Method for operating a fuel injector of an engine
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- PHINIA DELPHI LUXEMBOURG SARL
- Filing Date
- 2024-08-08
- Publication Date
- 2026-07-08
AI Technical Summary
The existing methods for operating fuel injectors in engines face challenges in optimizing the opening process, particularly in ensuring precise timing of the acceleration, deceleration, and retention phases to prevent excessive impact and maintain reliable injector operation.
A method for operating a fuel injector that involves a calibration process to optimize the acceleration duration by detecting the lift-stop feature and adjusting the current acceleration duration based on the comparison with the projected retention-start time, ensuring accurate timing of the injector phases.
This method enables individual calibration of the acceleration duration for specific fuel injectors, improving the accuracy and reliability of the injector opening process, thereby optimizing engine performance.
Smart Images
Figure EP2024072498_06032025_PF_FP_ABST
Abstract
Description
METHOD FOR OPERATING A FUEL INJECTOR OF AN ENGINETechnical Field
[0001] The invention relates to a method for operating a fuel injector of an engine, and to an injection system.Background Art
[0002] Fuel injectors are used in combustion engines to inject fuel e.g., into a runner of an air intake manifold ahead of a cylinder intake valve or directly into the combustion chamber of an engine cylinder. According to one known design, a pintle is disposed within an injector housing. The pintle is movable between a closed position, in which it (or a ball that is fixed to the pintle) closes a nozzle at one end of the injector housing, and an open position, in which it is moved away from the nozzle, thereby enabling fuel injection. The pintle is moved by an armature, which in turn is moved by a magnetic field generated by a magnetic coil. According to a common design, the magnetic coil magnetizes a pole piece, which then attracts the armature against the force of a return spring. While at the beginning of the opening stroke, it is desirable to accelerate the armature (and thus the pintle) in order to quickly open the injector, it is detrimental if the armature reaches its end position at an excessive speed. This would result in a severe impact between the armature and the pole piece.
[0003] It has therefore been proposed to initially apply a coil current that leads to an acceleration and then reduce or even completely cut the coil current so that the armature decelerates due to friction and the action of the return spring. Thus, the impact can be mitigated. As soon as the armature has reached the pole piece, though, it is desired to reliably keep it in its position, thereby keeping the injector fully open. Accordingly, the coil current is increased again to retain the armature. The timing of the three phases (acceleration, deceleration, and retention) is critical for optimum operation of the injector. Specifically, if the coil current is increased too early for the retention phase, the armature is accelerated again, which increases the impact force. If the coil current is increased too late, the armature may rebound from the pole piece before it can be retained in contact therewith. However, the proper timing will depend on various factors and may specifically be different not only for each type of injector but also for each individual injector of a certain type.Technical Problem
[0004] It is thus an object of the present invention to optimize the opening process of a fuel injector.
[0005] This problem is solved by a method of operating a fuel injector according to claim 1 and by an injection system according to claim 15.General Description of the Invention
[0006] The invention provides a method for operating a fuel injector of an engine (or a method of controlling injection in an engine). The engine is of course an internal combustion engine; the method being particularly applicable to hydrogen combustion engines (H2-ICE), but also to gasoline and diesel engines. Correspondingly, the fuel injector injects fuel, e.g., gaseous fuel such as hydrogen or liquid fuel such as gasoline or diesel, either directly into a combustion chamber or into an intake duct or the like, from where an air-fuel mixture is introduced into the engine.
[0007] The fuel injector comprises a housing extending axially along an injector axis from a proximal end to a distal end and having a nozzle at the distal end. One main function of the housing is to contain and guide fuel before it is ejected from the injector. Usually, the housing comprises a plurality of pieces or components that are stationarily connected with each other. The housing has a nozzle for ejecting the fuel, which nozzle is disposed at a distal end. The terms "distal" as well as "proximal" refer to the general flow direction of the fuel within the injector towards the distal end. In general, the distal end is the end of the injector that is closer to the nozzle and the proximal end is the end that is further away. The injector extends along an injector axis from the proximal end to the distal end. At least some parts of the injector can be symmetric with respect to the injector axis, but in general this injector axis only defines a reference frame, whereby an axial direction, a radial direction and a tangential direction are implicitly defined.
[0008] The injector further comprises a pintle being axially movable between an open position and a closed position in which it closes the nozzle. The pintle may have an axially extending pintle shaft that is normally cylindrical and elongate, with a length of the pintle shaft corresponding to e.g., more than 10 times its diameter.
[0009] In embodiments, the injector is designed such that the pintle opens outwardly. The pintle has a pintle head (typically radially protruding relative to the pintle shaft) that cooperates with an outwardly facing valve seat at the end of the gas / fuel passage. Such design may be particularly used for gas / hydrogen injectors.
[0010] In some embodiments, a ball is fixed to a distal end of the pintle. The ball may also be considered as a part of the pintle. In the closed position, the pintle (or the ball, respectively) closes the nozzle and prevents fuel from being ejected. In a typical embodiment, the ball engages a nozzle seat at the distal end of the housing, thereby closing the nozzle. The pintle can be moved axially to an open position in which the nozzle is open, and fuel can be ejected.
[0011] In case of an inward opening injector, the open position is a proximal position, and the closed position is a distal position. In case of an outward opening injector, the open position is a distal position, and the closed position is a proximal position.
