research 1..12 - American Chemical Society

Jan 11, 2018 - the authors, knowledge, the only experimental data available are found in the work of Vukadinovic et al.38 In ..... OH* chemiluminescen...
0 downloads 5 Views 805KB Size
Subscriber access provided by READING UNIV

Article

Experimental investigation of laminar flame speed measurement for kerosene fuels: Jet A-1, surrogate fuel and its pure components Yi Wu, Vincent Modica, Xilong Yu, and Frederic Grisch Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b02731 • Publication Date (Web): 11 Jan 2018 Downloaded from http://pubs.acs.org on January 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

Experimental investigation of laminar flame speed measurement for

2

kerosene fuels: Jet A-1, surrogate fuel and its pure components

3

Yi Wu1, 2*, Vincent Modica2, Xilong Yu1, Frédéric Grisch2

4 5 6

1

State Key Laboratory of High Temperature Gas Dynamics, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, China

7 8 9 10

2

CORIA-UMR 6614, Normandie Université, CNRS, INSA et Université de Rouen Campus Universitaire du Madrillet, 76800 Saint-Etienne du Rouvray, France

11 12 13 14 15 16 17

*Corresponding author: Dr. Yi WU Email: [email protected]

18 19 20 21 22

State Key Laboratory of High Temperature Gas Dynamics, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, China Phone: +86 132 6018 7692 Fax: +86 10 6256 1284

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

CORIA-UMR 6614, Normandie Université, CNRS, INSA et Université de Rouen Campus Universitaire du Madrillet, 76800 Saint-Etienne du Rouvray, France Phone: (+33) 6 51 26 39 65 Fax: (+33) 6 51 26 39 65

Correspondence to: [email protected]

ACS Paragon Plus Environment

1

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

38

Experimental investigation of laminar flame speed measurement for

39

kerosene fuels: Jet A-1, surrogate fuel and its pure components

40

Yi Wu1, 2*, Vincent Modica1, Xilong Yu1, Frédéric Grisch1

41 42 43

1

State Key Laboratory of High Temperature Gas Dynamics, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, China

44 45 46 47

2

CORIA-UMR 6614, Normandie Université, CNRS, INSA et Université de Rouen Campus Universitaire du Madrillet, 76800 Saint-Etienne du Rouvray, France

48 49 50

Abstract:

51

The present work investigated the laminar flame speed measurement of kerosene-relevant fuel, including Jet A-1

52

commercial kerosene, and surrogate kerosene fuel and its pure components (n-decane, n-propyl-benzene and

53

propyl-cyclohexane), using a high-pressure Bunsen flame burner. The OH* chemiluminescence technique and

54

the kerosene-PLIF technique were used for flame contours detection in order to calculate the laminar flame

55

speed. The experiments were firstly conducted for n-decane/air flame at T = 400 K, φ = 0.6 − 1.3 and

56

atmospheric pressure conditions in order to validate the whole experimental system and measurement

57

methodology. The laminar flame speed of Jet A-1/air, surrogate/air and pure kerosene component (n-decane,

58

n-propyl-benzene and propyl-cyclohexane) was then measured under large operating conditions, including

59

temperature T = 400-473 K, pressure P = 0.1-1.0 MPa and equivalence ratio φ = 0.7 − 1.3. It was found that

60

these three pure components of kerosene have very similar laminar flame speed. By comparing the experimental

61

results of surrogate kerosene and Jet A-1 commercial kerosene, it was observed that the proposed surrogate

62

kerosene, i.e., mixtures of 76.7wt%

63

can appropriately reproduce the flame speed property of Jet A-1 commercial kerosene fuel. The experimental

64

results were further compared with simulation results using a skeletal kerosene mechanism.

n-decane, 13.2wt% n-propyl-benzene and 10.1wt% propyl-cyclohexane,

65 66

Keywords: laminar flame speed, kerosene, Jet A-1, surrogate fuel, n-decane, n-propyl-benzene and

67

propyl-cyclohexane

2

ACS Paragon Plus Environment

Page 2 of 29

Page 3 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

68

1. Introduction

69

Understanding the fundamental combustion properties of commercial jet fuels is essential in the development of

70

low-cost and efficient aircraft engines. For this purpose, as well as improve the performance of aeronautical

71

propulsion systems, a comprehensive understanding of the chemical kinetic mechanisms of multicomponent

72

commercial jet fuels is needed [1]. However, the detailed chemical mechanisms of commercial jet fuels, such as

73

Jet A-1, consist of several hundreds of different species and about 1,000 elementary reactions. To perform

74

mechanical CFD (Computational Fluid Dynamics) simulations using detailed kinetic mechanisms, it needs to get

75

solutions of all transport equations for each species present in the detailed kinetic mechanism and calculate all

76

the source terms for each species. This kind of simulation of an industrial complex geometry using

77

aforementioned detailed kinetic mechanism remains elusive [2]. In practical investigations, an alternative

78

approach is using pure hydrocarbon fuels or composition-simplified surrogate fuels to simulate the

79

physicochemical characteristics of a practical kerosene fuels [3-8]. One of the evaluating indicator for the

80

validation of surrogate kerosene is the similarity between the combustion properties and those of

81

multicomponent practical fossil fuels [9]. Therefore, it is firstly of substantial interest to learn about the surrogate

82

fuel combustion characteristics and associated chemical mechanism in order to more effectively exploit

83

commercial jet fuels. Several methodologies have been used to quantify fundamental properties of surrogate jet

84

fuels, such as ignition [10-15], laminar flame speed [7, 16-22], flame stretch and extinction limit [11, 23-27].

85

Among these parameters, laminar flame speed is key parameter to understand the reactivity, diffusivity and

86

exothermicity of fuel/air mixtures and validating both kinetic chemical mechanisms and turbulent models [28,

87

29].

88

The aforementioned necessities have motivated numerous previous works, which have investigated the laminar

89

flame speed of composition-simplified surrogate kerosene fuel and their pure components [10, 19, 30-41].

