WO1995000833A1

WO1995000833A1 - Characterisation of internal combustion engines by the optical measurement of a plurality of quantities in the combustion chamber

Info

Publication number: WO1995000833A1
Application number: PCT/DE1994/000696
Authority: WO
Inventors: Peter Andresen; Gerd GRÜNEFELD; Volker Beushausen; Werner Hentschel
Original assignee: Peter Andresen
Priority date: 1993-06-24
Filing date: 1994-06-22
Publication date: 1995-01-05
Also published as: DE4320943A1; DE4320943C2; EP0674764A1

Abstract

The operation of internal combustion engines depends considerably upon quantities which cannot be set accurately enough from outside, e.g. the stoichiometry of the fuel-air mixture before ignition, the proportion of exhaust gas in this gas mixture and its temperature. The most simultaneous and precise measurement possible of these quantities in the combustion chamber helps greatly to explain engine problems. The novel laser Raman light scatter process, for instance, permits this measurement. The process is contactless and provides good time (i.e. crank angle) and position resolution. Intense pulsed u/v lasers are used as the excitation light source for the Raman and Rayleight scatter. The laser light (1) passes through a window (17) in the wall of the cylinder being measured into the upper part of the combustion chamber, especially to analyse the final gas before ignition. The laser-induced emissions (especially the Raman and Rayleigh scatter) can be brought out from the combustion chamber in various ways, e.g. via the same window (17) and a dichroic mirror (18). The various emissions, especially the Raman emission from fuel, oxygen, nitrogen, water, etc., are quantitatively and simultaneously measured by intensified short-time cameras (CCD) combined with upstream wavelength separation (spectrometer) (8). It is thus possible also to obtain local resolution along an axis in the combustion chamber. The excitation wavelength in the u/v permits high-precision single-pulse measurements so that the small cyclic fluctuations in the mixture formation (and combustion) in the engine can be resolved. In many cases, moreover, it is necessary to separate the Raman emission from interfering (fluorescent) emissions, and this is done with the aid of polarisation properties. The quantities relevant to combustion like stoichiometry and proportion of exhaust gas are found by calculating the ratio of Raman intensities, which produces high accuracy of measurement.

Description

Characterization of internal combustion engines by optical measurement of several sizes in the combustion chamber.

The way an engine works depends crucially on quantities that cannot be set from the outside precisely enough, such as the absolute value and the spatial distribution of the stoichiometry or the exhaust gas content. Although these sizes are roughly specified, fluctuations occur in the fine, which have a considerable influence on the operating conditions of the engine. In particular, in the case of the gasoline engine - for example by measuring the pressure curve - it is known that the power output is subject to cyclical fluctuations. These cyclical fluctuations are caused by changes in the percentage range in the amount of air supplied, the ^τ ge of the fuel supplied and changes in the portion of the exhaust gas uneven from the previous cycle. Therefore, these small fluctuations play a crucial role for the way of working and optimization, especially of gasoline engines.

Since the stoichiometry and the exhaust gas fraction have a decisive influence on the power output, the controlled setting of these variables plays an important role in the development. The effect of external measures on these variables can be checked by means of an exact simultaneous measurement of these variables and those resulting from other particle densities, for which a measuring arrangement is specified in the present patent specification.

With the measurement methods available to date, a simultaneous, precise measurement of these quantities in the combustion chamber of engines for a single nerburn cycle was not possible with sufficient precision. In the method proposed here, a measuring arrangement is presented with which it is possible to simultaneously determine the remaining exhaust gas content, the amount of air supplied, the amount of fuel supplied and the power output (via the pressure curve) for a single nerburn cycle with such great precision that the influence the cyclical fluctuation of the gas composition on the power output can be determined. It is also possible to record other quantities simultaneously.

It is important and patent-relevant for the process that this measurement is carried out with sufficient precision (around 1%) so that the influence of the small cyclical changes in the gas composition on the performance can be recorded. For example, The influence of the exhaust gas fraction, the stoichiometry - or other genes, such as the pressure curve of the previous cycle - on the power output are measured. Cycles with lean stoichiometry and a lot of exhaust gas lead e.g. to a reduced power output. It is important and patent-relevant that the influence of the measured variables on the power output can be differentiated by measurement - due to the high precision of the method.

