Sunteți pe pagina 1din 6

Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

A Comparative Study on Mean Value Modelling


of Two-Stroke Marine Diesel Engine
G. P. Theotokatos
Department of Naval Architecture
School of Technological Applications, TEI of Athens
St. Spyridonos Str., 12210 Egaleo, GREECE
mtheot@teiath.gr http://www.na.teiath.gr

Abstract: -. In the present paper, two mean value modelling approaches of varying complexity, capable of simulating
two-stroke marine Diesel engines, are presented. Both approaches were implemented in the computational environment
of MATLAB Simulink®. Simulation runs of transient operation cases of a large two-stroke marine Diesel engine were
performed. The derived results were validated against previously published data are used for comparing the two
modelling approaches and discussing the advantages and drawbacks of each.
Key-Words: - mean value modelling, marine Diesel engine, simulation, MATLAB Simulink®

1 Introduction results from both modelling approaches are validated


Due to the very high cost for the procurement of the against previously published data. In addition, they are
installation of a marine two-stroke Diesel engine, the compared against each other leading to conclusions on
major part of the engine control system development each approach advantages and disadvantages.
relies on appropriate simulation tools of varying
complexity, which can be categorized as transfer function 2 Mean Value Engine Modelling
models, cycle mean value models and zero or one- The two-stroke marine Diesel engine mean value
dimensional models. The representation of the real engine engine model is constructed by considering the processes
processes is enhanced as the complexity of the used occurring in its components. The main engine
simulation tool is increased (i.e. from transfer function components that have been mathematically modelled,
models to one-dimensional models), but at the same time, shown in Fig. 1, are the cylinders, the scavenging and
the required amount of input data as well as the model exhaust receivers, the compressor and turbine of the
execution time is also increased. The cycle mean value turbocharger and the engine air cooler. In addition, the
models compromise between the above mentioned engine exhaust pipe and the air filter, located upstream
contradictory factors and, therefore, are widely used compressor, are also modelled.
throughout the control system design procedure due to In both mean value engine modelling (MVEM)
their simplicity and their ability to sufficiently represent approaches presented bellow, the engine crankshaft and
the engine components behaviour [1,2]. turbocharger shaft speeds are calculated using the
In the present paper, two different ways of modelling following equations derived by applying the angular
a large two-stroke marine Diesel engine using cycle mean momentum conservation in the propulsion plant shaft
value models are presented. According to the first system and the turbocharger shaft, respectively:
approach, the engine crankshaft and turbocharger shaft dN E / dt = (η Sh QE − QP ) / ( I E + I Sh + I P ) (1)
speeds are obtained by solving the angular momentum
conservation differential equations. The other engine dNTC / dt = ( QT − QC ) / ITC (2)
variables are obtained as solutions to a nonlinear to ambient via engine
algebraic system of three equations corresponding to engine ambient exhaust piping system
energy and mass conservation across the engine.
According to the second approach, the engine scavenging compressor turbine
and exhaust receivers are modelled as open NTC
thermodynamic systems. The mass, temperature and
pressure of the working medium contained in the engine air cooler
receivers are calculated using the mass and energy
differential equations and the ideal gas law, respectively. scavenging exhaust
For representing the engine turbocharger compressor and receiver receiver
turbine, their maps, derived under steady state conditions
are used. The engine cylinders are modelled using a cycle
mean value modelling approach to calculate the average engine
mass flow and enthalpy rates of the exhaust gas exiting cylinders engine
NE crankshaft
the cylinders and entering the exhaust receiver. Engine
crankshaft and turbocharger shaft speeds are obtained Fig. 1: Components of a two-stroke marine Diesel
using angular momentum conservation. The simulation engine used for mean value modelling

ISSN: 1790-2769 107 ISBN: 978-960-474-120-5


Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

The engine cylinders of a two-stroke marine Diesel the average (per cycle) values of engine parameters are
engine are modelled considering a system comprising two calculated. In that respect, mean value modelling
orifices connected in series. Each one of the orifices approaches can not be used for calculating the
represents the cylinder intake ports and exhaust valve, instantaneous values (per degree of crank angle) of the
respectively. These two orifices can be combined in one engine parameters.
equivalent orifice producing the same mass flow rate for a The engine governor is modelled using a
given pressure ratio across engine cylinders and its proportional-integral (PI) controller law. Thus, the engine
geometric area can be estimated using the areas of intake governor rack position is calculated as follows:
ports and exhaust valves, as follows [3]: t
xr = xr , o + k p ∆N + ki ∫ ∆Ndt (13)
Aeq = ( zcyl / 2π ) ∫
0

