Part 4: Heat transfer, NOx and Soot Emissions

Reciprocating Internal Combustion Engines

Prof. Rolf D. Reitz

Engine Research Center
University of Wisconsin-Madison
2014 Princeton-CEFRC
Summer School on Combustion
Course Length: 15 hrs
(Mon.- Fri., June 23 – 27, 2014)

Copyright ©2014 by Rolf D. Reitz.

This material is not to be sold, reproduced or distributed without
prior written permission of the owner, Rolf D. Reitz.
1 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions

Short course outine:

Engine fundamentals and performance metrics, computer modeling supported
by in-depth understanding of fundamental engine processes and detailed
experiments in engine design optimization.
Day 1 (Engine fundamentals)
Part 1: IC Engine Review, 0, 1 and 3-D modeling
Part 2: Turbochargers, Engine Performance Metrics
Day 2 (Combustion Modeling)
Part 3: Chemical Kinetics, HCCI & SI Combustion
Part 4: Heat transfer, NOx and Soot Emissions
Day 3 (Spray Modeling)
Part 5: Atomization, Drop Breakup/Coalescence
Part 6: Drop Drag/Wall Impinge/Vaporization/Sprays
Day 4 (Engine Optimization)
Part 7: Diesel combustion and SI knock modeling
Part 8: Optimization and Low Temperature Combustion
Day 5 (Applications and the Future)
Part 9: Fuels, After-treatment and Controls
Part 10: Vehicle Applications, Future of IC Engines

2 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Challen, 1998

Engine heat transfer

Up to 30% of the fuel energy is lost to wall heat transfer
Can influence engine ignition/knock
Engine durability – catastrophic engine failure

Scorching Detonation Cracking

3 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Han, 1995
Wang, 2012
Heat transfer
Gas phase energy equation
 I 
    uI    p u   J    Qc  Q s  Q r
t
Radiation source term qw
  y   Dy u* / u* Dy
Q  r      I  r, Ω  dΩ  4 I b  r  
r

 Ω4 
wall
Wall heat flux (account for compressibility)
With radiation Without radiation
 g C p u*Tg ln Tg Tw    2.1y   33.34  G  u *  C p u*Tg ln Tg Tw 
qw  qw 
   2.5
2.1ln y 
 
2.1ln y   2.5

T 
dT qw  2.1 
2.1u * Tg ln  g 
  dT  Tw 
 2.1G  
dy  g C p  y  
 
dy  y   2.1ln y   2.5
 
G radiative heat flux = qwr
4 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Wiedenhoefer, 2003

Radiation modeling
Radiation Transfer Equation:
s
 Ω   I  r, Ω     anet   s  I  r, Ω    Ib r   S r, Ω 
4
anet net absorption coefficient,  s scattering coefficient
  anet   s extinction coefficient
 Tw4
Back body radiative flux (independent of angle) I b  r  

Scattering terms,  s , S ~ usually neglected compared to absorption

Radiation intensity at wall

G  qwr    n Ω I  r, Ω  dΩ   Tw4

n Ω 0

 surface emissivity
Discrete ordinates model
 nDir m m 
Q  r      I  r   4 I b  r  
r

 m1 

5 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Wiedenhoefer, 2003a

Soot and gas absorption

1
Total absorption coefficient ag T , P, Le   ln 1   g T , P, Le  
Le
anet  asoot  aCO2  H2O   1  1   fuel  1   CO  1   CO 1   
H 2O
2

Soot absorption
asoot  1260CsootT m-1 CO2 absorption bands

Wide band model for CO2 and H2O

 2C1 3 
 b band center , T    C2 T 
e  1 band center
Importance of soot:
1
agas  asoot  T
T

 
Q r  r   a   I  r, Ω  dΩ  4 I b  r     agas  asoot   T 4   Tgas
3
 Tsoot
5

 Ω4 

6 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Wiedenhoefer, 2000

Wall heat transfer

Conjugate heat transfer modeling
ERC - Heat Conduction in Components code (HCC)

Iterative coupling
between HCC
and CFD code

Unstructured
HCC Mesh
7 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Wiedenhoefer, 2000

Wall heat transfer

Cummins N14 engine

.........

