Fuel injection controlling method and system

Abstract

The invention relates to a method of controlling fuel injection in an internal combustion engine operating on gaseous fuel including: a fuel tank (10); a plurality of fuel injectors (1, 2, 3, 4); a pressure regulator (11 ); a shut-off valve (13) to isolate the fuel rail (12) from the fuel tank (10), thereby defining an isolated section downstream of the shut-off valve. A test routine is operated that comprises, during at least one first testing cycle:. a shutting-off step (401 ) for closing the shut-off valve (13) during a predetermined testing duration;. an injector actuating step (402) performing a series of N injections events by the injectors (1, 2, 3, 4) during said testing duration to inject a plurality of demanded quantities, the sum of which represents a global fuel quantity;. a pressure and temperature monitoring step (403) whereby the pressure and the temperature within the isolated section (12) are determined at predetermined moments; and. a computing step (404) that calculates the output quantity (OQ) of gaseous fuel as a function of the pressure and temperature measured within the isolated volume during the testing duration.

Description

[0001] FUEL INJECTION CONTROLLING METHOD AND SYSTEM

[0002] FIELD OF THE INVENTION

[0003] The present invention generally relates to a method and a system for controlling quantities of fuel gas, particularly hydrogen, supplied through injectors to an internal combustion engine.

[0004] BACKGROUND OF THE INVENTION

[0005] It is well known within the art that, in internal combustion engines, a crucial function is implemented by injectors through which gaseous fuel is supplied to cylinders enclosing the engine’s combustion chambers.

[0006] More particularly, the quantity of gaseous fuel injected in the engine has a direct impact on the torque developed by the engine, so that an accurate control of the engine relies on an accurate control of the quantity of gaseous fuel actually supplied to its cylinders.

[0007] Hydrogen injectors are typically driven by a solenoid actuator pulling a valve member (pintle) to a maximum stroke. The injected quantity depends on the drive pulse duration and fuel / gas pressure at injector inlet.

[0008] The engine control unit (ECU) typically operated fuel injection using so-called flow curves relating the fuel gas quantity and pulse with (injector on time).

[0009] As is known in the art, fuel injectors of a same design may have different flow behaviors, for example due to part-to-part variability as well as wear.

[0010] As indicated before, engine driveability depends on the control of injected quantity, but also emissions.

[0011] TECHNICAL PROBLEM

[0012] The object of the present invention is to provide an improved method of controlling fuel injection in an internal combustion engine operating on gaseous fuel, which addresses issues of fuel injector deviation. GENERAL DESCRIPTION OF THE INVENTION

[0013] This problem is solved by a method of controlling fuel injection in an internal combustion engine as claimed in claim 1.

[0014] The internal combustion engine comprises a fuel delivery system (SYST) including: . a fuel tank (10) for storing pressurized fuel that feeds a fuel rail (12);

[0015] . a plurality of fuel injectors (1 , 2, 3, 4) coupled to the fuel rail (12);

[0016] . a pressure regulator (11 ) to lower the gaseous fuel pressure in the fuel rail (12); fuel rail; and

[0017] . a shut-off valve (13) to isolate the fuel rail (12) from the fuel tank (10), thereby defining an isolated section downstream of said shut-off valve and including the fuel rail (12).

[0018] According to the present method, a test routine is operated that comprises, during at least one first testing cycle:

[0019] . a shutting-off step (401 ) for closing the shut-off valve (13) during a predetermined testing duration;

[0020] . an injector actuating step (402) performing a series of N injections events by the injectors (1 , 2, 3, 4) during said testing duration to inject a plurality of demand fuel quantities, the sum of which represents a global demand fuel quantity;

[0021] . a pressure and temperature monitoring step (403) whereby the pressure and the temperature within the fuel rail (12) are determined at predetermined moments during the testing duration; and

[0022] . a computing step (404) that calculates the output quantity (OQ) of gaseous fuel as a function of the pressure and temperature measured within the isolated section during the testing duration.

[0023] In the method according to the invention, the shut-off valve delimits an isolated section including the fuel rail when said shut-off valve is closed and the variations in the quantity of gaseous fuel enclosed within this isolated section will represent the output quantity which will have left the isolated section through the injectors during the testing duration, in the absence of any leak in the isolated section. This output quantity can then be compared to the global demand fuel quantity, which represents the sum of the fuel amounts commanded to the fuel injectors, so as to identify and quantify any deviation between both values. The knowledge of such deviation allows applying corrections coefficients to the fuel injectors in order to reduce the deviation.

