EP3526456B1 - Method for starting an internal combustion engine - Google Patents

Description

The invention relates to a method for starting an internal combustion engine by means of a compressed air starter system, in which the starter is engaged by means of compressed air in a first start sequence and the starter is acted upon with compressed air in a second start sequence.

An internal combustion engine is started either by means of an electrically operated starter or by means of a compressed air starter. Compressed air starting systems are for example from the DE 26 32 015 OS that U.S. 3,667,442 A and the U.S. 4,494,499 A known. Typically, a starting process in a compressed air starting system consists of a first and a second starting sequence. In the first start sequence the starter is engaged by means of compressed air and in the second start sequence the starter is set in a rotary motion by means of the compressed air. The second start sequence is ended when the internal combustion engine has reached an idling speed, for example 350 revolutions / minute. Then the internal combustion engine operation begins by injecting the fuel. In an internal combustion engine used as a ship propulsion system, the cylinders are equipped with decompression valves to relieve the cylinder working space. Any water that may have penetrated during the second start sequence is pumped away from the cylinder chamber via this. In practice, the problem arises that the starter has to apply a considerable breakaway torque in order to initially crank the internal combustion engine. If the breakaway torque is overcome, the internal combustion engine briefly rotates at high speed. In connection with residual water in the cylinder space, this is critical for the connecting rod.

The invention is therefore based on the object of providing an improved method for starting an internal combustion engine with a compressed air system.

The object is achieved in that a method having the features of claim 1 is created. Advantageous refinements result from the subclaims.

This object is achieved in particular by a method in which, in a first starting sequence, the starter is engaged by means of compressed air, a decompression valve is applied in the opening direction to relieve the cylinder working space, and the internal combustion engine is started by applying pulsed compressed air to the starter . In a second starting sequence, the decompression valve is then acted upon in the closing direction and the starter is acted upon with constant compressed air.

Here, a system controller defines a compressed air path for engaging the starter via an engagement valve and a compressed air path for cranking the starter in the first starting sequence and turning the starter in the second starting sequence is established using a starting valve. The pulsed compressed air is generated by controlling the start valve as a function of a target engine speed via a PWM signal during the first start sequence. In other words: the starter is continuously and gently turned on via the PWM signal and the pulsed compressed air. A hard transition from a stationary internal combustion engine to a rotating internal combustion engine is avoided.

In addition, it is provided that the target speed is increased in the form of a ramp from a first target speed value to a second target speed value. The first start sequence ends positively when a speed control deviation from the setpoint to the actual speed is detected within a tolerance band, for example 10 revolutions / minute.

Overall, the process offers a high level of process security and, as an additional security measure, allows sales-promoting argumentation. As a pure software solution, it is almost cost-neutral. In addition, the invention can be retrofitted without any problems, since the function only accesses the already existing components.

Figur 1

ein Systemschaubild,

Figur 2

einen Programm-Ablaufplan und

Figur 3

einen Ausschnitt aus dem Programm-Ablaufplan

A preferred embodiment is shown in the figures. Show it:

Figure 1

a system diagram,

Figure 2

a program schedule and

Figure 3

an excerpt from the program schedule

the Figure 1 shows a system diagram of an internal combustion engine 1 with a compressed air starting system 2. The compressed air starting system 2 comprises a compressed air storage device 10 for providing the compressed air, an engaging valve 5 and a starting valve 6. The engaging valve 5 and the starting valve 6 are designed as 2/2 valves. Alternatively, 3/2 valves can also be used. In the Figure 1 the engagement valve 5 is shown in position 1, so that there is a continuous compressed air path from the compressed air reservoir 10 via the engagement valve 5 to the starter 3. In this position the starter is engaged. The start valve 6 is shown in the zero position, in which the compressed air path from the compressed air reservoir 10 to the starter is blocked, that is, the starter does not rotate. The operating state of the overall system is determined by a system controller 4. An operator specifies his activation / deactivation request or his performance request via the system controller 4. A monitoring unit 7 (EMU), an interface unit 8 (EIM) and an engine control device 9 are connected to the system controller 4 via a CAN bus. The monitoring unit 7 in turn determines the switching state of the engagement valve 5 and the start valve 6. This is typically done via a PWM signal. The function of the monitoring unit 7 and the interface unit 8 are carried out in conjunction with the Figure 2 explained in more detail. The engine control unit 9 controls and regulates the state of the internal combustion engine 1. In internal combustion engine operation, these are, for example, a rail pressure, a start of injection and an end of injection. In the figure, the further input and output variables are shown with the reference symbol on / off, for example a switching signal for the switchable exhaust gas turbocharger in the case of register charging.

In the Figure 2 a program flow chart is shown. the Figure 2 consists of the sub-figures 2A, 2B and 2C. Here the Figure 2A the part of the program for preparation and testing of the start process, Figure 2B shows the program part of the first start sequence and Figure 2C the program part of the second start sequence. The program sequence in the monitoring unit 7 is identified by the reference symbol EMU. The sequence in the interface unit 8 is identified by the reference symbol EIM. The interface unit 8 (EIM) and the monitoring unit 7 (EMU) communicate via CAN bus. Information that is set or queried on the CAN bus is shown as dashed arrows. For example, in step S2A the air pressure sensor sets its status signal on the CAN bus, reference character B. This status signal is read in from the CAN bus, reference character B, in step S3 of the interface unit 8 (EIM).