[0012] The fuel injector further comprises an armature that is axially movable between a passive position and an active position and is adapted to move the pintle to the open position. The armature is movable along the injector axis between the passive position and the active position. Here and the following, "along the injector axis" particularly, but not exclusively, means "parallel to the injector axis". More generally, it means "at least partially in the direction of the injector axis". The armature, like the pintle, is disposed inside the housing. As the armature moves towards the active position, it moves the pintle towards the open position. As it moves towards the passive position, the pintle can move or is moved to the closed position. In case of an inward opening injector, the active position is a proximal position, and the passive position is a distal position. In case of an outward opening injector, the active position is a distal position, and the passive position is a proximal position. Preferably, the armature is moved to the passive position by the force of a return spring.
[0013] Also, the fuel injector comprises a magnetic coil for magnetically attracting the armature towards the active position and being operable by a coil current. Preferably, the magnetic coil magnetizes a pole piece which enhances the magnetic field to attract the armature. In such a configuration, the pole piece is disposed in the direction of the active position, i.e., proximal of the armature in caseof an inward opening injector. Either way, the armature is attracted to the active position when the magnetic coil is activated by a coil current of a sufficient magnitude.
[0014] The fuel injector is operated to perform injection events. During a regular injection (event), a profile of the coil current comprises an acceleration phase with a regular acceleration duration during which the coil current is adapted to accelerate the armature towards the active position, a deceleration phase during which the coil current is decreased to decelerate the armature, and a retention phase during which the coil current is increased to keep the armature in the active position. A “regular injection” is an injection that is performed during normal operation of the engine and is designed to achieve optimum performance. During a “regular injection”, the fuel quantity to be injected (QD) is typically defined by maps in function of a torque demand (TD) and calibrated in function of engine parameters (inter alia, e.g. engine speed, temperature, boost pressure, etc. ). For such a regular injection, the profile of the coil current has at least three phases. In a first phase, the acceleration phase, the coil current is strong enough to generate a magnetic field that moves the armature from the passive position and accelerates it towards the active position. The duration of the acceleration phase is herein referred to as the “acceleration duration”. Specifically, the duration used for the regular injection is the “regular acceleration duration”. This duration, like other characteristics of the profile, can be stored in a (preferably non-volatile) memory. In a second phase, the deceleration phase, the coil current is reduced so that the resulting magnetic field is insufficient to accelerate the armature. In particular, the coil current can be reduced to zero during the deceleration phase. Due to friction, a counteracting spring force, and / or other effects, the armature is decelerated during this phase. In a third phase, the retention phase, the coil current is increased with respect to the deceleration phase. The resulting magnetic field is strong enough to retain the armature in the active position, e.g., in contact with the pole piece. The coil current could be constant in each of the three phases but may also vary in at least one phase. It will be understood that the entire profile comprises at least one additional phase, during which the coil current is decreased again to release the armature from the active position. However, the further structure of the profile is not relevant for the present invention and therefore will not be discussed further.
[0015] The method comprises performing at least one calibration process, which comprises defining a current acceleration duration based on the regular acceleration duration and performing at least one calibration iteration. As will be understood from the following, the calibration process is a learning process, which serves to optimize the acceleration duration, i.e., to find an acceleration duration that is optimal for this specific fuel injector. At the start of the calibration process, a current acceleration duration is defined. The definition is based on the regular acceleration duration. For instance, the current acceleration duration may differ from the regular acceleration duration by a specific offset. With the current acceleration duration defined, at least one calibration iteration of the calibration process is performed. Preferably, each calibration iteration is performed during one engine cycle, so that the number of calibration iterations corresponds to the same number of engine cycles. The calibration iteration comprises at least the following steps, which may be performed in the sequence in which they are mentioned. However, it is possible that at least some steps may be performed simultaneously or in a different sequence.
[0016] In one step, a first calibration injection is performed by performing the acceleration phase using the current acceleration duration and afterwards reducing the coil current to decelerate the armature. The first calibration injection uses a profile that is similar to the regular injection in that it also comprises the acceleration phase. However, it employs the current acceleration duration, which normally differs from the regular acceleration duration used for the regular injection. After the acceleration phase, the coil current is reduced, preferably in the same way as in the deceleration phase of the regular injection. However, for the first calibration injection, there is preferably no retention phase. Accordingly, if the armature reaches the active position, it is not retained there but is allowed to return to the passive position.
[0017] In another step, a measurement is performed to detect a lift-stop feature indicating that the armature has reached the active position. The term “liftstop” hints to the fact that the armature moving towards the active position coincides with the pintle moving towards the open position, i.e., being “lifted” from the valve seat, and reaching the fully open position. The details of the lift-stop feature depend on various factors, in particular on the type of measurement that is performed.Generally, the lift-stop feature can be any detectable event or quantity that indicates that the armature has reached the active position. It should be noted that it is possible that no lift-stop feature is detected, e.g., because the armature does not reach the active position (i.e. a ballistic lift that does not fully open).
