EP1889024A1

EP1889024A1 - Method for determining an injection diesel engine combustion chamber noise level

Info

Publication number: EP1889024A1
Application number: EP06764845A
Authority: EP
Inventors: Jean-Pierre Chemisky
Original assignee: Peugeot Citroen Automobiles SA
Current assignee: PSA Automobiles SA
Priority date: 2005-06-07
Filing date: 2006-06-06
Publication date: 2008-02-20
Also published as: FR2886729B1; FR2886729A1; WO2006131673A1

Abstract

The invention concerns a method for determining combustion noise level (D3Ppil1, D3Ppal1) generated by an injection engine based on pressure measurements. Said method consists in: equipping the engine with a pressure sensor delivering a signal representing the pressure in at least one combustion chamber; applying a derivation process to the signal derived from the pressure sensor to form a signal (GP1) representing a pressure gradient in said combustion chamber; determining said noise level by evaluating the difference (D3Ppil1, D3Ppal1) between the value of the gradient at the local maximum (Mpil1, Mpal1) following an injection, and the gradient value during said injection (Ipil1, Ipal1). The invention is applicable to injection diesel engines.

Description

METHOD FOR DETERMINING THE COMBUSTION NOISE LEVEL OF AN INJECTION DIESEL ENGINE

The present invention claims the priority of the French application 0505772 filed on 07/06/2005 whose content (description, claims, drawings) is incorporated herein by reference.

The invention relates to a method for determining a combustion noise level value generated by an injection heat engine such as a diesel engine.

During the development of a combustion engine, the evaluation of the combustion noise level is carried out using anechoic bench tests. These tests are carried out by operating the engine according to different operating conditions, for example at different speeds and at different load levels for each regime. An expert panel listens to this engine to give it a note representative of the noise level, for each operating condition tested. The use of an anechoic bench and the need for a jury of experts, however, is an expensive and inflexible approach. In addition to the anechoic bench tests, therefore, tests based on instantaneous pressure measurements in the combustion chambers of the engine are also used to evaluate the level of combustion noise generated by the engine.

These additional tests include collecting curves of the pressure in one or more combustion chamber during several engine cycles, according to different operating conditions. The pressure curves obtained are averaged and derived to form representative curves of the pressure gradient during a cycle, and this, for each operating condition. The maximum value of the pressure gradient during a cycle is then determined for each tested operating condition. The value thus determined is the noise level of the engine, that is to say that it has a certain correlation with the scores given by the jury during the tests on anechoic bench. A calibration curve can thus be established on the basis of a first test on an anechoic bench during which pressure measurements are carried out in parallel with the rating established by the jury, for each operating condition tested. This calibration curve is intended to determine, from the maximum value of the pressure gradient during a cycle, the corresponding score that would theoretically have been awarded by the jury.

It is thus possible to modify the design of the engine or to change injection parameters to evaluate, thanks to new pressure measurements, notes representative of the noise level of this engine, without having to proceed to another bench test. anechoic with expert panel. More generally, the determination of the combustion engine noise level consists in using a method based on pressure measurements, which has the best possible correlation with anechoic bench tests with a panel of experts , to reduce the costs of developing an engine.

For this purpose, the subject of the invention is a method for determining a level of combustion noise generated by an injection-engine, consisting of: - equipping this engine with at least one sensor delivering a signal representative of the pressure in at least one minus a combustion chamber;

applying a bypass processing to the signal coming from the sensor to form a signal representative of the pressure gradient in this combustion chamber;

- determine the noise level by applying a process of identification of the difference between the value of the gradient when this gradient reaches a local maximum subsequent to an injection, and the value of the gradient at the instant of this injection.

According to one characteristic of the invention, the local maximum subsequent to the injection is detected by canceling the derivative of the pressure gradient.

According to another characteristic of the invention, the method is applied to a motor including several injections during each combustion cycle, and it consists in determining the noise level on the basis of the difference in gradient induced by each injection, and in retaining as the level of combustion noise, the maximum value among all the differences.

According to another characteristic of the invention, the method is applied to a motor including a so-called pilot injection and a so-called main injection during each combustion cycle, of determining the noise level on the basis of the gradient difference induced by the pilot injection and on the basis of the gradient difference induced by the main injection, and to retain as the level of combustion noise, the maximum value among these two differences.

