The invention will now be described, by way of example only, by reference to the accompanying drawings in which:
FIG. 1 is a schematic illustration of a known fuel injection system,
FIG. 2 is a schematic graph to illustrate the drive load characteristic of a known control method for the fuel injection system in FIG. 1, and for which the origin of mechanical noise is identifiable,
FIG. 3 is a schematic graph to illustrate the drive load characteristic of an alternative known control method for the fuel injection system in FIG. 1, and for which the origin of mechanical noise is also identifiable,
FIG. 4 shows (a) the drive current signal for the spill valve, (b) spill valve lift, (c) the injection pressure/drive load characteristic, (d) the drive current signal for the nozzle control valve, (e) nozzle control valve lift and (f) valve needle lift for a control method in accordance with the present invention,
FIG. 5 shows experimental results for (a) cylinder pressure, (b) injection pressure and (c) needle lift using the control method in FIG. 4,
FIG. 6 shows experimental results for (a) cylinder pressure, (b) injection pressure, (c) needle lift, (d) spill valve drive current signal and (e) nozzle control valve drive current signal for an alternative embodiment of the control method,
FIG. 7 shows a table of results corresponding to FIG. 6 to illustrate the smoke emissions benefits obtained using the control method illustrated in FIG. 6, and
FIG. 8 shows experimental results for (a) spill valve drive current signal and (b) injection pressure for an alternative embodiment of the control method.
FIGS. 2 and 3 show the drive load characteristic during an injection event using one known control method for the fuel injection system in FIG. 1. Between times t1 and t2, the spill valve 20 is closed and fuel pressure within the pump chamber 16 increases. Between times t2 and t3, the nozzle control valve 22 is opened to reduce fuel pressure within the control chamber 24, and the valve needle 12 starts to lift from its seating due to high fuel pressure within the delivery chamber which acts against the spring load 13, and injection commences. Between times t3 and t4 the nozzle control valve 22 is closed, and the flow of high pressure fuel from the supply line 18 into the control chamber 24 serves to urge the valve needle into engagement with its seating (at time t4) to terminate injection.
Between times t4 and t5, pressurisation of the control chamber 24 continues due to the continuous flow of high pressure fuel from the supply line 18, and the mechanical drive load increases. At time t5, the spill valve 20 is opened, so that fuel pressure within the pump chamber 16 and the supply line 18 is reduced, and a point is reached, just prior to time t6, at which the drive load is negative (i.e. herein referred to as “drive load overshoot”). At time t6 in the cycle the pump components are therefore caused to separate. Mechanical noise generated as a result of the drive components re-contacting one another as the drive load oscillates about zero (between times t6 and t7) is undesirable.
Using this injection control method, it will be appreciated that the end of injection is initiated by actuating the nozzle control valve 22 to increase pressure within the control chamber 24.
An alternative control method will now be described with reference to FIG. 3; which shows the drive load during an injection cycle in which injection is terminated by actuating the spill valve 20. As described previously, between times t1 and t2 the spill valve 20 is closed and fuel pressure within the pump chamber 16 increases. Between times t2 and t3 the nozzle control valve 22 is opened to reduce fuel pressure within the control chamber 24, and the valve needle 12 starts to lift from its seating to commence injection. At time t3, the spill valve 20 is opened to reduce the pressure of fuel delivered to the nozzle through the supply line 18, so that the force acting on the needle 12 is reduced and the valve needle 12 is urged against its seating by the nozzle spring load 13. There is a rapid decrease in drive load when the spill valve 20 is opened and, due to the relatively high rate of flow of spilled fuel to low pressure, the drive load overshoots (time t4). The valve needle is seated at time t3A. As described previously, the drive load overshoots until at time t4 the drive train components are caused to separate. Subsequent re-contact of the drive train components (between time t4 and time t5) gives rise to mechanical noise, which continues as the drive load oscillates about zero between times t4 and t6, and possibly beyond.
One aspect of the present invention, referred to as “synchronous end of injection” addresses the aforementioned problem and alleviates the disadvantageous effects of mechanical noise by near-synchronising operation of the spill valve 20 and the nozzle control valve 22 to terminate injection. FIG. 4 shows (a) the drive current signal for the spill valve 20, (b) the spill valve lift, (c) the drive load (as shown in FIGS. 2 and 3), (d) the drive current signal for the nozzle control valve 22, (e) the nozzle control valve lift and (f) the valve needle lift, for the synchronous end of injection method. FIG. 5 shows measurements of (a) cylinder pressure, (b) injection pressure and (c) valve needle lift using the synchronous end of injection method of FIG. 4.
At the start of an injection cycle, a drive current signal 30 for the spill valve 20 is switched on and ramps between times t1 to t2 to a first relatively high current level. The spill valve 20 is moved into its fully closed state at time t3, and fuel pressurisation within the pump chamber 16 commences, as illustrated in FIG. 4(c). At time t3, a drive current signal 40 is applied to the nozzle control valve 22 to lift (open) the nozzle control valve 22, thereby causing fuel pressure within the control chamber 24 to be reduced until at time t4 the valve needle 12 starts to lift from the seating, against the spring load 13, to commence injection.
