Engine Simulation & Optimization Software

Thermodynamic engine simulation tools

The thermodynamic engine simulation tools are most
applicable for general engine analysis and they are
widely used because do not require large resources.

How to use them for diesel combustion optimization to

meet emission regulations?

Gasexchange model Combustion model

GT-Power 1D Wiebe; User model; Link with CFD; Standard tool

(Gamma Technology) DI Jet model (Hiroyasu).
BOOST 1D Wiebe; User model; Link with CFD;
(AVL) Mix Control Combust (MCC) model.
NO Combustion
AMESim 1D Wiebe; User model; Link with CFD; Optimization
(LMS International) Mix Control Combust (MCC) model.

WAVE 1D Wiebe; User model; Link with CFD;

(Ricardo) Hiroyasu.
DIESEL-RK 1D Wiebe; RK-Model Fast simulation +
Optimization of Combustion
 Performance of Diesel Combustion Models
• Multi-Dimensional (CFD) Require too much computational time
Formal optimization is not possible.

From IVC till EVO

Time: 2 days Time: 10 hours Time: 2 hours

A B C

• Quasi-Dimensional, Multi-zone RK-Model

11 Zones of Spray

Time: 30 seconds
instead of 4 days
in case A
 Diesel combustion models
Zero-dimensional, Multi-Dimensional
Quasi-dimensional, Multi-zone
Single-zone (CFD)

Heat Release is specified by empirical Jung, Assanis

factors…
- Wiebe;
… + injection profile + ...
- Watson;
- Austen & Lyn; Chiu, Shahed, Lyn
- Shipinski;
- Whitehouse & Way;
- (MCC) model; etc.

Rakopoulos, Hountalas
Bi, Han, Yang

RK-Model Hiroyasu
 Workability of Diesel combustion models for
engineering tasks of emission control
Zero-Dimensional, Multi-Dimensional
Quasi-dimensional, Multi-zone
Single-zone (CFD)

No, due to insufficient capabilities May be acceptable, if improved Require too much resources

Even the most advanced Hiroyasu model has The existing Quasi-dimensional multi-zone models have
failings: limitations at resolving combustion optimization tasks due to
- Does not account piston motion; - Insufficiently detailed consideration of determining processes
- Supports only easy shapes of piston bowls; of mixture formation, combustion, emission formation;
- Supports only central location of injector; - as a result they have Insufficient accuracy of simulation of
- Does not account interaction among sprays; combustion and emission.
- Does not account mass-exchange among
packages; So, the most actual problems of engine simulation and
- Does not account hitting of fuel on cylinder their optimization are out of capabilities of existing
liner and head. simulation tools

We offer to use an another concept of RK-Model

Multi-Zone quasi-dimensional model
where sprays are divided on zones using
both geometrical fundamentals,
and mixture formation & evaporation
conditions.
 DIESEL – RK : combustion model possibilities
Advanced features of diesel combustion model:
1. Original multi-zone fuel spray combustion model (RK-model) which
accounts:
a - fuel properties including bio-fuels and blends of bio-fuels with diesel oil;
b - few fuel injection systems in one cycle of dual fuel engine;
c - detailed piston bowl shape;
d - swirl profile and swirl intensity;
e - injection profile, including multiple injection and PCCI / HCCI;
f - number, different diameters and directions of nozzles holes;
g - detailed interaction of sprays among themselves in volume and on walls
accounting local walls temperatures.

2. Detail Chemistry simulation at NOx and Ignition Delay prediction.

3. Model of Soot formation.
4. Simulation of Dual Fuel; Gas; HCCI, Assisted HCCI engine concepts.
 DIESEL - RK

Options of ICE simulation tool:

• "Fuel Spray Visualization" code (animation of the simulation results).

• Built-in procedures of Multiparameteric optimization (15 methods of the

nonlinear programming).

• Tool for express data file creation for different kinds of engines.

• Simulation of different combustion concepts:

- Dual Fuel;
- Gas;
- PCCIO / HCCI;
- Prechamber:
- Assisted HCCI.
 0.5 0.5
lwj  K j Bsw w

Original multi-zone fuel spray model (RK-Model)

Schematic Fuel spray structure

Publications:
• SAE 2005-01-2119;
• SAE 2006-01-1385;
• SAE 2007-01-1908;
• SAE 2009-01-1956;
• SAE 2013-01-0882;
• JSAE 20159169;
• JSAE 20159328.

