Method for determining a leak in the intake air duct of an internal combustion engine

Abstract

Method for determining a leakage in an intake air duct of an internal combustion engine with at least one combustion chamber, wherein the intake air duct is divided into at least two sections, comprising a section (A1, A2) upstream of a throttle valve (12) and a section (A3) downstream of the throttle valve (12), wherein - as part of an engine test, an intake manifold pressure variation is carried out with a first steady state (SZ1), in which a first intake manifold pressure is set in the intake air duct, and a second steady state (SZ2), in which a second intake manifold pressure is set in the intake air duct; - for the different intake manifold pressures in the first and second steady-state conditions (SZ1, SZ2) according to a predefined calculation procedure which processes the actual lambda value (λ), the value (LA) of a lambda adaptation and the value (LR) of a lambda controller, a weighting factor (LIM) is determined; and - a leakage in the intake air duct is inferred from the result of a comparison of the determined value of the evaluation factor (LIM) with a predetermined threshold (SW1, SW2).

Description

[0001] The invention relates to a method for determining a leak in an intake air duct of an internal combustion engine having at least one combustion chamber. The intake air duct is divided into at least two sections, comprising a section upstream of a throttle valve and a section downstream of the throttle valve.

[0002] A leak in the intake air duct of a combustion engine can lead to deficiencies in the vehicle's handling and to various error log entries that are relevant to emissions. A check of emissions-relevant functions is required by law, and a leak in the intake air duct, which leads to a change in the air mass-fuel mixture ratio from a specified target value, must be avoided at all costs.

[0003] A leak in the intake air duct can occur, for example, during vehicle assembly, as the intake air duct is composed of a number of different, complexly shaped components that are unique to each vehicle type. Vehicles with leaky intake air ducts then cause problems for the customer after just a short mileage, which require a visit to the workshop.

[0004] Finding leaks, both during the assembly of a new vehicle and in the workshop, is difficult and time-consuming. If leaks are not found or are thought to be located in the wrong places, repairs are not carried out properly, so the problems described above can continue to occur. This can lead to unacceptably high warranty costs.

[0005] DE 103 00 592 A1 discloses a method for operating an internal combustion engine with an internal combustion engine and with an actuator in an air supply for adjusting an air mass to be supplied to the internal combustion engine. A position of the actuator is adapted to a current operating state by means of a short-term adaptation, and a position of the actuator is adapted to compensate for longer-term influencing variables on the position of the actuator by means of a long-term adaptation. During the long-term adaptation, a long-term adaptation value for the position of the actuator is formed as a function of a minimum short-term adaptation value for the position of the actuator formed during the short-term adaptation. A leak in the air supply is diagnosed as a function of the long-term adaptation value.

[0006] DE 197 50 191 A1 discloses a method for monitoring the load detection of an internal combustion engine, in which a first value for the air mass flow supplied to the internal combustion engine is detected, in which a second value for the air mass flow is determined on the basis of the position of the throttle valve of the internal combustion engine, wherein an error is detected if the two signal values ​​deviate from each other to an inadmissible extent.

[0007] In the method described in DE 10 2007 023 850 B3, an intake tract and an exhaust tract of an internal combustion engine are connected to a combustion chamber depending on the switching position of the gas inlet valve or a gas outlet valve. The intake tract has an intake manifold, a throttle valve, and a load sensor. The exhaust tract has a lambda sensor, and actual values ​​of a load variable are determined by the load sensor, and estimates of the load values ​​are modeled based on an intake manifold model. A fault in the intake tract is identified depending on the smoothing function.

[0008] It is an object of the invention to provide a method and a device which enable the determination of a leak in an intake air duct of an internal combustion engine in a simple and robust manner.

[0009] This object is achieved by a method according to the features of patent claim 1 and a device according to the features of patent claim 16. Advantageous embodiments emerge from the dependent patent claims.

[0010] A method is proposed for determining a leak in an intake air duct of an internal combustion engine having at least one combustion chamber. The intake air duct is divided into at least two sections, comprising a section upstream of a throttle valve and a section downstream of the throttle valve. If the intake air duct includes a compressor, e.g., a turbocharger, the intake air duct is divided into at least three sections: a first section between the (air mass) sensor and the compressor, a second section downstream between the compressor and the throttle valve, and a third section downstream of the throttle valve.

[0011] In the method, an intake manifold pressure variation is carried out as part of an engine test with a first steady-state condition in which a first intake manifold pressure is set in the intake air duct, and a second steady-state condition in which a second intake manifold pressure is set in the intake air duct. The engine test can be carried out during vehicle production or in a workshop. The first and second steady-state conditions are predetermined time intervals within the overall duration of the engine test. The first and second steady-state conditions can comprise time intervals of the same length. The first and second steady-state conditions can comprise time intervals of different lengths. The length of the time intervals is in particular between 30 seconds and 2 minutes.