In the following, the program run of the monitoring unit (EMU) is described first. With S1A, the status of the decompression valve open / closed is determined and set as a value on the CAN bus, reference number A. With S2A, the state of the compressed air sensor and the compressed air is determined and set as a status value, reference character B, on the CAN bus. Steps S3A to S8A identify an error query and indicate the operational readiness of the monitoring unit. First, a check is made at S3A to determine whether an error has been detected. If an error is detected, query result S3A: yes, an alarm is displayed at S4A and this is set on the CAN bus, reference character C, for further processing. If the fault-free condition is determined at S3A, the function release is issued at S5A, reference character C, and then at S6A the status of the engagement valve ( Fig. 1 : 5), with S7A the status of the start valve ( Fig. 1 : 6) and the status of the speed sensor is queried at S8A. A branch is then made back to step S3A. Steps S9A to S11A characterize the procedure in the event of an aborted start. At S9A it is checked whether the monitoring unit (EIM) has set a start abort on the CAN bus, reference character D. If the start abort is initiated, the start valve is then deactivated at S10A and the engagement valve at S11A and this is done on the CAN bus for further processing indicated, reference symbol E.

The program run of the interface unit (EIM) begins at S1 with the query of the start mode. This is specified by the operator via the system controller. Accordingly, either the engine start using a generator, step S2, or a start using a compressed air system is selected. The start blocking is queried at S3. For this purpose, the set status of the decompression valve (reference symbol A), the air pressure sensor (reference symbol B) and an external stop signal are queried on the CAN bus. The stop signal, reference F, is set by the system controller on the CAN bus. The result of the start blocking is then queried at S4. If a shift lock is set, the start is aborted at S9 and displayed on the CAN bus, reference number D. If there is no shift lock, then at S5 a branch is made to the oil lubrication subroutine and then at S6 a check is made to see whether the oil pressure pÖL is greater than a Limit value GW is. In the event of an error, query result S6: no, an alarm is set for the operator at S7 and a branch is made to S8. With correct oil lubrication, Query result S6: yes, it is then checked at S8 whether the monitoring unit (EMU) is ready for operation. For this purpose, the operational readiness on the CAN bus, reference character C, is read out. If it was determined at S8 that the monitoring unit (EMU) is ready for operation, then it becomes Figure 2B branched. If the test result is negative, that is, the monitoring unit (EMU) is not ready for operation, a branch is made to S9, the start process is aborted and this status is set on the CAN bus, reference character D.

the Figure 2B shows the program part of the first start sequence. In the following, the program run of the monitoring unit (EMU) is described first. At S12A it is checked whether the actual speed nIST is greater than a limit value GW. The limit value here corresponds to the maximum permissible speed during cranking, for example 20 revolutions / minute. In addition, the status of the monitoring unit (EIM), reference G, is queried. If too high an actual speed was detected, query result S12A: yes, the program branches to the program part with steps S20A to S22A. If the actual speed nIST is not greater than the limit value GW, query result S12A: no, the engagement valve is activated at S13A, whereby the starter is pressurized with compressed air and engages. At S14A, a time step is run through which corresponds to the period of the meshing. A regulation is activated at S15A. The main features of this regulation are in the Figure 3 shown. The following input variables are available at a PI controller 11: the PWM frequency fPWM for controlling the engagement valve ( Fig. 1 : 5) and the start valve ( Fig. 1 : 6), a minimum pulse-pause ratio PWM (min), a maximum pulse-pause ratio PWM (max) to control the engagement and start valve, two speed setpoints nSL1 and nSL2, a tolerance band for the speed, a proportional coefficient kp and an integral coefficient ki. Typical values for these input variables are: fPWM = 8Hz, PWM (min) = 0%, PWM (max) = 20%, nSL1 = 2 1 / min; nSL2 = 10 1 / min and tolerance band = 10 1 / min. In addition, the PI controller 11 is supplied with the actual speed nIST, the value of which is available on the CAN bus, reference symbol K ( Figure 2B ). Alternatively, the monitoring unit can also use its own speed sensor. The output variables of the PI controller 11 are the cranking status and the position of the setpoint / actual deviation dn of the speed in relation to a first limit value GW1 and a second limit value GW2.

The output variables of the PI controller are now in step S16A Figure 2B further rated. If during a time dt the speed control deviation dn lies within the tolerance band TB, query result S16A: yes, the cranking is recognized as complete at S18A and set as a data value on the CAN bus, reference character J. If, however, the speed was not yet stable at S16A If the control deviation is detected, a time step t is compared with a limit value GW at S17A. If the timer t has expired, query result S17A: yes, the program sequence is continued at S20A. If, on the other hand, the time stage t is still running, query result S17A: no, the system branches back to S15A. If the cranking was set as complete in S18A, a timer is activated in S19A. During this time step it is checked whether from the first start sequence to the second start sequence ( Figure 2C ) should be changed, whether the timer has expired without result or whether the status should be set to idle. For this purpose, the status on the CAN bus, reference L, is queried during the time stage. If the time step has expired without result or if the status idling is set, the start valve is then deactivated at S20A, the engagement valve is deactivated at S21A and the cranking is ended at S22A.