[0018] Another step comprises, if the lift-stop feature is detected at a lift-stop time, comparing the lift-stop time with a projected retention-start time of the retention phase based on the current acceleration duration, and changing the current acceleration duration depending on the comparison. The lift-stop time is the time at which the lift-stop feature occurs. In practice, the lift stop may generally be detected in a predetermined time window starting when the acceleration duration is sufficient for the needle to reach the stop and finishing when the end of the acceleration phase reaches the lift stop event. The upper / lower bounds of the detection window may be adapted depending on the application. The lift-stop time is compared with a projected retention-start time of the retention phase. As mentioned above, the first calibration injection preferably does not comprise a retention phase, but the retention-start time can be determined with respect to the beginning of the acceleration phase. Insofar, the retention-start time is projected based on the current acceleration duration. This is the time at which the retention phase would start if the acceleration phase, having the current acceleration duration, was followed by the deceleration phase and the retention phase. Of course, the retention-start time also implies a deceleration duration of the deceleration phase, in that the retention-start time is the sum of the current acceleration duration and the deceleration duration. One could also say that the lift-stop time is compared with the projected retention-start time, which is the sum of the current acceleration duration and the deceleration duration.
[0019] If the lift-stop time is greater than the retention-start time, this would correspond to a scenario the armature reaches the active position (e.g., contacts the pole piece) after the retention phase has started. In a regular injection, the armature would be accelerated again, which is undesirable. If the lift-stop time is smaller than the retention-start time, this would correspond to the armature reaching the active position before the retention phase starts. In this case, the armature could rebound before the coil current increase can generate the necessary magnetic field to retain the armature. Accordingly, the current acceleration duration is adaptedbased on the comparison of the retention-start time and the lift-stop time. In general, this adaption may refer to an increase and / or a decrease, while a variety of possible strategies is conceivable. Some of these will be discussed below.
[0020] If a convergence criterion related to the current acceleration duration is fulfilled in one calibration iteration, the regular acceleration duration is re-defined based on the current acceleration duration and the calibration process is ended, and otherwise, a next calibration iteration is started. There are various possibilities for defining a convergence criterion. The convergence criterion either explicitly or implicitly depends on the current acceleration duration. If this convergence criterion is fulfilled, it means that the current acceleration duration is close enough to the optimum value. Also, the lift-stop time and the retention-start time are close enough together to end the calibration process. The regular acceleration duration is redefined based on the current acceleration duration of the current calibration iteration. In the simplest case, it could be defined to be identical to the current acceleration duration, but there are other, usually more adequate definitions.
[0021] The inventive method enables an individual calibration of the acceleration duration for a specific injector. As will become apparent in the following, the first calibration injection with the simplified current profile may facilitate detection of the lift-stop feature. Thus, the dedicated calibration process leads to a high degree of accuracy and reliability. The calibration can be performed for various operating points of the engine, wherein each operating point can be characterized by an engine speed and an engine torque.
[0022] Preferably, each calibration iteration comprises, if the lift-stop time is greater than the retention-start time, increasing the current acceleration duration. If the lift-stop time is greater than the retention-start time, this would correspond to a scenario the armature reaches the active position (e.g., contacts the pole piece) after the retention phase has started. In a regular injection, the armature would be accelerated again, which is undesirable. Accordingly, the test acceleration duration is increased, which also increases the retention-start time. Alternatively or additionally, each calibration iteration may comprise, if the lift-stop time is smaller than the retention-start time, reducing the current acceleration duration. This would correspond to the armature reaching the active position before the retention phase starts. In this case, the armature could rebound before the coil current increase cangenerate the necessary magnetic field to retain the armature. In this case, the current acceleration duration can be decreased, which also decreases the retentionstart time. There are various possible strategies how the current acceleration duration can be increased and / or decreased, some of which will be discussed below.
[0023] It is highly preferred that the calibration process comprises defining a demand injection quantity, and each calibration iteration comprises, if a first injection quantity of the first calibration injection is smaller than the demand injection quantity, performing a second calibration injection with a second injection quantity. The demand injection quantity is the total fuel quantity that the fuel injector needs to inject during one engine cycle. This will usually depend on the operation point of the engine. During each calibration iteration, if a first injection quantity of the first calibration injection is smaller than the demand injection quantity, a second calibration injection is performed with a second injection quantity. This serves to prevent an undersupply of the engine. As a rule, the first injection quantity, which is the fuel quantity delivered by the first calibration injection, is smaller than the demand injection quantity. Thus, the second calibration injection is normally necessary. Preferably, the second injection quantity corresponds to the difference between the demand injection quantity and the first injection quantity or is at least similar to this difference. It will be understood that minor undersupply or oversupply will not change the engine performance significantly, especially since the calibration process may only last for a few engine cycles. Therefore, any impact on engine performance will be limited to a short time interval. In any case the second calibration injection helps to prevent (significant) fuel undersupply to the engine, wherefore this embodiment greatly facilitates performing the calibration process while the engine is in use, e.g., while a vehicle powered by the engine is driving.
[0024] It is possible that in at least one calibration iteration, no lift-stop feature is detected. This may have various reasons, but in particular may be due to the armature not reaching the active position. This, in turn, is mostly due to an insufficient acceleration phase. It is thus preferred that each calibration iteration comprises, if no lift-stop feature is detected, increasing the current acceleration duration. The increase may be the same as for the abovementioned case where thelift-stop time is greater than the acceleration-start time, but it could be different (smaller or larger).
[0025] There are various options how the lift-stop feature can be detected. It would be possible to use a dedicated sensor for monitoring the movement or position of the armature. However, this would complicate the design of the injector and increase its cost. Preferably the coil current and / or the coil voltage is measured to detect the lift-stop feature. In some embodiments, only the coil current or only the coil voltage is measured. In other embodiments, both can be measured to increase the sensitivity or the reliability of the detection. The lift-stop feature can be represented by a discontinuity that is due to a sudden change in the inductance of the magnetic coil when the armature reaches the active position, where it may be stopped by the abovementioned pole piece.
