Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02306483 2000-02-28
Certified Translation from the German Language
Device for the analysis of motor vehicle exhaust emissions
1 Introduction
The exhaust emissions of passenger and commercial vehicles are
the cause of various types of harm to the environment. The
introduction of emission-limiting legislation has forced, and is
forcing, vehicle manufacturers to reduce the emissions of individual
vehicles by - for example - developing advanced engines and
exhaust systems.
One reason for a vehicle failing to conform to emission regulation is
worsening performance, in terms of a gradual increase in exhaust
emissions, as the vehicle ages. This is caused by wear and also, in
part, by the incorrect functioning of components in the drive and
emission-reduction systems.
The normal inspection procedure involves regular tests to attempt to
keep emissions at or near their original level. The disadvantage of
this method is that faults remain undetected until the next inspection,
and excessive emissions meanwhile continue to be produced.
In the first few seconds after the engine is started, the catalytic
converter - which has not yet reached its running temperature -
barely affects the level of harmful exhaust emissions. An engine
produces about 70% of its total emissions just after starting from
cold, so an ideal system for reducing harmful emissions would cover
this phase, which is precisely the phase that remains untouched by
current systems of emission control detection.
2 The state of the art
"On board diagnosis" (OBD) is one new system for reducing harmful
emissions. The term refers to an emission control system which
uses sensors to monitor the performance of those individual
components of a passenger or commercial vehicle that have a
bearing on exhaust emissions. An early version of OBD for
passenger cars - the OBD I Law - has already been in use for a
considerable time in the USA and is gradually being superseded b~~ ~h~: ~
;:,.,
the more stringent OBD II Law for models from 1995 onwards. VI(~e
OBD I only affected the performance monitoring of components~~ ~~~,2 .~c~
s,; ~ . .
,. . 1
.., ~;
CA 02306483 2000-02-28
2
forming part of an electronic engine control system, OBD II requires
the control of all components relevant to emissions. The law
expressly stipulates the monitoring of catalytic converters, lambda
probes, fuel systems, air injection systems, exhaust gas
recirculation, tank ventilation and the detection of misfiring. In the
event of a component breaking down or malfunctioning, a warning
lamp lights up on the dashboard and an error code is memorised.
The fault detected should be located as precisely as possible and
described. The information is then stored in order to permit swift
identification of the fault at the workshop (using a standard interface)
and to allow repair of the defective part.
A further step in this field is the use of "On Board Measurement"
(OBM). Systems for the direct analysis of vehicle emissions are
widely familiar. Some examples, among many others, are the
German public patents 32 32 416, 33 39 073, 36 08 122, 37 16 350,
39 32 838, 40 05 803, 41 24 116, 42 35 225, 43 07 190, the DE
specification 43 19 282 C1 and US patent specifications 4 803 052
and 5 281 817, along with further patent applications GB 2 264 170
A, EP 0 196 993 A2 and WO 94/09266. These documents should be
referred to for explanations in greater detail of items mentioned
here.
Applications I11 and 121 are concerned, in a narrow sense, with
subjects related to this field and are dealt with in greater detail for
this reason. Thus patent 111 describes an infrared measuring system
that monitors the operating condition of the catalytic converter via a
lateral access opening in the unit and measures the gases present
inside. In 121, the system in question is a rapid detector that uses
several infrared cells connected in series to permit a chronological
resolution of 0.1 - 0.2 sec. Neither source gives an indication of
continuous measurement of harmful exhaust system emissions
upstream of the catalytic converter.
Written sources reveal that no current measuring system is capable
of providing a continuous record of actual emissions, either in the
cold-start phase or during operation. Neither is it possible to detect
fluctuations or indicate faults.
.; '_.
.,;,.. w:.:,,- .
It
~.CS
.' j.
I
'.v ~.
CA 02306483 2000-02-28
3
3 Description of the object of the patent
Vehicles will, in the future, be fitted with an integrated OBM system
for the purposes of emission analysis. This system will analyse
certain elements of the exhaust gases, and a comparison of current
concentrations with a set of stored target values will permit the
detection of faults in the ignition system. A warning system will then
be activated whenever the "satisfactory" level specified for the
individual model of vehicle is exceeded, clearly and repeatedly, over
a period. "Over a period" means an extended length of time,
"repeatedly" signifies an excessive reading on not one, but various
occasions and "clearly" refers to a concentration that is outside the
margin of tolerance specified.