( i
2 2
e )
Ai (φ ) Ae (φ ) / A (φ ) A (φ ) dφ (3)
0
where ∆N=Nord-NE is the difference between the ordered
The cylinders air mass flow rate is calculated using engine speed and the actual engine speed. In addition,
the following equation, which has been derived according torque and scavenging pressure limiters have been also
to the quasi-one dimensional consideration in an orifice incorporated in the engine governor model as proposed
with subsonic flow [4]: and used by engine manufacturers [8] for protecting the
m a = cd Aeq pSC / RaTSC f ( prcyl , γ α ) , prcyl = pER / pSC , engine integrity during fast transients.
The compressor impeller absorbed and turbine wheel
(4)
(
f ( prcyl , γ α ) =  2γ α / ( γ α − 1)  prcyl 2 / γ α − prcyl ( α ) α)
γ +1 / γ delivered torques, required in equation (2), are calculated
by:
The mass and energy balances applied on engine QC = 30 PC / π NTC , QT = 30 PT / π NTC (14)
cylinders give: The compressor and turbine powers, absorbed and
m a + m f = m e (5) delivered to the turbocharger shaft respectively, are
(m a hSC + ηcomb m f H Lζ )ηex = m e hER (6) calculated depending on the modeling approach as
described in the next sections of this text.
where, ζ is fuel chemical energy proportion in the exhaust
For modelling the air cooler, its effectiveness and
gas entering turbine [3] and ηex is a correction factor used
pressure drop are required. The air cooler effectiveness is
to take into account the heat transfer from the exhaust gas
assumed to be a polynomial function of the air mass flow
to the ambient in the cylinder exhaust ports and exhaust
rate, according to the following equation:
receiver.
The proportion of the fuel chemical energy ε AC = k AC 0 + k AC1m a + k AC 2 m a2 (15)
contained in the exhaust gas is considered linear function whereas the pressure drop in the air cooler is calculated
of the engine mean effective pressure [3]: by:
ζ = k z 0 + k z1 pb (7) ∆p AC = f AC ( ρ AC w2 / 2 ) = f AC m a2 / ( 2 ρ AC AAC
2
) (16)
The engine brake mean effective pressure is The pressure after the turbine is calculated using the
calculated by subtracting the friction mean effective pressure increase of the exhaust piping system, which in
pressure from the indicated mean effective pressure: turn is regarded as proportional to the exhaust gas mass
pb = pi − p f (8) flow rate squared:
The indicated mean effective pressure is calculated pT , d = patm + ∆pep = patm + kep m e2 (17)
using the rack position, the maximum indicated mean The propeller torque, required in eq. (1), is
effective pressure of the engine and the combustion calculated according to the propeller law through the
efficiency, which in turn is regarded as function of air to engine maximum continuous rating (MCR) operating
fuel ratio [5]: point:
pi = xr pi ,maxηcomb (9)
QP = k P N E2 , k P = QE , MCR / N E2 , MCR (18)
The friction mean effective pressure is considered
function of the indicated mean effective pressure and the
engine crankshaft speed [6,7]: 2.1 1st MVEM Approach
The air and gas properties are considered to be
p f = k f 0 + k f 1 N E + k f 2 pi (10)
constant. Thus, the temperature of the working medium
The engine fuel mass flow rate is calculated using (air or gas) throughout the engine components is
the following equation, where of the injected fuel mass calculated using the following equation:
per cylinder and per cycle vs. fuel rack position is h = c pT (19)
provided as input:
The compressor is usually modelled using its
m f = zcyl m f , cy N E / ( 60 revcy ) (11) performance map. However, in marine propulsion plant
The engine torque, brake power, brake specific fuel systems, the engine is loaded according to the propeller
consumption and efficiency are calculated using the law and the compressor operating points under steady
following equations: state conditions lay on a single curve on the compressor
pbVD Q π NE m f Pb map [5]. Depending on the selected surge margin, the
QE = , Pb = E , bsfc = ,η b = (12) compressor operating points could be located on regions
2 π revcy 30 Pb m f H L
of constant efficiency or close to the optimum compressor
It must be noted that using the equations given above