[0024] According to an embodiment of the invention, the computing step (404) includes a mass difference computing step (434) for calculating the output quantity (OQ) as a difference between the masses (M1 , M2) of gaseous fuel enclosed within the isolated section at the beginning and at the end of the testing duration, said masses being calculated as a function of the pressures and temperatures (P1 , T1 ) (P2, T2) measured within the fuel rail at the beginning and at the end of the testing duration.

[0025] This embodiment of the invention enables to calculate the output quantity of gaseous fuel on the basis of pressure and temperature, which are parameters routinely measured / available in existing systems, and thus does not require the use of additional equipment for its operation.

[0026] In embodiments, the mass (M1 , M2) of the of gaseous fuel enclosed within the isolated section at a given instant (t1 , t2) is calculated as being equal to a product of the volume of the isolated section with a density (D1 , D2) of the gaseous fuel at said given instant (t1 , t2).

[0027] Although the density values to be used in this advantageous embodiment may be derived from existing abacuses / mappings by using the pressure and temperature values (P1 , T1 ) (P2, T2) as coordinates, the inventors have observed that the temperature variation between the beginning and the end of the testing duration may be too swift to be detected by conventional temperature sensors.

[0028] The inventors have thus developed an enhanced version of the first testing cycle, according to which the density (D2) of the gaseous fuel at end of the testing duration (t2) is computed by considering that the pressure and temperature (P, T) of the injection rail evolve according to an adiabatic process (or quasi-adiabatic) between the beginning and the end of the testing duration.

[0029] The density value obtained in this manner will be more accurate than that which may be derived from an existing abacus by using an imprecise temperature measurement.

[0030] According to a first variant of the invention, the computing step (404) includes a temperature calculating step in the course of which the temperature of the fuel rail at the end of the testing duration (t2) is calculated using the perfect gas law and the previously computed density (D2) at the end of the testing duration (t2).

[0031] Having the knowledge of the gaseous fuel temperature value at the end of the testing duration enables performing a correction of the demanded quantity to be taken into account for the calculation of the deviation. Indeed, during nominal operating conditions, the demanded quantity is a function of the measured temperature of the fuel rail. As already explained above, the temperature variation between the beginning and the end of the testing duration may be too swift to be detected, so that the demanded quantity will remain constant although the real value of the temperature of the fuel rail may have changed. In such a situation, the calculation of the deviation will be skewed. A possible correction may for example consist in computing an average value between the temperature measured at the beginning and the calculated temperature at the end of the testing duration, considering that the temperature evolution is linear during this duration.

[0032] According to a second variant of the invention, which may be used separately or cumulatively with the first variant, at least one second testing cycle is performed following each first testing cycle, said second testing cycle including:

[0033] . a shutting-off step (501 ) for closing the shut-off valve during a testing duration;

[0034] . a pressure and temperature monitoring step for determining the pressure and the temperature at predetermined moments within the isolated section during the testing duration;

[0035] . a leakage quantifying step (504) for calculating a leaked quantity (LQ) as a difference in the quantity of gaseous fuel enclosed within the isolated section between the beginning and the end of the testing duration; and

[0036] . a leakage correction step (534) for calculating an actual output quantity (AOQ) of gaseous fuel by subtracting the previously calculated output quantity (OQ) and the leaked quantity (LQ).

[0037] It should be noted that during this second test cycle no injection events are performed. That is, the ECU does not request / program any injection events, or inhibits any requests for injection events.

[0038] For example, the second test cycle can be performed during a foot off period, in particular where there is no torque request from the driver. The second testing cycle enables to calculate a missing quantity of gaseous fuel which will have left the isolated section without passing through the injectors.

[0039] Subtracting this leaked quantity from the output quantity previously calculated during the first testing cycle will enable to obtained an accurate value of the quantity of gaseous fuel actually outputted through the injectors, which will then be advantageously used for the purpose of quantifying the deviation with respect to the demanded quantity.

[0040] In this second variant, the second testing cycle will preferably further include a second pressure adjusting step (500) during which the pressure measured within the fuel rail is set at the predetermined test value (P1 ) by the pressure regulator before the shutting-off step (501 ).