Caterpillar SCOTE engine

Cut Plane

8 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Wiedenhoefer, 2003

Predicted piston temperature - CDC

F = 0.7
No Radiation
740 Run 1
Run 2 Effect of radiation on wall heat loss
Run 3
720
Total heat loss increased by 30% due to
With Radiation
Peak Piston Temperature [K]

700
Run 1 radiation.
680 Run 2
Run 3 34% - head, 19% - liner, 47% - piston.
660
Lowers bulk gas temperatures
640
Results in lower NOx and higher soot
620
NOx reduced by as much as 30% (ave)
600
16
580 F = 0.5
Uniform Temp / No rad
560 14 Non-uniform Temp / With Rad

-20 -15 -10 -5 0 F = 0.7

12 Uniform Temp / No rad
Start of Injection, ATDC Non-uniform Temp / With Rad
10
NOx [g/kWh]
8

-20 -15 -10 -5 0

Start-of-injection, ATDC

9 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Gingrich, 2014

Wall heat flux measurements

Dry Compressed Chilled

Air Water
Engine Geometry
AVL 415S Smoke
Choked Flow Meter
Base Engine GM 1.9L Diesel
Orifices Water Compression Ratio 16.3
Heater Displacement (Liters) 0.477
Horiba Stroke (mm) 90.4
Hydrocarbon Bore (mm) 82
Analyzer
EGR Heat Exchanger Intake Valve Closing -132° aTDC
Exhaust Valve Opening 112° aTDC
Horiba Analyzers Swirl Ratio 1.5 -4.8
NOx Stock (Re-
Air DC Dyno
CO
Heater O2 Piston Bowl Type entrant)
Exhaust CO2 Port Fuel Injectors
Port Intake CO2 Included Spray Angle 20°
Barrel Swirl #4
Fuel
Control Direct Injection Pressure (bar) 2 to 10
Heater Injectors
Valves #3 Injector Rated Flow (cc/sec) < 10
Intake Bosch Common Rail Injector
Exhaust Surge
Surge #2
Tank Number of Holes 7
Tank
#1 Hole Diameter (mm) 0.14
4-cylinder engine head Included Spray Angle 155°
cylinder #1,3,4 deactivated Injection Pressure (bar) 250 to 1000 bar

10 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Gingrich, 2014
Thermocouples

5 1 6 7 Power Piston
4 Converter
2 Transmitter
3 Primary Coil
(Engine
Mounted)

Inductive Data
Power Secondary Acqusition
Coil
Supply (Piston
Mounted)
Receiver
• Fourier analysis is applied to find dynamic heat flux
• Integral of the dynamic heat flux over the full cycle is zero
T (t )  Tm  [ An cos(nt )  Bn sin(nt )]
Steady Dynamic
k N
n
q  (Tm  Tl )  k  [( An  Bn ) cos(nt )  ( An  Bn )sin(nt )]
l n 1 2

11 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Gingrich, 2014

Combustion strategy effects - CDC / HCCI / RCCI

Mode 1 Mode 2 Mode 3 Mode 4

Speed (RPM) 1490 1900 2300 2300

IMEPg (bar) 4.2 5.7 5.7 8
CA50 (degATDC) 4 5 4.5 8
Swirl 1.5 1.5 1.5 1.5

Intake Temperature (C) 75 50 50 35

Intake Pressure (kPa) 115 130 130 188

ERG (%) 0 0 0 55

Regime Fuel
Fuels: HCCI 91PON Gasoline / n-heptane
RCCI F76 / 91PON Gasoline
CDC F76

12 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Gingrich, 2014
Heat Release Rate
150
Combustion strategy effects Heat release rate

