SAE TECHNICAL

PAPER SERIES 1999-01-0829

Dynamic Behaviour of a High Speed Direct

Injection Diesel Engine
R. S. Wijetunge, C. J. Brace, J. G. Hawley and N. D. Vaughan
University of Bath

R. W. Horrocks and G. L. Bird

Ford Motor Company

International Congress and Exposition

Detroit, Michigan
March 1-4, 1999

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760
The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sec-
tions 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as
copying for general distribution, for advertising or promotional purposes, for creating new collective works,
or for resale.

SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
orders to SAE Customer Sales and Satisfaction Department.

Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.

To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.

All SAE papers, standards, and selected

books are abstracted and indexed in the
Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.

ISSN 0148-7191
Copyright 1999 Society of Automotive Engineers, Inc.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 1999-01-0829

Dynamic Behaviour of a High Speed Direct Injection

Diesel Engine
R. S. Wijetunge, C. J. Brace, J. G. Hawley and N. D. Vaughan
University of Bath

R. W. Horrocks and G. L. Bird

Ford Motor Company

Copyright © 1999 Society of Automotive Engineers, Inc.

ABSTRACT yet a legislated emissions species, it is becoming

increasingly important owing to its connection with global
Many Diesel engine development programs concentrate warming. Limiting CO2 production can only be achieved
almost exclusively on steady state investigations to by improvements in fuel economy, or reducing the carbon
benchmark an engines performance. In reality, the inter- content of the fuel. The fact that fuel consumption figures
action of an engine’s sub-systems under transient evalu- must now be calculated on all new vehicles from 1 Janu-
ation is very different from that evident during steady ary 1997 [3] can be considered evidence of future legisla-
state evaluation. The transient operation of a complete tion of CO2 emissions. There is also a need for the
engine system is complex, and collecting test data is very passenger car Diesel engine to achieve gasoline engine
demanding, requiring sophisticated facilities for both con- levels of NVH and provide equivalent acceleration
trol and measurement. response on demand.
This paper highlights the essential characteristics of a Today’s Diesel engined cars demonstrate greater fuel
Diesel engine when undertaking testbed transient man- efficiency than ever before while simultaneously achiev-
ouevres. Results from simple transient sequences typical ing emissions levels some 50% lower than those of five
of on-road operation are presented. years ago [4]. This has been made possible by the appli-
cation of advanced technologies in the areas of EGR,
The tests demonstrate how transient behaviour of the VGT, multiple valves per cylinder, air flow events, oxida-
engine deviates greatly from the steady state optimum tion and de-NOx catalysts, and electronic fuel injection
settings used to control the engine. The operation of the systems (see Figure 1). These engine sub-systems have
EGR system and its interaction with other sub-systems, to be precisely controlled by software within the ECU.
in particular VGT, has a significant effect on emissions, These technological advances have led to a significant
fuel consumption and driveability, highlighting the need increase in the complexity, and hence cost, of the engine
for dynamic optimisation as an integrated system. and its control system. Consequently, the cost of a Diesel
engine to meet European Stage 3 emission legislation is
INTRODUCTION estimated to be 35-40% over its equivalent gasoline
counterpart. The emissions challenges facing the diesel
The high speed Diesel engine is now well established in passenger car for the proposed European Stage 4 are
the European passenger car market, with projected sales substantial, particularly in relation to PM for which a 50%
expected to encompass 25% of the market by 2005 [1]. reduction from Stage 3 levels is required [5,6]. Although
Direct injection , rather than indirect injection, will be the hardware improvements can still be made to engine
predominant future version [2]. The main attractive fea- design the greatest benefits will come from optimising the
ture of the Diesel engine, particularly in its turbocharged interaction of all the associated engine sub-systems.
form, is its high efficiency which can surpass 40%. This
enables Diesel engined vehicles to achieve lower fuel A current area of research at the University of Bath cen-
consumption figures than equivalently rated gasoline tres on the reduction of engine emissions and fuel con-
ones. For Diesel engines the development emphasis is sumption through the development of dynamic control
on reducing emissions of NOx and PM where these strategies. A dynamic testing programme has been per-
emissions are typically higher than those from equivalent formed in order to identify the characteristic behaviour
port injected gasoline engines equipped with 3-way cata- under transient operation of a HSDI Diesel engine and its
lysts, a summary of the current and impending require- sub-component interactions. Identification and evaluation
ments is given in Table 1. Although carbon dioxide is not of these characteristics are being used as a basis for pre-