[0026] At the start of the calibration process, the current acceleration duration can be defined as the difference between the regular acceleration duration and a predefined lower offset. Accordingly, the search for the optimal acceleration duration starts at a value below the regular acceleration duration. Thus, if the optimal acceleration duration, which is to be found by the calibration process, is not too far below the regular value, it will be found when the current acceleration duration is increased stepwise. One could say that the lower offset defines a lower limit of a search interval, i.e. , the search starts at the regular acceleration duration minus the lower offset.
[0027] The search for the acceleration duration should be limited by an upper limit, in order to prevent an “endless” search if the lift-stop feature is not detected. Failure to detect the lift-stop feature could be due to a malfunction of one component or due to some other reason. It is possible that the lift-stop feature remains undetected in one calibration process but will be detected in the next calibration process. Preferably, the calibration process is ended if the current acceleration duration is greater than the sum of the regular acceleration duration and a predefined upper offset. The upper offset defines an upper limit of the search interval. Once the current acceleration duration is greater than the regular acceleration duration plus the upper offset, the calibration process ends. It is possible, though, that another calibration process is performed later.
[0028] There are various concepts for increasing the current acceleration duration. Some of these concepts could take into account how great the difference between the lift-stop time and the retention-start time is. E.g., the greater the difference, the greater the increase of the current acceleration duration. According to a simpler and possibly more stable approach, a “constant-step” search is performed. In this embodiment, if the lift-stop time is greater than the retention-start time, the current acceleration duration is increased by a duration increase that is constant for one calibration process. The constant duration increase represents a step size of the search for the best acceleration duration. It should be noted that the duration increase is constant for one calibration process but could be changed for a subsequent calibration process. E.g., smaller steps could be used to determine the acceleration duration more precisely.
[0029] Preferably, the convergence criterion is fulfilled if the retention-start time is greater than the lift-stop time, and the regular acceleration duration is redefined to be between a current acceleration duration of the current calibration iteration and a previous acceleration duration of the previous calibration iteration. It will be understood that both the retention-start time and the lift-stop time depend on the current acceleration duration, so that the convergence criterion is related to the current acceleration duration. The goal of the calibration process is to bring the retention-start time as close as possible to the lift-stop time. In this embodiment, the retention-start time of the current calibration iteration is greater than the lift-stop time. On the other hand, the retention-start time of the previous calibration iteration can be assumed to be smaller than the lift-stop time, because otherwise, the calibration criterion would have been fulfilled in that previous calibration iteration. Accordingly, the optimal acceleration duration can be assumed to be between the acceleration durations of these two calibration iterations, i.e., the current acceleration duration and the previous acceleration duration. The regular acceleration duration can be set accordingly. For instance, it could be the arithmetic mean of the two current acceleration durations, i.e., these are weighted 0,5:0, 5. There are alternative approaches, though. For instance, a weighted average could be employed that accounts for the difference between the lift-stop time and the retention-start time in each calibration iteration. If, for instance, the difference for the previous calibration iteration is three times the difference for the current calibrationiteration, the (current and previous) acceleration durations should be weighted 0,25:0,75, i.e. , the current acceleration duration should have three times the weight of the of the previous acceleration duration.
[0030] In principle, the coil current for the second calibration injection could take various forms. Of course, the minimum requirement is that the profile is effective to open the injector. Also, it is desirable to control the movement of the armature and the pintle so that the second injection quantity can be delivered in a predictable way. One option is that the profile of the coil current for the second calibration injection corresponds to a regular injection that is aborted when the second injection quantity has been reached. In other words, the profile of the regular injection, which is generally considered optimal for the respective injector, is used for the second calibration injection. However, since the regular injection is intended for the demand injection quantity, it has to be modified. Specifically, the normalinjection profile is cut off or chopped when the second calibration quantity is reached. Accordingly, the armature and the pintle will move - or maintain their positions - in the same way as during the regular injection until a certain point, and then they will return to the passive position and the closed position, respectively, ahead of time. Accordingly, the profile for the coil current does not comprise the entire retention phase. Depending on the second calibration quantity, the profile may not comprise the retention phase at all. Also, the deceleration phase may be missing partially or entirely. Even the acceleration phase could only be partially present. If the acceleration phase and / or the deceleration phase are not fully present, this means that the injector is not fully opened during the second calibration injection.
[0031] It would be conceivable to not only vary the acceleration duration, but also the deceleration duration. However, it has been found that a deceleration duration can be found for a specific type of injector that will yield optimal or nearly optimal results for each individual injector of this type, if it is combined with the acceleration duration that is optimal for this injector. In practice, the deceleration duration may be optimized at design stage to reach pintle speed reduction target as well as maintaining robustness to part to part and environment conditions variations. Accordingly, a preferred embodiment provides that the deceleration duration of the deceleration phase is kept constant. In other words, the deceleration duration is the same in each calibration iteration and, if applicable, in each calibration process. Depending on theinjector type, the deceleration duration may be e.g., between 70 ps and 200 ps, but could also be shorter or longer. It will be understood that keeping the deceleration duration constant simplifies the calibration process, since there is only one parameter to be optimized, i.e. , the search is limited to a one-dimensional parameter space.
