DK178172B1

DK178172B1 - Improved monitoring of wear of bearings in a large two stroke diesel engine

Info

Publication number: DK178172B1
Application number: DK201000401A
Authority: DK
Inventors: Henrik Willads Christensen; Martin Olsson
Original assignee: Man Diesel & Turbo Deutschland
Priority date: 2010-05-07
Filing date: 2010-05-07
Publication date: 2015-07-20
Also published as: FI20125874L; JP5746547B2; FI123312B; FI125410B; FI20115427A; CH702986B1; KR101693755B1; JP2011237420A; CN102288408B; CH702986A2; DK201000401A; KR20110123669A; CN102288408A; FI20115427A0

Abstract

An apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre level in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition such as a speed; and determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold.

Description

IMPROVED MONITORING OF WEAR OF BEARINGS IN A LARGE TWO

STROKE DIESEL ENGINE

FIELD

The present application relates to a device and a method for measuring wear of bearings in a large two stroke diesel engine.

BACKGROUND

The main bearings, the crankpin bearings and the crosshead bearings of large two stroke diesel engines have a bottom and a top bearing shell that define a bearing surface for supporting a corresponding bearing journal. The bearing shells have a steel back with a layer of bearing metal thereon. The bearing metal is typically a tin-aluminium alloy or a white metal alloy in a layer of approximately 1,5 to 2mm thick.

Despite great care in production of these engines and operation there is a small failure rate. Since most large two stroke diesel engines are ocean going vessels engine failure due to a faulty bearing is highly undesirable.

Further, the costs of a bearing overhaul can be major, requiring for instance a regrinding of the crankshaft, which can lead to a substantial down time for the engine.

To prevent this from happening it is required by the classification societies to perform inspections every four or five years depending on the classification society, during which the bearings are dismounted and inspected.

These inspections are rather complex and require dismantling the bearings and then re-assembling the bearings again. This process includes a substantial risk for a build error, i.e. an error introduced during the overhaul.

To alleviate the need for inspections the prior art has proposed to monitor the wear of a bearing by measuring its distance.

US patent application US2007/0017280 discloses measuring the position of a guide shoe of a crosshead by using two sensors for each cylinder. This enables detection of wear in the main bearings, the crank-pin bearings and the cross-head bearings and also to determine if the wear is in the main bearings or in the crank-pin bearings and the cross-head bearings. These measurements are used to determine whether the wear is about to lead to failure or not and if a failure is imminent issue a warning.

As has been noted above the conseguences of a failure can be dire and an efficient and accurate manner of sensing the wearing of a bearing is thus of great use.

SUMMARY

On this background, it would be advantageously to provide a device and a method that improves the efficiency and accuracy of the prior art systems by providing a method and a device designed to calibrate and/or to interpret the sensor readings in an improved manner.

The primary aim of monitoring the wear of bearings is to detect a bearing failure before it develops to an extent where heat is causing damage to other parts than the bearing lining. Parts that could be affected are the crosshead, crank-pin, main bearing journal, or the bearing housing due to distortion. Such damage will generally be the result if the bearing lining is worn through and contact between the shaft and the bearing shell steel backing occurs.

It should be noted that even though the description below is focused on compensating for an engine speed the teachings herein also apply to compensating for a propeller pitch level and a load. These are examples of engine operating conditions affecting the bottom dead centre (BDC) for a movable part, such as a cross head or a guide shoe, to mention some examples, in an engine.

The disclosed embodiments provide an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe of a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine a deviation to accommodate for factors influencing the sensor signal; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold.

In one embodiment the engine operating condition is an engine speed.

In one embodiment the engine operating condition is a propeller pitch level.

In one embodiment the engine operating condition is a load.

By determining a deviation of a sensor signal the controller is able to accommodate for external influences that affect more than one sensor and that would otherwise be interpreted as a change in BDC level.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining a deviation to accommodate for factors influencing the sensor signal; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to change the threshold value dynamically according to a current operation of the engine.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre (BDC) for the crosshead of a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises changing the threshold value dynamically according to a current operation of the engine .

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine and wherein said at least two sensors are arranged so that there are two sensors in each cylinder, one fore and one aft sensor, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to: determine that a bearing is worn by combining the signal of the fore sensor of a first cylinder to the signal of the aft sensor of a second cylinder, and/or determine that a bearing is worn by combining the signal of the fore sensor of a first cylinder to the signal of the aft sensor of the same cylinder and to combine the two sensor signals and compare the result of the two sensor signals with said threshold level.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises determining that a bearing is worn out or in a critical condition by combining the signal of the fore sensor of a first cylinder to the signal of the aft sensor of a second cylinder, and/or determining that a bearing is worn out or in a critical condition by combining the signal of the fore sensor of a first cylinder to the signal of the aft sensor of the same cylinder and combining the two sensor signals and comparing the result of the two sensor signals with said threshold level.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to: detect a rapid change in BDC level by comparing the received sensor signals to an average level of signal values for a previous time period.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in an engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to: detect a rapid change in bottom dead centre level by updating a reference level for a sensor; updating a present level for a sensor and determining a reference value and determining whether the reference value exceeds a rapid change threshold level.

The aspects of the disclosed embodiments are also directed to providing a method for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in an engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to: detect a rapid change in bottom dead centre level by updating a reference level for a sensor; updating a present level for a sensor and determining a reference value and determining whether the reference value exceeds a rapid change threshold level.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises detecting a rapid change in BDC level by comparing the measured sensor values to an average level for a previous time period.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to generate an engine speed compensation table during operation of said engine during a learning phase by sampling a first number of sensor values in relation to a second number of speed points during operation of said engine and as the first number of samples have been received for a speed point determine a reference value by averaging the received sample values for that speed point.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises generating an engine operating condition compensation table during operation of said engine during a learning phase by sampling a first number of sensor values in relation to a second number of speed points during operation of said engine and as the first number of samples have been received for a speed point determine a reference value by averaging the received sample values for that speed point.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said controller is further configured to readjust a replaced or adjusted sensor by adjusting the signals from a readjusted or replaced sensor by compensating the signal value according to a operating condition compensation look-up table by calculating an average of an offset for the readjusted or replaced sensor over a period of time and offset the reference values for an affected sensor according to the calculated average offset.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises readjusting a replaced or adjusted sensor by adjusting the signals from a readjusted or replaced sensor by compensating the signal value according to an operating condition compensation look-up table by calculating an average of an offset for the readjusted or replaced sensor over a period of time and offset the reference values for an affected sensor according to the calculated average offset.

The aspects of the disclosed embodiments are also directed to providing an apparatus for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said apparatus further comprises a controller configured to: receive a signal from each sensor; compensate each signal depending on an engine operating condition; determine whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold, and wherein said controller is further configured to: generate trend curves showing if any wear has taken place during a time interval.

The aspects of the disclosed embodiments are also directed to providing a method for implementation in an apparatus having a controller arranged to execute instructions stored on a physical medium, said method being for monitoring wear of a crosshead bearing, a crankpin bearing and a main bearing in a large two stroke diesel engine, said apparatus comprising at least two sensors, wherein said sensors are arranged and configured for measuring a bottom dead centre for the crosshead or a guide shoe in a given cylinder relative to a fixed point of the engine, wherein said method comprises: receiving a signal from each sensor; compensating each signal depending on an engine operating condition; determining whether a threshold value has been exceeded and if so issue an indication of the exceeded threshold and wherein said method further comprises generating trend curves showing if any wear has taken place during a time interval.

The aspects of the disclosed embodiments are also directed to providing an engine comprising any apparatus according to above.

The aspects of the disclosed embodiments are also directed to providing a marine vessel comprising an engine according to above.

The aspects of the disclosed embodiments are also directed to providing a computer readable medium including at least computer program code for monitoring wear of a bearing in a large two-stroke diesel engine, said computer readable medium comprising software code for executing any one or a plurality of the methods according to above.

