RK21RDO
CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
PREDICTIONS OF THE PERFORMANCE AND EXHAUST
EMISSIONS PRODUCED BY SMALL LIGHT DUTY
VEHICLES POWERED BY PI AND IDI
DIESEL ENGINES
DP.81/297
-------
RG1RDO
CONSULTING ENGINEERS
Ricardo Consulting Engineers Ltd.
Bridge Works Shoreham-by-Sea
Sussex BN4 5FG England
19th February 1981
Report No. EPA-460/3-81-011
PREDICTIONS OF THE PERFORMANCE AND EXHAUST
EMISSIONS PRODUCED BY SMALL LIGHT DUTY
VEHICLES POWERED BY PI AND IDI
DIESEL ENGINES
DP.81/297
-------
RK2RDO
DP.81/297
Restricted CONSULTIHQ ENGINEERS
Report No. EPA-460/3-81-011
PREDICTIONS OF THE PERFORMANCE AND EXHAUST
EMISSIONS PRODUCED BY SMALL LIGHT DUTY
VEHICLES POWERED BY PI AND IDI
DIESEL ENGINES
SUMMARY
This report describes an exercise in which a computer simulation pro-
gram was used to predict the likely performance, fuel economy and exhaust
emission levels of light duty vehicles weighing from 1000 to 2000 Ib when
powered by naturally aspirated diesel engines of 0.8 litre displacement
having direct injection (DI) and indirect injection (IDI) combustion systems.
It was assumed that the vehicles were fitted with an efficient continuously
variable transmission.
Based on currently available engine data it appears that the 0.8 1 DI
engine would not comply with current U.S. exhaust emissions legislation
(.41/3.4/1.0 g/mile HC/CO/NOx) when fitted to vehicles weighing more than
1500 Ib, but other vehicle performance characteristics (apart from noise
and vibration) may be acceptable.
The predicted results suggest that the 0.8 1 IDI engine might prove to
be an acceptable power unit in terms of vehicle performance and exhaust emis-
sions in vehicles weighing up to 1750 Ib.
Predicted fuel economy over the FTP (urban cycle) for a 1000 Ib
vehicle using the DI and IDI engines was 94 and 86 miles/US gall. (Highway
results 86 and 79 miles/US gall.) respectively, in the 2000 Ib vehicle the
corresponding figures were 56 and 52 miles/US gall, (and 56 and 50 miles/
US gall, Highway).
-------
Restricted
CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
1. INTRODUCTION
Increasing emphasis is being placed on the need for improved vehicle
fuel economy. Two approaches currently being used in order to achieve this
end are the development of lighter vehicles and the use of diesel engines
in place of gasoline fuelled units. At present all production light duty
diesel engines employ combustion systems of indirect injection (IDI) con-
figuration; the relative advantage, from the point of view of better fuel
economy, of direct injection (DI) combustion systems is generally apprecia-
ted (1, 2, 3)* and research and development aimed at producing DI engines
which are acceptable in all respects for use in light duty vehicles is
being conducted.
The objective of the work reported here was to predict, by computer
simulation, the performance and exhaust emission characteristics of possible
future light-weight passenger cars powered by naturally aspirated DI and
IDI engines.
This work was requested by E.P.A. as a specific task to be carried out
under the 1980/1981 year consulting agreement and the choice of engine type
and size, and vehicle weights was agreed with E.P.A. at a meeting at Ann
Arbor on 12th November 1980, see Ricardo note DP.80/1904.
2. SIMULATION PROCEDURE
2.1 Simulation Program
The computer program used (CVSIM) is primarily designed to predict
the levels of exhaust emissions and fuel consumption to be expected from a
vehicle during operation over a prescribed velocity cycle (in this case the
LA4 and Highway drive cycles). Vehicle performance, in terms of accelera-
tion times, can also be predicted.
Essentially, the program analyses the driving cycle and, from a know-
ledge of vehicle characteristics, calculates the engine speed and brake mean
effective pressure (bmep) required to drive the vehicle over each velocity
increment in turn. Knowing these two parameters the levels of exhaust emis-
sions and fuel consumption are extracted from engine test bed performance
maps which are represented in the program input data by two dimensional
numerical arrays.
