely by the passage (hence the word “pe-ripheral”). On the side-channel pump, thetwo channels are located on each side of theimpeller ring adjacent to the blades.

Turbine pumps feature a low noise levelsince pressure buildup takes place continu-ously and is practically pulsation-free. Effi-ciency is between 10 % and about 20 %.Construction though is far simpler than thatof the positive-displacement pumps.

Single-stage pumps can generate systempressures of up to 450 kPa. In future, turbinepumps will also be suitable for the highersystem pressures that will be needed forbrief periods on highly supercharged en-gines and gasoline direct-injection engines.

For costs reasons, and due to their beingquieter, turbine pumps are used almost ex-clusively on newly designed gasoline-engineautomobiles.

Fuel supply Electric fuel pump 43

Figure 2a Roller-cell pump

(RZP)b Inner-gear pump

(IZP)c Peripheral pump

(PP)A Intake portB Outlet1 Slotted rotor

(eccentric)2 Roller3 Inner drive wheel 4 Rotor5 Impeller ring6 Impeller-ring blades7 Passage

(peripheral)8 “Stopper”

a 1

b

c

A

B

B

A

B

A

B A

A

B B

A

3 4

5 7 68 6 757

2

Principle of functioning of electric fuel pumps2

æU

MK

0267

-3Y

Robert Bosch GmbH

Fuel filterThe injection systems for automobile spark-ignition (SI) engines operate with extremeprecision. In order not to damage their pre-cision parts, it is imperative that the fuel isefficiently cleaned. Filters in the fuel circuitremove the solid particles which could causewear. Such filters are either replaceable in-line filters, or are integrated in the fueltank as “lifetime” in-tank filters. Apart fromthe filter’s purely straining or filtering effect,a number of different processes are appliedin order to remove the contaminants fromthe fuel. These include impact, diffusion,and blocking effects.

The filtration efficiency of the individual ef-fects is a function of the size and the flowspeed of the contaminant particles, and thefilter medium is matched to these factors.

Pleated paper, which is sometimes speciallyimpregnated, has come to the forefront asthe filter medium (Fig. 1, Pos. 3). The filtermedium is arranged in the fuel circuit sothat the velocity of the fuel flow through allsections of its surface is as uniform as possi-ble.

Whereas on manifold-injection systemsthe filter element has a mean pore size of10 µm, far finer filtering is needed for gaso-line direct-injection systems where up to85 % of the particles larger than 5 µm mustbe reliably filtered out of the fuel.

In addition, for gasoline direct injection,when a new filter is fitted the traces of cont-aminant remaining in the filter after manu-facture are an important factor: Metal, min-eral, plastic, and glass-fiber particles mustnot exceed 200 µm.

Depending upon the filter volume, the use-ful life (guaranteed mileage) of the conven-tional in-line filter is somewhere between37,500 and 55,000 miles (60,000...90,000km). Guaranteed mileages of 100,000 miles(160,000 km) apply for in-tank filters. Thereare in-tank and in-line filters available foruse with gasoline direct-injection systemswhich feature service lives in excess of150,000 miles (250,000 km).

Filter housings (2) are either steel, alu-minum, or plastic (100 % free from metal).Connections of the threaded, hose, or quick-connect type are used.

Filter efficiency depends on the throughflowdirection. When replacing in-line filters, it isimperative that the flow direction given bythe arrow is observed.

44 Fuel supply Fuel filter

Figure 11 Filter cover2 Filter housing3 Filter element 4 Support plate

1

2

3

4

Section through a fuel filter1

æU

MK

1779

Y

Robert Bosch GmbH

Fuel railManifold injectionThe fuel rail has the following assignments:

� Mounting and location of the injectors,� Storage of the fuel volume,� Ensuring that fuel is distributed evenly to

all injectors.

In addition to the injectors, the fuel rail usu-ally accomodates the fuel-pressure regulatorand possibly even a pressure damper. Localfuel-pressure fluctuations caused by reso-nance when the injectors open and closed, isprevented by careful selection of the fuel-raildimensions. As a result, irregularities in in-jected fuel quantity which can arise as afunction of load and engine speed areavoided.

Depending upon the particular require-ments of the vehicle in question, plastic orstainless-steel fuel rails are used. The fuel railcan incorporate a diagnosis valve for work-shop testing purposes.

Gasoline direct injectionOn gasoline DI systems, the rail is locateddownstream of the high-pressure pump, andis an integral part of the high-pressure stage.

Fuel-pressure regulator

Manifold injectionThe amount of fuel injected by the injector(injected fuel quantity) depends upon theinjection period and the difference betweenthe fuel pressure in the fuel rail and thecounterpressure in the manifold. On fuelsystems with return, the influence of pres-sure is compensated for by a pressure regu-lator which maintains the difference be-tween fuel pressure and manifold pressure ata constant level. This pressure regulator per-mits just enough fuel to return to the tankso that the pressure drop across the injectorsremains constant. In order to ensure that thefuel rail is efficiently flushed, the fuel-pres-

sure regulator is normally located at the endof the rail which leads the fuel tank.

On returnless fuel systems (RLFS), thepressure regulator is part of the in-tank unitinstalled in the fuel tank. The fuel-rail pres-sure is maintained at a constant level withreference to the surrounding pressure. Thismeans that the difference between fuel-railpressure and manifold pressure is not con-stant and must be taken into account whenthe injection duration is calculated.

The fuel-pressure regulator (Fig. 1) is of thediaphragm-controlled overflow type. A rub-ber-fabric diaphragm (4) divides the pres-sure regulator into a fuel chamber and aspring chamber. Through a valve holder (3)integrated in the diaphragm, the spring (2)forces a movable valve plate against the valveseat so that the valve closes. As soon as thepressure applied to the diaphragm by thefuel exceeds the spring force, the valve opensagain and permits just enough fuel to flowback to the fuel tank that equilibrium offorces is achieved again at the diaphragm.

Fuel supply Fuel rail, fuel-pressure regulator 45

Figure 11 Intake-manifold con-

nection2 Spring3 Valve holder4 Diaphragm5 Valve6 Fuel inlet7 Fuel return

1

2

5

6 6

7

34

Fuel-pressure regulator DR21

æU

MK

1781

Y

Robert Bosch GmbH

On multipoint fuel-injection systems, in or-der that the manifold vacuum can be ap-plied to the spring chamber, this is con-nected pneumatically to the intake manifoldat a point downstream of the throttle plate.There is therefore the same pressure ratio atthe diaphragm as at the injectors. Thismeans that the pressure drop across the in-jectors is solely a function of spring forceand diaphragm surface area, and thereforeremains constant.

Gasoline direct injectionOn gasoline direct-injection systems, it isnecessary to regulate the pressures in thehigh-pressure and the low-pressure stage,whereby the same fuel-pressure regulatorsare used for the low-pressure stage as formanifold injection.

Fuel-pressure damper

The repeated opening and closing of the in-jectors, together with the periodic supply offuel when electric positive-displacement fuelpumps are used, leads to fuel-pressure oscil-lations. These can cause pressure resonanceswhich adversely affect fuel-metering accu-racy. It is even possible that under certaincircumstances, noise can be caused by theseoscillations being transferred to the fuel tankand the vehicle bodywork through themounting elements of the fuel rail, fuellines, and fuel pump.

These problems are alleviated by the useof special-design mounting elements andfuel-pressure dampers.The fuel-pressuredamper is similar in design to the fuel-pres-sure regulator. Here too, a spring-loaded di-aphragm separates the fuel chamber fromthe air chamber. The spring force is selectedsuch that the diaphragm lifts from its seat assoon as the fuel pressure reaches its workingrange. This means that the fuel chamber isvariable and not only absorbs fuel whenpressure peaks occur, but also releases fuelwhen the pressure drops. In order to alwaysoperate in the most favorable range whenthe absolute fuel pressure fluctuates due to

conditions at the manifold, the spring cham-ber can be provided with an intake-manifoldconnection.

Similar to the fuel-pressure regulator, thefuel-pressure damper can also be attached tothe fuel rail or installed in the fuel line. Inthe case of gasoline direct injection, it canalso be attached to the high-pressure pump.

Fuel tank

As its name implies, the fuel tank is used asthe reservoir for the fuel. It must be non-corroding and must remain free of leaks atup to twice working pressure, or up to atleast 0.03 MPa (0.3 bar) gauge pressure.Openings or safety valves must be providedfor excess pressure to escape automatically.During cornering, on inclines, and in case ofshock or impact, no fuel may leak outthrough the filler cap or pressure-compensa-tion devices. The fuel tank must be situatedfar enough from the engine to avoid ignitionof escaping fuel in case of an accident.

Fuel lines

The fuel lines serve to carry the fuel fromthe fuel tank to the fuel-injection system.Seamless, flexible metal conduit or fuel-re-sistant hardly combustible material can beused for the fuel lines. These must be routedso that mechanical damage is avoided, andfuel which has evaporated or dripped as aresult of malfunctions cannot accumulate orignite. All fuel-carrying components mustbe protected against heat that could interferewith correct performance. Gravity feed mustnot be used in the fuel-supply circuit.

46 Fuel supply Fuel-pressure damper, fuel tank, fuel lines

Robert Bosch GmbH

6 4

1

2

1

4a

1

3

2

2

5

5

7

a K-/KE-Jetronicwith electric (in-line)fuel pump.

b L-Jetronic/Motronicwith electric (in-line)fuel pump.

c L-Jetronic/Motronicwith electric (in-tank)fuel pump.

d Mono-Jetronicwith electric (in-tank)fuel pump.

3

6 4

53

2

1

6

5

3

6

Fuel supply Fuel-supply systems 47

Figure 11 Fuel tank2 Electric fuel pump

(EKP)3 Fuel filter4 Fuel rail4a Fuel distributor

(K-/KE-Jetronic)5 Injector6 Pressure regulator7 Fuel accumulator

(K-/KE-Jetronic)

Development of fuel-supply systems (examples)1

æU

MK

1780

E

Robert Bosch GmbH

Manifold-injection engines generate the A/Fmixture in the intake manifold and not inthe combustion chamber. Since they wereintroduced to the market, these engines andtheir control systems have been vastly improved. Their superior fuel-meteringcharacteristics have enabled them to com-pletely supersede the carburetor enginewhich also operates with external A/F-mix-ture formation.

Overview

Regarding smooth running and exhaust-gasbehaviour very high demands are made onmodern-day vehicles which correspond tothe latest state-of-the-art. This leads to strictrequirements with respect to the composi-tion of the A/F mixture. Apart from the pre-cision metering of the injected fuel mass as afunction of the air drawn in by the engine, itis also imperative that injection of the fueltakes place at exactly the right instant intime.

As a direct result of increasingly severe emis-sion-control legislation, these technical stip-ulations are being increasingly tightened sothat fuel-injection system development isforced to keep pace.

In the manifold-injection field, the elec-tronically controlled multipoint fuel-injec-tion system represents the state-of-the-art.This system injects the fuel intermittently,and individually, for each cylinder directlyonto its intake valve(s) (Fig. 1).

Mechanically controlled continuous-injec-tion multipoint systems no longer have anysignificance for new developments in thisfield, nor do the single-point (TBI) systemswhich inject intermittently through a singleinjector into the intake manifold upstreamof the throttle valve.

48 Manifold fuel injection Overview

Manifold fuel injection

Figure 11 Cylinder with piston2 Exhaust valves3 Ignition coil with

spark plug4 Intake valves5 Injector6 Intake manifold

64

2

1

3

5

Manifold injection1æ

UM

K17

76Y

Robert Bosch GmbH

Operating conceptGasoline injection systems of the manifold-injection type are characterized by the factthat they generate the A/F mixture outsidethe combustion chamber, in other words, inthe intake manifold (Fig. 1), see “ExternalA/F-mixture formation”. The injector (5)sprays the fuel directly onto the intake valves(4) where together with the intake air itforms the A/F mixture which is then drawninto the cylinder (1) past the open intakevalves during the subsequent inductionstroke. One, two, or even three, intake valvescan be used per cylinder.

The intake valves are designed so that theengine’s fuel requirements are covered irre-spective of operating conditions – at fullload and at high engine revs.

A/F-mixture formationFuel injectionThe electric fuel pump delivers the fuel tothe injectors where it is then available for in-jection at system pressure. Each cylinder isallocated its own injector which injects in-termittently into the intake manifold di-rectly onto the intake valve (6). Here thefinely atomized fuel evaporates to a great extent, and together with the intake air en-tering via the throttle plate generates the A/Fmixture. In order that enough time is avail-able for the generation of the A/F mixture,the fuel is best sprayed onto the closed intake valve and “stored” there.

Some of the fuel is deposited as a film on themanifold walls in the vicinity of the intakevalves. The thickness of the film is a func-tion of the manifold pressure and, therefore,of engine load. For good dynamic engine response, the fuel mass in the wall film mustbe kept to a minimum. This is achieved byappropriate manifold design and fuel-spraygeometry. Since the injector is situated directly opposite the intake valve, the wall-film effects with multipoint injection systems are far less serious than they werewith the former TBI and carburetor systems.

Provided the A/F mixture is stoichiometric(λ = 1), the pollutants generated during thecombustion process can to a great extent beremoved using the three-way catalytic con-verter. At the majority of their operatingpoints, manifold-injection engines are there-fore operated with this A/F mixture compo-sition.