Further objects, features, advantages and properties of the apparatus, method and computer readable medium according to the present application will become apparent from the detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following detailed portion of the present description, the teachings of the present application will be explained in more detail with reference to the example embodiments shown in the drawings, in which:

Fig. 1 is diagrammatic view of an engine according to an embodiment,

Fig. 2 is a schematic view of a bearing according to an embodiment,

Fig. 3 a flow chart describing a general method according to an embodiment,

Fig. 4 is a flow chart describing a method according to an embodiment,

Fig 5 is a flow chart describing a method according to an embodiment,

Fig. 6 is a diagram showing a curve according to an embodiment,

Fig. 7 is a diagram showing a curve according to an embodiment,

Fig. 8 is a flow chart describing a method according to an embodiment,

Fig. 9 is a flow chart describing a method according to an embodiment, and

Fig. 10 is a diagram showing a curve according to an embodiment.

DETAILED DESCRIPTION

In the following detailed description, the device, the method and the software product according to the teachings of this application will be described by the embodiments in relation to a large two-stroke diesel engine. It should be noted that although only a two-stroke engine is described the teachings of this application can also be used in any engines such as four-stroke engines, two-stroke petrol engines and small two-stroke diesel engines.

FIG. 1 is a cut-through view of an engine 100. The engine has a crankshaft having a main journal 110 and one or more crank-pins 120 to which each a crosshead 130 is connected.

For each cylinder there are at least three bearings, a main bearing 140 (not visible in fig 1 which is indicated by the dashed line), a crank-pin bearing 150 and a crosshead bearing 160.

Two sensors 170 (only one is shown) are arranged for each cylinder. In one embodiment one sensor 170 is arranged aft of the crosshead 130 and one sensor 170 is arranged fore of the crosshead 130. The sensors 170 are arranged to measure a distance between the crosshead 130 and a fixed point. Alternatively the sensors are arranged to measure a distance between a point attached to the crosshead and to a fixed point.

Alternatively, the sensors are arranged to measure the distance for a guide shoe 135.

In figure 1 a sensor 170 is arranged on the engine structure 100 and measures the distance to a reflective plate 175 attached to the guide shoe 135. It should be noted that there exist different alternatives to what kind of sensor is used, how it is arranged and which distance it measures.

In a preferred embodiment the sensor 170 is arranged on the engine structure to generate the most reliable readings as the sensor is kept stationary.

In one embodiment the sensor plate 175 is arranged on the crosshead pin to ensure a reading with reduced variation due to angular movement of the guide shoe 135.

In one embodiment the sensors are arranged to determine the bottom dead centre level for a structure (for example the guide shoe as described above, but other structures are also possible) in the engine that is attached to or moving with the crosshead wherein an indirect reading of the bottom dead centre for the crosshead is determined. Thus measuring the bottom dead centre level for such a structure allows for determination of deviations and changes in bottom dead centre levels for a crosshead.

Alternatively the sensors 170 may be arranged differently and they may even be arranged in positions not in line with the shaft. Preferably though, they are arranged parallel with the shaft. Also, the sensors are preferably arranged axially spaced along a parallel to the shaft as this allows for determining that the wear is in the main bearing 140.

The sensors are thus configured to provide a measurement of the level of the lowest point, i.e. the Bottom Dead Centre for the cross-head 130 during one revolution of the engine 10 0.

As would be clear to a skilled person there are a number of alternatives for how to implement a sensor 170, such as an optical proximity sensor or a capacitance-measuring proximity sensor.

Fig 2. illustrates the general construction of a bearing 200 having a core 210 of a hard metal, such as steel, and a lining wall 220 of a softer metal, such as white metal or tin-aluminium.

The thickness of the lining is usually in the range 1.0 to 1.5 mm and the sensors 170 are designed to measure the distance or level repeatedly with an accuracy of +/- 0.01 mm.

A controller is connected to the sensors 170 and is arranged to receive the measurements from the sensors 170 and to process them to measure the wear of the bearings. The measuring of the wear is based on the fact that any change in bearing wall 220 thickness in the loaded part of one of these bearings 140, 150, 160 (due to wear or seizure) will result in a corresponding change of Bottom Dead Centre (BDC) level of one or more of the crossheads 130 relative to the structure of the engine 100. In case of crank pin 150 or crosshead bearing 160 wear, the BDC level of the cylinder in question will change, while in case of main bearing 140 wear, the BDC level of one sensor 170 in one cylinder and one sensor in the adjacent cylinder will change.

The signal from the sensors 170 will contain a certain level of noise which will affect the accuracy of the sensors' reading.

One other factor that affects the accuracy of the sensors is that the actual guide shoe or crosshead BDC level will vary slightly due to minor irregularities in engine parameters like firing pressure etc.

One other factor that affects the signal is the engine speed that i.e. the revolutions per minute (rpm) for the engine 10 0.

One other factor that affects the signal is the propeller pitch level.

One other factor that affects the signal is the load subjected to the engine. One example of such load is the cargo load carried by a vessel.

These are all examples of engine operating conditions and it should be noted that even though the below description is focused on compensating for engine speeds, the teachings may also be applied to compensating for the other operating conditions, alone or in combination.

Other factors such as hull deformation from different vessel draft and engine temperature may also have some effect on the BDC level as will variations in the manufacturing process of various engine parts.

Factors such as signal scatter of the bearings may also have an effect on the measured BDC levels.

The method and apparatus as disclosed herein are arranged to take factors such as above and others into account and provide a reliable measurement of the wear of bearings in an engine.

Figure 3 shows a flow chart for a general method of the embodiments herein.

In a first step 310 signals form the sensors 170 are acquired by the controller. The signals are then processed 320 by compensating for various factors and corrected by calculating deviations and the results are evaluated 330 upon which it is determined whether a warning should be given or not.

There are many factors affecting the wear of a bearing in an engine during operation. If an engine has a failure it is important to isolate the source for the breakdown. In the case of wearing of a bearing it is important to understand the cause of the increase wear of the bearing as a new replacement bearing would otherwise most likely also be worn out prematurely.

As is previously known and mentioned above the use of two sensors for each cylinder allows an operator to isolate the location of the bearing that is being worn out. Identifying which bearing needs to be replaced does help with the replacement as the engineers will now where to open up the engine and does not need to do unnecessary work trying to find which bearing.

However, as was realized by the inventors of this application, further information may be extracted from these two sensor readings which information is useful for identifying also the cause of the (increased) wear and this will allow an engineer to remedy the problem and prevent increased wear for the replacement bearing.

The further information may also be used to accommodate for situations where several or all of the sensors are affected without any damage being present.

As will be shown below, the extracted information can also be used to prevent that a bearing is worn out by giving an early warning which will allow operators to remedy the cause of the wearing before the bearing is worn out thereby preventing an engine failure and resulting down time for the engine.

The apparatus and method of this disclosure is thus highly advantageous as it allows for identifying the cause for the increased wear, it is able to predict failures and doing this without additional sensors.

The system can therefore easily be integrated into existing engines having two sensors installed by simply connecting an upgraded controller or possibly by updating the existing controller.

The most significant influence on the BDC level is the engine speed and the signal value received from the sensor (s) must be compensated for the engine speed or as the engine speed increases the BDC level will be lower and the wearing monitoring will falsely signal that the bearings are worn.

A variation of over 0.3 mm can exist in the BDC level depending on which engine speed is used, depending on engine size. A compensation for this influence is therefore necessary. Since the influence is individual for each installation, depending on layout and alignment of shaft line etc., the compensation must, for each sensor, be established individually for each installation .

The signal value is compensated for the engine speed through the use of a look up table having BDC levels for each sensor for a wide range of speeds.

In one embodiment a controller is configured to compensate for the speed based on a look-up table containing the normal or average signal values in dependency of engine speed or rpm (reference values).

