- Email: [email protected]
Self-ignition and pyrolysis of acetone behind reflected shock waves
Acta Astronautica xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Acta Astronautica journal homepage: www.elsevier.com/locate/actaastro
Self-ignition and pyrolysis of acetone behind reflected shock waves A.M. Tereza, S.P. Medvedev∗, V.N. Smirnov N.N. Semenov Federal Research Center for Chemical Physics, Russian Academy of Sciences, Russia
A R T I C LE I N FO
A B S T R A C T
Keywords: Self-ignition Pyrolysis Acetone Emission measurements Numerical simulation Chemical kinetics Shock waves Chemiluminescence
The self-ignition of a stoichiometric acetone−oxygen mixture diluted with argon was studied both experimentally and numerically. The experiments were performed behind reflected shock waves over a temperature range from 1280 to 1810 K at a total gas concentration of [M]50 ≈ 10−5 mol/cm3. The process was monitored by recording the signals of absorption by methyl radicals (λ = 216.5 nm) and emission from electronically excited OH* radicals (λ = 308 nm) and CO2* molecules (λ = 370 nm). Numerical simulations within the framework of various detailed kinetic mechanisms were performed to reproduce our own and published experimental data on the pyrolysis and self-ignition of acetone behind reflected shock waves, including the concentration profiles of acetone and CH3 in the ground state and electronically excited OH* and CO2*, as well as the temperature dependence of the self-ignition delay time. It has been established that various detailed kinetic mechanisms describe the kinetic characteristics of the pyrolysis of acetone in shock waves with varying degrees of accuracy. At the same time, all of them predicted the measured ignition delays within a factor of two. A sensitivity analysis showed that the reactions determining the pyrolysis of acetone produce only a slight effect on the branchedchain ignition process, so that the relevant rate constants only slightly affect the ignition delay time.
1. Introduction Studies of the kinetics of the decomposition and oxidation of acetone are motivated by its widespread use in various technological fields. As far back as the early 1960s, acetone was used as a component of liquid jet fuel for the Cricketsonde meteorological rocket [1]. Currently, patents on hypergolic systems involving acetone as a solvent in gels have been granted [2]. Based on the thermodynamic properties, the authors of [3], examined acetone as an alternative to existing rocket fuels. Since acetone is known to be formed in the atmosphere through the oxidation of various hydrocarbons [4,5], it is interesting to elucidate the role of acetone and other ketones in atmospheric chemistry, especially in the upper atmosphere, where it has been detected [6] in amounts capable of affecting the ozone layer [7]. The process of generating intense shock waves during the expansion of combustion products is of considerable interest for developing various detonation propulsion devices and ensuring safety of their operation [8,9]. A practically important case is the formation of spherical shock waves, with the conical shock tube (CST) being an effective tool for modeling such waves [10]. To optimize the operating parameters of the CST, it is advantageous to use a burning or detonating gas mixture as a driver medium. Acetone is one of the promising fuels for this purpose, because its mixtures with oxygen exhibit a high detonability
∗
even under conditions of rapid expansion [11]. Reliable information on the kinetics of its ignition at high temperatures is needed to develop a model of the combustion and detonation of acetone−oxygen−nitrogen mixtures in the driver section of the conical shock tube. Despite numerous studies on the pyrolysis of acetone and its selfignition in mixtures with oxygen and air, see, e.g., Refs. [12–21], a number of issues concerning the behavior of both the initial compound and various products of these processes remain unresolved [18–20]. The published detailed kinetic mechanisms (DKMs) are mainly aimed at modeling the temperature dependences of the ignition delays of various hydrocarbons in the presence of acetone additives [16,17]. At the same time, such data are of little use in simulating the time profiles of various products observed during the pyrolysis and self-ignition of acetone [19]. Therefore, in studying the kinetics of these processes, it is necessary to analyze various DKMs to select the most suitable for modification. In this regard, obtaining new kinetic data and constructing additional reaction blocks with corresponding rate constants can significantly improve the predictive capabilities of various DKMs [22]. On the other hand, the development of methods for recording ignition processes behind shock waves requires a comprehensive approach to interpreting experimental data by means of numerical simulations. An exhaustive DKM should include electronically excited species, a feature that makes it possible to simulate the time histories of
Corresponding author. E-mail addresses: [email protected] (A.M. Tereza), [email protected] (S.P. Medvedev), [email protected] (V.N. Smirnov).
https://doi.org/10.1016/j.actaastro.2020.03.045 Received 25 February 2020; Received in revised form 26 March 2020; Accepted 28 March 2020 0094-5765/ © 2020 IAA. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: A.M. Tereza, S.P. Medvedev and V.N. Smirnov, Acta Astronautica, https://doi.org/10.1016/j.actaastro.2020.03.045
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 1. Experimentally measured (solid lines) CH3 absorption, OH* emission, and pressure P time profiles at T50 = 1510 K and P50 = 0.126 MPa. The dashed and dash-dot lines represent the simulation results by the DKM from Ref. [30].