Measuring the air massIn order that the A/F mixture can be pre-cisely adjusted, it is imperative that the massof the air which is used for combustion canbe measured exactly. The air-mass meter issituated upstream of the throttle valve. Itmeasures the air-mass flow entering the in-take manifold and sends a correspondingelectric signal to the engine ECU. As an al-ternative, there are also systems on the mar-ket which use a pressure sensor to measurethe intake-manifold pressure. Together withthe throttle-valve setting and the enginespeed, this data is then used to calculate theintake-air mass. The ECU then applies thedata on intake air mass and the engine’s in-stantaneous operating mode to calculate therequired fuel mass.

Injection durationA given length of time is needed for the in-jection of the calculated fuel mass. This istermed the injection duration, and is a func-tion of the injector’s opening cross sectionand the difference between the intake-mani-fold pressure and the pressure prevailing inthe fuel-supply system.

Manifold fuel injection Operating concept 49

Robert Bosch GmbH

Electromagnetic fuel injectorsAssignmentThe electromagnetic (solenoid-controlled)fuel injectors spray the fuel into the intakemanifold at system pressure. They permitthe precise metering of the quantity of fuelrequired by the engine. They are triggeredvia ECU driver stages with the signal calcu-lated by the engine-management system.

Design and operating conceptEssentially. the electromagnetic injectors(Fig. 1) are comprised of the following com-ponents:

� The injector housing (9) with electrical(8) and hydraulic (1) connections,

� The coil for the electromagnet (4),� The movable valve needle (6) with sol-

enoid armature and sealing ball,� The valve seat (10) with the injection-ori-

fice plate (7), and the� Spring (5).

In order to ensure trouble-free operation,stainless steel is used for the parts of the in-jector which come into contact with fuel.The injector is protected against contamina-tion by a filter strainer (3) at the fuel input.

ConnectionsOn the injectors presently in use, fuel supplyto the injector is in the axial direction, thatis, from top to bottom (“top feed”). The fuelline is fastened to the injector using a specialclamp. Retaining clips ensure reliable align-ment and fastening. The seal ring (2) on thehydraulic connection (1) seals off the injec-tor at the fuel rail.

The injector is connected electrically tothe engine ECU.

Injector operationWith no voltage across the solenoid (sol-enoid de-energised), the valve needle andsealing ball are pressed against the cone-shaped valve seat by the spring and the forceexerted by the fuel pressure. The fuel-supplysystem is thus sealed off from the manifold.As soon as the solenoid is energised (excita-tion current), this generates a magnetic fieldwhich pulls in the valve-needle armature.The sealing ball lifts off the valve seat andthe fuel is injected. As soon as the excitationcurrent is switched off, the valve needlecloses again due to spring force.

Fuel outletThe fuel is atomized by means of an injec-tion-orifice plate in which there are a num-ber of holes. These holes (spray orifices) arestamped out of the plate and ensure that theinjected fuel quantity remains highly repro-ducible. The injection-orifice plate is insen-sitive to fuel deposits. The spray pattern ofthe fuel leaving the injector is a function ofthe number of orifices and their configura-tion.

The highly efficient injector sealing at thevalve seat is due to the cone/ball sealingprinciple.

The injector is inserted into the openingprovided for it in the intake manifold. Thebottom seal ring provides the seal betweenthe injector and the intake manifold. Essen-tially, the injected fuel quantity per unit oftime is determined by � The fuel-supply system pressure,� The counterpressure in the intake mani-

fold,� The geometry of the fuel-exit area.

Types of constructionIn the course of time, the injectors have beenfurther and further developed to matchthem to the ever-increasing demands re-garding engineering, quality, reliability, andweight. This has led to a variety of differentinjector designs.

50 Manifold fuel injection Electromagnetic fuel injectors

Robert Bosch GmbH

EV6 injectorThe EV6 injector is the standard injector fortoday’s modern fuel-injection systems (Figs. 1 and 2a). It is characterized by itssmall external dimensions and its lowweight. This injector therefore already pro-vides one of the prerequisites for the designof compact intake modules.

In addition, the EV6 is also outstandingwith regard to its hot-fuel behaviour, that is,there is very little tendency for vapor-bubbleformation when using hot fuel. This facili-tates the use of RLFS fuel-supply systems inwhich the fuel temperature in the injector ishigher than with systems featuring fuel return. Thanks to wear-resisting surfaces,the fuel quantities injected by the EV6 re-main highly reproducible over long periodsof time, and the injector features a long use-ful life.

Thanks to their highly efficient sealing,these injectors fulfill all future requirementsregarding zero evaporation. That is, no fuelvapor escapes from them.

The EV6 variant with “air shrouding” wasdeveloped especially to comply with require-

ments for even better fuel atomization.Finely vaporized fuel can be generated usingother methods: In future, in addition to 4-hole injection-orifice plates, multi-orificeplates with between 10 and 12 holes will beused. Injectors equipped with these multi-orifice plates generate a very fine fuel fog.

There are a wide variety of injectors avail-able for different areas of application. Thesefeature different lengths, flow classes, andelectrical characteristics. The EV6 is alsosuitable for use with fuels having an ethanolcontent of as much as 85 %.

EV14 injectorFurther injector development has led to theEV14 (Fig. 2b) which is based on the EV6. Itis even more compact, a fact which facili-tates its integration in the fuel rail.

The EV14 is available in 3 differentlengths (compact, standard, long). Thismakes it possible to adapt individually to theengine’s intake-manifold geometry.

Manifold fuel injection Electromagnetic fuel injectors 51

Figure 11 Hydraulic connec-

tion2 Seal rings

(O-rings)3 Filter strainer4 Coil5 Spring6 Valve needle with

armature and sealingball

7 Injection-orifice plate8 Electrical connec-

tion9 Injector housing10 Valve seat

Figure 2a EV6 Standardb EV14 Compact

1

2

3

9

2

8

10

7

4

5

6

Design of the EV6 electromagnetic injector1

æU

MK

1712

-3Y

a

b

Injector versions2

æU

MK

1786

Y

Robert Bosch GmbH

Spray formationThe injector’s spray formation, that is, itsfuel-spray shape, spray-dispersal angle, andfuel-droplet size, have a considerable influ-ence upon the generation of the A/F mix-ture. Different versions of spray formationare required in order to comply with the demands of individual intake-manifold andcylinder-head geometries. Fig. 3 shows themost important fuel-spray shapes.

“Pencil” sprayA thin, concentrated, and highly-pulsed fuelspray results from using a single-hole injec-tion-orifice plate. This form of spray practi-cally eliminates the wetting of the manifoldwall. Such injectors are most suitable for usewith narrow intake manifolds, and in instal-lations in which the fuel has to travel longdistances between the point of injection andthe intake valve.

The pencil-spray injector is only used inisolated cases due to its low level of fuel atomization.

Tapered sprayA number of individual jets of fuel leave theinjection-orifice plate. The tapered spraycone results from the combination of thesefuel jets.

Although engines with only 1 intake valveper cylinder typically use tapered-spray injectors, they are also suitable for engineswith 2 intake valves per cylinder.

Dual sprayThe dual-spray formation principle is oftenapplied on engines with 2 intake valves percylinder. Engines with 3 intake valves percylinder must be equipped with dual-sprayinjectors.

The holes in the injection-orifice plate areso arranged that two fuel sprays leave the injector and impact against the respectiveintake valve or against the web between theintake valves. Each of these sprays can beformed from a number of individual sprays(2 tapered sprays).

The spray offset angleReferred to the injector’s principle axis, thefuel spray in this case (single spray and dualspray) is at an angle, the spray offset angle(γ). Injectors with this spray shape aremostly used when installation conditions aredifficult.

Types of fuel injection

In addition to the duration of injection, afurther parameter which is important foroptimisation of the fuel-consumption andexhaust-gas figures is the instant of injectionreferred to the crankshaft angle. Here, thepossible variations are dependent upon thetype of injection actually used (Fig. 1).

The new injection systems provide for ei-ther sequential fuel injection or cylinder-in-dividual fuel injection (SEFI and CIFI re-spectively).

52 Manifold fuel injection Eletromagnetic fuel injectors, types of fuel injection

Figure 3a Pencil sprayb Tapered sprayc Dual sprayd Spray offset angle

α80: 80% of the injectedfuel is within theangle defined by α

α50: 50% of the injectedfuel is within theangle defined by α

�: 70% of the injectedfuel in a single sprayis within the angledefined by �

γ: Spray offset angle

a

α80

α50

b

c d

α80

β 7°γ

Fuel-spray shapes3

æU

MK

1774

Y

Robert Bosch GmbH

Simultaneous fuel injectionAll injectors open and close together in thisform of fuel injection. This means that thetime which is available for fuel evaporationis different for each cylinder. In order tonevertheless obtain efficient A/F-mixtureformation, the fuel quantity needed for thecombustion is injected in two portions. Halfin one revolution of the crankshaft and theremainder in the next. In this form of injec-tion, the fuel for some of the cylinders is notstored in front of the particular intake valvebut rather, since the valve has opened, thefuel is injected into the open intake port.The start of injection cannot be varied.

Group injectionHere, the injectors are combined to formtwo groups. For one revolution of the crank-shaft, one injector group injects the totalfuel quantity required for its cylinders, andfor the next revolution the second group in-jects.

This configuration enables the start of injec-tion to be selected as a function of engine-

operating point. Apart from this, the un-desirable injection into open inlet ports isavoided. Here too, the time available for theevaporation of fuel is different for eachcylinder.

Sequential fuel injection (SEFI)The fuel is injected individually for eachcylinder, the injectors being actuated one af-ter the other in the same order as the firingsequence. Referred to piston TDC, the dura-tion of injection and the start of injectionare identical for all cylinders, and the fuel isstored in front of each cylinder.

Start of injection is freely programmableand can be adapted to the engine’s operatingstate.

Cylinder-individual fuel injection (CIFI)This form of injection provides for thegreatest degree of design freedom. Com-pared to sequential fuel injection, CIFI hasthe advantage that the duration of injectioncan be individually varied for each cylinder.This permits compensation of irregularites,for instance with respect to cylinder charge.

Manifold fuel injection Types of fuel injection 53

Figure 1a Simultaneous fuel

injectionb Group fuel injectionc Sequential fuel injec-

tion (SEFI) andcylinder-individualfuel injection (CIFI)

Intake valve openInjectionIgnition

-360°Firing sequence

0°TDC cyl. 1

360° 720° 1080° cks

aCyl. 3Cyl. 4

Cyl. 1

Cyl. 2

Cyl. 1bCyl. 3Cyl. 4Cyl. 2

Cyl. 1cCyl. 3Cyl. 4Cyl. 2

Manifold fuel injection: Types of fuel injection1

æS

MK

1311

-1E

Robert Bosch GmbH

Gasoline direct-injection engines generatethe A/F mixture in the combustion cham-ber. During the induction stroke, only thecombustion air flows past the open intakevalve and into the cylinder. The fuel is in-jected directly into the cylinders by specialinjectors.

Overview

The demand for higher-power engines, cou-pled with the requirement for reduced fuelconsumption, were behind the “re-discov-ery” of gasoline direct injection. As far backas 1937, an engine with mechanical gasolinedirect injection took to the air in an air-plane. In 1952, the “Gutbrod” was the firstpassenger car with a series-production me-chanical gasoline direct-injection engine,and in 1954 the “Mercedes 300 SL” followed.

At that time, designing and building a di-rect-injection engine was a very complicatedbusiness. Moreover, this technology madeextreme demands on the materials used. Theengine’s service life was a further problem.

These facts all contributed to it taking solong for gasoline direct injection to achieveits breakthrough.

54 Gasoline direct injection: Overview

Gasoline direct injection

Figure 11 High-pressure pump2 Low-pressure

connection 3 High-pressure line4 Fuel rail5 High-pressure

injectors6 High-pressure

sensor7 Spark plug8 Pressure-control

valve9 Piston

1 2 34

6

78

9

5

Gasoline direct injection: Components1

æU

MK

1783

Y

Robert Bosch GmbH

Operating conceptGasoline direct-injection systems are charac-terized by injecting the fuel directly into thecombustion chamber at high pressure. Simi-lar to the diesel engine, A/F-mixture forma-tion takes place inside the cylinder (internalA/F-mixture formation).

Generation of high-pressure The electric fuel pump delivers fuel to thehigh-pressure pump (Fig. 1, Pos. 1) at a pri-mary pressure of 0.3...0.5 MPa (3...5 bar).Depending on the engine operating point(required torque and engine speed), thehigh-pressure pump then generates the sys-tem pressure which forces the fuel, which isnow at high pressure, into the rail (4) whereit is stored until required for injection.

The fuel pressure is measured by the high-pressure sensor (6) and adjusted to valuesbetween 5...12 MPa by the pressure-controlvalve (8).

The high-pressure injectors (5) are installedin the rail (also referred to as the “Commonrail”) and, when triggered by the engineECU, inject the fuel directly into the com-bustion chambers.

A/F-mixture formationThe injected fuel is finely atomized due tothe very high injection pressure. Togetherwith the drawn-in air, it forms the A/F mix-ture in the combustion chamber. Dependingupon the engine’s operating mode, the fuelis injected in such a manner that an A/Fmixture with λ ≤ 1 is evenly distributedthroughout the complete combustion cham-ber (homogeneous operation), or a strati-fied-charge A/F-mixture cloud (λ ≤ 1) isformed around the spark plug (lean-burnoperation or stratified-charge operation).During stratified-charge operation, the re-mainder of the cylinder is filled with eitherfreshly drawn-in air, with inert gas returnedto the cylinder by EGR, or with a very leanA/F mixture. The overall A/F mixture thenhas ltotal λtotal > 1.