The normal value may change significantly within small change in engine speed.

In one embodiment the look-up table has a resolution corresponding to dividing the speed range of nominal engine speed into sub-ranges, which are hereafter referred to as speed points.

In one embodiment the speed range corresponds to the range of 0% to 120% of the nominal engine speed.

In one embodiment the speed range corresponds to the range of 20% to 110% of the nominal engine speed.

In one embodiment the range below 20% of nominal engine speed is not considered for monitoring wearing in bearings .

In one embodiment the controller is configured to take into account whether the engine is being reversed or not.

Due to the speed compensation the measurement that is considered by the controller is not the actual BDC level, but the difference between the measured BDC level and the average BDC level at the current engine speed. This allows for detecting an abnormal or unexpected change in BDC level which is due to wearing of the bearings.

Thus processing of the signal received from the sensor should be based on a compensated value for the signal.

In one embodiment the controller is configured to determine a compensated signal value by subtracting a reference value from the received signal value:

Scomp — Sp — Sref

In one embodiment the controller is arranged to filter the sensor values.

To minimize the influence of noise the values provided by the sensors 170 are filtered in one embodiment using a low pass filter.

In one embodiment a simple filter wherein a moving average is updated for each sensor value, that is, once for each engine revolution. In one embodiment the filter is expressed as:

New Filtered Value = Old Filtered Value*(1-x) + Value * x

The value of x affects how quickly the controller will react to changes. A high value will cause the controller to be very sensitive to changes and therefore also to noise, whereas low values of x will result in slow reaction to an actual event. In one embodiment X is 0.05.

Some factors affect all sensors in common and the signals received from the affected sensors may be interpreted incorrectly.

One such a factor is a changed engine temperature.

To differentiate between such changes and a change due to an increased wear of a bearing a deviation for each sensor value is calculated.

In one embodiment the controller is configured to calculate a deviation of a sensor's signal compared to the other sensors' values by subtracting the average of the other signals from the individual signal.

The sensor deviation (d(S±)) for sensor 5 (S 5) from a total set of 8 sensors Si to S8 (corresponding to a four cylinder engine) can be expressed as: d(Ss) = S5 - (Si + S2 + S3 + S4 + S5 T S7 + Sg) /7

One other factor is the influence of a longitudinal shear formed deformation of the engine structure which is sometimes caused by a variation in thrust. In some cases this deformation causes the fore and aft sensor in a cylinder to show variations with regards to one another.

Through scrutinised monitoring and investigation of engines during operation and careful analysis of the findings it has been concluded that the variation for the two sensors is that the signal values are often in counter phase. This causes the signal value of one sensor to increase as the signal value for the other sensor decreases and vice versa. This is hereafter referred to as cylinder deviation.

In one embodiment the controller is configured to eliminate this variation by determining an average sensor value for the two sensors of one cylinder and compare them to the average of the other cylinders' sensors.

The cylinder deviation (d(cyli)) for cylinder 3 (cyls) from a total set of 4 cylinders together having 8 sensors Si to Ss can be expressed as: d(cyl3) = (S s + Sa)/2 - (Si + S2 + S3 + S4 + S7 + Ss)/6

Both the cylinder and sensor deviation reduces the influence of signal scatter. However, cylinder deviation is more efficient in reducing the signal scatter.

Calculating the cylinder deviation is thus useful for reducing the signal scatter.

Cylinder deviations are less sensitive for situations where only one sensor registers damage to a bearing as the registered change would be halved during the deviation calculation.

Therefore it is preferred to calculate both cylinder and sensor deviation to allow for a reliable detection of damage in a bearing.

In one embodiment a controller is configured to calculate both cylinder and sensor deviation.

One example of a situation where more than one sensor is affected is when more than one bearing is damaged. This may be the result of a common erosive process deteriorating several or all bearings. Such a situation may be the result of polluted oil or oil supply failure.

Such a situation will not be easily detectable from the cylinder and sensor deviations as the effects would be cancelled out by the averaging process of these calculations .

However, it is possible to detect such a situation from the engine speed compensated value for each signal.

In one embodiment a controller is configured to calculate cylinder deviation, sensor deviation and an engine speed compensated signal value.

In an embodiment according to above a controller is thus configured to extract information from the signal values taking into account several factors that affect the reliability of the sensor readings.

A controller of an apparatus according to herein is thus able to generate reliable readings for monitoring the wear of a bearing in an engine.

In one embodiment a controller is configured to evaluate the readings and determine if damage is imminent or about to occur.

In one embodiment a controller is configured to compare the current reading for a cylinder or bearing with an average value taken over a time period. In one embodiment the time period is 6 hours. In one embodiment the controller is configured to determine that the difference between the current reading and the time average is greater than a threshold level and if so activate a prewarning action. In one embodiment the pre-warning limit is +/- 0.25 mm.

If the pre-warning level is exceeded this is an indication that the condition of a monitored bearing is changing.

In one embodiment the controller is configured to store the pre-warning in a log file.

In one embodiment a controller is configured to receive input to reset the pre-warning action.

As the condition of the bearing is only detected to have started to change the engine may very well not be in any danger for at least several more hours and it is therefore possible to reset the pre-warning action and continue with the operation of the engine.

In one embodiment a controller is configured to compare the current reading for a cylinder or bearing with a alarm threshold. If such a threshold is exceeded the controller is configured to activate an alarm.

An alarm is an indication that the affected bearing is starting to wear out or close to being worn out and should be inspected.

In one embodiment the alarm threshold value for a sensor value is + /- 0.5 mm.

In one embodiment the alarm threshold value for a sensor deviation is + /- 0.4 mm.

In one embodiment the alarm threshold value for a cylinder deviation is + /- 0.3 mm.

In one embodiment the controller is configured to store the alarm action in a log file.

Should a larger change be detected a request to slow down the engine speed can be issued. As the engine speed is the greatest influence on the bearing load and the wear of the bearings it is in some situations possible to delay or prevent an engine failure resulting from damaged bearings by slowing down the engine. In severe cases the engine will have to be stopped completely to prevent failure .

In one embodiment a controller is configured to compare the current reading for a cylinder or bearing with a slow-down threshold. If such a threshold is exceeded the controller is configured to issue a request for a slow down. In one embodiment the threshold value for a sensor value is +/- 0.7 mm. In one embodiment the slow-down threshold value for a sensor deviation is + /- 0.5 mm.

In one embodiment a controller is configured to automatically slow down the engine speed upon such a slow-down reguest.

In one embodiment a controller is configured to automatically stop the engine upon such a slow-down reguest.

In one embodiment the controller is configured to store the slow-down reguest in a log file.

Fig 4 illustrates a general method of monitoring the wear of a bearing according to an embodiment of the present application .

In one embodiment a controller is configured to perform the steps of fig 4 in parallel for all sensors.

In one embodiment a controller is configured to perform the steps of fig 4 for one cylinder. In one such embodiment each cylinder has one controller. In one such embodiment the controller is configured to receive signal values from controllers for other cylinders.

In a first step 410 a value or signal, SN is received by the controller.

In one embodiment the controller is configured to calculate a compensation value through subtracting the received signal with a reference value (420) :

Scomp — Sp — Sref

In one embodiment the controller is further configured to apply a filter to reduce noise in the received signal 430 .

In one embodiment the controller is configured to calculate a sensor deviation, d(SsenSor):

In one embodiment the controller is also configured to calculate a cylinder deviation, d(cylinder):

In fig 4 both deviations are calculated in step 440.

In one embodiment the controller is configured to determine whether the alarm threshold for the sensor value is exceeded 450 and if so activate an alarm 460.

In one embodiment the controller is configured to determine whether the alarm threshold for the sensor deviation is exceeded 450 and if so activate an alarm 460 .