[0041] The first testing cycle, optionally followed by a second testing cycle, may be performed only once, or, preferably, several times, each iteration being performed at a specific predetermined operating pressure, in order to obtain several deviation values to be averaged in view of allowing optimal demand value compensation over the whole operating pressure range of the engine.

[0042] Generally speaking, the testing duration and the injector actuating signal (103) will preferably be configured in such a way that a total number N of injection events occurring during the testing duration is a multiple of the number of injectors connected to the fuel rail. Thereby, all injectors are operated in turn, a same number of times. Such a configuration will allow to easily average over each injector the total deviation value which will be calculated by the testing system according to the invention.

[0043] According to another aspect, the invention also relates to a fuel delivery system for delivering fuel to an internal combustion engine operating on gaseous fuel as claim in claim X.

[0044] To sum up, the invention provides a method of controlling fuel injection and related system which is designed to compare fuel quantities determined based, on the one hand, on fuel demand quantities relates commanding injector actuation and, on the other side, based on measured pressure variations in the fuel rail. A deviation between the demand / command quantities and actual quantities can thus be determined, and correction factors can thus be developed, to be used in fuel control strategies.

[0045] Brief Description of the Drawings

[0046] Further details and advantages of the present invention will be apparent from the following detailed description of several not limiting embodiments with reference to the attached drawings, wherein:

[0047] Fig. 1 is a schematic diagram of a fuel delivery system comprising a controller configured to operate an embodiment of the present method;

[0048] Fig. 2 is a graph illustrating the variation of pressure in the isolated section in an embodiment of the present test routine applied to the engine system of Fig.1 ;

[0049] Fig. 3 is a schematic diagram representing a leakage detection test;

[0050] Fig. 4 is a schematic diagram of a first testing cycle included in a method according to an advantageous embodiment of the invention;

[0051] Fig. 5 is a schematic diagram of a second testing cycle included in a method according to a second variant of the invention.

[0052] Description of Preferred Embodiments

[0053] Fig. 1 illustrates a fuel delivery system (SYST) for delivering fuel to an internal combustion engine operating on gaseous fuel, the system (SYST) comprising:

[0054] . a fuel tank (10) for storing pressurized fuel that feeds a fuel rail (12);

[0055] . a plurality of fuel injectors (1 , 2, 3, 4) coupled to the fuel rail (12);

[0056] . a pressure regulator (11 ), e.g an electronic regulator, to regulate the gaseous fuel pressure in the fuel rail (12) within a predetermined range; and

[0057] . a shut-off valve (13) to isolate the fuel rail (12) from the fuel tank (10).

[0058] It should be noted here that Fig.1 is a simplified diagram; a fuel delivery system (SYST) as used herein may include various additional elements that are not represented here in order to simplify the description (e.g. a circuit to fill the tank, filters, Pressure Relief Valves, Purge valves, etc.). Furthermore, although the engine includes four injectors in the present example, in other embodiments, the number of injectors may vary, e.g. ranging from three to eight. Such fuel delivery system, and its components, may be generally known and will not be further discussed. Typically, a controlling module is provided to operate the fuel injectors and the pressure regulator according to predetermined fuel strategies.

[0059] The controlling module may generally include a processor and a memory containing instructions to operate the delivery system. The controlling module can be partly or entirely integrated into the Engine Control Unit.

[0060] It will be appreciated that in system SYST, the controlling module (100) is configured to operate a test routine that comprises, during at least one first test cycle:

[0061] . generate a shutting-off signal (102) for closing the shut-off valve (13) during a predetermined testing duration, thereby defining an isolated section downstream of said shut-off valve and including the fuel rail (12);

[0062] . operating the injectors (1 , 2, 3, 4) to perform a series of N injections events during said testing duration, to inject a plurality of demand fuel quantities into the engine;

[0063] . measure the pressure and the temperature within the isolated section (12) at predetermined moments during the testing duration; and

[0064] . calculate the output quantity (OQ) of gaseous fuel exiting the isolated section as a function of the pressure and temperature measured within the isolated section during the testing duration.