2.2 Simulated Engine/Vehicle Combinations
Two naturally aspirated diesel engines were employed in the simulation
exercise. Both were of 0.8 litre total swept volume, in two cylinders; one
had an IDI (Ricardo Comet) combustion system, the other was of DI configura-
tion. The maximum power outputs of the IDI and DI engines were respectively:
25 bhp @ 4600 rev/min and 23 bhp @ 4400 rev/min. During simulation runs
each engine was fitted in turn to vehicles of 1000, 1250, 1500, 1750 and
2000 Ibs weight.
*Numbers in parentheses refer to references listed in Section 6.
-------
RK2RDD
DP.81/297
Restricted COMSULTIMO ENGINEERS
Report No. EPA-460/3-81-011
2.3 Input Data
The LA4 drive cycle was used by the program so that exhaust emission
and fuel economy data corresponding to vehicle tests to the U.S. Federal
Test Procedure would be generated. The Highway fuel economy test drive
cycle was also used to produce appropriate estimates of fuel economy.
Estimates of performance data for each engine, i.e. torque curves and
steady state emissions and fuel consumption maps, were based largely on
actuar measurements made on light duty IDI and DI engines; in general
these engines had not been optimised with regard to particular parameters,
e.g. it was assumed that EGR, which might be used to reduce NOx emissions or
exhaust after treatment for reduction of HC or particulate emissions were
not applied. Where necessary some allowance was made for the effects of
differences in the numbers of cylinders and cylinder swept volume, since
most available data related to larger engines with a greater number of
cylinders.
With regard to vehicle characteristics the inertia and road load set-
tings specified in the Federal Register (Vol. 42, No. 124, 28th June 1977)
for each vehicle weight were used since only chassis dynamometer tests were
being simulated.
The driving tyres (BR50-13) were assumed to have a rolling radius of
0.28 m (11 in). In order to derive final drive (axle) ratios it was assumed
that maximum vehicle speeds occur at maximum power engine speed when using
a primary drive (gearbox) ratio of 1:1. This produced the final drive
ratios shown in Table 1 (calculated maximum speeds and acceleration times,
based on the vehicle road load characteristics given in the Federal Register,
for the various vehicle/engine combinations are also shown).
An automatic, continuously variable, transmission (CVT) was used in
all the simulation exercises, this was assumed to have an overall span of
5:1. An overdrive top ratio of 0.8:1 was specified, the bottom ratio then
became 4.0:1. It was assumed that ratio changes from one extreme to the
other could be accomplished in 1 second. A gearbox transmission efficiency
as shown in Fig. 1 was assumed. The final drive transmission efficiency was
set at a constant value of 92%.
The polar moment of inertia of the various drive train components were
assumed to be:-
driving wheels and axle - 1.6 kg.m2
engine and CVT - 0.15 kg.m2
3. RESULTS
The results produced by the simulation program for each vehicle/engine
combination are listed in Table 2. Figs. 2 and 3 are graphs indicating the
variation of exhaust emissions and vehicle performance with vehicle weight.
Due to the uncertainties noted below the level of confidence in the
predicted results is rather low. With regard to the accuracy of the results
predicted by the simulation program the following confidence levels may be
considered realistic:-
-------
RI0RDO
DP.81/297
Restricted CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
Vehicle Performance ±10%
(0-50 mile/h acceleration time)
Fuel Economy ±10%
Exhaust Emissions HC +75, -40%
CO ±10%
NOx ±10%
Particulates ±15%
These levels do not include the element of uncertainty relating to the per-
formance and emissions data used as input for the simulation. Especially
in the case of the DI engine future research and development may result in
significant improvements in all aspects of engine operation.
4. DISCUSSION
When considering the validity of the results produced by this exercise
the following points should be noted:-
i] The computer program produces simulated results of transient tests
using engine performance and emissions data derived under steady
state conditions, it is likely that under true transient operation
engine performance and emissions levels will show some variations
from predicted results.
ii] All engine data used as input is nominally acquired at normal operat-
ing temperatures. In actual FTP vehicle tests the engine starts from
cold and hence its performance and emissions during the early part of
the test may be considerably different to what is predicted.