Mode 3 Location 100

3 Temperature
150

AHRR [J/deg]
5.7 bar IMEPg CDC
5 deg ATDC CA50 145 HCCI
50
2300 rev/min RCCI

Temperature [C]
140
0
135 -20 -10 0 10 20 30 40
Crank Angle [deg]
6
x 10 Location 3 6 Location 7
5 130 x 10
5
Tm=169.1C Tm=191.8C
4 Tm=155.4C 125 4 Tm=182.1C
Heat Flux [W/m2]

Tm=140.6C

Heat Flux [W/m2]

Tm=158.5C
3 120 3
-400 -200 0 200 400
2 Crank2Angle [deg]

1 1

0 0
-20 -10 0 10 20 30 40 -20 -10 0 10 20 30 40
Crank Angle [deg] Crank Angle [deg]
13 CEFRC2-4, 2014
 6000
Part 4: Heat transfer, NOx and Soot Emissions Gingrich, 2014
5000
Combustion strategy effects - CDC / HCCI / RCCI

Integrated Heat Flux [J/m2]

Location 3 4000 Location 7
6000 6000
CDC CDC
3000
HCCI HCCI
5000 RCCI 5000 RCCI
Integrated Heat Flux [J/m2]

Integrated Heat Flux [J/m2]

2000
4000 4000

1000 CDC
3000 3000
HCCI
RCCI
0
1 2000
2 3 4 1 2000
2 3 4
Mode Mode

1000 1000

0 0
1 2 3 4 1 2 3 4
Mode Mode
Heat losses significantly less with low temperature combustion strategies

14 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Hendricks, 2014

Heavy-duty diesel heat flux data

Compare CDC and RCCI
combustion at matched CDC
CA50, load, Φg
(4.6°CA ATDC, 0.35)
RCCI piston heat flux
measured to be lower
than CDC
Area integrated HX and
temp. determined
RCCI
CDC RCCI

∫Piston HX fuel 7.7 5.9

energy (%)
GTE (%) 51.2 52.7

15 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions
Engine emissions - transportation & toxic air pollutants
Criteria air contaminants (CAC), or criteria pollutants
- air pollutants that cause smog, acid rain and other health hazards.
EPA sets standards on:
1.) Ozone (O3),
2.) Particulate Matter (soot):
PM10, coarse particles: 2.5 micrometers (μm) to 10 μm in size
PM2.5, fine particles: 2.5 μm in size or less
3.) Carbon monoxide (CO), 4.) Sulfur dioxide (SO2),
5.) Nitrogen oxides (NOx), 6.) Lead (Pb)
Toxic air pollutants - Hazardous Air Pollutants or HAPs known to cause or
suspected of causing cancer or other serious health ailments.
- Clean Air Act Amendments of 1990 lists 188 HAPs from transportation.
In 2001, EPA issued Mobile Source Air Toxics Rule:
- identified 21 MSAT compounds.
- a subset of six identified having the greatest influence on health:
benzene, 1,3-butadiene, formaldehyde, acrolein, acetaldehyde,
and diesel particulate matter (DPM).
Harmful effects on the central nervous system:
BTEX/N/S - benzene, toluene, ethylbenzene, xylenes, Naphthalene, Styrene
16 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Curtis, 2014

Engine emissions - transportation & toxic air pollutants

17 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions
Diesel emission solutions
– Selective Catalytic Reduction (SCR) and Diesel Particulate Filter (DPF)
US EPA 2010 HD soot: 0.0134 g/kW-hr
NOx: 0.2682 g/kW-hr.
1.) EGR?
Navistar – no SCR
Enabling technologies (Cost?):
Improved combustion bowl design - PCCI
Improved EGR valves, air-handling, VVA
Twin-series turbochargers, inter-stage cooling
High-pressure CR fuel injection (31,800 psi)
2.) SCR?
Cummins Cu-Zeolite with DEF for 2010
Claim 3-5% fuel economy gain (Class 8 truck 1% ≈$1,000 per year)
“StableGuard Premix” dose rate ~2% of fuel consumption rate
Cost? $3/gal? AdBlue at pump in Germany $12/gal
Volvo announced surcharge of $9,600 for 2010 compliance
(complex – dosing rate, DEF freezes at 12F, gasifies at 130F)
Plus $7,500 for 2007 compliance  AT system cost equals cost of engine!