90

Typically, commercial kerosene, such as Jet A-1, involves around 50 - 65% paraffin, 10 - 20% aromatics and 20

91

– 30 % naphthenic, which cause the elucidation of the main components of previously proposed surrogate

92

kerosene [42]. In the early stages of a kerosene surrogate investigation, the composition of the surrogates

93

generally comprised more than 10 kinds of hydrocarbons fuels [30, 35]. Investigations were then performed to

94

reducing the number of compositions to less than 10, because it was commonplace to compose the surrogates in

95

using some compositions in practical fuels, including low-concentration components, such as naphthalene and

96

several hydrocarbons of the same category [36]. More recently, the composition of surrogate fuels has

97

continually been simplified to culminate in one-, two- or three-component surrogate fuels [10,30-34]. A review 3

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

98

of literature reveals that, the Aachen surrogate [5] and the UCSD surrogate [34] can well reproduce the property

99

of auto ignition and extinction rate of commercial kerosene fuels. Laminar flame speeds measurement of the

100

Aachen surrogate fuel have been performed with the spherically expanding flame method under conditions of P

101

= 0.1 MPa and T = 473 K. Holley et al. [25] conducted measurements of the extinction strain rate and ignition

102

temperature of the surrogate fuels with a counter flow burner under atmospheric pressure and an elevated

103

preheating temperature conditions. In their work, they also compared single-component hydrocarbon with

104

surrogate jet fuel of JP-8. More recently, Comandini et al. [7] measured laminar flame speeds of surrogate fuel

105

composed by n-decane, n-butyl-benzene and n-propyl-benzene using the spherically expanding flame method

106

under conditions of T = 403 K and P = 0.1 MPa. Laminar flame speeds of surrogate mixtures were compared

107

with each pure component and discussed in detail [7]. Besides the aforementioned Aachen surrogate [5] and the

108

surrogate referred in the work of Comandini et al. [7], few investigations of laminar flame speeds measurements

109

for other surrogate fuels can be found in literature. Comparisons of laminar flame speeds between surrogate fuel

110

and commercial kerosene fuels have been rarely performed. In the present work, surrogate kerosene consisting of

111

n-decane (76.7wt%), n-propyl-benzene (13.2wt%) and propyl-cyclohexane (10.1wt%) are investigated for their

112

laminar flame speeds using a high-pressure Bunsen flame burner [43, 44]. This composition of surrogate for

113

kerosene was initially referred in the work of Dagaut et al. [8] and Luche et al. [6], who developed skeletal

114

kinetic mechanism targeting to emulate the combustion properties of commercial kerosene fuels. However, to

115

the best of our knowledge, no measurements of laminar flame speeds were performed and compared with

116

commercial kerosene fuels for this three-component surrogate kerosene. In the present work, in order to evaluate

117

the performance of this surrogate fuel, its laminar flame speed will be measured under large range of conditions,

118

including variations of temperature, pressure and equivalence ratio, and compared with the laminar flame speed

119

of commercial kerosene fuel (Jet A-1).

120

In order to compare the flame speed of the surrogate kerosene proposed in the present work with commercial

121

kerosene, laminar flame speeds measurements of Jet A-1 kerosene fuel/air mixtures under the same conditions as

122

for the aforementioned surrogate kerosene have been performed. A review of the literature reveals that, even

123

though commercial kerosene (such as Jet A-1) fuel has been used for decades in aeronautics, it has only been in

124

recent years that its laminar flame speeds have been measured with reasonable accuracy, owing to the

125

development of the measurement devices used [18, 19, 37-41]. For instance, Kumar et al. [37] measured laminar

126

flame speeds of Jet A-1 using the counter flow flame method under atmospheric pressure conditions and

127

temperatures of 400, 450 and 470 K. Hui et al. [19] also measured the laminar flame speed of Jet A-1 using the 4

ACS Paragon Plus Environment

Page 4 of 29

Page 5 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

128

counter-flow method, in their work efforts were focused on measurements in higher pressure conditions of P =

129

0.1 - 0.3 MPa, with preheated temperatures between 350 and 470 K and equivalence ratio from 0.7 to 1.3. In

130

their work, the un-stretched laminar flame speed was determined by the linear extrapolation of the stretched

131

laminar flame speeds. While both of the measurements [19, 37] were performed using the same counterflow

132

flame burner, large data deviation was observed. Other laminar flame speed measurements of Jet A-1 were

133

depicted in the investigation of Chong et al. [18]. The PIV optical technique has been applied on a jet wall

134

stagnation flame to determine the laminar flame speed of Jet A-1. Regarding the laminar flame speed of Jet A-1

135

measured under pressure conditions higher than 0.5 MPa, to the best of the authors’ knowledge, the only

136

experimental data available are found in the work of Vukadinovic et al. [38]. In their study, the laminar flame

137

speeds were measured at a pressure of up to 0.8 MPa using the spherically expanding flame method. Generally,

138

as a summary of the above literature analysis, large data scattering (up to 15 cm/s) are observable between the

139

different results of laminar flame speeds for Jet A-1 fuel [19, 37, 39-41]. Even though part of the discrepancies

140

are due to the fact that the composition of Jet A-1 varies among different production regions, further

141

investigation is still necessary to restrain the uncertainties introduced by methodologies, such as linear or

142

nonlinear correlations, and by measurements devices. Moreover, measurements under high pressure and elevated

143

temperature conditions in the propulsion system are still deficient, such that further investigations into accurate

144

laminar flame speed measurements of kerosene fuels are still necessary.

145

In the present work, laminar flame speeds of kerosene fuels, including Jet A-1 commercial kerosene, as well as

146

Luche surrogate fuel and its pure components, i.e., n-decane, n-propyl-benzene and propyl-cyclohexane, are

147

measured under large operating conditions, including variations in temperature, pressure and equivalence ratio

148

based on a high-pressure Bunsen flame burner. The OH* chemiluminescence optical technique and the

149

kerosene-PLIF technique were used for flame contours detection in order to calculate the laminar flame speed.