[0065] Fig. 2 illustrates the operation of a system shown in the previous figure during a first testing cycle having a testing duration starting at a first instant t1 , where the fuel rail pressure is equal to a first value P1 , and ending at a second instant t2, where the fuel rail pressure is equal to a second value P2.

[0066] The above-described shut-off valve 13 is closed between the first and second instants (t1 , t2). Before t1 and after t2 the rail pressure is regulated at a predetermined level. Regulation is typically also stopped during the test cycle.

[0067] During the test period, a number N of injection events are performed. That is, fuel is injected through the injectors into the engine, and the fuel rail is thus emptied, whereby the second value P2 of the rail pressure is lower than its first value P1 . The square wave signal 103 at the bottom of Fig.2 represents the injectors command sequence, each on-time period corresponding to actuation of one injector, i.e. injector event. Typically, the engine combustion cycle follows a predetermined firing order, and the injectors are thus actuated in the same order (alternately, one after another). The duration of each injector event (with of the on-time pulse) is the so- called pulse width, which determines the injected fuel quantity.

[0068] As can be seen, during the test period, a total number of 12 injection events are performed, i.e. N=12. Since the engine has 4 cylinders, each injector will typically have performed three injection events during the test period.

[0069] In practice, each injector event is performed to inject an individual, demand fuel quantity Qdem, whereby a corresponding pulse width PW is determined from injector flow mappings (relating Qdem, Pressure and PW).

[0070] Accordingly, the resulting amount of fuel injected during the test period may be referred to as global (or cumulative) demand fuel quantity, noted QG, and represents the sum of all Qdem corresponding to the performed injection events.

[0071] In embodiments, as shown in Fig.2, the same fuel demand Qdem is used for all injection events, accordingly we have: QG=N*Qdem.

[0072] In other situations, Qdem could be variable, in which case QG may be computed as:

[0073] In the present method, the shut-off valve delimits an isolated section including the fuel rail when said shut-off valve is closed, whereby variations in the quantity of gaseous fuel enclosed within this isolated section will represent the output quantity which will have left the isolated section through the injectors during the testing duration (t2-t1 ), in the absence of any leak in the isolated section.

[0074] This output quantity can be compared to the global demand quantity QG as ordered by the controlling module, so as to identify and quantify any deviation between both values. Based on the deviation between these two parameters, injection control strategies may be adapted to improve the convergence between the demand quantities and the actually injected quantity. In particular, based injector correction factor(s) may be developed based on such deviation.

[0075] The testing duration (t2-t1 ) and the injector actuating signal (103) will advantageously be configured in such a way that a total number N of control pulses occurring during said testing duration (t2-t1 ) is a multiple of the total number of injectors to be driven by said pulses (and involves all injectors), which will in turn ensure that the number N of injection events to be performed during said testing duration (t2-t1 ) is a multiple of the number of injectors connected to the fuel rail. Such a configuration will allow to easily average over each injector the total deviation value which will be calculated by the testing system according to the invention.

[0076] Fig. 3 illustrates the operation of a system shown above during a second testing cycle starting at a third instant t3, where the fuel rail pressure is equal to a third value P3 which should preferably be equal to the first value R1 , and ending at a fourth instant t-4, where the fuel rail pressure is equal to a fourth value P4, and having a same duration (t4-t3) as the duration (t2-t1 ) of the first testing style shown in the previous figure. During the second testing cycle, the shut-off valve is closed between third and fourth instants (t3, t-4) , but the injector actuating signal (103) doesn’t feature any control pulse over the whole testing duration (t4-t3) . That is, no injection events are performed during the second testing cycle, whereby any pressure drop between the third and fourth values (P3, P4) of the rail pressure during said second testing cycle will be due to leaks, the magnitude of which can be quantified during the second testing cycle as will be explained hereinafter. This second testing cycle can be performed during a foot-off phase, where there is no torque demand from the driver or engine systems.

[0077] The second testing cycle enables to calculate a missing quantity of gaseous fuel which will have left the isolated section without passing through the injectors due to leakage in this section or calculate an additional / excess gas quantity resulting from a leaky shut off valve upstream of this section.

[0078] Subtracting / adding this leaked / excess quantity from / to the output quantity previously calculated during the first testing cycle will enable to obtain an accurate value of the quantity of gaseous fuel actually outputted through the injectors, which may then be advantageously used for the purpose of quantifying the deviation with respect to the demanded quantity.