ISSN: 1790-2769 108 ISBN: 978-960-474-120-5


Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

efficiency parabola. In that respect, the compressor simulation time, the values for the three independent
pressure ratio can be modelled as function of the variables (air mass flow rate, exhaust gas pressure and
turbocharger shaft speed, whereas the compressor temperature) are initialized. For each time step, taking
efficiency can be taken either as constant or on the into account the values for engine speed, turbocharger
optimum efficiency curve of the compressor map. shaft speed (their initial values are taken into
According to the modelling approach presented in consideration at the start of the simulation time) and rack
this section, the compressor pressure ratio and efficiency position, the required engine parameters are calculated
are regarded as second order polynomial functions of the and the no-linear algebraic system is solved to provide the
turbocharger shaft speed and are calculated using the air mass flow rate and the exhaust gas pressure and
following equations: temperature. Then, the remaining engine parameters, as
prC = 1 + kC1 NTC + kC 2 NTC
2
(20) well as the time derivatives of the engine speed and the
turbocharger shaft speed are also derived. The engine
ηC = kCef 0 + kCef 1 NTC + kCef 2 NTC
2
(21) speed and the turbocharger shaft speed are calculated
The compressor exiting air temperature is calculated using a fourth order Runge-Kutta integration method with
using the following equation, which has been derived fixed time step. The above described procedure is
using the compressor efficiency definition equation [5]: repeated for every time step till the end of the simulation
( (
TC = Ta 1 + prC (
γ a −1) / γ α
) )
− 1 / ηC (22) time.

The pressure and temperature of the air contained in 2.2 2nd MVEM Approach
the engine scavenging receiver are derived as follows: According to the second modelling approach, the
TSC = TC (1 − ε AC ) + ε AC Tw, AC (23) two-stroke marine Diesel engine is modelled using flow
pSC = pC − ∆p AC = prC patm − ∆p AC (24) receivers (control volumes) interconnected between flow
elements. Fixed fluid elements with constant pressure and
The engine exhaust gas mass flow rate is calculated
temperature are used for modelling the engine
by the following equation derived applying a quasi-one
boundaries. Shaft elements are used for calculating the
dimensional approach [4,5] in the turbine:
engine crankshaft and turbocharger shaft rotational
2γ e
m e = AT ,eff
pER
ReTER (γ e − 1)
(
prT2 / γ e − prT( e ) e )
γ +1 / γ speeds by solving the differential equations (1) and (2),
respectively. The engine governor rack position and
(25) propeller torque are calculated using equations (13) and
γe
 
prT = max  pT , d / pex , ( 2 /(γ e + 1) ) e 
γ − 1 (18) in the PI governor and propeller elements,
  respectively.
The turbine effective flow area is calculated from the The engine scavenging and exhaust receivers are
turbine geometric area and the turbine flow coefficient. modelled as flow receivers, whereas the engine
The turbine flow coefficient and efficiency are derived compressor, cylinders and turbine are modelled as flow
from turbine steady state performance maps and are elements. The model was also implemented in the
considered polynomial functions of the turbine pressure MATLAB/Simulink environment following a modular
ratio, i.e.: way as shown in Figure 2. Thus, the modelled engine
AT , eff = α T AT , geo , αT = f ( prT ), ηT = f ( prT ) (26) elements form discrete subsystems, which exchange the
Combining the equations (4)-(11), (15)-(17) and required variables through appropriate connections. The
(19)-26), for a given set of engine speed, turbocharger flow elements use as input the pressure, temperature and
shaft speed and engine governor rack position, (NE, NTC, the properties of the working medium (air or gas)
xr), a non-linear algebraic system is solved to obtain the contained in the adjacent elements (flow receiver(s) or
three independent variables, i.e. the air mass flow rate, the fixed fluid), whereas their output includes the mass flow
exhaust receiver pressure, and the exhaust receiver and energy rates entering and exiting the flow element as
temperature. Then, the remaining engine parameters are well as the absorbed (for the case of compressor) or
calculated from the respective equations given above. produced torques. The former are provided as input in the
Taking into account eq. (19), the power of adjacent flow receiver elements, whereas the latter is
compressor and turbine required according to eq. (2) and required as input in the shaft elements. The output of
(14) for calculating turbocharger shaft speed, are shaft elements i.e. the engine crankshaft and turbocharger
PID governor time