Fig. 1 shows an example of how the concentration of harmful
substances is affected by faults in the ignition system (caused by
misfiring (1 ) in this case).
The measurement of emissions is hindered by the fluctuating
conditions present in the vehicle. A measuring system must on one
hand keep to the general margins of tolerance and specifications
valid for the vehicle while, on the other hand, it is precisely the
exhaust-related elements of pressure, moisture, temperature and
flow rate that are subject to sharp fluctuations. In order to solve this
problem, especially robust micro system components are required -
both for exhaust gas processing and for the detection of the
elements of which the gas consists.
One device for the analysis of vehicle exhaust gases is already
familiar in the shape of DE 196 05 053 A1. Problems related to
specified operation have however been encountered with this
device, as they have with measuring devices described in other
documents. The presence of vibrations in the vehicle requires that
measuring systems be of highly stable construction and also
resistant to soot, dust and aerosol precipitation. They must
furthermore attain a high level of resolution, as the constituent
components of the exhaust gas being analysed - e.g. carbon
monoxide (CO), hydrocarbons (HC), and oxides of nitrogen (NO) -
are present in extremely low concentrations, precisely in those
petrol-driven vehicles that are fitted with a catalytic converter. -.
-:": ~ c:.-;~,,
The emission analysis method used in the new OBM system,;'
submitted here is an infrared gas absorption process. This in:yentton~y.:--:-~
..
..
CA 02306483 2000-02-28
4
is based the assumption that, in order to obtain the required
resolution, optical path length must be increased. The optical cell
can thus be fitted to a vehicle if it is incorporated at the construction
stage.
4 Technical design of the measuring system
The main assembly of the OBM system in a vehicle is shown in Fig.
2, along with the main components of the ignition system. The
engine (2) produces exhaust fumes as it burns fuel. In the catalytic
converter (3), harmful elements are transformed into less toxic
substances. The vehicle OBM system consists of the sampling point
{4), exhaust gas processing unit (5), analysing device (6), exhaust
system (7) and data cable (8) that provides the link between the
display unit (9) and analysing device (6).
Gas is extracted from the exhaust system upstream of the catalytic
converter, as this is the only way in which an evaluation of the
condition of the ignition system as a whole can be made.
Exhaust gas processing is illustrated in the gas flow diagram (Fig.
3). Soot and particles are removed from the exhaust gas using a
disposable filter (11 ). A solenoid valve (12) is used to change over
between exhaust gas and calibration gas (see chapter 7). The
measuring gas pump (13) sends the gas to be measured to the
analysing device (6) via the pressure reducer (14) and flow meter
(15).
Exhaust gas analysis is carried out in the analysis device (optical
cell) following the principle of infrared gas absorption. This device
consists of an infrared source (transparent tube), the radiation from
which is directed to the measuring head via a measured length
(optical cell). The optical cell can consist of one straight, highly
reflective tube or of several tubes with reflective heads. The two
pyroelectric measuring sensors fitted to the measuring head are
equipped with various optical filters and produce one signal that
depends on measurements and another which acts as a reference
signal. The ratio formation of these signals reduces the disturbing
influences (temperature, pressure, contamination) acting on the
measuring signal. The use of the pyroelectric principle requires a .. ,
synchronised radiation source. Electrical timing of the radiation.~.,<::~~~
~~%-,~~w-
source avoids delicate mechanical components (chopper). Tl~e
measuring system is thus rendered more robust, with an optical selJ..:,..~,~.~
'r,.-
..
'. .; ..,.
~ ilv::.~I~_f J'
... y.
~. ~.~\~: :.r ~ !'v!~ .
CA 02306483 2000-02-28
(measured length) made of stainless steel. In the event of the device
being contaminated or suffering component faults, the advantage of
the modular construction of the unit becomes clear. Single
components such as filters can simply be replaced.
5 Modification kit for emission analysis on older vehicles
In the case of older vehicles, which have not been fitted by the
manufacturer with an OBD or OBM system, engine and exhaust gas
processing performance cannot be measured other than by analysis
of the exhaust gas itself. For this reason, a modification option
should be available.
The disadvantage of carrying out modifications with an OBD system
is the large number of transducers, for which there is neither
sufficient room nor electronic connections. It is therefore more
convenient to install an on board measuring system.
Fig. 4 shows a modular OBM modification system of this type.
