Experimental Measurement of Ignition Delay Times of Thermally

Jan 27, 2017 - †Key Laboratory for Green Chemical Technology of Ministry of Education, School of Chemical Engineering and Technology and ‡Collabor...
0 downloads 29 Views 1MB Size
Article pubs.acs.org/EF

Experimental Measurement of Ignition Delay Times of Thermally Cracked n‑Decane in a Shock Tube Shanshan Pei,† Hongyan Wang,† Xiangwen Zhang,†,‡ Shengli Xu,§ and Guozhu Liu*,†,‡ †

Key Laboratory for Green Chemical Technology of Ministry of Education, School of Chemical Engineering and Technology and ‡ Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin University, Tianjin 300072, China § School of Aerospace Engineering, Tsinghua University, Beijing 100084, China S Supporting Information *

ABSTRACT: The ignition characteristics of endothermic hydrocarbon fuels (with different pyrolysis degrees) were investigated in a shock tube using n-decane as model compound. Six component surrogates (CH4/C2H4/C2H6/C3H6/C3H8/n-C10H22, marked as cracked n-decane) for thermally cracked n-decane were proposed based on the chemical compositions from the thermal stressing of n-decane on electrically heated tube under 5 MPa. Ignition delay times were measured behind reflected shock waves over a temperature range of 1296−1915 K, a pressure of 0.1−0.2 MPa, and equivalence ratios of 0.5−2.0. n-Decane showed a shorter ignition delay time than cracking gas at 0.1 MPa, demonstrating higher reactivity. For cracked n-decane, it was found that thermal cracking could improve the ignitability under certain conditions to a limited degree, i.e., at T > 1480 K for x = 37.97% and x = 17.61% and at T < 1480 K for x = 62.15% (x represents the conversion of thermal cracking of n-decane) in this work. Unimolecular decomposition reactions of n-decane producing active radicals and H atom would help chain initiation via H-abstraction reaction for unreacted fuels. This initial stage might accelerate ignition by activating cracking gas at these conditions. The empirical correlations for the ignition delay time of cracking gas and n-decane were also analyzed and compared with previous works. Two models were also used to simulate the experimental data of n-decane and cracking gas and showed good agreement with experimental results, and a combined model was utilized to predict results of cracked n-decane.

1. INTRODUCTION Endothermic hydrocarbon fuels have been proposed to be used in advanced aircraft as both propellants and coolants to solve thermal management by undergoing endothermic reactions.1−3 The endothermic hydrocarbon fuel can move waste heat through both sensible heat sink and endothermic reactions bringing out extra chemical heat sinks.4−6 Then fuel experiences thermal decomposition to produce smaller hydrocarbon species (such as methane, ethane, ethylene, and propylene) and hydrogen in the heat exchanger before combustion.4,7,8 That is to say, it is a mixture of fuel and cracking products that undergoes combustion in the combustor. Therefore, it is important to investigate the ignition characteristics of cracked fuel for designing and developing the combustor in the high-speed propulsion system. There were several investigations focusing on the ignition of cracked fuels. Puri et al.9 studied the ignition delay time of two different compositions of cracked JP-7 fuel in both air and oxygen by different kinetic models over a wide range of temperatures (1200−2000 K), pressures (0.1−2 MPa), and equivalence ratios (0.5−1.5). The work analyzed the influence of ignition in both air and oxygen, and the impact of cracked fuel composition on ignition delay time. The correlations of ignition delay time of cracked JP-7 fuel were developed, concluding that cracked JP-7 fuel showed a longer ignition delay time in air, and the addition of C3 species (i.e., propane and propene) could reduce the induced time of cracked JP-7 with a limit. Castaldi et al.10 studied theoretically the impact of pyrolysis on ignition delay time of JP-8 at stoichiometric ratio based on pyrolysis calculations. Ignition delay times of three modified fuels, including propene converted to propane, toluene to methylcyclohexane, methane to ethane © XXXX American Chemical Society

and hydrogen, all based on pyrolysis products (unreacted fuel included), were compared with that of JP-8 and pyrolysis products (unreacted JP-8 including) at 0.1, 0.3 MPa, and 1000 K. Pyrolysis products ignited faster at 0.1 MPa but slower at 0.3 MPa. Only when all the methane of pyrolysis products was converted to ethylene and hydrogen, the ignition delay time would be reduced, from 121 to 107 ms at 0.1 MPa. More recently, Nakaya et al.11 conducted the effects of the thermal cracking component from n-dodecane on supersonic combustion using scramjet model combustors. A mixture consisting of H2/CH4/C2H6/C2H4 was chosen for surrogate fuel of thermally cracked n-dodecane. Five other mixtures (removal of H2, increase of CH4 or C2H4 concentration) were also added to investigate the sensitivity of components on ignition. They concluded that the ignition delay time showed little change compared with basic fuels. Colket and Spadaccini7 studied the effect of cracking on ignitability by measuring the ignition delay time of fuel/product mixtures, with 0.3CH4/0.6C2H4/0.1C7H16 (baseline-fuel) as a surrogate for endothermic reaction products, and two other mixtures added heptane and hydrogen into baseline-fuel to simulate low cracking conversion (80% C7H16) and dehydrogenation (10% H2), respectively. All these experiments were measured at a temperature from 1201 K to 1455 K, equivalence ratios from 0.5 to 1.0, and pressure of 0.7 MPa. They found that all of these three mixtures ignited more easily than heptane, and addition of heptane to baseline fuel increased ignition delay, while hydrogen Received: December 7, 2016 Revised: January 20, 2017 Published: January 27, 2017 A