150

OH* chemiluminescence is a convenient and widely applied combustion diagnostic technique; however, for

151

Bunsen flames with large molecular weight fuels, flame tip opening phenomenon occurs under slight fuel-rich

152

conditions (around φ > 1.15~1.2). Under these conditions, OH* chemiluminescence has difficulty in obtaining

153

the whole flame contours due to local quenching at the cusp of the flame [45-48]. Kerosene-PLIF has the

154

advantage of visualizing the contours of the unburned fronts of kerosene/mixtures, such as Jet A-1, which is free

155

from the influence of local quenching located in the burned gas [49, 50]. Hence, in the present work,

156

kerosene-PLIF is applied to validate the suitable flame contours definition for the OH* chemiluminescence

157

technique when the flame-opening phenomenon conditions are present. 5

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

158

Experimental results of n-decane/air mixtures using the OH* chemiluminescence technique will first be

159

presented to validate the accuracy of the whole experimental system and the determination methodology. For

160

further investigating the effect of the flame-opening phenomenon on the laminar flame speed measured by OH*

161

chemiluminescence, comparison of the laminar flame speeds obtained by OH* chemiluminescence and the

162

kerosene-PLIF technique for Jet A-1/air mixtures under the same operating conditions are performed.

163

Measurements are then extended to higher temperatures T = 400, 423 and 473 K for pure surrogate components,

164

i.e., n-decane, n-propyl-benzene and propyl-cyclohexane under atmospheric pressure conditions. Finally, the

165

laminar flame speed of surrogate kerosene, i.e., mixtures of n-decane (76.7 wt %), n-propyl-benzene (13.20 wt

166

%) and propyl-cyclohexane (10.1 wt %), are measured under conditions of T = 400-473 K, P = 0.1-1.0 MPa

167

and φ = 0.7 − 1.3, then compared with the experimental results of commercial kerosene (Jet A-1) measured

168

under identical conditions. The laminar flame speed of surrogate fuel is also simulated with skeletal kerosene

169

mechanism. Further analysis and discussions will be detailed.

170

2. Experimental methodology

171

2.1 High pressure burner and gas feeds

172

In the present study, a high pressure Bunsen flame burner was developed to measure laminar flame speed of

173

kerosene fuels. The detailed description of the experimental apparatus can be found in Wu et al.’s previously

174

published work [43, 44]; as such, only a brief introduction is given here.

175 176

Figure 1: Schematic of the experimental setup 6

ACS Paragon Plus Environment

Page 6 of 29

Page 7 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

177

As illustrated in Figure 1, the whole experimental apparatus consists of a high-pressure Bunsen flame burner,

178

liquid fuel vaporization with a gas feed system and an optical diagnostic set-up. With this system, a steady

179

conical flame can be generated on the outlet of the contoured nozzle that has an outlet diameter of 7 mm. A

180

concentric contoured nozzle with an outlet diameter of d2 = 7.6 mm, which surrounds the central nozzle, is used

181

to produce a pilot flame in order to stabilize the flame under high-pressure operating conditions. The equivalence

182

ratio and the flow rate of the pilot flame are regulated by two mass flow controllers. The whole burner is placed

183

in an N2 (guard flow) ventilated high-pressure chamber. A liquid flow controller (Bronkhorst mini-CORIFLOW)

184

is used to deliver liquid kerosene fuel to the Controlled Evaporator and Mixer (CEM, Bronkhorst), where the

185

liquid fuel and N2 carrier gas are mixed and heated to a preset temperature. The CEM system allows to vaporize

186

liquid kerosene fuel and two additional flow controller were used to deliver nitrogen and oxygen to mix at the

187

outlet of CEM to reproduce the synthetic species’ composition of air and control the equivalence ratio of

188

vaporized fuel/air mixtures. The whole chamber and all the gas feed lines are preheated with electrical wire

189

heaters. Five thermocouples located in different positions of the high pressure chamber are used to measure and

190

monitor the uniformity of the temperature throughout of the pressure chamber. The pressure in the chamber is

191

measured with a piezoelectric transducer located at the outlet of the high pressure chamber. Two circulation

192

heaters are used to heat the guard flow and the vaporized fuel/air mixture to avoid condensations of the fuel

193

vapor. There are four UV quartz windows at the high pressure chamber allowing probing the flame with optical

194

diagnostics.

195 196

2.2 Optical diagnostic setup 2.2.1 OH* chemiluminescence

197

One of the optical techniques used in the present work to detect the flame contour is the OH*

198

chemiluminescence imaging technique. A thermoelectrically cooled, 16-bit intensified CCD camera (PI-MAX 3,

199

Roper Scientific) was used to record the OH* chemiluminescence flame image. With an achromatic UV lens

200

(CERCO)of f/2.8, f = 100 mm, and an interference band pass filter centered at 310 nm, the camera can visualize

201

a 40 × 40 mm2 area of the flame with a 1024×1024 array so that the spatial resolution is about 40 µm per pixel.

202

The exposure time to record the OH* chemiluminescence image is controlled by opening the intensifier gate 1

203

µs. The acquisition repetition rate of the camera is set to 10 Hz.

204

2.2.2 Kerosene-PLIF

205

It is well known that the composition of kerosene commonly used in civil engines (i.e. Jet A1) consists of a wide

206

range of hydrocarbon molecules. Among these, aromatic molecules are known to have UV and visible transitions 7

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

207

exhibiting strong fluorescence. Therefore, kerosene fuel benefits from its broadband fluorescence emission

208

between the 260-420 nm spectral domains [49, 50]. This fluorescence emission results from the excitation of

209

aromatics (i.e., mono- and di-aromatics) allows visualizing the fresh gas region of the flame. Although aromatics

210

may be excited on a wide range of wavelengths, for a flame to visualize only the unburned gas, the excitation

211

wavelength of kerosene must be selected carefully to avoid overlaps with the resonance of an OH radical

212

transition located in the burned gas. An observation of the fluorescence spectrum of the OH transition reveals

213

that the Q1(5) and Q1(6) rotational lines are well separated, thus presenting the possibility to tune the excitation

214

wavelength of aromatics to 282.85 nm in a spectral region in which only the aromatics fluorescence will be

215

collected.