REPLACEMENT LEAF The measurements can also be averaged over different cycles. In this case - without knowing the individual fluctuations - it can be determined at which average values of stoichiometry and exhaust gas content the engine works. When varying engine conditions, the influence, for example, of the influence of different injection variants, the load or the ignition timing on the gas mixture before ignition and the power output can be determined.

The measurement of the amounts of air, fuel and the amount of the remaining exhaust gas fraction is carried out according to the patented method using optical measurement technology.

One of the methods is Raman scattering. It is well known that Raman scattering can be used to measure the density of majority species (Eckbreth: Laser Diagnostics for Combustion: Temperature and Species. AEGupta, DG Liley eds. Vol. 7 of Energy and Engineering Sciences, Abacus Press Cambridge, MA 1988). However, due to the relatively weak intensity of the Raman scattering and the simultaneous occurrence of other light phenomena caused by the laser, it is difficult (I) to obtain enough signal to distinguish the small cyclical fluctuations and (II) to distinguish the Raman emission from the other emissions. It is well known that the use of Raman scattering - e.g. in the combustion of hydrocarbons - is significantly restricted in accuracy by these other emissions (see Eckbreth).

Regarding (I): The generation of a sufficiently large signal is achieved in that the Raman scattering is carried out with intense pulsed lasers in the deep UV. It is well known that the intensity of the Raman scatter increases with the fourth order of frequency and is therefore particularly intense in the deep UV.

Regarding (II): The distinction from other lighting phenomena is achieved, among other things, by utilizing the polarization of the Raman scattering. Since the other lighting phenomena caused by the laser are mostly unpolarized, the background U is obtained by setting the electrical vector of the laser light in the direction of observation - in simple terms - and by adjusting the electrical vector of the laser light perpendicular to the direction of observation the sum S _r + U is obtained from the Raman signal S _r and background U and can then determine the Raman signal separately by forming the difference S _r = (S _r + U) - U. The background can be determined in an analogous manner - without rotating the electrical vector of the laser light - by analyzing the emission in the direction of observation for polarization components (for example using polarization filters). Again the one delivers

REPLACEMENT LEAF Direction of polarization the background and the other direction of polarization signal + background, so that the Raman signal can be determined separately. The advantage of the latter variant is that both components can be determined in one and the same measuring process. It is therefore possible to separate the Raman scattering from the other interfering light phenomena and to selectively determine the densities of the gas components, ie without interfering external emissions.

The analysis of polarized emissions is important and patent-relevant, since it allows the differentiation of the emissions and thus a selective measurement of the individual gas components in gas mixtures with a complex composition. The high intensity of the Raman scattering is important and relevant to the patent because it makes the measuring precision so high that the cyclical fluctuations in the gas composition that actually occur can be measured.

In addition to the Raman scattering, other optical measurement techniques for determining the particle densities of fuel, air and residual gas can be used to generate characteristic maps, such as Absorption or laser induced emissions or FTIR.

In absorption spectroscopy, the density of the particles - integrated over locations along the absorption light path - in the combustion chamber is determined by the decrease in the incident light. The "direct" absorption by the naturally present molecules or the absorption by tracer can be used.

For direct absorption, the water density (for determining the residual gas content) and the fuel density e.g. can be determined via IR absorption. The oxygen content can e.g. can be determined via the absorption in the deep UV (e.g. Schumann Runge bands). Another absorption technique results from the selective addition of substances ("tracers"). A tracer (e.g. acetone) can be added to the fuel and the density of the tracer (e.g. the acetone density) can be determined via the absorption of the tracer and the fuel density can be determined. Accordingly, other absorbent molecules can be added to the air to measure the air density. The residual gas content in the fresh charge can also be determined by injecting a tracer into the combustion chamber at the time of gas discharge (e.g. in the area of top dead center). The stoichiometry and the residual gas content can in turn be determined from the data on fuel, air and residual gas by forming the ratio.