The various methods of running the engineas listed above are referred to as the engine’soperating modes. On the one hand, the se-lection of the operating mode to be appliedis a function of engine speed and desiredtorque, and on the other it depends uponfunctional requirements such as the regener-ation of the accumulator-type catalytic con-verter.

TorqueDuring stratified-charge operation, the in-jected fuel mass is decisive for the generatedtorque. The excess air permits “unthrottled”operation, also at part load, with the throttleopened wide. This measure reduces thepumping (exhaust and refill) work, andtherefore also serves to lower the fuel con-sumption.

In homogeneous and lean-burn operation atλ > 1 and homogeneous A/F-mixture distri-bution, “unthrottling” also results in fuelsavings, although not to the same extent asin stratified-charge operation.

In homogeneous operation at λ ≤ 1, thegasoline direct-injection engine for the mostpart behaves the same as a manifold-injec-tion engine.

Exhaust treatmentThe catalytic converter is responsible for re-moving the pollutants from the exhaust gas.In order to operate with maximum effi-ciency, the 3-way catalytic converter needs astoichiometric A/F mixture. Due to excessair, lean-burn operation results in increasedlevels of NOx emissions which are storedtemporarily in an accumulator-type NOx

catalytic converter. These are then reducedto nitrogen, carbon dioxide and water byrunning the engine briefly with excess air.

Gasoline direct injection: Operating concept 55

Robert Bosch GmbH

RailThe rail stores the fuel delivered by the high-pressure pump and distributes it to thehigh-pressure fuel injectors. The rail’s vol-ume is sufficient to compensate for pressurepulsations in the fuel circuit.

An aluminum rail is used. Design and con-struction (volume, dimensions, weight etc.)are specific to the engine and the system.

The rail is provided with connections for anumber of the injection-system components(high-pressure pump, pressure-controlvalve, high-pressure sensor, high-pressureinjectors). Construction guarantees thatthere are no leaks in the rail itself, nor at itsinterfaces.

High-pressure pumpsAssignmentIt is the job of the high-pressure pump(HDP) to compress the fuel delivered by theelectric fuel pump at a primary pressure of0.3...0.5 MPa. It must provide enough fuel atthe pressure (5...12 MPa) needed for thehigh-pressure injection.

Initially, when starting the engine, the fuel isinjected at the primary pressure. The highpressure is built up when the engine runs upto speed. The minimal level of pumping-flow pulsation means so that there is verylittle pulsation in the rail.

In order to prevent the fuel mixing with anoil lubricant, the high-pressure pump iscooled and lubricated by fuel.

56 Gasoline direct injection: Rail, high-pressure pump

Figure 11 Eccenter cam2 Sliding block3 Pumping element

with pump piston(hollow piston, fuelinlet)

4 Sealing ball5 Outlet valve6 Inlet valve7 High-pressure con

nection to the rail8 Fuel inlet

(low presure)9 Eccenter ring10 Axial face seal11 Static seal12 Driveshaft

1

8

5

9

10

2

3

4

11

13

7

6

12

Three-cylinder high-pressure pump HDP1 (axial section)1

æU

MK

1785

Y

Robert Bosch GmbH

Three-cylinder high-pressure pumpHDP1There are many different types of high-pres-sure pumps available. Fig. 1 shows an axialsection, and Fig. 2 a cross section throughthe HDP1 three-cylinder radial-pistonpump. Driven by the engine camshaft, thedriveshaft (12) rotates with the eccenter cam(1) which is responsible for the up anddown motion of the pistons (3) in theircylinders. When the piston moves down-ward, fuel flows at the primary pressure of0.3...0.5 MPa from the fuel line through thehollow pump piston and the inlet valve (6)into the pump cylinder. When the pistonmoves upward this volume of fuel is com-pressed and when the rail pressure isreached the outlet valve (5) opens and thefuel is forced out to the high-pressure con-nection (7).

The use of three pump cylinders at an angleof 120° to each other results in very low lev-els of residual pulsation in the rail. The de-livery quantity is proportional to the rota-tional speed. So that there is always enoughfuel available, and in order to limit the fuelwarm-up in the rail, at maximum deliverythe high-pressure pump delivers slightlymore fuel than the maximum needed by theengine. The pressure-control valve releasesthe pressure of the excess fuel and then di-rects this into the return line.

Single-cylinder high-pressure pumpHDP2The HDP2 single-cylinder pump is a cam-driven radial-piston pump with variable de-livery quantity. When the piston movesdownward, fuel flows at the primary pressureof 0.3...0.5 MPa from the fuel line, throughthe inlet valve and into the pump cylinder.When the piston moves upwards this volume

Gasoline direct injection: High-pressure pump 57

Figure 2(Position numbersidentical to Fig. 1)1 Eccenter cam2 Sliding block3 Pumping element

with pump piston5 Outlet valve6 Inlet valve9 Eccenter ring

2 cm

1

10

4

3

7

6

2

Three-cylinder high-pressure pump HDP 1 (cross-section) 2

æU

MK

1784

Y

Robert Bosch GmbH

of fuel is compressed and forced into the railas soon as it exceeds the rail pressure.

The pump chamber and the fuel inlet areconnected with each other through a trig-gerable delivery-quantity control valve. Ifthis valve is triggered and opens before theend of the delivery stoke, the pressure in thepump chamber collapses and the fuel flowsback into the fuel inlet. This means that thedelivery-control valve has the same functionas the pressure-control valve in the three-cylinder HDP1 pump.

In order to adjust the delivery quantity, thequantity control valve remains closed frompump-cam BDC until a given stroke hasbeen completed. Once the required rail pres-sure is reached the valve opens and preventsfurther pressure increase in the rail.

The maximum delivery quantity (l/h) is afunction of the rotational speed, the numberof cams, and the cam lift. The delivery quan-tity can be adjusted to comply with require-ments by triggering the control valve ac-cordingly.

The non-return valve between the pumpchamber and the rail prevents the rail pres-sure dropping when the delivery-quantitycontrol valve opens.

Pressure-control valveAssignmentThe pressure-control valve is located be-tween the rail and the low-pressure side ofthe HDP1 high-pressure pump. It adjuststhe required pressure in the rail by changingthe flow cross-section. The excess fuel deliv-ered by the HDP1 flows into the low-pres-sure circuit.

Design and operating conceptThe solenoid is triggered by a pwm signal(Fig. 1, Pos. 3). The valve ball (7) lifts fromthe valve seat (8) and in doing so changesthe valve’s flow cross-section as required.

With no current flowing, the pressure-con-trol valve is closed. This is a safety measureto ensure adequate rail pressure in case ofmalfunction in the electrical-triggering circuit. A pressure-limiting function is incorporated to prevent excessive rail pres-sure which could otherwise damage thecomponents.

58 Gasoline direct injection: High-pressure pump, pressure-control valve

Figure 11 Electrical connec-

tion2 Spring3 Solenoid coil4 Solenoid armature5 Seal rings (O-rings)6 Outlet passage7 Valve ball8 Valve seat9 Inlet with inlet

strainer

2

3

1

4

5

576

89

Section through the pressure-control valve1

æS

MK

1812

Y

Robert Bosch GmbH

Rail-pressure sensorsAssignmentThe rail-pressure sensors used in the Com-mon Rail and MED-Motronic systems mea-sure the fuel pressure in the high-pressurefuel reservoir (fuel rail). Precisely maintain-ing the stipulated fuel pressure in the rail isof extreme importance with respect to theengine’s power output, toxic emissions, andnoise. Fuel pressure is controlled in a specialcontrol loop, deviations from desired valuebeing compensated for by an open-loop orclosed-loop pressure-control valve.

Very tight tolerances apply to the rail-pres-sure sensors, and in the main operatingrange, the measuring accuracy is below 2%of the measuring range.

Rail-pressure sensors are used with the fol-lowing engine systems:

� Common Rail diesel accumulator-type in-jection systemThe maximum working pressure pmax

(rated pressure) is 160 MPa (1600 bar).

� Gasoline direct injection MED-MotronicThe working pressure in such a gasolinedirect injection system is a function of thetorque and engine speed.It is 5 ... 12 MPa (50 ... 120 bar).

Design and operating conceptA steel diaphragm is at the heart of the rail-pressure sensor. Deformation-dependentmeasuring resistors are vapor deposited onthe diaphragm in the form of a bridge cir-cuit (Fig. 1, Pos. 3). The sensor’s measuringrange is a function of the diaphragm thick-ness (thicker diaphragms are used for higherpressures, and thinner ones for lower pres-sures).As soon as the pressure to be measured isapplied to one side of the diaphragm via thepressure connection (4) the deformation-dependent resistors change their values due

to the bending of the diaphragm (approx.20 µm at 1500 bar). By means of collectingleads, the 0 ... 80 mV output voltage signalgenerated by the bridge is transferred to anevaluation circuit (2) in the sensor and am-plified to 0 ... 5 V. This is then passed on tothe ECU which uses it, together with astored characteristic curve, to calculate thepressure (Fig. 2)

Gasoline direct injection: Rail-pressure sensors 59

Fig. 11 Electrical connec-

tion (plug) 2 Evaluation circuit 3 Steel diaphragm

with deformation-dependent resistors

4 Pressure connection 5 Mounting thread

0

4.5

0.5

Pressure

Out

put v

olta

ge

V

pmax

Rail-pressure sensor: Charcteristic curve (example)2

æU

MK

0719

-2E

2

3

4

5

p

1

2 cm

Rail-pressure sensor (design)1

æU

MK

1576

-1Y

Robert Bosch GmbH

High-pressure injectorAssignmentThe high-pressure injector represents the in-terface between the rail and the combustionchamber. Its job is to meter the fuel, and bymeans of the fuel’s atomisation achieve con-trolled mixing of the fuel and air in a spe-cific area of the combustion chamber. De-pending upon the required operating mode,the fuel is either concentrated in the vicinityof the spark plug (stratified charge), orevenly distributed throughout the combus-tion chamber (homogeneous distribution).

Design and operating conceptThe high-pressure injector (Fig. 1) com-prises the following components:

� Injector housing (5),� Valve seat (7),� Nozzle needle with solenoid armature (6),� Spring (3), and� Solenoid (4).

A magnetic field is generated when the sole-noid coil is energized (current flows). Thislifts the valve needle from the valve seatagainst the force of the spring and opens theinjector outlet passage (8). Fuel is then in-jected into the combustion chamber due tothe difference between rail pressure andcombustion-chamber pressure.

When the energising current is switchedoff, the spring forces the needle back downagainst its seat and injection stops.

The injector opens very quickly, guaran-tees a constant opened cross-section duringthe time it is open, and closes against the railpressure. Taking a given opened cross-sec-tion, the injected fuel quantity is thereforedependent upon the rail pressure, thecounter-pressure in the combustion cham-ber, and the length of time the injector remains open. Excellent fuel atomisation isachieved thanks to the special nozzle geome-try at the injector tip.

Compared to manifold injection, gasolinedirect injection can boast faster injection,improved precision of spray alignment, andbetter formation of the fuel spray.

Technical requirementsCompared with manifold injection, gasolinedirect injection differs mainly in its higherfuel pressure and the far shorter time whichis available for directly injecting the fuel intothe combustion chamber.

60 Gasoline direct injection: High-pressure injector

Figure 11 Fuel inlet with fine

strainer2 Electrical

connections3 Spring4 Solenoid 5 Injector housing6 Nozzle needle with

solenoid armature7 Valve seat8 Injector outlet

passage

6

7

8

1

3

2

5

4

High-pressure injector (HDEV): Design1

æU

MK

1782

Y

Robert Bosch GmbH

Fig. 2 underlines the technical demandsmade on the injector. With manifold injec-tion, two revolutions of the crankshaft areavailable for injecting the fuel into the mani-fold. At an engine speed of 6,000 min–1, thiscorresponds to 20 ms.

In the case of gasoline direct injectionthough, considerably less time must suffice.During homogeneous operation, the fuelmust be injected in the induction stroke. Inother words, only half a crankshaft rotationis available for the injection process.Referred to the same engine speed as withmanifold injection (6,000 min–1), this corresponds to an injection duration of only5 ms.

For gasoline direct injection, the fuel requirement at idle referred to that at WOTis far lower than with manifold injection (factor 1:12). At idle, this results in an injec-tion duration of approx. 0.4 ms.

Triggering the HDEV high-pressureinjector The injector must be triggered with a highly

complex current characteristic in order tocomply with the requirements for defined,reproducible fuel-injection processes (Fig. 3). The initial triggering signal fromthe microcontroller in the engine ECU issimply a digital signal (a). A special trigger-ing module uses this signal to generate theactual triggering signal (b) with which theHDEV driver stage triggers the injector.

A booster capacitor is used to generate the50...90 V trigger voltage which is highenough to provide a high current at the startof the switch-on process so that the valveneedle can lift off of the valve seat veryquickly (c). Once the valve needle has lifted(maximum needle lift), only a very low trig-gering current suffices to maintain the nee-dle at a constant opened position. With theneedle’s opened position constant, the in-jected fuel quantity is proportional to the in-jection duration (d).