Character zones

Before spray and wall impingement: Additional zones

1. Dense axial core of free spray. 8. Fuel allocated on cylinder Head surface.
2. Dense forward front. 9. Fuel allocated on cylinder Liner surface.
3. Dilute outer sleeve of free spray. 10. Fuel allocated in crossing of NWF cores
formed by adjacent sprays.
After spray and wall impingement: 11. Fuel allocated in crossings of Fronts
4. Axial conical core of NWF. and Cores of free sprays.
5. Dense core of NWF.
6. Dense forward front of NWF. * NWF is the so-called Near-Wall Flow of
7. Dilute outer surroundings of NWF. air with high density of fuel drops
Representation of spray zones and piston bowl geometry
1. Analytical: - Piston bowl is set of straight cones and straight truncate cones.
- Spray zones are sets of sloping cones and loping truncate cones.

2. As a 3D mesh of cubic cells. Number of cells: ~ 80 per Cylinder Diameter

Piston crown with grooves for

injectors in OP diesel 88-Г

… corresponding 3D
mesh with cubic cells

Spray is a set of cone and truncate cones

A volume of every spray zone is a sum of
Volumes of all cells included into the zone.

The cells included into zones of few sprays

simultaneously form zone of sprays intersection.
 Spray tip penetration modeling
Modified Lyshevski’s equations
using dimensionless parameters

Penetration
at break up:
Penetration
at main phase:

where:

14.21
DS 
  1.3511 
Df   0.68764exp(d n )  0.88869 ln( d n )
This study  exp(d n ) 
dn – nozzle bore, mm.
Lyshevski

Experimental data:
SAE Pap. N 1999-01-0200.
SAE Pap. N 2000-01-0287.
SAE Pap. N 2002-01-0946
ILASS – Europe 2011, Estoril, Portugal,
Sept. 2011.
 Free spray contour angle modeling

Lyshevski’s equations

Penetration at break up: ga

Penetration at main phase: gb

The free spray contours obtained by different ways: where:

a) calculated with KIVA by Reitz and Bracco [33];
b) measured by Dan [34]; Fs = 0.00750.009
c) calculated by Jung and Assanis [35] using Usage of dimensionless parameters allows
Hiroyasu and Arai equations [36]; account properties of alternative fuels in
d) this study.
simulation.
33. Reitz, R. D. and Bracco, F. B. On the Dependence of Spray Angle and Other Spray Parameters on Nozzle Design and
Operating Conditions // SAE Paper 790494, 1979.
34. Dan T. The Turbulent Mechanism and Structure of Diesel Spray. Ph. D. Thesis, Toshisya University, 1996.
35. Dohoy Jung and Dennis N. Assanis. Multi-zone DI Diesel Spray Combustion Model for Cycle Simulation Studies of Engine
Performance and Emissions // SAE Paper No 2001-01-1246, 2001.
36. Hiroyasu, H., and Arai, M. Fuel Spray Penetration and Spray Angle of Diesel Engines // Trans. of JSAE, Vol. 21, 1980, pp.
5-11.
 0.5 0.5
lwj  K j Bsw w
Simulation of the fuel sprays in the swirling air flow

Phenomenological Model of Interaction of

Spray and their Near Wall Flow with Swirl
and Walls. Penetration of spray tip and boundaries of
NWF as functions of time
Frame #

Photo-record obtained by V.V.Gavrilov

Effect of impingement
angles gj Swirl profiles

Effect of
local swirl
velocity Wt

l wj  K j Bsw w
0.5 0.5

w

 Wt  U t cos   w ;
1.5
 C wj   0.5d 32 lb  Bs0.5 s0.5 ;
  
0
 w   s   sw ; la  As Э 0.35 exp  0.2  s  g ;

K j  sing 1 sing 3  1.2 1  sing j  2cosg j 3
 Allocation of air in the character zones
Scheme of air flows in a diesel spray
Air motion around fuel spray

Motion of Elementary Fuel Mass (EFM) from

injector to spray front zone lk and spray tip lm.

3
 U  2
l
   1
 Uo  lm

Mass of entrained air Dma for every EFM Dmf

is defined from momentum conservation:
U 0 Dm f  U k Cl Dm f  Dma 
 Preprocessor for Piston Bowl Design Specification
Specification by main dimensions Specification by coordinates of points

3D mesh is used
for piston bowl
specification

Detailed geometry of piston bowl and configuration of nozzles holes allows definition
of Coordinates and Time of spray with wall impingement.
 Allocation of fuel in the character zones
Truck diesel Yamz: Locomotive diesel Д49
S/D=140/130, RPM=1700 S/D = 260/260,
RPM=1000, BMEP=15 bar
Visualization of sprays evolution with account the swirl

Experiment:
Tractor diesel СМД
4L D/S = 120/140
RPM=1800,
BMEP = 8 bar.
 0.5 0.5
lwj  K j Bsw w

