A Shock Tube Study of the Auto-Ignition of Toluene/Air Mixtures at

Experimental Study of theOxidation, Ignition, and Soot Formation

Characteristics of Jet Fuels

Matthew Oehlschlaeger,Hsi-Ping Shen, Jeremy Vanderover, Shane Daley, and Andrew Berkowitz

Mechanical and Aerospace EngineeringRensselaer Polytechnic Institute

Sponsor: U.S. Air Force Office of Scientific Research

Period: 2/07-present

Outline

• Motivation• Objectives• New Shock Tube Facility• Ignition time measurements and

kinetics discussion – iso-octane, n-alkanes, cycloalkanes, aromatics

• Summary

Motivation

• Developing and improving combustion systems requires multi-physics models– One important piece of these models is the

description of fuel oxidation (kinetic mechanism)

• Development of kinetic mechanisms requires validation targets from controlled experiments– Shock tubes, rapid compression machines, flow

reactors, jet-stirred reactors, flames, etc.

• These target measurements provide valuable insight into fuel kinetics

Objectives• Provide validation targets and kinetic insight for

larger hydrocarbon fuels of relevance to jet fuels– Needed for the development of kinetic mechanisms

and surrogate representations of jet fuels– Little data is available for many larger compounds

Recent and current activitiesIgnition delay time measurements

Hydrocarbon fuel components (the focus of this talk)Jet fuels and surrogate mixtures

• Future activities– Carbon monoxide measurements (additional kinetic

information, particularly regarding early time oxidation in the NTC regime)

– Soot formation

Jet fuel composition

n-alkanes, 28%

iso-alkanes, 29%

cycloalkanes, 20%

aromatics, 18%

other, 5%

Ignition delay studiesStatus of ignition exps

1-methylnaphthalene X

Fuel Complete In progress Future

iso-octane X

heptamethylnonane X

toluene X

Jet fuels (JP-8, Jet A)

X

Jet fuel surrogate mixtures

X

xylenes X

ethylbenzene X

tetralin X

• Recent and current efforts are focused on characterizing the ignition delay of hydrocarbon components relevant to jet fuel

• In the near future investigations of jet fuels and surrogate mixtures will begin

Xdecalin

Xethylcyclohexane

Xmethylcyclohexane

Xcyclohexane

Xcyclopentane

Xn-hexadecane

Xn-tetradecane

Xn-dodecane

Xn-decane

Xn-heptane

In progressCompleteFuel

Status of ignition exps

RPI shock tube facility

Heated and insulatedmixing vessel

Driver

Heated and insulateddriven section

Test location w/optical access

Shock velocitydetection

Diaphragm

• Inner diameter = 5.7 cm• Reflected shock pressures up to 200 atm• Initial temperatures up to 180 ºC• Designed for kinetic studies of low vapor pressure

fuels at engine-like conditions

Mixing manifold Vacuum section

Heated shock tube uniformity

0 50 100 150 200 250 300 350 40020

40

60

80

100

120

140

160

180

200

diaphragm@ x = 411 cmIn

tern

al W

all T

empe

ratu

re [°

C]

Axial Position [cm]

Heated Shock Tube Temperature Profiles

shock tubeendwall @ x = 0

Ignition time measurements

• Shock velocity is measured over last meter to determine endwall conditions

• Endwall ignition time measured using OH* emission

to DAQ

to counter-timers for velocity

Kistler transducerto DAQ

incidentshock

reflectedshock

PZTs

Si photodetector (OH* emission)

filter (UG-5 Schott glass)

Incident shock velocity

• Attenuation requires multiple accurate velocity measurements• Uncertainties

– Room temperature shock tube: 1% & 1.5% in T & P respectively– Heated shock tube: 1.5% & 2.0% in T & P respectively

0 200 400 600 8000.92

0.93

0.94

0.95

0.96

0.97

0.98

Velo

city

[mm

/μs]

Axial position [mm]

n-decane/air, phi = 0.5Shock tube heated to 60°C Reflected shock condition: 1082 K, 49.1 atmIncident shock attenuation: 3.5%/m

endwall

Shock tube non-ideal gasdynamics

0 1000 2000 3000 4000

0.0

0.5

1.0

1.5

2.0

Sig

nal [

arbi

trary

uni

ts]

Time [μs]

pure N2 driven20% N2/He driverInitial reflected shock conditions:

961 K, 9.7 atmdP/dt = ~2% / ms

Estimated conditions at 4 ms980 K, 10.5 atm

• Boundary layer induced attenuation causes P & T rise over time• To date, we have chosen to only report ignition measurements

for test times shorter than ~3 ms to avoid influence from condition changes

Shock tube ignition example

0 200 400 600 800 10000.0

0.5

1.0

1.5

2.0

2.5

endwall OH*emission

Toluene/air, φ = 1.01082 K and 46.9 atm

Sig

nal [

arb.