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

incident shock wave sent to the endwall. The fuel aerosol is rapidly evaporated (at approximately 500−800 K) and diffusively mixed as the incident shock wave passes through the aerosol.14,15 Following the incident shock wave reflected from the endwall, the reflected shock wave is formed and it further heats the fuel gaseous mixture to ignite. Reflected-shock temperatures and pressures were calculated based on the measured incident shock speed extrapolated to the endwall and the ideal frozen-chemistry shock relations (assuming vibrationally equilibrated test gas in both the incident and reflected regions). The uncertainty in the reflected shock temperatures is within 2.0%, while the uncertainty of ignition delay time is within 15%. The ignition delay time τig in this work is defined as the time interval between the arrival of the reflected shock wave at the endwall and the time derived from back the extrapolation of the maximum slope of the whole chemiluminescence signal to the baseline, as illustrated in Figure 1. This definition tends to exhibit consistency with

decreased it a small amount. In addition, small changes of reformed fuel caused no dramatic change in ignition delay times. Obviously, the ignition of the cracked fuel is a far better understood problem involving both experimental and theoretical analysis. However, up to now the only experimental work by Colket and Spadaccini7 has been reported but with very limited surrogate components (CH4 and C2H4). Therefore, it is still necessary to do more experimental study on the ignition behaviors of cracked fuel with suitable surrogate components for further understanding of the ignition and combustion of endothermic hydrocarbon fuel in the combustor. In this paper, n-decane is selected as a surrogate of endothermic hydrocarbon fuel to investigate the ignition characteristics of cracked fuels. High-pressure pyrolysis of n-decane was first carried out in an electrically heated tube reactor to get the detailed and thus the surrogate components. Then, a shock tube was used to measure the ignition delay time of n-decane, pyrolysis products, as well as the cracked fuel behind reflected shock waves in the temperature range of 1296−1915 K, pressure of 0.1−0.2 MPa, and equivalence ratios of 0.5−2.0. The experimental results were also compared with the theoretically predicted and previous experimental results.

2. MATERIALS AND METHODS 2.1. Materials. n-Decane (mass fraction >99%) was obtained from Beijing Chemical Reagent Company, China. The cracking gas was obtained by mixing pure components including methane, ethane, ethylene, propane, and propene with molar ratios of 23/19/32/7/18 (with a purity of 99.99% for each component, 75% N2 used as diluted gas) by Beijing AP BAIF Gases Industry Co., Ltd., with an uncertainty of 1%. 2.2. Thermal Cracking of n-Decane by Electrically Heated Tube Tests. Thermal cracking of n-decane in this study was conducted by the electrically heated tube, which has been introduced in detail in our previous work.12,13 Therefore, only a brief description of the apparatus and product analysis methods is given here. The tubes (2.0 mm o.d. and 1.0 mm i.d., 100 cm in length) made of GH3128 nickel-based super alloy are purchased from China Iron & Steel Research Institute Group (Beijing, China). The back-pressure valve was used to maintain a system pressure at 5.0 MPa. The mass flow rate of 0.6 g/s was controlled by a P230 high-pressure liquid chromatography pump (Dalian Elite Analytical Instrument Co., Ltd.). An Agilent 5795 gas chromatograph with a flame ionization detector and a mass spectrometry was used to analyze the composition of the liquid products. The gaseous products were analyzed with a multichannel gas chromatograph (Agilent Micro GC 3000A) equipped with three thermal conductivity detectors and three analytical columns, i.e., Plot U (10 m × 30 μm), molecular sieve (10 m × 12 μm), and alumina (10 m × 8 μm).13 2.3. Ignition Delay Time Measured by Shock Tube. Measurements of ignition delay time were carried out behind reflected shock waves using a helium-driven, stainless-steel aerosol shock tube apparatus at Tsinghua University. The shock tube with an internal diameter of 90 mm has a 3.46-m long driven section and a 3.79-m long driver section, separated by polycarbonate diaphragms. Helium (99.999%) works as the driver gas filling in the driver section. The incident shock velocity was determined using three pressure transducers (PCB Piezotronics, M111A22, rise time of 1 μs) linearly located 61.2, 27.9, and 2 cm from the endwall. The high-speed Nicolet storage oscilloscope was used for monitored signals from transducers, with a time resolution of 1 μs per point. The fuel-air-aerosol is prepared in a 250 L mixing tank utilizing a pneumatic nebulization. The mixed-fully aerosol is slowly pulled into the test section through a plastic hose with a large diameter, low flow velocities, and small-angle to avoid n-decane droplets attaching to the pipe wall. The diaphragm dividing the driven and driver section is burst using a pin controlled by pneumatic solenoid valves, creating an