Exhaust gas sampling is carried out using a sampling sensor (16)
fitted to the end of the exhaust pipe. The gas is cleaned and dried in
the exhaust gas processing unit (17) and then pumped onwards to
the analysing device (6). The display unit (9) on the dashboard then
indicates information about the status and operation of the OBM
system.
The fitting of the modification kit to the vehicle is illustrated in Fig. 5.
This involves attaching the sampling sensor (16) to the end of the
exhaust pipe, while the analysing device (6) and gas processing unit
(17) can be installed in the car boot. The display unit (9) can be
hung from a ventilator grille or fitted elsewhere on the dashboard.
6 Cold-start measurement and adsorption trap
The cold-start phase (see Fig. 6) is when 70% of total engine
emissions are produced, and an on board measurement system
calibrates these gases. These readings can be used to activate an
HC adsorption trap (10, see Fig. 2) used for collecting cold-start
emissions in the exhaust flow path. Emission measuring makes it
possible to synchronise the adsorption trap in the exhaust flow path....
to activate at exactly the right moment, or to start desorption.
".
li
Desorption of the small amount of retained hydrocarbons until;'the
' /Cw
/7
..
s
t
f.
~;,:.7.. r. . "f: V
CA 02306483 2000-02-28
6
catalytic converter reaches a temperature at which satisfactory
conversion is guaranteed.
The energy consumption of the cold-start measuring system is
extremely low, so it can enter operation before the cold-start phase
has actually commenced. Control can be carried out, for example,
by means of either a seat occupation sensor or a sensor on the
ignition lock, which can also be used to activate the HC adsorption
trap in the exhaust flow path.
7 Reference line calibration
The measuring principle of infrared gas absorption is sufficiently well
known. The problems with this measuring principle with regard to
fluctuating ambient conditions have already been described in
section 3. We will now examine the various measurement correction
methods used.
The most common problem is the shifting of the zero point - i.e. the
reading for uncontaminated gas is not zero. This problem can be
solved by calibrating the system with ambient air, proceeding as
follows:
The solenoid valve (12) in the exhaust gas processing unit (5,17) is
automatically switched over either after a pre-set period or as a
result of detected external factors, allowing ambient air to enter the
analysis device (6). The concentrations of CO, HC and NO present
in the ambient air are so low that they can safely be regarded as
zero. The use of mathematical compensation allows the zero line to
be calibrated. After this has been carried out, the sensitivity of the
device usually recovers its original levels and the system begins
once again to display reproducible readings. Fig. 7 shows the effect
of a zero line correction. The graph illustrates how the zero line (18)
has been displaced by temperature drift and also shows the re-
corrected measuring curve (19) produced after calibration. This
procedure, with an interruption in emission recording, has no
influence in terms of nominal values on the meaningfulness of
measurements, whose purpose is - in any case - the detection of
faults in the exhaust system rather than the providing of continuous
monitoring.
~ ~ ~c;~ -:::,;. .
~C~J - y
a
v;~. ,
,,
CA 02306483 2000-02-28
7
8 Adjusting the sensitivity of the measuring signals using
the C42 concentration of the ambient air
The zero-calibration procedure described in section 7 has the
advantage of avoiding the need for constant sensitivity adjustment,
as this procedure also produces the right correction for the
sensitivity level (and thus all others). A sensitivity check can
nevertheless be carried out as follows:
The atmosphere in all parts of the world (with clean non-city air) has
an average C02 concentration of 350ppm. This fact can be used to
check sensitivity, as this concentration matches the measuring
ranges of the components normally detected in the stream of
exhaust gas. CO, HC and - above all - NO in fact have weaker
absorption bands than C02, but with correspondingly higher peak
concentrations. According to the Lambert-Beerschen equation, the
same optical cell length or - in practical terms - the same optical cell,
can thus be used. If uncontaminated ambient air is now fed into the
exhaust gas analysing device (6), the system should show the
average C02 concentration - once the zero point reset procedure
described above has been carried out. One can now be sufficiently
sure that the sensitivity level of the other measurement factors is
also correct.
The disadvantage of the above procedure is that local C02 concen-
trations fluctuate sharply due to external influences. This is especial-
ly true in densely populated areas, where road traffic can produce
extremely high concentrations of CO2. Fig 8 shows the carbon di-
oxide concentration of the ambient air during a test drive. After ad-
justment of the zero point using synthetic air (20), the vehicle was
driven through a small municipality (21 ) where the C02 concentra-
tion was relatively constant. A test drive through a larger town (22)
with crossings and traffic lights reveals high, sharply-fluctuating C02
concentrations. Finally, a measurement carried out in a quiet interior
courtyard (23) is closer to a natural C02 concentration.