216 217

Figure 2: Jet A-1/air flames under the condition of T = 400 K, P = 0.1 MPa for equivalence ratios φ = 0.8, 1.0

218

and 1.3: (a) OH* chemiluminescence technique; (b) Kerosene-PLIF technique

219

In the present work, the kerosene planar laser-induced fluorescence system consists of a cluster system, which

220

includes an Nd:YAG laser (Quanta-ray Pro-190), a dye laser (Sirah PRSC-G-300) and a high-resolution ICCD

221

camera (PI-MAX 3, Roper Scientific). A frequency-doubled, Q-switched Nd:YAG laser was used to pump a dye

222

laser, which was then frequency-doubled to obtain wavelengths in the 280-290 nm spectral range. The dye laser

223

was tuned to 282.85 nm to excite the aromatics in the Jet A-1 fuel. The laser energy was regulated to 5 mJ in

224

order to keep the fluorescence of kerosene within the linear regime, thus maintaining the proportionality between

225

the intensity of fluorescence and the concentration of kerosene fuel. The UV laser beam at the exit of the laser

226

source was transported via optical mirrors. It was then transformed into one 50 mm collimated sheet using a set

227

of converging and divergent cylindrical lenses and a spherical lens. The two cylindrical lenses formed a

228

telescope, which sped up the beam into a collimated sheet to cross the target flame. The kerosene fluorescence

229

image was then recorded by the ICCD camera (the same one used for the OH* chemiluminescence technique).

8

ACS Paragon Plus Environment

Page 8 of 29

Page 9 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

230

The intensifier gate width was set to 1 µs and the framing rate of the acquisition of fluorescence images was 10

231

Hz. The camera was equipped with the same optical lens with OH* chemiluminescence, while only the optical

232

filter was changed to 300-550 nm. As shown in Figure 2, images of Jet A-1/air flames using OH*

233

chemiluminescence technique and kerosene-PLIF technique are presented.

234

2.3 Fuel properties

235

The chemicals used in the different mixtures were n-decane (C10H22, Sigma–Aldrich, 99+ %), n-propyl-benzene

236

(C9H12, Aldrich, 99+ %), and n-propyl-cyclohexane (C9H18, Aldrich, 99+ %). The Luche surrogate fuel for

237

emulating the commercial jet fuel is a mixture of these three components by n-decane (76.7wt %),

238

n-propyl-benzene (13.2wt %) and propyl-cyclohexane (10.1wt %). The average chemical formula is C9.74H20.05

239

with a molar mass equal to 136.93 g/mol. Meanwhile, the target commercial kerosene fuel used in the present

240

work was analyzed by gas chromatography and its chemical composition is shown in Table 1. The chemical

241

composition of this fuel consists mainly of paraffin, naphthenic and alkyl-benzenes. In addition, there is no

242

hydrocarbon poly-aromatic hydrocarbon heavier than naphthalene. The average chemical formulation of Jet A-1

243

is C11.16H20.82 with a molar mass equals to 154 g/mol. Table 2 summarizes the properties of studied fuels in the

244

present work.

245

Species

Proportion (wt%)

Paraffin

46.5

Non-condensed naphthene

22.3

Condensed naphthene

13.2

Alkylbenzenes

13.2

Indanes and tetralines

2.8

Naphthalenes

1.8

Acenaphthenes and diphenyls

0.1

Benzothiophenes

0.1

Saturated

82

Monoaromatic

16

Diaromatic

1.9

Sulfur

0.1

Table 1: Chemical composition of Jet A-1 in the present study

246 247 248 9

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

249

Page 10 of 29

Fuel

Formula

Boiling point (K)

Density (kg/m3)

n-decane

C10H22

447.3

724.0

n-propyl benzene

C9H12

432.4

825.7

propyl cyclohexane

C9H18

429.9

793.0

Luche surrogate

C9.74H20.05

432~447

747.0

Jet A-1

C11.16H20.82

423~533

815.9

Table 2: Fuel property of n-decane, n-propylene, propyl cyclohexane, Luche surrogate and Jet A-1 fuel.

250 251

2.4 Extraction of flame speed

252

The Bunsen flame method is a convenient and widely used approach for laminar flame speed measurements

253

because of its advantages of simplicity and a well-defined flame structure. Laminar flame speed is defined as the

254

velocity at which a planar flame front travels relative to the unburned gas in the direction of the flame surface

255

[51]. For a conical flame, with the hypothesis that the flame speed is identical all over the entire surface area of

256

the flame, the laminar flame speed can be calculated, based on the mass conservation between the outlet nozzle

257

and the flame front [44, 52-56]. As shown in Figure 3, the average flame speed in the transverse plane is

258

expressed by:

259

   =  →  =  /( )

Eq. 1

260

Where  is the total mass flow rate of the fuel/air gaseous mixture, and ρu is the density of fresh gas. This

261

method requires the knowledge of the total area of the flame surface A, which is determined in the present

262

investigation by analyzing the OH* chemiluminescence image and the kerosene-PLIF image (for Jet A-1 fuel) of

263

the flames. Kerosene-PLIF is applied to validate the OH* chemiluminescence image-processing algorithm for

264

the cases when the flame tip-opening phenomenon is present. For large molecule weight fuel, such as kerosene,

265

the flame tip-opening phenomenon appears when the equivalence ratio is higher than around φ > (1.15~1.2).

266

The tip opening of the laminar Bunsen flame involves local combustion extinction, which occurs at the top part

267

of the flame where a strong negative flame stretch presents [45,46]. When the OH* chemiluminescence

268

technique is applied, as illustrated in Figure 2 (for the case φ = 1.3), the tip opening phenomenon greatly

269

increases the uncertainties of flame contours determination of the top part of the flame. Fortunately,

270

Kerosene-PLIF, as applied in the current work, has the advantage of directly imaging the unburned fuel/air

271

mixture contour, which is free from the influence of the tip-opening phenomenon. The image-processing

272

procedures are detailed in the following sections.

273

10

ACS Paragon Plus Environment

Page 11 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

274

Energy & Fuels

2.4.1 OH* chemiluminescence

275

In order to obtain the two-dimensional flame front of the reaction zone of the OH* chemiluminescence, the

276

recorded image is transformed by Abel inversion using a MATLAB program [57]. By taking the inside flame

277

front of image after Abel inversion, the surface A can be calculated. The inside boundary is used because its

278

laminar flame speed is defined by the velocity relative to the unburned premixed gas in the direction of the flame

279

surface. As the inside boundary is more attracted to fresh gas, it is more appropriate to use it as the flame

280

boundary in order to calculate flame speed.