The calculations for the duration of injec-tion take into account the premagnetisationtime before the valve needle actually lifts.

Gasoline direct injection High-pressure injector 61

Figure 2Injected fuel quantity asa function of the durationof injection

Figure 3a Triggering signalb Injector current

characteristicc Needle liftd Injected fuel quantity

0.4 3.5 5Duration of injection in ms

Idle

WOT

20

Gasoline direct

injection

Manifold injection

Inje

cted

fuel

qua

ntity

Comparison between gasoline direct injection andmanifold injection

2

æU

MK

1777

E

Duration of injection

Inje

cted

fu

elqu

antit

y

Ivm

tvm

ton0

0

1

0

b

c

d

Cur

rent

N

eedl

e lif

t

Premagnetization Ivm, tvm

toff

a

Imax

Ih

Signal characteristic for triggering the HDEV high-pressure injector

3

æS

MK

1772

E

Robert Bosch GmbH

Combustion processThe combustion process is defined as theway in which A/F-mixture formation andenergy conversion take place in the combus-tion chamber.

Depending upon the combustion processconcerned, flows of air are generated in thecombustion chamber. In order to obtain therequired charge stratification, the injector in-jects the fuel into the air flow in such a man-ner that it evaporates in a defined area. Theair flow then transports the A/F-mixturecloud in the direction of the spark plug sothat it arrives there at the moment of ignition.

Two basically different combustionprocesses are possible:

Spray-guided combustion processThe spray-guided process is characterised bythe fuel being injected in the spark plug’simmediate vicinity where it also evaporates(Fig. 1a). In order to be able to ignite the A/Fmixture at the correct moment in time (ig-nition point), it is imperative that spark plugand injector are exactly positioned, and thatthe spray direction is precisely aligned.

With this process, the spark plug is sub-jected to considerable thermal stressingsince under certain circumstances the hotspark plug can be directly impacted by therelatively cold jet of injected fuel.

Wall-guided combustion processIn the case of the wall-guided process, onedifferentiates between two possible flows ofair which are the result of specific intake-port and piston design. The injector injectsinto this air flow which transports the resulting A/F mixture to the spark plug inthe form of a closed A/F-mixture cloud.

Swirl air flowThe air drawn by the piston through theopen intake valve and into the cylinder gen-erates a turbulent flow (rotational air move-ment) along the cylinder wall (Fig. 1b). Thisprocess is also designated “swirl combustionprocess”.

Tumble air flowThis process produces a cylindrical air flow,or tumbling air flow, which in its movementfrom top to bottom is deflected by a pro-nounced piston recess so that it then movesupwards in the direction of the spark plug(Fig. 1c).

62 Gasoline direct injection Combustion process

Figure 1a Spray-guidedb Wall-guided swirl air

flowc Wall-guided tumble

air flow

a

b

c

Air-flow conditions for the various combustionprocesses

1

æU

MK

1778

Y

Robert Bosch GmbH

A/F-mixture formationAssignmentIt is the job of the A/F-mixture formation toprovide a combustible A/F mixture which isto be as homogeneous as possible.

Technical requirementsIn the “homogeneous” mode of operation(homogeneous λ ≤ 1 and homogeneouslean-burn), this A/F mixture is distributedhomogeneously throughout the whole of thecombustion chamber. During stratified-charge operation on the other hand, the A/Fmixture is only homogeneous within a re-stricted area, while the remaining areas ofthe combustion chamber are filled with in-ert gas or fresh air.

All fuel must have evaporated before a gasmixture or gas-vapor mixture can be termedhomogeneous. A number of factors influ-ence this process:

� Combustion-chamber temperature,� Fuel-droplet size, and� The time which is available for fuel evapo-

ration.

Influencing factorsTemperature influenceDepending upon temperature, pressure, andcombustion-chamber geometry, an A/Fmixture (air/gasoline) is combustible at λ = 0.6...1.6. Since gasoline cannot evaporatecompletely at low temperatures, this meansthat under these conditions more fuel mustbe injected in order to obtain a combustibleA/F mixture.

A/F-mixture formation in the homogeneousoperation mode The fuel is injected as soon as possible sothat the maximum length of time is avail-able for formation of the A/F mixture. Thisis why the fuel is injected in the inductionstroke during homogeneous operation. Theintake air can then assist in achieving rapidevaporation of the fuel and efficient ho-mogenisation of the A/F mixture.

A/F-mixture formation in the stratified-charge mode The configuration of the combustible A/F-mixture cloud which is in the vicinity of thespark plug at the instant of ignition is deci-sive for the stratified-charge mode. This iswhy the fuel is injected during the compres-sion stroke so that a cloud of A/F mixture isgenerated which can be transported to thevicinity of the spark plug by the air flows inthe combustion chamber, and by the pistonas it moves upwards. The ignition point is afunction of the engine speed and the required torque.

Penetration depthThe fuel-droplet size in the injected fuel is afunction of injection pressure and combus-tion-chamber pressure. Higher injectionpressures result in smaller droplets whichthen vaporize quicker. Taking a constantcombustion-chamber pressure, the so-calledpenetration depth increases along with increasing injection pressure. The penetra-tion depth is defined as the distance trav-elled by the individual fuel droplet before itvaporizes completely.

The cylinder wall or the piston will bewetted with fuel if the distance needed forfull vaporization exceeds the distance fromthe injector to the combustion-chamberwall. If the fuel on the cylinder wall and pis-ton has not vaporized before the ignitionpoint, either no combustion takes place atall, or it is incomplete.

Gasoline direct injection A/F-mixture formation 63

Robert Bosch GmbH

Operating modesThere are six operating modes in use withgasoline direct injection (Fig. 1):

� Stratified-charge mode,� Homogeneous mode,� Homogeneous and lean-burn,� Homogeneous and stratified-charge,� Homogeneous/anti-knock,� Stratified-charge/cat-heating.

These operating modes permit the best-pos-sible adaptation for each and every engineoperating state. During actual driving, thedriver does not notice the change-overs be-tween operating modes since these takeplace without torque surge.

The lines in the diagram (Fig. 1) showwhich operating modes are passed throughwhen accelerating strongly (pronouncedchanges in torque with at first unchangedengine speed), and when accelerating gently(slight changes in torque with increasing en-gine speed).

Stratified-charge modeIn the lower torque range at speeds up to ap-prox. 3000 min–1, the engine is operated inthe stratified-charge mode. Here, the injec-tor injects the fuel during the compressionstroke shortly before the ignition point.During the brief period before the ignition

point the air flow in the combustion cham-ber transports the A/F mixture to the sparkplug. Due to the late injection point, there isnot sufficient time to distribute the A/F mix-ture throughout the complete combustionchamber.

Referred to the combustion chamber as awhole, the A/F mixture is very lean in thestratified-charge mode. The untreated NOx

emissions are very high when the excess-airlevel is high. In this operating mode, the bestremedy is to use a high EGR rate, wherebythe recirculated exhaust gas reduces thecombustion temperature and, as a result,lowers the temperature-dependent NOx

emissions.

The parameters “Engine speed” and“Torque” define the limits for stratified-charge operation. In the case of excessivetorque, soot is generated due to zones of lo-cal rich-mixture. If engine speed is too high,charge stratification and efficient transportof the A/F mixture to the spark plug can nolonger be maintained due to excessive tur-bulence.

Homogeneous modeFor high torques and high engine speeds theengine is operated in the homogeneousmode λ = 1 (in exceptional cases with λ < 1)instead of in the stratified-charge mode. In-jection starts in the induction stroke, so thatthere is sufficient time for the A/F mixtureto be distributed throughout the whole ofthe combustion chamber. The injected fuelmass is such that the A/F mixture ratio isstoichiometric or, in exceptional cases, hasslightly excessive fuel (λ ≤ 1).

Since the whole of the combustion cham-ber is utilised, the homogeneous mode is re-quired when high levels of torque are de-manded. In this operating mode, emissionsof untreated exhaust gas are also low due tothe stoichiometric A/F mixture.

In homogeneous operation, combustionto a great extent corresponds to the combus-tion for manifold injection.

64 Gasoline direct injection Operating modes

Figure 1A Homogeneous

operation with λ = 1,this operating modeis possible in alloperating ranges

B Lean-burn or homo-geneous operation,λ = 1 with EGR; this operating mode is possible in area Cand area D

C Stratified-chargeoperation with EGR

Operating modeswith dual injection:

C Stratified-charge/cat-heating mode,same area asstratified-chargeoperation with EGR

D Homogeneous and stratified-charge

E Homogeneous/anti-knock

Engine speed n

Torq

ue M

B

C

AE

DD

Acce

lera

tion

Road-resis

tance

curves

Operating-mode characteristic curves for gasolinedirect injection

1

æS

MK

1773

E

Robert Bosch GmbH

Homogeneous and lean-burn modeIn the transitional range between stratified-charge and homogeneous mode, the enginecan be run with a homogeneous lean A/Fmixture (λ>1). Since the pumping losses arelower due to “non-throttling”, in the homo-geneous and lean-burn mode, fuel con-sumption is lower than in the homogeneousmode with λ ≤ 1.

Homogeneous and stratified-chargemodeThe complete combustion chamber is filledwith a homogeneous lean A/F mixture. Thismixture is generated by injecting a smallquantity of fuel during the induction stroke.

Fuel is injected a second time (dual injec-tion) during the compression stroke. Thisleads to a richer zone forming in the area ofthe spark plug. This stratified charge is easilyignitable and then ignites the rest of the ho-mogeneous mixture in the remainder of thecombustion chamber.

The homogeneous and stratified-chargemode is activated for a number of cyclesduring the transition between stratified-charge and homogeneous mode. This en-ables the engine management system to bet-ter adjust the torque during the transition.Due to the conversion to energy of the verylean A/F mixture λ > 2, the NOx emissionsare also reduced.

The distribution factor between each in-jection is approx. 75 %. That is, 75 % of thefuel is injected in the first injection which isresponsible for the homogeneous basic A/Fmixture.

Steady-state operation using dual injec-tion at low engine speeds in the transitionalrange between stratified-charge and homo-geneous mode reduces the soot emissionscompared to stratified-charge operation, aswell as lowering fuel consumption com-pared to homogeneous operation.

Homogeneous/anti-knock modeIn this operating mode, since the chargestratification hinders knock, the use of dualinjection at WOT, together with ignition-an-gle shift in the retard direction as needed toavoid “knock”, can be dispensed with. At thesame time, the favorable ignition point alsoleads to higher torque.

Stratified-charge/cat-heatingAnother form of dual injection makes itpossible to quickly heat up the exhaust-gassystem, although this must be optimized before this solution can be applied. Here,therefore, in stratified-charge operation withhigh levels of excess air, injection takes placeonce in the compression stroke (similar tothe “stratified-charge mode”), and thenagain in the combustion (power) cyclewhereby the fuel injected here combustsvery late and thus heats up the exhaust sideand the exhaust system to a very high temperature.

A further important application is for heat-ing up of the NOx catalytic converter to tem-peratures in excess of 650 °C as needed toinitiate the desulphurization of the catalyticconverter. Here, it is imperative that dual injection is used since with conventionalheating methods the high temperaturewhich is required here cannot always bereached in all operating modes.

Gasoline direct injection Operating modes 65

Robert Bosch GmbH

The Otto-cycle engine is a gasoline internal-combustion engine with spark ignition (SI).An ignition spark is used to ignite the com-pressed A/F mixture in the combustionchamber and thus initiate its combustion.The ignition spark is in the form of a sparkdischarge between the spark-plug electrodeswhich extend into the combustion chamber.The ignition system is not only responsiblefor generating the high voltage needed forthis spark discharge, but also for triggeringthe ignition spark at exactly the right in-stant in time.

Survey

The most important characteristic values forthe ignition of the A/F mixture are:

� Ignition angle, and� Ignition energy.

The ignition angle is referred to crankshaftTDC. It defines the ignition point and there-fore the inflammation or burning of the A/Fmixture. It also has considerable influenceon the gasoline engine’s output power andits exhaust-gas emissions.

A given voltage across the spark-plug elec-trodes, the ignition voltage, must be ex-ceeded in order to generate the ignitionspark in the combustion chamber. Depend-ing on the engine’s operating point and thecondition of the spark plug, voltages as highas 30,000 V (turbocharged engine) areneeded. Following the spark discharge, thespark energy is transferred to the A/F mix-ture and initiates the combustion process.

Inductive (coil) ignition systems have cometo the forefront in passenger-car applica-tions. In this form of ignition, the ignitionenergy is temporarily stored in the ignitioncoil’s magnetic field and after having beentransformed to a high enough voltage it istransferred to the A/F mixture at the igni-tion point.

Ignition systems with capacitive high-powerenergy storage are available for use with rac-ing and high-performance engines. Thesestore the ignition energy in the magneticfield of a capacitor.

Ignition systems: Development

Since they first came onto the market, therehas been no letup in ignition-system devel-opment. This was, and is, the result of theever-increasing demands made for higherengine outputs and improved exhaust-gasemissions. Here, electronics is continuing toplay a more and more important role (Fig. 1).