[0012] In the next step, a weighting factor is determined for the different intake manifold pressures in the first and second steady-state conditions according to a predefined calculation rule. The calculation rule processes the actual lambda value, the value of a lambda controller, and the value of a lambda adaptation.

[0013] The actual lambda value is determined by the combustion engine's lambda sensors, which are installed in a catalytic converter and measure the residual oxygen content in the exhaust gases. Using the residual oxygen content of the exhaust gases, an engine control system can determine whether the combustion air ratio in the combustion engine is correctly adjusted and whether legal emissions limits can be met. A stoichiometric mixture is described by an actual lambda value of 1 (λ = 1) and is created during combustion at a mass fraction of fuel to approximately 14.7 mass fractions of air (conventional regular fuel).

[0014] The lambda controller is a calculated variable of the engine control system that ensures that the actual lambda value always corresponds to a specified lambda target value and that stoichiometric combustion takes place. If the fuel-air mixture is too lean, the lambda controller increases the amount of fuel supplied to combustion. If the fuel-air mixture is too rich, the amount of fuel supplied to combustion is reduced.

[0015] Lambda adaptation is a learned parameter of the engine control system that reflects a permanent control deviation of the lambda controller. In other words, lambda adaptation represents a low-pass filtered signal of the lambda controller signal and enables faster control processes.

[0016] According to a further step, a leak in the intake air duct is concluded from a result of a comparison of the determined value of the evaluation factor with a predetermined threshold value.

[0017] Since the second steady-state condition involves increased negative pressure in the intake air duct, the lambda controller will jump in the event of a leak in the intake air duct and increase its value, as more unregistered false air enters the intake air duct. Without a leak, the lambda controller's value will not jump, as no false air enters the combustion engine's intake air duct.

[0018] The evaluation factor is, or enables, a statement about whether the combustion engine, when negative pressure is present in the intake air duct, is introducing additional false air into the intake air duct, which could enter through a leak in the intake air duct. By comparing the determined value of the evaluation factor with a predetermined threshold value, which can be determined through tests with vehicles that have a leak in the intake air duct and vehicles that do not have a leak in the intake air duct, a statement about the presence of a leak in the intake air duct can then be made with high reliability.

[0019] The proposed method enables a targeted, fast, and simple diagnosis of whether or not a leak is present in the intake air duct of a vehicle's internal combustion engine. The diagnosis can be performed automatically, enabling a fast and robust repair process in a workshop. Due to the short time required to perform the method, it can be used both in existing processes during vehicle production and in stationary operation in a workshop.

[0020] The information determined and evaluated in the method is generated during operation of the combustion engine and processed by an engine control unit (Electronic Control Unit). To implement the method, these measured values ​​simply need to be read from the engine control unit and processed, e.g., by a processing unit, in the manner proposed here. A further advantage of the method is that the leak location can be roughly categorized for rework. In particular, it can be determined with a high degree of certainty whether or not a leak exists in the section downstream of the throttle valve. The method thus makes it possible to increase the probability of a successful repair.

[0021] According to an expedient embodiment, in the first steady-state state, based on a reference intake manifold pressure difference when the engine is idling, an intake manifold pressure difference is induced that is smaller than the reference intake manifold pressure difference, and in the second steady-state state, an intake manifold pressure difference is induced that is larger than the reference intake manifold pressure difference. In other words, this means that during the first steady-state state, an intake manifold pressure prevails in the intake air duct that, in absolute terms, is greater than the intake manifold pressure in the second steady-state state.

[0022] The reference intake manifold pressure difference is the pressure difference between the ambient pressure and the air pressure prevailing in the intake manifold (part of the intake air duct) when the engine is running, particularly when idling. In the first steady-state condition, the intake manifold pressure can be brought approximately to the ambient pressure in the intake air duct. In the second steady-state condition, in contrast, there is an increased negative pressure in the intake air duct. It is expedient if the engine load point (speed, torque) is kept approximately the same during the intake manifold pressure variation. It is expedient if the difference between the intake manifold pressure prevailing when the engine is idling and the intake manifold pressure in the first steady-state condition, as well as the difference between the intake manifold pressure prevailing when idling and the intake manifold pressure in the second steady-state condition, are approximately the same.Depending on the engine, the differences can also take on different values ​​if this is appropriate for meaningful measurement results.

[0023] It is also advisable for the first steady-state condition to occur before the second steady-state condition. Depending on the engine, however, the sequence may be reversed.

[0024] It is also expedient if the intake manifold pressure variation is carried out in the section downstream of the throttle valve. In the first steady-state state, this can be achieved, for example, by reducing the lift of the intake valve of the at least one combustion chamber of the internal combustion engine when the throttle valve is open compared to conventional operation. This allows the pressure in the section downstream of the throttle valve to approach ambient pressure because less air is sucked into the combustion chamber due to the operation of the internal combustion engine. During the second steady-state state, in which the lift of the intake valve of the at least one combustion chamber is greater than during the first steady-state state, the throttle valve can be throttled, thereby enabling the intake manifold pressure in the section downstream of the throttle valve to be reduced.