Termination of the fuel injection event is achieved in the following manner. At time t5, the drive current signal 30 is switched off, defining a drive signal off-pulse, and a short time later (time t8) the spill valve 20 starts to move towards its open state. At time t7, just after the drive current signal 30 for the spill valve 20 is switched off, a drive current signal 40 for the nozzle control valve 22 is switched off and the nozzle control valve 22 starts to move to its closed state (time t8) to re-establish high fuel pressure within the control chamber 24. The nozzle control valve 22 is therefore actuated within a spill valve opening period. The near-synchronisation of opening of the spill valve 20 and closure of the nozzle control valve 22 to terminate injection (time t9) has the effect that the drive load decays to zero, but that oscillations about zero and drive load overshoot are avoided. This advantageous effect is illustrated in FIG. 4(c) between times t9 and t10.
Immediately after injection has been terminated, between times t9 and t10, the rate of flow of fuel through the open spill valve 20 to low pressure will be reduced compared to that using the conventional methods (FIGS. 2 and 3), due to the additional positive (reflected) pressure wave generated upon closure of the valve needle 12 which occurs as the spill valve 20 is still moving towards its fully open position. It is therefore an important feature of the invention that at the moment the valve needle 12 actually closes (time t9) at the end of pressure decay, the spill valve 20 is still moving towards its open position (see FIG. 4(b)) i.e. during spill. By controlling the spill and nozzle control valves 20, 22 such that they are phased in this manner, the problems associated with drive load overshoot, and hence mechanical noise, may be alleviated. In practice, the optimum relative timing between actuation of the valves will be determined by parameters of the system, such as actuator delays and pipe lengths and volumes. Conveniently, the relative timing of switching of the valves 20, 22 to terminate injection may be determined by reference to pre-calibrated data, using look up tables or data maps.
In an alternative embodiment, the rate of spill through the open spill valve 20 to terminate injection may be modified further by reconfiguring certain flow passages/restrictions, thereby further reducing the rate of decay of pressure between times t9 and t10 and, hence, further reducing mechanical noise generated by the drive train components at the end of injection.
A further aspect of the invention permits the smoke emissions level to be reduced at the end of an injection event by selecting particular drive current signal timings for the spill valve 20 and the nozzle control valve 22. It has been found that for lower values of needle lift (e.g. at the end of injection) the associated fuel spray form is less desirable. Thus, by minimising the fuel quantity and the injection pressure at such lower lift values, the smoke level in the exhaust steam can be reduced. It is important, however, that the injection pressure is maintained at a sufficiently high level during closure of the valve needle to avoid blowback of the cylinder gases into the nozzle.
Additionally, as described previously, it is beneficial to reduce the rate of decrease of injection pressure (times t9 to t10 in FIG. 4(c)) at the end of injection so as to reduce the effects of mechanical noise generated by the drive train components. To compromise between achieving a relatively low smoke emission level, no nozzle blowback and low mechanical noise, it has been found that the spill valve drive current signal 30 should be switched off (i.e. to open the spill valve 20) between 0.1 and 1 millisecond before the nozzle control valve drive current signal 40 is switched off to close the nozzle control valve 22 and, thus, seat the valve needle 12.
It is also desirable in some operating conditions for a pilot injection of fuel to be delivered to the engine, prior to a main injection, or for a main injection of fuel to be followed by a post injection. For example, the use of a pilot injection prior to a main injection can be used to reduce combustion noise, and the use of a post injection shortly after a main injection has combustion benefits for soot reduction.
In order to optimise the combustion benefits of applying a main injection followed by a post injection, it is desirable to vary the injection pressure for the post injection of fuel such that the post injection pressure is high. This may be achieved using the method illustrated in FIG. 6, which shows the drive current signals 30, 40 for the spill valve 20 and the nozzle control valve 22 respectively. The injection pressure is illustrated by three traces 50A, 50B and 50C, having decreasing peak injection pressure values. The bottom trace in FIG. 6 illustrates valve needle lift, and shows a main injection (60A) of fuel followed by a post injection (60B) within the same injection cycle.
To terminate the main injection of fuel prior to the post injection, the spill drive current signal 30 is switched off (corresponding to time t8, as shown in FIG. 4(b)) to move the spill valve 20 towards its open state. The nozzle control valve 22 is in its closed state at the end of the main injection event. It will be appreciated that the drive current signals 30, 40 in FIG. 6 are shown to be switched negative to activate the respective valve, but equally the electrical connections to the valves may be configured such that the drive current signals 30, 40 are switched positive to actuate the respective valves.
To initiate the post injection of fuel, the drive current signal 30 for the spill valve 20 is switched on to reduce the rate at which fuel can escape from the supply line 18 and pump chamber 16 to low pressure, and fuel pressure in the nozzle delivery chamber starts to increase as pumping continues. When the post injection of fuel is required, the drive current signal 40 is switched on to move the nozzle control valve 22 into its open state to relieve fuel pressure within the control chamber 24, and the valve needle 12 is caused to lift again.