66 Ignition: An overview Survey, ignition-systems development

Ignition: An overview

Inductive ignition systems

Mechanical Electronic

Switch ignition-coil current

Ignition-angle adjustment (timing)

High-voltage distribution

Conventional coil ignition (CI)

Transistorized ignition (TI)

Electronic ignition (EI)

Distributorless semiconductor ignition

αz

The development of the ignition system1

æU

MZ

0307

E

Robert Bosch GmbH

Conventional coil ignition (CI)(1934 ... 1986)Mechanical breaker points in the ignitiondistributor control the flow of currentthrough the ignition coil (charge coil and ig-nition). A mechanical (flyweight) advancemechanism and a vacuum unit define the ig-nition angle as a function of engine speedand load (mechanical ignition timing).

A mechanical rotor which rotates insidethe ignition distributor is responsible fordistributing the high voltage to the sparkplugs (rotating high-voltage distribution).

Transistorized ignition (TI)(1965 ... 1993)The mechanical breaker points were re-placed here by a non-wearing power transis-tor mounted in a transistorized trigger box.The transistor is triggered by an inductive orHall sensor. The use of a transistor forswitching means that the disadvantages dueto wear at the mechanical breaker points areavoided.

Electronic ignition (EI)(1983 ... 1998)Although high-voltage distribution is stillmechanical, the mechanical ignition timingis dispensed with. Engine speed and load aremeasured electronically and used as the in-put variables for an ignition map stored in asemiconductor memory. An ignition ECUwith microcontroller is needed for triggeringand control.

Distributorless semiconductor ignition (1983 ... 1998)With this ignition system, voltage distribu-tion is no longer mechanical, but is per-formed electronically by the ignition ECU(static voltage distribution). This means thatthe ignition system no longer contains anycomponents which are subject to wear.

As from 1998, all newly designed engineshave been equipped with an engine ECUwhich combines distributorless semiconduc-tor ignition and gasoline injection(Motronic, Fig. 2).

Ignition: An overview Ignition-systems development 67

Figure 21 Spark-plug

ignition coil2 Spark plug

1 2

Section through a 4-cylinder engine with gasoline direct injection and distributorless semiconductor ignition2

æU

MM

0561

Y

Robert Bosch GmbH

The gasoline engine’s (inductive) coil igni-tion system is responsible for the spark dis-charge at the spark-plug electrodes, and forthe provision of enough energy for a power-ful spark.

Survey

The ignition circuit of the coil ignition com-prises the following components:

� Ignition driver stage (Fig. 1, Pos. 1),� Ignition coil (2),� High-voltage distributor,� Spark plug (4), and� Connecting devices and interference

suppressors.

Modern ignition systems with static voltagedistribution are no longer equipped withhigh-voltage distributors.

Using a coil ignition system with static volt-age distribution and double-ended ignitioncoil as an example, Fig. 1 shows the principle design of the ignition circuit.

Ignition driver stageAssignmentIt is the job of the ignition driver stage toswitch the ignition-coil current.

Design and operating conceptThese driver stages are mostly in the form ofa 3-stage power transistor. The “primary-current limitation” and “primary-voltagelimitation” functions are integrated mono-lithically in the driver stage and serve to pro-tect the ignition components against over-load.

During operation, driver stage and igni-tion coil both heat up. In order not to exceedthe permissible operating temperatures, it isnecessary that appropriate measures aretaken to ensure that the power loss is reliablydissipated to the surroundings even whenoutside temperatures are high. Primary-cur-rent limitation is only needed for limitationof current in case of fault (e.g. short circuit).

Internal and external ignition-driver stagesare available. The former are integrated onthe engine ECU printed-circuit board, andthe latter are located in their own housingoutside the engine ECU. Due to the costs in-volved, external driver stages are no longerused on new developments.

In addition, it is becoming increasinglycommon for the driver stages to be incorpo-rated in the ignition coil.

68 Coil ignition Survey, ignition driver stage

Coil ignition

Figure 11 Ignition driver stage2 Ignition coil3 EFU diode (EFU =

Switch-on sparksuppression)

4 Spark plug

15,1,4,4a terminal designationsTriggering for the ignition driver stage

14

3

2

15 4

4a1

12V

Using a coil ignition system with static voltagedistribution and double-ended ignition coils as anexample, Fig. 1 shows the basic design of theignition circuit

1

æU

MZ

0308

Y

Robert Bosch GmbH

Ignition coilAssignmentThe ignition coil stores the required ignitionenergy and generates the high voltage for thespark flashover at the ignition point.

DesignToday’s state-of-the-art ignition coils arecomprised of two magnetically-coupledcopper windings (primary and secondarywindings), an iron core assembled fromsheet-metal laminations, and a plastic case.Depending upon design, the core can be ofeither the closed type (compact coil), or ofthe rod type (rod-type coil). The arrange-ment and location of the primary and sec-ondary windings depends upon the coil’sshape. In order to increase the insulation re-sistance, the secondary winding can be de-signed as a disc or chamber winding.

So as to ensure efficient insulation be-tween primary and secondary winding, andbetween the windings and the case, the caseis filled with epoxy resin. The design andconstruction of the ignition coil are adaptedto the application in question.

Operating conceptThe ignition coil functions according toFaraday’s Law. The energy stored in the pri-mary winding’s magnetic field is transferredto the secondary winding by magnetic in-duction. Depending upon the turns ratio,voltage and current are transferred from theprimary to the secondary winding (Fig. 2).

On the single-ended ignition coils for sys-tems with rotating high-voltage distribution,one of the primary-winding terminals isconnected to one of the secondary-windingterminals and then to Terminal 15 of thedriving switch (economy connection). Theother end of the secondary winding is con-nected to the ignition driver stage (Terminal1). The secondary winding’s other connec-tion goes to the ignition distributor (Termi-nal 4). On the double-ended and dual-sparkignition coils used on ignition systems with

static voltage distribution, primary and sec-ondary windings are not connected. On thedouble-ended ignition coil, one end of thesecondary winding (Term. 4a) is connectedto ground, while the other end is directlyconnected to the spark plug. On the dual-spark ignition coil, each secondary-windingterminal is connected to a spark plug.

High-voltage generationOn modern ignition systems, the engine-management ECU switches on the ignitiondriver stage for the calculated dwell period,during which the coil’s primary current in-creases to its desired value and in the processgenerates a magnetic field.

The magnitude of the primary current,together with the ignition coil’s primary inductance, are decisive for the energystored in this magnetic field.

Coil ignition Ignition coil 69

Figure 2For rotating high-voltagedistribution:a Single-ended

ignition coil

For static high-voltagedistributionb Double-ended

ignition coilc Dual-spark ignition

coil

A Primary windingB Secondary winding

a

A B

b c

15 15 154a

1 1 14 4b

4a

4

Ignition coils: Schematic representations2

æU

MZ

0257

-2Y

Robert Bosch GmbH

At the moment of ignition (ignition point)the ignition driver stage interrupts the cur-rent flow. The resulting change in magneticfield induces the secondary voltage in thecoil’s secondary winding. The maximumpossible secondary voltage is a function ofthe energy stored in the ignition coil, thewinding capacitance, the coil’s turns ratio,the secondary load (spark plug), and the pri-mary-voltage limitation of the ignition dri-ver stage.

The secondary voltage must in any case ex-ceed the voltage level required for theflashover between the spark-plug electrodes(required ignition voltage). There must beadequate spark energy available to ignite theA/F mixture even when follow-up sparks aregenerated. These occur when the ignitionspark is diverted by mixture turbulence and“breaks off” as a result.

When the primary current is switched on,an undesirable voltage of approx. 1...2 kV isinduced in the secondary winding (switch-on voltage). This is of opposite polarity tothe high voltage. Spark discharge at thespark plug (switch-on spark) must beavoided at all costs at this point.

On systems with rotating high-voltagedistribution, the switch-on spark is effec-tively suppressed by the upstream distribu-tor-rotor spark gap. In the case of static volt-age distribution with double-ended ignitioncoils, a diode (EFU diode, Fig. 2b) in thehigh-voltage circuit stops the switch-onspark. With static voltage distribution, whendual-spark ignition coils are used, theswitch-on spark is effectively suppressd bythe high flashover voltage required for theseries connection of two spark plugs. Addi-tional measures need not be taken.

When the primary current is switched off, a200...400 V self-induced voltage is generatedin the secondary winding.

High-voltage distributionAssignmentAt the ignition point, the high voltage in-duced in the ignition coil must be availableacross the electrodes of the correct sparkplug. This is the responsibility of the high-voltage distribution.

Rotating high-voltage distributionIn this form of high-voltage distribution, thevoltage generated by a single ignition coil(Fig. 3a, Pos. 2) is mechanically distributedto the individual spark plugs (5) by an igni-tion distributor (3).

This form of distribution no longer has anysignificance for modern engine-manage-ment systems.

70 Coil ignition Ignition coil, high-voltage distribution

Figure 3a Rotating high-volt-

age distributionb Static high-voltage

distribution withdouble-ended ignition coils

1 Ignition lock2 Ignition coiI3 Ignition distributor4 Ignition cable5 Spark plug6 ECU7 Battery

7

7

1

6 5

5

2

6

1 2

3 4

Principle of high-voltage distribution3

æU

MZ

0309

Y

Robert Bosch GmbH

Static voltage distributionMechanical components are dispensed withon distributorless (electronic or static) high-voltage distribution (Fig. 3b). The ignitioncoils are connected directly to the sparkplugs and voltage distribution takes place atthe ignition-coil primary side. This permitswear-free and loss-free voltage distribution.There are two versions of this form of volt-age distribution.

Installations with double-ended ignitioncoils Each cylinder is allocated its own spark-plugignition coil and ignition driver stage. Theengine ECU triggers the driver stage in accordance with the firing sequence.

Since there are no distributor losses, theseignition coils can be designed to be verysmall. Preferably, they are mounted directlyabove the spark plug.

The static voltage distribution with dou-ble-ended ignition coils can be applied uni-versally irrespective of the number of enginecylinders. There are no limitations on the ignition-timing adjustment range, althoughthis system must also be synchronized to thecamshaft by means of a camshaft sensor.

Installations with dual-spark ignition coilsOne ignition driver stage and one coil are allocated to two cylinders. The ends of thesecondary winding are each connected to aspark plug in different cylinders. The cylin-ders have been chosen so that when onecylinder is in the compression stroke theother is in the exhaust stroke (applies onlyfor engines with an even number of cylin-ders). Spark discharge takes place at eachspark plug at the moment of ignition (igni-tion point). Care must be taken that thespark which takes place during the exhauststroke does not ignite residual gas or freshgas which has just been drawn in. Althoughthis precautionary measure leads to a limita-tion in the the ignition-timing adjustmentrange, it is not necessary to synchronize thesystem to the camshaft.

Spark plugs AssignmentThe spark plug generates a spark which ignites the A/F mixture in the combustionchamber.

Design and operating conceptThe spark plug (Fig. 4) is a ceramic-insu-lated, gastight high-voltage lead-throughinto the combustion chamber. It is providedwith a ground electrode (2) and a centerelectrode (1).

The type of spark is determined by the posi-tion of the ground electrode(s). If this is opposite to the center electrode one speaksof an air-gap spark plug (a). When theground electrode(s) is/are located to the sideof the center electrode, this results in a side-electrode air-gap spark plug (b) or in thesurface air-gap spark plug (c) or the purelysurface-gap spark plug (d).

Coil ignition Voltage distribution, spark plugs 71

Figure 41 Center electrode2 Ground electrodea Air spark gap with

front electrodeb Air spark gap with

side electrodec Surface air-gap

(air spark or surfacespark possible)

d Surface spark gapEA Spark gap

1

EA2

a

b

c

d

Spark plug (partial section) and spark gap4

æU

MZ

0129

-1Y

Robert Bosch GmbH

After interrupting the primary current at themoment of ignition (ignition point), thevoltage in the ignition coil’s secondarywinding increases very rapidly (approx.30 µs, Fig. 5) to the ignition voltage. As soonas the required ignition voltage is exceeded,the spark gap between center and groundelectrode becomes conductive. The capaci-tances in the secondary circuit which havecharged up to ignition voltage (spark plug,ignition cable, and ignition coil) dischargeabruptly in the form of a spark across theelectrodes. Within a typical spark durationof 1...2 ms, the energy stored in the ignitioncoil is converted in a glow discharge (sparktail). The residual energy in the ignition coilthen decays completely in a post-oscillationphase.

Spark-plug wearDuring normal engine operation, the spark-plug electrodes are subject to wear as a resultof the erosion stemming from the spark cur-rent and corrosion due to the hot gases inthe combustion chamber. This wear enlargesthe spark gap and the required ignition volt-age increases as a result. Independent of theoperating mode, up until the end of the pre-

scribed spark-plug replacement interval,there must always be adequate secondaryvoltage available from the ignition system toreliably provide for this ignition voltage.

Electrical connection and interference-suppressor devicesIgnition cableThe high voltage generated in the ignitioncoil must be delivered to the spark plug.Special, plastic-insulated, high-voltage-proofcables with special plugs for contacting thehigh-voltage components, are used with ig-nition coils which are not mounted directlyon the spark plug.

Since, for the ignition system, each high-voltage line represents a capacitive loadwhich reduces the available secondary volt-age, the ignition cables must be kept as shortas possible.

Interference suppressors, shieldingThe pulse-shaped discharge which occurs atevery spark flashover at spark plug or igni-tion distributor (in the case of rotating high-voltage distribution) is a source of interfer-ence. Interference suppression resistors inthe high-voltage circuit limit the dischargepeak current. In order to minimise the inter-ference radiation from the high-voltage cir-cuit, the suppression resistors should be in-stalled as close as possible to the interferencesource.