[0025] A further advantageous embodiment provides that the calculation rule includes a change in the value of a lambda controller and the value of a lambda adaptation between the first and second steady-state conditions. The change in the value of the lambda controller and the value of the lambda adaptation results from the changes in the lambda controller and the lambda adaptation that occur during the first and second steady-state conditions as a function of a leak.

[0026] A further advantageous embodiment provides that the mean value of the actual lambda value is processed in the specified calculation rule during an intake manifold pressure variation. In particular, the mean value of the actual lambda value in the first steady-state condition is processed in the specified calculation rule.

[0027] A further expedient embodiment provides that the specified calculation rule processes respective mean values ​​of the lambda controller values ​​and the lambda adaptation values ​​in the first steady-state state and in the second steady-state state. In order to actually determine the mean values ​​of the lambda controller and the lambda adaptation in the first and second steady-state states, respectively, it is expedient to collect the mean values ​​from the time-dependent measured values ​​in a respective averaging period of the first and second steady-state states, which follow a respective preceding stabilization period. This ensures that, due to the sufficiently long time intervals of the first and second steady-state states, a steady-state value of the lambda controller or the lambda adaptation value is present.

[0028] According to a further expedient embodiment, the specified calculation rule processes respective difference values ​​of the lambda controller value and the lambda adaptation value between the assigned mean values ​​in the first steady-state state and the second steady-state state. The difference values ​​of the lambda controller value and the lambda adaptation value between the assigned mean values ​​in the first steady-state state and the second steady-state state depend on the size of the leak. Taking this change in the lambda controller and lambda adaptation values ​​into account enables a significant change in the evaluation factor in the event of a leak in the intake air duct of the combustion engine. Since there is an increased negative pressure in the intake manifold in the second steady-state state, the lambda controller performs a jump in the event of a leak and increases its value because more unregistered false air enters the intake air duct.In contrast, the jump of the lambda regulator does not occur without leakage in the intake air duct, since no false air gets into the intake air duct.

[0029] The evaluation factor is calculated according to a further expedient design using the following formula LIM=λSZ1*(LRSZ1+LASZ1)*(ΔLR+ΔLA+1), where λ SZ1 the mean value of the actual lambda value in the first steady state (measure of mixture quality), LR SZ1 the mean value (LR) of the lambda controller in the first steady state (measure of short-term mixture change), LA SZ1 the mean value (LA) of the lambda adaptation in the first stationary state (measure of long-term mixture change), ΔLR is the amount of the difference between the mean values ​​of the lambda controller values ​​between the first and the second stationary state, and ΔLA is the amount of the difference between the mean values ​​of the lambda adaptation values ​​between the first and the second stationary state.

[0030] This formula makes it clear that the difference between the mean values ​​of the lambda controller or the lambda adaptation in particular has a significant influence on the value of the evaluation factor LIM.

[0031] According to a further expedient embodiment, if a first threshold value is exceeded, a conclusion is drawn that there is a suspected leak in the intake air duct, in particular in the section downstream of the throttle valve, and if a second threshold value, which is higher than the first threshold value, is exceeded, a conclusion is drawn that there is a reliably detected leak in the intake air duct, in particular in the section downstream of the throttle valve. By providing two threshold values, a distinction can be made between a suspected and a definitely existing leak.

[0032] It is also expedient if, when deciding whether there is a leak in the intake air duct, a difference between the measured or modeled air mass taken in during the second steady-state and the first steady-state is processed, and a leak in the intake air duct is concluded from the result of a comparison of the determined difference in the air mass with a predetermined threshold value. The air mass flowing into the intake air duct can be determined using one or more hot-film sensors. In the event of a leak in the intake air duct, the air mass taken in will be lower in the second steady-state than in the first steady-state, since in the second steady-state there is already enough air for combustion due to the increased false air in the intake air duct and the throttle valve reduces its opening angle.A small difference in the intake air mass in the first steady state to the second steady state indicates that there is no leakage in the intake air duct.

[0033] According to a further advantageous embodiment, it is expedient for the predetermined threshold values ​​to be determined depending on the oxygen content of the fuel. This embodiment is based on the consideration that, depending on the percentage of ethanol content in the fuel, the higher the ethanol content, the less fresh air needs to be supplied for combustion.

[0034] Furthermore, a system for determining a leak in an intake air duct of an internal combustion engine having at least one combustion chamber is proposed. The intake air duct is divided into at least two sections, comprising a section upstream of a throttle valve and a section downstream of the throttle valve. The system is designed to carry out an intake manifold pressure variation as part of an engine test with a first steady-state condition in which a first intake manifold pressure is set in the intake air duct, and a second steady-state condition in which a second intake manifold pressure is set in the intake air duct. The system is further designed to determine an evaluation factor for the different intake manifold pressures in the first and second steady-state conditions according to a predetermined calculation rule that processes the actual lambda value, the value of a lambda adaptation, and the value of a lambda controller.Finally, the system is designed to conclude that there is a leak in the intake air duct based on the result of a comparison of the determined value of the evaluation factor with a predetermined threshold value.