18 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Yoshikawa, 2008

NOx modeling
Zeldo’vich thermal NOx mechanism

Rate controlling step due to high N2 bond strength

ERC 12-step NOx model is based on GRI-Mech v3.11 and includes:

Thermal NOx Zeldovich, 1946

Prompt NOx around 1000 K. Fenimore, 1979

Extensions
NO can convert HCN and NH3 Eberius, 1987
Interaction between NO and Soot Guo, 2007

19 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Kong, 2007

ERC 12 step NOx Mechanism

SENKIN2 used to
predict species
histories.
XSENKPLOT used to
visualize reaction
pathways and identify
important reactions
and species.
Reduced mechanism
validated for test
temperatures from
700K to 1100 K and
equivalence ratios
from 0.3 to 3.0.
Four additional species
(N, NO, N2O, NO2)
and 12 reactions
added to ERC PRF
mechanism

Detailed mechanism: Smith, GRI-mech, 2005

20 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Kong, 2007

ERC 12 step NOx mechanism

Diesel spray computations

Comparison of NOx predictions (T=900K, P=3.7MPa)

N  NO  N 2  O
N  O 2  NO  O
N + OH  NO  H
N 2 O  O  N 2  O 2
N 2 O  O  2NO
N 2 O  H  N 2  OH
N 2 O  OH  N 2  HO 2
N 2 O  M  N 2  O  M
HO 2  NO  NO 2  OH
NO  O  M  NO 2  M
NO 2  O  NO  O 2
NO 2  H  NO  OH
GRI mechanism results Reduced mechanism results

Detailed mechanism: Smith, GRI-mech, 2005

21 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Yoshikawa, 2008

CH radical and HCN bridge in fuel-rich regions

Constant
volume
SENKIN Competing?
analysis with
ERC n-heptane
mechanism &
CxHy group
GRI ver.3 NOx
mechanism
φ=1.0 φ=3.0
N group
Absolute Flux
normalized to
NO by
XSENKPLOT

Tini=769K
Pini=40bar
Time=100ms

22 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions

Influence of soot radiation on combustion and NOx

BW: measured
Musculus, 2005

Colored: prediction
Yoshikawa, 2009

23 CEFRC2-4, 2014 23
 Part 4: Heat transfer, NOx and Soot Emissions Yoshikawa, 2009

Influence of soot radiation on combustion and NOx

140 8

120 Musculus (2005) soot

6
100

Max SINL [a.u.]

NOx [g/kgfuel]

80

60
“NOx bump” 4

40 Model w/ radiation
2
20
Measured Musculus (2005)
0 0
-15 -10 -5 0 5 10 15 20 -15 -10 -5 0 5 10 15 20
SOI [CAD] SOI [CAD]
80
Model w/o radiation
70 Model w/ radiation
“NOx bump” not observed in prediction, but
60 Model w/o soot and radiation reduction in predicted NOx seen with retard of
NOx [g/kgfuel]

50 SOI (~ SOI=8 CAD ATDC)

40 Radiation lowers predicted NOx ~ 7.5 %
30 Absence of soot lowered predicted NOx ~ 2.5 %
20
NOx model underpredicts measured NOx
10
Predicted Magnitude sensitive to turbulent Schmidt #
0
-15 -10 -5 0 5 10 15 20
SOI [CAD]
24 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Kittelson, 1998

Particulate emissions Regulated emissions PM2.5

New challenge - engines must meet

particulate number-based regulations (PN).
Euro 6:
PN limit 6.0e11 particles/km for vehicles
produced after 2017.
California Air Resources Board (CARB)
LEV III:
Total PM mass: 3.8 mg/km for 2014
and 1.9mg/km for 2017
PN: 3.8e12 and 1.9e12 particle/km.