The particle densities can also be determined by adding tracers that can be excited to glow with lasers (laser-induced emissions). So you can add different tracers to the fuel, air and residual gas and select the excitation wavelength of the laser and the emission wavelengths so that one

REPLACEMENT LEAF spectral separation of the emissions is possible and thus the components are detected simultaneously. The density of fuel, air and residual gas can then be determined from the intensity of the laser-induced emissions, and the stoichiometry and the residual gas content can be determined by forming the ratio.

The creation of maps and the use of the are essential and patent-relevant

Maps to characterize and optimize the way motors work. From the

Measurement data about the particle densities and possibly other sizes (pressure, temperature,

Pollutants) can be determined under which conditions the engine works and whether

Improvements can be made. So you get e.g. by applying the

Lambda values (from the relative intensity of O2 to fuel, see below) on the X axis, the

Exhaust gas quantity (via the quantity of water and O2) on the Y axis and the associated one

Power output on the Z-axis shows the dependence of the power output on stoichiometry and the proportion of exhaust gas. This comparison of cause and effect gives a direct one

Regulation for an optimization, according to which e.g. an injection system can be specifically modified and it can also be checked whether a specific change has the desired effect.

The simultaneous acquisition of the quantities [θ ₂ ], [N ₂ ], [H ₂ θ] and [f] ([x] is the density of the

Particle type x, f = fuel) before a combustion cycle after compression together with the pressure curve of the previous and current combustion cycle provides e.g. following information:

1. Stoichiometry of the gas mixture

2. Exhaust gas fraction from the previous cycle in the gas mixture

3. Air intake

4. The amount of fuel injected

5. the performance achieved with the gas composition

6. the influence of the previous cycle on gas exchange

The amount of exhaust gas in the engine combustion chamber from the previous cycle can be determined from the ratio of H2O / N2 (since the amount of water in the air supplied is negligible) or from the ratio O2 N2.

The simultaneous, precise acquisition of various measured variables makes it possible to determine causal relationships in individual combustion cycles and thereby characterize the operating conditions of the engine and use the knowledge as a basis for optimization. The simultaneous measurement of the stoichiometry and the proportion of exhaust gas in a certain measurement volume before the ignition simultaneously with the recording of the pressure curve for the current combustion cycle may be mentioned as an example. This

REPLACEMENT LEAF Measurement makes it possible to link the cause (mixture composition before ignition) with the effect (power output). By simultaneously measuring other variables (e.g. NO density via LIF, temperatures, ...) in the same cycle, further statements about cause and effect can be made. It is also essential for the present invention that the small fluctuations in the mixture composition can be detected precisely enough before ignition.

Ntersuchungen results of on ^u

Figure 1 shows two emission spectra from the combustion engine. If the electrical vector of the laser light is in the direction of observation (see Fig. 4), the background caused by external emissions becomes visible. If the electrical vector of the laser light is perpendicular to the direction of observation, the emission caused by the Raman scattering and the background becomes noticeable. The difference (put simply) is the Raman signal. The picture clearly shows that the polarization of the Raman scattering can be used to determine the background. In the quantitative interpretation, it must be taken into account that depolarized parts of the Raman scattering also occur. The results of the preliminary investigations in Fig. 2 and 3 show that it is possible - for a given operating condition of the engine - to measure the small fluctuations in stoichiometry mentioned in single shot. In the tests with a VW transparent engine, e.g. out that the fluctuation range of the stoichiometry in the range of ΔΛ =

0.2 lies, i.e. e.g. that with an externally set Λ number of 1.0, the measured Λ number fluctuates between 0.9 and 1.1. These fluctuations are shown in Fig. 2. The fact that these fluctuations are measured precisely enough to determine the influence of the fluctuation of the gas mixture on the performance can be seen from the pressure measured simultaneously: for the cycles with rich mixtures, the maximum pressure is high and fluctuates little, for the lean cycles it is Maximum pressure is lower and fluctuates considerably more. Fig. 3 shows the results of measurements which, in addition to stoichiometry, also show the additional influence of the exhaust gas content. The picture shows the maximum pressure reached as a function of the stoichiometry and the proportion of exhaust gas and is referred to here as a map. Obviously, larger proportions of exhaust gas lead to lower pressures and thus less power output.

The map, which is here for the way the Mc> works, depends on the way the engine works. It can be used to differentiate between different ways of working.