In one embodiment the controller is configured to determine whether the alarm threshold for the cylinder deviation is exceeded 450 and if so activate an alarm 460 .

In fig 4 all three determinations relating to the alarm limits are performed 450.

If the controller determines that an alarm is to be activated the controller is further configured to store the event in a log file 460.

In one embodiment the controller is configured to determine whether a slow down limit for a sensor deviation is exceeded 480 and if so activate a request for a slow down procedure 490.

In one embodiment the controller is configured to return to step 410 for receiving a new signal value.

In one embodiment the controller is configured to perform 410 to 440 for each revolution.

In one embodiment the controller is configured to perform 450 to 490 for each revolution.

In one embodiment the controller is configured to perform 450 to 490 at intervals. In one embodiment the interval is in the range of 1 to 50 revolutions. In one embodiment the interval is in the range of 10 to 30 revolutions. In one embodiment the interval is 30 revolutions. This is indicated in fig 4 by the dashed line.

In one embodiment the controller is configured to recalculate the average for the signal values for each sensor at intervals of 50 hours at each speed point. This allows the system to adapt and react to changes in the structure of the engine.

In one embodiment the controller is configured to determine whether a reference value for any engine speed is changed by a value greater than an update threshold when compared to the first obtained valid compensation value and if so activate and alarm indicating this. In one embodiment the update threshold is 0.2 mm.

In one embodiment the controller is configured to change the threshold values for at least one of the pre-warning, the alarm and the slow-down dynamically according to the current operation of the engine.

During normal operation of a large two-stroke diesel engine the engine is set to run at one engine speed for a long period of time. During this time period the environment of the engine is rather stable. For example as there are no accelerations there is less deformations due to thrust.

As the engine speed changes the environment becomes less stable which affects the BDC level of the bearings and these BDC levels will change and/or fluctuate accordingly. During such time periods the controller is configured to raise the threshold levels to accommodate for such changes without raising any alarms.

One example of another situation that affect the engine environment or the operation of the engine is when the load of a vessel being propelled by the engine increases and the vessel sits deeper in the water.

As the operation of the engine becomes stable again, possibly at a new engine speed, the controller is configured to lower the threshold values.

In one embodiment the controller is configured to lower the at least one threshold level after a delay time period has lapsed. This allows for after effects of the change in operation to be accommodated for without raising alarms unnecessarily.

In one embodiment there are arranged two sensors in each cylinder, one fore and one aft sensor, and as has been described above the controller is configured to determine whether a main bearing or a crosshead and/or crank pin bearing is worn. To determine that a main bearing is worn out or in a critical condition the fore sensor of a first cylinder is compared to the aft sensor of a second cylinder. To determine that a crosshead and/or crank pin bearing is worn out or in a critical condition the fore sensor of a first cylinder is combined with the aft sensor of the same cylinder.

There are thus two sensor readings available for the main bearing and for the crosshead and/or crank pin bearing.

In one embodiment the controller is configured to combine the two sensor readings and compare the result of the two sensor readings with an alarm threshold level.

For an embodiment having 4 cylinders each arranged with two sensors giving a total of 8 sensors labelled (SF4; SAi) , (Sf2; Sa2) , (SF3; SA3) and (SF4, SA4) with SF4 being the fore sensor for cylinder 1 and SA4 being the aft sensor for cylinder 1, wherein cylinder 1 is the cylinder that is closest to the engine output (in other words cylinder 1 is the aftmost cylinder in this example), the combined sensor value for the main bearing between cylinder 1 and 2, SMi2 is :

Smi2 = SF4 + SA2and the combined sensor value for the crosshead and/or crank pin bearing of cylinder 1, Scci is:

In one embodiment the sensor values are compensated as before by subtracting a reference value. In one such embodiment the controller is configured to sum up the values of the compensated sensor values and calculate the absolute value of the sum.

For the main bearing between cylinder 1 and 2, SMi2 is then :

By comparing the sum of two sensors the sensitivity of the measurement is further increased as any change in BDC level will be doubled before comparison which is beneficial since the differences are so small.

It should be noted that the threshold levels for the alarm, pre-warning and slow down are not the same for single sensor values as they are for combined sensor values .

In one embodiment the controller is configured to detect a rapid change in the sensor values.

During operation of the engine a small change of the BDC level is to be expected over time as the bearings will become worn. However, these changes occur very slowly and are difficult to differentiate from other changes such as changed engine environment (temp, thrust etc) and the controller is configured to focus on the BDC level instead of the change of level.

However a rapid change of BDC level may be an indication that something is about to go wrong. One such situation is when some pollution has entered the oil system and the bearings are being worn out more rapidly.

Even though the change is still within the tolerable limits (below the threshold limits for the various alarms) the speed of the change may lead to severe damage if not stopped in time and in some cases it may be too late when the alarm is raised if the change is rapid.

In one embodiment the controller is configured to detect a rapid change in BDC level by comparing the sensor values to a previous, but recent level. This enables the controller to determine if there is a significant change in the last time period even if this change is well within the tolerance limits.

In one embodiment the controller is configured to determine whether a rapid change is happening by performing an arithmetic analysis by comparing the sensor values to at least one previous time period (see below).

One advantage of the arithmetic analysis is that the compensation table is finely tuned.

In one embodiment the controller is configured to determine whether a rapid change is happening by performing an exponential analysis by comparing the sensor values to a floating reference level (see further on) .

One advantage of the exponential analysis is that it does not require the storing of many values and is thus faster and more resource friendly.

In one embodiment the controller is configured to detect a rapid change in BDC level by comparing the sensor values to an average level for a previous short time period.

In one embodiment this time period is in the interval 1 to 20 minutes. In one embodiment this time period is in the interval 1 to 10 minutes. In one embodiment this time period is in the interval 5 to 10 minutes. In one embodiment this time period is in 10 minutes. In one embodiment this time period is 5 minutes.

In one embodiment the time period compared to precedes the measuring by a second time period. In one embodiment this second time period is in the interval 1 to 20 minutes. In one embodiment this time period is in the interval 1 to 10 minutes. In one embodiment this time period is 5 minutes. In one such embodiment the controller is configured to compare the current measurements with an average level of a first time period that precedes the current time by a second time period. In one example where the first time period is 10 minutes and the second time period is 5 minutes a measurement at time T is compared to the average during T-5 and T-15.

In order to accommodate for erroneous readings and other short fluctuations a controller is configured to compare the average of the last readings with the average of sensor signals of the previous time period in one embodiment.

In one embodiment said set comprises a number of last received sensor values, wherein said number is in the range of 5 to 10. In one embodiment said set comprises the last five of the received sensor values. In one embodiment said set comprises the last ten of the received sensor values.

In order to accommodate for erroneous readings and other short fluctuations a controller is configured to compare the average of the last five readings with the average of the previous time period in one embodiment.

The controller is configured to determine rate of change from the difference between the average level of a previous time period and the current (average) level.

In one embodiment the controller is configured to determine whether the difference exceeds a threshold level and if so the controller is configured to issue a notification to an operator. In one embodiment the threshold level is 0.2 mm.

In one embodiment the controller is configured to determine whether the difference exceeds a threshold level and if so the controller is configured to issue a notification to a log.

In one embodiment the controller is configured to determine whether the difference exceeds a threshold level and if so the controller is configured to raise an alarm.

In one embodiment the controller is configured to determine whether the difference exceeds a threshold level and if so the controller is configured to reguest a slow down.

In such situations an early indication can thus be giving which may potentially prevent a serious engine failure.

Detecting rapid changes through exponential analysis In one embodiment the controller is configured to perform an exponential analysis of the sensor values and determine whether an alarm should be issued or not.

In one embodiment the controller is configured to determine a reference level and to determine a present state level by calculating exponential averages.

In one embodiment the controller is configured to update a reference level, Srefievei by updating an old reference with a compensated value SCOmp using a lowpass filter with an update factor.