281

A specific high-frequency filtering function was used to smooth the resulting curve plotted along the red line in

282

Figure 3 and obtain the final flame front. For equivalence ratio fuel-lean conditions when the flame tip-opening

283

phenomenon has not occurred, the flame contours can be accurately detected using the above-described

284

image-processing method. However, for cases where the flame-opening phenomenon is present, due to the local

285

quenching of the OH* signal, the flame contours are difficult to detect at the flame tip part. In such cases, the

286

flame area that is used for flame speed determination is obtained by extending the linear peripheral contour to

287

the tip of the flame with the application of a high-order interpolation method, as illustrated in Figure 3. The

288

accuracy of the OH* chemiluminescence image-processing method will be further evaluated by the

289

kerosene-PLIF technique (Part 3.1).

290 291 292 293 294

Figure 3: Determination of the flame surface area: (a) initial OH* chemiluminescence image and (b) left part of the Abel-inverted image (the red line is the initial flame contours determined by taking the inside front of the Abel-inverted image, while the dashed line is the flame contours used in flame speed calculations made following the smoothing function)

295

After obtaining the flame contours, flame area A can be calculated by pivoting flame front profile  () along the

296

burner axis using the following equation:

297



 = 2   () 1 ! | # ()| $ %

Eq. 2

298

Where & and ' are the boundary limits of the flame contours, and  () is the flame contour profile function

299

obtained by image processing. Assuming the axisymmetric condition of the flame, the final flame speed is the 11

ACS Paragon Plus Environment

Energy & Fuels

300

averaged value of the flame surface area, obtained from each half of the recorded images. It should be noted that

301

the flame speed measured using the aforementioned equation is an averaged value of the entire flame surface,

302

which is different from the unstretched flame speed, as this conical flame is affected by its strain and curvature

303

effects. To avoid the effect of the flame, the stretch and the flame tip-opening phenomenon, for each case, the

304

measured flame was controlled in relation to the maximum height allowed by the current experimental

305

configuration in order to alleviate the tip of the flame. Fortunately, the difference between both values was found

306

to tiny and negligible, which is consistent with the results in [44, 52, 56].

307

2.4.2 Kerosene-PLIF

308

Compared to the processing procedures of OH* chemiluminescence imaging, the data processing of

309

kerosene-PLIF is largely simplified. The outer edge of the fresh gases is defined as the position at which the

310

fluorescence of the aromatics molecules disappears (Figure 4). This location corresponds to the frontier

311

delimiting a chemical transformation of these organic molecules through the action of chemical reactions.

312

Typically, these organic molecules disappear before the OH radical starts to be optically detected. Once the

313

flame front is obtained, the laminar flame speed is determined using the same calculation (Eq. 2) as the OH*

314

chemiluminescence technique. 18000 16000 14000 12000 10000

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

8000 6000

B'

A'

4000 2000 0 0

200

400

A

600

800

1000

B

315 316 317

Figure 4: Illustration of the image processing of kerosene-PLIF techniques

2.5 Measurements uncertainties

318

For all the experimental results presented in this work, 30 single-shot images are systematically recorded and the

319

resulting laminar flame speed is determined by the data processing of averaged images, as determined from the 12

ACS Paragon Plus Environment

Page 13 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

320

set of single-shot images. The measurement uncertainties are mainly associated with the gas/liquid flow delivery

321

system (U)* ) and the uncertainty of the calculated flame area image (U)* ), as determined by the camera

322

resolution. The uncertainty of the total flow rate comes from the mass flow controller uncertainty (0.5% of

323

reading + 0.1% full scale), which is estimated to be ~2%, while uncertainty derives from the camera’s spatial

324

resolution, which is estimated to be ~3%. The overall uncertainty is calculated to be within ~4% (from the

325

$ relation+,- ! ,.$ ) for all the laminar flame speeds recorded under atmospheric pressure conditions. For

326

high-pressure measurements P > 0.5 MPa, the measurement uncertainty can be amplified due to flame stability.

327

Indeed, the fluctuation of the position of the flame displays an artificial thickening of the flame front during the

328

time integration of the signal on the camera. This effect modifies the position of the flame contours (2~3 pixels),

329

giving a maximum uncertainty of ~ 7% in the worst situation.

330

3. Experimental results and discussion

331

Experimental results of n-decane/air mixtures using the OH* chemiluminescence technique will first be

332

presented in order to validate the accuracy of the aforementioned image-processing methodology and the whole

333

experimental system. To further investigate the effect of the flame-opening phenomenon on the laminar flame

334

speed measured by OH* chemiluminescence, a comparison of the laminar flame speeds obtained by OH*

335

chemiluminescence and the kerosene-PLIF technique for Jet A-1/air mixtures under the same operating

336

conditions is performed. The laminar flame speed of pure surrogate components (n-propyl-benzene and

337

propyl-cyclohexane) is then measured under atmospheric pressure conditions at different temperatures, that is,

338

400, 423 and 473 K. The laminar flame speed of the kerosene surrogate (mixtures of n-decane, n-propyl-benzene

339

and propyl-cyclohexane) and the Jet A-1/air mixture are then subsequently measured at T = 400, 423 and 473 K.

340

High-pressure measurements from 0.1-1.0 MPa are performed with an equivalence ratio ranging from 0.7 to 0.8.

341

For high-pressure measurements, the preheated temperature is fixed at T = 423 K, with only the equivalence

342

ratio for the fuel-lean side performed. These temperature and equivalence ratio conditions are selected in order to

343

assure a stabilization of the flame at the nozzle outlet during the experiments and to prevent fuel pyrolysis for

344

temperatures higher than 423 K, especially under elevated pressure conditions. This is essential to guarantee

345

measurement accuracy and make sure that no chemical reaction modifies the fuel composition during the

346

transport of the fuel/air mixture inside the combustion chamber. Experiments are then compared with data found

347

in the literature and the simulation results using the Luche skeletal kerosene mechanism [6].