behind incident and reflected shock waves. The endwall was located 1.1 cm downstream of the windows. Shock waves were generated by the spontaneous rupture of one or a stack of 0.005-cm-thick aluminum diaphragms, depending on the conditions required. The driver gas was helium. Since a diaphragm of a given thickness ruptures at about the same driver gas pressure, in experiments in which it was necessary to vary the temperature behind the shock wave (incident or reflected) while maintaining the shock-compressed gas density at approximately constant level, the shock wave was generated by helium−air mixtures of different compositions. All other things equal, the higher the air content in the driver mixture, the lower the temperature behind the shock wave. We recorded the chemiluminescence signals from electronically excited OH* radicals and CO2* molecules and the absorption time profile of CH3 radicals. The time histories of the emissions from OH* and CO2* were monitored at wavelength of λ = 308 ± 2.5 nm and λ = 363 ± 3.5 nm, respectively. The signal from OH* was detected using a DMR-4 two-prism quartz monochromator in conjunction with a FEU-39A photomultiplier, whereas the emission from CO2* was directed into a FEU-52 photomultiplier through a combined filter composed of an interference filter and UF-2 (ultraviolet) and SZS-21 (blue green) 0.3-cm-thick optical glasses. The absorption of CH3 radicals was recorded at a wavelength of λ = 216.5 nm using a DMR-4 monochromator in combination with a FEU-39A photomultiplier. The light source was a high-frequency electrodeless copper lamp powered by a PPBL-3V generator. Such a lamp gives a discrete spectrum of the glow of copper atoms [27]. The concentration of CH3 radicals was determined according to the Beer−Lambert−Bouguer law:
their concentrations based on the corresponding chemiluminescence signals. Since emission measurements are highly sensitive and unobtrusive, they are widely used in studies on the self-ignition of fuel−oxidizer mixtures behind shock waves [23–25]. The aim of the present work is to study the kinetics of the pyrolysis and self-ignition of acetone−oxygen−argon mixtures behind reflected shock waves by performing numerical simulations within the framework of various DKMs and comparing the results with the available experimental data on the time profiles of CH3 absorption and the emission of electronically excited OH* and CO2*. 2. Experimental The experimental setup was described in detail elsewhere [26] with only some general features given below. The experiments were performed in a stainless steel shock tube with an inner diameter of 7.5 cm, driver section length of 150 cm, and driven section length of 320 cm. The driven section was evacuated to 10−2 Torr with two 2-NVR-5D forepumps and then to 10−3 Torr with a NVDS-100 diffusion pump. The vacuum/gas-handling system had two liquid-nitrogen-cooled traps. The degree of evacuation was controlled with a VIT-2 ionization-thermocouple vacuum gauge (here and below, the capitalized abbreviations are the transliterations from Russian of the types of instruments and devices). The driver section was evacuated with a 2-NVR-5D forepump through a liquid-nitrogen-cooled trap to a pressure of 0.1 Torr. The rate of leakage into the driven section did not exceed ~10−4 Torr/min. Before each experiment, the driven section was twice purged with argon used to prepare test mixtures with intermediate evacuations to 10−1 Torr. The total density and temperature of the gas behind the incident and reflected shock waves were calculated from the composition of the test mixture, initial pressure, and velocity of the incident shock wave by using ideal shock tube theory. The incident shock wave velocity was measured over two measurement intervals formed by three sequentially flush-mounted pressure gauges G1, G2, and G3. The distances between the gauges were G1−G2 = 52.8 cm and G2−G3 = 28.1 cm, with the endwall-nearest gauge G3 mounted 4.0 cm upstream of the observation windows. The fourth pressure sensor, 0.7 cm in diameter, located in the same cross section as the observation windows, was intended to record the pressure profiles
[CH3] = (1/ σ l)ln(I0/ I ), where I0 and I are the initial (before the arrival of the shock wave) and current signal intensity, l = 7.5 cm is the internal diameter of the shock tube, and σ is the absorption coefficient. The absorption coefficient was taken from [28],
σ = 3.25 107 T −0.36 exp(500/T )exp(−T /2500) cm2 /mol.
3. Experimental results The self-ignition of a 0.5%(CH3)2CO−2%O2−Ar stoichiometric 2
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 2. Experimental (solid lines) and simulated (interrupted lines) time profiles of electronically excited ОН* radicals and CO2* molecules. T50 = 1638 K, P50 = 0.133 MPa. The dashed and dash-dot lines represent the simulation results by the DKM from Ref. [30].
the DKMs describe the pyrolysis of acetone behind reflected shock waves. For this, the experimental results on measuring the yield of CH3 radicals during the pyrolysis of a 0.25% (CH3)2CO−Ar mixture [19] were considered. Fig. 3 compares the experimental data from Refs. [19] and the DKM calculations from Refs. [30,32–35] at three different temperatures. As can be seen from Fig. 3a, at T50 = 1469 K (Fig. 3a), the profiles calculated using the DKMs from Refs. [33,34] significantly differ both in the time of reaching the maximum concentration and in the rate of consumption of CH3 radicals. At the same time, the time profiles simulated using the DKM from Ref. [32] predict well the maximum concentration of CH3, but show a very rapid decrease in the concentration of methyl radicals, resulting in its value at late times (after 600 μs) approximately twofold less than that observed in the experiment. A similar rate of CH3 absorption decay is predicted by the DKM given in Ref. [33], which enables it to describe well the level of CH3 concentration at late pyrolysis times, despite a discrepancy between the calculated and measured concentrations of CH3 radicals. Calculations using the DKM from Ref. [35] and the extended mechanism from Ref. [30], although predicting different decay rates after reaching the maximum CH3 concentration, nevertheless, closely reproduce the concentration of CH3 in late stages of the pyrolysis. At lower temperatures (Fig. 3b), the description of the CH3 concentration time profile within the framework of the DKMs from Refs. [32–35] is slightly worse compared to the predictions of the DKM from Ref. [30]. This DKM, although it does not quite accurately describe the maximum CH3 concentration, but rather well reproduces the decay rate and final concentration of methyl radicals. With a further decrease in temperature (Fig. 3c), all the examined DKMs describe the temporal profiles of the concentration of CH3 measured in Ref. [19] only qualitatively, probably due to a higher error in the measurements of the absolute concentration of methyl radicals at low temperatures. To find out why the different DKMs produce different results, we analyzed the sensitivity of the leading stages of acetone pyrolysis for the conditions specified in the caption of Fig. 3a. Since the DKMs used in this work contain nearly the same blocks of leading reactions, which differ only in the corresponding rate constants, the sensitivity analysis was carried out using only the DKM from Ref. [30]. The time dependence of the local sensitivity d(ln[CH3])/d(lnki) [37] is