In particular, the pressure curve of the previous cycle also provides important information e.g. about the history of gas exchange in the combustion chamber. If e.g. the pressure in the previous cycle was high, you can find it in the following cycle - for a certain operating condition -

REPLACEMENT LEAF less fuel. This is explained by the pressure that has not yet been completely reduced during the fuel inlet. The correlation between the previous pressure curve and the fuel content therefore provides indirect information about the gas exchange in the combustion chamber. The absolute N2 quantity provides information about the total amount of air admitted and thus about the throttle valve position, which regulates the pressure in the intake manifold. In this way - via the gas exchange characterized by the majority densities - indirect conclusions about the gas dynamic processes are not only obtained in the combustion chamber.

The knowledge that emerges from these results is essential: by precisely measuring the stoichiometry and the proportion of exhaust gas, it is possible to record a map (Fig. 3) that allows the engine's power output to be optimized so that the pressure - as a function of the two important parameters stoichiometry and exhaust gas content - is maximum. This should e.g. happen via the variation of the injection or other relevant influencing parameters.

The measurements can also be averaged over different cycles if the signal in the single shot does not provide enough intensity. In these measurements, the small fluctuations in the stoichiometry and in the exhaust gas are eliminated, so that in a representation as in Fig. 3 the fluctuations become much smaller and reproducible characteristic data are obtained. One can then examine as a function of other engine parameters in which area of this map the engine works, i.e. Cover a specific area for each operating condition in Fig.3.

Measurements with Raman and Rayleigh scattering.

Due to the Raman emissions of nitrogen [N ₂ ] _> oxygen [O ₂ ], water [H ₂ O] and fuel (eg [CgHig]), which occur separately at different wavelengths, it is possible to measure the intensity of these emissions simultaneously. The basis of the methods is completely known and also that shocks at high pressures play a subordinate role. The following describes briefly the basis on which the data are evaluated here.

The following applies: Ij _ram = Sσj _ram [i] E ₀

with the energy EQ of the laser pulse, the Raman section σi _ram for particles i and the partial density [i] of the particles i. S is the sensitivity of the measuring system for the detection of particles i. S is in principle to be determined separately for each particle type and each location,

REPLACEMENT LEAF is assumed to be constant here for the sake of simplicity. The following applies to the relative measurement of the densities of the particles i, j:

li ram "j ram ⁼ Si'Sj σj _ra / σj _ram [i] / D] ie, from the relative intensity for the two components - without the laser: ~ gie or eg the detection sensitivity - the ratio of the densities [ i] / [j] Since the stoichiometry results from the relative intensity of O2 to fuel, the stoichiometry must be recorded with greater precision than the individual components.

Furthermore, it is possible to measure the intensity of the Rayleigh scatter at the same time and - in connection with the Raman scattering - to obtain information about the total density (= sum of the partial densities of the majority species).

The following applies:

ray ^{= s E} 0 ^∑ i ^σ i ray P] and because

P] ⁼ Α ram / ^σ i ram ^E 0

Iray ⁼ S ∑j (σj _ra y σi ram) - * i ram

It is therefore possible to evaluate the Rayleigh scattering over the sum of the Raman intensities. The laser energy does not drop out here because it is implicitly contained in the Raman intensity. It is well known that Rayleigh scattering can be used to determine the temperature via the gas law - in conjunction with the pressure at the time of the measurement, which is known from the measured pressure curve. Rayleigh scattering is also polarized and its intensity increases with the fourth power of the frequency. Since the Rayleigh scattering is much more intense than the Raman scattering, it provides a high measuring accuracy for the entire density or the temperature.

In this way, the partial densities of fuel, air and water can be determined separately in connection with calibration measurements. In addition, the temperature can be determined via combined measurement of the pressure curve and the Rayleigh scatter.