calculated by: engpar


To Workspace FR
Nord
Neng

PC = m a c pa (TC − Tatm ) , PT = m e c pe (TER − TT , d ) (27)


pscav Nord
schedule
OUT_FF

where, the temperature of the exhaust gas exiting turbine


OUT_u

OUT_d
FR
INP_u

Neng

INP_d

INP_u

INP_d
Ntc

fixed
OUT_FF

OUT
INP_u

INP_d
Ntc

fixed Open Open fluid


flluid Thermo- Thermo- exhaust
engine dynamic
is calculated using the following equation, which is ambient ambient
OUT_shaft

dynamic
System- turbine
Sum_out

cylinders
Sum_out

Sum_in

OUT_u

OUT_d

System-
Sum_in
OUT_u

Qcomp

OUT_d

OUT_u

OUT_d

Qturb

scavenging exhaust
compressor receiver
derived by the turbine efficiency definition equation [5]: receiver

( (γ e −1) / γ e 
)
T2

TT , d = TER 1 − ηT 1 − ( pT , d / pER )
T1
N_tc

Neng

OUT Neng

(28)
Q_comp

T/C
INP_load
INP_eng

 
Q_turb

Engine
shaft propeller
crankshaft

The above presented equations were implemented in


the MATLAB/Simulink® environment. The calculation Fig. 2: Model of two-stroke Diesel engine implemented
procedure takes place as follows. At the start of the in MATLAB/Simulink environment

ISSN: 1790-2769 109 ISBN: 978-960-474-120-5


Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

shaft rotational speeds, are supplied as input in the 3 Results and Discussion
respective flow controller. The above described MVEM approaches were
The flow receiver elements are modelled using the applied for examining the transient behaviour of the
open thermodynamic system concept [4,5]. The working MAN B&W 9K90MC two-stroke marine Diesel engine.
medium mass, temperature and pressure are calculated That engine was also used in previous research studies [8-
using the following equations derived from mass and 10], where significant amount of experimental data and
energy balances and the ideal gas law, respectively: simulation results were published. The main engine
dm / dt = m in − m out (29) characteristics as well as the required input data and
( ht in in out out )
dT / dt = Q + m h − m h − udm / dt / ( mc ) (30)