[0079] It should be noted here that the first testing cycle, optionally followed by a second testing cycle, may be performed only once, or, preferably, several times, each iteration being performed at a specific predetermined operating pressure, in order to obtain several deviation values to be averaged in view of allowing optimal demand value compensation over the whole operating pressure range of the engine.

[0080] Fig. 4 illustrates a first testing cycle included in a method according to an advantageous embodiment of the invention.

[0081] This first testing cycle includes a first pressure adjusting step (400) during which the pressure measured within the fuel rail is set (through pressure regulator 11 ) at a predetermined test value (P1 ) before the shut-off valve is closed in a first shutting- off step (401 ), said predetermined test value (P1 ) being chosen so that an average value of the measured pressure of the fuel rail over the testing duration will be substantially equal to a first known usage pressure.

[0082] The first testing cycle depicted here further includes:

[0083] . an injector actuating step (402) by which a series of N injection events are performed by the injectors (1 , 2, 3, 4). Accordingly, the controlling module operates N injection events by sending individual control pulses of predetermined pulse width PW to cause the injection of corresponding demand fuel quantities Qdem. Thus, during the test cycle a cumulative amount QG is injected, which is the sum of the individual demand fuel quantities Qdem;

[0084] . a pressure and temperature monitoring step (403) for determining the pressure and the temperature within the fuel rail (12) at the beginning and end of the testing duration; and

[0085] . a computing step (404) for calculating the output quantity (OQ) of gaseous fuel as a function of the pressure and temperature measured within the testing duration.

[0086] The shut-off valve delimiting an isolated section including the fuel rail when said shut-off valve is closed, the computing step (404) includes a mass difference computing step (434) for calculating the output quantity (OQ) as a difference between the masses (M1 , M2) of gaseous fuel enclosed within the isolated section at the beginning and at the end of the testing duration, said masses being calculated as a function of the pressures and temperatures (P1 , T1 ) (P2, T2) measured within the fuel rail at the beginning and at the end of the testing duration:

[0087] OQ=M1 -M2 During a mass computation step (424), the mass (M1 , M2) of the of gaseous fuel enclosed within the isolated section at a given instant (t1 , t2) is calculated as being equal to a product of the volume V of the isolated section and a density (D1 , D2) of the gaseous fuel at said given instant (t1 , t2):

[0088] M1 =V.D1 and M2=V.D2, so that the output quantity (OQ) can be written as:

[0089] 0Q=M1 -M2=V.D1 -V.D2=V.(D1 -D2)

[0090] Although the density values to be used in this advantageous embodiment may be derived from existing abacuses by using the pressure and temperature values (P1 , T1 ) (P2, T2) as coordinates, the inventors have observed that the temperature variation between the beginning and the end of the testing duration may be too swift to be detected by conventional temperature sensors due to their inertia.

[0091] The inventors have thus developed an enhanced version of the first testing cycle, according to which the density (D2) of the gaseous fuel at end of the testing duration (t2) is computed during a density computation step (414) by considering that the pressure and temperature (P, T) of the injection rail evolve according to an adiabatic process between the beginning and the end of the testing duration.

[0092] Accordingly, P2 / (D2)Y=P1 / (D1 )Y, where y is known as the adiabatic coefficient of Laplace and essentially equal to 1 .4 for the purpose of the present calculation.

[0093] The density value (D2) obtained in this manner will be more accurate than that which may be derived from an existing abacus by using an imprecise temperature measurement. This allows a more precise calculation of the output quantity OQ as follows:

[0094] OQ = V.D1.[1 -(P2 / P1 )(1 / Y)]

[0095] In a deviation computation step (444), the deviation DEV between the total demand quantity QG and the output quantity can then be calculated as:

[0096] DEV = QG - V.D1.[1 -(P2 / P1 )(1 / Y)]

[0097] As already explained above, the temperature variation between the beginning and the end of the testing duration may be too swift to be detected, so that the demanded quantity will remain constant although the real value of the temperature of the fuel rail will have changed between the first and second instants (t1 , t2). In such a situation, the calculation of the deviation DEV will be skewed. A possible correction may for example consist in computing an average value between the temperature (T1 ) measured at the first instant (t1 ) and a calculated temperature (T2) at the end of the testing duration (t2), considering that the temperature evolution is linear during this duration.