mixture was carried out behind reflected shock waves in the temperature range of 1280–1720 K and a total gas concentration of [M]50≈10−5 mol/cm3 (hereinafter, the subscript 50 denotes the initial gas conditions behind the reflected shock wave). Fig. 1 shows typical OH* emission and CH3 absorption time profiles. The self-ignition delay τ was defined as the time interval between the arrival of the reflected shock wave and the time of maximum OH* emission intensity (Fig. 1). As can be seen from Fig. 1, the concentration of СН3 radicals passes through a maximum, reaches a slowly declining plateau, and then decreases rapidly to a nearly zero level at the time the OH* luminescence signal passes through its maximum. This suggests that, by the time the (CH3)2CO−O2−Ar mixture self-ignites, most of the CH3 radicals formed during the decomposition of acetone have reacted. As can be seen, no OH* emission is observed nearly throughout the induction period−only a slight increase in the OH* emission is observed immediately before the ignition. Fig. 2 displays OH* and CO2* emission time profiles typical of hightemperature experiments. It can be seen that the maxima of the OH* and СО2* luminescence signals practically coincide, which was also observed for the self-ignition of stoichiometric mixtures of other hydrocarbons [29,30]. Note also that, at high temperatures, a pre-peak step is observed (Fig. 2), which is absent at lower temperatures (Fig. 1). Figs. 1 and 2 show that, after reaching a maximum, the OH* and CO2* signals decrease sharply to a slowly declining level. The experimentally recorded time profiles of CH3 absorption and OH* and CO2* emissions were numerically simulated using various detailed kinetic mechanisms. 4. Numerical simulations All the simulations were performed using the CHEMKIN III software package [31] under the condition of constant volume (V = const). The DKMs were taken from Refs. [30] (ChemphysMech 2.0), [32] (LLNL 2.0), [33] (JetSurf 2.0), [34] (Aramco 1.3), [35] (ERC), and [36] (CRECK), with the corresponding NASA thermodynamic polynomials given in these studies. The choice of these DKMs was motivated by the presence of reactions involving acetone in them. Before simulating our own experimental results, we examined how 3
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
C2H6+H = C2H5+H2
(3)
(СН3)2СО = CH3CO + CH3
(4)
In this case, the primary decomposition reaction of acetone (4) rapidly shifts to the left after ~200 μs; i.e., when the starting acetone is nearly completely consumed. A further rapid decomposition of CH3CO leads to the formation of a second CH3 radical, which participates in the recombination (1) and disproportionation (2) reactions. Clearly, atomic hydrogen can contribute to the formation and consumption of CH3 radicals: (СН3)2СО+Н = CH3COСН2+H2 → CH3+COСН2+H2
(5)
СН3+Н(+М) = СН4(+М)
(6)
Note, however, that the estimates made by the brute force method [37] showed that the influence of reactions (5) and (6) on the profiles in Fig. 3 is insignificant. In the considered temperature range, the second fragment of the dissociation of the acetone molecule, the CH3CO radical, decays quite quickly, CH3CO = СН3+СО,
forming another CH3 radical and a CO molecule. Reaction (7) proceeds so quickly that it could be combined with reaction (4). It can be expected that a reliable determination of the rate constants of reactions (1)−(4) will provide a fairly good description of the CH3 profiles in Fig. 3. It should be noted that substantial difficulties are encountered in determining the rate constants of the unimolecular dissociation reactions, (1), (4), and (6). These reactions occur in a falloff pressure region, with their rate constants being dependent on a much larger number of parameters than those for bimolecular reactions (2) and (3), expressed in a three-parameter Arrhenius form [38]. In the present study, we did not refine the rate constants of reactions (1)−(4), (5) and (6) given in Refs. [13,30]. The time histories of CH3 absorption and OH* luminescence for the self-ignition of acetone−oxygen–argon mixtures calculated using various DKMs is displayed in Figs. 1, 5–7, which demonstrate that all the DKMs give approximately same time profiles of the formation and consumption of CH3 radicals. For relatively high temperatures (Figs. 1, 5, and 6), all the DKMs predict a fast formation of CH3 radicals, as observed in experiment. After passing the maximum, the CH3 concentration decreases to the moment of self-ignition, corresponding to the maximum of the OH* emission signal. This is clearly demonstrated by calculations based on the DKMs from Refs. [30,34]. The fact that the OH* emission time profile calculated by the DKM from Ref. [33] (Fig. 5) differs markedly can be explained solely by the ratio between the contribution from the different reactions of OH* excitation at the early and late stages of the self-ignition of acetone, but not by any drawbacks of the DKM proposed in Ref. [33]. An inspection of the times of occurrence of the OH* emission maximum and the sharp drop in the СН3 concentration to zero predicted by the DKM from Ref. [33] show that they are nearly identical, as in the case of the DKMs from Ref. [30,34]. The behavior of OH* emission profiles calculated by the DKM reported in Ref. [33] for the ignition of hydrocarbons was analyzed in Ref. [30]. Since we did not examine the OH* emission profiles calculated by the DKM from Ref. [33] in detail, they are not represented in Figs. 6 and 7. The CH3 profiles calculated by the DKMs from Refs. [32, 35] behave very similar to those simulated using the DKMs reported in [30, 33, 34]. Since the DKMs from [32, 35] include no reaction blocks involving OH*, Figs. 5-7 show only the CH3 concentration profiles calculated by these DKMs. While the DKMs from Refs. [30,32–35] satisfactorily reproduce the measured time profiles of CH3 absorption at relatively high temperatures (Figs. 1, 5, and 6), this is not the case at low temperatures (Fig. 7). After a sharp initial rise, the measured CH3 concentration continues to increase at a slower rate, reaching a maximum at a time equal to approximately one third of the observed ignition delay (Fig. 7).
Fig. 3. Comparison of experimental [19] and DKM-calculated СН3 profiles for a 0.25% acetone−argon mixture. Experimental condition: (а) T50 = 1469 K, Р50 = 0.148 MPa; (b) T50 = 1393 K, Р50 = 0.155 MPa; and (c) T50 = 1273 K, Р50 = 0.165 MPa. The solid line represents the experimental data, and the interrupted lines are the corresponding calculated profiles by the DKMs from Refs. [30,32–35].
displayed in Fig. 4. Fig. 4 shows the main pathways for the formation and consumption of CH3 radicals during acetone pyrolysis: CH3+CH3(+M) = C2H6(+M)
(1)
CH3+CH3 = C2H5+H
(2)
(7)
4
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 4. Analysis of the local sensitivity of the main pathways of the formation and consumption of CH3 radicals during the pyrolysis of a 0.25%(CH3)2CO−Ar mixture. The experimental conditions are specified in the caption of Fig. 3 (panel a).
approaches and passes through its maximum, indicative of the occurrence of self-ignition (Figs. 1, 5–7). Moreover, even if the description of the CH3 absorption time profile is not satisfactory (Fig. 7), this nonzero level of the CH3 concentration is much the same as the experimentally observed. The results of a sensitivity analysis for the experimental conditions specified in Fig. 1 are displayed in Figs. 8 and 9. In contrast to Fig. 4, which shows the sensitivity to the reactions involved in the pyrolysis of acetone, Fig. 8 demonstrates the sensitivities to a number of other reactions that determine the kinetic behavior of CH3 radicals, in particular, to reaction (5), the dissociation of the ethyl radical, and the chain branching reaction
After reaching this maximum, the concentration of CH3 slowly decreases to a non-zero value up to the time of sharp rise in the OH* emission. As soon as the OH* emission reaches the maximum, the CH3 concentration rapidly drops to zero, as at relatively high temperatures (Figs. 5 and 6). At the same time, calculations for all the DKMs under consideration [30,32–35] show a decrease in the CH3 concentration immediately after it reaches a maximum after a rapid rise behind the shock wave front. Thus, the patterns of the formation and consumption of methyl radicals predicted by the simulations differ significantly from those observed in experiment. At the same time, all the DKMs used [30,32–35] correctly predict the value of the CH3 concentration from which it begins to decrease rapidly as the OH* emission signal
Fig. 5. Comparison of the measured (solid) and calculated (with the DKMs from Refs. [32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission for the experimental conditions specified in Fig. 1. 5
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 6. Comparison of the measured (solid) and calculated (with DKMs from Refs. [30,32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission for T50 = 1620 K, P50 = 0.14 MPa.