3D Fuel Spray Visualization code

3D visualization allows rotate animation, zoom and highlight sprays and zones

Computational time of spatial 7 sprays evolution simulation

(in thermodynamic cylinder model) is about 1 minute !
 0.5 0.5
lwj  K j Bsw w

Fuel Spray Visualization code

3D visualization of sprays evolution in 2 stroke large marine engine with 2
injectors in cylinder. Yellow bullets mark spatial intersection of sprays

#8
#1

#2 #7

#6
#3
#5

#4

Dark Green bullets mark spray # 3 & # 7; Blue bullets mark Near Wall Flows on cylinder head;
Dark Blue – their intersections
 0.5 0.5
lwj  K j Bsw w

3D Fuel Spray Evolution

3D visualization of sprays evolution in 2 stroke large
marine engine with 3 injectors in cylinder.
Yellow bullets mark spatial
Sprays settings window
intersection of sprays

Dark Green bullets mark spray # 4, #9 & # 14; Blue bullets mark Near Wall Flows on cylinder
head; Dark Blue – their intersections
Effect of Spatial intersection of sprays on HRR in engine with side
injection system
View from bottom
(through piston)
Red, Yellow, Green
5
and Blue are cells of
sprays core. 8 6
3 7
2 4
3610 1
Brown: cells of spray
Experiment
front. 3630
Swirl
3630
Cells of NWF on cylinder
liner

3670
Black: cells of sprays
cores intersection.

3710
Cells of Near Wall Flow
on the piston surface
Simulation of combustion in OP engine
18 L D/S = 230 / 2х300 6700 kW @ 900 RPM

Animation shows only 4 sprays from 1 injector.

■ Green bullets show intersection of the sprays

Simulation of fuel spray motion and combustion in
two-stroke diesel with side injection system

Engine: Mitsubishi UEC 45 LA

D = 450 mm S = 1350 mm
RPM = 158;
2 injectors: 4 x 0.75
Angles of holes in above view:
500 , 350, 90, -10
 Results of simulation of fuel sprays evolution with
DIESEL-RK software in comparison with published CFD
simulation and experiment

H.Nakagawa, Y.Oda, S.Kato, M.Nakashima and

M.Tateishi: "Fuel Spray Motion in Side Injection
Combustion System for Diesel Engines",
International Symposium COMODIA 90, pp. 281-286,
1990.
 3D Fuel spray visualization
Mitsubishi UEC 45 LA
D = 450 mm S = 1350 mm
RPM = 158;
2 injectors: 4 x 0.75
Angles of holes in top view:
500 , 350, 90, -10

Intersections of sprays:
(Yellow markers).
Red, Yellow, Green & Blue bullets 3D visualization of sprays evolution
are sprays core zones cells.
in diesel with side injection system
Bottom view through
piston surface
3610

Brown bullets are

3630 Sprays Front Zones 3630
Cells.

Zones of Near Wall Flow Near Wall

on the Cylinder Liner. Flow of “light
blue” spray
3670
Black bullets are the
cells where the spray
cores intersect each
other.
3710
3710 Prune Bullets are
Zones of Near Wall
Flow on the piston
surface Near Wall Flow of
“blue” spray
 Calculation of the zone temperature
Energy balance equation for every zone of the spray:

DU a  DU lf  DU vf  DQa IN  DQlf IN  DQvf IN  DQa OUT  DQlf OUT  DQvf OUT 

 pDV  H evap  DQ X
where: DU is difference of internal energy at the end and start of time step;
DQIN is energy, delivered into zone; DQ OUT is energy, removed from zone;
p is a pressure, DV is variation of the zone volume; Hevap is a heat for droplet evaporation;
DQ X  mvf 2  b HU is heat of combustion of fuel vapor in the zone.
DQX  mvf 2 b HU
Indexes: a – gas (air); lf – liquid fuel; vf – fuel vapor; fe – evaporated fuel;
1 and 2 mean start and end of time step; IN and OUT are delivering and removing.

d 32 2  d 32 mix  K i d
The diameter of the fuel droplets after evaporation 2
during time step d:

The diameter of the fuel droplets (SMD) in the 3

N1 d 32  N d 3

d 32 mix 
1 IN 32 IN
zone after their mixing; here N is a number of
droplets in zone.
2
N1 d 32 1  N d 2
IN 32 IN

  d 
3

Dm fe  mlf 1  Dmlf IN 
The mass of evaporated fuel is calculated using
1   32 2
 
  d 32 mix  
diameters of fuel droplets prior and after
evaporation:  
 Modeling of evaporation
d 32 2  d 32 mix  K i d
2
Evaporation rate of droplet is described by Sreznevski’s equation:
where d32 is a current Sauter Mean Diameter of droplets;
d is a time step.
Ki is evaporation constant (every i-zone has own Ki) Ki  4  106 NuDi D pi pSi  f

NuD is Nuselt number for diffusion process (Sherwood number). Every zone has own NuD.