1
dictive modelling, leading to new control strategies which TRANSIENT TESTING
are able to exploit the potential benefits of transient opti-
misation. A selection of results from this test program is Evaluation of the performance of automotive engines is
presented in this paper. based, to a considerable extent, on the dynamic charac-
teristics when changing state. To quantify these charac-
STEADY STATE TESTING AND ENGINE teristics, especially in relation to emissions, fuel
CALIBRATION consumption and driveability, transient testing and analy-
sis is required. The purpose being to study the overall
Steady state testing involves measuring and analysing response of the engine, the interaction of its sub-systems
engine performance at discrete points within the operat- and the engine management system to the many kinds of
ing range of the engine. The engine is allowed to settle at transient to which it will be subjected in service. Typical
each operating point before data are logged, large matri- examples of aspects of engine performance under tran-
ces can be constructed that map the steady state behav- sient conditions that require investigation include:
iour of the engine. These matrices form the basis of the
1. Response of the engine and control to changes, initi-
engine calibration, by analysing the maps optimum set-
ated either by the driver or as a result of changes in
tings for the engine sub-systems can be chosen from the
road load
map for any engine operating point to deliver maximum
torque for the best compromise between low emissions 2. Interactive behaviour of fuel system, turbocharger
and fuel consumption. The selected settings are stored and EGR under accelerating conditions
as maps within the ECU where they are used to com- 3. Short-term changes in exhaust emissions accompa-
mand the sub-systems via control algorithms, such as nying changes in operating regime
Proportional Integral Derivative (PID). The aim of the 4. Study of the response of different engine control
controller is to attempt to keep the settings as close to the strategies to the whole range of possible transient
mapped steady state values as possible. For example conditions
Figure 2 is a boost pressure map taken from the calibra-
tion for a prototype HSDI engine. At any operating point Transient testing is highly demanding and requires com-
(determined here by engine speed and fuelling quantity) plex and sophisticated facilities. For this reason, most
the ECU looks up a target boost pressure which it relays typical engine development programs concentrate almost
to VGT vane position, EGR valve position and other sub- exclusively on steady state performance optimisation. In
system parameters. reality, the interaction of engine sub-systems under
steady state evaluation is very different during transient
Steady state mapping provides a straightforward route to operation and this does create dynamic control problems.
engine calibration. However, it must be remembered that
very little real-life driving is steady state, even on motor- Further possibilities of reductions in fuel consumption
way journeys the seemingly constant engine conditions and emissions via steady state techniques are now very
are permeated with changes in speed and load alongside limited, it is necessary to look at the transient behaviour
many other variables. The outstanding characteristic of of the engine to identify areas where such reductions can
all emissions test procedures (drive cycles) is that they be made. For example, previous transient test work per-
involve transient operation of the engine, even though the formed at the University of Bath with a 1.8 litre HSDI with
object is to establish the cumulative mass of each nomi- Continuously Variable Transmission (CVT) indicated that
nated exhaust gas species emitted during the test. Such 50% of NOx emissions during the European Drive Cycle
tests do not indicate where in the cycle these emissions stem from periods of acceleration and deceleration. Fig-
are being produced, and as such, it is difficult to attempt ure 3 illustrates an acceleration transient performed on
any form of optimisation without a detailed knowledge of the 1.8 litre HSDI with CVT, where both engine speed
how the engines sub-components react under dynamic and load are allowed to increase during the fuelling step-
conditions. Although steady-state optimisation is carried up transient. The plot shows transient uHC (measured
out it can, at best, be considered as a baseline for tran- with a fast hydrocarbons detector) overlaid onto the
sient investigations, the main reasons for this being: steady state map for fuelling and engine speed vs. hydro-
carbon emissions . The start of the step is indicated by
1. Steady state measurements give a good indication of (1) and the finish by (2). As can be seen, the deviation
how an engine will perform in vehicle emission test from mapped values during the transient is significant,
cycles. and both the start and the finish of the transient ‘hover’
2. The facilities required to undertake testbed engine above the map surface. This can be attributed to differ-
transient performance investigations are expensive ences in engine temperature between the steady state
and the interpretation of transient test results can be and transients tests, which strongly affect uHC emis-
difficult. sions.
3. There are severe time and cost constraints on the
development process.

