/25
Stabilized three-stage oxidation of gaseous
n-heptane/air mixture in a micro flow reactor
with a controlled temperature profile
August 5, 2010
The 33rd International Symposium on Combustion
Akira Yamamoto, Hiroshi Oshibe, Hisashi Nakamura,
Takuya Tezuka, Susumu Hasegawa and Kaoru Maruta
Institute of Fluid Science, Tohoku University
(4C01)
/25Institute of Fluid Science, Tohoku Univ. 2
It is important to know the ignition characteristics of various fuels.
Typical ignition process of practical hydrocarbon fuels
Time
Cool flame
Hot flame
Two-stage ignition at
different temperature range
BackgroundH
eat
rele
ase
rate
Cool flame
• Low temperature oxidation (LTO)
• Small heat release
Hot flame
• High temperature oxidation (HTO)
• Main heat release
Two-stage ignition phenomena strongly affect the combustion control.
⇒ Improvement of internal combustion engines.• Spark ignition engines
• Compression ignition (Diesel, HCCI) engines
• Gas turbines etc.
/25Institute of Fluid Science, Tohoku Univ. 3
Major approach: e.g., Rapid Compression Machines (RCMs)
Approaches to investigate the ignition characteristics
• Heat loss to chamber wall
• Roll-up vortices by piston motion
• Ignition in local spot → Propagation
Disparities with
the modeling that assumes
homogeneous combustion
Difficulties in RCM experiments…
Auto-ignition of methane/air mixture, Strozzi et al., CST 180 (2008)
Fuel/air
mixture
Rapid compression
Multi-zone/dimensional model should be required to reproduce the
experiments more accurately. But it needs higher computational cost.
/25Institute of Fluid Science, Tohoku Univ. 4
• Imposed wall-temperature profile in the flow direction
• Inner diameter of the tube < Ordinary quenching diameter
• Laminar flow (Re ≈ 1 ~ 100)
• Constant pressure
※Maruta et al., PCI 30, 32
Fuel/Air d < Quenching
diameter
x
Tw(x)
Test section0
Flame
External heat
Wall temperature profile
Quartz tube
Micro flow reactor with controlled temperature profile
/25Institute of Fluid Science, Tohoku Univ. 5
At extremely low velocity Gas-phase temperature ≒ wall temperature
※Maruta et al., PCI 30, 32
Fuel/Air d < Quenching
diameter
x
Tw(x)
Test section0
Flame
External heat
Wall temperature profile
Quartz tube
Micro flow reactor with controlled temperature profile
CH4/air, 1atm, U = 0.2 cm/s
Tsuboi et al., PCI 32
Stable weak flames exist at certain positions in temperature profile.
/25Institute of Fluid Science, Tohoku Univ. 6
Objectives
n-Heptane (n-C7H16)
– A basic liquid hydrocarbon fuel that shows typical multi-stage ignition
– One of the primary reference fuel of automotive gasoline
Investigate ignition and combustion characteristics of
gaseous n-heptane/air mixture using a micro flow reactor
with a controlled temperature profile.
/25Institute of Fluid Science, Tohoku Univ. 7
n-C7H16/Air
Flat flame
for heating
H2/Air
d = 1-2 mm
x
Tw
300 K
1300 K
Test section0
Quartz tube
Flame
Experimental setup
• Stoichiometric gaseous n-heptane/air mixture, Pressure = 1-4 atm
• Flame images were taken by CH-filtered digital still camera
• Gas sampling analysis by GC
Pressure
Regulator
/25Institute of Fluid Science, Tohoku Univ. 8
Flame code PREMIX-based 1-D steady code
Reaction scheme n-Heptane, Reduced mechanism from LLNL
(159 species, 1540 steps) ※Seiser et al., PCI 28 (2000)
Conditions • Stoichiometric gaseous n-heptane/air mixture
• Experimentally measured wall-temperature profile
as a boundary condition
Gas-phase energy equation
21 1
1 4( ) 0
K K
k k pk k k k w g
k kp p p p
dT d dT A dT A A NuM A Y V c h W T T
dx c dx dx c dx c c d
Heat transfer with the wall
Flame position Peaks in heat-release-rate (HRR) [W/cm3] profile
Computational method
/25Institute of Fluid Science, Tohoku Univ. 9
Combustion characteristics at atmospheric pressure
Results and Discussion
/25
400
600
800
1000
1200
1400
1
10
100
-14 -12 -10 -8 -6 -4 -2 0 2 4 6
Wal
l te
mp
erat
ure
[K
]
Mea
n f
low
vel
oci
ty [
cm/s
]
x [mm]
Normal flame FREI_ignition FREI_extinction1st weak flame 3rd weak flame Wall temp.