Dp is Diffusion Coefficient (every zone has own Dpi) : D pi  D po Tki To  po p 

Dpi depends on Equilibrium Evaporation Temperature Tki and current pressure p;
pSi is Saturated Vapor Pressure at the temperature Tki (every zone has own pSi).

Tki of i - zone is calculating using energy balance around a droplet (express. of Virubov D.N.):
750
700
Ti  Tki 
l a Ti  Tki   D pi pSi C f Tki  T f   hevap  C fv
 650

2 
600


Т ki, K
550
500
450 p=3 bar
where: l is heat conductivity at Tki ; Ti is character 400
p=100 bar
350
temperature of i-zone; Cf and Cfv are heat capacity of
300
fuel and fuel vapor, Tf is injected fuel temperature. 300.0 500.0 700.0 900.0 1100.0
Тi, К
 Validation of results of numerical modeling
1 cyl MAN test engine D/S=320/440; RPM=750; BMEP=6.54 bar [*]

Сгорание начинается при СА ≈ 358.5 град. Однако, в расчете не получается

столь резкого скачка температуры во фронте струи, как это фиксирует
измерение, возможно в алгоритме расчета не достаточно оценена степень
выгорания паров топлива b в начальный момент объемного сгорания. Тем не
менее, расчетная температура в зоне фронта струи близка к
экспериментальному значению. Позднее, при СА > 360 град. относительно более
горячий передний фронт струи уходит из зоны измерения вперед, и его место
замещает более холодное ядро струи. (Задняя граница зоны фронта струи
удаляется более чем на 55 мм от форсунки.) Фиксируемая температура в зоне
измерения в это время остается высокой, она заметно превышает среднюю
расчетную температуру ядра, по крайне мере до момента времени СА ≈ 362
град., рис. 13. Отличие температур в данном случае объясняется тем, что
температура в ядре не равномерно распределена по его длине: чем ближе к
фронту струи, тем выше температура. А именно головная часть ядра попадает в * Fridolin Unfug, Uwe Wagner, Kai W. Beck, Juergen Pfeil, Ulf
зону измерения до момента СА ≈ 362 град., что подтверждается и результатами
визуализации развития струи. В расчетных же данных фигурирует средняя
Waldenmaier, Oguz Celik, Johannes Jaeschke and Juergen Metzger.
температура по объему зоны. Позднее, при СА > 362 град., когда в зону Investigation of Fuel Spray Propagation, Combustion and Soot
измерения попадает уже основной объем (срединная часть) ядра струи, Formation/Oxidation in a Single Cylinder Medium Speed Diesel Engine //
расчетная средняя температура ядра струи практически совпадает с ASME 2012 Internal Combustion Engine Division Fall Technical
результатами измерений. В результате следует отметить, что расчет достаточно
Conference, Vancouver, BC, Canada, September 23–26, 2012.
точно отражает температуру внутри струи, а значит и процессы массообмена,
 Improved ignition delay calculation
For engines with PCCI / HCCI the existing empirical equations for Ignition Delay prediction can
not be used and Detailed Chemistry Model was developed and implemented.

The Lawrence Livermore National Laboratory (LLNL) mechanism is used for diesel fuel.
At every time step the delay is calculated taking into account:
■ Pressure, ■ Temperature, Calculation for
■ Burnt Gas Fraction (EGR), ■ Air/Fuel Ratio. n-heptane (Diesel)
10
Surrogate Chemical Calculated
fuel mechanism
Measured (Fieweger
et al. 1997)

Ignition Delay, ms
Diesel 160 species
n-heptane
fuel 1540 reactions
1

n-heptane +
49 species
B100 Methyl
144 reactions
butanoate 0,1
1,0 1,1 1,2 1,3 1,4 1,5
1000 K/T
 Low temperature combustion simulation
Low Temperature Combustion (LTC) Model is used when High Temperature Combustion
(HTC) ignition delay exceeds some value. For engines with PCCI / HCCI the LTC delay QiLTC
is function of HTC delay QiHTC and EGR fraction C:

Fraction of fuel burning by LTC mechanism can be calculated with expression derived by
processing published data:

where Q = MAX (6.7, QiLTC ).

Heat release of LTC can be approximated with Wiebe expression, as a function of crank angle
j varied from the beginning of LTC (where j = 0) up to jz .

where: mv = 1.2+0.69 C is a mode of Wiebe function;

jz = 6…8 CA deg is a duration of the LTC.