According to another characteristic of the invention, the process for determining the derivative of the pressure determines the derivative of the pressure with respect to an angle of rotation of the engine.

The invention also relates to a method for determining, in an injection engine, injection parameters leading to a noise level lower than a predetermined value, including a step of determining the combustion noise according to one of the preceding claims.

The invention will now be described in more detail and with reference to the accompanying drawings which illustrate an embodiment thereof by way of non-limiting example. Figure 1 is a diagram giving a first pressure gradient curve of a combustion chamber;

FIG. 2 is a diagram giving a second pressure gradient curve of a combustion chamber;

Fig. 3 is a diagram showing expert panel notes as a function of the gradient maximum values according to the state of the art; Fig. 4 is a diagram showing expert panel scores as a function of differences in gradient values according to the invention;

Figure 5 is a motor map based on the gradient maximum according to the state of the art; Figure 6 is a motor map based on gradient differences according to the invention.

FIGS. 1 and 2 each comprise a diagram representative of the pressure gradient in a combustion chamber of a motor, each diagram including a curve, respectively GP1, GP2 giving the value of the pressure gradient in the ordinate, as a function of the angle rotation of the motor on the abscissa.

These curves are representative of the pressure gradient during a time interval beginning slightly before an injection and ending after this injection, the abscissa being graduated between -

50 and +50 degrees of rotation angle of the engine, the value of 0 degrees corresponding to a reference point of a combustion cycle, namely a top dead center of high pressure loop of the engine cycle.

The pressure gradient values result, for example, from signals from one or more pressure sensors in one or more combustion chambers, a filtering and bypass treatment having been applied to these signals, either in real time or after acquisition. It is advantageously mean curves, ie resulting for example from pressure values measured over one hundred engine cycles, these values being derived from different combustion chambers. The GPl and GP2 curves corresponding to two different engines, or to different operating conditions of the same engine.

These two curves GP1 and GP2 correspond to multiple injection engines, in which an engine cycle comprises on the one hand a pilot injection marked respectively by Ipill and Ipil2, followed by a main injection respectively marked by Ipall and Ipal2.

The pilot injection is an injection of a small amount of fuel made a few milliseconds before the main injection. It reduces the pressure gradient in the combustion chamber, which has the effect of reducing the combustion noise.

As seen in these two diagrams, each injection, whether pilot or main, is followed by a significant increase in the pressure gradient, and then a decrease in this gradient. Each injection is thus followed by a local maximum of the pressure gradient, the local maximums following the injections Ipill, Ipall, Ipil2 and Ipal2 being respectively marked by Mpill, Mpall, Mpil2 and Mpal2.

According to the invention, the level of combustion noise is evaluated by determining a difference in values of the pressure gradient, referred to as the pressure derivative of the premix peak, denoted D3P. This premixing peak D3P is the difference, for an injection, between the value of the pressure gradient at the local maximum following this injection and the pressure gradient value at the instant of the injection. D3P is thus an indicator of the violence of self-inflamation.

It has indeed been found that this difference in gradient values has an excellent correlation with the marks awarded by an expert jury on an anechoic test bench.

The injection taken into account to determine this difference in gradient values may be one or the other of the injections that occur in a cycle of a multiple injection engine. Several injections can also be taken into account to determine the noise level.

In the diagrams of Figure 1 and Figure 2, D3Ppill, D3Ppall, and D3Ppil2, D3Ppal2, have been identified as the pressure gradient difference induced by the Ipill, Ipall, Ipil2, and Ipal2 injections, respectively.

In the case of the diagram of FIG. 1, the noise level value D3Ppall for the main Ipall injection is greater than that of the D3Ppill noise level for the pilot injection. However, in the case of the diagram of Figure 2, it is the opposite, and the noise level D3Ppil2 for the pilot injection, has a greater value than the noise level D3Ppal2 corresponding to the main injection.