2
TRANSIENT TEST CELL the AFR ratio will decrease sharply (see Figure 6b Nomi-
nal AFR). To exacerbate the problem further there is the
The transient engine test facility is shown schematically effect of EGR, which is well known to have adverse con-
in Figure 4. The basic design was described by Dorey sequences on smoke, hydrocarbons, particulates and
and Guebeli [7] and it was used by Brace et al [8] for a brake specific fuel consumption. The air fuel ratio derived
research program with a continuously variable transmis- from exhaust gas analysis (Figure 6b Exhaust Gas Anal-
sion powertrain. The engine is a prototype 2.0 litre HSDI ysis AFR) takes into account EGR. The data have been
Diesel engine equipped with VGT, a distributor pump fuel corrected for transport lags in the exhaust system but not
injection system and a standard 5-speed manual gear- for the response of the analysers, hence the fast dynam-
box. Physical inputs such as accelerator position, clutch ics shown in the nominal plot are not reflected in the true
and gearchange controls can be operated manually or plot. As is to be expected, the initial and final values of
automatically via a PC equipped with the Ricardo Engine air fuel ratio are lower due to the presence of EGR before
Management Prototyping System (EMPS) environment, and after the transient.
used here for driver emulation purposes. This system is
also used to control the engine air and water tempera- EGR CONTROL DURING TRANSIENTS – Figure 6e
tures. The engine management system can be interro- shows the position of the EGR valve during the transient.
gated and altered via the Kleinknecht MCS interface and Initially the valve is open, then at the start of the transient
Gredi software which also permits data logging of engine it shuts rapidly in an attempt to maintain a good air fuel
management variables. General data acquisition is per- ratio. Immediately after the valve closure there will still be
formed by the HPVee DTVee software with 12 bit ADC. exhaust gas residuals present in the inlet manifold, par-
The hydraulic dynamometer can be operated in constant ticularly so if the exhaust manifold pressure rises sharply
speed, constant load or vehicle emulation modes. whilst the valve is closing (see Figure 6f Exhaust Mani-
fold Pressure) due to the flow restriction of the closed
The results presented here were generated from pedal
VGT. The peak in pressure causes a brief high EGR flow
steps, these are performed at constant speed and show
into the inlet manifold, which in turn results in very low
how the engine responds to step changes in fuelling. The
transient air-fuel ratios. Poor air-fuel ratios are associated
dynamometer is used to maintain constant engine speed
with poor combustion, reflected by the 90% opacity spike
(1500, 2500 and 3500 rev/min) and 2 torque levels are
on the smoke plot (Figure 6c). Large smoke spikes such
manually set using fuelling potentiometers. A toggle
as these are clearly undesirable, and are associated with
switch allows switching between the 2 settings to emulate
similar spikes in hydrocarbons and particulates [9]. Also
a step change.
of note is the brief opening of the EGR valve during the
transient, this is the result of control being derived directly
TRANSIENT RESULTS from steady state maps. At the start of the transient, the
boost error exceeds acceptable limits and this in itself
STEADY STATE VS. TRANSIENT BOOST PRESSURE – disables EGR, however the ECU still looks up the new
Figures 5a and 5b illustrate the deviation of inlet manifold EGR valve position demand based on the new fuelling
pressure from steady state map values during 3 pedal and speed conditions. As soon as the boost error falls to
steps (10-170Nm at 1500 rev/min, 10-170Nm at 2500 within acceptable limits, control is handed back to the
rev/min and 10-135Nm at 3500 rev/min). The boost pres- EGR map-based system, which in turn opens the valve
sure map is taken direct from the ECU calibration and is because in the steady state condition EGR is scheduled
used in the VGT vane position control loop with boost for the current engine speed and fuelling.
pressure feedback. During the pedal steps, the fuelling
quantity achieves the target value far faster than the air FUEL INJECTION QUANTITY LIMITING – As a mea-
charge, the transient does not follow the surface of the sure to prevent excessively low air-fuel ratios a maximum
map. In a turbo-charged engine undergoing such a tran- fuelling limit can be imposed, in the case of the control
sient, the first thing that happens is the quantity of fuel strategy shown the manifold pressure is used to calculate
injected increases rapidly, this in turn puts more energy the air charge density which, in conjunction with engine
into the exhaust gases which then accelerate the turbo- speed, determines the maximum fuelling. This method
charger. As the turbocharger accelerates, the flow and does not take account of residual EGR which explains
pressure to the inlet manifold increase which in turn per- why it permits the large smoke peaks. Figure 6a shows
mits more fuel to be injected. the fuelling limitations applied during the transient, with
Figures 6a - 6f detail the 2500 rev/min 10-170Nm tran- the dotted line representing the actual pedal demand and
sient and are described below. This type of transient is the solid line the command sent to the fuel pump. Fuel
particularly severe and simulates the onset of hard accel- limiting involves a fine compromise between emissions
eration, such as during a ‘foot-down’ overtaking manoeu- and driveability, limiting fuel until all EGR residuals have
vre. passed through the system would eliminate black smoke
but give very sluggish response to accelerator demand.
TRANSIENT AIR FUEL RATIO (AFR) – During the initial In order to maximise transient response, as high an air
stages of a large transient when the fuelling is rising rap- fuel ratio as possible should be maintained at all times to
idly, the lack of immediate boost (‘turbo-lag’) means that allow injection of as much fuel, hence generation of as