Citation: Kuleshov A.S. Multi-Zone DI Diesel Spray Combustion Model

for Thermodynamic Simulation of Engine with PCCI and High EGR Level //
SAE Tech. Pap. Ser. – 2009. – N 2009-01-1956. – P. 1-21.
 Engine simulation software possibilities
Full cycle thermodynamic engine simulation tool DIESEL-RK has
following features for combustion optimization:
- Any location of sprayers.

Dual Fuel Injection System

- Arbitrary piston bowl shape.

- Arbitrary
sprays
configuration.
2. … by coordinates of points.
Data base of piston bowls is supported.
 Interface for specification of few Fuel Injection
Systems in one engine

2 stroke marine diesel

Every injector has own

injection profile
 Simulation of combustion in
Dual Fuel Engine
Diesel

Methanol

Every system supplies own fuel:

А – Diesel oil
В – Methanol
Diesel Methanol
DIESEL-RK allows control
sprays 3D evolution &
Injection profiles of both systems
intersections

Computational time ~ 1…2 min.

 Simulation of
combustion of
Methanol in Dual Fuel
Marine diesel W32

Experim. Simulat.
BMEP, 20.85 20.65
bar
рmax, бar 160 164

Spray tip penetration [mm]

SMD [micron]

Injection profiles

37
 DIESEL-RK capabilities
Full cycle thermodynamic engine simulation tool DIESEL-RK has
following features for combustion optimization:

-Any multiple injection strategy.

- Injection profile may be specified:

● as diagram; ● parametrically (for optimization of the shape).

- Effect of high injection pressure.

 Soot formation model
Phenomenological simulation method takes into
account features of sprayed fuel burning. It is
assumed, the soot is formed mainly by two ways:

- As a result of chain destructive transformation of

molecules of fuel diffusing from the surface of
drops to the front of a flame.
- Owing to high-temperature thermal
polymerization and dehydrogenization of a vapor-
liquid core of evaporating drops.

In parallel to this, the process of burning of soot

particles and reduction of their volumetric
concentration owing to expansion occurs.

Sauter Mean Diameter (SMD) of droplets is calculating

during injection of every portion of multiple injection.
Evaporation constants are calculated as functions of
pressure and temperature of zones.

Diagrams show soot formation in z-engine at Max

Torque point @1500 RPM having split injection:
Pilot injection is 15% and sepatation is 4 deg.
Injection pressure is a pressure before nozzles.
AUMET OY
 Simulation of soot emission in the diesel
over the whole speed range
Power Truck diesel S/D=120/120

Power

Comparison between calculated

and experimental data SFC

SFC

Measurement Simulation
 Illustration of high accuracy of ICE simulation
over the whole operating range (1)

Power
Truck diesel: S/D=140/130

Power

Click picture to zoom and start visualization

 Illustration of high accuracy of ICE simulation
over the whole operating range (2)
Comparison between calculated and experimental data
Power D is the relative error
Truck diesel: S/D=140/130 Measur. Calcul. D,%
Pow 244.3 252 3.1
Measur. Calcul D,% SFC 240 232 3.3
Pow 180.3 178. 1.2 Power NOx 1920 1869 2.6
SFC 219 222 1.4
NOx 2160 1990 7.9
Measur. Calcul. D,%
Pow 122.2 122. 0
Measur. Calcul D,% SFC 258 258 0
Pow 90.2 86 4.6 NOx 980 930 5.1
SFC 223 235 5.4
NOx 1430 1023 28
Measur. Calcul. D,%
Measur. Calcul. D,%
Pow 24.4 25.1 2.9
Pow 18 16.2 10
SFC 577 560 2.9
SFC 411 456 11
NOx 240 260 8.3
NOx 280 320 14
 Illustration of high accuracy of ICE simulation
over the whole operating range (3)

Characteristic of locomotive diesel S/D=260/260

Power,

Power

SFC,

SFC

Experiment Simulation

Click picture to zoom and start visualization

 Illustration of high accuracy of ICE simulation
over the whole operating range (4)
Comparison between calculated and experimental data
D is the relative error.
Characteristic of locomotive diesel S/D=260/260
Power SFC Air Flow Tt Smoke NO
Power,
D% 2.5 1.9 1.9 3.3 0 0.6

Power SFC Air Flow Tt Smoke NO Power

D% 0.7 0 6.2 0.9 14.2 2
SFC,

SFC

Experiment Simulation
Power SFC Air Flow Tt Smoke NO Air Flow is the Air flow rate;
D% 3.6 3.5 1 1.2 7.1 0.7 Tt is Turbine inlet temperature;
Smoke is the Bosch smoke number.
 0.5 0.5
lwj  K j Bsw w

Advanced NOx Formation Model

- Detail Kinetic Mechanism

for advanced diesel engines:
 with Multiple Injection or / and with high EGR
 working on alternative fuels: DME, Biofuel
The detail kinetic mechanism consists of two blocks:
- initial disintegration of a fuel molecule, consisting of 40 reactions with
participation of 10 species;
- the detail kinetic mechanism of methane oxidation and NOx formation,
consisting of 199 reactions and 33 species.