Normally, the interference resistors are in-tegrated in the spark-plug connectors, in theplugs at the other end of the ignition cableand, when high-voltage distribution is used,in the distributor rotor. Spark plugs are alsoavailable which feature an integral suppres-sion resistor. Increasing the secondary-cir-cuit resistance, though, leads to increasedenergy losses in the ignition circuit andtherefore to lower levels of spark energy atthe spark plug.

Interference radiation can be even furtherreduced by partially or completely screeningthe ignition system.

72 Coil ignition Spark plug, electrical connection and interference-suppressor devices

Figure 5K Spark headS Spark tailtF Spark duration

kV

15

10

5

0

0Time

1.0 2.0 3.0 ms

K

S

Vol

tage

Approx 30 s

tF

Voltage curve at the spark-plug electrodes5

æU

MZ

0044

E

Robert Bosch GmbH

Ignition voltageThis is the level of voltage across the spark-plug electrodes required to cause spark dis-charge between them. It depends upon anumber of factors:

� Density of the A/F mixture in the com-bustion chamber, and therefore also theignition point,

� Composition of the A/F mixture (excess-air factor, Lambda value),

� Flow velocity and turbulence,� Electrode geometry,� Electrode material,� Electrode gap.

Care must be taken that the ignition systemprovides the required ignition voltage irre-spective of operating conditions.

Ignition energy

The breaking current and the ignition-coilparameters define the energy stored by theignition coil and then made available as ig-nition energy in the ignition spark. The igni-tion energy has a decisive influence upon the

ignition of the mixture. Good A/F-mixtureignition is the prerequisite for high-perfor-mance engine operation coupled with lowlevels of toxic emissions. These requirementsplace high demands on the ignition system.

Energy balance of a single ignitionprocess

The energy stored in the ignition coil is re-leased as soon as the ignition spark is initi-ated. This energy is divided into two differ-ent sections.

Spark headThe energy E which is stored in the ignitioncircuit’s secondary-side capacity C, is re-leased abruptly at the ignition point, and in-creases as the square of the applied voltage U(E = 1/2 CU2). Fig. 6 therefore shows asquare-law curve.

Coil ignition Ignition voltage, ignition energy 73

Figure 6The energy values applyfor an imaginary ignitionsystem with an ignition-coil capacity of 35 pF, anexternal load of 25 pF,and a secondary induc-tance of 15 H.

mJ

Ignition voltage U

Ene

rgy

E

10

20

30

40

0 403530252015105 kV

Available energy

Spark head,capacitive discharge

Spark tail,inductive discharge

Energy balance of an ignition process without shunt, resistance and Zener losses6

æS

MZ

0310

E

Robert Bosch GmbH

Spark tailThe rest of the energy stored in the ignitioncoil (inductive share) is then released. Thisenergy is the difference between the total energy stored in the ignition coil, and theenergy released by capacitive discharge.

This means that the higher the required ig-nition voltage, the larger is the proportion oftotal energy in the spark head.

In certain cases, when the required igni-tion voltage is very high due for instance tobadly worn spark plugs, the energy stored inthe spark tail no longer suffices to com-pletely burn an already ignited A/F mixtureor, by means of follow-up sparks, re-ignite aflame that has been extinguished.

Further increases in the required voltagelead to the misfire limit being reached. Theavailable energy in the spark head no longersuffices to generate a spark discharge anddecays away as a damped oscillation (ignition misfire).

Shunt lossesFig. 6 on the previous page shows a simpli-fied representation of the existing condi-tions. The suppression resistors themselves,and the ohmic resistances in the ignition coiland ignition lines, cause losses which arethen not available as ignition energy.

Further losses result from shunt resis-tances which can be caused by contamina-tion at the high-voltage connections, as wellas by deposits and soot on the parts of thespark plug projecting into the combustionchamber.

The severity of the shunt losses dependsupon the required ignition voltage. Thehigher the voltage applied to the spark plug,the higher are the currents which are lostthrough shunt resistances.

Igniting the A/F mixtureUnder ideal conditions, provided that theA/F mixture is stationary, homogeneous andstoichiometric, for each individual ignitionprocess an energy of approx. 0.2 mJ is re-quired to ignite the mixture by means ofelectric spark. Under such conditions, richor lean mixtures need more than 3 mJ.

The energy that is actually required to ignitethe A/F mixture (the ignition energy) is onlya fraction of the total energy in the ignitionspark. On conventional ignition systems,when high break-down voltages are con-cerned, energies in excess of 15 mJ areneeded to generate the high-voltage sparkdischarge at the ignition point. Further en-ergy is required to compensate for losses,due for instance to contamination shunts atthe spark plugs, and in order to maintain thespark for a given period of time. These re-quirements amount to ignition energies ofat least 30...50 mJ, a figure which corre-sponds to an energy level of 60...120 mJstored in the ignition coil.

A/F-mixture turbulences such as occur inthe stratified-charge mode with gasoline di-rect injection, can divert the ignition sparkto such an extent that it extinguishes. Anumber of follow-up sparks are then neededto ignite the A/F mixture, and this energymust also be provided by the ignition coil.

The more air there is in a lean A/F mixture,the more difficult it is to ignite it. This factleads to a particularly high level of energybeing needed on the one hand to cover thehigher ignition-voltage requirements, andon the other to ensure that spark duration isas long as necessary.

If insufficient energy is available, the A/Fmixture does not ignite and this leads tocombustion misses.

74 Coil ignition Ignition energy

Robert Bosch GmbH

These facts mean that enough ignition en-ergy must be made available so that the A/Fmixture ignites reliably even under the mostadverse conditions. In such cases, igniting asmall A/F-mixture cloud in the vicinity ofthe spark plug can suffice to initiate ignitionand combustion of the rest of the A/F mix-ture in the cylinder.

Influences on the ignition characteristicEfficient mixture formation and ease of ac-cess to the ignition spark improve the igni-tion characteristic, as do an extended sparkduration, a long spark, and a wide electrodegap. Mixture turbulence can also be an ad-vantage provided enough energy is availablefor follow-up ignition sparks should thesebe needed. Turbulence supports rapidflame-front distribution in the combustionchamber, and with it the rapid and completecombustion of all the A/F mixture.

Spark-plug contamination is also of consid-erable importance. If the spark plugs arevery dirty, energy flows from the ignitioncoil and through the spark-plug shunt (de-posits) during the time in which the highvoltage is being built up. This reduces thehigh voltage, shortens the spark duration,and has a negative effect upon the exhaustgas. In extreme cases, this can lead to igni-tion misfire if the spark plugs are badly con-taminated or wet.

Ignition misfire leads to combustion misswhich increases both fuel consumption andexhaust-gas emissions. The catalytic con-verter can also be damaged.

Ignition point

About two milliseconds elapse between themoment the ignition spark is generated andcomplete combustion. These figures apply aslong as the A/F mixture composition re-mains unchanged. Therefore, along with in-creasing engine speed, ignition must alsotake place at an earlier and earlier point re-ferred to the crankshaft angle.Poor cylinder charge means that the A/Fmixture’s ignition characteristic deterioratesaccordingly. This leads to increased ignitionlag so that the ignition point has to be ad-vanced even further. For the best-possibletorque output, the ignition angle must bechosen so that main combustion, and with itthe peak pressure, takes place after Top DeadCenter (TDC), whereby care should betaken that the engine does not knock (Fig. 7).

In the stratified-charge mode (gasoline di-rect injection), the range for the variation ofthe ignition point is limited due to the endof injection and the time needed for A/F-mixture formation during the compresionstroke.

Coil ignition Ignition energy, ignition point 75

Figure 71 Ignition Za at the

right moment in time2 Ignition Zb too early

(combustion knock)3 Ignition Zc too late

bar

40

20

075° 50° 25° 0° -25° -50° -75°

BTDC ATDC

Ignition angle αZ

21

3

60

ZbZa ZcC

ombu

stio

n-ch

ambe

r pre

ssur

e

Pressure curve in the combustion chamber for different ignition angles (ignition points)

7

æU

MZ

0001

E

Robert Bosch GmbH

Emission-control legislation defines thelimits for the toxic agents generated duringthe combustion process in the spark-igni-tion engine. Catalytic treatment of the ex-haust gas is necessary in order to complywith these limits.

Overview

Before leaving the exhaust pipe, the exhaustgas flows through the catalytic converter in-stalled in the exhaust-gas tract (Fig. 1,Pos. 3). Inside the converter, special coatingsensure that the toxic agents in the exhaustgas are chemically converted to harmlesssubstances. Lambda oxygen sensors (2, 4)are used to measure the residual-oxygencontent in the exhaust gas. These measuredvalues are then applied in adjusting the A/Fmixture so that the catalytic converter canwork at maximum efficiency.

A number of different catalytic-converterconcepts were applied in the past years. Thethree-way catalytic converter represents thestate-of-the-art for engines with homoge-neous A/F mixture distribution and opera-tion at λ = 1. Engines which run with a leanA/F mixture also require a NOx accumula-tor-type catalytic converter.

Oxidation-type catalytic converterIn this type of catalytic converter, the hydro-carbons and the carbon monoxide in the ex-haust gas are converted by oxidation (burn-ing) into water vapor and carbon dioxide.The oxygen needed for the burning processis already present in the case of a lean A/Fmixture (λ > 1) or by blowing air into theexhaust-gas tract upstream of the converter.The oxidation converter cannot convert theoxides of nitrogen (NOx).

Oxidation-type catalytic converters werefirst introduced in 1975 in order to complywith the exhaust-gas legislation in force inthe USA at that time. Today, catalytic con-verters which operate exclusively with oxida-tion principles are used only very rarely.

76 Catalytic emissions control Overview, oxidation-type catalytic converter

Catalytic emissions control

Figure 11 Engine 2 Lambda oxygen

sensor upstream ofthe catalytic con-verter (two-step sen-sor or broad-bandsensor dependingupon system)

3 Three-way catalyticconverter

4 Two-step lambdaoxygen sensordownstreaam of thecatalytic converter(only on systemswith lambda two-sensor control)

1

2

3 4

Exhaust-gas tract with Lambda oxygen sensors and a three-way catalytic converter installed in the immediate vicinityof the engine

1

æU

MA

0029

Y

Robert Bosch GmbH

Three-way catalytic converterThe three-way catalytic converter is installedin the exhaust-emission control systems ofmanifold-injection engines and gasoline direct-injection engines.

Assignment Three toxic components are generated dur-ing the combustion of the A/F mixture: HC(hydrocarbons), CO (carbon monoxide),and oxides of nitrogen (NOx). It is the job ofthe three-way catalytic converter to convertthese into harmless components. The prod-ucts which result from this converion areH2O (water vapor), CO2 (carbon dioxide),and N2 (nitrogen).

Operating conceptThe toxic components are converted in twophases: Firstly, the carbon monoxide and thehydrocarbons are converted by oxidation(Fig. G, Equations 1 and 2). The oxygenneeded for the oxidation process is availablein the exhaust gas in the form of the residualoxygen resulting from incomplete combus-tion, or it is taken from the oxides of nitro-gen whereby these reduce as a result (Fig. G, Equations 3 and 4).

The concentration of the toxic substances inthe untreated exhaust gas is a function of theexcess-air factor λ (Fig. 2a). For carbonmonoxide and hydrocarbons (HC), the con-version level increases steadily along with in-creasing excess-air factor (Fig. 2b). At λ = 1,there is only a very low level of toxic compo-nents in the untreated exhaust gas. Withhigh excess-air factors (λ > 1), the concen-tration of these toxic components remains atthis low level.

Conversion of the oxides of nitrogen(NOx) is good in the rich range (λ < 1) . Thelowest levels of NOx are present during stoi-chiometric operation (λ = 1). Even a smallincrease in the exhaust-gas oxygen contentas caused by operation at λ > 1 impedes thenitrogen reduction and causes a sharp in-crease in its concentration.

In order to maintain the three-way catalyticconverter’s conversion level for all threetoxic substances at as high a level as possible,these must be present in a chemical balancein the exhaust gas. This means that the A/Fmixture composition must have a stoichio-metric ratio of λ = 1, so that the “window”for the A/F mixture ratio l is necessarily veryrestricted. A/F mixture formation must becontrolled by a Lambda closed-loop controlcircuit.

Catalytic emissions control Three-way catalytic converter 77

Figure 2a Before catalytic

aftertreatment (untreated exhaust gas)

b After catalytic after-treatment

c Voltage characteris-tic of the two-stepLambda sensor

Lambda-Regelbereich(Katalysatorfenster)

Excess-air factor λRich Lean

0.975 1.0

NOX

NOX

CO

CO

Lambda control range (catalytic-converter window)

a

b

c

HC

HC

1.025 1.05

Uλ

Toxic components in the exhaust gas2

æU

MK

0876

-3E

Reaction equations in the three-way catalyticconverter

G

(1) 2 CO + O2 ➞ 2 CO2

(2) 2 C2H6 + 7 O2 ➞ 4 CO2 + 6 H2O

(3) 2 NO + 2 CO ➞ N2 + 2 CO2

(4) 2 NO2 + 2 CO ➞ N2 + 2 CO2 + O2

Robert Bosch GmbH

Design and constructionThe catalytic converter (Fig. 3) comprises asteel casing (6), a substrate (5), and the ac-tive catalytic noble-metal coating (4).

