You are on page 1of 10

MAN Diesel & Turbo

Info no. Sub category Ident. no.

3093109-4
Subject
Troubleshooting using HCU Events
Title Page no.

Examples of HCU Events from ME-ECS 1(19)


Replacement for Ident. no.
Date Des. Chk. Appd. A. C. Change/replacement C. no.
2012-01-26 MSJ Initial draft for PrimeServ Academy. Internal use only 0
1
2
3
4
5
6
7
8
Replaced by Ident. no. 9

Introduction and Layout


The following is a collection of some of the cases where the HCU Events have
been used as part of the troubleshooting process.
It is intended as a guide both in interpreting HCU Events and also in
understanding the usefulness of taking manual logs on a regular basis.
Incidentally the examples also show that many cases which would previously
lead to unnecessary exchange of FIVA and/or MPC can be solved more
correctly using the HCU Events.

Examples included
1. MECHANICAL FAULT ON EXHAUST VALVE ACTUATOR 2
2. WORN BEARINGS ON A CW FIVA 6
3. PROBLEMS WITH EXHAUST VALVES 3 AND 6 8
4. CHANNEL 70 DISCONNECTION 12
5. TOO LOW SETTINGS OF SAFETY VALVES FOR AIR SPRING 14
6. INSUFFICIENT FUEL PLUNGER MOVEMENT 16
1. Mechanical fault on Exhaust Valve Actuator

The alarm "Exhaust Valve Stroke Low (SlowDown)" was raised by CCU1.
Incorrect movement of the exhaust valve can be due to many things: Problems with the MPC, the FIVA,
the actuator, the exhaust valve, the sensors, the cabling, the hydraulic oil etc.
The system automatically logged the below HCU Event:

As can be seen it was during starting of the engine since no fuel was injected prior to the alarm being
raised.
Zooming on the FIVA signals shows that the FIVA followed the set point:
In case there is doubt about whether the FIVA behaviour was acceptable it is possible to compare
with HCU Events taken on days where there were no problems:

On the above image the exhaust valve is opening and closing just as intended and there are
no problems.
Zooming on this HCU Event it can be seen that the FIVA behaviour in this case is similar to
the behaviour when the exhaust valve failed to move:

Based on this examination of the HCU Events it is possible to start the troubleshooting
somewhere elsethan the FIVA. For instance testing of the exhaust valve position feedback
sensor (possibly by exchanging with the sensor from another unit) would be a somewhat easy
place to start.
In the actual case it turned out that the problem was mechanical trouble with the exhaust valve
actuator.
2. Worn bearings on a CW FIVA
A routine manual log gave the below HCU Event:

Note that in this case the plant is equipped with FIVA's manufactured by Curtiss-Wright, which means
that the amplifier is physically placed with the MPC and the output from the amplifier is also included
in the HCU Event (4: "Amplifier Current Ch33").
Zooming reveals large differences between the FIVA set point and the FIVA feedback - also
during the steady states (where the set point is a horizontal line):

The optimal action would be to compare with earlier HCU Events and corresponding
performance measurements; preferably from sea trial or maiden voyage.
The original plastic bearings in the Curtiss-Wright FIVA valves can get worn over time and this
was the case here. MDT offers a procedure for reconditioning the valve by replacing the plastic
bearings with self-lubricating ones made of bronze.
3. Problems with exhaust valves 3 and 6
Routine manual logs showed the below when all 7 units were plotted together:

Plotting several plots taken at the same time (i.e. with the same engine conditions) gives the possibility of
comparing exhaust valve movement, fuel plunger movement and main spool movement between units in
a qualitative manner.
Note that due to the plot being logged manually the angular spread of the logs will not be evenly divided
across the engine revolution; that is why for instance the fuel plungers are clustered in a group.
Zooming makes it easier to compare the shapes of the exhaust valve movement:

In this case it can be seen that two of the exhaust valves have a sharper cut-off during opening.
By plotting the logs two at a time it is possible to single out the units that differ from the rest.
Unit 1 and 3:

Unit 1 and 6:

Through this process it is possible to see that exhaust valve 3 and 6 behave differently than the rest.
The next course of action would then be to compare the performance parameters and see if
the exhaust valve behaviour has any negative influence on these.
When convenient it would be wise to disassemble a couple of exhaust valves and compare these.
The differences in behaviour could be due to mechanical problems in e.g. the actuator damper or
the damper in the exhaust valve.
4. Channel 70 disconnection
The alarm "Exhaust Valve Stroke Low (SlowDown)" was raised by CCU1. The system
automatically logged the below HCU Event:

As can be seen this happened either during starting or an air-run since no fuel was injected prior to the
alarm being raised. Also note that the engine is equipped with 'low force' exhaust valves as can be seen
from the overshoot in position during the opening of the valves.
Zooming on the three signals concerning the FIVA valve (set point, feedback and controller) gives the
below image:

It can be seen that approximately half a second prior to the alarm is raised, the FIVA main spool moves
to its fail safe position, however ECS continues to attempt to control the FIVA which can be seen by the
control signal increasing in magnitude until it is basically just an offset version of the set point. This is
correct behaviour since the FIVA position controller is mainly a P-controller.
The FIVA moving to the fail safe position can be caused by the pilot valve loosing its signal, in this case
it was caused by a loose connection in the cable from the CCU channel 70.
This is also an example where a malfunction in one component (the wire for the FIVA) is detected by
another component (the supervision of the exhaust valve stroke), giving an alarm that needs interpreting
together with the HCU Event before the root cause is found.
5. Too low settings of safety valves for air spring
A manually logged HCU event during engine start gave the below image:

This engine is equipped with low force exhaust valve and therefore it was expected that the opening of
the exhaust valve would have an overshoot and a settling period (similar to the behaviour displayed on
the HCU Event shown on page 12 in this document). However this seems to not be the case.
Zooming on the exhaust valve opening makes the lack of overshoot more evident; the top of the signal
where the exhaust valve is open is almost completely flat.

Since the signal also includes a small undershoot (at app. 3.20sec) the flatness is not due to the exhaust
valve hitting the physical maximum movement. Rather it is due to saturation in the position sensor: If
the exhaust valve moves too far then the sensor reaches its maximum value and then gives a constant
value.
Comparing with earlier HCU Events would reveal if this was a new situation or one that had persisted for
some time.
In the actual case it turned out that it had been the case since the sea trial and it was due to a too low
setting of the safety valve for the exhaust valve air spring. A too low opening pressure means that the
exhaust valve moves further than intended.
6. Insufficient fuel plunger movement
As part of investigating intermittent start problems on an engine, the below HCU Event was manually
logged:

As can be seen this was during the starting of the engine - actually the very first injection on that unit.
The fuel equipment was suspected as the root cause for the starting troubles.
Zooming on the HCU Event is was possible to compare the FIVA main spool movement with the fuel
plunger movement:

In the above case the main spool moves to approximately 2000 'counts' and the fuel plunger moves
approximately 400 'counts'.
With the engine running another HCU Event was logged manually:

Zooming makes it apparent that the fuel plunger moves much further in this case:

Since the manual log was taken when the index was more or less the same as during starting, the main
spool moves to approximately the same value (2000 'counts') but this time the fuel plunger moves
almost 2000 'counts'.
This indicates that one of the fundamental prerequisites for the FIVA control was not fulfilled: It is
required that the same movement of the FIVA valve results in the same fuel plunger stroke.
It is worth noting that the starting position of the fuel plunger is the same in both cases; i.e. the fuel
plunger was returned as it should be prior to injection.
The possible causes for this kind of behaviour could be malfunctioning suction valves in the fuel oil
pressure booster or air in the hydraulic oil.
It turned out that it was air in the hydraulic oil and a change to the mechanical design was introduced.

You might also like