Н+О2 = О+ОН
value of τ as compared to the consumption of the methyl radical itself. Fig. 9 shows that, in addition to (8), all other reactions have approximately equal effects on the duration of the induction period, much smaller than that of reaction (8). Note that the total effect of reactions (1)−(4) on the formation of the ground-state OH radical is even smaller than effects of all other reactions (Fig. 9). This led us to conclude that the uncertainty in the values of the rate constants of the reactions controlling the pyrolysis of acetone has a not so significant effect on the formation of the ground-state OH radical and, accordingly, the ignition delay time τ. Therefore, despite the fact that all the DKMs considered in this work do not accurately describe the yield of CH3 during the pyrolysis of acetone (Fig. 3), nevertheless, up to a factor of 2, they are able
(8)
It should be noted that, throughout the induction period, the sensitivities to reactions (1) and (4) remain high. However, for simulating the self-ignition of acetone, OH* luminescence is more informative. Fig. 9 shows the sensitivity to the most important pathways of its formation and consumption. Note that, for the ground-state OH radical, the effect of reaction (8) is significantly greater (Fig. 9), as are the effects of reactions of the interaction of H atoms with acetone and its pyrolysis products increases. An inspection of Figs. 8 and 9 suggests that the contribution of the reactions that determine the kinetics of CH3 has a less significant effect on the formation of the ground-state OH radical and thereby on the
Fig. 7. Comparison of the measured (solid) and calculated (with the DKMs from Refs. [30,32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission at T50 = 1410 K, P50 = 0.12 MPa. 6
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 8. Analysis of the local sensitivity of the main pathways of formation and consumption of the CH3 radical during the oxidation of a 0.5%(CH3)2CO−2%O2−Ar mixture. The experimental conditions correspond to Fig. 1.
to reproduce the τ value (Figs. 10 and 11). Figs. 10 and 11 show the temperature dependences of the ignition delay time for a 0.5%(CH3)2CO−2%O2−Ar mixture measured in the present work at a pressure of ~0.1 MPa and at 0.5 MPa in Ref. [12]. In the present work, the value of τ was determined from the time of maximum luminescence of OH* and CO2*, whereas in Ref. [12], τ was determined from the maximum of the OH* luminescence signal and the time interval between the arrival of the reflected shock wave and the point of intersection of the maximum-slope tangent to the ground-state
CO2 concentration time profile and the base line. Figs. 10 and 11 also show the temperature dependences of the values of τ calculated using the DKMs from Refs. [30,33,34,36]. In Ref. [12], the temperature dependence of τ was also described with the DKM from Ref. [36]. Although this DKM does not contain electronically excited species, it includes the reactions involved in the pyrolysis and oxidation of acetone, which makes it possible to describe the formation of ground-state CO2 molecules. Figs. 10 and 11 show that the DKMs from Refs. [30,33,34,36] closely reproduce the temperature dependence of τ. At
Fig. 9. Sensitivity analysis of the main pathways of the formation and consumption of OH radical during the oxidation of a 0.5%(CH3)2CO−2% O2−Ar mixture. The experimental conditions are specified in Fig. 1. 7
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 10. Temperature dependence of the ignition delay time. Triangles (OH*) and circles (CO2*) represent experimental data from the present work (~0.1 MPa) for a 0.5%(CH3)2CO−2% O2−Ar mixture.
Fig. 11. Temperature dependence of the ignition delay time for a 0.5%(CH3)2CO−2%O2−Ar mixture. Squares represent experimental data from Ref. [12] (0.5 MPa).
CH + O2 → OH* + CO,
T > 1600 K, the DKM proposed in Ref. [33] (Fig. 10) describes the ignition delay time determined from the maximum of OH* luminescence somewhat worse, which is caused not by defects of the DKM itself, but by an overestimated rate constant of the О + Н = ОН* reaction, which determines the steady-state level of OH* after the emission signal drops [39]. Since OH* and CO2 * are formed in reactions of CH with molecular oxygen [29,30],
the characteristics of the maximum and the first quasistationary level, as well as the rates of decay of the OH* and CO2* emission signals, are determined by the formation of CH radicals. The concentration of CH radicals is determined by that of CH3 radicals. Thus, it can be argued that the accuracy of the simulation of the OH* and CO2* luminescence signals is determined by the kinetics of CH3 radicals. Therefore, a further improvement in the description of the chemiluminescence of OH* and CO2* is directly related to the quality of the simulation of the
CH + O2 → CO2* + H, 8
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
behavior of methyl radicals. This requires additional experimental data on the time history of methyl radicals not only in the induction period, but also in the region of decrease of the CH3 concentration to zero (Fig. 1). These data are expected to provide new quantitative information on the rate constants of reactions (1)−(3), (6), which determine the OH* and СО2* emission profiles at all stages of the selfignition, combustion, and burnout of various hydrocarbon fuels.
[12] T. Tsuboi, K. Ishii, S. Tamura, Thermal oxidation of acetone behind reflected shock wave, Nihon Kikai Gakkai Ronbunshu B Hen Part B Trans. Jpn. Soc. Mech. Eng. 67 (2001) 2797–2804. [13] K. Sato, Y. Hidaka, Shock-tube and modeling study of acetone pyrolysis and oxidation, Combust. Flame 122 (2000) 291–311. [14] D. Yua, Zh.-Y. Tian, Z. Wang, Yu-X. Liu, L. Zhou, Experimental and theoretical study on acetone pyrolysis in a jet-stirred reactor, Fuel 234 (2018) 1380–1387. [15] S. Pichon, G. Black, N. Chaumeix, M. Yahyaoui, J.M. Simmie, J.H. Curran, R. Donohue, The combustion chemistry of a fuel tracer: measured flame speeds and ignition delays and a detailed chemical kinetic model for the oxidation of acetone, Combust. Flame 156 (2009) 494–504. [16] Ch.T. Chong, S. Hochgreb, Measurements of laminar flame speeds of acetone/methane/air mixtures, Combust. Flame 158 (2011) 490–500. [17] S. Saxena, J.H. Kiefer, S.J. Klippenstein, A shock-tube and theory study of the dissociation of acetone and subsequent recombination of methyl radicals, Proc. Combust. Inst. 32 (2009) 123–130. [18] A.M. Tereza, V.N. Smirnov, P.A. Vlasov, V.V. Shumova, A.A. Garmash, Emission of OH* and CO2* during the high-temperature oxidation of acetone in reflected shock waves, J. Phys.: Conf. Ser. 946 (2018) 012071. [19] K.-Y. Lam, Shock Tube Measurements of Oxygenated Fuel Combustion Using Laser Absorption Spectroscopy, Dissertation. Stanford University., 2013. [20] D.F. Davidson, S.C. Ranganath, K.