SubstratesTwo substrate systems have come to theforefront

Ceramic monolithsThese ceramic monoliths are ceramic bodiescontaining thousands of narrow passagesthrough which the exhaust gas flows. Theceramic is a high-temperature-resistantmagnesium-aluminum silicate. The mono-lith, which is highly sensitive to mechanicaltension, is fastened inside a sheet-steel hous-ing by means of mineral swell matting (2)which expands the first time it is heated upand firmly fixes the monolith in position. Atthe same time the matting also ensures a100 % gas seal.

Ceramic monoliths are at present themost commonly used catalyst substrates.

Metallic monolithsThe metallic monolith (metal catalytic con-verter) is an alternative to the ceramicmonolith. It is made of finely corrugated,0.05 mm thin metal foil which is wound andsoldered in a high-temperature process.Thanks to its thin walls, more passages canbe accomodated inside the same area, whichmeans less resistance to exhaust-gas flow, afact which is important in the case of high-performance engines.

CoatingThe ceramic and metallic monoliths requirean aluminum oxide (Al2O3) substrate coat-ing, the so-called “Washcoat” (4). This coat-ing serves to increase the converter’s effec-tive surface area by a factor of around 7000.On the oxidation catalytic converter, the ef-fective catalytic coating applied to the sub-strate contains the noble metals platinumand/or palladium. On the three-way con-verter, rhodium is also applied. Platinumand palladium accelerate the oxidation ofthe hydrocarbons (HC) and of the carbonmonoxide. Rhodium accelerates the reduc-tion of the oxides of nitrogen (NOx).

Depending upon the engine’s displace-ment, a catalytic converter contains about1...3 g of noble metal.

78 Catalytic emissions control Three-way catalytic converter

Figure 31 Lambda oxygen

sensor2 Swell matting 3 Thermally insulated

double shell4 Washcoat (Al2O3

substrate coating)with noble-metalcoating

5 Substrate (monolith)6 Housing

4

5

6

1 2 3

HC + CO + NO2

Three-way catalytic converter with Lambda oxygen sensor3

æU

MA

0027

-1Y

Robert Bosch GmbH

Operating conditionsOperating temperatureThe catalytic converter’s temperature plays adecisive role in emission-control efficiency.Considering a three-way catalytic converter,no worthwhile conversion of toxic sub-stances takes place until temperature exceeds300 °C. Operation within a temperaturerange of 400...800 °C is ideal with regard tohigh conversion levels and a long service life.

At temperatures between 800...1000 °C,thermal aging is accelerated due to the sin-tering of the noble metals and of the Al2O3

substrate layer, and this leads to a reductionof the effective surface. The time spent at800...1000 °C is of vital importance, andabove 1000 °C thermal aging increases dras-tically and leads to the catalytic converterbecoming practically 100 % ineffective.

Engine malfunction (ignition misfire) cancause the temperature inside the catalyticconverter to exceed 1400 °C. Since such tem-peratures melt the substrate and completelydestroy the catalyst, it is imperative that theignition system is highly reliable and main-tenance-free. Modern engine-managementsystems are able to detect ignition and com-bustion miss, and in such cases interrupt thefuel injection to the cylinder concerned sothat unburned A/F mixture cannot enter theexhaust-gas tract.

Unleaded fuelAnother prerequisite for long-term opera-tion is the use of unleaded fuel. Otherwise,lead compounds are deposited in the poresof the active surface and reduce their num-ber. Residues from the engine oil can also“poison” the catalyst and damage it so farthat it becomes ineffective.

Installation pointStrict emissions-control legislation demandsspecial concepts for heating the catalyticconverter when the engine is started. Thecatalytic converter’s installation point is de-termined by such concepts (for instance,secondary-air injection, shift of the timing

in the “retard” direction). The three-way cat-alytic converter’s sensitivity regarding oper-ating temperature limits the choice of instal-lation point. The temperature conditionsneeded for a high conversion level make itabsolutely imperative that the three-wayconverter is installed close to the engine.

In the case of the three-way catalytic con-verter, a configuration featuring a “pre-cat”near the engine followed by a second (main)underfloor catalytic converter has come tothe forefront. Catalytic converters near theengine demand that their coating techniquesbe optimized to provide for high-tempera-ture stability. Underfloor converters on theother hand, require optimisation in the so-called “low light-off” direction (low start-uptemperature) and good NOx conversioncharacteristics.

An alternative is available with just one“overall” catalytic converter which is theninstalled close to the engine.

EffectivenessFor a spark-ignition engine with homoge-neous mixture distribution operating at λ = 1, catalytic treatment of the exhaust gasusing a three-way catalytic converter is atpresent the most effective emission-controlmethod. Included in this system is theLambda closed-loop control which monitorsthe composition of the A/F mixture. Usingthe three-way catalytic converter, the pollu-tant emissions of carbon monoxide, hydro-carbons, and oxides of nitrogen can be prac-tically eliminated provided the engine oper-ates with homogeneous A/F-mixturedistribution and at stoichiometric A/F ratio.Notwithstanding the fact that it is not alwayspossible to comply fully with these operatingrequirements, one can still presume an aver-age pollutants reduction of more than 98 %.

Catalytic emissions control Three-way catalytic converter 79

Robert Bosch GmbH

NOx accumulator-type catalytic converterAssignmentDuring lean-burn operation, it is impossiblefor the three-way catalytic converter to com-pletely convert all the oxides of nitrogen(NOx) which have been generated duringcombustion. In such cases namely, the oxy-gen that is needed for the oxidation of thecarbon monoxide and of the hydrocarbonsis not split off from the oxides of nitrogenbut instead is taken from the high level ofresidual oxygen in the exhaust gas. The NOx

accumulator catalytic converter reduces theoxides of nitrogen in a different manner.

Design and special coatingThe NOx accumulator-type catalytic con-verter is similar in design to the conven-tional three-way converter. In addition tothe platinum and rhodium coatings, theNOx converter is provided with special addi-tives which are capable of accumulating ox-ides of nitrogen. Typical accumulator ma-terials are the oxides of potassium, calcium,strontium, zirconium, lanthanum, and bar-ium.The coating for NOx accumulation and forthe 3-way catalytic converter can be appliedon a common substrate.

Operating conceptAt λ = 1, due to the noble-metal coating theNOx converter operates the same as a three-way converter. In lean exhaust gases thoughit also converts the non-reduced oxides ofnitrogen. This conversion is not a continu-ous process as it is with the hydrocarbonsand the carbon monoxide, but instead takesplace in three distinct phases:

1. NOx accumulation (storage),2. NOx release, and3. Conversion.

NOx accumulation (storage)On the surface of the platinum coating, theoxides of nitrogen (NOx) are oxidized cat-alytically to form nitrogen dioxide (NO2).The NO2 then reacts with the special oxideson the catalyst surface and with oxygen (O2)to form nitrates. For instance, NO2 com-bines chemically with barium oxide (BaO)to form barium nitrate (NO3)2 (Fig. G,Equation 1). This enables the NOx converterto accumulate the oxides of nitrogen whichhave been generated during engine opera-tion with excess air.

There are two methods in use to determinewhen the NOx converter is full and the accu-mulation phase has finished:

� Taking the catalyst temperature into ac-count (Fig. 1, Pos. 4), the model-based method calculates the quantity of stored NOx.

� An NOx sensor (6) downstream of theNOx converter continually measures theNOx concentration in the exhaust gas.

NOx removal and conversionThe more NOx that is stored, the less theability to chemically bind further nitrogensof oxide. This means that regeneration musttake place as soon as a given level is ex-ceeded, in other words the accumulated ox-ides of nitrogen must be released and con-verted. To this end, the engine is run brieflyin the rich homogeneous mode (λ < 0.8).The processes for releasing the NOx andconverting it to nitrogen and carbon dioxidetake place separately from each other. H2,HC, and CO are used as reducing agents. Re-duction is slowest with HC and most rapidwith H2. NOx release takes place as follows,whereby the following description applieswith carbon monoxide (CO) as the reducingagent: The carbon monoxide reduces the ni-trate (e.g. barium nitrate Ba(NO3)2 to an ox-ide (e.g. barium oxide BaO). This leads tothe generation of carbon dioxide (CO2) andnitrogen monoxide (NO) (Fig. G, Equa-tion 2).

80 Catalytic emissions control NOx accumulator-type catalytic converter

Robert Bosch GmbH

Subsequently, using the carbon monoxide(CO), the rhodium coating reduces the NOx

to nitrogen and carbon dioxide (CO2) (Fig. G, Equation 3).

There are two different methods for deter-mining the end of the NOx-release phase:

� The model-based method calculates thequantity of NOx still held by the con-verter.

� A Lambda oxygen sensor (Fig. 1, Pos. 6)downstream of the converter measuresthe exhaust-gas oxygen concentration andoutputs a voltage jump from “lean” to“rich” when conversion has finished.

Operating temperature and installationpointThe NOx converter’s ability to accumulate/store NOx is highly dependent upon temper-ature. Accumulation reaches its maximum

between 300 and 400 °C, which means thatthe favorable operating-temperature range ismuch lower than that of the three-way cat-alytic converter. For catalytic emissions con-trol, therefore, two separate catalytic con-verters must be installed - a three-way pre-cat near the engine (Fig. 1, Pos. 3), and anNOx accumulator-type main converter (5)remote from the engine (underfloor cat).

Sulphur in the NOx accumulator-typecatalytic converterThe sulphur in gasoline presents the accu-mulator-type catalytic converter with aproblem. The sulphur contained in the ex-haust gas reacts with the barium oxide (ac-cumulator material) to form barium sul-phate. The result is that, over time, theamount of accumulator material availablefor NOx accumulation diminishes. Bariumsulphate is extremely resistant to high tem-peratures, and for this reason is only de-graded to a slight degree during NOx regen-eration. When sulphurized gasoline is usedtherefore, desulphurization must be carriedat regular intervals. Here, selective measuresare applied to heat the converter to between600 and 650 °C. For instance, the engine canbe run in the “stratified-charge/cat-heatingmode”. Rich (λ = 0.95) and lean (λ = 1.05)

Catalytic emissions control NOx accumulator-type catalytic converter 81

Figure 11 Engine with EGR

system2 Lambda oxygen sen-

sor upstream of thecatalytic converter

3 Three-way catalyticconverter (pre-cat)

4 Temperature sensor5 NOx accumulator-

type catalytic con-verter (main cat)

6 Two-step Lambdaoxygen sensor, op-tionally available withintegral NOx sensor

Reaction equations for the NOx accumulation phase (1), removal phase (2), and conversion phase (3)

G

(1) 2 BaO + 4 NO2 + O2 ➞ 2 Ba(NO3)2

(2) Ba(NO3)2 + 3 CO ➞ 3 CO2 + BaO + 2 NO

(3) 2 NO + 2 CO ➞ N2 + 2 CO2

1

2

3 4

5 6

Exhaust-gas system with three-way catalyic converter as pre-cat, and downstream NOX accumulator-type converterand Lambda oxygen sensors

1

æU

MA

0030

Y

Robert Bosch GmbH

exhaust gases are then passed through thecat one after the other. The barium sulphatereduces to barium oxide as a result.

Lambda control loop

Assignment For systems which operate with only a singlethree-way catalytic converter, the pollutantsmust be in a state of chemical balance in or-der that the conversion level for all threepollutant constituents is as high as possible.This necessitates a stoichiometric A/F-mix-ture composition with λ = 1.0, which meansthat the “window” in which the A/F ratiomust be located is very narrow. The only so-lution is to apply closed-loop control to theadjustment of the A/F mixture ratio. Open-loop control of fuel metering is not accurateenough.

Direct-injection gasoline engines are runwith A/F mixtures which deviate from stoi-chiometric. Closed-loop control can also beused on these systems.

Design and constructionA Lambda oxygen sensor (Fig. 1, Pos. 3a) islocated upstream of the pre-cat (4). The sen-

sor signal USa is inputted to the engine ECU(7). In order to do so, either a two-stepLambda sensor (two-step control) or abroad-band Lambda sensor (continuous-ac-tion Lambda control) must be used. A fur-ther Lamda oxygen sensor (3b) can be situ-ated downstream of the main catalytic con-verter (5). This is always a two-step sensor,and it delivers the sensor signal USb. Thisform of control is known as two-sensor con-trol.

Operating conceptUsing the Lambda control loop, deviationsfrom a specific A/F-ratio can be detectedand corrected. The control principle is basedon the measurement of the residual oxygenin the exhaust gas. This is a measure for thecomposition of the A/F mixture supplied tothe engine (2).

Two-step control The sensor voltage USa generated by the two-step Lambda oxygen sensor upstream of thepre-cat (4) is high in the rich range (λ < 1)and low in the lean range (λ > 1). Since thesensor voltage jumps abruptly at λ = 1, thetwo-step Lambda oxygen ensor can only dif-ferentiate between rich and lean A/F mixtures.

