RELIABILITY ASSESSMENT OF MARINE DIESEL
FUEL INJECTION SYSTEM ON THE BASIS OF THE
TORQUE PULSATION
OCENA NIEZAWODNOŚCI UKŁADU WTRYSKOWEGO
SILNIKA OKRĘTOWEGO NA PODSTAWIE PULSACJI
MOMENTU OBROTOWEGO
Charchalis Adam, Rosłanowski Jan
Gdynia Maritime University, The Faculty of Marine Engineering Morska Street 83, 81-225 Gdynia
e-mail: achar@am.gdynia.pl rosa@am.gdynia.pl
Abstract: The method of calculating the number of average value exceedings ship’s propulsion engine torque signal is described in probabilistic aspect. The number of exceendings can be used as an indicator of engine injection system efficiency. Instantaneous torque values measured on intermediate shaft in various sailing conditions determine ergodic realization of stationary stochastic process.. Reliable marine engine in directly propulsion system determines the safety of the ship's movement. Traffic safety of the ship prevent pollution of the marine environment by its wreck.
Keywords: marine diesel, torque signal, the safety of the ship’s movement
Streszczenie: Przedstawiono metodę obliczania liczby przewyższeń wartości średniej przez sygnał momentu obrotowego silnika napędowego statku. Liczba przewyższeń może być wykorzystana jako wskaźnik prawidłowości działania układu wtryskowego silnika. Chwilowe wartości momentu mierzone na wale napędowym stanowią ergodyczną realizację stacjonarnego procesu stochastycznego.. Niezawodność okrętowego silnika w bezpośrednim układzie napędowym statku decyduje o bezpieczeństwie jego ruchu. Bezpieczeństwo ruchu statku zapobiega zanieczyszczeniu środowiska morskiego przez jego zatonięcie
Słowa kluczowe: silnik okrętowy, sygnał momentu obrotowego silnika, bezpieczeństwo ruchu statku
1. Introduction
In steering system of ship’s propulsion engines, torque meters are used in order to protect them against overloading by torque. Torque meters with small-constant take measurements characteristic deflected by accidental disturbances. Such signals can be used to control engine’s injection systems during its settled work by measuring the number of torque exceedings in relation to average value in time unit. Thus allow us to assess the correctness of the injection system and detect its failure. This contributes to improved durability and reliability of the ship's engine and the cleanliness of the exhaust. In addition, increased vessel traffic safety with direct-drive system which forms a serial reliability structure. In such a system failure of the propulsion engine of the vessel loses its steering.
2. The number of given level exceedings by random torque function
Efficient engine performance is connected with torque of with instantaneous values in time of settle work make a signal being a random function. Such a signal includes random components originating from accidental disturbances of combustion process in engine’s cylinders and anti-torque changes. Signals of instantaneous torque values are ergodic realizations of stationary stochastic process [ 1, 3, 5 ]. For stationary processes, distribution density of instantaneous torque values f(M) as well as bivariate distribution density of torque value probability and its change rate f(M, v) do not depend on time.Probability of exceeding average by torque signal is determined by the formula:
dt M p M t M dt t V M P{ ( ) ( ) } ( ) (1) where:-
P{ } - probability of occurrences of exceeding average value by torque signal,
M - average value estimated approximately, M(t) - torque value in given time t,
V(t) - change of torque value in given time t, (torque derivative after time),
) (M
p - temporary probability density of exceeding average value by torque signal in time unit,
dt - time interval i.e. time difference between taking torque value reading named sampling step or discretisation time
Discretisation time dt is selected so as to avoid physical possibility of occurrence signal components of higher frequencies than limit frequency fg which satisfies a condition: g f dt 2 1 (2) where: - notations as above.
Probability of exceeding average value by torque signal in infinitesimal discretisation time dt is proportional to the length of the interval [ 1, 2 ]. The sum of probabilities of exceeding average value Mby torque signal upwards p(M)
and downwardsp* M( ) in time until is determined by the formula:
) ( ) ( ) ( ) , ( ) ( * ) ( M f M t M t V E dv v v M f M p M p
(3) where:M - average torque value estimated approximately, M(t) - torque value in given time t,
E[ ] - symbol of average value,
) , (M v
f - bivariate distribution of probability density of average torque value and its change rate,
v
- absolute value of torque signal change rate,) (M
f - density of probability distribution of average torque value,
)
(t
V
- absolute value of instantaneous change rate of torque.The number of exceedings NMof average value M by torque signal is proportional to selected observation time interval T and is determined by the formula:-
dv
v
M
f
v
T
N
M
0)
,
(
(4) where:- - notations as above.While selecting torque observation time T the following condition should be satisfied:- min max 1 f f T (5) where: max
f
- the largest expected frequency of torque signal in [Hz],min
f
- the smallest expected frequency of torque signal in [Hz].Torque observation time interval T determines discrimination degree of harmonic components while discretisation time dt determines maximum frequency which can still be discriminated in torque signal.
The average number of exceedings of the value appointed by torque signal in time unit
N
Mdefines so-called its “apparent” frequency and is determined by theformula:
dv
v
M
f
v
T
N
N
M M
0)
,
(
(6) where: - notations as above.The exactness of the average number of exceedings in time unit
N
M is related toterror mean-square which is made when the number is being determined. Terror mean-square
M
N
made when determining the average number of exceedingsM
N
in time unit for observation length of torque signal T is calculated by means of the expression [ 1 ]:
T M M NMN
T
N
T
T
t
P
t
d
t
0 2)
(
)
(
)
(
)
(
(7) where: MN
- average number of exceedings in time unit for observation ( recording ) length of torque signal T in [s],) ( t
P - probability of occurrence of torque signal exceeding in relation to average value in time instances which are distant from each other by
t
seconds,M
N - the number of torque signal exceedings in relation to average value. Terror mean-square
M
N
tends towards zero for observation (recording) times of torque signal tending to infinity.3. Measurements of quantities of average value exceedings by torque
signal on intermediate shaft
Measurements of instantaneous torque values can be taken by means of torque meter and transformer method with non-contact transmission of signal from rotating shaft [ 3 ]. Measurements of quantities of average value exceedings by torque signal hale been in two stages:
1. The first stage of measurements comprises recordings of torque signal during work of ship’s propulsions engine in different signal during settled work of ship’s propulsion engine in different sailing conditions are shown.