-Y. Lam, M. Liaw, Z. Hong, R.K. Hanson, Ignition delay time measurements of normal alkanes and simple oxygenates, J. Propul. Power 26 (2010) 280–287. [21] H. Minwegen, U. Burke, K.A. Heufer, An experimental and theoretical comparison of C3-C5 linear ketones, Proc. Combust. Inst. 36 (2017) 561–568. [22] H. Matsui, M. Koshi, M. Oya, K. Tsuchiya, Improvement of chemical kinetic data at high temperatures by piston actuated shock tube, excimer laser photolysis, and atomic resonance absorption spectrometry, Shock Waves 3 (1994) 287–292. [23] B. Rotavera, E.L. Petersen, Model predictions of higher-order normal alkane ignition from dilute shock-tube experiments, Shock Waves 23 (2013) 345–359. [24] V.V. Leschevich, · V.V. Martynenko, O.G. Penyazkov, K.L. Sevrouk, S.I. Shabunya, · Auto-ignitions of a methane/air mixture at high and inter mediate temperatures, Shock Waves 26 (2016) 657–672. [25] A.M. Tulgestke, S.E. Johnson, D.F. Davidson, R.K. Hanson, High-speed imaging of inhomogeneous ignition in a shock tube, Shock Waves 28 (2018) 1089–1095. [26] P.A. Vlasov, V.N. Smirnov, A.M. Tereza, Reactions of initiation of the self-ignition of H2–O2 mixtures in shock waves, Russ. J. Phys. Chem. B. 10 (3) (2016) 456–468 https://DOI:10.1134/S1990793116030283. [27] B.V. L'vov, Atomic Absorption Spectrochemical Analysis, Hilger, London, 1970. [28] Yu.P. Petrov, Yu.K. Karasevich, S.V. Тuretskii, Decomposition of azomethane in shock tube as the example of “concerted” decomposition mechanism, 22nd ICDERS. Minsk. Belarus. 2009 Paper ICDERS2009-0133. [29] V.N. Smirnov, A.M. Tereza, P.A. Vlasov, I.V. Zhiltsova, Luminescent characteristics of the shock−wave ignition of an ethylene−oxygen mixture, Combust. Sci. Technol. 189 (2017) 854–868 http://old.chph.ras.ru/lab_0114.html. [30] A.M. Tereza, S.P. Medvedev, V.N. Smirnov, Experimental study and numerical simulation of chemiluminescence emission during the self-ignition of hydrocarbon fuels, Acta Astronaut. 163 (2019) 18–24, https://doi.org/10.1016/j.actaastro.2019. 03.001. [31] R.J. Kee, F.M. Rupley, E. Meeks, J.A. Miller, CHEMKIN III: Tech. Report No. SAND96–8216, (1996). [32] C.K. Westbrook, W.J. Pitz, O. Herbinet, H.J. Curran, E.J. Silke, A detailed chemical kinetic reaction mechanism for n-alkane hydrocarbons from n-octane to n-hexadecane, Combust. Flame 156 (2008) 181–199. [33] H. Wang, S.J. Warner, M.A. Oehlschlaeger, R. Bounaceur, J. Biet, P.A. Glaude, An experimental and kinetic modeling study of the self-ignition of α-methylnaphthalene/air and α-methylnaphthalene/n-decane/air mixtures at elevated pressures, Combust. Flame 157 (2010) 1976–1978 http://ignis.usc.edu/USC_Mech_II.htm. [34] W.K. Metcalfe, S.M. Burke, S.S. Ahmed, H.J. Curran, A hierarchical and comparative kinetic modeling study of C1−C2 hydrocarbon and oxygenated fuels, Int. J. Chem. Kinet. 45 (2013) 638–675 http://www.nuigalway.ie/c3/Mechanism_ release/frontmatter.html. [35] H. Wang, A.B. Dempsey, M. Yao, M. Jia, R.D. Reitz, Kinetic and numerical study on the effects of di-tert-butyl peroxide additive on the reactivity of methanol and ethanol, Energy Fuels 28 (8) (2014) 5480–5488. [36] E. Ranzi, A. Frassoldati, R. Grana, A. Cuoci, T. Faravelli, A.P. Kelley, C.K. Law, Hierarchical and comparative kinetic modeling of laminar flame speeds of hydrocarbon and oxygenated fuels, Prog. Energy Combust. Sci. 38 (4) (2012) 468–501. [37] M. Frenklach, W.C. Gardiner (Ed.), Modeling, Combustion Chemistry, SpringerVerlag., 1984, pp. 423–453. [38] W.C. Gardiner Jr., J. Troe, W.C. Gardiner (Ed.), Rate Coefficients of Thermal Dissociation, Isomerization, and Recombination Reactions, Combustion Chemistry, Springer-Verlag., 1984, pp. 173–196. [39] V.N. Smirnov, A.M. Tereza, P.A. Vlasov, I.V. Zhiltsova, A study of the chemiluminescence of CH*, OH*, C2*, and CO2* during the ignition of C2H2/O2/Ar mixture behind reflected shock waves, Proc. Int. Symp. Shock Waves. 31 (1) (2017) 159–167, https://doi.org/10.1007/978-3-319-91020-8_17.
5. Conclusions Thus, an experimental study of the self-ignition of a stoichiometric acetone−oxygen−argon mixture behind reflected shock waves in conjunction with numerical simulations showed the following. The published DKMs do not adequately describe the production of methyl radicals during acetone pyrolysis. At the same time, this fact does not substantially affect their ability to predict the temperature dependence of the ignition delay time for acetone−oxygen−argon mixtures. Simulations demonstrated that the reactions that control the pyrolysis of acetone influence only the initial stage of the induction period for the ignition of acetone−oxygen−argon mixtures. Nevertheless, to improve the quality of description of acetone self-ignition, it is necessary to refine the kinetic parameters of the interaction of methyl radicals with each other and with H atoms and OH radicals. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgment This work was performed under financial support from the Russian Science Foundation (Grant no. 19-19-00554). References [1] W. Viezee, J. Oblanas, Lidar-observed haze layers associated with thermal structure in the lower atmosphere, J. Appl. Meteorol. 8 (3) (1969) 369–375. [2] G. Silva, K. Iha, Hypergolic systems: a review in patents, J. Aerosp. Technol. Manag. 4 (4) (2012) 407–412, https://doi.org/10.5028/jatm.2012.04043812. [3] S. Matsumura, J.J. Rusek, Thermodynamic Design of an Alternative Monopropellant for Emergency Power Units, 23-rd Intern. Congress of Aeronautical Sciences, (2002) Paper ICAS 2002-8.3.4. [4] H.B. Singh, D. O'Hara, D. Herlth, W. Sachse, D.R. Blake, J.D. Bradshaw, M. Kanakidou, P.J. Crutzen, Acetone in the atmosphere: distribution, sources, and sinks, J. Geophys. Res. Atmos. 99 (1994) 1805–1819. [5] M. Yujing, A. Mellouki, The near-UV absorption cross sections for several ketones, J. Photoch. Photobiol. A Chem. 134 (2000) 31–36. [6] G.Yu. Grigoriev, A.S. Lagutin, Sh.Sh. Nabiev, A.A. Vasiliev, O.I. Orlov, L.N. Mukhamedieva, A.A. Pakhomova, A.V. Rodin, V.M. Semenov, D.B. Stavrovskii, M.G. Golubkov, Acta Astronaut. 163 (2019) 112–119. [7] J.-S. Chern, B. Wu, Y.-S. Chen, A.-M. Wu, Suborbital and low-thermospheric experiments using sounding rockets in Taiwan, Acta Astronaut. 70 (2012) 159–164, https://doi.org/10.1016/j.actaastro.2011.07.030. [8] N.N. Smirnov, V.B. Betelin, V.F. Nikitin, Yu.G. Phylippov, Koo Jaye, Detonation engine fed by acetylene–oxygen mixture, Acta Astronaut. 104 (2014) 34–146, https://doi.org/10.1016/j.actaastro.2014.07.019i. [9] N.N. Smirnov, O.G. Penyazkov, K.L. Sevrouk, V.F. Nikitin, L.I. Stamov, V.V. Tyurenkova, Onset of detonation in hydrogen-air mixtures due to shock wave reflection inside a combustion chamber, Acta Astronaut. 149 (2018) 77–92. [10] S.P. Medvedev, A.N. Polenov, B.E. Gelfand, Simulation of non-ideal explosions in a conical shock tube, in: R. Brun, L.Z. Dumitrescu (Eds.), Shock Waves @ Marseille IV, Proc. 19th Int. Symp. Shock Waves, Springer, Berlin, Heidelberg, 1995, pp. 381–386. [11] S.P. Medvedev, A.N. Polenov, S.V. Khomik, B.E. Gelfand, Initiation of upstreamdirected detonation induced by the venting of gaseous explosion, Proc. Int. Symp. Combust. 25 (1994) 73–78, https://doi.org/10.1016/S0082-0784(06)80630-2.