13

ACS Paragon Plus Environment

Energy & Fuels

348

3.1 OH* chemiluminescence methodology validation

349

Preliminary measurements using OH* chemiluminescence are performed for n-decane/air mixtures under

350

condition φ = 0.7 − 1.3, T = 400 K and atmospheric pressure conditions to validate the whole experimental

351

setup and measurement methodology. Laminar flame speeds calculated in the present work are compared with

352

literature found experimental results. A review of the literature reveals that n-decane is one of the main

353

components in kerosene fuels, while its laminar flame speed has been extensively explored. Figure 5 illustrates

354

the comparison between the experimental results in the present work and the experimental data found from the

355

literature. The simulated flame speeds using detailed kinetic mechanism JetSurF 2.0 are plotted at the same

356

figure as well. JetSurf 2.0 is a detailed mechanism for kerosene fuels, which consists of 348 species and 2,162

357

reactions, as established by Wang et al. [58]. Except for the results presented by Hui et al. [19] and those

358

reported by Kumar et al. [37], which shows significant difference from the other measurements, there is a

359

general agreement between our laminar flame speeds and those in the literature, albeit with a better agreement

360

between our data and the JetSurF 2.0 simulation results. Our experimental results show good agreement with the

361

laminar flame speed measured using spherical flames as well [7, 59, 60]. However, a tiny difference between our

362

experimental results and the numerical predictions is observed at condition of equivalence ratio φ = 1.3. 70 60 50

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

40 30

Jet surf 2.0 Hui et al. 2013 Singh et al. 2011 Kumar et al. 2007 Comandini et al. 2015 Kim et al. 2015 Munzar et al. 2013 Present work

20 10 0 0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Equivalence ratio

363 364 365 366

Figure 5: Comparison between our experimental results and the data from the literature of n-decane/air

367

Figure 6 shows experimental results for higher temperature conditions of T = 423 K and 473 K. As expected,

368

the laminar flame speed increased with the preheating temperature of the mixture. The conditions where the

369

flame tip-opening phenomenon was present are also labeled in the figure. In order to further clarify the influence

370

of the flame tip-opening phenomenon on the determination of the flame speed using the OH*

371

chemiluminescence

mixtures: φ = 0.6-1.5, T = 400K and P = 0.1 MPa

method,

experiments are

performed with both the

14

ACS Paragon Plus Environment

kerosene-PLIF and OH*

Page 15 of 29

372

chemiluminescence techniques for Jet A-1/air mixtures at T = 400 K, P = 0.1 MPa and φ = 0.6 − 1.3. As

373

depicted in Figure 7, laminar flame speeds derived from the kerosene-PLIF technique show general agreement

374

with those determined with OH* chemiluminescence, with a maximum difference of 3 cm/s on the fuel-lean side.

375

There is a better agreement between the results of the OH* chemiluminescence technique and the kerosene-PLIF

376

technique under approaching stoichiometric conditions. For the fuel-rich side, where the flame tip-opening

377

phenomenon is present, the difference between these two techniques is less than 2 cm/s. This illustrates that the

378

OH* chemiluminescence image-processing method proposed in the present work is able to determine the flame

379

speeds with reasonable accuracy when the flame tip-opening phenomenon is present. The experimental results

380

presented in the following sections were obtained using OH* chemiluminescence technique considering the

381

simplicity of this methodology.

382 90 85

n-decane T = 400 K n-decane T = 423 K n-decane T = 473 K

80 75 70

SL (cm/s)

65 60 55 50 45 Flame tip opening

40 35 30 25 20 0.4

0.5

0.6

0.7

0.8

0.9

383 384 385

1.0

1.1

1.2

1.3

1.4

1.5

ϕ

Figure 6: Laminar flame speed of n-decane /air mixture at T = 400, 423 and 473 K, P = 0.1 MPa and φ = 0.65-1.3 60 55 50

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

45 40

OH* Chemiluminescence Kereosene-PLIF

35 30 0.5

386 387 388 389

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

Equivalence ratio

Figure 7: Comparison between laminar flame speeds determined from OH* chemiluminescence and Kerosene-PLIF of Jet A-1/air mixture, T = 400 K, P = 0.1 MPa, φ = 0.65-1.3

15

ACS Paragon Plus Environment

Energy & Fuels

390

3.2 Laminar flame speed of neat surrogate components 3.2.1 N-propyl-benzene

391 392

N-propyl-benzene is a large molecular weight fuel that can be largely found in kerosene fuels. However, its

393

laminar flame speed has not been widely studied. The only literature results can be found is the measurements

394

recently performed with the stagnation flame [11] and the heat flux methods [61]. Figure 8 plotted the

395

experimental results obtained in the present work and those presented in [11] [61]. It shows that our

396

experimental results are in accordance with those reported in the work of Mehl et al. [61] for all the equivalence

397

ratio conditions investigated. The maximal difference observed from this comparison is around 5 cm/s, which is

398

a value that remains comparable with our measurement uncertainty. Meanwhile, our experimental results are

399

shows lower flame speeds compared with the results obtained by Hui et al. [11]. As shown in Figure 9, in order

400

to investigate the effect of preheating temperature on laminar flame speeds, experiments were performed with

401

higher temperature T = 423 K and 473 K. As expected, the experiments results show that the laminar flame

402

speed increases with higher temperature conditions. 70 65 60 55 50

SL (cm/s)

45 40 35 30 25 20

Hui et al. 2012, T=400 K Mehl et al. 2015, T = 403 K Present work

15 10 5 0 0.4

0.6

0.8

1.0

403 404

1.2

1.4

1.6

ϕ

Figure 8: Laminar flame speed of n-propyl-benzene at T = 400 K, φ = 0.6-1.3 and P = 0.1 MPa [11] [61] propylbenzene T=400K propylbenzene T=423K propylbenzene T=473K

80

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

60

Flame tip opening

40

20 0.4

0.6

0.8

1.0

405 406 407

1.2

1.4

1.6

ϕ

Figure 9: Laminar flame speed of n-propyl-benzene at T = 400 K, 423 K and 473 K, φ = 0.6-1.3 and P = 0.1 MPa 16

ACS Paragon Plus Environment

Page 17 of 29

3.2.2 Propyl-cyclohexane

408

Propyl-cyclohexane is one of the main cycloalkanes that are usually present in kerosene fuels. Different from

410

n-alkanes, branched alkanes or aromatics, laminar flame speeds measurements of propyl-cyclohexane over a

411

large range conditions of preheating temperature and equivalence can be rarely found in literature. In the present

412

work, experimental measurements were firstly performed at T = 400 K, P = 0.1 MPa and φ = 0.65-1.3, then

413

compared with the data found in the literature. As shown in

414

Figure 10, at the equivalence ratio lean side our experimental results are in accordance both with the results

415

presented in the work of Comandini et al. [7] (nonlinear extrapolation at a zero stretch rate) and with the results

416

presented in the work of Dubois et al. [62] (linear extrapolation at a zero stretch rate). However, discrepancy was

417

observed at fuel rich side that our experimental results give higher laminar flame speed compared with literature

418

found results. This discrepancy could come from the influence of the tip-opening phenomenon, when active for

419

these operating conditions of φ = 1.15-1.3. Experiments were also performed for higher temperature conditions

420

of T = 423 K and 473 K, as plotted in Figure 11.