At S10, the interface unit (EIM) sets the following states on the CAN bus, reference symbol G: No injection, activate decompression valve, i.e. actuate in opening positions and turn on a state variable CTS. It is then checked at S11 whether the turning is in progress. For this purpose, the corresponding value, reference symbol H, is read in on the CAN bus. If the test result is negative, the cranking is aborted and a branch is made to S10. If the cranking was recognized as activated at S11, query result S11: yes, the status variable CTS is set accordingly at S12 and a check is made at S13 to determine whether the cranking has taken place completely. During this test, the status of the monitoring unit (EMU), reference character J, is queried. If the turning has not yet been completed, a branch is made back to S12. In addition, there is an error query, which can cause the start to be aborted. If the cranking is ended, query result S13: yes, the decision is made at S14 whether the second starting sequence is after Figure 2C should take place or whether the run variable CTS should be set to idle at S15. If the cranking is to be ended, then the status variable CTS is set to idle at S15 and, in addition, is set on the CAN bus, reference character L. Then at S16 the actual speed nIST is checked for standstill (nIST = 0). With negative Test result, that is, the internal combustion engine is already turning, the program sequence is aborted, step S19 and reference symbol M. If the test at S16 is positive, the decompression valve is actuated in the closing direction at S17 and the cranking is set to be ended at S18.

the Figure 2C shows the program parts of the second start sequence. First, the program run of the monitoring unit (EMU) is described. At S23A, the second start sequence is set and set as the status on the CAN bus, reference number N. Then, at S24A, the start valve is activated, with the pulse-pause ratio being set to one hundred percent (PWM = 100%). As a result, the starter is now fully pressurized with compressed air. At S25A it is checked whether the actual speed nIST is greater than the idling speed LL, for example LL = 350 rpm. If this is not yet the case, query result S25A: no, a time step t, for example t = 20s, is set at S26A. If this time stage has not yet expired, a branch is made back to S25A. Otherwise the program sequence is continued with S27A. If it was recognized in S25A that the actual speed is greater than the idle speed LL, then the start valve is deactivated in S27A, the engagement valve deactivated in S28A and the second start sequence set as complete in S29A, reference symbol O. In S30A, this program sequence is ended .

At S20, the interface unit (EIM) deactivates the decompression valve, that is, the decompression valve is actuated in the closing direction. At S21 the status variable CTS is set to the status Start. It is then checked at S22 whether the second start sequence is running. For this purpose, the status on the CAN bus, reference N, is taken into account. If the start process has not yet been set, a branch is made back to S21. If an error was detected in S22, the start process is aborted with S27. If it was recognized at S22 that the starting process is running, then at S23 the status variable CTS is set to Start and at S24 the starting process is set as completed. In S24, the status on the CAN bus, reference symbol O, is also taken into account. The status is then set to idling at S25, the starting process is ended with S26 and the engine is switched to operation.

Reference number

1

Internal combustion engine

2

Compressed air starting system

3

starter

4th

System controller

5

Engagement valve

6th

Start valve

7th

Monitoring Unit (EMU)

8th

Interface Unit (EIM)

9

Engine control unit

10

Compressed air storage

11th

PI controller

Claims (5)

1. Method for starting an internal combustion engine (1) by means of a compressed air starting system (2), in which in a first starting sequence an engagement of the starter (3) is effected by means of compressed air, a decompression valve for relieving the cylinder working chamber is impinged in the opening direction and a turning-on of the internal combustion engine (1) is initiated, in that the starter (3) is applied with pulsed compressed air, and in which, in a second starting sequence, the decompression valve is impinged in the closing direction and the starter (3) is applied with constant compressed air, wherein a compressed air path for engaging the starter (3) is defined by a system controller (4) via an engagement valve (5) and a compressed air path for turning-on the starter (3) in the first starting sequence and for rotating the starter (3) in the second starting sequence is defined via a starting valve (6), characterised in that during the first starting sequence the starting valve (6) is controlled as a function of a setpoint engine speed (nSL) via a PWM signal.
2. Method according to claim 1, characterised in that the setpoint engine speed (nSL) is increased in a ramp-like manner from a first setpoint speed value (nSL1) to a second setpoint speed value (nSL2).
3. Method according to claim 2, characterised in that a speed control deviation (dn) is calculated from the setpoint speed (nSL) to the actual speed (nIST) and the first starting sequence is terminated positively when the speed control deviation (dn) is determined within a tolerance band (TB).
4. Method according to claim 3, characterised in that a time period of the speed control deviation (dn) is additionally checked.
5. Method according to one of the preceding claims, characterised in that during the second starting sequence the actual speed (nIST) is compared with a no-load value (LL), when the no-load value (nIST>LL) is exceeded the second starting sequence is terminated positively and a change is made to internal combustion engine operation.

Images