In one embodiment this is expressed as:

where x is the update factor.

The reference level is a representation of the current normal level and should thus react relatively slow. Therefore the update factor should be chosen to be small. In one embodiment the update factor x is chosen to be 0.0001.

In one embodiment the controller is configured to update a present state level, Spres by updating an old present state with a compensated value Scomp using a lowpass filter with an update factor.

In one embodiment this is expressed as:

where y is the update factor.

The present state level is a representation of the present state of the BDC level and should thus react relatively fast to react to rapid changes. Therefore the update factor should be chosen to be large. In one embodiment the update factor y is chosen to be 0.2.

In one embodiment the controller is configured to also calculate a reference value SrefVaiue· In one embodiment the reference value is the difference between the reference level and the present state:

In one embodiment the controller is configured to analyse the reference value to determine whether an alarm should be raised.

Figure 9 shows a method for detecting rapid changes in BDC levels in bearings using exponential algorithms according to the embodiments herein.

A signal representing a BDC level is received and compensated according to an engine operating condition compensation table in steps 910 and 920.

In step 930 the calculations for the reference level, the present state and the reference value as described above are performed.

In one embodiment the controller is configured to analyse the reference value of a single sensor to determine that a rapid change is taking place. The rapid change is determined to take place if the reference value exceeds a threshold value. In figure 9 these calculations are performed in step 960.

By determining wear from only one sensor's reading allows for detection of uneven (in the direction of the axial) wear of a bearing.

In one embodiment the threshold level for only one sensor is 120.

In one embodiment the threshold level for only one sensor is above 120.

In one embodiment the threshold level for only one sensor is above 110.

In one embodiment the controller is configured to analyse the sum of the reference values of sensors of a single cylinder to determine that a rapid change is taking place. The rapid change is determined to take place if the sum exceeds a threshold value. The rapid change could be in the crank pin or the cross head bearing of the cylinder. In figure 9 these calculations are performed in the (alternative) step 940 and the determination whether an alarm limit is exceeded or not in step 960.

By determining wear from two sensors' readings allows for fast detection of wear of a bearing as the two readings are growing simultaneous and their sum thus grows twice as fast. To accommodate for the sum being larger than the two individual sensor readings the threshold level for a sum is greater than the threshold level for a single sensor in one embodiment.

In one embodiment the threshold level for sensors in a cylinder is 170.

In one embodiment the threshold level for sensors in a cylinder is above 170.

In one embodiment the threshold level for sensors in a cylinder is above 160.

In one embodiment the controller is configured to analyse the sum of a reference value of one sensor from one cylinder with a reference value of one sensor in a neighbouring cylinder to determine that a rapid change is taking place. The rapid change is determined to take place if the sum exceeds a threshold value. The rapid change could be in the main bearing between the two cylinders. In figure 9 these calculations are performed in the (alternative) step 950 and the determination whether an alarm limit is exceeded or not in step 960.

By determining wear from two sensors' readings from two different cylinders allows for even faster detection of wear of a main bearing as the two readings seem to be growing faster than the sensors in the cylinder sum. This thus provides an extra safety. To accommodate for the sum growing faster than the cylinder sum the threshold level for the neighbouring cylinders is greater than the threshold level for the cylinder sum in one embodiment.

In one embodiment the threshold level for sensors in neighbouring cylinders is 220.

In one embodiment the threshold level for sensors in neighbouring cylinders is above 220.

In one embodiment the threshold level for sensors in neighbouring cylinders is above 210.

It should be noted that all these steps (940 and 950) need not be performed.

In one embodiment the controller is configured to issue an alarm reguest (not shown) if it is determined that a rapid change has taken place (step 965).

In one embodiment the controller is configured to issue a slow down reguest (step 970) if it is determined that a rapid change has taken place (step 965) . As the changes are rapid it is important that measurements are taken guickly to prevent damage from happening and a slowdown provides for the fastest remedy to the rapid change.

In one embodiment the controller is also configured to cause information on the rapid changes to be stored in a log file (step 980) .

This enables a controller to react quickly to rapid changes without having to store multiple values for each sensor and to prevent further wear or damage to a bearing.

To account for variations in BDC levels due to engine speed changes a dynamic alarm limit or threshold level may be used when monitoring for rapid changes.

In one embodiment the dynamic alarm limit or threshold level is calculated based on the change in engine speed RPM (Revolutions Per Minute) and the dynamic threshold level increases when the engine speed increases.

In one embodiment the controller is configured to receive an input indicating the current or new speed, denoted RPMn.

The controller is configured to update a reference speed based on the new speed. In one embodiment the controller is configured to update the reference speed through a exponential moving average as in:

The controller is also configured to calculate a change in RPM, denoted ARPM by calculating the absolute value of the difference between the current speed RPMN and the reference speed:

In one embodiment ARPM is set to be 3 for all values larger than 3. This ensures that the alarm limit increases without getting too high.

A basic threshold level or basic alarm limit, denoted alarmbasic is determined to be one of the thresholds listed above depending on what sensor combination is to be monitored (single sensor, same cylinder or neighbouring cylinder).

In one embodiment the controller is further configured to calculate a first candidate dynamic threshold level (denoted alarm 1) based on an amplification constant, denoted k, as in:

In one embodiment the controller is determined to use the first candidate threshold level as the dynamic threshold level.

In one embodiment the controller is further configured to calculate a second candidate dynamic threshold level (denoted alarm 2) based on the number of engine speed readings received from the last time the threshold level was increased, denoted H, an amplification constant, denoted k, and a delay constant, denoted β, as in:

In one embodiment the controller is determined to use the second candidate threshold level as the dynamic threshold level.

The higher the amplification constant is the faster the alarm level increases. Examples of the amplification constant are: 0.1, 0.15, 0.2, 0.25 and 0.3. In one embodiment the amplification constant is in the range 0.05 to 0.4.

Examples of the delay constant are: 190, 200, 210. In one embodiment the delay constant is in the range 150 to 250.

In one embodiment the controller is configured to determine a highest threshold level of the candidate threshold levels and use that as the dynamic threshold level:

Dynamic threshold level = max(alarm 1, alarm 2).

Thus a controller is able to adapt to variations and still be able to detect rapid changes of the wear of a bearing.

It should be noted that the exponential analysis of the rapid changes as discussed above is also suitable for four-stroke engines. In one embodiment the engine is a two-stroke engine. In one embodiment the engine is a four-stroke engine.

To account for all the variable factors in monitoring of an engine the monitoring apparatus is first calibrated during a learning phase.

The look-up table of engine speed compensation must be acquired during a "learning process" performed by the controller .

In one embodiment the learning phase has a duration of 500 hours.

In one embodiment the controller is configured to establish a speed compensation table for each sensor 170.

In case of a wearing monitoring system fitted to a new-built engine, running shop test or performing sea trial, it is important that the system can offer protection of the bearings from the earliest possible time. Therefore, a rough calibration curve will be established as soon as a 10 minutes average value from one fixed engine speed is achieved. This curve is then re-adjusted so three fixed engine speeds are used to start with during the learning process .

Figure 5 shows a general flowchart for providing a rough calibration. The controller is configured to first create a rough estimate of the engine speed compensation by approximating the table values with a curve of the BDC level vs the engine speed.

This rough estimation allows for some control during the learning phase.

In one embodiment the controller receives the values for each sensor at one given engine speed (or rpm) 510. The values are averaged 520 and a preliminary curve is created 530 from the averaged engine speed-BDC-level coordinate or control point and an expected change depending on engine size over the speed range. In one embodiment the expected change is pre-calculated and stored in a table such as table 1.

Table 1 Expected change for engine types.

Fig 6 shows a curve of a preliminary curve or a rough estimate of the curve 600 having a control point 610 for the average BDC-level for a certain engine speed.

The curve 600 is a rough estimate and the controller is configured to use it as a reference only during the establishment of the rough calibration curve.