v
engine steady state performance data were extracted from
the engine manufacturer project guide [11].
p = mRT / V (31) For each one of the MVEM approaches, the
No heat transfer is taken into account for the following steps were followed. Initially, the mean value
scavenging receiver, whereas the transferred heat from engine model was set up and the required constants were
the gas contained in the exhaust receiver to the ambient is calibrated so as the simulation results under steady state
calculated using the following equation: conditions are in good agreement with the respective ones
Q ht = kht Aht (TER − Tatm ) (32) presented by the engine manufacturer in [11]. Then, the
calibrated mean value engine model was used to perform
The working medium properties are considered to be simulation runs under engine transient operating
functions of temperature and fuel to air equivalence ratio. conditions.
The compressor is modelled using its steady state The transient engine operation case selected to be
performance map, which is provided as input in a investigated was the first one presented in [8]. According
digitized form containing lines of the turbocharger speed, to that, the engine operation for 100 seconds with changes
pressure ratio, corrected flow rate and efficiency. Given of the ordered engine speed from 94 rpm to 69 rpm at the
the turbocharger speed and pressure ratio, the corrected 10th sec and back to 94 rpm at 60th sec was examined. The
flow rate and efficiency are calculated using changes in ordered speed correspond to engine load
interpolation. The compressor pressure ratio is calculated changes from 100% load to 40% load. Such changes are
by the following equation using the pressures of the fixed considered very abrupt and are usually avoided in real
fluid and the scavenging receiver connected upstream and engine operation, but engine operation response under
downstream the compressor, respectively, the pressure similar load changes are used for the control system
drops in the air filter and air cooler and the pressure design. The results for the above mentioned transient run
increase in the auxiliary blower: presented in [8] were derived using a zero-dimensional
prC = ( pSC − ∆p AC + ∆pBL ) / ( patm − ∆p AF ) (32) engine simulation code, which was extensively validated
The air filter pressure drop is considered to be under steady state and transient conditions [8-10] and,
function of the compressor mass flow rate as follows: therefore, are considered of high reliability. So, they are
∆p AF = k AF m C2 (33) used in the Fig. 3 and 4 presented below as “reference”
data.
The blower pressure increase is taken as function of
A set of results derived using the 1st MVEM
the volumetric flow rate:
approach, including the engine crankshaft speed, the
∆pBL = f (VBL ) (34) turbocharger shaft speed, the scavenging receiver
The air cooler is modelled using eq. (15) and (16). pressure and the exhaust receiver temperature are shown
The temperature of the air exiting the compressor is in Fig. 3 (labelled as “model 1–T/C Inertia x 1”). As it is
calculated using eq. (22). The compressor absorbed deduced from Fig. 3, the engine speed response is
power is calculated by: adequately predicted during the first applied ordered
PC = m C ( hC − hatm ) (35) speed change (after the 10th sec). However, there is a
where the enthalpies of the air exiting and entering the slight deviation between the predictions for the engine
compressor are calculated from the respective speed and the respective reference data during the second
temperatures. applied ordered speed change (after 60th sec), where,
The turbine is modelled using its swallowing according to the predictions, the scavenging pressure
capacity and efficiency maps, which must be provided in limiter is not engaged as it occurs in the reference case.
digitized form. Given the turbine pressure ratio, the That is explained as follows. As it is shown in Fig. 3, the
turbine mass flow rate and efficiency are calculated using turbocharger speed, predicted using the 1st MVEM
interpolation. The temperature of the gas exiting turbine approach, varies more rapidly than the respective one of
is calculated by equation (28). The turbine power is the reference case, owing to the unmodelled dynamic of
derived by: the scavenging and exhaust receivers’ volumes. Thus, the
predicted scavenging receiver pressure during the second
PT = m T ( hER − hT , d ) (36) ordered speed change (after 60th sec) was greater than the
The model includes six differential equations (eq. respective one of the reference case, and in consequence
(29),(30) for engine scavenging and exhaust receivers and the scavenging pressure limiter was not engaged. Finally,
eq. (1),(2)), which are solved using a fourth order Runge- although the exhaust receiver temperature response is
Kutta integration method with fixed time step. qualitatively predicted, there is a noticeable deviation

ISSN: 1790-2769 110 ISBN: 978-960-474-120-5


Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

between the predictions and the reference data. value of the turbocharger polar moment of inertia is used.
A way to improve the model accuracy is through On the contrary, the prediction of the exhaust receiver
increasing the value of the turbocharger rotating parts temperature exhibits greater deviation from the respective
polar moment of inertia by an additional quantity. That reference data.
added inertia corresponds to the turbocharging system A set of simulation results derived from a) the 1st
volume inertia (scavenging and exhaust receiver volumes, MVEM approach using 75% increase in the actual value
compressor and turbine piping). The simulation run was of the turbocharger polar moment of inertia (model 1), b)
repeated increasing the actual value of the turbocharger the 2nd MVEM approach (model 2) and the respective
polar moment of inertia by 50%, 75% and 100%. The reference data (reference) are presented in Fig. 4. As it is
respective results are also presented in Fig. 3. As it is observed from Fig. 4, the engine parameters response
shown in Fig. 3, the responses of the engine speed, including the exhaust receiver temperature are predicted
turbocharger speed and scavenging receiver pressure are with very satisfactory accuracy using the 2nd MVEM
adequately predicted when 75% increase in the actual approach. This is attributed to the more detailed
100 13000
95
reference
model 1-T/C inertia x 1 12000
reference
model 1-T/C inertia x 1 modelling of engine receivers and, in a lesser extent, to
model 1-T/C inertia x 1.5 model 1-T/C inertia x 1.5
engine speed (rpm)

model 1-T/C inertia x 1.75 model 1-T/C inertia x 1.75


the use of complete compressor map. Adequate
T/C speed (rpm)