3
much torque, before air density limitations cut-in. VGT FURTHER WORK
permits higher air fuel ratios at lower engine speeds by
virtue of increased boost, and can therefore improve Now that areas for improvement have been identified and
driveability. the behaviour of various components characterised,
dynamic models of the engine and sub-systems are
VGT TRANSIENT BEHAVIOUR – The use of VGT can being developed for ultimate use in the design of control
minimise the boost pressure response time. The optimis- algorithms. Once validated on computer simulation,
ation of flow across the turbine for a wider range of these strategies will be evaluated on the transient test
engine conditions increases the acceleration of the rotor facility as well as in-vehicle using a chassis dynamome-
and hence gives a rapid boost pressure rise time. Never ter.
the less, the boost pressure dynamics are far slower than
the fuelling dynamics which is why there is such a large CONCLUSION
deviation from steady state map values (figure 5a and
5b). Figure 6d shows the VGT vane position during the A definition study of the dynamic behaviour of a High
transient, with 100% corresponding to the smallest flow Speed Direct Injection Diesel engine has been per-
area. As can be seen, the flow area is kept at a minimum formed. The interactions between EGR and other sub-
until boost pressure builds up sufficiently then the vanes components such as VGT and fuel injection have been
are opened to prevent overboosting. A dynamic problem shown to have significant effects on the dynamic behav-
associated with this particular type of VGT is illustrated in iour of the engine. The most important performance
this test, figure 6f shows large overshoots in boost pres- parameter is the AFR, which indicates the combustion
sure and airflow, this is due to hysteresis in the VGT vane quality and hence dictates the emissions behaviour, fuel
mechanism. The vacuum to the vane actuator (figure 6d) consumption and performance. Control of EGR needs to
reduces steadily but the response of the vanes them- optimised for dynamic operation as the current steady
selves is non-linear, which leads to a period of stagnation state approach allows too much degredation of the AFR
followed by a sudden flip open. The stagnation causes an by residual EGR at the onset of a transient. Also, better
overshoot in pressure and airflow, then the rapid opening co-ordination between the sub-systems to prevent events
leads to undershoot, settling only after several seconds. such as EGR valve opening momentarily during the tran-
VGT’s are typically controlled by PID controllers which sient, combined with more intelligent transient fuelling
react to errors occurring between demanded steady state strategies, is necessary in order to reduce transient
values and actual values. PID control is not ideal for non- smoke and give benefits to fuel consumption. Attention
linear systems as can be seen by the response of the must also be paid to the dynamics of particular devices,
VGT. such as the hysteresis of the VGT mechanism or the
inherent transport lags in the EGR system.
ABSENCE OF EGR – Figures 7a - 7f illustrate the same
transient with the EGR valve disabled, comparisons It has been shown that steady state characterisation of
between the 2 tests highlight the significant effect EGR an engine cannot be applied to transient behaviour. Fur-
has on the dynamic behaviour of the engine. The initial ther improvements in economy and emissions will require
AFR (Figure 7b) is much higher as expected, though the precise control of the engine during transients, this is not
final AFR is very similar with and without EGR, this is can achievable through steady state based techniques alone.
be explained by the small pressure difference between
final inlet and exhaust manifold pressures in the EGR-on REFERENCES
test. This pressure differential drives the EGR through
the valve, as it is small the EGR flow is small, in this situ- 1. Cichocki, R., “Legislative challenges for the HSDI Diesel
ation VGT can be used to increase the exhaust backpres- engine in the next decade” Engine & Environment AVL Pro-
sure hence EGR flow. ceedings Grazer Congress 4,5 Sept1997
2. [Belaire, R.C., Boggs, D.L., Bartunek, B., Durnholz, M.,
The smoke response (Figure 7c) shows a dramatic
Strobel, M., Ecker, H., “The development of Small Direct
improvement, from which it can be concluded that resid-
Injection Diesel Engines for Europe and the US”, Engine &
ual EGR was the most significant source of smoke in this
Environment AVL Proceedings Grazer Congress 4,5
particular test. VGT response (Figure 7d) is more con-
Sept1997
trolled, the VGT is operating around mid vane position as
opposed to the extremities of its travel and therefore 3. Porter, B. Ross-Martin, TJ., Truscott, AJ., “Control Technol-
remains out of the hysteresis band. There is still over- ogy for future low emissions Diesel passenger cars” Paper
boost, but this settles rapidly, the absence of the EGR C517/035/96, IMechE International Seminar on Application
leakage path also allows boost pressure to rise faster of Powertrain and Fuel Technologies to meet Emissions
(Figure 7f). Standards, London, 24-26 June 1996.
4. Horrocks R.W., “Light duty Diesel - the emissions chal-
lenge” ImechE Proceedings Vol 206, pp249-255, 1992.