In one embodiment a second speed is selected and the controller receives the values for each sensor at the second engine speed 540. The values are averaged 550 and the curve is updated 560 to account for the second averaged engine speed-BDC-level coordinate or control point.

In one embodiment the controller is configured to repeat steps 540 to 560 for a third speed, in Fig 5 this is indicated by a dashed line.

In step 570 the controller completes the rough calibration curve (700) by interpolating between the control points (710, 720 730) and extrapolating to cove the full range of engine speeds.

In one embodiment the engine speed range is from 0-120%. In one embodiment the engine speed range is from 20-110%.

Fig 7 shows an example curve of a rough calibration 700 having three control point 710, 720 and 730 for the average BDC-level for a three engine speeds.

In one embodiment the engine is run at each speed for 10 minutes each. It should be noted that other running times are also available.

In one embodiment the controller is configured to select the three speeds according to table 2.

Table 2. Speeds and corresponding speed intervals.

In one embodiment the controller is configured to space the three speeds chosen at intervals of at least 20 % of the nominal speed to ensure that the control points are properly spaced apart in the curve 700.

In one embodiment the controller is configured to use rough calibration curve 700 during the remainder of the learning phase to allow for some reference of the wear during the learning phase.

In one embodiment the controller is further configured to receive sensor values during free operation of the engine at various speeds and to update the table of BDC-levels vs engine speeds.

Fig 8 shows a method of an embodiment of the present application for completing the engine speed-BDC table.

In one embodiment the steps of fig 8 are performed in parallel for all sensors.

In a first step 810 a value or signal, SN is received by the controller. The controller then determines whether the compensation is valid or not for the current speed 815.

As a pre-determined number of samples have been received for one engine speed the controller is configured to calculate a reference value for that engine speed by averaging the received samples for that sensor at that speed and generate a first valid compensation value and the engine speed compensation table is updated.

In one embodiment the controller is configured to receive and sample 1000 samples for each engine speed.

If it is determined that the compensation is valid the controller is configured to calculate a compensation value through subtracting the received signal with a reference value (step 820):

If it is determined that the compensation is not valid the controller is configured to calculate a compensation value through subtracting the received signal with a reference value (step 825):

In one embodiment the controller is further configured to apply a filter to reduce noise in the received signal 830 .

In one embodiment the controller is configured to next calculate a sensor deviation, d(SsenSor):

In fig 8 both deviations are calculated in step 840.

In one embodiment the controller is configured to determine whether the alarm threshold for the sensor value is exceeded 850 and if so activate an alarm 860.

In one embodiment the controller is configured to determine whether the alarm threshold for the cylinder deviation is exceeded 850 and if so activate an alarm 860 .

In one embodiment the controller is configured to determine whether the alarm threshold for the sensor deviation is exceeded 850 and if so activate an alarm 860 .

In fig 8 all three determinations relating to the alarm threshold are performed in step 850.

In one embodiment the alarm threshold value for a sensor value during the learning phase is + /- 0.8 mm.

In one embodiment the alarm threshold value for a sensor deviation during the learning phase is + /- 0.5 mm.

In one embodiment the alarm threshold value for a cylinder deviation during the learning phase is + /- 0.4 mm.

If the controller determines that an alarm is to be activated the controller is further configured to store the event in a log file 870.

In one embodiment the controller is configured to determine whether a slow down limit for a sensor deviation is exceeded 880 and if so activate a request for a slow-down procedure 890.

In one embodiment the slow-down threshold value for a sensor value during the learning phase is + /- 0.9 mm. In one embodiment the slow-down threshold value for a sensor deviation during the learning phase is + /- 0.7 mm.

In one embodiment the controller is configured to return to step 810 for receiving a new signal value.

In one embodiment the controller is configured to perform 810 to 840 for each revolution.

In one embodiment the controller is configured to perform 850 to 890 for each revolution.

In one embodiment the controller is configured to perform 850 to 890 at intervals. In one embodiment the interval is in the range of 1 to 50 revolutions. In one embodiment the interval is in the range of 10 to 30 revolutions. In one embodiment the interval is 30 revolutions. This is indicated in fig 8 by the dashed line.

To ensure that the full engine range is covered during the learning phase the controller is configured to receive samples for different engine speeds over the range of available engine speeds.

In one embodiment the controller is configured to receive samples for 100 engine speeds.

In one embodiment the controller is configured to receive samples from each sensor for each engine speed, that is, to receive samples for each possible revolution per minute in the engine's range.

In one embodiment the controller is configured to determine a reference point for each speed point that has not been validated during the learning phase. In one embodiment the controller is configured to determine the reference value by interpolating from the reference values of the adjacent speed points. In one embodiment the controller is configured to determine the reference value by extrapolating from the reference values of the reference points of the rough estimate.

To ensure a reliable monitoring of the wearing of a bearing during an extensive learning phase the learning phase is divided into three steps. The first step is to create a very rough preliminary estimate which is used as a reference for the second step in which a complete rough estimate is completed. The complete rough estimate is then used through the remainder of the learning phase (step 3).

This ensures that wearing of the bearings is monitored even during start-up of the engine. This allows for detecting increased wear due to for example faulty installations and irregularities which leads to an increased protection against down time for the engine.

A straightforward calibration would not be able to take such start-up problems into account and the learning phase as disclosed herein is thus highly advantageous.

During the lifespan of an engine it will undergo a series of maintenance operations and overhauls. For example if a bracket supporting the sensor (s) has been removed due to overhaul of a main bearing. When refitting the bracket, it must be expected that the sensor (s) will locate differently. There is also a possibility that a sensor needs to be replaced after having been damaged or otherwise rendered functionless. A sensor may also need to be re-adjusted if the bracket holding the sensor is slightly bent.

This will necessitate a re-adjustment of the speed compensation table or curve for one or several of the sensors .

The adjustment is effected through an adjustment of the offset of an individual sensor.

In one embodiment a controller is configured to adjust the signals form the sensor by compensating the signal value according to the existing speed compensation lookup table or curve. During normal operation the difference between the measured BDC level and the reference value is zero (at least on average) . The compensated value will therefore reflect the offset for the sensor as the absolute value of the compensated value will now be greater than zero (at least on average).

In one embodiment the controller is configured to calculate an average of the offset over a period of time. In one embodiment the time period is 50 hours.

In one embodiment the controller is configured to offset the reference values for an affected sensor according to the calculated average offset.

However, as there is a time lapse between a sensor failure and sensor replacement and this time lapse may in some circumstances be relatively long there is a risk of undetected wearing of a bearing during this time lapse.

To accommodate for any such wearing of a bearing a controller is configured to base the re-adjustment on a combination of the signals from the neighbouring sensor and an average taken over a period of time before and after the replacement. In one embodiment the time period is 500 hours.

Figure 10 shows a diagram of the speed compensation curves for two sensors, wherein: aiti + bi is the curve for the first sensor during the 500 hours prior to the failure of the sensor; a2t2 + b2 is the curve for the first sensor during the 500 hours after the replacement of the sensor; and a3t3 + b3 is the curve for the neighbouring sensor during the time the first sensor is broken.

In these formulas 'a' denotes the slope, 't' is the time and 'b' is a constant indicating the starting point for the curve for t=0.

In one embodiment a controller is configured to generate these three lines by performing 3 different best fit Root Mean Square lines based on the average of compensated and filtered sensor values over 6 hours.

In one embodiment a controller is configured to use the speed compensation for the neighbouring sensor while readjusting the sensor.

After the time period for re-adjusting the sensor has lapsed the offset (0) for the replaced sensor can be calculated.

In one embodiment the controller is configured to calculate the offset for the replaced sensor, and in one embodiment this is done through: 0 = Oi + a3tb + Ta2

Where : Οι is the offset during the time the sensor was broken; tb is the time the sensor was broken; and T is the time period (in the embodiment above T = 500 hours).

