PVM modeling for multi-mode combustion
PVM modeling for multi-mode combustion
Contents
• Introduction
• Motivations
• Approach
• Results
• Conclusions
2
Introduction – Challenges to Modern Engine Design
• Reduce emissions and increase efficiency
– Modern engines organize combustion in different ways, examples:
Mixed mode, Wider range Air/Fuel/EGR ratio. Multi-Fuel ……
• A more sophisticated analysis tool is needed to:
– Predict/simulate Combustion
– Understand Combustion
– Design/control Combustion
• Designers expect more from analysis:
– More information: flame structure, location, rates, species, auto-
ignition/quench, energy release, emission formations….
– Wider operating region: lean-to-rich, low/high-Temp., fast/slow
reactions….
– More accuracy: physical processes, chemical processes,
interactions between physical and chemical processes
3
Contents
• Introduction
• Motivations
• Approach
• Results
• Conclusions
4
Motivations
• Detailed Chemistry Based Combustion
Modeling Approach
• Capable of Solving Different Types of
Combustion and Mixed-Mode of Combustion
• Capable to Cope with Multi-Fuel
• Low Computational Cost
5
Contents
• Introduction
• Motivations
• Approach
• Results
• Conclusions
6
Approach – Pre-Solved PVM Library
• Govern equations
• PVM-Library Input/Output:
– A set of pre-solved 0-D solutions is re-organized as
follow function in table form
– The independent variables and dependent variables
7
)(xFy
outputquantityspecies
propertiesthermal
ngtransportireaction
Y
AAM
C
y
EQ
EGR
C
UH
P
x
i
W
/
/
~~,
)(
)(
61
ii w
dt
dY
dt
dp
dt
dH
Approach – Pre-Solved PVM Library cont.
– Lookup Library
» CFD solver provides to Library, the library
returns any elements of after interpolating
8
x
y
Approach – CFD Flow Solver
• Govern Equations:
9
spray
i
i mx
U
t
~
spraym
i
ij
ji
jijm
xx
p
x
UU
t
U,
~~~
sprayH
it
t
ii
i mx
H
xt
p
x
HU
t
H,)
~
(
~~~
sprayii
i
j
t
t
ii
jijmw
x
Y
xx
YU
t
Y,)
~
(
~~~
kk
ik
t
ii
i DGx
k
xx
kU
t
k
)(
~
DG
xxx
U
t i
t
ii
i
)(
~
Approach – Combustion Modeling
• Progress Variable and Its Transport Equation
• Thermodynamic Property determination
10
C
it
t
ii
i SCx
C
xx
CU
t
C
)(
~
)()(
)()()(
0298298
0298298
tthtth
tththtC
TendT
TT
i
iiT KThYh )298(298
)//(4
5
3
4
2
321 KkgJTATATATAACP
)/(5432
6
554433221 kgJAT
AT
AT
AT
ATAH
)//( KkgJM
RR
W
universal
Approach – Combustion Modeling cont.
• HCCI combustion (base scheme):
– Solve progress variable transport equation with
library source term
– Lookup PVM-Lib to obtain thermal properties
and species
• Premixed combustion (premixed scheme):
– G-eq determines flame location in SP-
ignition/main combustion stages
– Model C distribution based on G distribution
– Lookup PVM-Lib to obtain thermal properties
and species
•
11
dt
dCC
Approach – Combustion Modeling cont.
• Non-premixed combustion (non-premixed
scheme):
– Return to flamelet concept - strain-effect model
– Solve progress variable transport equation with
library source term
– Lookup PVM-Lib to obtain thermal properties
and species
• Mixed Types of Combustion
– Fuel Scalar Dissipation Rate Level Controls
12
Approach – Combustion Modeling cont. Base Scheme
• For HCCI Combustion
13
CFD Solver
),,,,( CHPx i
PVM-Library
C iBY
Solve Progress Variable C Transport equation
Flamelet Solver for Emissions
61,....., AAMW
Approach – Combustion Modeling cont.
Premixed Scheme
• Flame location
• G-equation
• Flame thickness
• Flame Propagation Speed
14
GSGUt
Gt
1G
zoneburninglGl
zoneunburnedlG
zoneburnedlG
:5.05.0
:5.0
:5.0
5.1kCCl l
)( ltl SSSS
Approach – Combustion Modeling cont.
Premixed Scheme
• For Premixed Combustion
15
CFD Solver
),,
,,(
C
HPx
i
PVM-Library
C
iBY
Solve
Progress
Variable C
Transport
equation in
sub-zone of
G>0.5l and
G<-0.5l
Flamelet Solver for Emissions
61,....., AAMW
Solve G- Equation and
flame thickness l
Linearly distribute C
value inside flame
layer -0.5l<G<0.5l
Approach – Combustion Modeling cont.
Non-Premixed Scheme
• Return to Flamelet Concept
– Strain-Effect Model
– Scalar Dissipation Rate
16
spray
it
t
ii
i mx
Z
xx
ZU
t
Z
)
~
(
~~~
0))((
C
t
C
2"~~
~~ Z
kcx
~)
~
(2
)
~
(
~~~2
2"2"2"
2"~
it
t
i
t
ii
i
x
Z
x
Z
xx
ZU
t
Z
Z
Approach – Combustion Modeling cont.
Non-Premixed Scheme
• For non-Premixed Combustion
17
CFD Solver
),,
,,(
C
HPx
i
PVM-Library
C iBY
Solve
Progress
Variable C
Transport
equation
Flamelet Solver for Emissions
61,....., AAMW
Scalar Dissipation
Solver
Contents
• Introduction
• Motivations
• Approach
• Results
• Conclusions
18
Results – Natural Gas Engine Parameters
19
Ignition Mode SI MP
Bore/stoke/rod length 250/250/500mm
Engine speed 750rpm
Premixed fuel EQ and
EGR Conc.
0.4914/0.1
Assistant fuel in pre-
chamber
NG Diesel
Sweep Volume 12.3liters
Pre-chamber volume % 1.5%
Assistant fuel injection
masses
4.35mg 3.03mg
Assistant fuel injection
timings
652.6 708CA
Results
20
Results
21
Results– Animation
• MP mode – fuel field distribution
CH4 Diesel
Results– Animation
• MP mode – combustion progress variable and flame front
represented by iso-surface of G=0 Progress variable Flame Front
Results– Animation
• SI mode – combustion progress variable and flame front
Progress variable Flame Front
Contents
• Introduction
• Motivations
• Approach
• Results
• Conclusions
25
Conclusions
• 1. PVM-MF has been developed. Its main features are
– Detailed chemistry based
– Capable different types and mixed-mode of combustion
– Capable of multi-fuel operation
– Low computational cost
• 2. Chemistry solution stored in PVM-library. Couple
CFD/chemistry solution through progress variable
• 3. PVM -MF saves computational cost in two ways:
– Only solve Base species/Prog. Variable transport Eqs
– Pre-calculated library for Reaction/Properties
• 4. Practices proves PVM-MF works well
26
Thank You !
27