A specific arrangement for measuring is shown in Fig. 4 as one example from many. In this case, the intense pulsed UV laser beam (1) is shot through the combustion chamber of an engine. For this purpose, UV-transmitting windows (2) are used in the upper part. In addition to the combustion chamber with the spark plug (3) and the inlet valve (4), part of the intake manifold with the injection valve (5) is also shown in cross section. From a section

REPLACEMENT LEAF With a finite length from the combustion chamber along the laser beam, the scattered Raman light is directed via a window in the bulb (6) and a deflecting mirror (7) to the spatially resolving optical multi-channel analyzer (8). This is described in more detail in Fig.5. An imaging optical system (9) images the laser beam axis (16) on the slit (10) of a spectrograph. This usually consists of several mirrors (11) and a dispersion grating (12). In the spectrograph, the different emissions are separated by wavelength (along the axis (14)) and detected with an intensified, short-opening CCD camera (13). The structure also enables a local assignment of the Raman emission along the section from the combustion chamber along the axis (15). For example, the Rochon prism can be used for separation to determine the polarization of the different emissions. Except for the polarization analysis, the arrangement described is usually referred to with the keyword "locally resolved optical multi-channel analyzer".

Fig.6 shows another realization of the measuring technique, the backscattering. Here, the optical access to the motor is only realized through small windows (17) on the side. One of the windows (17) through which the laser exits the combustion chamber in Fig. 6 can optionally be dispensed with. The laser (1) is coupled into the combustion chamber via a dichroic mirror (18). The Raman emission - and also other emissions - are coupled out via one of the windows on the side (17) and in turn detected using an optical multi-channel analyzer (see Fig. 5) - and a polarization analyzer (8). In this case, (somewhat) local resolution may also be obtained in the medium if the window is extended in one direction.

The measurement can also be carried out with other spectral filters and other polarization analyzers.

REPLACEMENT LEAF

Claims

Expectations

1 method and measuring arrangement for the characterization and optimization of internal combustion engines by measuring the gas composition in the combustion chamber and possibly other sizes, dad "ch characterized that several particle densities are measured by optical measurement technology, and possibly additional gas pressure and temperature, and that from these measured variables and Correlations derived from this are created in order to reveal possible causal relationships of engine combustion.

2 The method according to claim 1, characterized in that the particle densities and possibly pressure and temperature are measured simultaneously in individual cycles.

3 The method according to claim 1, characterized in that important quantities such as the air ratio (e.g. from concentrations of oxygen to fuel density) and / or the residual gas content (e.g. from concentrations vc wet to nitrogen density) are determined by the combination of simultaneously measured particle densities.

4 procedures n ,. . Claim 1, characterized in that the simultaneous measurement of the particle densities takes place by means of laser-induced emissions (e.g. Raman scattering or LIF from selectively added tracers) or absorption spectroscopy in UV, VtS or IR (possibly with selectively added tracers) or FTIR.

5 The method according to claim 1, characterized in that by determining the air ratio and / or residual gas content from the particle densities these quantities can be determined so precisely that the small cyclical fluctuations of these quantities can be resolved during operation of the internal combustion engine.

6 The method according to claim 1, characterized in that the laser Raman scattering with an intensive laser in the UV and an optical multi-channel analyzer and a wavelength-selective component (eg spectrometer) and possibly polarization properties (possibly by simultaneous measurement of both polarization components) are used to suppress interfering emissions .

7. The method according to claim 1, characterized in that the measurement of the particle densities with spatial resolution (e.g. along an axis) or averaged over local areas or at individual special locations (e.g. on the spark plug) are carried out.

8. The method according to claim 1, characterized in that mean values and / or fluctuation ranges of the measured variables and / or the quantities derived therefrom are determined from the individual cycle measurements or the measurements are carried out averaged over cycles from the outset.

9. The method according to claim 1, characterized in that the causal relationships are identified via maps by the application of e.g. Stoichiometry and residual gas against the subsequent performance or by plotting the

REPLACEMENT LEAF Performance in the pre-cycle on the subsequent charge composition ('maps') and that these maps, possibly as a function of the engine parameters, are used in order to achieve an optimization (eg of the injection system) based on the resulting knowledge.

10 The method according to claim 1, characterized in that other measured variables (e.g. temperature, pollutants such as NO or HC) may be later measured in the same cycle and used to further characterize the operating conditions.

11. The method according to claim 1, characterized in that in addition to the Raman scattering, the Rayleigh scattering is also measured in order to determine the overall density (and thus the temperature).

12 The method according to claim 1, characterized in that the laser-induced emission in backscatter is observed with or without local resolution.

REPLACEMENT LEAF