9
Contents lists available at ScienceDirect
Acta Astronautica journal homepage: www.elsevier.com/locate/actaastro
Self-ignition and pyrolysis of acetone behind reflected shock waves A.M. Tereza, S.P. Medvedev∗, V.N. Smirnov N.N. Semenov Federal Research Center for Chemical Physics, Russian Academy of Sciences, Russia
A R T I C LE I N FO
A B S T R A C T
Keywords: Self-ignition Pyrolysis Acetone Emission measurements Numerical simulation Chemical kinetics Shock waves Chemiluminescence
The self-ignition of a stoichiometric acetone−oxygen mixture diluted with argon was studied both experimentally and numerically. The experiments were performed behind reflected shock waves over a temperature range from 1280 to 1810 K at a total gas concentration of [M]50 ≈ 10−5 mol/cm3. The process was monitored by recording the signals of absorption by methyl radicals (λ = 216.5 nm) and emission from electronically excited OH* radicals (λ = 308 nm) and CO2* molecules (λ = 370 nm). Numerical simulations within the framework of various detailed kinetic mechanisms were performed to reproduce our own and published experimental data on the pyrolysis and self-ignition of acetone behind reflected shock waves, including the concentration profiles of acetone and CH3 in the ground state and electronically excited OH* and CO2*, as well as the temperature dependence of the self-ignition delay time. It has been established that various detailed kinetic mechanisms describe the kinetic characteristics of the pyrolysis of acetone in shock waves with varying degrees of accuracy. At the same time, all of them predicted the measured ignition delays within a factor of two. A sensitivity analysis showed that the reactions determining the pyrolysis of acetone produce only a slight effect on the branchedchain ignition process, so that the relevant rate constants only slightly affect the ignition delay time.
1. Introduction Studies of the kinetics of the decomposition and oxidation of acetone are motivated by its widespread use in various technological fields. As far back as the early 1960s, acetone was used as a component of liquid jet fuel for the Cricketsonde meteorological rocket [1]. Currently, patents on hypergolic systems involving acetone as a solvent in gels have been granted [2]. Based on the thermodynamic properties, the authors of [3], examined acetone as an alternative to existing rocket fuels. Since acetone is known to be formed in the atmosphere through the oxidation of various hydrocarbons [4,5], it is interesting to elucidate the role of acetone and other ketones in atmospheric chemistry, especially in the upper atmosphere, where it has been detected [6] in amounts capable of affecting the ozone layer [7]. The process of generating intense shock waves during the expansion of combustion products is of considerable interest for developing various detonation propulsion devices and ensuring safety of their operation [8,9]. A practically important case is the formation of spherical shock waves, with the conical shock tube (CST) being an effective tool for modeling such waves [10]. To optimize the operating parameters of the CST, it is advantageous to use a burning or detonating gas mixture as a driver medium. Acetone is one of the promising fuels for this purpose, because its mixtures with oxygen exhibit a high detonability
∗
even under conditions of rapid expansion [11]. Reliable information on the kinetics of its ignition at high temperatures is needed to develop a model of the combustion and detonation of acetone−oxygen−nitrogen mixtures in the driver section of the conical shock tube. Despite numerous studies on the pyrolysis of acetone and its selfignition in mixtures with oxygen and air, see, e.g., Refs. [12–21], a number of issues concerning the behavior of both the initial compound and various products of these processes remain unresolved [18–20]. The published detailed kinetic mechanisms (DKMs) are mainly aimed at modeling the temperature dependences of the ignition delays of various hydrocarbons in the presence of acetone additives [16,17]. At the same time, such data are of little use in simulating the time profiles of various products observed during the pyrolysis and self-ignition of acetone [19]. Therefore, in studying the kinetics of these processes, it is necessary to analyze various DKMs to select the most suitable for modification. In this regard, obtaining new kinetic data and constructing additional reaction blocks with corresponding rate constants can significantly improve the predictive capabilities of various DKMs [22]. On the other hand, the development of methods for recording ignition processes behind shock waves requires a comprehensive approach to interpreting experimental data by means of numerical simulations. An exhaustive DKM should include electronically excited species, a feature that makes it possible to simulate the time histories of
Corresponding author. E-mail addresses: [email protected] (A.M. Tereza), [email protected] (S.P. Medvedev), [email protected] (V.N. Smirnov).
https://doi.org/10.1016/j.actaastro.2020.03.045 Received 25 February 2020; Received in revised form 26 March 2020; Accepted 28 March 2020 0094-5765/ © 2020 IAA. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: A.M. Tereza, S.P. Medvedev and V.N. Smirnov, Acta Astronautica, https://doi.org/10.1016/j.actaastro.2020.03.045
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 1. Experimentally measured (solid lines) CH3 absorption, OH* emission, and pressure P time profiles at T50 = 1510 K and P50 = 0.126 MPa. The dashed and dash-dot lines represent the simulation results by the DKM from Ref. [30].