[0098] In such a first variant of the invention, the method described above may advantageously include a temperature calculating step (not shown) in the course of which the temperature of the fuel rail at the second instant (t2) is calculated using the perfect gas law and the previously computed density (D2) at the end of the testing duration (t2), which gives:

[0099] P1 / (D1 ,T1 ) = P2 / (D2.T2), and thus

[0100] T2 = (D1.T1.P2) / (D2.P1 )

[0101] A new, corrected demanded quantity value Qcor to be used for the calculation of the deviation value DEV can then be defined as

[0102] Qcor = QG.Comp(Tavgd) / Comp(T1 ) where Comp(T) is a correction factor depending of the temperature T and Tavgd is the average temperature value between the first and second instants (t1 , t2), which can be written as:

[0103] Tavgd = (T1 +T2) / 2 = [T1 +(D1.T1.P2) / (D2.P1 )] / 2, where P1 , P2 and T1 have been measured during the pressure and temperature monitoring step (403) and D1 and D2 have been calculated during the density computation step (414).

[0104] Fig. 5 illustrates a second testing cycle included in a method according to an advantageous embodiment of the invention.

[0105] This second testing cycle includes a second pressure adjustment step (500), a shutting-off step (501 ), a pressure and temperature monitoring (503) and a second computation step (504) similar to the first pressure adjustment step (400), first shutting-off step (401 ), first pressure and temperature monitoring step (403) and first computation step (404) discussed hereinbefore.

[0106] As explained above, the injector actuating signal doesn’t feature any control pulse over the whole testing duration (t4-t3). No gaseous fuel delivery being thus allowed through the injectors during the second testing style, any pressure drop between the third and fourth values (P3, P4) of the rail pressure during said second testing cycle will be due to leaks. The second testing cycle described here enables to calculate a quantity variation of gaseous fuel resulting from gas (i) having left, i.e. leaked out of, the isolated section without passing through the injectors; and / or (ii) having been added to the isolated section due to leaky shut off valve.

[0107] Subtracting or adding this leaked quantity from the output quantity previously calculated during the first testing cycle will enable to obtain an accurate value of the quantity of gaseous fuel actually outputted through the injectors, which will then be advantageously used for the purpose of quantifying the deviation with respect to the demanded quantity.

[0108] To this end, as already explained hereinbefore, the leaked / added quantity LQ will be calculated during a second mass difference computing step (534) as a difference between the mass values (M3, M4) of the gaseous fuel enclosed within the isolated section at the third and fourth instants (t3, t4), respectively, computed during a second mass value computation step (524) on the basis of respective third and fourth density values (D3, D4) calculated during a second density computation step (514).

[0109] During a second subtracting step (554), the leaked / added quantity (LQ) is then subtracted from the output quantity (OQ) previously calculated during the first testing cycle, which gives an actual output quantity (AOQ) to be used in a second deviation computation step (544), where the deviation DEV between the demanded quantity Qdem or, optionally the corrected demanded quantity Qcor, and the output quantity can then be calculated.

Claims

Claims1 . A method of controlling fuel injection in an internal combustion engine operating on gaseous fuel, said internal combustion engine comprising a fuel delivery system (SYST) including:. a fuel tank (10) for storing pressurized fuel that feeds a fuel rail (12);. a plurality of fuel injectors (1 , 2, 3, 4) coupled to the fuel rail (12);. a pressure regulator (11 ) to regulate the gaseous fuel pressure in the fuel rail (12);. a shut-off valve (13) to selectively isolate the fuel rail (12) from the fuel tank (10), thereby defining an isolated section downstream of said shut-off valve and including the fuel rail (12), wherein a test routine is operated that comprises, during at least one first testing cycle:. a shutting-off step (401 ) for closing the shut-off valve (13) during a predetermined testing duration;. an injector actuating step (402) performing a series of N injections events by the injectors (1 , 2, 3, 4) during said testing duration to inject a plurality of demanded quantities, the sum of which represents a global fuel quantity;. a pressure and temperature monitoring step (403) whereby the pressure and the temperature within the isolated section (12) are determined at predetermined moments; and. a computing step (404) that calculates the output quantity (OQ) of gaseous fuel as a function of the pressure and temperature measured within the isolated volume during the testing duration.