Greatest health risk - fine particles

can lodge deeply into the lungs

25 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions

Soot modeling at the ERC

Soot models

Multi-step
Two-step model Phenomenological PAH chemistry
(MSP) model

Patterson, SAE 940523 Kazakov & Foster, SAE 982463 Vishwanathan &
Kong, ASME 2007 Tao, 2009 Reitz, CST 2010
Vishwanathan & Reitz, SAE Tao, SAE 2006-01-0196
2008-01-1331 Tao, 2006
Vishwanathan & Reitz, 2009 Vishwanathan & Reitz, SAE 2008-
01-1331

Models of soot formation/oxidation – Kennedy, Prog. Energy Comb. Sc., 1997

Soot processes in engines - Tree and Svenson, Prog. Energy Comb. Sc., V2007

26 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions
Two-step model “tuning” constant
Hiroyasu soot
formation Msf =Asf  P0.5  exp(-Esf /RT)  MC2H2
Nagle and Strickland- 6
Constable (NSC) Mso =  Wnsc  Ms
oxidation ρs Dnom
d(Ms )
Net soot mass =Msf -Mso
dt

C2H2 soot precursor

Hiroyasu & Kadota, SAE 760129
Nagle & Strickland-Constable, 1962
ρs = Soot density = 2 g/cm3
Dnom = assumed nominal soot diameter
= 25 nm
Wnsc = NSC oxidation rate/area
Mc2h2 = C2H2 Mass, Ms = Mass of soot

27 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2008

Performance of two-step soot model

SANDIA spray chamber: Pickett, 2004

Soot mass comparison

Model predicted soot inception location
2 → Lift-off length position
Soot Mass (micro gms)

1.5
Expt.
1
Model
0.5
0
0 20 40 60 80 100 120
Distance from Injector (mm)

C2H2 inception occurs at lift-off location

Inclusion of PAH chemistry needed for accurate prediction of soot form/oxid.

28 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2008

Performance of two-step soot model

10 mm 90 mm

Sandia experiment
Pickett, 2004
Heptane
injection

HCHO flame soot Model

Vishwanathan, 2008

0 mm 100 mm

C2H2 inception occurs at lift-off location

Inclusion of PAH chemistry needed for accurate prediction of soot form/oxid.

29 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions

Phenomenological
soot models

30 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2009

Reduced PAH mechanism

Reduced PAH mechanism of Xi & Zhong, 2006 based
on detailed mechanism of Wang &Frenklach, 1997 was
integrated (20 species and 52 reactions)
A1 formation through propargyl radical (C3H3)
Higher aromatics formed through HACA scheme
(hydrogen abstraction, carbon addition)

Reaction Arrhenius parameters-A, n, E. (Units of A in mole-

cm-sec-K and units of E in cal/mole)
C3H3 + C3H3 → A1 2.0E+12, 0.0, 0.0
A1- + C4H4 ↔ A2 + H 2.50E+29, -4.4, 26400.0
A1+ A1-↔ P2 + H 1.10E+23, -2.9, 15890.0
A2-1 + C4H4 ↔ A3 + H 2.50E+29, -4.4, 26400.0
A1C2H* + A1 ↔ A3 + H 1.10E+23, -2.9, 15890.0
A3-4 + C2H2↔ A4 + H 3.00E+26, -3.6, 22700.0

A1 = benzene, A2 = naphthalene, P2 = biphenyl, A3 = phenanthrene, A4= pyrene, A1- = phenyl,