Figure 1. An example of oscilloscope traces illustrating the definition of ignition delay time. the definition that time from the beginning of shock heating and the onset of pressure rise is used to define τig.16 The onset of a pressure rise in the endwall pressure trace gives the true ignition delay time if it is available.17 However, the endwall pressure trace is ready to be influenced by potential nonideal gas dynamic effects, and the onset of the pressure rise at the point of ignition becomes very gradual which will prevent an explicit determination of the ignition time.18 In this work, the whole chemiluminescence signal exhibiting stronger signal is consistent with the onset of pressure rise measurement. Therefore, it is reasonable to determine the ignition delay time by the whole chemiluminescence signal. In addition, for hydrocarbons, ignition delay time is affected by many factors, such as the ignition temperature, pressure, concentration, and composition of the reactant mixture. Therefore, it can be expressed utilizing the empirical formula with the Arrhenius form:

τig = A exp(E /RT )[fuel]α [O2 ]β [Ar]γ

(1)

where τig is the ignition delay time (in s), E is the global activation energy (in kcal/kmol), A is an empirically constant, and R is the universal gas constant, [fuel], [O2], [Ar] are the fuel, oxygen, and argon concentrations (in mol/mL) in the combustible mixture, respectively. The empirical exponents α, β, and γ are represented of how much the ignition delay time dependent on fuel, oxidizer, and argon concentrations. Most hydrocarbon fuels exhibit strong dependence (β < −1) on oxygen concentration and weak dependence (0 < α < 1) on fuel concentration.7 Equation 1 also can be expressed in the following form: τig = APaϕbXOc 2 exp(E /RT )

(2)

where p is the pressure (in MPa). Both correlation eqs 1−2 essentially result in the same dependent degree of pressure, which may be expressed by τig ∝ pα+β.7,17 B

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels Simulations of ignition delay time were performed using Chemkinpro19 package assuming constant volume and adiabatic conditions in closed homogeneous batch reactor. The reflected shock pressure (p5) and temperature (T5) were used as the initial pressure and temperature for calculation, respectively. The calculated ignition delay time is defined as the time interval between the beginning of the simulation and the maximum rate of temperature rise (max dT/dt).7,20

3. RESULTS AND DISCUSSION 3.1. Pyrolysis of n-Decane. Thermal cracking of n-decane was carried out in an electrically heated tube reactor (2.0 mm o.d. and 1.0 mm i.d., made of GH3128 nickel-based super alloy) with a flow rate of 0.60 g/s at 5 MPa. All the experimental data of n-decane pyrolysis were shown in the Supporting Information, as shown in Figures S1−S3. n-Decane began to crack when the bulk temperature was greater than 470 °C and conversion reached 83% at 687 °C. The endothermic reaction took place at more than 550 °C with the heat sink increasing rate enhanced. There were more than 30 components produced by n-decane pyrolysis, making it difficult to investigate ignition delay time for cracked n-decane containing all the pyrolysis products. To simplify this problem, a mixture of representative components has been proposed as surrogate for pyrolysis products of n-decane in the current study. CH4, C2H6, C2H4, C3H8, and C3H6 have been chosen as surrogate for pyrolysis of kerosene21 and JP-79 in the reported papers, listed in Table 1. Moreover,

Figure 2. Thermally cracked components of n-decane at various conversions.

Table 2. Compositions (% Molar Fraction) of Fuel/Air Mixtures in the Present Study

compound fuel

current work n-decane

hydrogen (H2) methane (CH4) ethane (C2H6) ethylene (C2H4) propane (C3H8) propene (C3H6)

0.230 0.190 0.320 0.070 0.190

a

a

ref 21 kerosene

ref 23 JP-7

0.35 0.20 0.15 0.15 0.15

0.281 0.215 0.159 0.135 0.210

XC1−C3 (%)

XO2 (%)

XN2 (%)

0.5 1 0.5 1 2 1

17.62

3.228 6.254 0.000 0.000 0.000 1.186

0.000 0.000 2.784 5.008 8.342 0.576

22.549 21.844 18.655 16.781 13.977 20.266

74.223 71.902 78.561 78.211 77.681 77.972

1 1 1

37.97 61.25 81.04

0.946 0.589 0.279

1.487 2.847 4.028

19.549 18.479 17.550

78.018 78.085 78.143

n-decane

cracked n-decane

ref 9 JP-7 0.0523 0.5029 0.2053 0.2359

Xn‑decane (%)

φ

cracking gas

Table 1. Surrogate for Components of Cracked Fuels in Different Works

conversion (%)

fuel

listed in Table 2, and measured ignition delay times n-decane, cracking gas, and cracked n-decane are tabulated in Table S1 in the Supporting Information. The experimental data are also compared with results from the literature and predictions using mechanisms. 3.2.1. n-Decane. Ignition-delay experiments were conducted for n-decane in air over the range of pressures of 0.1−0.2 MPa, temperature of 1310−1915 K, and equivalence ratios of 0.5 and 1.0. Measurements of ignition delay time for n-decane in the reflected shock tube have been published by several investigators;17,25−29 nevertheless, there have few ignition delay times conducted at low pressures.17,25 The previous experimental results from Chen et al.25 using aerosol shock tube apparatus, and Horning et al.17 utilizing a shock tube were compared with data in the current work, as presented in Figure 3. As indicated in Figure 3, the present work shows good agreement with the result from Chen et al.25 at almost the same conditions. Compared with Horning et al.,17 it can be noted that the current experimental data exhibit a longer ignition delay time. The different oxygen and n-decane concentrations, and definition of ignition delay time could result in some of the differences for the experimental data.7,30 For n-decane in the current work, the ignition-delay correlation could be expressed as follows:

With 75% N2 as diluted gas.