2nd reaction zone
Institute of Fluid Science, Tohoku Univ. 10
(a) Normal flame
(b) FREI
(c) Weak flames
Experimental flame positions against various flow velocities
Three different flame responses were observed by changing inlet flow velocities.
/25
400
600
800
1000
1200
1400
1
10
100
-14 -12 -10 -8 -6 -4 -2 0 2 4 6
Wa
ll tem
pera
ture [K
]
Mea
n f
low
velo
cit
y [cm
/s]
x [mm]
Normal flame FREI_ignition FREI_extinction1st weak flame 3rd weak flame Wall temp.
2nd reaction zone
Institute of Fluid Science, Tohoku Univ. 11
(a) Normal flame
(b) FREI
(c) Weak flames
Normal flame in high velocity condition
Flow
Normal flame Local mixture flow velocity = Flame propagating velocity
30
/25
400
600
800
1000
1200
1400
1
10
100
-14 -12 -10 -8 -6 -4 -2 0 2 4 6
Wa
ll tem
pera
ture [K
]
Mea
n f
low
velo
cit
y [cm
/s]
x [mm]
Normal flame FREI_ignition FREI_extinction1st weak flame 3rd weak flame Wall temp.
2nd reaction zone
(c) Weak flames
Institute of Fluid Science, Tohoku Univ. 12
(a) Normal flame
(b) FREI
Flow
t [msec] 0
1
2
3
4
6
10
23
IgnitionExtinction
FREI Unstable Flames with Repetitive Extinction and Ignition
Unstable FREI in middle velocity condition
30
4
/25
400
600
800
1000
1200
1400
1
10
100
-14 -12 -10 -8 -6 -4 -2 0 2 4 6
Wa
ll tem
pera
ture [K
]
Mea
n f
low
velo
cit
y [cm
/s]
x [mm]
Normal flame FREI_ignition FREI_extinction1st weak flame 3rd weak flame Wall temp.
2nd reaction zone
Institute of Fluid Science, Tohoku Univ. 13
(a) Normal flame
(b) FREI
(c) Weak flames
Weak flames in low velocity condition
Flow
2nd reaction zone
Three reaction zones in the flow directionStable weak flames
4
/25Institute of Fluid Science, Tohoku Univ. 14
Three HRR peaks in the flow direction
Observed three weak flames were reproduced.
U = 2.0 cm/s
Twall
HRR
Computational profiles of Tg and HRR of weak flames
Tgas
/25
300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 15
Form CH2O, H2O2, CO, CH4 etc.
U = 2.0 cm/s
Oxidation of fuel (n-C7H16)
Computational species profiles -1st reaction-
Low temperature oxidation
/25
300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 16
U = 2.0 cm/s
Computational species profiles -2nd reaction-
CH2O + OH ⇒ HCO + H2OH2O2 (+M) ⇒ 2OH (+M)
H2O2 decomposition
HCO + O2 ⇒ CO + HO2
/25
300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 17
U = 2.0 cm/s
Computational species profiles -3rd reaction-
Oxidation of CH4 & CO to CO2: CO + OH ⇒ CO2 + H
/25
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0
5
10
15
20
25
300 500 700 900 1100 1300
Ma
ss c
on
cen
tra
tio
n o
f C
H2O
[%
]
Vo
lum
etr
ic c
on
cen
tra
tio
n [
%]
Wall temperature [K]
O2
CO2
CH2O
CO
CH4×20 300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 18
Measurement with GC (U = 2.0 cm/s)
Computation (U = 2.0 cm/s)
Comparison of experimental & computational species profiles
Three-stage oxidation process was experimentally confirmed
by gas sampling analysis with GC.