[0032] One embodiment provides that the acceleration phase comprises a first boost phase during which the coil voltage at least temporarily corresponds to a boost voltage above a battery voltage, and optionally a first normal phase during which the coil voltage is limited by the battery voltage. Although the term “battery voltage” is not limited to this meaning, it preferably corresponds to a voltage supplied by a battery. The boost voltage, on the other hand, is above the battery voltage, possibly even by a factor of 2, 3 or more. The boost voltage may be generated using a boost converter that is generally known in the art. During the first boost phase, the coil current and the magnetic field are maximized to provide a maximum acceleration for the armature. Optionally, the magnetic field can be reduced after an initial part of the acceleration phase, namely during a first normal phase. During this normal phase, the coil voltage may be equal to the battery voltage at least temporarily. However, it may also be below the battery voltage during at least a portion of the first normal phase. The function of the first normal phase is primarily to preserve energy. However, in some cases, it may be more advantageous to omit the first normal phase, so that the first boost phase is directly followed by the deceleration phase. During the deceleration phase, the coil voltage may be inverted and may e.g., correspond to the negative battery voltage.
[0033] In one embodiment, the retention phase comprises a second boost phase during which the coil voltage at least temporarily corresponds to the boost voltage, and at least a second normal phase during which the coil voltage is limited by the battery voltage. Again, the coil voltage may be changed intermittently between the boost voltage and the battery voltage during the second boost phase. The resulting coil current and magnetic field serve to provide an increased retention force for the armature during the first part of the retention phase. Afterwards, the coil voltage and the coil current are reduced to preserve energy. Specifically, the coil voltage may be equal to the battery voltage during the at least one normalphase. It may also be below the battery voltage during at least a portion of at least one second normal phase.
[0034] In some cases, it may be acceptable and sufficient to perform only one calibration process, whereafter the regular acceleration duration for this injector remains unchanged for the lifetime of the injector. However, it may also be beneficial if a plurality of calibration processes is performed, and a plurality of normal injections is performed between each two calibration processes. In other words, each two calibration processes are separated by a time period during which the injector is operated normally, i.e., it only performs normal injections. Specifically, several thousand or even several millions of normal injections may be performed between two calibration processes. There may also be a predefined time interval between two calibration processes, e.g., several days, weeks or months. After the first calibration process, additional calibration processes may be necessary due to properties of the injector changing over its lifetime. As mentioned above, calibration processes may be performed for different operating points. In this case, a plurality of calibration processes can be performed for each operating point.
[0035] The upper offset and the lower offset together define the size of a search interval for the acceleration duration. Under normal circumstances, a first calibration process has to be performed for an individual injector due to e.g., production tolerances. After the first calibration process, it can be assumed that the properties of the injector change only slowly over time, and only to a minor extent. Accordingly, the search interval can be reduced for the following calibration process(es). One embodiment therefore provides that the lower offset and / or the upper offset are reduced after the first calibration process. It is also possible that the lower offset and / or the upper offset are increased if the convergence criterion is not fulfilled in a complete calibration process. This failure could be attributed to a search interval that is too small.
[0036] In one embodiment of the invention, the armature is fixedly connected to the pintle and may even be formed as a single piece therewith. In this case, the passive position corresponds to the closed position and the active position corresponds to the open position. In another embodiment, the armature is separate from the pintle and pushes the pintle into the open position. In other words, as the armature moves towards the active position, it engages the pintle and pushes ittowards the open position. The pintle may comprise a pintle shaft that passes through an opening in the armature. The pintle may be biased towards the closed position by a pintle spring while the armature may be biased towards the passive position by an armature spring. However, it would also be conceivable that a pintle spring biases the pintle towards the closed position and the pintle pushes the armature toward the passive position.
[0037] The invention further provides an injection system according to claim 15. The injection system comprises a fuel injector of an engine, which fuel injector comprises:- a housing extending axially along an injector axis from a proximal end to a distal end and having a nozzle at the distal end,- a pintle being axially movable between an open position and a closed position in which it closes the nozzle,- an armature being axially movable between a passive position and an active position and adapted to move the pintle to the open position, and- a magnetic coil for magnetically attracting the armature towards the active position and being operable by a coil current, wherein the injection system is adapted so that during a regular injection, a profile of the coil current comprises an acceleration phase with a regular acceleration duration during which the coil current is adapted to accelerate the armature towards the active position, a deceleration phase during which the coil current is decreased to decelerate the armature, and a retention phase during which the coil current is increased to keep the armature in the active position.
[0038] The injection system is further adapted to perform at least one calibration process, which comprises defining a current acceleration duration based on the regular acceleration duration, and performing at least one calibration iteration with the following steps:- performing a first calibration injection by performing the acceleration phase using the current acceleration duration and afterwards reducing the coil current to decelerate the armature,- performing a measurement to detect a lift-stop feature indicating that the armature has reached the active position, and- if the lift-stop feature is detected at a lift-stop time, comparing the lift-stop time with a projected retention-start time of the retention phase based on the current acceleration duration, and if the lift-stop time is greater than the retention-start time, increasing the current acceleration duration, and otherwise reducing the current acceleration duration, wherein, if a convergence criterion related to the current acceleration duration is fulfilled in one calibration iteration, the regular acceleration duration is re-defined based on the current acceleration duration and the calibration process is ended, and otherwise, a next calibration iteration is started.
[0039] The injection system may comprise a control device that is adapted to control the coil current and / or the coil voltage during the regular injection and during the calibration process. The control unit may also be adapted to detect the lift-stop feature. The injection system may also comprise an energy source that is at least indirectly connected to the magnetic coil and is adapted to provide the coil voltage and the coil current.