CH4, C2H6, C2H4, C3H8, and C3H6 are major gaseous pyrolysis products of most endothermic hydrocarbon fuels, such as n-dodecane,11,22 JP-7,23 JP-10,24 JP-8 + 10024 and n-decane in current study (see Figure S3 in Supporting Information). Thus, the mixture of CH4, C2H6, C2H4, C3H8, and C3H6 is proposed as a surrogate for pyrolysis products of n-decane in this work. Figure 2 indicates the distribution of these five major gaseous products, occupied 70−85% in mole fractions and 48−69% in mass fraction of the whole products under our interested conditions (conversion from 17.61% to 82.15%). Furthermore, there was no dramatic change of the distribution for CH4, C2H6, C2H4, C3H8, and C3H6 at different conversions. Then the fixed proportion of 0.23CH4/0.19C2H6/0.32C2H4/0.07C3H8/ 0.19C3H6 is proposed by averaging the proportion of each component with minor adjustments to simplify experiment operation. Besides, the mixture with 75%N2 used as diluted gas was obtained by Beijing AP BAIF Gases Industry Co., Ltd., with an uncertainty of 1%. 3.2. Ignition Delay Time of Fuel/Air Mixtures. In this section, ignition delay times from the shock tube for n-decane, cracking gas, and cracked n-decane are analyzed to determine whether and how thermal cracking affects the ignition delay time on the parent fuel. Compositions of fuel mixtures are

τig = 1.90 × 10−13 exp(41246/RT )[fuel]0.77 [O2 ]−1.41

(3)

2

with an R correlation coefficient of 0.99505. The correlations of ignition delay time of n-decane from previous and current work are summarized in Table 3, including expressions of n-heptane.7,27 As listed in Table 3, all the ignition correlations C

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

However, it underestimated the measured data by a factor of 2 at 0.2 MPa. 3.2.2. Cracking Gas. The ignition delay times for cracking gas/air mixtures were measured using an aerosol shock tube at 0.1 and 0.2 MPa, over the temperature range of 1296−1775 K with equivalence ratios of 0.5 to 2. Comparing the current experimental data with results from previous investigations seems to be difficult for there are few experiments conducted at the similar conditions.9,33 Only Liao33 investigated the ignition delay time of cracking product of kerosene21 (composition listed in Table 1) using an aerosol shock tube at 0.1 MPa, temperature ranging over 1348−1940 K. Comparing these two results, it can be noted that the ignition delay times of the cracking product of kerosene were much larger than that of n-decane, as shown in Figure 5. For the high-temperature ignition of hydrocarbons, the production of methyl radicals promotes chain termination though R1: CH3 + CH3(+M) = C2H6(+M), inhabiting the overall ignition rate, while the production of the vinyl radical (C2H3) enhances the ignition rate by the chain branching reaction R2: O2 + C2H3O + CH2CHO, which becomes more relevant with a large ethylene content.34,35 Compared to the present study, higher methane and lower ethylene content in Liao’s work result in more CH3 and less C2H3 radicals at the initial stage, retarding the ignition reactivity. Similar to Gascoin’s study,4 kerosene presented a lower ignition delay for high ethylene content, while n-dodecane had a higher ignition delay time for a high methane content. In addition, propane is known to exhibit low-temperature chemistry but at a higher temperature, chain branching depends on oxygen concentration but not on propane concentration.36 Besides, a similar content of ethane and propene has no significant impact on different results of ignition delay time. Furthermore, simulations of pyrolysis products were added in Figure 5 by AramcoMech2.0,37 proving the same conclusion

Figure 3. Comparison of ignition delay time of n-decane between present and previous results17,25 at P = 0.1 MPa.

for n-heptane and n-decane express stronger dependence on oxygen concentration than on fuel concentration. In addition, the present data express a stronger dependency both on fuel and oxygen than that of n-heptane. Compared with present work, correlation from Olchanski et al.27 exhibits a similar dependence of n-decane and oxygen concentration for n-decane, indicating a similar dependence on pressure (with τ ∝ p−0.64 for current work and τ ∝ p −0.705 for Olchanski et al.,27 compared with τ ∝ p −0.625 from Haylett et al.31), but a different global activation energy which can be explained by different conditions in these two investigations. LLNL model32 developed by Lawrence Livermore National Laboratory for the detailed mechanism of n-alkanes consisting of 693 species and 3249 reactions, was utilized to simulate for n-decane as shown in Figure 4. LLNL model achieved a reasonable agreement with the measured ignition delay times for the stoichiometric and fuel-lean mixtures at 0.1 MPa.