behind incident and reflected shock waves. The endwall was located 1.1 cm downstream of the windows. Shock waves were generated by the spontaneous rupture of one or a stack of 0.005-cm-thick aluminum diaphragms, depending on the conditions required. The driver gas was helium. Since a diaphragm of a given thickness ruptures at about the same driver gas pressure, in experiments in which it was necessary to vary the temperature behind the shock wave (incident or reflected) while maintaining the shock-compressed gas density at approximately constant level, the shock wave was generated by helium−air mixtures of different compositions. All other things equal, the higher the air content in the driver mixture, the lower the temperature behind the shock wave. We recorded the chemiluminescence signals from electronically excited OH* radicals and CO2* molecules and the absorption time profile of CH3 radicals. The time histories of the emissions from OH* and CO2* were monitored at wavelength of λ = 308 ± 2.5 nm and λ = 363 ± 3.5 nm, respectively. The signal from OH* was detected using a DMR-4 two-prism quartz monochromator in conjunction with a FEU-39A photomultiplier, whereas the emission from CO2* was directed into a FEU-52 photomultiplier through a combined filter composed of an interference filter and UF-2 (ultraviolet) and SZS-21 (blue green) 0.3-cm-thick optical glasses. The absorption of CH3 radicals was recorded at a wavelength of λ = 216.5 nm using a DMR-4 monochromator in combination with a FEU-39A photomultiplier. The light source was a high-frequency electrodeless copper lamp powered by a PPBL-3V generator. Such a lamp gives a discrete spectrum of the glow of copper atoms [27]. The concentration of CH3 radicals was determined according to the Beer−Lambert−Bouguer law:
their concentrations based on the corresponding chemiluminescence signals. Since emission measurements are highly sensitive and unobtrusive, they are widely used in studies on the self-ignition of fuel−oxidizer mixtures behind shock waves [23–25]. The aim of the present work is to study the kinetics of the pyrolysis and self-ignition of acetone−oxygen−argon mixtures behind reflected shock waves by performing numerical simulations within the framework of various DKMs and comparing the results with the available experimental data on the time profiles of CH3 absorption and the emission of electronically excited OH* and CO2*. 2. Experimental The experimental setup was described in detail elsewhere [26] with only some general features given below. The experiments were performed in a stainless steel shock tube with an inner diameter of 7.5 cm, driver section length of 150 cm, and driven section length of 320 cm. The driven section was evacuated to 10−2 Torr with two 2-NVR-5D forepumps and then to 10−3 Torr with a NVDS-100 diffusion pump. The vacuum/gas-handling system had two liquid-nitrogen-cooled traps. The degree of evacuation was controlled with a VIT-2 ionization-thermocouple vacuum gauge (here and below, the capitalized abbreviations are the transliterations from Russian of the types of instruments and devices). The driver section was evacuated with a 2-NVR-5D forepump through a liquid-nitrogen-cooled trap to a pressure of 0.1 Torr. The rate of leakage into the driven section did not exceed ~10−4 Torr/min. Before each experiment, the driven section was twice purged with argon used to prepare test mixtures with intermediate evacuations to 10−1 Torr. The total density and temperature of the gas behind the incident and reflected shock waves were calculated from the composition of the test mixture, initial pressure, and velocity of the incident shock wave by using ideal shock tube theory. The incident shock wave velocity was measured over two measurement intervals formed by three sequentially flush-mounted pressure gauges G1, G2, and G3. The distances between the gauges were G1−G2 = 52.8 cm and G2−G3 = 28.1 cm, with the endwall-nearest gauge G3 mounted 4.0 cm upstream of the observation windows. The fourth pressure sensor, 0.7 cm in diameter, located in the same cross section as the observation windows, was intended to record the pressure profiles
[CH3] = (1/ σ l)ln(I0/ I ), where I0 and I are the initial (before the arrival of the shock wave) and current signal intensity, l = 7.5 cm is the internal diameter of the shock tube, and σ is the absorption coefficient. The absorption coefficient was taken from [28],
σ = 3.25 107 T −0.36 exp(500/T )exp(−T /2500) cm2 /mol.
3. Experimental results The self-ignition of a 0.5%(CH3)2CO−2%O2−Ar stoichiometric 2
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 2. Experimental (solid lines) and simulated (interrupted lines) time profiles of electronically excited ОН* radicals and CO2* molecules. T50 = 1638 K, P50 = 0.133 MPa. The dashed and dash-dot lines represent the simulation results by the DKM from Ref. [30].
the DKMs describe the pyrolysis of acetone behind reflected shock waves. For this, the experimental results on measuring the yield of CH3 radicals during the pyrolysis of a 0.25% (CH3)2CO−Ar mixture [19] were considered. Fig. 3 compares the experimental data from Refs. [19] and the DKM calculations from Refs. [30,32–35] at three different temperatures. As can be seen from Fig. 3a, at T50 = 1469 K (Fig. 3a), the profiles calculated using the DKMs from Refs. [33,34] significantly differ both in the time of reaching the maximum concentration and in the rate of consumption of CH3 radicals. At the same time, the time profiles simulated using the DKM from Ref. [32] predict well the maximum concentration of CH3, but show a very rapid decrease in the concentration of methyl radicals, resulting in its value at late times (after 600 μs) approximately twofold less than that observed in the experiment. A similar rate of CH3 absorption decay is predicted by the DKM given in Ref. [33], which enables it to describe well the level of CH3 concentration at late pyrolysis times, despite a discrepancy between the calculated and measured concentrations of CH3 radicals. Calculations using the DKM from Ref. [35] and the extended mechanism from Ref. [30], although predicting different decay rates after reaching the maximum CH3 concentration, nevertheless, closely reproduce the concentration of CH3 in late stages of the pyrolysis. At lower temperatures (Fig. 3b), the description of the CH3 concentration time profile within the framework of the DKMs from Refs. [32–35] is slightly worse compared to the predictions of the DKM from Ref. [30]. This DKM, although it does not quite accurately describe the maximum CH3 concentration, but rather well reproduces the decay rate and final concentration of methyl radicals. With a further decrease in temperature (Fig. 3c), all the examined DKMs describe the temporal profiles of the concentration of CH3 measured in Ref. [19] only qualitatively, probably due to a higher error in the measurements of the absolute concentration of methyl radicals at low temperatures. To find out why the different DKMs produce different results, we analyzed the sensitivity of the leading stages of acetone pyrolysis for the conditions specified in the caption of Fig. 3a. Since the DKMs used in this work contain nearly the same blocks of leading reactions, which differ only in the corresponding rate constants, the sensitivity analysis was carried out using only the DKM from Ref. [30]. The time dependence of the local sensitivity d(ln[CH3])/d(lnki) [37] is