90 11000 model 1-T/C inertia x 2


model 1-T/C inertia x 2
85 10000

80 9000
agreement in the prediction of engine parameters (except
75 8000 from exhaust receiver temperature) is also obtained using
70 7000
the 1st MVEM approach. The execution time of the
65 6000
0 10 20 30 40 50 60
time (s)
70 80 90 100 0 10 20 30 40 50 60
time (s)
70 80 90 100
simulation run using the 2nd MVEM approach increased
in comparison to the respective one of the 1st MVEM
4 900
reference reference
exh. receiver temperature (K)
scav. receiver pressure (bar)

model 1-T/C inertia x 1 850 model 1-T/C inertia x 1


3.5 model 1-T/C inertia x 1.5
approach, but it was reasonable. Thus, the 2nd MVEM
model 1-T/C inertia x 1.5 800
model 1-T/C inertia x 1.75 model 1-T/C inertia x 1.75
model 1-T/C inertia x 2 750 model 1-T/C inertia x 2
3
700
650
approach can be reliably used in advanced engine control
2.5
600
550
studies, e.g. including, apart from engine speed control,
2
500 turbocharger speed or scavenging receiver pressure
1.5 450
0 10 20 30 40 50 60
time (s)
70 80 90 100 0 10 20 30 40 50 60
time (s)
70 80 90 100
control etc.
Fig. 3: Simulation results using the 1st modelling However, in cases of slow engine transients, which
approach and comparison with reference data taken from are usual for the ship propulsion plant operation, the
[8] engine parameters are also predicted with enough
accuracy using the 1st MVEM approach, as inferred by
1.2 100
engine speed (rpm)
rack position (-)

1 reference reference
90
0.8
model 1
model 2 80
model 1
model 2
comparing the results shown in Fig. 5, which are derived
0.6
70
from a slow transient run conducted with 1st and 2nd
0.4
0 10 20 30 40 50 60 70 80 90 100 60
MVEM approaches.
5000 time (s) 0 10 20 30 40 50 60 70 80 90 100
time (s)
engine torque (kNm)

12000
reference
T/C speed (rpm)

reference
4000 model 1
model 2
10000 model 1 4 Conclusions
model 2
3000 8000 The modelling of a large two-stroke marine Diesel
2000 6000
engine was presented by using two different mean value
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
approaches. The simulation results were used for
exh. receiver temperature (K)
scav. receiver pressure (bar)

time (s) time (s)


4 1000
reference
3
reference
model 1 800
model 1 revealing the advantages and drawbacks of each
model 2 model 2
approach.
2 600
The main findings derived from this work are
1
0 10 20 30 40 50 60 70 80 90 100
400
0 10 20 30 40 50 60 70 80 90 100
summarised as follows.
time (s) time (s)

Fig. 4: Simulation results using the 1st and 2nd modelling The first mean value engine model is simpler,
approaches and comparison with reference data taken requires smaller amount of input data and takes less
from [8] computing time for execution. It adequately predicts the
100 11500 engine speed response and in that respect, it can be
model 1 model 1
95
model 2
11000
model 2 reliably used in the engine speed control design process.
engine speed (rpm)

T/C speed (rpm)

10500
90
10000
In order to represent with acceptable accuracy the
85
9500 turbocharger shaft speed response during fast transients,
80
9000 the value of the turbocharger rotating parts polar moment
75 8500
of inertia, provided to the model as input, must be
70 8000

4
0 50 100 150 200 250 300 350 400 450 500
time (s)
800
0 50 100 150 200 250 300 350 400 450 500
time (s) increased by an additional amount for taking into account
model 1 the effect of the volume inertia of the engine scavenging
exh. receiver temperature (K)
scav. receiver pressure (bar)

model 1
3.5 model 2 750 model 2
700
and exhaust receivers. In such a case, additional data
3
650
(experimental data or simulation results from more
2.5
600 detailed models) are required in order to adjust the value
2
550 of turbocharger inertia. Thus, for advanced engine control
1.5
0 50 100 150 200 250 300 350 400 450 500
500
0 50 100 150 200 250 300 350 400 450 500
studies (e.g. including, apart from engine speed control,
time (s) time (s) turbocharger speed or scavenging receiver pressure
Fig. 5: Simulation results using the 1st and 2nd modelling control), such a model should be carefully treated.
approaches for a slow engine transient operation case The second mean value engine model requires

