In-cylinder method for air/fuel ratio control

Abstract

The invention relates to a Method of purging a lean NOx trap (LNT) coupled downstream of a direct injection internal combustuion engine with post injection for achieving a purge air/fuel (A/F) ratio. At least two separate post injections are applied in the expansion stroke after a main injection train, so that a torque produced is decoupled from the A/F-control.

Description

FIELD OF INVENTION

The Invention relates to a method of purging a lean NO_x trap (LNT) coupled downstream of a direct injection internal combustion engine, and more particularly to using post injection to achieve a purge air/fuel (A/F) ratio.

BACKGROUND AND SUMMARY OF THE INVENTION

A typical lean burn internal combustion engine is equipped with an exhaust gas aftertreatment device, such as a lean NOx trap (LNT) that absorbs and stores emissions of oxides of nitrogen (NOx) during the lean phase. When saturated with NOx molecules, a rich operation phase (order of few seconds) is required to purge the trap. This allows the release of the stored NOx molecules and its reduction into non-polluting components, mainly nitrogen, carbon dioxide, and water vapour. The frequency of this purging action is determined by the engine out NOx emissions and the storage capacity of the LNT which is also dependent on the temperature of the exhaust gas, typically a loading cycle will span few minutes of lean (normal operation mode for a diesel engine) which is followed by few seconds in the purging rich mode. Hence, for secure LNT operation, there must be means to increase the fuel to air ratio (F/A) in the exhaust gas to lambda (lambda is the air to fuel ratio relative to stoichiometric value) between 0.90 and 0.98 (oxygen <=1%) with sufficient level of reducing agents (HC and CO) under all engine operating conditions.

In-cylinder post injection (i.e., injection of additional fuel into all or some cylinders after the respective main injection but still during the power stroke) is an efficient method not only for increasing exhaust gas temperature but also for achieving a purge A/F ratio. Depending on the quantity and start of injection of the post injection, a fraction of injected fuel may burn in the cylinder and contribute both to engine torque as well as to engine out temperature, whereas the remaining fraction will evaporate and leave the engine as unburned hydrocarbons (HC). In that sense, post injection is an effective way to achieve the two basic functionalities (reduction of oxygen level in the exhaust stream as well as providing reducing agents HC and CO through partial combustion) required during the rich purging phase of the LNT. The quantity and especially timing of the post injection have to be calibrated very carefully in order to achieve desired air/fuel ratio.

The sensitivity of post injection combustion to injection timing is especially critical at lower engine loads. Under those conditions, the energy released by the combustion of the main fuel quantity is rather low. If the post injection is timed too late, only unburned HCs will be generated. On the other hand, if it is timed too early, it will burn completely and increase the torque output of the engine thus violating the torque neutrality limit. This torque increase can be compensated for by reducing the main quantity, however, there is a lower limit on injection quantities below which the fuel injection hardware does not operate sufficiently accurately, and this lower limit is close to the main fuel quantity at low engine load. At very low engine loads, it may not be possible to achieve sufficiently the desired air/fuel ratio due to the fact that the required amount of post injected fuel cannot be injected without violating either the HC concentration limits or the torque increase limits mentioned above.

A further problem becomes apparent during transient operation especially at low vehicle speeds (with frequent decelerations and idle periods (urban driving)). As opposed to in-direct injection, homogeneous mixture combustion typical to Otto engine, it is extremely difficult to achieve the desired lambda control (subject to the tail pipe emission constraint on HC and CO) during the transient trajectory because of the non-homogeneous mixture combustion characteristic of direct injection Diesel engines. Moreover, a prolonged attempt for the rich pulse will result in excessive rise in the LNT temperature resulting in a rapid release of the trapped NOx without conversion to the non-polluting elements. The importance of this should not be underestimated especially when one considers more transient regulatory cycles, such as the FTP75 required for the US market as opposed to the European regulatory NEDC test cycle with more steady-state phases.

With current technology, it is possible to purge the LNT by making use of post injection in the expansion stroke. However, complex control strategies are required to deal with the strong coupling between the two functionalities required, mainly, maintaining desired air/fuel ratio and sufficient level of the reduction gases (HC and CO) subject to neutrality of torque, tail pipe emissions of HC and CO, as well as LNT temperature levels. These limitations result in an overall limited trapping efficiency, inability to purge the trap under low load conditions, as well as excessive calibration effort.