A2-1 = 1-naphthyl, A3-4 = 4-phenanthryl, A1C2H* = phenylacetylene radical

31 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2009

PAH species
Reduced PAH mechanism implemented considering up to 4 aromatic rings (pyrene)
- A3 (Phenanthrene) used as precursor for soot formation model

C2H2 Expt. soot mass

 g)

1
Sandia expts: Pickett & Idicheria, 2006

0.1
15% O2, A3 is precursor
Soot/Carbon mass in precursors (

A1

0.01 Expt.

1E-3 A2 15% O2

Soot mass fraction

1E-4 A3 CFD
A4
1E-5
0 20 40 60 80 100 X=0 mm ~ peak of 0.016 ppm X=85 mm
Distance from Injector (mm.)
Improvement in soot location
Amount of dry-carbon mass locked-up in aromatic precursors small compared to
measured soot

32 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

Soot model implementation

ω
1. Soot inception through A4: C16H10 (A4 ) 1
 16C(s) + 5H2 Graphitization

ω1 = k1 [A4 ], k1 = 2000 {s-1}

ω
2. C2H2 assisted surface growth: Leung, 1991 C(s) +C2 H2 
 3C(s) + H2
2

4 -1
ω 2 = k 2  [C 2 H 2 ], k 2 = 9.0  10 exp(-12100/T)  S {s }
YC(S) = soot mass fraction
2 -1 Surface area per
S = πd p N {cm } N = soot number density (per cc)
unit volume
1/3 ρC(S) = 2.0 gm/cm3
 6Yc(s)ρ 
dp =   {cm} Particle size MC(S) = MW of carbon
 πρc(s) N 
  Kbc = Boltzmann’s constant

Mono –disperse locally: All soot in a comp. cell have same diameter
Ca = agglomeration constant = 9
ω
3. Soot coagulation: Leung, 1991 nC(s) 
 C(s)n
3

1/6 1/2
 6Mc(s)   6K T  ρY
ω3 = 2Ca     bc   [ c(s) ]1/6  [N]11/6 {particles cm-3s-1}
 πρc(s)   ρc(s)  Mc(s)
   
33 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

Soot model implementation

1 ω
4. O2 assisted soot oxidation (NSC model): C(s) + O2 4
 CO
2
 K P  
 A O2   x+K P  (1-x)   S {mol cm-3 s-1}

12
ω4 = 
  B O2
Mc(s)   1+K Z PO  
 2 
x = PO (PO + (K T /K B ))
2 2 x = fraction of A sites
K A = 30.0  exp (-15800/T) {g cm-2 s-1 atm-1} (1-x) = fraction of B sites

K B = 8.0 10-3  exp (-7640/T) {g cm-2 s-1 atm -1} PO2 = partial pressure of O2

K T = 1.51 105  exp (-49800/T) {g cm-2 s-1} KA,B,T,Z = rate constants

XOH = mole fraction of OH
K Z = 27.0  exp (3000/T) {atm-1}
γOH = OH collision efficiency = 0.13
Fenimore, 1967

5. OH assisted oxidation (Modified Fenimore and Jones model):

ω 1
C(s) + OH 5
 CO + H 2 ω5 = (12) 10.58  γOH  XOH  T-1/2  S {mol cm-3 s-1}
2

34 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

Soot model implementation

6. PAH-assisted surface-growth
j
C(s) + PAH k,j 
ω6
 C(s+k) + H 2 , ω6 = γ ks  β ks  PAH k,j   N
2
π  K bc  T
β ks = 2.2   (d p + d PAH ) 2 cm3s -1
2  μ i,j

2mi
d PAH = d A 
3

k = number of carbon atoms and j = number of hydrogen atoms,

γks = 0.3 is the collision efficiency between soot and PAH, βks = collision frequency,
di = collisional diameter of PAH, dA = size of single aromatic ring = 1.393√3 Ǻ,
μi,j = Reduced mass of colliding species = Mass of PAH,
mi = mass of PAH expressed in terms of number of carbon atoms - k
Most models consider only mono-aromatic benzene as growth species.