[0040] All other terms have been explained above with respect to the inventive method and therefore will not be explained again. Preferred embodiments of the inventive injection system correspond to those of the inventive method.Brief Description of the Drawings
[0041] Preferred embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:Fig.1 is a schematic view of an inventive injection system;Fig.2 is diagram showing time-dependent quantities during a regular injection;Fig.3A-3B are diagrams showing time-dependent quantities during a calibration iteration;Fig.4 is another diagram showing time-dependent quantities during a calibration iteration; andFig.5 is a flow chart of an inventive method.Description of Preferred Embodiments
[0042] Figs.1 shows an inventive injection system 1 . The system 1 comprises a fuel injector 10, e.g., a for injection of gaseous fuel in an engine of a car. The fuel / gas injector 10, which is greatly simplified in fig.1 , comprises a housing 11 defining an internal gas passage 11.1 , that extends from a proximal end 10.1 to a distal end 10.2 where a nozzle 12 of the fuel injector 10 is disposed. A pintle 13 (having a pintle shaft 13.1 and pintle head 13.2) is disposed within the housing 11 and is movable between a closed position B (as shown in Fig.1 ) and an open position C (shown in dashed lines in fig. 1 ). In the closed position B, the pintle head engages a outwardly facing valve seat 11.2 that surrounds an outlet orifice of the gas passage 11.1 The pintle 13 may be biased towards the closed position B in which it closes the nozzle 12, by a spring element 17. Furthermore, an armature assembly 14, or simply armature, is axially movable within the housing 11 between a passive position D and an active position E (shown in dashed lines in fig. 1 ). When the armature 14 moves towards the active position it, it engages the pintle 13 and moves it towards the open position C. The armature 14 can be moved by a magnetic field that is generated by a magnetic coil 16 (or solenoid) disposed inside the housing 11 . When the coil 16 is activated / energized, it magnetizes a pole piece 15, which attracts the armature 14 to its active position E. The magnetic coil 16 is connected to a control device 20 which can apply a coil voltage you see to the coil 16, thereby generating a coil current lc. The control device 20 is connected to a battery 22 which can provide a battery voltage. However, the control device 20 comprises a boost converter 21 that can generate a boost voltage to temporarily increase the coil current lc.
[0043] Fig. 2 shows a diagram with a regular injection of the fuel injector 10. Apart from the coil current lc, an armature displacement dA of the armature 14 in relation to the passive position D is shown, which also reflects the pintle position. The term “regular injection” is the injection strategy that is applied during standard engine operation (i.e. typically more than 90% of the operating time). The fuel quantity to be injected QD is determined based on current torque demand. Theinjector is then actuated to discharge the fuel quantity QD, which is referred to as injection event. During such regular injection a pulse profile as shown in Fig.2 is applied. The most important phases of the profile of the coil current lc are an acceleration phase PA, during which the armature 14 is accelerated from the passive position D towards the active position E, a deceleration phase PD, during which the coil current is reduced to zero so that the armature 14 is decelerated, and a retention phase PR, during which the coil current is increased in order to retain the armature 14 in the active position E. In this example, the acceleration phase PA comprises a first boost phase PBI , during which a boost voltage is applied, and a first normal phase PNI , during which the coil voltage Uc is limited by the battery voltage. The retention phase PR comprises a second boost phase PB2, during which a boost voltage is applied, a second normal phase PN2 and a third normal phase PN3, during both of which the coil voltage Uc is limited by the battery voltage. After a pause, the coil current lc is advantageously increased for a soft-landing phase PSL in order to dampen the impact of the pintle 13 when it returns to the closed position B.
[0044] In fig. 2, a retention-start time tRs, which is the start of the retention phase PR, is marked as well as a lift-stop time Ls, which is the time at which the armature 14 reaches the active position E and engages the pole piece 15. For a regular injection, the retention start time tRs is the sum of a regular acceleration duration TA,R of the acceleration phase PA and a deceleration duration TD of the deceleration phase PD. In the example of fig. 2, the lift-stop time ti_s is somewhat greater than the retention-start time tRs. Ideally, the lift-stop time ti_s and the retention-start time tRs should be as close together as possible in order to avoid any acceleration of the armature 14 after the deceleration phase PD while at the same time avoiding a possible rebound of the armature 14, further taking into account possible magnetic delay.
[0045] As illustrated by figs. 3A to 3D, it is possible to measure the lift-stop time ti_s by detecting a lift-stop feature LS of the coil voltage Uc. All figures show a current profile that only comprises the acceleration phase PA, whereafter the coil current is reduced to zero. Also, the acceleration phase PA only comprises a boost phase, which is not followed by a normal phase. It has been found that with the full profile shown in fig. 2, it is very difficult to detect the lift-stop feature LS. Fig. 3A shows a situation in which the armature 14 does not reach the active position E butreturns to the passive position D due to an insufficient acceleration phase PA. As can be seen in fig. 3A, the coil voltage Uc increases sharply at the end of the acceleration phase PA and afterwards decreases relatively smoothly. Fig. 3B shows a situation where the armature 14 reaches the active position E. The coil voltage Uc shows a lift-stop feature LS in the form of a discontinuity. Fig. 3C shows a situation where the acceleration phase PA has been increased with respect to fig. 3B, but the lift-stop time ti_s is still after the end of the acceleration phase PA and the discontinuity of the lift-stop feature LS can be clearly detected. Fig. 3D shows a situation where the acceleration phase PA has been further increased, so that the lift-stop time ti_s is shortly before the end of the acceleration phase PA. In this case, the lift-stop feature LS cannot be detected.