As can be seen from the three traces 50A, 50B and 50C, the period of time after termination of the main injection at which the drive current signal 30 for the spill valve 20 is switched on to cause the valve 20 to start to close determines the rate of decay of fuel pressure between the main injection and the post injection, and thus determines the injection pressure for the post injection event. It has been found that there are considerable benefits for emissions if the post injection pressure is high, and by careful selection of the relative timing between opening of the spill valve 20 to terminate the main injection event and closure of the spill valve 20 for the post injection event, a high pressure post injection of fuel can be achieved. By way of example, FIG. 7 shows numerical test results corresponding to FIG. 6, from which it can be seen that the level of smoke in the exhaust emissions is considerably reduced for higher post injection pressures. For a post injection pressure greater than 1700 bar (50B), the level of emitted smoke is greatly reduced compared with a lower post injection pressure (50C), and a post injection pressure in excess of 2000 bar (50A) improves the smoke emissions level further. It can also be seen from FIG. 6 that only for trace C is the injection pressure reduced to substantially zero between the main and post injections of fuel.
It has previously been proposed to achieve a main injection of fuel followed by a post injection of fuel by operating only the nozzle control valve 22. Using this technique, the spill valve 20 remains closed between the main and the post injection and instead just the nozzle control valve 22 is opened to relieve fuel pressure in the control chamber 24 to initiate the post injection of fuel, and is closed at the end of post injection to re-establish high fuel pressure in the control chamber 24. As a consequence of using this method, however, the level of injection pressure in the post injection is not under independent control and may be excessively high, especially if wider separation times between the main and post injection events are required. The benefit of the present invention is that the injection pressure can be made independent of the separation between the main and post injection.
FIG. 8 illustrates experimental results achieved using a further alternative embodiment of the invention, in which the spill valve drive signal 130 includes two pulses, one to initiate spill valve opening and an additional pulse that is applied later, but still during the spill valve opening phase. In the following example only the termination of a main injection of fuel is described, although it will be appreciated that a subsequent post-injection of fuel may follow, as described previously. It will be appreciated that the spill valve drive signal 130 in FIG. 8 is switched off to open, as in the FIG. 6 example, although the current axis is inverted compared to that in FIG. 6.
The spill valve drive signal 130 is initially switched off when it is required to initiate pressure decay, causing the spill valve 20 to start to move towards its open state. The initial spill valve actuation pulse is identified by pulse 131 in FIG. 8. At time t8 (as described previously), the spill valve 20 is fully open. It is a particular feature of this embodiment that the spill valve drive signal 130 is modified during the spill valve opening phase by switching it on, and then off again, prior to full opening of the spill valve 20. This is indicated by the presence of a second spill valve actuation pulse, identified as 132.
As before, the nozzle control valve drive signal is de-actuated to cause the nozzle control valve 22 to close during the spill valve opening phase (i.e. synchronised end of injection), prior to full opening of the spill valve 20. For simplicity the nozzle control valve drive signal is not shown in FIG. 8. The additional spill valve actuation pulse 132 occurs during the spill valve opening period, and thus substantially simultaneously with actuation of the nozzle control valve 22.
With reference to FIG. 8, the time at which the additional actuation pulse 132 is provided may be determined by means of a glitch detection technique. Glitch detection techniques are well known in fuel injection control methods and will be familiar to those skilled in this field, for example see EP 0857251. Such methods are used to provide an accurate indication of when the spill valve 20 reaches a fully open (or closed) state.
In practice, the additional spill valve actuation pulse 132 will be implemented for most, but not all, injection events. For those events where it is not applied, a current glitch or discontinuity in the current signal can be detected to provide an indication of the correct time at which the additional spill valve actuation pulse 132 should be applied for subsequent events. The current signal glitch detection discontinuity is identified at 70 in the current signal shown as a solid line in FIG. 8 (i.e. without the second pulse 132, and similar to drive current signal 30 in FIG. 6). It can be seen that for events where the second actuation pulse 132 is applied, it is timed to occur just prior to the point at which the glitch detection discontinuity 70 occurs.
The benefit of providing an additional spill valve actuation pulse 132 during the spill valve opening period can be seen by comparing the injection pressure for the two-pulse drive signal 130, represented by “Injection Pressure A”, with the “Injection Pressure B” obtained for the single pulse drive signal 30. It can be seen that the provision of the additional pulse 132 results in a reduction in the injection pressure decay rate compared to that obtained for the single pulse drive signal 30, therefore providing a noise benefit upon termination of injection.
The glitch detection method may be implemented by means of an adaptive control algorithm. Alternatively, data maps or look-up tables may be used.
As an alternative to using the glitch detection technique, data mapping techniques may be used to determine the correct timing of the second actuation pulse 132.
Although the fuel system shown in FIG. 1 shows the spill valve 20 and the nozzle control valve 22 arranged remotely from one another, the invention is equally applicable to unit pump injection systems in which the spill and nozzle control valves are arranged within a common housing, or in which the spill valve and the nozzle control valve functions are under the control of a common actuator.
Patent Application
20080092853