35 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

Soot model implementation

7. Transport equations:
M  μ M μ T  πη
    M  v         ξM   SM ξ= 0.75 (1+ ) , η = 0.9
t  SC ρ ρ T
    8

convection diffusion Thermophoresis Source terms

M = ρYc(s) (soot species density) and N (number density) with N being treated
as passive species
Thermophoresis term implemented as a source term
SM = 16ω1 + 2ω2 +6ω6 - ω4 - ω5   Mc(s) {g cm-3 s-1} for ρYc(s)

 Mc(s) 

SM = 16ω1  - ω3  {particles cm-3 s-1} for N
 M nuci 
 
π
M nuci =  d3nuci  ρc(s)
6
dnuci = 1.25 nm (~100 carbon atoms)

36 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

Soot mass and particle diameter prediction

Premixed charge SI engine particulate modeling

70,000 cells at BDC,

including the intake
and exhaust manifolds
and cylinders.

Spark plug: at center of cylinder head.

Completely homogeneous fuel/air mixture at IVC
Experiment: EPA Tier II EEE certification fuel, 28% aromatics.
ERC KIVA code simulations:
DPIK ignition model, G-Equation combustion model.
Fuel: iso-octane/28% toluene by volume.
MultiChem mechanism:
ic8h18/nc7h16/c7h8/PAH (79 species & 379 reactions)

37 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

Soot formation prediction

0.03
F; C/O
0.78; 0.26
Incylinder soot (g/kg-f)

0.98; 0.33
1.2 ; 0.41
0.02 1.3 ; 0.44
1.4 ; 0.48
1.5 ; 0.51

0.01

0.00
-40 -20 0 20 40 60 80
Crank Angle (deg)

CAD 680 700 720 740 760 780 800

Predicted soot mass no longer reduces significantly after 80 ATDC.

Soot produced at 80 ATDC increases with increase of φ.
Soot formation dominates first and then soot oxidation begins to
play a key role. Peak in-cylinder soot mass increases w/ an
increase of φ.
•38

38 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

TDC φ =1.5

3500 4
Incylinder temperature
Temperature (K)

3000 TDC 3
G (-)
2500 2

G (-)
2000 1
1500 0
----------------burnt-----------------
1000 -1
500 -2
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm)
A4 mass fraction (-)

C2H2 mass fraction (-)

2x10-9 6x10-4
TDC A4 mass fraction
C2H2 mass fraction 5x10-4
2x10-9
4x10-4
-9
C2H2
1x10 3x10-4
2x10-4
5x10-10 A4
----------------burnt----------------- 1x10-4
0 0
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm)
39 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

TDC φ =1.5

O2, OH mass fraction (-)

Number density (#/cm3) Soot mass fraction (-)
4x10-6 101
TDC Soot mass fraction
O2 mass fraction O2
100
3x10-6 10-1
OH mass fraction
10-2
2x10-6
10-3
soot OH 10-4
1x10-6
---------------burnt------------- 10-5
0 10-6
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm)
1x1012 400
Number density

Particle size (nm)

1x1010 TDC Particle size
1x108 nd 300
1x106 200
1x104 ---------------burnt-------------
1x102 dp 100
1x100
1x10-2 0
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm)
40 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

800 ATDC φ =1.5

O , OH mass fraction (-)

Number density (#/cm3) Soot mass fraction (-) 1x10-6 1x10-4

2
Soot mass fraction
-7 800 aTDC
8x10 O2 mass fraction 8x10-5
6x10-7 soot OH mass fraction 6x10-5
OH
4x10-7 4x10-5
2x10-7 2x10-5
0 O2
0
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm) 500
8
Number density

Particle size (nm)

1x10 800 aTDC Particle size 400
1x106
300
1x104 nd
1x102 200
dp 100
1x100
1x10-2 0
-4 -3 -2 -1 0 1 2 3 4
Radial position (cm)
41 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Jiao, 2014