(These two points have been confirmed in previous work in which simu-
lation results were compared with measured data when some divergence,
especially in the case of HC emissions, has been observed).
iii] The input data relating to engine performance and emission levels
were based on results achieved by engines operating over a fairly wide
speed range (typically 1000-4500 rev/min) as required by current pro-
duction engines having conventional transmission systems. When using
a CVT it may prove advantageous in terms of fuel economy and emissions
to optimise engine operation over a narrower speed range, most varia-
tion in vehicle speed then being achieved by employing a greater span
of transmission ratios.
iv] All the input data relating to engine performance and emission levels
were estimated based on a knowledge of the figures achieved by actual
engines which generally had greater numbers of cylinders (in most
cases four), and in some cases had larger cylinder displacements (up
to 0.6 1).
v] Development of small DI engines suitable for use in light duty vehicle
applications is at an early stage. Future work especially with regard
to improved combustion systems and fuel injection equipment will pro-
bably lead to considerable reductions in exhaust emissions from the
levels assumed in this exercise.
-------
DP.SV2S7 RK21RDO
Restricted CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
vi] The basic performance and emissions data used as input were typical
average levels representative of current engines not optimised for
any specific parameter. Lower emission levels may be achievable by
employing exhaust after treatment, EGR, etc., but may incur penalties
in terms of reduced engine performance and/or higher fuel consumption.
With regard to small DI engines few data concerning the performance of
current units are available and hence there is considerable doubt as
to the fuel economy and emissions levels which may ultimately be
achieved.
vii] Production versions of continuously variable transmissions having the
characteristics assumed for this exercise are not yet available.
viii] The vehicle road load characteristics are assumed to be as set out in
the Federal Register. In real vehicles aerodynamic drag and rolling
resistance may be significantly different so that performance levels
may show considerable variations.
As would be expected the results of this exercise suggested that in-
creased vehicle weight caused increases in exhaust emission levels and fuel
consumption while vehicle performance, in terms of acceleration time deteri-
orated.
An interesting result of this exercise is that predicted fuel economy
over the LA4 (Urban) drive cycle is superior to that achieved during the
Highway cycle. Previous simulation exercises of this type conducted by
Ricardo (which in all cases have been run on vehicles more similar, in terms
of vehicle weight, engine displacement and with use of fixed transmission
ratios, to current conventional vehicles) have produced results suggesting
that in all cases Highway fuel economy is better than that returned during
the Urban cycle. This tendency is also confirmed in actual tests on real
vehicles. Computer simulation runs with the vehicles used in this exercise,
but fitted with larger capacity engines (retaining the same fuel consumption
maps) and lower final drive ratios, so that shorter 0-50 mile/h acceleration
times (^12 s) and higher top speeds (%80 mile/h) could be achieved indicated
that Highway economy was then better than that returned over the Urban cycle.
It therefore appears that the apparent anomaly, revealed in the present
exercise is a real phenomenon in the particular cases of engine/transmission/
vehicle concepts used.
The DI engine option displayed a fuel economy advantage of 7-11% over
the IDI unit but was inferior in terms of exhaust emissions and vehicle
performance.
With regard to exhaust emissions use of a 0.8 1, naturally aspirated,
IDI engine in vehicle weighing 1000-1750 Ib produced simulated exhaust
emission levels within the 1982 standard (.41/3.4/1.0/0.6 g/mile HC/CO/NOx/
particulates). At all vehicle weights above 1000 Ib the predicted levels
using the DI engine exceeded the standards for particulates and (at vehicle
weights greater than 1500 Ib) NOx. Predicted acceleration times for all
engine/vehicle builds were rather poor; taking a 0-50 mile/h time of 25
seconds as the criterion of acceptability for vehicles of this type the IDI
engine would appear to be acceptable for vehicles of up to 1750 Ib while
the DI unit appears to be feasible in vehicles up to 1500 Ibs.
-------
•"••"/*" RIG1RDO
Restricted CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
Apart from performance, fuel economy and exhaust emissions character-
istics there are several other parameters of importance when assessing the
suitability of an engine for use in a particular vehicle. The level of
noise and vibration produced by diesel engines, particularly of DI configura-
tion, is quite high; in two cylinder engines vibration problems are likely
to be very pronounced and even the use of balancer shafts and special
purpose engine mounts may not reduce the problem to acceptable levels
especially in the light weight vehicles considered here.