- Thermal Zeldovich’s mechanism

for conventional diesel engines
• Temperature in a zone of combustion is defined by zone model.
• On each step the equilibrium composition of 18 species is defined in a
zone of combustion.
O, O2 , O3 , H, H2 , OH, H2O,
• The calculation of NOx formation is carried
C, CO, CO2 , CH4 , N, N2 ,
out with the kinetic equation. NO, NO , NH , HNO , HCN
2 3 3
 0.5 0.5
lwj  K j Bsw w
Advanced NOx Formation Model
- Thermal Zeldovich’s mechanism can not be used for engines with large EGR.
- Detail Kinetic Mechanism (Basevich’s scheme)
DKM is intended for engines:
 with Multiple Injection or / and with massive EGR or/and with PCCI;
 working on alternative fuels: DME, Biofuel, etc.
The detail kinetic mechanism consists of two blocks:
1) The Initial disintegration of a fuel molecule, consisting of 40 reactions with 10 species;
2) The detail kinetic mechanism of methane oxidation and NOx formation, consisting of
199 reactions with 33 species.
• Temperature in a zone of combustion is defined by zone model.

Measured NOx and simulated NOx with Zeldovich and DKM

a) for 1 cyl. diesel engine S/D=66/82 mm) and 3600 RPM. b) 4cyl. 2 liters light duty diesel with
max BMEP=26 bar (massive EGR)
 Simulation of combustion in diesel with
different strategies of fuel injection
Comparison between calculated data and experimental ones
published by M. Bakenhus & R.Reitz: SAE pap. N 1999-01-1112

Caterpillar 3401
D/S=137/165; e=16.5
BMEP=10 bars
RPM=1600,
Injector: 6x0.259x125
 Simulation of NOx formation in diesel with
different strategies of fuel injection

Comparison calculated data with experimental ones published

by M. Bakenhus & R.Reitz: SAE pap. N 1999-01-1112
Caterpillar 3401
12
D/S=137/165; e=16.5
BMEP=10 bars Experiment
RPM=1600, 10

NOx emission, g/kWh

Calculation
Injector: 6x0.259x125
8

0
Single Single Single Double Double Triple Triple
Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7

www.diesel-rk.bmstu.ru
 PCCI modeling

Engine:
Peugeot DW10-ATED4
D = 85 mm S = 88 mm
RPMnom = 4000;
injector: 6 x 0.14

Triple pilot injection;

pilots fuel fraction: 28%

RPM=2600 BMEP=8.7 bar

Experimental data were published by:

Gary D. Neely, Shizuo Sasaki and Jeffrey A. Leet "Experimental
Investigation of PCCI-DI Combustion on emissions in a Light-Duty Diesel
Engine" SAE Pap N 2004-01-0121, 2004
 PCCI modeling

It is possible to define
duration and fraction
of each pilot to avoid
the hitting of the fuel
on the liner

Peugeot DW10-ATED4
(4L8.5/8.8)

RPM=2600

LTC: Low
Temperature
Combustion

HTC: High
Temperature
Combustion
 PCCI modeling
If Large Drops injected at the end of every portion have not enough time to be evaporated
completely the Air/Fuel eq. ratio being responsible for ignition delay is 1 (left diagram).
If the Large Drops are evaporated The Air/Fuel eq. ratio starts to grow up to total value being
character for whole cylinder; it results in: preparation of fuel to selfignition slows down. First portion
being ignited will have Integral reached 1 first.,
Peugeot DW10-ATED4 (4L8.5/8.8) RPM=2600; BMEP = 8.7 bar; Triple pilot: 28%
Injection timing : 70 deg .BTDC Injection timing : 90 deg. BTDC

i
Livengood – Wu integral of d
Ignition Delay: 
0 ign
1
 PCCI modeling

Peugeot DW10-
ATED4
(4L8.5/8.8)

RPM=2600
Double pilot 15%

LTC: Low
Temperature
Combustion

HTC: High
Temperature
Combustion

Experimental data were

published by:
Gary D. Neely, Shizuo Sasaki and Jeffrey A. Leet "Experimental Investigation of PCCI-DI
Combustion on emissions in a Light-Duty Diesel Engine" SAE Pap N 2004-01-0121, 2004
 Data base of fuels and Gas engines simulation
User can create own fuel and save one in the data base.
-- Blends of biofuels with diesel oil are supported.
-- Arbitrary mixed of gases are supported for gas engine. Properties of mixture are calculated
automatically

It is possible to set individual

fuel for every operating
mode. It allows presentation
of engine parameters as
function of fuel composition.