Table 3. Summary of the Correlations for Different Fuels Given by Different Authors parameters A

fuels n-heptane n-decane n-decane methane ethylene cracking gas23 cracking gas cracking gas

6.76 1.00 1.90 2.21 2.82 1.66 8.34 1.55

× × × × × × × ×

10−15 10−12 10−13 10−14 10−17 10−13 10−13 10−15

conditions

E

α

β

γ

P (MPa)

T (K)

φ

dilution

source

40,160 34,240 41,246 45,000 35,000 37,388 34,775 43,215

0.40 0.60 0.77 0.33 0.00 0.1886 0.1357 0.44

−1.20 −1.305 −1.41 −1.05 −1.20 −0.9087 −0.7607 −1.28

0.00 0.08 0.00 0.00 0.00 0.00 0.00 0.00

0.14−1.37 0.17−0.10 0.1−0.2 0.30−1.52 0.5−0.8 0.1−2 0.1−2 0.1−0.2

1080−1670 1239−1616 1310−1915 1250−2500 1125−1410 1200−2000 1200−2000 1296−1775

0.5−2.0 0.33−1.46 0.5−1.0 0.8−2.17 0.5−1.0 0.5−1.5 0.5−1.5 0.5−2.0

70−97.7 75.25−92.01 air 81.3−90.9 95−97.2

ref 7 ref 27 current work ref 7 ref 7 ref 9 ref 9 current work

Figure 4. Comparison between experimental data and prediction by LLNL model for n-decane. (a) Different equivalence ratios (0.5, 1.0) at 0.1 MPa; (b) different pressures (0.1, 0.2 MPa) at Φ = 1. Symbols: experimental data; lines: predicted data. D

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

compared with experimental data and showing good agreement. The activation energy of 43 215 kcal/kmol for cracking gas is reasonable compared with most hydrocarbon fuels taken from results from Colket and Spadaccini7 (summarized in Table 3). AramcoMech2.0 is also used as a chemical kinetic mechanism for cracking gas, as shown in Figure 7. AramcoMech2.0 predicted good agreement with the present work, except a distinct under-prediction for experimental data at φ = 2, over 1300− 1650 K within a factor of 2. 3.2.3. Cracked n-Decane. Experiments on the ignition delay time of cracked n-decane were carried out to evaluate how fuel cracking affects ignition. As mentioned above, cracking gas consisting of CH4, C2H6, C2H4, C3H8, and C3H6 with 75% N2 diluted specified as a surrogate for pyrolysis products of n-decane, added with n-decane with different proportions to represent different pyrolysis degrees (conversion of 17.61%, 37.97%, 62.15%, and 81.04% to be investigated) is listed in Table 2. Measurements for ignition delay times were conducted for stoichiometric cracked n-decane/air in the aerosol shock tube over the range of temperature 1305−1867 K at 0.1 MPa, which is shown in Figure 8 with linear regression, compared with ignition-delay data of n-decane and cracking gas. As shown in Figure 8, cracking gas shows lower reactivity than pure n-decane, with a longer ignition delay time at 0.1 MPa. This result was consistent with the conclusion drawn by Liao33 that showed ignition delay time of pyrolysis gas was longer than kerosene. For cracked n-decane/air, it can be found that specific features were exhibited, but not simple enhancement or inhibition for cracked n-decane compared with n-decane. Cracked n-decane/ air at x = 81.04% ignited more slowly than the other mixtures except cracking gas over the temperature range. Furthermore, thermal cracking of n-decane accelerated ignition at specific conditions: cracked n-decane with x = 17.61% and x = 37.97% at T > 1480 K and x = 62.15% at T < 1480 K, with a limited degree. Colket and Spadaccini7 focused on the effects of thermal and catalytic cracking of n-heptane with 0.3CH4/0.6C2H4/0.1C7H16 (i.e., 0.9gases/0.1C7H16), concluding that fuel cracking improved ignition at temperatures from 1201 K to 1455 K, equivalence ratios from 0.5 to 1.0 and pressure of 7 atm. In present work, experimental results of cracked n-decane showed that thermal cracking would improve ignition of n-decane at T < 1480 K for conversion of 61.25% of pyrolysis (i.e., 0.83 cracking gas/0.17 n-decane, see in Table 2), which is consistent with the result from Colket and Spadaccini7 at the same temperature range for a similar proportion of gas/fuel. For simulation for cracked n-decane, a mechanism which could well simulate that both light hydrocarbons and n-decane are needed. But LLNL model32 could not predict light hydrocarbons very well especially for ethylene and propane. To solve this problem, the Combined Mech consisting of 875 species and 4270 reactions, merging from AramcoMech2.0 and LLNL model, was obtained by ANSYS Workbench.38 Combined Mech could predict the ignition delay time of both n-decane and light hydrocarbon (CH4, C2H6, C2H4, C3H8, and C3H6) at the same conditions in present work. The merging principles and validation of Combined Mech are in the Supporting Information, as shown in Figure S4 and Figure S5. Combined Mech was applied to simulate the ignition delay time of cracked n-decane, as illustrated in Figure 9. The difference for cracked n-decane and pure n-decane is very small

Figure 5. Comparison and simulation for cracking gas from Liao33 and current work at 0.1 MPa. Symbols: experimental data; lines: simulation data.

that ignition delay times of cracking product of kerosene was larger than that of n-decane. However, the difference in experimental data of these two studies is much larger than the difference in simulation. By carefully rechecking Liao’s experiments, a rectangular tube was used to measure ignition delay time in their study. Compared to the shock tube in current work, the uneven boundary layer and flow field distribution in rectangular tube might be responsible for the large difference in experimental results. The difference in experimental equipment between Liao’s work and the current study might be the major cause for a much larger ignition delay time of the cracking product of kerosene. Puri9 established theoretical calculation for two kinds of cracked JP-7 fuel (summarized in Table 1), correlations of which are also listed in Table 3 (where the preexponential has been modified here to obtain the same units with others), indicating a weaker dependency on fuel and oxygen concentrations than data from current measurement. The empirical correlation of experimental data in this work could be expressed as follows: τig = 1.55 × 10−15 exp(43215/RT )[fuel]0.44 [O2 ]−1.28