SL (cm/s)

409

90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Dubois et al. 2009, T=403K Ji et al. 2011, T=403K Comandini et al. 2015, T=403K

0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

ϕ

421 422

Figure 10: Laminar flame speed of propyl-cyclohexane at T = 400 K, P = 0.1 MPa, φ = 0.65-1.3 [7] [62] 90 85

T=473K T=423K T=400K

80 75 70 65

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

60 55 50 45

Flame tip opening

40 35 30 25 0.3

0.4

0.5

0.6 0.7 0.8

0.9

423 424 425

1.0

1.1

1.2 1.3 1.4 1.5

1.6

ϕ

Figure 11: Laminar flame speed of propyl-cyclohexane/air mixture at T = 400, 423 and 473 K, P = 0.1 MPa and φ = 0.65-1.3 17

ACS Paragon Plus Environment

Energy & Fuels

426

3.2.3 Comparison between the pure components

427

Figure 12 plots the laminar flame speeds of the three component in Luche surrogate fuel at condition of T = 400

428

K, P = 0.1 MPa and / = 0.65 − 1.3. It shows that for all equivalence ratio conditions investigated in the present

429

work, the three pure components have quite similar laminar flame speed values. However, a more careful

430

observation of the results reveals that n-decane has slightly higher flame speed values (about a maximum of 2.5

431

cm/s at fuel lean side) than n-propyl-benzene. This tiny difference can be explained by the different molecular

432

structure between n-decane and n-propyl-benzene. As discussed in the work of Comandini et al. [7],

433

n-propyl-benzene is an aromatic molecule who has a ring chemical structure which potentially have an influence

434

on the combustion reaction. Indeed, a sensitivity analysis explained that the flame speed is affected by the ring

435

structure, in particular, by the reactions involving radical intermediates, such as phenoxy, benzyl and the phenyl

436

radicals. Moreover, the work of Mehl et al. [61] also mentioned that the increase of benzyl radicals in the

437

pre-heating zone restrains flame propagation and can decrease flame speed. The above analysis indicates that the

438

laminar flame speed of propyl-cyclohexane is higher than that of n-propyl-benzene. 70 65 60 55 50

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

45 40

Propylbenzene Propylcyclohexane n-decane

35 30 25 20 0.5

0.6

0.7

0.8

0.9

1.0

439 440 441

1.1

1.2

1.3

1.4

1.5

1.6

ϕ

Figure 12: Laminar flame speeds of n-decane, n-propyl-benzene, and n-propyl-cyclohexane at T=400 K, P=0.1 MPa and / = 0.65 − 1.3

442 443

As n-decane and propyl-cyclohexane both belong to the alkane category, they can be directly compared. Our

444

experiments results show that laminar flame speed of n-decane (n-alkane) is almost identical but uniformly

445

higher than those of propyl-cyclohexane (mono-alkylated cyclohexane). This observation is with agreement of

446

the work of Ji et al. [63] and Wu et al. [64]. In their study, it shows that event though laminar flame speed of

447

n-decane is influenced by the small intermediates, the decomposition of linear structure of n-alkane (such as

448

n-decane) is faster than the decomposition of the ring structure exist in mono-alkylated cyclohexane (such as

18

ACS Paragon Plus Environment

Page 19 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

449

propyl cyclohexane). Therefore, these experimental results shed light on the fact that laminar flame speeds are

450

influenced by both the small intermediates involved in the combustion reactions and the initial step in the fuel

451

decomposition. [7]. For lean equivalence ratio conditions, above mentioned tendency is well observed in Figure

452

12: n-decane > propyl-cyclohexane > n-propyl-benzene. However, discrepancies among these measurements are

453

limited (about 2-3 cm/s), meaning that it is difficult to estimate this tendency with accuracy. According to our

454

measurement results, laminar flame speed of these three components is basically the same at fuel-lean conditions.

455

For fuel-rich mixtures, the propyl-cyclohexane obviously shows higher laminar flame speeds than those

456

observed for n-decane, which contradicts the results found in the literature. These discrepancies could be caused

457

by the aforementioned flame tip-opening phenomenon, which makes the flame contour determination more

458

difficult. Consequently, the measurement uncertainties increase, making the elucidation of the evolution of the

459

different laminar flame speeds more complex. For greater clarity, error bars of the propyl-cyclohexane

460

measurements in fuel-rich conditions have been added in Figure 12, representing a global uncertainty of around

461

4.5%.

462 463

3.3 Laminar flame speed of surrogate kerosene

464

In this part, laminar flame speeds of the Luche surrogate fuel consisting of n-decane (76.7 wt%),

465

n-propyl-benzene (13.2 wt%) and propyl-cyclohexane (10.1wt%), are investigated in a large rang of working

466

conditions, including variation of temperature (400-473 K), pressure (0.1-1.0 MPa) and equivalence ratio

467

(/ = 0.6 − 1.3). The measurements are compared with simulation results using the Luche kinetic mechanism

468

and compared to experimental results of neat component of surrogate Luche. Experimental data of Luche

469

surrogate is firstly plotted with neat component of surrogate in Figure 13. It depicts that the difference between

470

laminar flame speed of Luche surrogate and its three component is tiny, meanwhile its laminar flame speed is

471

more approaching to n-decane.

