Cool flames in micro channel

Recent progress of weak flames in a micro flow reactor

with a controlled temperature profile

The 1st International Workshop on Near Limit Flames, July 29-30, 2017, Boston, MA, USA

Kaoru Maruta1,2, Ryota Tatsumi1, Hisashi Nakamaura1, Takuya Tezuka1

1: Tohoku Univ., Japan

2: Far Eastern Federal Univ., Vladivostok, Russia

A. Yamamoto, M. Hori, P. Grajetzki, T. Okuno, A.K. Dubey, T. Onda, Y. Kizaki

Contents

・ Outline of a micro flow reactor with

a controlled temperature gradient

・ Separated cool flame

・ Recent progress of weak flame studies

For larger fuel diversity, understandings of ignition and

combustion characteristics of various practical fuels are

important. Ex. bio-, synthesized-, reformed-fuels, etc.

A micro flow reactor with a controlled temperature profile

Background

Shock tube, RCM, burners have

been extensively used but

simple and compact method for

examining reaction dynamics is

required. TimeHe

at re

lea

se

ra

te

Cool flame

Hot flame

Two-stage

ignition

Microcombustion projects

Swiss roll

microcombustors

for heat sourcesMicro flow reactor with

controlled temp. profile

Microcombustion for fundamental combustion studies

4

Micro flow reactor with a controlled temperature profile

5

Stationary wall-temperature profile

Diameter of tube < conventional quenching diameter

Gas phase temperature dominated by wall temperature profile

Laminar flow and constant pressure

T

xoTest section xo

TmaxT

xoTest section xo

Tmax

dd

T

Mixture

Room temp

Tmax = 1300 K

d < Quenching

diameter

External heat source

Mixture

Flame behaviors in a micro flow reactor with a controlled temperature profile

Normal flame

Oscillatory flame

High velocity region

Weak flame

Low velocity region

Intermediate velocity region

6

Three kinds of flame responses

(1) Normal flame

(3) Weak flame

(2) Oscillatory flame

V=50cm/s

V=20cm/s

V=0.2cm/s-12 -10 -8 -6 -4 -2 0 2 4 6 8 10

0

20

40

60

80

100

120

= 1.0

Normal flame

FREI

Estimated points

of ignition

Location (mm)

Mean f

low

velo

city (

cm

/s)

400

600

800

1000

1200

1400

1600

Wa

ll Te

mp

era

ture

, Tw (K

)Ignition

Extinction

Tw

Flow

Normal flame, oscillatory flame, weak flame

CH4/air mixture

7Maruta et al., PCI 30

Theoretical S-shaped solutions

V

v

Stable

xStable

v

Unstable

Flow

Two stable and one unstable solutions predicted theoretically

Weak flame corresponds to ignition branch of S-curve

8Minaev et al., CTM 2007

experiments

9

Stable branch

Stable branch: weak flame

Unstable branch

timeChemical

time residense Flowa D

• Normal flame:

preheated premixed

flame

• Oscillatory flame:

flames with

repetitive extinction

and ignition, FREI

• Weak flame: stable

weak flame which

represent ignition

Analogy between S-curve and conventional Findell-curve

Da: Damköhler number

Max

imu

m t

emp

erat

ure

Ign

itio

n

Exti

nct

ion

Three flames are utilized for:

(1) Normal flame

(3) Weak flame

(2) Oscillatory flame

10

Measurements of laminar burning

velocity of highly preheated mixtures

Investigations of non-linear dynamics of

given fuels

Investigations of ignition relevant

properties of various fuels, validation

and modifications of kinetics,

particularly at low temperature

Triple weak flames, n-heptane

U = 3 cm/s

Triple stationary weak flames observed

Weak flame location insensitive to flow velocity

= 1

11

Yamamoto et al., PCI 33

Code PREMIX-based

Reaction scheme: n-heptane, reduced mechanism from LLNL

(159 species, 1540 steps)

Conditions: Stoichiometric gaseous n-heptane/air mixture

Experimental wall temperature profile was provided as Tw(x)

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 of heat-release-rate (HRR) [W/cm3] profile

Computations (one-dimensional)