82 Catalytic emissions control Lambda control loop

Figure 11 Air-mass meter2 Engine3a Lambda oxygen

sensor upstream ofthe pre-cat (two-stepLambda sensor, orbroad-band Lambdasensor)

3b Two-step Lambdasensor downstreamof the main catalyticconverter (only if required; ongasoline directinjection with integralNOx sensor)

4 Pre-cat (three-waycatalytic converter)

5 Main cat (On mani-fold injection: three-way converter; ongasoline direct injec-tion: NOx accumula-tor-type converter)

6 Injectors7 Engine ECU8 Input signals

US Sensor voltageUV Injector-triggering

voltageVE Injected fuel quantity

Air Exhaust gas

FuelVE

3a 3b

UV USa

1 2 4 5

6

7

8

USb

Functional diagram of the Lambda closed-loop control1

æU

MK

1642

-1E

Robert Bosch GmbH

The sensor output signal is converted to abinary signal in the engine ECU and used asthe input signal for the Lambda closed-loopcontrol as implemented using software. TheLambda control has a direct influence on theA/F mixture formation and sets the correctA/F ratio by adapting the injected fuel quan-tity. The manipulated variable comprises astep change and a ramp, and its control di-rection changes with each jump of the sen-sor voltage. In other words, a jump of themanipulated variable causes the A/F mixtureto change. This change is first of all veryabrupt, and then it follows a ramp. With ahigh sensor voltage (“rich” A/F mixture), themanipulated variable adjusts in the “lean”direction, and for a low sensor voltage(“lean” A/F mixture) in the “rich” direction.This so-called two-step control enables A/Fmixture to be closed-loop controlled to val-ues around λ = 1.

Shaping the manipulated variable’s char-acteristic curve asymmetrically compensatesfor the Lambda sensor’s typical false signalcaused by variations in A/F mixture forma-tion (rich/lean shift).

Continuous-action Lambda controlThe broad-band Lambda sensor outputs acontinuous voltage signal USa. This meansthat not only the Lambda area (rich or lean)can be measured, but also the deviationfrom λ = 1 so that the Lambda control canreact more quickly to an A/F mixture devia-tion. This leads to better control behaviourwith highly improved dynamic response.

The broad-band Lambda oxygen sensor canmeasure A/F mixtures which deviate fromλ = 1. This means that (in contrast to thetwo-step control), such A/F mixtures canalso be controlled. The control range coversλ = 0.7...3.0 so that continuous Lambdacontrol is suitable for the “rich” and “lean”operation of engines with gasoline direct in-jection.

Two-sensor controlWhen it is situated upstream of the pre-cat,the Lambda oxygen sensor (3a) is heavilystressed by high temperatures and untreatedexhaust gas, and this leads to limitations inaccuracy. On the other hand, locating thesensor downstream of the main catalyticconverter (3b) means that these influencesare considerably reduced.

The only problem here though is that a sin-gle downstream sensor would be far too“sluggish” due to the exhaust gases taking solong to reach it. The principle of two-sensorcontrol relies upon the upstream sensorcontrolling the “lean” and “rich” shift, whilethe downstream sensor is part of a “slow”corrective closed control loop responsiblefor additive changes.

Lambda closed-loop control of gasoline di-rect injectionThe NOx accumulator-type catalytic con-verter has two different functions. Duringlean-burn operation, NOx accumulation andCO oxidation must take place. In addition,at λ = 1, a stable three-way function isneeded which provides for a minimum levelof oxygen-accumulation.The Lambda sensor upstream of the cat-alytic converter monitors the stoichiometriccomposition of the A/F mixture.Together with the integrated NOx sensor, thetwo-step Lambda sensor downstream of theNOx accumulator converter not only takespart in the two-sensor control but also mon-itors the behaviour of the combination O2

and NOx accumulator (detection of the endof the NOx release phase).

Catalytic emissions control Lambda control loop 83

Robert Bosch GmbH

Catalytic-converter heatingIgnition timing towards “retard”In order to keep the pollutant concentrationin the exhaust gas down to a minimum, it isnecessary that the catalytic converter reachesits operating temperature as soon as possi-ble. One method is to adjust the ignitiontiming towards “retard”.

This step lowers the engine efficiency, andin doing so leads to hotter exhaust gaseswhich then heat-up the converter.

Secondary-air injectionThe unburnt components of the A/F mix-ture still present in the exhaust gas are burntin the thermal afterburning process. With“lean” A/F mixtures, the oxygen required forthis afterburning process is available in theexhaust gas in the form of residual oxygen.With “rich” A/F mixtures, as often needed

for an engine which has not yet reached op-erating temperature, extra air (secondaryair) is injected into the exhaust-gas passageto speed-up the catalytic-converter heating.

On the one hand, this exothermic reactionreduces the hydrocarbons and the carbonmonoxide. On the other, afterburning alsoheats up the catalytic converter so that itquickly reaches its operating temperature.During the warm-up phase, this processconsiderably increases the conversion rate sothat the catalytic converter is quickly readyfor operation. Fig. 1 shows the curves of thehydrocarbon and carbon monoxide emis-sions in the first seconds of an emissionstest, with and without secondary-air injec-tion.

In line with present state-of-the-art, electricsecondary-air pumps are used for sec-ondary-air injection.

Post injection (POI)On gasoline direct-injection engines, an-other method can be used for quickly bring-ing the catalytic converter up to tempera-ture. In the “stratified-charge/cat-heating”operating mode, during stratified-chargeoperation with high levels of excess air a sec-ond injection of fuel takes place during theengine’s power cycle. This fuel is combustedlate and causes considerable heat-up of theengine’s exhaust side and of the exhaustmanifold. This means, that in those cases inwhich conventional measures (adjust igni-tion timing in the”retard” direction) do notsuffice for complying with the stipulated ex-haust-gas limits, the secondary-air pumpused for manifold injection can be dis-pensed with.

84 Catalytic emissions control Catalytic-converter heating

Figure 11 Without secondary-

air injection2 With secondary-air

injectionn Vehicle speed

1000

2000

3000

0

km/h

1

1

2

2

40

Time

80 120 s0

CO

em

issi

ons

υ

100

200

300

ppm

ppm

0

50

0

HC

em

issi

ons

Influence of secondary-air injection on CO and HCemissions

1

æU

MK

1711

-1E

Robert Bosch GmbH

Technical terms

AA/F mixture, 6A/F ratio, 15A/F-mixture cloud, 63A/F-mixture distribution, 6Air bypass actuator, 20Air charge, 12Air-mass meter, 49Auto-ignition, 19

BBoost pressure, 29Bottom Dead Center (BDC), 4Broad-band Lambda oxygen sensor,

83

CCamshaft changeover, 23Camshaft phase adjustment, 22Canister-purge valve, 41Carbon canister, 41Carbon dioxide, 77Carbon monoxide, 77Catalytic converters, 76Catalytic emissions control, 76Center electrode (spark plug), 71Centrifugal turbo-compressor, 29Combustion knock, 19Combustion process, 62Common Rail, 55Compression ratio, 6Compression stroke, 4Compressor, 29Continuous-action Lambda control, 83Conventional coil ignition (CI), 67Conversion of toxic components, 77Cylinder charge, 12Cylinder (engine), 4Cylinder-individual fuel injection (CIFI),

53

DDelivery-quantity control valve, 56Displacement-type compressor, 27Distributorless semiconductor ignition,

65Down-sizing, 33Dual injection, 65Dual spray, 52Dual-spark ignition coil, 69Dwell angle, 19Dynamic supercharging, 26

EEfficiency, 8EGR valve, 25Electric fuel pump, 36, 42Electromagnetic fuel injectors, 50Electronic ignition, 67Electronic throttle control (EGAS), 21Emission-control legislation, 76Evaporative-emissions control system,

41Excess-air factor (Lambda), 6Exhaust stroke, 5Exhaust valve, 5Exhaust-gas recirculation (EGR), 25Exhaust-gas turbine, 30Exhaust-gas turbocharging, 30External EGR, 25Externally supplied ignition, 66

FFollow-up spark, 74Four-stroke principle, 4Fresh gas, 12Frictional losses, 9Fuel consumption, 16, 25Fuel filter, 36, 44Fuel lines, 46Fuel rail, 37, 45Fuel supply, 36Fuel tank, 36, 46Fuel-pressure damper, 46Fuel-pressure regulator, 36, 45Fuel-supply system, 37

GGas-exchange valves, 4Gasoline direct injection, 54Ground electrode (spark plug), 71Group fuel injection, 53

HHigh-pressure injectors, 60High-pressure pumps, 56High-voltage distribution, 70High-voltage generation, 69Homogeneous mode, 64Homogeneous/anti-knock mode, 65Homogeneous and lean-burn mode,

65Homogeneous and stratified-chargemode, 65Hydrocarbons (HC), 77

IIgnition angle, 18Ignition cable, 72Ignition coil, 69Ignition distributor, 70Ignition driver stage, 68Ignition energy, 73Ignition map, 18Ignition point, 18, 70, 75Ignition timing, 84Ignition voltage, 72, 73Induction stroke, 4Inductive (coil) ignition system, 68Inert gas, 13Infinitely-variable valve timing, 24Injection valves, 49Injection-orifice plate, 50Inner-gear pump, 42Intake manifold, 26Intake valve, 4, 49In-tank unit, 40Intercooling, 33Interference-suppression resistor, 72Internal EGR, 14, 23

KKnock control, 19

LLambda closed-loop control, 83Lambda control loop, 82Lambda oxygen sensor, 76, 82Lambda, 6Lean-burn limit, 15Low-pressure circuit, 39

MManifold chamber, 26Manifold fuel injection, 48Mechanical supercharging, 29Monoliths (catalytic converter), 78Multipoint fuel-injection systems, 34

NNitrogen, 77Noble-metal coating, 78Non-return valve, 36, 58NOx accumulator-type catalytic con-verter, 80NOx emissions, 25

Index of technical terms 85

Index of technical terms

Robert Bosch GmbH

OOperating modes, 64Output power, 7, 16Overrun fuel cutoff, 17Overrun, 17Oxidation, 76Oxides of nitrogen (NOx), 76Oxydation-type catalytic converter,

76

PPalladium, 78Pencil spray, 52Peripheral pump, 43Platinum, 78Positive-displacement pump, 42Post injection (POI), 84Power (combustion) stroke, 4Pre-cat, 79Pressure-control valve, 55, 58Presupply pump, 42Primary pressure, 56Primary winding (ignition coil), 69Primary-current limitation, 68Pumping losses, 9p-V diagram, 8

RRail, 55, 56Rail-pressure sensor, 59Ram-tube supercharging, 26Residual exhaust gas, 13Rhodium, 78Roller-cell pump, 42Rotary-screw supercharger, 29Rotating high-voltage distribution, 70

SSecondary winding (ignition coil), 69Secondary-air injection, 84Sequential fuel injection, 53Shunt losses, 74Side-channel pump, 43Simultaneous fuel injection, 53Single-cylinder high-pressure pump,

57Single-point injection (TBI), 35Single-spark ignition coil, 69Spark duration, 75Spark head, 73Spark length, 75Spark plug, 71Spark-plug ignition coil, 69Spark tail, 74Spiral-type supercharger, 29Spray formation, 52

Spray offset angle, 52Static voltage distribution, 71Stoichiometric ratio, 15, 82Stratified charge, 6Stratified-charge mode, 64Stratified-charge/cat-heating mode,

65Sulphur charge, 81Swirl air flow, 62System pressure, 37

TTapered spray, 52Thermal losses, 9Three-cylinder high-pressure pump, 57Three-way catalytic converter, 76, 77Throttle device, 21Throttle valve, 20Throttling losses, 22Top Dead Center (TDC), 4Torque, 7Trailing throttle, 17Transistorized ignition (TI), 67Tumble air flow, 62Tuned-intake-tube charging, 27Turbine pump, 43Turbo flat spot, 33Two-sensor control, 83Two-step control, 82Two-step Lambda oxygen sensor, 83Types of injection, 53

UUnderfloor catalytic converter, 79

VValve overlap, 14Valve timing, 5Variable valve timing, 22Variable intake-manifold geometry, 27Volumetric efficiency, 14VST supercharrger, 32VTG supercharger, 31

WWall film, 17Wall wetting, 17Washcoat, 78Wastegate supercharger, 31

86 Index of technical terms

Robert Bosch GmbH

Abbreviations

AATL: Exhaust-gas turbocharger

BBDC: Bottom Dead CenterBPS: Boost-Pressure Sensor

CCI: Coil IgnitionCIFI: Cylinder-Individual Fuel InjectionCO: Carbon monoxideCO2: Carbon dioxide

DDI: Direct InjectionDR: Pressure regulator

EECU: Electronic Control UnitEGAS: = ETCEGR: Exhaust-Gas RecirculationEI: Electronic IgnitionEKP: Electric fuel pumpETC: Electronic Throttle ControlEI: Electronic Ignition

HHC: HydrocarbonsHDEV: High-pressure injectorHDP: High-pressure pump

IIV: Intake Valve

LLML: Lean Misfire Limit

MMPI: Multi-Point InjectionMSV: Delivery-quantity control valve

NNOx: Oxides of nitrogen

PPOI: Post injection

RPP: Peripheral PumpRLFS: Returnless Fuel SystemROV: Rotating high-voltage

distributionRUV: Static voltage distributionRZP: Roller-cell pump

SSEFI: Sequential Fuel InjectionSI: Spark IgnitionSRE: Manifold fuel injection

TTBI: Throttle-Body InjectionTDC: Top Dead CenterTI: Transistorized Ignition

VVST: Variable Sleeve TurbineVTG: Variable Turbine GeometryVZ: Distributorless ignition

ZZP: Inner-gear pump

Index of technical terms Abbreviations 87

Robert Bosch GmbH