Particulate size distributions

Experiment [1] Simulation
10 10
1x10 1x10
reference FC/O


Averaged particle number density (#/cm )

9 expt

3
9
1x10 1x10

8 8

1x10 1x10 
dN/dlog(dp) (#/cm )


3

7
1x10 1x10
7

6
1x10 1x10
6

5 FC/O
1x10 
5
1x10
4 
1x10  4
1x10
3 
1x10  3
1x10
2 
1x10
2
10 100 1x10
dp (nm) 10 100
dp (nm)

Nearly identical PSDs until about φ =1.3, For φ <1.4, shape of PSDs is very flat and broad,
nd sharply declines with increase of dp. which is different from experiment, but looks
like PSDs for A/F of 14.6 for engine loads
When φ >1.3, nd consistently increases lower than 4 bar in Ref. [2].
with increasing φ , and decreases gradually For φ =1.4 and 1.5 , magnitude of nd of small
with increasing dp. particles are well represented, nd decreases
with increasing dp.

[1] Hageman, 2013. [2] Maricq, 1999

42 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

Soot in stratified charge engines

HTC-diff./ LTC
SANDIA optical engine – HTC/LTC premixed Early/late

Engine Parameter Amb. O2 % 21 12.7

SOI -7/-5 -22/0
Bore x stroke (cm) 13.97 x 15.24
Pin (bar) 2.33/1.92 2.14/2.02
Speed (rpm) 1200 Tin (C) 111/47 90/70

Compression ratio (CR) 11.2:1 Fuel (mg) 61 56

Pinj (bar) 1200 1600
Swirl ratio 0.5

Number of nozzle holes 8

Orifice diameter (mm) 0.196

Included angle 152°

Fuel Diesel #2

Sector angle 45

Expt. Data: Singh, 2007

43 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions Vishwanathan, 2010

SNL optical engine – HTC/LTC

10 Expt. (Singh et al. 2007) 1000
Model Predicted
In-cylinder soot formation/oxidation
HTC-Diffusion 900
8 800
Difference in HTC and LTC soot amounts well
700
captured
Pressure (MPa)

HRR (J/Deg)
6 600
500
20 HTC-Diffusion
4 400
HTC-Long ignition delay
300 18 LTC-Early injection
2 200 16 LTC-Late injection

In-cylinder soot (g/kg-f)

100
14 Solid - Expt. (Singh et al. 2007)
0 0 Dashed - Model Predicted
-50 -40 -30 -20 -10 0 10 20 30 40 50 12
CAD ATDC
10
10 Expt. (Singh et al. 2007) 2400 8
Model Predicted
2200
LTC-Early inj. 6
2000
8
1800 4
Pressure (MPa)

1600
6 2
HRR (J/Deg)

1400
1200 0
4 1000 -10 -5 0 5 10 15 20 25 30 35 40 45 50
800 CAD ATDC
600
2
400
200
Diffusion to premixed combustion, soot ↓
0 0 HTC to LTC, soot ↓
-50 -40 -30 -20 -10 0 10 20 30 40 50
CAD ATDC
Expt. Data: Singh, 2007
44 CEFRC2-4, 2014
 Part 4: Heat transfer, NOx and Soot Emissions

Summary and current directions

Integration of soot model with multi-component vaporization and chemistry models
Fuel breakdown + fuel
aromatic led PAH growth

H2 +
Inception

Co
a
gu
C2H2 assisted surface Oxidation by O2

l at
io
growth

n
Fue
l-ar
sur omat
Gasoline/Diesel fac ic a
eg s
row siste
th d + CO

n
tio
+ H2

la
a gu
Oxidation by OH

Co
H2 +

Need development/improvement

Extension to GDI and H/P/RCCI

Organic fraction modeling:
OF correlates with premixedness
Soot diameter comparisons with TEM
measurements obtained from various
combustion modes

45 CEFRC2-4, 2014