(4)

τig = 1.16 × 10−10ϕ0.61 exp(41948/ RT )(p = 0.1 MPa)

(5)

2

with R correlation coefficient 0.98221 and 0.98033 for eqs 4 and 5, respectively. The correlation (4) is plotted in Figure 6,

Figure 6. Correction compared with experimental data of cracking gas. E

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Figure 7. Measured and predicted ignition delay times for cracking gas: (a) different equivalence ratios (0.5−2.0) at 0.1 MPa; (b) different pressures (0.1, 0.2 MPa) at Φ = 1. Symbols: experimental data; lines: calculated data.

Westbrook et al.,39 while different tendencies were predicted by JetSurf2.0,40 Zeng et al.,29 and Wang et.al.41 The possible reasons to these results could be explained as follows. For a 81.04% cracked mixture, cracking gas dominated the ignition delay time by occupying an almost 93.5% mole proportion in the mixture. Hence, a 81.04% cracked mixture showed a similar and limited shorter ignition delay time with cracking gas. For cracked n-decane at x = 17.61%, x = 37.97% at T > 1480 K and x = 62.15% at T < 1480 K, the initial step was unimolecular decomposition reactions of n-decane producing active radicals and H atom, which would help chain initiation via H-abstraction reaction of unreacted fuels and participate chain propagation reactions.42 This initial stage might accelerate the ignition delay time by activating cracking gas at these conditions. The kinetic mechanism should get further specified and modified by experimental data for cracked fuel and provide explanations for experimental results. And more efforts should be focused on experiments of ignition delay time for cracked fuel under various conditions.

Figure 8. Comparison of measured ignition delay times of stoichiometric n-decane/air, cracking gas/air, and cracked n-decane/ air mixtures at 0.1 MPa. Symbols: experimental data, lines: linear regressions.

from the simulation. Cracked n-decane/air at x = 81.04% ignited more slowly than the other mixtures except cracking gas over the temperature range. At lower temperature, the predicted tendency by Combined Mech is consistent with the experimental data; i.e., thermal cracking could enhance the ignitability of n-decane at certain conditions at T < 1480 K for x = 62.15%. But at higher temperature, simulation shows a different tendency compared with experimental results. Compared with computations performed by other kinetic models, simulated features were reproduced by a reaction mechanism of

4. CONCLUSIONS The ignition characteristics of thermal cracking of endothermic hydrocarbon fuels were investigated using n-decane as a model compound. First, high-pressure pyrolysis of n-decane was carried out in an electrically heated tube reactor to get the detailed and thus the surrogate components. A mixture of 0.23CH4/ 0.19C2H6/0.32C2H4/0.07C3H8/0.19C3H6 (marked as cracking gas) was selected as a surrogate of production, which occupied

Figure 9. Comparison between experimental data and prediction by the LLNL model for cracked n-decane. Symbols: experimental data; lines: predicted data. F