[0035] The system according to the invention has the same advantages as those described above in connection with the method according to the invention.

[0036] The system may further comprise further means for carrying out the method described in this description.

[0037] The invention further proposes a computer program product that can be loaded directly into the internal memory of a digital computer and comprises software code sections that execute the steps of the method described herein when the product is run on a computer. The computer program product can be embodied, for example, on a CD-ROM, a DVD, a USB stick, or a signal that can be loaded over a network.

[0038] The invention is explained in more detail below using an exemplary embodiment in the drawing. They show: Fig. 1 is a schematic representation of an internal combustion engine illustrating the intake air duct; Fig. 2 a diagram showing the time course of relevant measured values ​​for determining a leak in an intake air duct of an internal combustion engine when no leak is present; Fig. 3 a diagram showing the time course of relevant measured values ​​for determining a leak in an intake air duct of an internal combustion engine when a leak is present; Fig. 4 is a flow chart showing the essential steps for determining a leak in an intake air duct of an internal combustion engine; and Fig. 5 a flowchart showing the steps for determining a leak in an intake air duct of an internal combustion engine.

[0039] Fig. Figure 1 shows a schematic representation of an intake air duct in an internal combustion engine, which has, for example, four cylinders as combustion chambers. Fig. 1 denote: 1 a charge air cooler, 2 a bypass valve, 3 an intake silencer (filter), 4 a hot-film air mass meter (sensor), which measures the air mass flowing into the combustion engine in kg / h immediately after the intake silencer 4, 5 an exhaust turbocharger, 6 a wastegate valve, 7 a lambda sensor upstream of a catalytic converter 8, 8 the catalytic converter, 9 a lambda sensor downstream of the catalytic converter 8, 10 digital engine electronics (DME), 11 an intake manifold pressure sensor, which provides a signal indicating the throttling, 12 a throttle valve and 13 a charge air temperature / pressure sensor. The intake air duct of the Fig. The intake air duct of the internal combustion engine shown in Figure 1 essentially comprises three sections: a first section A1 is formed between the hot-film air mass meter 4 or intake silencer (filter) 3 and the compressor 5. A second section A2 is formed between the compressor 5 and the throttle valve 12. A third section A3 is formed between the throttle valve 12 and a respective inlet of the combustion chamber of the internal combustion engine. If the internal combustion engine does not have a compressor, the intake air duct only comprises sections upstream and downstream of the throttle valve 12.

[0040] The intake air duct formed from these three sections is a complex structure made up of various individual parts, various plug / hose connections and sealing surfaces, which serves to supply filtered fresh air to the combustion chambers in the internal combustion engine. Fig. In the turbocharged internal combustion engine shown in Figure 1, the intake air duct is exposed to a wide pressure range. This ranges from ambient pressure conditions in the first section A1 (i.e., between the hot-film air mass sensor 4 / intake silencer (filter) 3 and the compressor (exhaust turbocharger) 5) to negative pressure in naturally aspirated operation in the third section A3 behind the throttle valve 12 or to positive pressure in turbocharged operation downstream of the compressor (exhaust turbocharger) 5. Inadequate installation of the intake air duct risks potential leakage.

[0041] Such a leak can flow in various directions depending on the pressure gradient. So-called "false air" can enter the intake air duct and increase the air mass in the intake air duct. So-called "leakage air" can escape from the intake air duct and reduce the air mass in the intake air duct. Both mass flows generate a deviation from the desired combustion air ratio, with corresponding negative effects. In particular, this can lead to an error log entry, which, if not discovered during production, can cause a malfunction indicator lamp (MIL) to illuminate during customer operation.

[0042] The complex assembly of the intake air duct components requires a detailed leak test during vehicle production. A purely visual inspection, even with a leak detection spray, is not always successful, at least for small leaks. It is also time-consuming and error-prone. The procedure described below enables an automatic test for the air duct's leak tightness. This test is simple and highly robust. The result of the procedure allows for a classification into "good" or "poor" and for the location of the leak to be narrowed down to one of the sections A1, A2, or A3.

[0043] The procedure for determining a leak in the intake air duct of the internal combustion engine includes an engine test in which the internal combustion engine is operated at idle (so-called idle test).

[0044] By controlling the air mass in a combustion chamber, the desired power requested by a driver based on the accelerator pedal position is controlled. The air mass flow flowing into the intake air duct is measured at the beginning of the intake air duct (i.e., at the inlet of the first section A1) by the hot-film air mass sensor 4.