In one embodiment a controller is configured to exclude a replaced sensor when calculating sensor deviation and/or cylinder deviation for the other sensors during the time period after replacement of or the re-adjustment time period for the sensor.

In one embodiment a controller is configured to not calculate sensor deviation and/or cylinder deviation for a replaced sensor during the re-adjustment time period for the sensor.

In one embodiment a controller is configured to refrain from activation an alarm or issuing a slow-down request for sensor values resulting from the replaced sensor during the re-adjustment time period.

This allows for a learning phase for only the affected or replaced sensor and does not necessitate a complete rerun of the learning phase for all sensors. This allows for safer monitoring during the re-adjustment phase than a full rerun of the learning phase for all sensors would have provided.

Data storage in the bearing wear monitoring system serves two purposes. The first purpose is to be able to retrieve the data should any bearing damage have taken place. In one embodiment this is implemented as a "black box" function .

The second purpose relates to the inspection of the bearings. Traditionally, bearings are opened up for routine inspections on a time-based schedule. E.g. every bearing must be opened at, say, four or five year intervals. To avoid unnecessary opening of the bearings, it is desired to shift from time-based inspections to condition-based inspections. For this purpose, the data stored can be used to generate trend curves showing if any wear has taken place during the inspection interval. Such curves may be presented to a class surveyor.

In one embodiment a controller is configured to facilitate storage for each sensor of filtered values with time stamp available from latest 24 hours. In one embodiment such filtered values are stored in a shortterm storage. In one embodiment one set of data for every 30 engine revolutions is required. In case an alarm limit is exceeded, a controller is configured to store copy of the short-term storage separately as a "frozen" copy. In one embodiment the controller is configured to include data for 5 minutes after the time of the alarm in the copy.

In one embodiment a controller is configured to facilitate storage for each sensor of maximum, minimum, and average filtered values. In one embodiment such maximum, minimum, and average values are stored in a long-term storage. In one embodiment the Maximum, minimum, and average filtered values are stored for every 6 running hours. In one embodiment the maximum, minimum, and average filtered values are stored with a time stamp.

In one embodiment an event log is also stored. In one embodiment an event log contains the following information : any alarms, slow downs or pre-warnings released or changes during the learning process; possible replacement and/or offset adjustment of any sensors; change of sensor reference from rough calibration to fine calibration; and any reset of reference level for pre-warning.

In one embodiment all information is time stamped and all storage and event logs are kept in a non-volatile memory.

In one embodiment an apparatus is configured for downloading of stored data and event log to an external device like a PC. This data can be for the purpose of presentation to surveyors from classification societies.

In one embodiment the data includes:

Engine info part. This contains information of the ship and engine. In one embodiment the information is as below.

Name of Ship = XXXXXXX

IMO Number = XXXXXXX

Class Register No. = XXXXXXX

Component = XXXXXXX

Engine licenser = XXXXXXX

Engine maker = XXXXXXX

Engine type = XXXXXXX

Engine serial no. = XXXXXXXX

CM System Type = Bearing Wear Monitoring System

CM System Maker = XXXXXXXX

CM System Hardware = XXXXXX

CM System Software = XXXXXXX

Period of Trend Data From = YYYY-MM-DD

Period of Trend Data to = YYYY-MM-DD

Engine Operating hrs. From = 99999

Engine Operating hrs. To = 99999

Log part of all changes in system status and sensor status according to the event log. The format is [DATE][TIME][EVENT].

Filtered value trend part. Derived from "Long time storage". Following data to be presented for each sensor every six engine operating hours: Time stamp, Engine operating hours, and 6 h average filtered value. The format is in one embodiment [DATE AND TIME: YYYY-MM-DD hh: mm: s s]; [ENGINE OPERATING HOURS: h]; [DISTANCE: mm] .

Status part. The purpose of this file is to provide a quick overview of each cylinder showing if opening of the bearings related to this cylinder for inspection is justified due to detected wear or due to other circumstances. These circumstances could be detection of wear exceeding the "significance limit" during exchange of a sensor or loss of reference. In one embodiment the status is shown as a four level presentation. The four levels being:

Normal "N" indicating that no wear has been detected beyond the pre-warning limit;

Pre-Warning "W" indicating that wear has been detected beyond the pre-warning limit;

Alarm "A" indicating that an alarm has been released from sensors connected to this cylinder; and

Unknown "U" indicating that a sensor of this cylinder has lost reference or its reference curve has been corrected due to wear detected by its "neighbour" during exchange. The same indication is to be given if a damaged sensor remains unchanged.

In one embodiment an apparatus is also configured to provide the complete data of long term storage, the short term storage and the reference curves of each sensor.

The various aspects of what is described above can be used alone or in various combinations. The teaching of this application may be implemented by a combination of hardware and software, but can also be implemented in hardware or software. The teaching of this application can also be embodied as computer readable code on a computer readable medium.

The teaching of the present application has numerous advantages. Different embodiments or implementations may yield one or more of the following advantages. It should be noted that this is not an exhaustive list and there may be other advantages which are not described herein. For example, one advantage of the teaching of this application is that an apparatus according to herein provides a reliable monitoring accommodating for several (external) factors.

Another exemplary advantage of the teaching of the present application is that the wear monitoring accommodates for deformation caused by thrust.

Another exemplary advantage of the teaching of the present application is that the monitoring of wear in bearings is reliable also during a calibration phase.

Another exemplary advantage of the teaching of the present application is that a replacement of a sensor can be effected and re-adjusted in a more reliable manner.

Claims

1. Apparat til monitorering af slitage af et krydshovedleje (160), et krumtapleje (150) og et hovedleje (140) i en stor totaktsdieselmotor (100), hvilket apparat omfatter mindst to sensorer (170), hvor sensorerne er indrettet og konfigureret til at måle et nedre dødpunktsniveau i en given cylinder i forhold til et fast punkt for motoren, hvor apparatet endvidere omfatter en styreenhed, der er konfigureret til at: modtage et signal fra hver sensor (170); kompensere hvert signal afhængigt af en motordriftstilstand; bestemme, om en tærskelværdi er overskredet og i givet fald udsende en indikation af den overskredne tærskelværdi, kendetegnet ved, at styreenheden endvidere er konfigureret til at generere en motordriftstilstandskompensationstabel under drift af motoren i en indlæringsfase ved sampling af et første antal sensorværdier i forhold til et andet antal motordriftstiistandspunkter under drift af motoren, og når de første antal prøver er modtaget for en motordriftstilstand, for at bestemme en referenceværdi ved at tage gennemsnittet af de modtagne prøveværdier for dette hastighedspunkt.An apparatus for monitoring wear of a cross head bearing (160), a crank bearing (150) and a head bearing (140) in a large two-stroke diesel engine (100), comprising at least two sensors (170), wherein the sensors are arranged and configured to measuring a lower dead center level in a given cylinder relative to a fixed point of the engine, the apparatus further comprising a control unit configured to: receive a signal from each sensor (170); compensate each signal depending on a motor operating mode; determine if a threshold is exceeded and, if applicable, send out an indication of the exceeded threshold, characterized in that the control unit is further configured to generate a motor operating state compensation table during operation of the motor in a learning phase by sampling a first number of sensor values second number of engine operating points during operation of the engine, and when the first number of samples is received for a motor operating mode, to determine a reference value by taking the average of the sample values received for that speed point.