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

(5) Edwards, T. I. M. Cracking and deposition behavior of supercritical hydrocarbon aviation fuels. Combust. Sci. Technol. 2006, 178 (1−3), 307−334. (6) Jiang, R.; Liu, G.; You, Z.; Luo, M.; Wang, X.; Wang, L.; Zhang, X. On the Critical Points of Thermally Cracked Hydrocarbon Fuels under High Pressure. Ind. Eng. Chem. Res. 2011, 50 (15), 9456−9465. (7) Colket, M. B.; Spadaccini, L. J. Scramjet Fuels Autoignition Study. J. Propul. Power 2001, 17 (2), 315−323. (8) Ponnuthurai, G.; Casey, F.; Michael, K.; Yangye, Z.; David, F. D.; Ronald, K. H.; Barry, V. K., Ignition of Light Hydrocarbon Mixtures Relevant to Thermal Cracking of Jet Fuels. In 54th AIAA Aerospace Sciences Meeting, American Institute of Aeronautics and Astronautics, 2016. (9) Puri, P.; Ma, F.; Choi, J.-Y.; Yang, V. Ignition characteristics of cracked JP-7 fuel. Combust. Flame 2005, 142 (4), 454−457. (10) Castaldi, M.; Leylegian, J.; Chinitz, W.; Modroukas, D., Development of an Effective Endothermic Fuel Platform for Regeneratively-Cooled Hypersonic Vehicles. In 42nd AIAA/ASME/ SAE/ASEE Joint Propulsion Conference & Exhibit, American Institute of Aeronautics and Astronautics, 2006. (11) Nakaya, S.; Tsue, M.; Kono, M.; Imamura, O.; Tomioka, S. Effects of thermally cracked component of n-dodecane on supersonic combustion behaviors in a scramjet model combustor. Combust. Flame 2015, 162 (10), 3847−3853. (12) Liu, G.; Cao, Y.; Jiang, R.; Wang, L.; Zhang, X.; Mi, Z. Oxidative desulfurization of jet fuels and its impact on thermal-oxidative stability. Energy Fuels 2009, 23 (12), 5978−5985. (13) Jiang, R.; Liu, G.; Zhang, X. Thermal cracking of hydrocarbon aviation fuels in regenerative cooling microchannels. Energy Fuels 2013, 27 (5), 2563−2577. (14) Haylett, D. R. The Development and Application of Aerosol Shock Tube Methods for the Study of Low-Vapor-Pressure Fuels, Stanford University, 2011. (15) Davidson, D. F.; Haylett, D. R.; Hanson, R. K. Development of an aerosol shock tube for kinetic studies of low-vapor-pressure fuels. Combust. Flame 2008, 155 (1−2), 108−117. (16) Tsang, W.; Lifshitz, A. Shock tube techniques in chemical kinetics. Annu. Rev. Phys. Chem. 1990, 41 (1), 559−599. (17) Horning, D. C.; Davidson, D.; Hanson, R. Study of the hightemperature autoignition of n-alkane/O/Ar mixtures. J. Propul. Power 2002, 18 (2), 363−371. (18) de Vries, J.; Hall, J. M.; Simmons, S. L.; Rickard, M. J. A.; Kalitan, D. M.; Petersen, E. L. Ethane ignition and oxidation behind reflected shock waves. Combust. Flame 2007, 150 (1−2), 137−150. (19) ANSYS CHEMKIN-Pro 17.2; ANSYS, 2016. (20) Zhang, J.; Hu, E.; Zhang, Z.; Pan, L.; Huang, Z. Comparative Study on Ignition Delay Times of C1−C4 Alkanes. Energy Fuels 2013, 27 (6), 3480−3487. (21) Fan, X.; Yu, G.; Li, J.; Lu, X.; Zhang, X.; Sung, C.-J. Combustion and ignition of thermally cracked kerosene in supersonic model combustors. J. Propul. Power 2007, 23 (2), 317−324. (22) Jiang, R.; Liu, G.; He, X.; Yang, C.; Wang, L.; Zhang, X.; Mi, Z. Supercritical thermal decompositions of normal-and iso-dodecane in tubular reactor. J. Anal. Appl. Pyrolysis 2011, 92 (2), 292−306. (23) Edwards, T.; Anderson, S. Preliminary results of high temperature JP-7 cracking assessment. Preprints-American Chemical Society, Division of Petroleum Chemistry 1992, 37 (2), 549−559. (24) Huang, H.; Spadaccini, L.; Sobel, D. In Endothermic heat-sink of jet fuels for scramjet cooling, 38th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2002; p 3871. (25) Chen, Q.; Dou, Z.; Li, L. Experimental Study of Measuring the Ignition Delay Time of Toluene and n-Decane by a Shock Tube (in Chinese). Chin. J. Energetic Mater. 2015, 23 (10), 971−976. (26) Zhukov, V. P.; Sechenov, V. A.; Starikovskii, A. Y. Autoignition of n-decane at high pressure. Combust. Flame 2008, 153 (1−2), 130− 136. (27) Olchanski, E.; Burcat, A. Decane oxidation in a shock tube. Int. J. Chem. Kinet. 2006, 38 (12), 703−713.

70−85% of products and more than 92% of gaseous products in mole fractions. Then, a shock tube was used to measure the ignition delay time of n-decane, cracking gas, as well as the cracked fuel behind reflected shock waves in the temperature range of 1296−1915 K, pressure of 0.1−0.2 MPa, and equivalence ratios of 0.5−2.0. n-Decane showed a higher reactivity time than cracking gas at 0.1 MPa. For cracked n-decane, it was found that thermal cracking could enhance the ignitability of n-decane at certain conditions to a limited degree, i.e., at T > 1480 K for x = 37.97% and x = 17.61% and at T < 1480 K for x = 62.15% in this paper. Earlier decomposition of n-decane producing H atom and alkyl radicals might be attributed to this improvement. For x = 81.04%, ignition delay time was similar to cracking gas, showing a longer ignition delay time than other mixtures except cracking gas. The experimental results were also compared with previous and theoretical predicted experimental results. Simulation from LLNL model and AramcoMech2.0 gave good agreements with the current data of n-decane and cracking gas, respectively. The Combined Mech merging with the LLNL model and AramcoMech2.0 was used to simulate the ignition delay time of n-decane. Pyrolysis extent and temperature has an impact on the ignition of cracked fuels in the current work. Further efforts should be focused on more detailed and systemic experiments under wider conditions. Furthermore, the kinetic mechanism should get further specified for both light and heavy hydrocarbon, modified by experimental data for cracked fuel and provide detailed theoretical explanations for experimental results.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.energyfuels.6b03242. Experimental results including the wall and bulk temperature, pyrolysis conversion, the total heat sink and major species of n-decane cracking (Figures S1−S3); validation of Combined Mech (Figure S4 and Figure S5); measured ignition delay times of n-decane, cracking gas, and cracked n-decane in air (Table S1) (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (21522605). REFERENCES