List of gases

H2 Hydrogen
O2 Oxygen
N2 Nitrogen
H2O Water Vapor
CO2 Carbon Dioxide
CH4 Methane
C2H6 Ethane
C3H8 Propane
C4H10 Buthane
CH3OH Methanol
CH3-O-CH3 Dimethyl Ether
C2H5OH Ethanol
 Variable Valve Flow Coefficient as
a function of Valve
Lift / Timing Analysis Lift

Valve Lift Diagram with variable

valve actuation can be set
individually for every operating
mode. Resulted Effective flow
area diagrams:
 Detail temperature fields of engine components
Mesh is generated
automatically

Account of
walls local
temperatures
at in-cylinder
processes
simulation.

Simultaneous
simulation of
thermo-
dynamic
processes
with Finite
Element
Analysis

Drag & drop to Boundary conditions Result temperature

Data base of engine assemble any and materials field is used for
parts is included combination of parts properties data base evaporation simualtion
 Link DIESEL-RK with another Simulation Tools

Run DIESEL-RK kernel under the

control of external codes

Interface Input
text
file

Diesel-RK
External code
solver

Output
text
file
 Engine parameters optimization problem
Optimization objectives: 1. Decrease of SFC Z1 = SFC = f(X) MIN
2. Decrease of particulate matter emission (РМ) and
nitrogen oxides emission (NOx) together.

Z2 = SE MIN
where index “0”
means required
3. … etc. values.

Arguments:

CR - Compression ratio; 
(independent

n, dn - Number and Diameter of injector nozzles;

variables) 
φ, θ - Injection Duration and Injection Timing; 
X  PR, EGR, Valve timing, Bypasses, etc. 

InjProf - Injection profile including strategy and parameters of multiple
injection;

 PistBowl - Piston bowl shape; 
a, - Injector nozzles directions in both planes.

The structured arguments: Injection profile, Piston bowl shape, Injector nozzles design are
assigned by user and may be varied by sequential retrieval.
Limits:
(restrictions)  Pz - Maximum cylinder pressure (Pz < 150 bar);

Y  Pinj - Maximum injection pressure (Pinj < 1500 bar);

 
Tt - Temperature before turbine;
SFC, etc.
 Solution of engine parameters optimization problem

1D problem: Z1 = SFC = f(X1) MIN;

example
X1 = EVO
Method: 1D scanning Decision is made by user

2D problem:
Z2 = SE (PM, NO) = f(X1, X2) MIN;
example
X1 = φ inj dur ; X2 = θ inj tim ; Y1= pinj < 1000 bar;…
Method: 2D scanning Decision is made by user
DIESEL-RK carries out the Drag and drop
simulation of ICE in the nods of technique to plot 3D
orthogonal grid. diagrams and plot
isolines.

Number of nods and space are selected by user.

 2D scanning results presentation
The results of scanning may be displayed as 3-D diagram
and isolines
 Multidimensional optimization of engine parameters
Engine 8 parameters
optimization at full load point.

8D optimization of engine
parameters.

Limitations: Pmax < 200 bar.

dp/dCA < 5 bar/deg.
 Solution of engine parameters optimization problem
Multiparametrical optimization
nD problem:
MIN;
example
Method:
Multiparametric
optimization by
means of nonlinear
programming

Library of DIESEL-RK
includes:
- 15 Procedures for
Multidimensional
optimum search and Decision is made by
- 4 Procedures for
optimization procedure
One-dimensional
(because graphic
search.
interpretation of result is
impossible).
 Calibration of the combustion model of light duty diesel

Comparison of
experimental and
measured HRR
and in-cylinder
pressure at
different 10
engine operating
points.

All empirical
coefficients are
same for each
point.
 Calibration of the combustion model of light duty diesel
Comparison of
experimental
and measured
HRR and in-
cylinder
pressure at
different 10
engine
operating
points.

All empirical
coefficients are
same for each
point.
 Calibration of the combustion model of light duty diesel
Comparison of
experimental
and measured
engine
parameters at
different 137
engine
operating
points.