uni

ts]

Time [μs]

τ = 757 μs

sidewall pressure

Example ignition data set

0.7 0.8 0.9 1.0 1.1101

102

103

Typical parameter space• Fuel/air mixtures at Φ = 0.25, 0.5, and 1.0

• Two pressures (10-15 atm and 40-50 atm)

• Variable temperature range (limits of 700 and 1400 K)

methylcyclohexane / air, φ = 0.25 12 atm data 50 atm data Pitz et al. mechanism, 12 and 50 atm Orme et al. mechanism, 12 and 50 atm

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

(a)

0.7 0.8 0.9 1.0 1.1101

102

103


Igni

tion

Tim

e [μ

s]

1000/T [1/K]

(b)

0.7 0.8 0.9 1.0 1.1 1.2 1.3101

102

103


Igni

tion

Tim

e [μ

s]

1000/T [1/K]

(c)

Iso-octane: influence of diluent gas (LLNL simulations)

0 500 1000 1500 2000

1000

1010

1020

1030

0 1000 2000 3000 4000

1000

1500

2000

2500

3000

3500

Time [μs]

Tem

pera

ture

[K]

Time [μs]

10152025303540

0 1000 2000 3000 4000

iso-octane/O2/diluent1000 K, 10 atmφ = 1.0, 20.66 mol-% O2

constant volume constraint Ar diluent N2 diluent

Pre

ssur

e [a

tm]

• Argon mixtures have shorter (~15%) ignition times relative to N2 mixtures due to lower specific heat

Iso-octane: influence of diluent gas (experiment)

0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10101

102

103

104

iso-octane/O2/diluent10 atm, φ = 1.0, 20.66 mol-% O2

current data, N2 bath current data, Ar bath Curran et al. mechanism, N2 and Ar Glaude et al. mechanism, N2 and Ar

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

N2

Ar

• Measurements have been made for iso-octane for both Ar and N2bath gases

• Ar mixtures display ignition times that are 15-20% shorter than N2mixtures

• Mechanisms predict 15% difference

Influence of diluent

0 200 400 600 800 1000 1200

05

10152025 Simulated pressure for above experiment

Constant volume constraintCurran et al. mechanism

Time [μs]

Pres

sure

[atm

]Time [μs]

13.0 atm

05

10152025

0 200 400 600 800 1000 1200

induction peroidheat release

iso-octane / airφ = 1, 20.66 mol-% O2

1126 K, 8.3 atm

Pre

ssur

e [a

tm]


~13.0 atm

Measured sidewall pressure

reflected shockbifurcation

0 500 1000 1500 20000

10

20

30 Simulated pressure for above experimentConstant volume constraintCurran et al. mechanism

Time [μs]

Pres

sure

[atm

]

Time [μs]

14.7 atm

0

10

20

30

0 500 1000 1500 2000


iso-octane / O2 / Arφ = 1, 20.66 mol-% O2

1069 K, 9.2 atm

Pre

ssur

e [a

tm]


~14.7 atm

Measured sidewall pressure

• Measured induction period heat release (pressure rise) similar to that predicted by LLNL mechanism

• Mechanism overpredicts ignition time

0.8 0.9 1.0 1.1

102

103

104

iso-octane/air, φ = 1 current study, 10 atm Davidson et al., 10 atm Fieweger et al., 12.8 atm

Ig

nitio

n Ti

me

[μs]

1000/T [1/K]

Iso-octane: Comparison with previous data

0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5

102

103

104

iso-octane/air, φ = 1current study, 25 atmFieweger et al., 16.8 atmFieweger et al., 33.6 atm

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5

102

103

104

iso-octane/air, φ = 1 current study, 50 atm Davidson et al., 50 atm Fieweger et al., 39.5 atm Fieweger et al., 44.4 atm

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

• Agreement with previous Aachen and Stanford data

N-alkane ignition

0.7 0.8 0.9 1.0 1.1 1.2101

102

103

φ = 0.5Current study, 12 atm

n-heptane n-decane n-dodecane n-tetradecane data bands

Igni

tion

Tim

e [μ

s]

1000/T [1/K]0.8 0.9 1.0 1.1 1.2

101

102

103

φ = 0.5Current study,40 atm

n-heptane n-decane n-dodecane n-tetradecane data bands

Igni

tion

Tim

e [μ

s]1000/T [1/K]

• Measurements have been made for n-heptane, n-decane, n-dodecane, and n-tetradecane at ~12 atm and ~40 atm and for Φ = 0.25, 0.5, and 1.0 (20 data sets)

• The ignition times for all alkanes are essentially the same, within measurement uncertainties, for mixtures with common carbon content (true for all conditions studied)

x2 in τ

x2 in τ

N-alkanes: all available data

0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6101

102

103

104

105

current study, n-heptane current study, n-decane current study, n-dodecane current study, n-tetradecane Curran et al., n-heptane, 13.5 bar Westbrook et al., n-tetradecane,