This invention is directed to a direct injection internal combustion engine equipped with an after-treatment system that trap emissions of NOx under lean operation (normal mode) and will periodically require achieving rich purge to convert the trapped NOx into non-polluting gases. An engine management system includes the ability of multiple fuel injections, notably injections during the expansion stroke and after the main injection (post-injection). Preferably separate multiple (at least two) post injections after the main injection train are applied in the expansion stroke.

The introduction of multiple post injections (two or more) is intended to achieve the following functionalities:

• • Pre-conditioning of the state of gas in-cylinder through partial combustion of the current injection in preparation for the next injection.
  • Stimulation the completion of the in-cylinder combustion of the previous injection through the generation of a strong mixing effect resulting from the increased turbulence level caused by the high kinetic energy of the current injection.
  • Proportionally splitting the total quantity required on several injections.
  • Decoupling functionality of upstream trap combustion (either in-cylinder or in the exhaust manifold) and HC production. In this sense, for non-close coupled injections, the timing of injection is placed far enough to ensure within a pre-determined range of change in its quantity to supply the reducing agent (HC and CO as a result of the partial combustion of the posted injected fuel) and at the same time torque neutrality is sustained. The lower range of the quantity change is determined by the torque neutrality criterion and the upper limit is determined by the concentration of reducing agent in the exhaust gas.

The realization of such functionalities through multiple injections in the expansion stroke results in:

• • Improved ability to achieve the target of one percent oxygen level or less in the exhaust gas. The splitting of the required post injection quantity on multiple injections allows calibration of quantity and timing of each injection while respecting the boundary conditions for neutral engine torque (same torque interpretation for a given driver demand), cycle-to-cycle variability, and HC limitation. On the other hand, attempting to achieve the same result with a single post injection may prove to be more difficult (time-intensive), if not impossible, because of the inability to achieve stable in-cylinder combustion of a large single post quantity. In general, tight A/F ratio control for stoichiometric or less values (lambda 0.9 to 1.0) can be achieved over a large area of the engine map.
  • Robust in-cylinder combustion of the multiple injection train for large deviations in the air path, engine thermal level e.g. cold start, environmental deviations such as sub-zero ambient temperature and low ambient pressure at high altitude. This eliminates the requirement for inter-cooler bypass for the case of turbocharged engines as well as the requirement for intake air heating.
  • Limiting the level of fuel oil dilution by splitting a single large quantity of post injection on two or more post injections.
  • Improved A/F ratio control under real world transient drive cycles as a result of the late combustion in the expansion stroke without perceivable torque production. This is expected to improve the overall trapping efficiency by limiting the non-desired rapid NOx desorption if trap temperature exceeds a pre-defined limit.
  • For high sulphur content fuels, heating strategies required for desulphurization of the LNT can very easily make use of wobbling approaches (lean-to-rich-to-lean) by closed loop control of the non-torque producing post injections to achieve the required lambda control without the need for excessive calibration effort of two modes (lean/rich) with the challenging task of retaining neutral torque requirement during transitions.

The above advantages and other advantages, and features of the present invention will be readily apparent from the following detailed description of the preferred embodiments when taken in connection with the accompanying drawings, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

The objects and advantages described herein will be more fully understood by reading an example of an embodiment in which the invention is used to advantage, referred to herein as the Description of Preferred Embodiment, with reference to the drawings, wherein:

FIG. 1 is a strategy block diagram for determining of injection signal, and

FIG. 2 is an Engine out oxygen concentration on CEC cycle test.

DESCRIPTION OF PREFERRED EMBODIMENT(S)

An implementation of the multiple post strategy 1 is shown in FIG. 1. In its simplest form, a base steady state calibration 2 (Feed Forward Calibration Steady state) as a function of engine state and corrected for environmental deviation (timing and quantity maps as function of engine speed and driver demand (total fueling or indicated torque)), including correction maps/curves based on engine thermal level (coolant temperature/cylinder head temperature) and environmental deviation (ambient temperature and pressure) (Feed Forward Ambient Correction 3). In this implementation, the injection signal is controlled in an open loop manner 4.