[0045] The lambda value λ represents the ratio of air mass to stoichiometric air consumption during combustion superimposed for the corresponding combustion chambers by measuring the residual oxygen content in the exhaust gas and comparing it with that of the ambient air. The output of the lambda controller, which is intended to regulate a deviating mixture ratio, is designated LR. The controller's value is written over time into the so-called lambda adaptation, which can be found as LA in the digital engine electronics. The use of these two values ​​for the leak test arises from the fact that the leakage air changes the air to fuel ratio and this change is visible in the lambda value and lambda controller in the short term and in the lambda adaptation in the long term. This circumstance is exploited in the idle test, which can be used in a workshop, for example, because the lambda adaptation (LA) is already available. This workshop test allows more precise statements for naturally aspirated engine operation.

[0046] The basic principle for determining a leak in the intake air duct is to perform an intake manifold pressure variation in section A3 after throttle valve 12 during the engine test. This involves recording various measured variables in two steady-state conditions, a first steady-state condition SZ1 and a second steady-state condition SZ2, from which a weighting factor LIM is then determined. The measured variables recorded during the intake manifold pressure variation and the two steady-state conditions SZ1 and SZ2 include the actual lambda value λ, the value LR of the lambda controller, and the value LA of the lambda adaptation. Fig. 2 and Fig. In the diagrams shown in Figure 3, the time profiles of the actual lambda value λ are shown with a dashed line, the value LR of the lambda controller with a dash-dotted line, and the value LA of the lambda adaptation with a dash-double-dotted line. Furthermore, the intake manifold pressure PSR generated by the intake manifold pressure variation is shown with a dotted line.

[0047] In the Fig. 2 and Fig. 3, the first stationary state SZ1 comprises a period between times t3 and t1 (ie T SZ1 = t3 - t1). The first steady-state condition SZ1 lasts approximately one minute, for example only. In the first steady-state condition SZ1, a first intake manifold pressure p1 ≈ 955 mbar is set, for example only.

[0048] The second stationary state SZ2 follows the first stationary state SZ1 in time. This means that the second stationary state SZ2 begins at time t3, with an end of the second stationary state SZ2 in the Fig. 2 and Fig. 3 is not shown. The duration of the second steady-state state SZ2 can approximately correspond to the duration of the first steady-state state SZ1. During the second steady-state state SZ2, a second intake manifold pressure of, for example, p2 ≈ 860 mbar is set in the intake air duct, which is lower than the first intake manifold pressure p1 during the first steady-state state SZ1. The absolute value of the second intake manifold pressure p2 results, for example, from the fact that a difference of, for example, 100 mbar is set compared to the first steady-state state SZ1.

[0049] Not shown in the Fig. 2 and Fig. 3 is an ambient pressure which, depending on the ambient conditions and the altitude at which the engine test is carried out, is e.g. 970 mbar (or slightly lower or higher).

[0050] Before time t1, i.e. before the first steady-state state SZ1, the combustion engine is idling, with a pressure p3 ≈ 870 mbar being established in the intake air duct, for example, which pressure lies between the pressure p1 during the first steady-state state SZ1 and the second pressure p2 during the second steady-state state SZ2. The respective intake manifold pressures p1, p2 and p3 and the ambient pressure result in an intake manifold pressure differential, with the difference between the ambient pressure and the pressure p3 when the engine is idling representing a reference intake manifold pressure differential. In the first steady-state state SZ1, an intake manifold reference is thus produced which is smaller than the reference intake manifold pressure difference, whereas in the second steady-state state SZ2 there is a larger intake manifold pressure difference than the reference intake manifold pressure difference.In other words, during the steady-state condition SZ1 there is approximately ambient pressure in the intake air duct, while in the steady-state condition SZ2 there is an increased negative pressure in the intake air duct.

[0051] The above-mentioned evaluation factor LIM, which is determined from the actual lambda value λ and the changes in the value of the lambda adaptation LA and the value of the lambda controller LR, makes it possible to determine whether the engine draws additional false air into the intake air duct when there is a negative pressure in section A3.

[0052] The actual lambda value λ is determined by the lambda sensors 7, 9, which are installed on the catalytic converter 8 and monitor the residual oxygen content in the exhaust gases. Using the residual oxygen content of the exhaust gases, the digital engine electronics 10, which serves as the engine control unit, can check whether combustion was stoichiometric and whether legal emission limits can be met. A stoichiometric mixture is described with an actual lambda value of 1 and is created during combustion with a mass fraction of fuel to approximately 14.7 mass fractions of air (common standard fuel).

[0053] The lambda controller is a calculated variable in the engine control system that ensures that the actual lambda value always corresponds to the target lambda value and that stoichiometric combustion takes place. If the air-fuel mixture is too lean, the lambda controller increases the amount of fuel supplied to the combustion chamber. If the air-fuel mixture is too rich, the amount of fuel is reduced.

[0054] The lambda adaptation is a learned variable of the engine control, which reflects the permanent control deviation of the lambda controller.