9 Sensitivity adjustment via the C02 concentration in the
exhaust gas
One possible way of avoiding the problems resulting from the
fluctuations from natural C02 concentrations described in section _8,: :<:
is the monitoring of the C02 concentration in the vehicle exhaust:v .
The ignition process makes this value relatively stable, so tha~t~this
CA 02306483 2000-02-28
8
concentration can be used as a reference value for adjusting the
sensitivity of the individual meter flumes. However, the high concen-
tration (12% by volume) of C02 in the exhaust gas means that the
COZ beam path in the optical measuring cell must be arranged
differently to that used for other harmful gases. The optical path for
C02 measurement basically has to be shorter than that used for the
contaminants CO, NO and HC.
Correction - using a software-controlled filter - of zero line
of measuring signal displaced due to temperature
fluctuations
A measurement value is normally determined by the production of a
ratio from the signal for the contaminant present (measuring signal)
and the reference signal.
The signal progressions for measuring signals and reference signals
reveal a great similarity. A ratio procedure can thus be modified if a
certain margin of tolerance is determined around the signal pro-
gression and the ratio is set to "one" within this margin. This allows a
range for zero concentration to be obtained, and only in the event of
this margin of tolerance being exceeded will a concentration cor-
responding to the values of the then determined real ratio be dis-
played. Note: the concentration "zero" need not necessarily corres-
pond to the ratio "one", but it does obtain the best measuring result.
11 Compensation of temperature drift by examination of the
dynamics of the signal progressions
Experience shows that extreme dynamic conditions are present in
motor vehicles (brusque momentary system alterations, compared
to the cycle period of the radiation emitter). This means that it is
easy to differentiate between genuine measuring signals (i.e. those
produced by the exhaust gas) and the slower-fluctuating variations
that depend on temperature. To carry out correction, the first deriva-
tion of the concentration process must be produced according to
time. The first derivation records only genuine step functions that
occur, for example, when the vehicle accelerates. Fig. 9 shows an
actual measuring value progression. The first derivation (25) was
produced from the original measuring signal of contaminant HC (24).
It can be clearly seen that measurement signal fluctuations (26) .
provoked by the influence of temperature approach zero in the. ~~
derivation (25).
'.,. Lc, .,
r~n~
CA 02306483 2000-02-28
9
Once the step function places from the first derivation have been
found according to time, the points can be recognised with clear
step characteristics. If such a genuine step function appears, i.e. if a
measurement value exceeds the previously defined margin of
tolerance by a permitted amount relative to the differential curve, this
point must be used as a reference relative to the actual concen-
tration curve used for evaluation. When the first derivation returns to
zero, the software-controlled filter once again emits the zero line as
an unaltered, stable line. Thus you have at your disposal during the
test drive one of two things. The first possibility is an absolute zero
line - without fluctuations, as no step functions have appeared and
the fluctuations caused by temperature are ignored. The other
possibility is that whenever real, dynamic step functions occur, such
as when accelerating, changing gear, braking, etc., the original
measuring signals {obtained from the concentration curve) are
observed according to the first derivative.
12 Setting the original signal strengths in the channels of the
IR gas analyser
A further correction method involves resetting the signal strength by
means of an electronically regulated amplification controller.
Since the margin of reference for infrared gas absorption is set in
such a way that virtually no absorption takes place at this limit, the
infrared detector reference channel measuring signal should always
maintain its original strength. The effects of temperature and vehicle
wear do however cause noticeable fluctuations in this signal.
In order to compensate for the signal fluctuations caused by
temperature conditions, the possibility exists to monitor continuously
the reference signal by means of a measurement, control and
regulating device built into the system. Whenever the reference
signal deviates by a pre-defined margin from the value originally
adjusted at initial calibration, all signals are realigned - using an
electronically regulated amplification controller - with the original
signal strength. Fig. 10 shows the original curve for the reference
signal (27), the weakened curve resulting from wear or temperature
drift (28) and the curve that has been corrected by electronically
controlled amplification (29). This method retains the full range of
signal dynamics.
~~ - Z . 5 ~.
' > i'.
I ~ -~~