13.5 bar Data bands illustrating a factor of

three in ignition time (Δlog(τ) = +/-0.238)

φ = 1.0 experimental studiesall data scaled to 12 atm using τ ~ P-1

Gauthier et al., n-heptane Ciezki et al., n-heptane Zhukov et al., n-decane Pfahl et al., n-decane Kumar et al., n-decane Vasu et al., n-dodecane

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

x3 in τ

N-alkane kinetics

Measured and predicted ignition times are essentially independent of alkane length• High-T: all alkanes lead to similar pool of C3H6, C2H4, CH3, and H; H + O2 controls• Moderate-T: all alkanes lead to similar alkenes; H2O2 + M controls• Low-T: size of R does not significantly influence the rates of peroxy reactions• Note: small increase in reaction rate in the NTC regime with increasing alkane

length observed experimentally and predicted by kinetic simulations (isomerization slightly easier for longer chains)

RH

R

RO2

H,O,OH,HO2

O2

QOOHO2

OOQOOH

OH + R’OOH R’O + OH

Cyclic ether + OH

R1 + alkene

HO2 + alkene

High temperature (T > 900-1000 K)

Moderate temperature (700-1000 K)

Low temperature (500-800 K)

RH + HO2 R + H2O2

H2O2 + M 2OH + M

Further H-abstraction and β-scissionRH

R

RO2

H,O,OH,HO2

O2

QOOHO2

OOQOOH


Cyclic ether + OH

R1 + alkene

HO2 + alkene




RH + HO2 R + H2O2

H2O2 + M 2OH + M

RH

R

RO2

H,O,OH,HO2

O2

QOOHO2

OOQOOH


Cyclic ether + OH

R1 + alkene

HO2 + alkene




RH + HO2 R + H2O2

H2O2 + M 2OH + M

Further H-abstraction and β-scission

Cycloalkanes

0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4

102

103

cycloalkanes/air, φ = 1.0 and 50 atm decalin cyclopentane methylcyclohexane ethylcyclohexane cyclohexane

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

• Measurements have been made for five cycloalkanes at ~15 atm and ~50 atm and for Φ = 0.25, 0.5, and 1.0 (28 data sets)

• Ignition times vary in the order:decalin (longest τ) > cyclopentane > methylcyclohexane > ethylcyclohexane > cyclohexane (shortest τ)

Cyclohexane and methylcyclohexane comparisons

0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.510-1

100

101

102

103

104

105

MCH / air, φ = 1.0, 50 atm current 12 atm data scaled to 50 atm current 50 atm data Vasu et al., 45 atm scaled to 50 atm Pitz et al., 15 atm scaled to 50 atm Pitz et al., 20 atm scaled to 50 atm Orme et al., 1 atm, 1% MCH/O2/Ar

scaled to 50 atm 1.962% MCH best fit to all 50 atm scaled data Pitz et al. mechanism, 50 atm

Igni

tion

Tim

e [μ

s]1000/T [1/K]

0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5102

103

104

105

Cyclohexane, φ = 1.0, 12.5 atm current 15 atm data scaled to 12.5 atm current 50 atm data scaled to 12.5 atm Lemaire et al., 11-14 atm Sirjean et al. mechanism, 12.5 atm Buda et al. mechanism, 12.5 atm Silke et al. mechanism, 12.5 atm

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

• Cyclohexane and methylcyclohexane– Agreement with previous shock tube and RCM data– Mechanisms overpredict measured ignition times

• No previous experiments at high pressure to compare with cyclopentane, ethylcyclohexane, and decalin data

Cycloalkane reactivity:cyclopentane vs cyclohexane

ethylene

2 ethylene

allyl

vinyl

opening a strained ring

opening an unstrained ring

+

+ + +

+X

+M

+X

+M

+M

+M

Primary reaction pathways above 900-1000 K

• Cyclopentane > cyclohexane– Higher activation energy for ring opening

(based on QC calculation)– Lower reactivity of β-scission products

C2H4 + aC3H5 vs 2C2H4 + C2H3

X: H,O,OH,HO2M: third body

Cycloalkane reactivityPrimary productsof ring opening

CH3+

+ +

CH3++

CH3++

H ++

CH3+ + +

+ +

CH3+

+ +

CH3++

CH3++

H ++

CH3+ + +

+ +

• Methylcyclohexane >> cyclohexane– C3H5 and C4H6 vs C2H4 and C2H3

• Ethylcyclohexane more reactive than methylcyclohexane due to secondary H-atoms

• Decalin leads to less reactive aromatics (benzene, toluene, xylene, styrene, ethylbenzene), Violi and co-workers (UMich)– Least reactive compound studied