ISSN: 1790-2769 111 ISBN: 978-960-474-120-5


Proceedings of the 2nd International Conference on Maritime and Naval Science and Engineering

additional input data, including the compressor and HL : fuel lower heating value
turbine maps, the volume of the engine receivers and h : specific enthalpy
initial values for the pressure and temperature of the gas k : coefficients, constants
contained in the engine receivers. The engine physical m : mass
processes are more accurately represented since the m
 : mass flow rate
second approach includes the more detailed modelling of N : rotational speed
the engine scavenging and exhaust receivers. In that P : power
respect, the model gives better predictions for the engine p : pressure
operating parameters response during fast transients. In pr : pressure ratio
addition, the model execution time, although greater than p : mean effective pressure
the respective one of the first model, is reasonable Q : torque
enough. As a result, the second modelling approach is R : gas constant
more appropriate for advanced engine control system revcy: revolutions per cycle
design studies. T : temperature
t : time
References: u : specific internal energy
[1] Guzzella L. and Onder Ch.H., “Introduction to VD : displacement volume
Modeling and Control of IC Engine Systems”, Springer, w : velocity
2007. xr : rack position
[2] Eriksson L., “Modeling and Control of Turbocharged zcyl : number of engine cylinders
SI and DI engines”, Oil & Gas Science and Technology- αT : turbine flow coefficient
Rev. IFP, Vol.62, No. 4, pp. 523-538, 2007. γ : ratio of specific heats
[3] Meier E., “A Simple Method of Calculation and ∆p : pressure drop
Matching Turbochargers”, BBC Brown Boveri & ε : effectiveness
Company Ltd, Publication No. CH-T 120 163 E. η : efficiency
[4] Heywood J.B, “Internal Combustion Engines ηex: correction factor for the temperature of the
Fundamentals”, Mc-Graw Hill, 1988. exhaust receiver
[5] Watson N., and Janota M. S., “Turbocharging the
ρ : density
Internal Combustion Engine”, Macmillan Press, 1982.
ζ : proportion of the chemical energy of the fuel
[6] Theotokatos G., “A Modelling Approach for the
contained in the exhaust gas
Overall Ship Propulsion Plant Simulation”, 6th WSEAS
International Conference on System Science and φ : crank angle
Simulation in Engineering (ICOSSSE ’07), November
21-23, Venice, Italy, 2007. Subscripts
[7] Xiros N., “Robust Control of Diesel Ship
Propulsion”, Springer, 2002. AC : air cooler eq : equivalent
[8] Kyrtatos N.P., Theodossopoulos P., Theotokatos G., AF : air filter f : fuel, friction
atm : ambient ht : heat transfer
Xiros N., “Simulation of the overall ship propulsion plant
BL : blower i : indicated, intake ports
for performance prediction and control”, MarPower99
Conference, Newcastle upon Tyne, UK, 1999. b : brake MCR : maximum continuous
[9] Kyrtatos N.P., Theotokatos G., Xiros N., Marec K., C : compressor rating
and Duge R., “Transient Operation of Large-bore Two- comb: combustion o : initial conditions
stroke Marine Diesel Engine Powerplants: Measurements cy : cycle P: propeller
and Simulations”, 23rd CIMAC World Congress, cyl: cylinder SC : scavenging receiver
Hamburg, Germany, 2001. d : downstream Sh : shaft
[10] Kyrtatos N.P., Theotokatos G., and Xiros N.I., E: engine TC : turbocharger
(2002), “A Virtual Experiment Tool for Marine Diesel ER : exhaust receiver T : turbine
Engine Powerplant Analysis”, IMAM 2002, Rethymnon, e: exhaust gas, exhaust valve w : cooling water
Crete, Greece, May 13-17, 2002. ep : exhaust pipe
[11] MAN B&W, K90MC Mark 6 Project Guide Two-
Stroke Engines, 5th Edition, November 2000.

NOMENCLATURE
A : area
bsfc: brake specific fuel consumption
cd : discharge coefficient
cp : specific heat at constant pressure
cv : specific heat at constant volume
f : friction factor, fuel
Ι : polar moment of inertia

ISSN: 1790-2769 112 ISBN: 978-960-474-120-5