[0055] The evaluation factor LIM is determined according to the formula LIM=λSZ1*(LRSZ1+LASZ1)*(ΔLR+ΔLA+1), where: λ SZ1 the mean value of the actual lambda value λ in the first steady state SZ1, LR SZ1 the mean value of the lambda controller LR in the first steady state SZ1, LA SZ1the mean value LA of the lambda adaptation in the first steady state SZ1, ΔLR is the difference between the mean values ​​of the LR values SZ1 , LR SZ2 of the lambda controller between the first and the second stationary state SZ1, SZ2 (ie LR SZ2 -LR SZ1 ), ΔLA is the difference between the mean values ​​of the LA values SZ1 , LA SZ2 the lambda adaptation between the first and the second stationary state SZ1, SZ2 (ie LA SZ2 -LA SZ1 ) are.

[0056] This formula calculates the mean value of the actual lambda value λ SZ1 at the stationary state SZ1 (ie during the reduction of the intake manifold pressure difference) with the sum of the mean values ​​LR SZ1 , LA SZ1 of lambda controller and lambda adaptation in the steady state SZ1. The mean values ​​LR SZ1 , LA SZ1The lambda controller and the lambda adaptation are recorded in a so-called averaging period of the first steady state SZ1, which occurs after a preceding stabilization period. The averaging period in the first steady state SZ1 includes the time difference between t2 and t3 ( Fig. 3). The averaging period of the second steady state SZ2 comprises the time difference between t4 and the end of the second steady state SZ2 (not shown). The stabilization period in the first steady state SZ1 comprises the time difference between t1 and t2 ( Fig. 3). The stabilization period in the second steady state SZ2 comprises the time difference between t3 and t4. During the stabilization period, the values ​​of the lambda controller LR and the lambda adaptation LA have not yet stabilized. After the stabilization period, stable values ​​can be assumed. This procedure ensures that the mean values ​​are not transiently distorted.

[0057] The term thus determined is then multiplied by the sum of the differences between the lambda controller and the lambda adaptation, which result between the stationary states SZ1 and SZ2. The respective difference results from the subtraction of the mean values ​​of the lambda adaptation and the lambda controller (ie ΔLR = LR SZ2 - LR SZ1 and ΔLA = LA SZ2 - LA SZ1 ).

[0058] Since in the second steady state SZ2 there is an increased vacuum in the intake manifold, the lambda controller performs a jump in the event of a leak and increases its value, since unregistered false air is now entering the system. This can, for example, be the Fig. 3, which shows the time course of the measured values ​​when a leak is present. Fig. As can easily be seen from Figure 3, the lambda controller LR reacts at time t1, i.e., with the onset of the intake manifold pressure variation. A further change in the value LR of the lambda controller can then be detected at time t3, i.e., with the transition from the first steady-state state SZ1 to the second steady-state state SZ2. The lambda adaptation LA, which reflects a permanent control deviation of the lambda controller and is determined by extended low-pass filtering of the signal from the lambda controller LR, shows no significant changes. In contrast, the actual lambda value λ reacts according to the change in the signal from the lambda controller LR.

[0059] In contrast, Fig. 2 For comparison, the time course of the measured values ​​described above when there is no leak in the intake air duct. Fig. 2 shows that without a leak in the intake air duct, the jump in the signal of the lambda controller LR does not occur, since no false air enters the intake air duct.

[0060] An intake air duct without leaks has an actual lambda value of approximately 1 in stoichiometric operation. The values ​​of the lambda controller LR and the lambda adaptation LA are also neutral at 1 (no intervention in the injection times), which results in a theoretical evaluation factor LIM = 2 according to the above formula. Tests have shown that the actual evaluation factor LIM for vehicles without leaks in the intake air duct varies slightly around this theoretical value LIM = 2, approximately in the range between 1.95 and 2.1. In contrast, the evaluation factor LIM for vehicles with leaks can increase to as much as 3.1.

[0061] It should be noted that the method can also be used for non-stoichiometric engine concepts, such as diesel engines, lean-burn (gasoline) engines, etc., in which case different lambda values, controller and adaptation values ​​result in a different theoretical evaluation factor than LIM = 2 for an intake air flow without leakage.

[0062] Through tests, a first threshold value SW1 was established for a suspected leak in the intake air duct, and a second threshold value SW2 was established for a clear conclusion about a leak in the intake air duct. By comparing the value of the evaluation factor LIM determined according to the above formula with the previously determined threshold values ​​SW1 and SW2, a conclusion can then be drawn about a suspected or definitely existing leak in the intake air duct.

[0063] A more precise statement regarding the presence of a leak as well as additional information about the location of the leak in the intake air duct is obtained if a difference between the intake air mass in the second steady state SZ2 and the first steady state SZ1 is also determined. In the event of a leak in the intake air duct, the intake air mass, which is determined by means of sensor 4, is lower than in the steady state SZ1, since in the second steady state SZ2 there is already sufficient air for combustion due to the increased false air in the intake air duct. As a result, the throttle valve 12 will reduce its opening angle in order to supply less air to the third section A3 in the intake air duct. Threshold values ​​can be determined through tests; if these values ​​are exceeded, a leak is suspected, whereby an additional threshold value can be used to reliably conclude that there is an uncertainty.The determination is thus again made by comparing the result of the determined difference value of the air mass with the previously determined threshold values.