Profile of each species qualitatively agreed well.
/25
300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 19
Present three-stage oxidation
Interpretation of the three-stage oxidation
Typical two-stage oxidation: Cool flame + Hot flame
Present three-stage oxidation: Cool flame + Separated hot flames
Time
Con
centr
atio
n
Cool flame Hot flame
CO
H2O2
CH4CH2O
Typical two-stage oxidation
From the species profiles…
(Blue flame & Hot flame)
/25
300
400
500
600
700
800
900
1000
1100
1200
1300
0
5
10
15
20
25
3.5 4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Mo
le f
ra
cti
on
[%
]
x [cm]
O2
CO2
CH2O×10 CO
CH4×20
n-C7H16×10
H2O2×10
Tw
Institute of Fluid Science, Tohoku Univ. 20
Present three-stage oxidation
Interpretation of the three-stage oxidation
Time
Con
centr
atio
n
Cool flame Hot flame
CO
H2O2
CH4CH2O
Typical two-stage oxidation
Ordinary approach
Strongly transient ignition
Micro flow reactor
Stabilized spatial profile
Micro flow reactor can contribute to chemical kinetic studies.
/25Institute of Fluid Science, Tohoku Univ. 21
Pressure dependence of three-stage oxidation
Results and Discussion
/25
500
600
700
800
900
1000
1100
1200
1300
0
20
40
60
80
100
4 4.5 5 5.5 6
Wa
ll t
em
pera
ture
[K
]
Hea
t re
lea
se r
ate
[W
/cm
3]
x [cm]
HRR (6 atm) Tw
HRR×6 (1 atm)
Institute of Fluid Science, Tohoku Univ. 22
1 atm: 3rd HRR peak is the strongest.
6 atm: 1st and 2nd HRR peaks are stronger than 3rd peak.
U = 1.0 cm/s
Pressure dependence of HRR profile (Computation)
Main heat release shift to the lower temperature side at high pressure.
/25
500
600
700
800
900
1000
1100
1200
1300
0
1
2
3
4
5
6
7
4 4.2 4.4 4.6 4.8 5 5.2 5.4 5.6
Wa
ll tem
pera
ture [K
]
Press
ure [a
tm]
x [cm]
1st HRR peak 2nd HRR peak 3rd HRR peak
Tw
Institute of Fluid Science, Tohoku Univ. 23
U = 1.0 cm/s
Increase of pressure • 1st & 2nd peak shift to lower temperature side.
• 3rd peak shifts to higher temperature side at 2-5 atm.
Pressure dependence of HRR peak positions (Computation)
/25Institute of Fluid Science, Tohoku Univ. 24
Pressure
U = 3.0 cm/s, d = 1 mm
Low and middle temperature reactions affect the whole ignition process
more significantly at higher pressure conditon.
Flow
• 1st and 2nd flames become strong, 3rd flame becomes weakened.
• 2nd and 3rd flames are clearly separated.
Weak flame images at high pressures
As increasing pressure from 1 to 4 atm…
1 atm
2 atm
3 atm
4 atm
/25Institute of Fluid Science, Tohoku Univ. 25
Ignition and combustion characteristics of a gaseous n-heptane/air
mixture were examined using a micro flow reactor with a
controlled temperature profile.
• Three different flame responses were observed by changing an inlet mean
flow velocity. Especially at the low velocity condition, three reaction
zones were observed in the flow direction.
• Three peaks of heat release rate in the flow direction were obtained in the
computation. Computational species concentration profiles agreed well
with the results of gas sampling analysis.
Conclusions
• Under the high pressure condition, 1st and 2nd flames become stronger,
and 3rd flame becomes weakened. This indicates that low and middle
temperature reactions affect the whole ignition process more significantly
at higher pressure condition.
/25Institute of Fluid Science, Tohoku Univ. 26
Thank you for your kind attention!