[0046] Fig. 5 shows a flowchart of an inventive method for operating the fuel injector 10, or generally of controlling fuel injection in the engine. The method may be implemented by the control device 20 (which may typically comprise a processor adapted to implements code instructions to perform the method). The method can be performed while the engine is operating. After the start, it is determined at 100 whether a calibration of the injector timing needs to be performed. This can be determined depending on various criteria, e.g., whether a calibration of the fuel injector 10 has been performed before and if so, how much operating time and / or engine cycles have passed since the last calibration. Also, the calibration may have to be performed for various operating points of the engine, so it may be decisive if a calibration has been performed before for the current operating point or at least a similar operating point.
[0047] If no calibration is deemed necessary at 100, a regular injection as shown in fig. 2 is performed at step 110 and the method returns to 100. If a calibration has been performed before, it will be understood that several thousand or even several million regular injections may be performed before another calibration is deemed necessary.
[0048] If a calibration is found necessary at 100, the method enters a calibration process 150 which starts at step 160 with a definition of a current acceleration duration TA,C, which is defined as the difference between the regular acceleration duration TA are and a lower offset TLO. Also, a demand injection quantity QD is defined, which is the quantity of fuel that needs to be injected perinjector during one engine cycle at the specific operating point. Then, the method enters a first calibration iteration 200 of the calibration process 150. At 210, a first calibration injection is performed. Similar to figs. 3A to 3D, the profile of the first calibration injection comprises only an acceleration phase PA, whereafter the coil current lc is reduced to zero. An example for such a first calibration injection is shown in the first part of the diagram of fig. 4. The method continues at 220, where it is determined whether a first injection quantity Qi, which is injected during the first calibration injection, is smaller than the demand injection quantity QD. AS a rule, this will be the case since the first calibration injection is rather short. In such a case, a second calibration injection is performed at step 230. The profile of the second calibration injection can be seen in fig. 4. The profile of the coil current lc during the second calibration injection is similar to the regular injection of fig. 2. However, the second calibration injection is ended when a second injection quantity Q2 has been injected that is equal to the difference between the demand injection quantity QD and the first injection quantity Qi (i.e. we have QD = Qi + Q2).
[0049] At 240, it is determined whether the lift-stop feature LS has been detected. It should be noted that this step could also be performed before or during step 220 and / or 230. If there has been a detection, the lift-stop time ti_s is compared (at 250) with a projected retention-start time tps, which is based on the current acceleration duration TA,C. More specifically, the projected retention-start time TRS is the sum of the current acceleration duration TA,C and the deceleration duration TD, which is constant in this example of the method. If the lift-stop time ti_s is greater than the retention-start time tRs, the method continues at step 260, where a previous acceleration duration TA,P is defined as the current acceleration duration TA,C, while the current acceleration duration is increased by a duration increase TINC. In other words, the current acceleration duration TA,C is increased, but its value before the increase is stored for the next calibration iteration as the “previous acceleration duration” TA,P. The method also continues with step 260 if no detection is found in step 240. In a following step, at 270, it is determined whether the current acceleration duration TA,C is smaller than the sum of the regular acceleration duration TA,R and an upper offset Tuo. If so, the method returns to step 210 and enters a next calibration iteration 200. If not, the calibration process 150 is ended and the method returns to step 100. It will be understood that the lower offset TLOand the upper offset Tuo together define a search interval for the acceleration duration.
[0050] If the lift-stop time ti_s is not greater (i.e., normally smaller) than the retention-start time tps, this represents a convergence criterion. Since this is fulfilled, no further calibration iterations are necessary. In a following step 280, the regular acceleration duration TA,R is redefined as the arithmetic mean of the current acceleration duration TA,C and the previous acceleration duration TA,P. Then, the calibration process 150 is ended and the method returns to step 100. Optionally, the lower offset TLO and the upper offset Tuo can be reduced in a step 290.Legend of Reference Numbers:1 injection system 22 battery10 fuel injector A injector axis10.1 proximal end B closed position10.2 distal end C open position11 housing D passive position12 nozzle E active position13 pintle14 armature15 pole piece16 magnetic coil20 control unit21 boost converter
Claims
Claims1. A method for operating a fuel injector (10) of an engine, which fuel injector (10) comprises:- a housing (11 ) extending axially along an injector axis (A) from a proximal end (10.1 ) to a distal end (10.2) and having a nozzle (12) at the distal end (10.2),- a pintle (13) being axially movable between an open position (C) and a closed position (B) in which it closes the nozzle (12),- an armature (14) being axially movable between a passive position (D) and an active position (E) and adapted to move the pintle (13) to the open position (C), and- a magnetic coil (16) for magnetically attracting the armature (14) towards the active position (E) and being operable by a coil current lc, wherein, during a regular injection (110), a profile of the coil current lc comprises an acceleration phase PA with a regular acceleration duration TA,R during which the coil current lc is adapted to accelerate the armature (14) towards the active position (E), a deceleration phase PD during which the coil current lc is decreased to decelerate the armature (14), and a retention phase PR during which the coil current lc is increased to keep the armature (14) in the active position (E), wherein the method comprises performing at least one calibration process (150), which comprises defining a current acceleration duration TA,C based on the regular acceleration duration TA,R and performing at least one calibration iteration (200) with the following steps:- performing (210) a first calibration injection by performing the acceleration phase PA using the current acceleration duration TA,C and afterwards reducing the coil current lc to decelerate the armature (14),- performing a measurement to detect a lift-stop feature LS indicating that the armature (14) has reached the active position (E), and- if the lift-stop feature LS is detected at a lift-stop time tis, comparing the liftstop time ti_s with a projected retention-start time tRs of the retention phase PR based on the current acceleration duration TA,C, and changing (260) the current acceleration duration TA,C depending on the comparison,wherein, if a convergence criterion related to the current acceleration duration TA,C is fulfilled in one calibration iteration (200), the regular acceleration duration TA,R is re-defined (280) based on the current acceleration duration TA,C and the calibration process (150) is ended, and otherwise, a next calibration iteration (200) is started.