In an alternative, more sophisticated implementation, the post injection strategy includes a feedback term, which corrects the post injection such that allowable F/A (or A/F) ratio levels are achieved. As shown in FIG. 1, a feed forward transient correction 6 as a function of engine state is further corrected as a function of the deviation in the air path (error signal for manifold pressure/mass air flow/manifold temperature)(Feed Forward Air path deviation correction 7). The purpose of this part is to achieve a steady state engine-dyno calibration with modified target F/A ratio and modified boundary conditions. The closed loop controller 8 responds to the error between desired and measured (or estimated) target F/A ratio by trimming the transient correction part (Feed back control Correction factor 9). The control authority of the transient correction factor is calculated as a function of the measured exhaust gas temp (turbine inlet temperature, catalyst/Nox trap temperature, catalyst/NOx trap exotherm, etc.) (Feed back control maximum limiter 11).

The function of the controller x-x is implemented as a multiplication i.e. the output of the closed-loop control is multiplied by the summation (Feed Forward transient correction plus Feed Forward air path deviation correction).

For an injection system capable of preferably two post injections, two implementations of the above structure are given below:

Case I: Rich Mode Purging of LNT

• • The steady state feed forward calibration 2 will target the minimum trap temperature required to keep the trap above its light off temperature and will be carried for the first and second post injection quantity and timing.
  • The transient feed forward correction 6 calibration will target the desired F/A ratio and will be carried for the first and second post injection quantity, whereby the closed loop correction will only be applied to the second post injection (minimal contribution to torque production).
  • The air path deviation will be calibrated such that for large deviations, the final output will reduce to temperature control (sustain LNT above light off temperature) and for smaller deviation to trimming feed-forward term for F/A ratio control. The second post injection quantity will be corrected in closed-loop to achieve the desired F/A ratio value using a classical controller (PID), which is limited.
    Case II: Heating Mode for Desulphurization of LNT
  • The steady state feed forward calibration 2 will target a lean heating mode with exhaust oxygen percent greater than three percent and will be carried for the first and second post injection quantity and timing.
  • The transient feed forward correction calibration 6 will target a rich heating mode with exhaust oxygen percent less than one percent. The desired F/A ratio will be carried for the first and second post injection quantity, whereby the closed loop correction will only be applied to the second post injection (minimal contribution to torque production).
  • The air path deviation will be calibrated such that for large deviations, the final output will reduce to the lean mode and for smaller deviation to trimming feed-forward term for F/A ratio control. The second post injection quantity will be corrected in closed-loop to achieve the desired F/A ratio value using a classical controller (PID), which is limited.

One important aspect in the above implementations is that the pre-requirement for multiple post injection to implement such a strategy 1 since one should be able to achieve dynamic correction of the quantity of post injection without perceivable intervention from the driver to achieve same torque for a given demand. In this sense, the first post injection (close coupled to main injection) is not trimmed using the feedback control. On the other hand, the second injection is placed far enough not to produce any torque and thus for corrected dynamically.

The result of multiple post injection for rich purge mode on the low speed 12 (32 km/hr) ECE part of the regulatory cycle is shown in FIG. 2. The oxygen concentration is compared for the rich purge pulse (in thick red 13) to the normal lean mode (in thin blue 14). The desired level of one percent of the oxygen level is achieved at such low speeds as a result of the double post injection approach as outlined earlier.

Therefore, a method for purging and disulphating an LNT using at least double post injection to achieve coordinated temperature and air-to-fuel ratio control and transient correction as a function of air path deviations, is disclosed.

This concludes the description of the invention. The reading of it by those skilled in the art would bring to mind many alterations and modifications without departing from the spirit and the scope of the invention. Accordingly, it is intended that the scope of the invention be defined by the following claims:

Claims

1. A method of purging a lean NOx trap (LNT) coupled downstream of a direct injection internal combustion engine, comprising:

applying at least two separate post injections in the expansion stroke after a main injection train such that there is substantially no torque generated due to said two separate post injections.

2. The method according to claim 1, further comprising a pre-conditioning of the state of gas in-cylinder through partial combustion of the current injection in preparation for the next injection.

3. The method according to claim 2, further comprising a stimulation the completion of the in-cylinder combustion of the previous injection through the generation of a strong mixing effect resulting from the increased turbulence level caused by the high kinetic energy of the current injection.

4. The method according to claim 3, wherein a total injection quantity required is proportionally split into several injections.

5. The method according to claim 4, wherein a lower range of the quantity change is determined by the torque neutrality criterion whereby an upper limit is determined by the concentration of reducing agent in the exhaust gas.

6. The method according to claim 5, wherein the injection signal is controlled in an open loop manner.

7. The method according claim 6, wherein a closed loop controller responds to an error between desired and measured target F/A (or A/F) ratio by trimming a transient correction part.