[0064] Tests have shown it to be useful to use two threshold values ​​to assess leaks: If the difference in the air mass DIFF_HFM is greater than a third threshold value SW3 (which is selected, for example, to be SW3 = -0.5 kg / h), the intake air duct is assumed to be leak-tight. If the difference in the air mass DIFF_HFM is less than a fourth threshold value SW4 (which is selected, for example, to be SW4 = -1 kg / h), a leak of more than 20 l / min in the intake air duct is assumed. If the difference in the air mass DIFF_HFM is between the third and fourth threshold values, ie SW4 < DIFF_HFM < SW3, a leak is suspected depending on the value of the evaluation factor LIM.

[0065] For specific leak locations in the intake air duct, different values ​​are set for the LIM evaluation factor and the air mass differential. The various pairings can be stored in a table or matrix for computer-aided comparison with the values ​​determined during the engine test. This can then narrow down the likely location of a leak.

[0066] In general, it may also be necessary to specify predetermined thresholds for the LIM evaluation factor and the thresholds for the air mass differential values ​​depending on the oxygen content in the fuel. Fuel grades with different octane ratings and ethanol contents are available in different countries. Some fuels are available with 85% (E85) and 100% (E100) ethanol content. A sensor can detect the ethanol content of the known fuel. Above 50% ethanol content, the engine becomes so lean that it would no longer be able to run without adjustments. Since ethanol contributes its own oxygen to the combustion process, the stoichiometric air ratio changes depending on the ethanol content.

[0067] The combination of the LIM assessment factor and the air mass difference serves to consolidate the damage assessment and avoid misinterpretations.

[0068] Fig. Figure 4 shows a flowchart illustrating the essential steps for determining a leak in the intake air duct of an internal combustion engine. In step S41, the intake manifold pressure differential is reduced. In step S42, the intake manifold pressure differential is increased. In step S43, the measured values ​​are evaluated as described above.

[0069] Fig.Figure 5 shows a more detailed flowchart illustrating the steps for determining a leak in the intake air duct of the combustion engine. After the engine test has started, step S51 is used to wait until the catalytic converter 8 has heated up sufficiently and is ready for operation. To minimize the influence of the operating history of the vehicle under test, step S52 optionally deletes the lambda adaptations (reset value = 1) at the beginning of the measurements if this makes sense in relation to surrounding diagnostic procedures. In step S53, the tank ventilation valve is closed. During normal operation, the escaping fuel gases from the tank, as well as the blow-by gases, are recirculated to a collector behind the throttle valve 12 and would thus potentially distort the responses of the lambda controller. In step S54, the idle speed is set to a fixed value to stabilize the engine.In step S55, the engine fan is activated to keep the engine at low temperatures for as long as possible. The reason for this is that as the engine warms up, leakage rates decrease and thus become more difficult to detect. In step S56, the intake manifold pressure difference is reduced and the measured values ​​λ are determined. SZ1 , LR SZ1 , and LA SZ1 , in order to be increased in step S57 and to be able to determine a change in the intake false air mass flow and thus a jump in the lambda controller (ΔLR, ΔLA). In step S58, the evaluation factor LIM is then determined and evaluated. List of reference symbols 1 intercooler 2 diverter valve 3 intake silencers 4 hot-film air mass meter (sensor) 5 exhaust gas turbochargers 6 Wastegate valve 7 Lambda sensor in front of the catalytic converter 8 Catalyst 9 Lambda sensor after the catalyst 10 digital engine electronics (DME) 11 Intake manifold pressure sensor 12 Throttle valve 13 Charge air temperature pressure sensor A1 first section of the intake air duct A2 second section of the intake air duct A3 third section of the intake air duct SZ1 first steady state SZ2 second steady state LR value of a lambda control LA value of a lambda adaptation PSR intake manifold pressure S41 - S43 Process step S51 - S58 Process step