4
5. Dunne, J.M., Greening, P.J., “European Emissions Stan- GLOSSARY OF ABBREVIATIONS
dards for the year 2000” Worldwide Engine Emission Stan-
dards and How to Meet Them, IMechE proceedings 25-26 AFR – Air to Fuel Ratio
May 1993
ECU – Engine Control Unit
6. Greening, P.J., “The Future of European Emission Regula-
tions”, Engine & Environment AVL Proceedings Grazer EGR – Exhaust Gas Recirculation
Congress 4,5 Sept1997 HSDI – High Speed Direct Injection
7. Dorey, R.E., Guebeli, M., “Real Time Powertrain Simulation
IDI – Indirect Injection
for Dynamic Engine Testing Using a Hydrostatic Dyna-
mometer” IEE Colloquium on Powertrain Control, 16 May NOx – Oxides of Nitrogen
1990
NVH – Noise Vibration & Harshness
8. Brace, C.J., Deacon, M., Horrocks, R.W., Vaughan, N.D.,
Burrows, C.R., “Impact of alternative controller strategies PM – Particulate Matter
on emissions from a diesel CVT powertrain-preliminary VGT – Variable Geometry Turbocharging
results” IMechE Paper C517/044/96
uHC – unburned Hydrocarbons
9. Greeves, G., Wang, C.H.T., “Origins of Diesel Particulate
Mass Emission” SAE paper 810260

Exhaust Gas
Aftertreatment

Variable
Geometry
Intercooler Turbocharger
(VGT)

Exhaust Gas
EGR Cooler
Recirculation
(EGR)

4 Valves per
cylinder with Electronic
port- Fuel Injection
deactivation System

Figure 1. Schematic represetation of a modern high speed direct injection diesel engine

5
 Boost Pressure Map

Boost
Pressure

Fuelling
Engine Speed

Figure 2.

Transient HC emissions
(data points+line)
compared to steady state
emissions map (mesh)
1500

Hydrocarbons [ppm C3H8]

1000

500
(2)

05000
4000
120
(1) 3000 m]
100 d [rp
Eng 80 60 2000 ee
in e T e Sp
orqu 40 20 1000 gin
e [N 0 En
m]

Figure 3. Comparison of transient and steady state hydrocarbon emissions

6
Figure 4. Schematic of the University of Bath Transient Engine Test Facility

7
 Pedal Step Transients compared with Steady State Boost Map

Boost Pressure
1500 rpm
2500 rpm

Speed
3500 rpm

Engine
Fuelling

(a)

P edal Step Transients com pared w ith S teady State Boost M ap

B oost
P ressure

1500 rpm
2500 rpm

3500 rpm
Engine
Speed

Fuelling

(b)

Figure 5. Deviation of boost pressure from steady state behaviour (alternative views of same plot)

8
9
10