5. CONCLUSIONS
Based on current knowledge of engine characteristics the results of
this simulation exercise suggest that a naturally aspirated IDI diesel
engine of 0.8 1 could be an acceptable power unit with respect to accelera-
tion time and exhaust emissions for light duty vehicles weighing up to
1750 Ib. Above this weight vehicle performance and emissions would probably
prove to be unacceptable.
A 0.8 1 DI diesel engine could provide reasonable performance in
vehicles up to 1500 Ib but some reductions in emissions, especially parti-
culates are necessary if U.S. requirements are to be satisfied.
For the 1000 Ib vehicle predicted fuel economy over the FTP (Urban
cycle) was 94 and 86 miles/US gall. (86 and 79 miles/US gall, Highway) for
the DI and IDI engines respectively.
Some factors likely to have a considerable influence on engine/
vehicle acceptability, e.g. noise, vibration, cost, have not been addressed
in this study.
6. REFERENCES
1. Downs, D. & French, C.C.J.
DEVELOPMENT OF ALTERNATIVE ENGINES AND THEIR FUEL REQUIREMENTS
(Tenth World Petroleum Congress, Bucharest, 9-14 September 1979)
2. French, C.C.J.
FUEL EFFICIENT ENGINES FOR LIGHT DUTY VEHICLES
(VDI-Berichte, 1980, No. 370, pp.81-86, XVIII FISITA, Hamburg,
5-8 May 1980)
3. Monaghan, M.L.
THE HIGH SPEED DIRECT INJECTION DIESEL FOR PASSENGER CARS
(SAE 810477)
CRM/BEW
-------
DP.81/297
Restricted
RlflRDO
CONSULTING ENGINEERS
Report No. EPA-460/3-81-011
TABLE 1
Characteristics of Vehicles Powered by Small Diesel Engines
Vehicle
Weight
(lb)
1000
1250
1500
1750
2000
800 cc IDI Diesel Engine
Final
Drive
Ratio
4.05
4.20
4.35
4.50
4.65
Maximum
Speed
mile/h
69.5
67.4
65.6
63.9
62.3
0-50
mile/h
time
(sec)
14.3
17.5
20.7
24.9
28.8
800 cc DI Diesel Engine
Final
Drive
Ratio
4.05
4.20
4.35
4.50
4.65
Maximum
Speed
mile/h
68.1
66.1
64.3
62.5
60.8
0-50
mile/h
time
(sec)
15,6
19.5
23,3
27.7
31.9
-------
TABLE 2
Predicted Exhaust Emissions & Fuel Economy
Vehicle
Weight
(Ib)
1000
1250
1500
1750
2000
800 cc IDI Diesel Engine
FTP Exhaust Emissions g/mile
HC CO NOx Parti-
culates
.11 .72 .57 .35
.13 .88 .64 .42
.15 1.10 .73 .49
.17 1.31 .81 .57
.19 1.52 .89 .62
Fuel Economy
mile/US gall
LA4 Highway
85.6 78.6
75.0 70.4
65.7 62.9
58.2 56.0
52.3 50.2
800 cc DI Diesel Engine
FTP Exhaust Emissions g/mile
HC CO NOx Parti-
culates
.23 .88 .77 .50
.26 1.17 .86 .61
.29 1 .53 .97 .71
.33 1 .94 1 .08 .82
.37 2 .38 1 .20 .92
Fuel Economy
mile/US gall
LA4 Highway
94 .2 85 .7
82 .1 76 .7
71 .9 68 .6
63 .0 61 .5
55 .9 55 .7
oo
70
70
-------
/.#.
FIG. No. f
Drg. No. O43731
Date f£B' 81
ASSUMEO A4E.CHA^JICAL C.VT
-------
RK2RDO
P«£DlCTEO EXHAUST E.MIS5IOJJ LEVELS
e.g.M. /
-------
RI0RDO
PREDICTED VEHICLE PERFORMANCE CHARACTERISTICS
C.. R. M . I »V. g^^
FIG. No. 3
Drg. No. O 4 44 33
Date 2 Lo - 5 - S1
+
• •
O • & LITRE I Dl
O • & LITRE Dl
LA 4 CuR&A>j2 RESULTS
AV/CW*V4y RZSULT3
------- |