2. Apparat ifølge krav 1, hvor styreenheden er konfigureret til at bestemme et råt kalibreringsestimat for en motordriftstilstandskompensation ved af: modtage signalværdier fra mindst én sensor (170) over et tidsrum for en første motordriftstilstand; beregning af gennemsnittet af signalværdierne for at generere en første referenceværdi; og ekstrapolering af fra den første referenceværdi for motordriftstilstanden til andre motordriftstilstande ved anvendelse af en forhåndsestimeret ændringsfaktor; hvor styreenheden er konfigureret til at kompensere yderligere modtagne signalværdier ifølge det rå kalibreringsestimat i indlæringsfasen.Apparatus according to claim 1, wherein the controller is configured to determine a raw calibration estimate for a motor operating mode compensation by: receiving signal values from at least one sensor (170) over a period of a first motor operating state; calculating the average of the signal values to generate an initial reference value; and extrapolating from the first reference value of the motor operating mode to other motor operating modes using a pre-estimated change factor; wherein the controller is configured to compensate for further received signal values according to the raw calibration estimate during the learning phase.

3. Apparat ifølge krav 2, hvor styreenheden endvidere er konfigureret til at opdatere råkalibreringsestimatet ved at modtage signalværdier fra mindst én sensor over et tidsrum for en anden motordriftstilstand; beregning af gennemsnittet af signalværdierne for at generere en anden referenceværdi; interpolering mellem første og anden referenceværdier for hastigheder mellem første og anden hastighed; og ekstrapolering fra den første og anden referenceværdi for motordriftstilstande, der ligger uden for første og anden motordriftstilstand, ved anvendelse af den forhåndsestimerede ændringsfaktor; hvor styreenheden er konfigureret til at kompensere yderligere modtagne signalværdier ifølge det rå kalibreringsestimat i en indlæringsfase.Apparatus according to claim 2, wherein the controller is further configured to update the raw calibration estimate by receiving signal values from at least one sensor over a period of another motor operating state; calculating the average of the signal values to generate another reference value; interpolating between first and second reference values for speeds between first and second speeds; and extrapolating from the first and second reference values for motor operating states which are outside the first and second motor operating states using the pre-estimated change factor; wherein the controller is configured to compensate for further received signal values according to the raw calibration estimate in a learning phase.

4. Apparat ifølge krav 1, hvor driftstilstanden er en motorhastighed og motordriftstilstandspunktet er et hastighedspunkt.Apparatus according to claim 1, wherein the operating state is a motor speed and the motor operating state is a speed point.

5. Apparat ifølge krav 1, hvor driftstilstanden er et bladstigningsniveau.Apparatus according to claim 1, wherein the operating state is a leaf rise level.

6. Apparat ifølge krav 1, hvor driftstilstanden er en belastning.Apparatus according to claim 1, wherein the operating condition is a load.

7. Apparat ifølge krav 4, hvor styreenheden er konfigureret til at modtage signaler for tre motorhastigheder, hvilke tre motorhastigheder er: en første motorhastighed taget fra intervallet 20 til 50 % af en nominel motorhastighed; en anden motorhastighed taget fra intervallet 50 til 80 % af en nominel motorhastighed og en tredje motorhastighed taget fra intervallet 80 til 100 % af en nominel motorhastighed.Apparatus according to claim 4, wherein the control unit is configured to receive signals for three engine speeds, which three engine speeds are: a first engine speed taken from the range of 20 to 50% of a nominal engine speed; a second engine speed taken from the range 50 to 80% of a nominal engine speed and a third engine speed taken from the range 80 to 100% of a rated engine speed.

8. Apparat ifølge krav 3, hvor styreenheden endvidere er konfigureret til at opdatere råkalibreringsestimatet for en motordriftstilstandskompensationstabel under indlæringsfasen.Apparatus according to claim 3, wherein the control unit is further configured to update the raw calibration estimate of an engine operating mode compensation table during the learning phase.

9. Apparat ifølge krav 2, hvor styreenheden endvidere er konfigureret til igen at bestemme referenceværdien for hvert motordriftstilstandspunkt med et tidsinterval og opdatering af motordriftstilstandskompensationstabellen med de nye referenceværdier.Apparatus according to claim 2, wherein the control unit is further configured to again determine the reference value of each motor operating state with a time interval and updating of the motor operating state compensation table with the new reference values.

10. Apparat ifølge krav 9, hvor styreenheden endvidere er konfigureret til at bestemme, om en referenceværdi for et motordriftstilstandspunkt er ændret af en faktor sammenlignet med en første opnået referenceværdi og i givet fald aktivere en alarm.Apparatus according to claim 9, wherein the control unit is further configured to determine whether a reference value for a motor operating state is changed by a factor compared to a first obtained reference value and, if necessary, activate an alarm.

11. Apparat ifølge krav 8, hvor styreenheden endvidere er konfigureret til at opdatere motordriftstilstandskompensationstabellen efter indlæringsfasen ved at interpolere og ekstrapolere mellem værdier fra motordriftstilstandskompensationstabellen og råt kalibreringsestimat.The apparatus of claim 8, wherein the controller is further configured to update the motor operating state compensation table after the learning phase by interpolating and extrapolating between values from the motor operating state compensation table and raw calibration estimate.

12. Fremgangsmåde til implementering i et apparat med en styreenhed indrettet til at udføre instruktioner lagret på et fysisk medium, hvilken fremgangsmåde er til monitorering af slitage af et krydshovedleje (160), et krumtapleje (159) og et hovedleje (150) i en stor totaktsdieselmotor (100), hvilket apparat omfatter mindst to sensorer (170), hvor sensorerne (170) er indrettet og konfigureret til at måle et nedre dødpunktsniveau i en given cylinder i forhold til et fast punkt for motoren (100), hvor fremgangsmåden omfatter: modtagelse af et signal fra hver sensor (170); kompensering af hvert signal afhængigt af en motordriftstilstand; bestemmelse af, om en tærskelværdi er overskredet og i givet fald udsende en indikation af den overskredne tærskelværdi og hvor fremgangsmåden er kendetegnet ved endvidere at omfatte: generering af en motordriftstilstandskompensationstabel under motorens drift i en indlæringsfase ved sampling af et første antal sensorværdier i forhold til et andet antal motordriftstilstandspunkter under drift af motoren og når de første antal prøver er modtaget for en motordriftstilstandspunkt bestemme en referenceværdi ved at tage gennemsnittet af de modtagne prøveværdier for dette motordriftstilstandspunkt.A method of implementation in an apparatus with a control unit adapted to execute instructions stored on a physical medium, which is to monitor the wear of a cross head bearing (160), a crank bearing (159) and a main bearing (150) in a large a two-stroke diesel engine (100), comprising at least two sensors (170), wherein the sensors (170) are arranged and configured to measure a lower dead center level in a given cylinder relative to a fixed point of the engine (100), the method comprising: receiving a signal from each sensor (170); compensating each signal depending on a motor operating mode; determining whether a threshold has been exceeded and, if appropriate, emitting an indication of the exceeded threshold and wherein the method is characterized by further comprising: generating a motor operating state compensation table during operation of the motor in a learning phase by sampling a first number of sensor values relative to a second number of engine operating states during operation of the engine and when the first number of samples is received for a motor operating state, determine a reference value by taking the average of the sample values received for that motor operating state.

13. Fremgangsmåde ifølge krav 12, hvilken fremgangsmåde indbefatter mere end to sensorer (170), hvor styreenheden er konfigureret til at bestemme en sensorafvigelse for hver sensor (170) baseret på de modtagne sensorværdier for at bestemme hver sensors afvigelse fra gennemsnittet af de andre sensorer (170) .The method of claim 12, which comprises more than two sensors (170), wherein the controller is configured to determine a sensor deviation for each sensor (170) based on the received sensor values to determine each sensor's deviation from the average of the other sensors. (170).

14. Motor (100) omfattende et apparat ifølge et hvilket som helst af kravene 1 til 11, hvor motoren (100) er en stor totaktsskibsdieselmotor.An engine (100) comprising an apparatus according to any one of claims 1 to 11, wherein the engine (100) is a large two-stroke diesel engine.

15. Skib omfattende en motor (100) ifølge krav 14.A ship comprising an engine (100) according to claim 14.