(1) Huang, H.; Spadaccini, L. J.; Sobel, D. R. In Fuel-Cooled Thermal Management for Advanced Aero Engines, ASME Turbo Expo 2002: Power for Land, Sea, and Air, 2002; American Society of Mechanical Engineers; pp 367−376. (2) Edwards, T. Liquid Fuels and Propellants for Aerospace Propulsion: 1903−2003. J. Propul. Power 2003, 19 (6), 1089−1107. (3) Gong, C.; Ning, H.; Xu, J.; Li, Z.; Zhu, Q.; Li, X. Experimental and modeling study of thermal and catalytic cracking of n-decane. J. Anal. Appl. Pyrolysis 2014, 110, 463−469. (4) Gascoin, N.; Abraham, G.; Gillard, P. Synthetic and jet fuels pyrolysis for cooling and combustion applications. J. Anal. Appl. Pyrolysis 2010, 89 (2), 294−306. G

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels (28) Zhang, C.; Li, B.; Rao, F.; Li, P.; Li, X. A shock tube study of the autoignition characteristics of RP-3 jet fuel. Proc. Combust. Inst. 2015, 35 (3), 3151−3158. (29) Zeng, M.; Yuan, W.; Wang, Y.; Zhou, W.; Zhang, L.; Qi, F.; Li, Y. Experimental and kinetic modeling study of pyrolysis and oxidation of n-decane. Combust. Flame 2014, 161 (7), 1701−1715. (30) Xu, C.; Konnov, A. A. Validation and analysis of detailed kinetic models for ethylene combustion. Energy 2012, 43 (1), 19−29. (31) Haylett, D. R.; Davidson, D. F.; Hanson, R. K. Ignition delay times of low-vapor-pressure fuels measured using an aerosol shock tube. Combust. Flame 2012, 159 (2), 552−561. (32) Sarathy, S. M.; Westbrook, C. K.; Mehl, M.; Pitz, W. J.; Togbe, C.; Dagaut, P.; Wang, H.; Oehlschlaeger, M. A.; Niemann, U.; Seshadri, K.; Veloo, P. S.; Ji, C.; Egolfopoulos, F. N.; Lu, T. Comprehensive chemical kinetic modeling of the oxidation of 2methylalkanes from C7 to C20. Combust. Flame 2011, 158 (12), 2338−2357. (33) Qin, L. Experimental studies on autoignition phnomena of kerosene and cracked kerosene in a shock tube (in Chinese). University of Science and Technology of China, 2009. (34) Ranzi, E.; Frassoldati, A.; Grana, R.; Cuoci, A.; Faravelli, T.; Kelley, A. P.; Law, C. K. Hierarchical and comparative kinetic modeling of laminar flame speeds of hydrocarbon and oxygenated fuels. Prog. Energy Combust. Sci. 2012, 38 (4), 468−501. (35) Liu, W.; Kelley, A.; Law, C. Flame propagation and counterflow nonpremixed ignition of mixtures of methane and ethylene. Combust. Flame 2010, 157 (5), 1027−1036. (36) Healy, D.; Curran, H. J.; Simmie, J. M.; Kalitan, D. M.; Zinner, C. M.; Barrett, A. B.; Petersen, E. L.; Bourque, G. Methane/ethane/ propane mixture oxidation at high pressures and at high, intermediate and low temperatures. Combust. Flame 2008, 155 (3), 441−448. (37) Li, Y.; Zhou, C.-W.; Somers, K. P.; Zhang, K.; Curran, H. J., The oxidation of 2-butene: A high pressure ignition delay, kinetic modeling study and reactivity comparison with isobutene and 1-butene. Proc. Combust. Inst. 2016, DOI: 10.1016/j.proci.2016.05.052 (38) ANSYS WORKBENCH 17.2; ANSYS, 2016. (39) Zhang, K.; Banyon, C.; Togbé, C.; Dagaut, P.; Bugler, J.; Curran, H. J. An experimental and kinetic modeling study of n-hexane oxidation. Combust. Flame 2015, 162 (11), 4194−4207. (40) Wang, H.; Dames, E.; Sirjean, B.; Sheen, D. A.; Tango, R.; Violi, A.; Lai, J. Y. W.; Egolfopoulos, F. N.; Davidson, D. F.; Hanson, R. K.; Bowman, C. T.; Law, C. K.; Tsang, W.; Cernansky, N. P.; Miller, D. L.; Lindstedt, R. P.A high-temperature chemical kinetic model of n-alkane (up to n-dodecane), cyclohexane, and methyl-, ethyl-, n-propyl, and nbutyl-cyclohexane oxidation at high temperatures, JetSurF version 2.0, September 19, 2010, http://web.stanford.edu/group/haiwanglab/ JetSurF/JetSurF2.0/index.html. (41) Wang, H.; Jin, J.; Wang, J.; Tan, N.; Li, X. Combustion Mechanism of n-Decane at High Temperatures and Kinetic Modeling of Ignition Delay for Aviation Kerosene. Chem. J. Chin. University-Chin. 2012, 33 (2), 341−345. (42) Yang, F.; Deng, F.; Zhang, P.; Hu, E.; Cheng, Y.; Huang, Z. Comparative Study on Ignition Characteristics of 1-Hexene and 2Hexene Behind Reflected Shock Waves. Energy Fuels 2016, 30 (6), 5130−5137.

H

DOI: 10.1021/acs.energyfuels.6b03242 Energy Fuels XXXX, XXX, XXX−XXX