2. The method as claimed in claim 1 , wherein, the computing step (404) includes a mass difference computing step (434) for calculating the output quantity (OQ) as a difference between the masses (M1 , M2) of gaseous fuel enclosed within the isolated section at the beginning and at the end of the testing duration, said masses being calculated as a function of the pressures and temperatures (P1 , T 1 ) (P2, T2) measured within the fuel rail at the beginning and at the end of the testing duration.

3. The method as claimed in claim 2, wherein the mass (M1 , M2) of the of gaseous fuel enclosed within the isolated section at a given instant (t1 , t2) is calculated as being equal to a product of a volume of the isolated section with a density (D1 , D2) of the gaseous fuel at said given instant (t1 , t2).

4. The method as claimed in claim 3, wherein the density (D2) of the gaseous fuel at the end of the testing duration (t2) is computed by considering that the pressure and temperature (P, T) of the fuel rail (12) evolve according to an adiabatic process between the beginning and the end of the testing duration.

5. The method as claimed in claim 5, wherein the computing step (404) includes a temperature calculating step in the course of which the temperature of the fuel rail (12) at the second instant (t2) is calculated using the perfect gas law and the previously computed density (D2) at the end of the testing duration (t2).

6. The method as claimed in any one of claims 1 to 5, wherein the method includes a deviation calculating step (444), whereby an output quantity of gaseous fuel is subtracted to a demanded quantity defined by an injection controlling module as a function of the temperature of the fuel rail.

7. The method as claimed in claim 6, wherein the method includes a demanded quantity compensating step for adjusting said demanded quantity by taking into account the fuel rail’s temperature variation between the beginning and the end of the testing duration.

8. The method as claimed in any one of claims 1 to 7, wherein the first testing cycle further includes a first pressure adjusting step (400) during which the pressure measured within the fuel rail is set at a predetermined test value (P1 ) by the pressure regulator before the shutting-off step, said predetermined test value (P1 ) being chosen so that an average value of the measured pressure of the fuel rail over the testing duration will be substantially equal to a first known usage pressure.

9. The method as claimed in claims 1 to 8, wherein the method includes at least one second testing cycle following each first testing cycle, said second testing cycle including:. a shutting-off step (501 ) for closing the shut-off valve during a testing duration; . a pressure and temperature monitoring step for measuring the pressure and the temperature within the fuel rail during the testing duration;. a leakage quantifying step (504) for calculating a leaked quantity (LQ) as a difference in the quantity of gaseous fuel enclosed within the isolated section between the beginning and the end of the testing duration; and. an-leakage correction step (534) for calculating an actual output quantity (AOQ) of gaseous fuel equal to a subtraction between the previously calculated output quantity (OQ) and the leaked quantity (LQ).

10. The method as claimed in claim 9, wherein the second testing cycle further includes a second pressure adjusting step (500) during which the pressure measured within the fuel rail is set at the predetermined test value (P1 ) by the pressure regulator before the shutting-off step (501 ).

11. The method as claimed in any one claims 1 to 10, wherein the method includes a plurality of sequences of first testing cycles, the quantity of fuel ordered for each injection during a given first testing cycle being different than those ordered for each injection during other first testing cycles.

12. The method as claimed in any one of claims 9 or 10, wherein the method includes a plurality of sequences of pairs of first and second testing cycles, the quantity of fuel ordered for each injection during a given first testing cycle being different than those ordered for each injection during other pairs of first and second testing cycles.

13. The method as claimed in any one of claims 1 to 12, wherein the number N of injections to be performed during each testing duration is a multiple of the number of injectors connected to the fuel rail.

14. A fuel delivery system (SYST) for delivering fuel to an internal combustion engine operating on gaseous fuel, said system (SYST) comprising:. a fuel tank (10) for storing pressurized fuel that feeds a fuel rail (12);. a plurality of fuel injectors (1 , 2, 3, 4) coupled to the fuel rail (12);. a pressure regulator (11 ) to regulate the gaseous fuel pressure in the fuel rail (12); and a shut-off valve (13) to configured to selectively isolate the fuel rail (12) from the fuel tank (10), thereby defining an isolated section downstream of said shut-off valve and including the fuel rail (12); characterized by a controlling module (100) that is configured to implement the method as claimed in any of the preceding claims.

Images