Advantageously, the level of combustion noise selected is the maximum value among the gradient differences induced by each injection considered. In the case of a motor comprising a pilot injection and a main injection as illustrated on the diagrams, it is therefore the maximum between D3Ppill and D3Ppall for the diagram of FIG. 1, and the maximum between D3Ppil2 and D3Ppal2 in FIG. the case of the diagram in Figure 2. The diagrams in Figures 3 and 4 show the relationship between measurements from values of combustion pressure and the corresponding ratings given by an expert panel on anechoic bench. They give rating values assigned by an expert jury in ordinate, as a function of pressure gradient values on the abscissa. A high score of the jury corresponds to a quiet engine, that is to say in which the pressure gradients in the combustion chamber have low values. In the case of FIG. 3, the pressure gradient value considered is that of the state of the art, namely the maximum of the pressure gradient over one cycle. In the case of FIG. 4, the gradient value considered is the gradient difference induced by an injection, according to the invention.

The diagram of FIG. 3 comprises a first calibration curve marked by EtA, and a second curve marked by QpilA. Each of these two curves is the result of tests on anechoic bench during which an engine has been tested under different operating conditions.

With respect to the EtA curve, each tested operating condition is represented by a rhombic point, with the abscissa as the value of the pressure gradient maximum, and for ordinate the score given by the expert panel. The EtA curve itself is a median curve of the various points thus determined. The different operating conditions have been obtained here by varying a first adjustment parameter. This parameter is for example a value of injection pressure, also called injection rail pressure, or a quantity of recycled exhaust gas, also known as EGR.

(Exhausted Gas Recycling). The second curve, identified by QpilA, is representative of another test carried out on this same engine, but in which the different operating conditions were obtained by varying a parameter other than that which was used to establish the EtA curve. In the present case, this other parameter is the quantity of fuel injected during the pilot injection. Similarly, each condition of tested operation corresponds to a point marked by a triangle giving the assigned rating and the value of the maximum pressure gradient resulting from the measurements.

As seen in Fig. 3, the QpilA curve is disjunct from the EtA calibration curve, particularly with respect to a set of injected fuel amount values that correspond to one end of the QpilA curve. Thus, the correlation between the EtA calibration curve and the QpilA curve is insufficient to reliably estimate the level of combustion noise based on pressure measurements alone.

Indeed, the QpilA curve comprises for example a triangular point corresponding to 6.5 as the maximum pressure gradient, and to 4 as a note actually given by the panel of experts. If one relied on the EtA calibration curve to predict the score given by the jury, this curve gives a score of 2.5 for a 6.5 gradient, which would be an error of more than 30% on the evaluation of the jury's score. As in the diagram of FIG. 3, the diagram of FIG. 4 comprises a calibration curve EtB established from an anechoic bench test during which several operating conditions, for example several engine control parameters, have been determined. have been tested. For each operating condition, the value of the pressure gradient difference subsequent to an injection according to the invention was noted, and the score actually assigned by the panel. A set of points, identified by QpilB and represented in the form of lozenges is representative of another test in which the different operating conditions result from the modification of a motor parameter other than the parameter which has been modified to establish the curve. EtB calibration. This other parameter is the amount of fuel injected during the pilot injection. As can be seen in this FIG. 4, the QpilB points are very close to the EtB curve, which is indicative of a very good correlation of the criterion according to the invention with the expert panel's scores with regard to a scanning of values. quantities of fuel injected during the pilot injection.

In particular, the improvement of the noise level, for a large quantity of fuel injected by the pilot injection, is perceptible in the diagram of FIG. 4, in accordance with the invention, whereas it is not visible in the diagram. of Figure 3 which corresponds to the state of the art. In fact, in the diagram of FIG. 3 the curves EtA and QpilA are completely disjoint for high values of quantity of fuel injected during the pilot injection whereas they are substantially identical for small quantities of fuel injected during pilot injections. .

It is thus possible to reliably determine the level of combustion noise generated by the engine on the basis of the pressure gradient difference resulting from an injection, in accordance with the invention, since the QpilB points are substantially merged with the EtB calibration curve. Figures 5 and 6 are two diagrams giving a map of an engine, i.e. pressure gradient values for different operating conditions. Each diagram includes on the abscissa the engine speed in revolutions per minute, and on the ordinate the load applied to the engine, expressed as the average effective pressure in bar.