2. The method according to claim 1 , wherein each calibration iteration comprises, if the lift-stop time ti_s is greater than the retention-start time tps, increasing (260) the current acceleration duration TA,C.
3. The method according to any one of the preceding claims, wherein the calibration process comprises defining (160) a demand injection quantity QD, and each calibration iteration comprises, if a first injection quantity Qi of the first calibration injection is smaller than the demand injection quantity QD, performing (230) a second calibration injection with a second injection quantity Q2.
4. The method according to any one of the preceding claims, wherein each calibration iteration (200) comprises, if no lift-stop feature LS is detected, increasing (260) the current acceleration duration TA,C.
5. The method according to any one of the preceding claims, wherein the coil current lc and / or the coil voltage Uc is measured to detect the lift-stop feature LS.
6. The method according to any one of the preceding claims, wherein, at the start of the calibration process (150), the current acceleration duration TA,C is defined as the difference between the regular acceleration duration TA,R and a predefined lower offset TLO.
7. The method according to any one of the preceding claims, wherein the calibration process (150) is ended if the current acceleration duration TA,C is greater than the sum of the regular acceleration duration TA,R and a predefined upper offset Tuo.
8. The method according to any one of the preceding claims, wherein, if the lift-stop time ti_s is greater than the retention-start time tRs, the current acceleration durationTA,C is increased (260) by a duration increase TINC that is constant for one calibration process (150).
9. The method according to any one of the preceding claims, wherein the convergence criterion is fulfilled if the lift-stop time ti_s is greater than the retentionstart time tRs, and the regular acceleration duration TA,R is re-defined (280) to be between a current acceleration duration TA,C of the current calibration iteration (200) and a previous acceleration duration TA,P of the previous calibration iteration (200).
10. The method according to any one of the preceding claims, wherein the profile of the coil current lc for the second calibration injection corresponds to a regular injection that is aborted when the second injection quantity Q2 has been reached.11 . The method according to any one of the preceding claims, wherein a deceleration duration TD of the deceleration phase PD is kept constant.
12. The method according to any one of the preceding claims, wherein a plurality of calibration processes (150) is performed, and a plurality of normal injections is performed (110) between each two calibration processes (150).
13. The method according to any one of the preceding claims, wherein the lower offset TLO and / or the upper offset Tuo are adapted (290) after the first calibration process (150).
14. The method according to any one of the preceding claims, wherein the armature (14) is separate from the pintle (13) and pushes the pintle (13) into the open position (C).
15. An injection system (1 ) with a fuel injector (10) of an engine, which fuel injector comprises:- a housing (11 ) extending axially along an injector axis (A) from a proximal end (10.1 ) to a distal end (10.2) and having a nozzle (12) at the distal end (10.2),- a pintle (13) being axially movable between an open position (C) and a closed position (B) in which it closes the nozzle (12),- an armature (14) being axially movable between a passive position (D) and an active position (E) and adapted to move the pintle (13) to the open position (C), and- a magnetic coil (16) for magnetically attracting the armature (14) towards the active position (E) and being operable by a coil current lc, wherein the injection system (1 ) is adapted so that during a regular injection (110), a profile of the coil current lc comprises an acceleration phase PA with a regular acceleration duration TA,R during which the coil current lc is adapted to accelerate the armature (14) towards the active position (E), a deceleration phase PD during which the coil current lc is decreased to decelerate the armature (14), and a retention phase PR during which the coil current lc is increased to keep the armature (14) in the active position (E), wherein the injection system (1 ) is further adapted to perform at least one calibration process (150), which comprises defining a current acceleration duration TA,C based on the regular acceleration duration TA,R and performing at least one calibration iteration (200) with the following steps:- performing (210) a first calibration injection by performing the acceleration phase PA using the current acceleration duration TA,C and afterwards reducing the coil current lc to decelerate the armature (14),- performing a measurement to detect a lift-stop feature LS indicating that the armature (14) has reached the active position (E), and- if the lift-stop feature LS is detected at a lift-stop time tis, comparing the liftstop time ti_s with a projected retention-start time tRs of the retention phase PR based on the current acceleration duration TA,C, and changing (260) the current acceleration duration TA,C depending on the comparison, wherein, if a convergence criterion related to the current acceleration duration TA,C is fulfilled in one calibration iteration (200), the regular acceleration duration TA,R is re-defined (280) based on the current acceleration duration TA,C and the calibration process (150) is ended, and otherwise, a next calibration iteration (200) is started.