Claims

[1] Method for determining a leak in an intake air duct of an internal combustion engine having at least one combustion chamber, wherein the intake air duct is divided into at least two sections, comprising a section (A1, A2) upstream of a throttle valve (12) and a section (A3) downstream of the throttle valve (12), in which - as part of an engine test, an intake manifold pressure variation is carried out with a first stationary state (SZ1), in which a first intake manifold pressure is set in the intake air duct, and a second stationary state (SZ2), in which a second intake manifold pressure is set in the intake air duct; - a weighting factor (LIM) is determined for the different intake manifold pressures in the first and second stationary states (SZ1, SZ2) according to a predetermined calculation rule which processes the actual lambda value (λ), the value (LA) of a lambda adaptation and the value (LR) of a lambda controller; and - a leak in the intake air duct is concluded from the result of a comparison of the determined value of the evaluation factor (LIM) with a predetermined threshold value (SW1, SW2). [2] Method according to claim 1, in which in the first stationary state (SZ1), starting from a reference intake manifold pressure difference when the engine is idling, an intake manifold pressure difference which is smaller than the reference intake manifold pressure difference is brought about, and in the second stationary state (SZ2) an intake manifold pressure difference which is larger than the reference intake manifold pressure difference is brought about. [3] Method according to claim 1 or 2, wherein the first stationary state (SZ1) is temporally prior to the second stationary state (SZ2). [4] Method according to one of the preceding claims, in which the intake manifold pressure variation is carried out in the section (A3) after the throttle valve (12). [5] Method according to one of the preceding claims, in which the calculation rule comprises a change in the value (LR) of the lambda controller and the value (LA) of the lambda adaptation between the first and the second stationary state (SZ1, SZ2). [6] Method according to one of the preceding claims, in which the mean value of the lambda actual value (λ) is processed in the predetermined calculation rule during an intake manifold pressure caused by the intake manifold pressure variation. [7] Method according to claim 6, wherein the mean value of the actual lambda value in the first steady state (SZ1) is processed in the predetermined calculation rule. [8] Method according to one of the preceding claims, in which in the predetermined calculation rule respective mean values ​​of the value (LR) of the lambda controller and of the value (LA) of the lambda adaptation in the first stationary state (SZ1) and in the second stationary state (SZ2) are processed. [9] Method according to claim 8, wherein the mean values ​​are collected from the time-dependent measured values ​​in a respective averaging period of the first and second stationary states (SZ1, SZ2) which follow a respective preceding stabilization period. [10] Method according to one of the preceding claims, in which in the predetermined calculation rule respective difference values ​​of the value (ΔLR) of the lambda controller and the value (ΔLA) of the lambda adaptation between the assigned mean values ​​in the first stationary state (SZ1) and in the second stationary state (SZ2) are processed. [11] Method according to one of the preceding claims, in which the evaluation factor (LIM) is calculated according to the following formula, LIM=λSZ1*(LRSZ1+LASZ1)*(ΔLR+ΔLA+1), where: λ SZ1 the mean value of the actual lambda value in the first steady state SZ1, LR SZ1the mean value (LR) of the lambda controller in the first steady state SZ1, LA SZ1 the mean value (LA) of the lambda adaptation in the first steady state SZ1, ΔLR is the amount of the difference between the mean values ​​of the lambda controller values ​​between the first and the second stationary state SZ1, SZ2 (LR SZ2 - LR SZ1 ), ΔLA is the amount of the difference between the mean values ​​of the lambda adaptation between the first and the second stationary state SZ1, SZ2 (LA SZ2 - LA SZ1 ). [12] Method according to one of the preceding claims, in which, when a first threshold value (SW1) is exceeded, a conclusion is drawn as to a suspected leak in the intake air duct, in particular the section (A3) after the throttle valve (12), and, when a second threshold value (SW2) which is higher than the first threshold value is exceeded, a conclusion is drawn as to a reliably detected leak in the intake air duct, in particular the section (A3) after the throttle valve (12). [13] Method according to one of the preceding claims, in which a difference between the intake air mass in the second stationary state (SZ2) and the first stationary state (SZ1) is processed in the decision as to whether there is a leak in the intake air duct, and a leak in the intake air duct is concluded from a result of a comparison of the determined difference value of the air mass with a predetermined threshold value. [14] Method according to one of the preceding claims, in which the predetermined threshold values ​​are determined as a function of an oxygen content in the fuel. [15] A computer program product capable of being loaded directly into the internal memory of a digital computer and comprising software code portions for carrying out the steps of any preceding claim when the product is run on a computer. [16] System for determining a leak in an intake air duct of an internal combustion engine having at least one combustion chamber, wherein the intake air duct is divided into at least two sections, comprising a section (A1, A2) in front of a throttle valve (12) and a section (A3) after the throttle valve (12), which is designed to - to carry out, as part of an engine test, an intake manifold pressure variation with a first stationary state (SZ1), in which a first intake manifold pressure is set in the intake air duct, and a second stationary state (SZ2), in which a second intake manifold pressure is set in the intake air duct; - to determine an evaluation factor (LIM) for the different intake manifold pressures in the first and second stationary states (SZ1, SZ2) according to a predetermined calculation rule which processes the actual lambda value (λ), the value (LA) of a lambda adaptation and the value (LR) of a lambda controller; and - to conclude that there is a leak in the intake air duct from the result of a comparison of the determined value of the evaluation factor (LIM) with a predetermined threshold value (SW1, SW2). [17] A system according to claim 16, comprising further means for carrying out the method according to any one of claims 2 to 14.

Images