Fig. 1. Torque signals of fixe frequency f=10Hz on intermediate shaft measured in diffrent saling conditions
2. The second stage of measurements comprises recordings of torque signal while introducing disturbances in fuel injection to one of the engine’s cylinders. Disturbances were introduced by gradual opening injector’s deaerating value in torque signals while taking such measurements are shown.
The charter of establishing the number of torque signal exceeding of average value according to measurement time was checked experimentally. His check results are show after work [ 3 ] on chart 1. On its basis were made diagrams presenting correlation between average number of torque signal exceedings of average value in time unit and quantity of their occurrences, which are shown in fig. 3. The average number of exceedings stabilizes satisfactorily on the segment of torque signal measured in one-second time.
Fig. 2. Torque signals at times when different quantities of fuel were injected to a cylinder of ship’s propulsion engine in its settled motion.
Legend: 1- At time of undisturbed engine’s work; 2- on the time of dropping fuel by deaerating injector; 3- complete lack of fuel injection to engine’s cylinder.
4. Conclusions
Measurements of number torque signal exceedings of average allow us to ascertain the following:
1) One-second-recording of torque during normal operation of five-cylinder engine at rotational speed varying from 2,1[s-1] to 2,25 [s-1] keeps the number of
2) Frequency of torque signals is determined by frequency of ignition occurrences in engine’s cylinders,
3) The numbers of torque signal exceedings of average value in time unit depends only on rotational speed of the engine and is independent of quantities of amplitudes and signal phases connected with sailing conditions of the ship (fig. 1.),
4) Disturbances developed by dropping fuel in injector deaerating process cause decrease in time unit; For conditions shown in point 1, the number of exceedings of torque signal being recorder in one-second time was kept within the limit (from 6 to 8),
5) For undisturbed five-cylinder engine operation at rotational speed 1,1[s-1] i.e. „very slowly ahead” the number of torque signal exceedings of average value in one second time is equal to11.
Fig. 3. Correlation between the average number of exceedings of a given value by torque signal
N
M and observed quantity of their occurrencesNM during normal and disturbedOn the basis of number of torque signal exceedings of average value in time unit, the statements above make possible to establish the limit value at which condition. This limit value can be a unitary number of exceedings at settled work of the engine operating at minimum rotational speed [ 4, 5 ]. Decrease in number of torque signal exceedings in time unit below the limit value informs about malfunction of the injection systems of ship’s propulsion engine
References
[1] Bendat J. S., Piersol A.G.: Random data: Analysis and measurement procedures. PWN Warsaw 1976.
[2] Fisz M.: Probability calculus and mathematical statistics. PWN, Warsaw 1969. [3] Rosłanowski J.: An attempt to apply autocorrelation analysis into supervision of
shipping combustion engine on the basis of changes in rotatory moment caused by the disturbances of fuel injection, Doctoral thesis, Posen Polytechnic 1982.
[4] Jeszke T., Rosłanowski J., Wiśniewski T.: Insection system of controlling regularity of heat generation process in propulsion combustion engine, especially marine no.137905.( Certificate no. 222158), Patented information 6/87.
[5] Rosłanowski J.: Estimation of heat generation in engines cylinders on the bases of autocorrelation of rotatory moment’s temporary values. Journal of EXPLO-DIESEL & GAS TURBINE ’01 Vol 1. Gdańsk – Międzyzdroje – Kopenhaga 2001.
[6] Komorska I.: The diagnosticmodelproposition oft he enginevibration signal. Journal of Kones Powertrain and Transport, vol 15 No. 2,Warsaw 2008.
[7] Charchalis A., Pawletko R. The use of expert system of marine diesel engine diagnosis. Journal of KONBiN, No. 1-2, Warsaw 2009.
[8] Borowczyk H.,. Linsted P., Manerowski J./: A metod of parametic evaluatiopnof thetechnical object reliability. Journal of KONBiN, No. 1-2, Warsaw 2009.
[9] Borkowski T., Kowalski P.: In cylinder pressure assessment from marine diesel engine. Journal of Kones Powertrain and Transport, vol 16 No. 3,Warsaw 2009.
Prof. D.Sc Hab. Eng. CHARCHALIS Adam, Dean of Faculty of Marine Engineering, Director of Repair Technology of Ship and Harbour Equipment. Professor Charchalis was graduated from Polish Naval Academy in 1971. His got a D. Sc degree in 1978, habilitation degree in 1984 and was made a professor in 1994. Professor Charchalis was a dean of Faculty of Mechanical and Electrical Engineering in Polish Naval Academy in 1994-2003. Employed as a professor Mr Charchalis works at Gdynia Maritime University since 1999. In his scientific work he deals with the problems of power plant energy of seagoing vessels, propulsion devices, ships designing, exhaust gas turbines, marine units diagnosis. Prof Charchalis created and implemented main propulsion diagnosis system of marine ships equipped with exhaust gas turbines. Prof Charchalis is an author of 3 monographs, 8 textbooks, 250 research works and thesis supervisor of 13 PhD’s degrees.