Seiser et al., PCI 28 (2000)

12

Triple weak flames, n-heptane

U = 3 cm/s

Three-stage heat releases13

Computational

U = 3 cm/s

Three-stage heat releases14

Computational

Triple weak flames, n-heptane

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

U = 2.0 cm/s

Computational species profiles

15Three peaks of heat release rate

Yamamoto et al., PCI 33

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

LTO: CH2O, H2O2, CO, CH4 produced

U = 2.0 cm/s

Computational species profiles

First HRR peak

16

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

U = 2.0 cm/s

CH2O + OH ⇒ HCO + H2O

H2O2 (+M) ⇒ 2OH (+M) HCO + O2⇒ CO + HO2

Computational species profiles

Partial oxidations:

Second HRR peak

17

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

U = 2.0 cm/s

Full oxidations: CO + OH ⇒ CO2 + H

Computational species profiles

Third HRR peak

18

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

Measurement by GC

Measurements and computations

Three-stage oxidation process was

confirmed by gas sampling experiment

(U = 2.0 cm/s)

Computations

19

Yamamoto et al., PCI 33

Interpretation of triple weak flames

Conventional two-stage ignition converted into

steady, three-stage, spatially-separated weak

flames in MFR

Cool flame

Hot flame

Hea

t re

leas

e ra

te

time

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

n-heptane/air = 1

1 mm

Ignition in

RCM and ST

(transient)

Cool flame Separated hot flames

20

MFR applied for gasoline PRF

n-heptane + iso-octane (PRF)

Appearances of multiple weak flames

represent Research Octane Number21

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

= 1

U0 = 2 cm/s

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

1 mm

* Hori, et al., CNF 2012

U=1.2 cm/sec

Weak flames at different RON

Significant LTO in smaller RON

Weak flame behaviors reproduced22

Computation

* Hori, et al., CNF 2012

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200

Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

(a)

(d)

Flow direction

700 800 900 1000 1100 1200Wall temperature (K)

2 mm

120

Flow direction

PRF0

PRF20

PRF50

PRF100

700 800 900 1000 1100 1200

Wall temperature (experiment) [K]

-1.5 -1.0 -0.5 0 0.5Location [cm]

methanetoluene

propane

ethane

n-heptane, PRF 0

iso-octane, PRF 100n-butane

PRF 50

PRF 20

120

112

108

100

94

50

20

0

RON

Weak flame appearances for various RON

Weak flame locations: monotonic function of RON

Second weak flame: visible when RON < 100

First weak flame: visible when RON < 20

P = 1 atm

23

Structure of a separated cool flame in MFR

24

0

0.1

0.2

0.3

0.4

0

0.05

0.1

0.15

0.2

0.25

5 6 7 8 9 10 11 12 13

Mole

fra

ctio

n (

-)

HR

R (

J/s/

cm3)

1st flame

nC7H16×10

O2

H2O2×10

CH2O×10

CO×10

CO2×10

(cool flame)

KUCRS mech. Location (cm)

600 700650 600550

Wall Temperature [K]

Tatsumi et al., Heat Transfer Symposium 2017.

Also, will be presented at ICDERS 2017.

Vertical MFR

Small temp. grad.

Sym. & Higher res.

KUCRS mech.

25

Structure of a cool flame, computation

Tatsumi et al., Heat Transfer Symposium 2017.

Also, will be presented at ICDERS 2017.

600 700650 600550

Conditions : nC7H16/air, = 1, P = 1 (atm), V = 2 (cm/s), Exposure time 30 (minutes)

Wall temperature (K)

0

0.2

0.4

0.6

0

0.01

0.02

5 6 7 8 9 10 11 12 13 HR

R (

J/s/

cm3)

0

0.2

0.4

0.6

0

0.01

0.02

5 6 7 8 9 10 11 12 13

Mo

le f

ract

ion (

-)

Location (cm)

00.6

00.02

KUCRS mech.