All empirical
coefficients are
same for each
point.
 Calibration of the model performed by GM

Heat Release for Case 1 Heat Release for Case 2

0.04 0.035
Comparison of 0.035 0.03
experimental and 0.03
Diesel RK Case 1 Diesel RK Case 2
Experimental 0.025 Experimental
measured 0.025
0.02
parameters at 0.02
0.015

different 10 engine 0.015

0.01
0.01
operating points 0.005
0.005

0 0
340 360 380 400 420 440 340 360 380 400 420 440
-0.005 -0.005

Heat Release for Case 3 Heat Release for Case 4 Heat Release for Case 5

0.04 0.08 0.04

0.07 0.035
0.035
Diesel RK Case 3
0.06
0.03 0.03
Experimental
0.05
0.025 0.025
0.04
0.02 0.02 4
Case Ca
0.03
0.015 Exp
0.015 Ex
0.02
0.01 0.01
0.01
0.005 0 0.005
340 360 380 400 420 440
0 -0.01 0
340 360 380 400 420 440 340 360 380 400 420 440
-0.005 -0.02 -0.005
 Calibration of the model performed by GM
Heat Release for Case 6 Heat Release for Case 7
Comparison of
experimental and 0.07 0.08

0.07
measured 0.06
0.06
parameters at 0.05
0.05
different 10 engine 0.04
0.04 Case
Case 6
operating points 0.03
Exp 0.03 Exp

0.02
0.02

0.01 0.01

0 0
340 360 380 400 420 440 340 360 380 400 420 440
-0.01 -0.01

Heat Release for Case 8 Heat Release for Case 9 Heat Release for Case 10

0.08 0.04 0.1

0.07 0.035
0.08
0.06 0.03
0.05
0.025 0.06
0.04
0.02
Case 8 Case 9 C
0.03 0.04
Exp
0.015 Exp E
0.02
0.01 0.02
0.01

0 0.005
0
340 360 380 400 420 440
-0.01 0 340 360 380 400 420 440
340 360 380 400 420 440
-0.02 -0.005 -0.02
 180
AirFlow (kg/hr)
Calibration of the model
160
performed by GM
R2 = 0.9988

140 Comparison of experimental

Model

and measured parameters at

120
different engine operating points
100
89 experimental points were used
80
IMEP : Indicated mean effective
80 100 120 140 160 180

Experimental
pressure

Nox (ppm) IMEP (bar)

2500 28

26
2000 R2 = 0.9859
2
R = 0.8857 24

1500 22
Model

Model

20
1000
18

16
500
14

0 12
0 500 1000 1500 2000 2500 12 14 16 18 20 22 24 26 28
Experimental Experimental
 Comparison of the thermodynamic engine simulation programs
Accessible Functions for Engine Analysis  Express engine analysis (function of automatic
engine design prediction & empiric coefficients
• Overall Engine Analysis setting for the case of data deficit)
• Steady state operating modes Gas SI engines with prechamber (arbitrary gas)
• Difference between • Turbocharging analysis  Automatic Multi Dimensional Optimization
the cylinders • Gas Exchange analysis  2 parametrical researches
• Transient operating • Heat Exchange analysis  Advanced multi-zone DI diesel spray
modes simulation • Valve Timing optimization combustion model:
• Analysis of Noise • 4 stroke & 2 stroke engines.  Optimiz. of Piston Bowl Shape (& Data Base
• List of easy diesel • Junkers and OPOC engines. of piston bowls & advanced graphic interface)
combustion models, • Zeldovich NO formation model  Optimiz. of Injector design including central &
including Hiroyasu • Thermodynamic EGR analysis non-central sprayer as well side injection
model & user model. • Export/Import data via clipboard system (& 3D Fuel spray evolution visualization)
• Link with CFD spray • 1 parametrical researches.  Account of adjacent sprays interaction in volume
model using KIVA • Account of the swirl at and near the wall.
code. spray behavior simulation  Optimiz. of multiple injection strategy and
• Link with valve train • Phenomenological Soot model PCCI strategy (& advanced graphic interface)
simulators, etc.  Detail Kinetic Mechanism for NO formation
(199 reactions 33 spec.)
 Bio-Fuels and blends & Data base of fuels
 Detail Chemistry (LLNL mech. 1540 reactions) at
Ignition Delay simulation (PCCI / HCCI).
 Run under control of another software tools
 Coupled thermodynamic simulations with FEA
Existing (account how the local wall temperature effect
commercial in fuel evaporation)
engine
simulation tools
Common functions DIESEL-RK:
Specific functions Specific functions
 Additional options of DIESEL-RK
Simulation of GAS and DUAL FUEL ENGINES.
- Injection of WATER;
- Ignition by pilot diesel injection into PRECHAMBER