Each diagram comprises different areas marked by the letters a to p which correspond to pressure gradient values, according to the correspondence table below which indicates these pressure gradients, in bar per degree of rotation angle of the motor. ab C defgh

7, 5 - 8 7 - 7, 5 6, 5 - 7 6 - 6, 5 5, 5 - 6 5 - 5, 5 4, 5 - 5 4 - 4, 5

i j k 1 m n O P

3, 5 - 4 3 - 3, 5 2, 5 - 3 2 - 2, 5 1, 5 - 2 1 - i, 5 o, 5 - 1 0 - o, 5

In the case of the diagram of FIG. 5, the pressure gradient values are the maximum of the pressure gradient during a cycle, according to the state of the art. In the case of the diagram of FIG. 6, the values taken into account are the difference in pressure gradient subsequent to an injection, according to the invention.

The mapping of FIG. 5 illustrates the result of choice of parameters, based on the criterion of the state of the art, and after reworking the parameters by ear on the vehicle. This did not make it possible, in particular, to reach sufficiently low noise levels in certain regions, for example for regimes below 3000 rpm for load levels higher than 9, where the evaluated noise level is greater than 3, 5.

The implementation of the method according to the invention is intended to overcome this step of reworking the parameters to the ear on the vehicle.

The mapping of FIG. 6, based on pressure differences in accordance with the invention, illustrates the result of choosing injection parameters made on the basis of the noise level evaluation method according to the invention.

This map shows on the one hand an area delimited symbolically by a square double line, corresponding to regimes below 3000 rpm for a level of load less than 9 bar effective mean pressure, and which is an area in which the criterion according to the invention made it possible to adjust the quantity of pilot injection fuel to respect a compromise between noise level and pollution level, the noise level remaining below 3.

It also shows an area delimited by a frame in broken lines, which corresponds to a regime located between 2000 and 3000 rpm for a level of load greater than 9, in which the noise level is less than 2.5. , the fuel quantity of the pilot injection having been increased to reduce the noise level.

On the other hand, it is visible that beyond 3000 rpm, D3P becomes relatively high, and at least higher than 2.5 for all load levels. This corresponds to an area in which, for the case considered, the injection technology used does not allow to maintain a pilot injection beyond 3000 rpm.

Claims

A method for determining a combustion noise level (D3Ppill, D3Ppall, D3Ppil2, D3Ppal2) generated by an injection heat engine, comprising:

- Equip the engine with at least one sensor delivering a signal representative of the pressure in at least one combustion chamber;

- Apply a bypass processing to the signal from the sensor to form a signal (GP1, GP2) representative of the pressure gradient in the combustion chamber;

- determine the noise level by applying a difference identification treatment (D3Ppill, D3Ppall, D3Ppil2, D3Ppal2) between the gradient value when this gradient reaches a local maximum (Mpill, Mpall, Mpil2, Mpal2) subsequent to an injection, and the value of the gradient at the instant of this injection (Ipill, Ipall, Ipil2, Ipal2).

2. Method according to claim 1, wherein the local maximum (Mpill, Mpall, Mpil2, Mpal2) subsequent to the injection (Ipill, Ipall, Ipil2, Ipal2) is detected by canceling the derivative of the pressure gradient (GPl, GP2).

A method according to claim 1 or 2, applied to a motor including several injections (Ipill, Ipil2, Ipall, Ipal2) during each combustion cycle, of determining the noise level on the basis of the gradient difference induced by each Injection (Ipill, Ipil2), and to retain as combustion noise level (D3Ppill, D3Ppall, D3Ppil2, D3Ppal2), the maximum value among all the differences.

4. Method according to one of claims 1 to 3, applied to a motor including a so-called pilot injection (Ipill, Ipil2) and a so-called main injection (Ipall, Ipal2) during each combustion cycle, consisting of determining the noise level based on the gradient difference induced by pilot injection

(Ipill, Ipil2) and on the basis of the gradient difference induced by the main injection (Ipall,

Ipal2), and to remember as the level of combustion noise (D3Ppill, D3Ppall, D3Ppil2, D3Ppal2), the maximum value among these two differences.

5. Method according to one of the preceding claims, wherein the process for determining the derivative of the pressure determines the derivative of the pressure (GP1, GP2) with respect to a rotation angle of the engine.

6. Method for determining in a fuel injection engine injection parameters leading to a noise level lower than a predetermined value, including a step of determining the combustion noise according to one of the preceding claims.