LLNL mech.

nC7H16

H2O2CH2O

nC7H16 H2O2

CH2O

Reaction paths in a cool flame

RH

R

ROO

QOOH

OOQOOH

OQ-HOOH Cyclic ethers

Aldehydes + ester or ketone group

Alkenes + ester or ketone group

CH2O, H2O2

R : C7H15

Q : C7H14

Q-H : C7H13

(Ketohydroperoxide)

OQ-HO

← OH

← O2

→ OH

→ OH← OH

← O2

→ OH

Radical branching path

Radical propagation path

CO, CO2

27

Reactions in radical branching path

O C·C

C

C

C

CCO O C·C

C

C

C

CCO

O C·C

C

C CCCO

O C·C

C

C

C

C

CO

O C·C CC CCCO

C CC

O

C

O·

CCC C CC

O

C

O·

CCC C CC

O

C

O·

CCC C CC

O

C

O·

CCC

72 %

C2H5CHO + C2H5COCH2

C3H7CHO + CH3COCH2

CH3CHO + C4H9CO

C4H9CHO + CH2CHO

72 %

OQ-HO (C7H13O2)

OOQOOH

Aldehydes

+ ester or ketone group

QOOH (C7H14OOH)

OQ-HOOH

Conditions : nC7H16/air, = 1, P = 1 (atm), V = 2 (cm/s), Tw,max = 700 (K) KUCRS mech.

28

Reactions in radical propagation path

O C·C

C

C

C

CCO

O CC

C

C CCC·O

O CC

C

C

C

CC·OO CC

C

C

C

C·O

C

O CC

C

C CC·O C

C

C

C

C CC C

OCC

CO

C CCC

C

C

C CC C

O

C

93 %

C3H6 + C2H5COCH2

or

C4H8 + CH3COCH2

C3H6 + nC3H7CO

or

C5H10 + CH3CO

C4H8 + C2H5CO

80 %

QOOH (C7H14OOH)

Cyclic ethers

Alkene

+ ester or ketone group

Conditions : nC7H16/air, = 1, P = 1 (atm), V = 2 (cm/s), Tw,max = 700 (K) KUCRS mech.

33

PRFs, ultra lean, higher resolution

Vertical-type reactorGrajetzki, et al., 36 Symp, WIP.

34

CH4, ultra lean and Xe dilution

Okuno et al., PCI 36.

Conducted for analyzing

microgravity experiments

on interaction between

flameball and

counterflow flames

Fuel: CH4; O2/Inert = 0.141; = 0.3

35

Syngas, elevated pressures

H2:CO

H HO2

H2O2 OH

O2(+M)

+H

+O2

+M

Relevant to H2-O2 explosion peninsula

Tanaka et al., 36symp. WIPP.

Soot and PAH formations at temperature below 1400 K

38

CH4 CH3

CH2O

HCO

CO

CO2

CH3O

C2H6

C2H3

C2H2

C2H4

C2H5

CH2CHO

CH2COHCCO

+ CH3(+M)

+H

+OH

+ CH3

+ HO2

+O

+OH

+H

+O

+ HO2

+O2

+O2

+(M)

+H2O

+OH

+HO2

+O2

+(M)

+OH

+H

+ CH3

+O

+OH

+H

+CH3

+O

+(M)

+O2

+(M)

+O2

+O2

+H

+OH

+(M)+O

+O

+ HO2

* Dubey et al., CNF 2017.

Soot and PAH formations at temperature below 1400 K

39* Dubey et al., CNF 2017.

40

Effect of radical quenching at wall

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

0.7 0.8 0.9 1 1.1 1.2 1.3

Mo

le f

ract

ion

[-]

φ

Inert wall

Quench wall

Experiment

Inert wall (computation)

Quench wall (computation)

Experiment

• CO mole fraction of burned gas:

Quench case > Inert case

• Experimental results agree with Inert case

→ radical quenching effect negligible

Inert: GRI3.0

Quench: GRI 3.0 + Raimondeau

CH4/air; = 1; d = 1.5 mm, 1 atm

Kizaki, et al., PCI35 (2015)

Conclusions

42

Micro flow reactor with prescribed temperature profile

can be a new platform for study on chemical kinetics

particularly for low temperature conditions

Acknowledgements

IHI, IIC, HONDA R&D, MAZDA

IHI, NEDO, HITACHI, TOKYO GAS, JSPS, JST