+X

+X

+X

+X

+X

+M

+M

+M

+M

+M

+M

+M

vs

Aromatics

0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10101

102

103

hydrocabon/air, φ = 1.0, 40 atm toluene p-xylene m-xylene o-xylene ethylbenzene

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

• Measurements have been made for five aromatics at ~12 atm and ~40 atm and for Φ = 0.5 and 1.0 (20 data sets)

• Ignition times vary in the order:Toluene ≈ p-xylene >≈ m-xylene > o-xylene > ethylbenzene

LLNL toluene modeling

• The LLNL group has revisited their toluene mechanism• Updated mechanism provides fairly accurate predictions of RPI shock tube

and Case Western RCM data when heat loss is accounted for in RCMsimulations

0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05101

102

103

104

105

106

Current study50 atm

Igni

tion

Tim

e [μ

s]

1000/T [1/K]

Current study12 atm

Mittal and Sung, RCM44 bar, 0.962% toluene

LLNL predictions

Comparison with previous o-xylene and ethylbenzene RCM measurements

• Extrapolation of RPI data to lower temperatures shows intersection with Roubaud et al. (Lille) RCM data

• o-xylene RCM data shows NTC, ethylbenzene does not• No previous data for m- and p-xylene for comparison

– Lille group observed no ignition in RCM for T < 900 K

0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6

102

103

104

105

o-xylene, φ = 1.0 current data, scaled to 16 atm Roubaud et al. (2000) RCM data, 13.8-18.8 atm best fit to both data sets

Igni

tion

Tim

e [μ

s]

1000/T [1/K]0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20

101

102

103

104

105

106

ethylbenzene, φ = 1.0 current data, scaled to 15 atm Roubaud et al. (2000) fit to RCM data, 12.5-17.3 atm best fit to both data sets

Igni

tion

Tim

e [μ

s]1000/T [1/K]

Aromatic kinetics: toluene vs ethylbenzene

Toluene >> ethylbenzene• Weaker C-C bond in ethylbenzene side

chain• Secondary H-atoms in ethylbenzene

side chain• Ethylbenzene is consumed more rapidly

• Toluene exclusively yields stable benzyl• Ethylbenzene can yield stable

C6H5CHCH3 and more reactive C6H5CH2CH2 radicals

+ CH3+ CH3

+X

+M

+M

+X

+X

104 kcal C-C

78 kcal C-C

Aromatic kinetics: xylenes

p-xylene >≈ m-xylene > o-xylene– o-xylene is slightly more

reactive due to possible RO2 isomerization due to the proximity of the methyl side chains

– o-xylene is the only xylene to exhibit NTC behavior

OO

OO

OO

OO

OO

OO

H

H

HHH

OO

OO

OO

OO

OO

OO

H

H

HHH

+X

+X

+X

+O2

+O2

+O2

113 kcal C-H

113 kcal C-H

90 kcal C-H

Summary• A new shock tube has been constructed for the kinetic study

of low-vapor pressure fuels and fuel components at elevated pressures

• The ignition of several larger hydrocarbons of relevance to jet fuel have been investigated– Many new targets for kinetic modeling efforts are available– Diluent gas has a 15-20% influence on ignition for larger hydrocarbons– Larger n-alkanes display ignition times essentially independent of chain length

for mixtures with common carbon content– The influence of structure on reactivity has been quantified for cycloalkanes

and aromatics

• Near term investigations:– Ignition of the diesel cetane reference fuels– Ignition of jet fuel surrogates and jet fuels (JP-8 and Jet A)

• Longer term investigations:– New kinetic targets including CO concentration measurements– Soot formation studies

Publications• H.-P. S. Shen and M.A. Oehlschlaeger, “The Ignition of C8H10

Aromatics at Elevated Pressures,” submitted.

• H.-P. S. Shen, J. Vanderover, and M. A. Oehlschlaeger, “A Shock Tube Study of Iso-Octane Ignition at Elevated Pressures: the Influence of Diluent Gases,” Combustion and Flame, to appear.

• H.-P. S. Shen, J. Vanderover, and M. A. Oehlschlaeger, “A Shock Tube Study of the Auto-Ignition of Toluene/Air Mixtures at High Pressures,” Proceedings of the Combustion Institute 32, to appear.

• J. Vanderover and M.A. Oehlschlaeger, “Ignition Time Measurements for Methylcyclohexane- and Ethylcyclohexane-Air Mixtures at Elevated Pressures,” International Journal of Chemical Kinetics, to appear.

• S.M. Daley, A.M. Berkowitz and M. A. Oehlschlaeger, “A Shock Tube Study of Cyclopentane and Cyclohexane Ignition at Elevated Pressures,” International Journal of Chemical Kinetics 40 (2008) 624-634.

Documents

A Shock Tube Study of the Auto-Ignition of Toluene/Air Mixtures at