472

0.6 - 1.3 conditions are compared and plotted in Figure 14, which shows that the model developed by Luche is

473

effective at predicting laminar flame speeds. For all equivalence ratio investigated, simulation results using

474

Luche mechanism give slightly lower values compared with experimental results. Meanwhile, it should be noted

475

that this difference is quite close to measurement uncertainties. A comparison of the different temperature

476

conditions also illustrated that the Luche kinetic mechanism effectively predicts the effect of preheating

477

temperature on laminar flame speed that higher temperature increases the laminar flame speed. However, the

Simulation results of Luche surrogate at T= 400 - 473 K, P = 0.1 MPa and φ =

19

ACS Paragon Plus Environment

Energy & Fuels

478

mechanism underestimates flame speed of Luche surrogate about 2-3 cm/s, compared with our measurements for

479

all temperature conditions investigated. 65 60 55

SL (cm/s)

50 45 40

Propylbenzene Propylcyclohexane n-decane LUCHE experimental results

35 30 25 0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

ϕ

480 481 482

Figure 13: Laminar flame speeds of n-decane, n-propyl-benzene, propyl-cyclohexane and the LUCHE surrogate at condition of T=400 K, P=0.1 MPa and. / = 0.65 − 1.3

80

60

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

40 LUCHE mechanism T = 400 K LUCHE mechanism T = 423 K LUCHE mechanism T = 473 K present work measurements T = 400 K present work measurements T = 423 K present work measurements T = 473 K

20

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

ϕ

483 484 485 486

Figure 14: Comparisons between experimental results and simulation results of Luche surrogate fuel with

487

In order to evaluate the capability of Luche kinetic mechanism under high pressure conditions, laminar flame

488

speed measurements of Luche surrogate fuel are performed at T = 423 K, φ = 0.7 and 0.8, and under pressure

489

conditions from 0.1 to 0.8 MPa. In Figure 15, experimental results of Luche surrogate fuel under high pressure

490

conditions are compared with simulation results using Luche skeletal mechanism in a logarithmic graphic. The

491

same as observed at atmospheric pressure conditions, the Luche mechanism gives lower flame speed values

492

compared with our measurements for both equivalence ratios. Meanwhile, Luche mechanism can well reproduce

493

the effect of pressure on laminar flame speed that the higher pressure gives lower flame speed values for

temperature variations (T = 400, 423, 473 K, P = 0.1 MPa and / = 0.6 − 1.3)

20

ACS Paragon Plus Environment

Page 21 of 29

494

condition of at φ = 0.8, while a discrepancy in the dependence of pressure is observed for conditions of

495

equivalence ratio φ = 0.7. With the increments of pressure, the discrepancy between experiments and simulations

496

tends to increase. However, aforementioned discrepancy must be put into perspective because the maximum

497

difference between the simulation and the experimental results does not exceed 5 cm/s, which is a value

498

comparable to the uncertainty of our measurements under high pressure conditions. This indicates that the Luche

499

skeletal mechanism performs better under higher rather than lower pressure conditions.

500 60 55 50 45 40

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

35 30 25 20 Experimental results ϕ = 0.8 Experimental results ϕ = 0.7 LUCHE mechanism ϕ = 0.8 LUCHE mechanism ϕ = 0.7

15 10 5 0 0.1

501 502 503 504 505

0.2

0.3

0.4

0.5

0.6 0.7 0.8 0.9 1

Pressure (MPa)

Figure 15: Comparison of laminar flame speeds of the Luche surrogate between experimental results and simulation results at condition of T= 423 K, φ = 0.7 and 0.8, P = 0.1 – 0.8 MPa

3.4 Laminar flame speed of Jet A-1

506

In this part, laminar flame speed measurements of commercial kerosene fuel Jet A-1 are conducted over a large

507

range of working conditions. Measurements are also performed at high temperature and elevated pressure to

508

investigate temperature and pressure dependency on laminar flame speed of kerosene Jet A-1. An empirical

509

correlation of laminar flame speed in taking account of temperature and pressure dependency is then proposed.

510

As illustrated in Figure 16, the general tendency of the effect of the preheating temperature on the laminar flame

511

speed is in agreement with theoretical predictions, i.e., the laminar flame speed increases in line with the

512

preheating temperature. According to the power law correlation theory of temperature dependency  =

513

 0 1 4

514

flame speed at the reference condition of temperature and pressure ( 0 ) and multiplied by correction factors

515

displaying the temperature and pressure dependencies. As shown in Figure 17, laminar flame speeds are plotted

516

as a function of the preheating temperature using log-log scales, while straight lines for all the equivalence ratios

517

tested are observed.

2 5

23

[44], the laminar flame speed under higher temperature conditions can be expressed by a laminar

21

ACS Paragon Plus Environment

Energy & Fuels

85 80 75 70 65

SL (cm/s)

60 55 50 45 40 35

T = 400K T = 423K T = 445K T = 473K

30 25 20 15 0.4

0.5

0.6

0.7

0.8

519 520 521

0.9

1.0

1.1

1.2

1.3

1.4

1.5

ϕ

518

Figure 16: Laminar flame speed of Jet A-1/air with temperature variation – T = 400 K, 423 K, 445 K and 473 K with φ = 0.6-1.3, P = 0.1MPa

522 1.9

log(SL)

1.8

1.7

1.6

0.7 0.9 1.1

0.8 1.0 1.2

1.5 1.0

1.1

1.2

log (T/T0)

523 524 525

Figure 17: Log-log plot of the laminar flame speed of Jet A-1/air under atmospheric pressure and different preheating temperatures (the symbols are the experiments and the lines are linear fits) 60 55 50 45 40

SL (cm/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 29

35 30 25 20 Jet A-1 ϕ = 0.7 Jet A-1 ϕ = 0.8

15 10 5 0 0.1

526 527 528 529

0.2

0.3

0.4

0.5

0.6 0.7 0.8 0.9

Pressure (MPa)

Figure 18: Laminar flame speed of Jet A-1/air mixture at T = 423 K, P = 0.1 – 0.8 MPa and φ = 0.7 and 0.8 (the points are our measurements and dash lines are the linear fits)

22

ACS Paragon Plus Environment

Page 23 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

530

With the experimental results obtained in the present work, an empirical temperature dependency correlation was

531

proposed by using the power law equation S7 = S7 0 1 4 , where S7 0 is defined as the laminar flame speed at

532

T = 400 K and P = 0.1 MPa. The effect of the equivalence ratio is then taken into account by using the extended

533

formulation proposed by Metghalchi and Keck [65] :

534 535

2 5

23

S0 (φ) = S73,9:; ! S73,; (φ − 1) ! S7,$ (φ − 1)$ ! S7,< (φ − 1)