mixture was carried out behind reflected shock waves in the temperature range of 1280–1720 K and a total gas concentration of [M]50≈10−5 mol/cm3 (hereinafter, the subscript 50 denotes the initial gas conditions behind the reflected shock wave). Fig. 1 shows typical OH* emission and CH3 absorption time profiles. The self-ignition delay τ was defined as the time interval between the arrival of the reflected shock wave and the time of maximum OH* emission intensity (Fig. 1). As can be seen from Fig. 1, the concentration of СН3 radicals passes through a maximum, reaches a slowly declining plateau, and then decreases rapidly to a nearly zero level at the time the OH* luminescence signal passes through its maximum. This suggests that, by the time the (CH3)2CO−O2−Ar mixture self-ignites, most of the CH3 radicals formed during the decomposition of acetone have reacted. As can be seen, no OH* emission is observed nearly throughout the induction period−only a slight increase in the OH* emission is observed immediately before the ignition. Fig. 2 displays OH* and CO2* emission time profiles typical of hightemperature experiments. It can be seen that the maxima of the OH* and СО2* luminescence signals practically coincide, which was also observed for the self-ignition of stoichiometric mixtures of other hydrocarbons [29,30]. Note also that, at high temperatures, a pre-peak step is observed (Fig. 2), which is absent at lower temperatures (Fig. 1). Figs. 1 and 2 show that, after reaching a maximum, the OH* and CO2* signals decrease sharply to a slowly declining level. The experimentally recorded time profiles of CH3 absorption and OH* and CO2* emissions were numerically simulated using various detailed kinetic mechanisms. 4. Numerical simulations All the simulations were performed using the CHEMKIN III software package [31] under the condition of constant volume (V = const). The DKMs were taken from Refs. [30] (ChemphysMech 2.0), [32] (LLNL 2.0), [33] (JetSurf 2.0), [34] (Aramco 1.3), [35] (ERC), and [36] (CRECK), with the corresponding NASA thermodynamic polynomials given in these studies. The choice of these DKMs was motivated by the presence of reactions involving acetone in them. Before simulating our own experimental results, we examined how 3
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
C2H6+H = C2H5+H2
(3)
(СН3)2СО = CH3CO + CH3
(4)
In this case, the primary decomposition reaction of acetone (4) rapidly shifts to the left after ~200 μs; i.e., when the starting acetone is nearly completely consumed. A further rapid decomposition of CH3CO leads to the formation of a second CH3 radical, which participates in the recombination (1) and disproportionation (2) reactions. Clearly, atomic hydrogen can contribute to the formation and consumption of CH3 radicals: (СН3)2СО+Н = CH3COСН2+H2 → CH3+COСН2+H2
(5)
СН3+Н(+М) = СН4(+М)
(6)
Note, however, that the estimates made by the brute force method [37] showed that the influence of reactions (5) and (6) on the profiles in Fig. 3 is insignificant. In the considered temperature range, the second fragment of the dissociation of the acetone molecule, the CH3CO radical, decays quite quickly, CH3CO = СН3+СО,
forming another CH3 radical and a CO molecule. Reaction (7) proceeds so quickly that it could be combined with reaction (4). It can be expected that a reliable determination of the rate constants of reactions (1)−(4) will provide a fairly good description of the CH3 profiles in Fig. 3. It should be noted that substantial difficulties are encountered in determining the rate constants of the unimolecular dissociation reactions, (1), (4), and (6). These reactions occur in a falloff pressure region, with their rate constants being dependent on a much larger number of parameters than those for bimolecular reactions (2) and (3), expressed in a three-parameter Arrhenius form [38]. In the present study, we did not refine the rate constants of reactions (1)−(4), (5) and (6) given in Refs. [13,30]. The time histories of CH3 absorption and OH* luminescence for the self-ignition of acetone−oxygen–argon mixtures calculated using various DKMs is displayed in Figs. 1, 5–7, which demonstrate that all the DKMs give approximately same time profiles of the formation and consumption of CH3 radicals. For relatively high temperatures (Figs. 1, 5, and 6), all the DKMs predict a fast formation of CH3 radicals, as observed in experiment. After passing the maximum, the CH3 concentration decreases to the moment of self-ignition, corresponding to the maximum of the OH* emission signal. This is clearly demonstrated by calculations based on the DKMs from Refs. [30,34]. The fact that the OH* emission time profile calculated by the DKM from Ref. [33] (Fig. 5) differs markedly can be explained solely by the ratio between the contribution from the different reactions of OH* excitation at the early and late stages of the self-ignition of acetone, but not by any drawbacks of the DKM proposed in Ref. [33]. An inspection of the times of occurrence of the OH* emission maximum and the sharp drop in the СН3 concentration to zero predicted by the DKM from Ref. [33] show that they are nearly identical, as in the case of the DKMs from Ref. [30,34]. The behavior of OH* emission profiles calculated by the DKM reported in Ref. [33] for the ignition of hydrocarbons was analyzed in Ref. [30]. Since we did not examine the OH* emission profiles calculated by the DKM from Ref. [33] in detail, they are not represented in Figs. 6 and 7. The CH3 profiles calculated by the DKMs from Refs. [32, 35] behave very similar to those simulated using the DKMs reported in [30, 33, 34]. Since the DKMs from [32, 35] include no reaction blocks involving OH*, Figs. 5-7 show only the CH3 concentration profiles calculated by these DKMs. While the DKMs from Refs. [30,32–35] satisfactorily reproduce the measured time profiles of CH3 absorption at relatively high temperatures (Figs. 1, 5, and 6), this is not the case at low temperatures (Fig. 7). After a sharp initial rise, the measured CH3 concentration continues to increase at a slower rate, reaching a maximum at a time equal to approximately one third of the observed ignition delay (Fig. 7).
Fig. 3. Comparison of experimental [19] and DKM-calculated СН3 profiles for a 0.25% acetone−argon mixture. Experimental condition: (а) T50 = 1469 K, Р50 = 0.148 MPa; (b) T50 = 1393 K, Р50 = 0.155 MPa; and (c) T50 = 1273 K, Р50 = 0.165 MPa. The solid line represents the experimental data, and the interrupted lines are the corresponding calculated profiles by the DKMs from Refs. [30,32–35].
displayed in Fig. 4. Fig. 4 shows the main pathways for the formation and consumption of CH3 radicals during acetone pyrolysis: CH3+CH3(+M) = C2H6(+M)
(1)
CH3+CH3 = C2H5+H
(2)
(7)
4
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 4. Analysis of the local sensitivity of the main pathways of the formation and consumption of CH3 radicals during the pyrolysis of a 0.25%(CH3)2CO−Ar mixture. The experimental conditions are specified in the caption of Fig. 3 (panel a).
approaches and passes through its maximum, indicative of the occurrence of self-ignition (Figs. 1, 5–7). Moreover, even if the description of the CH3 absorption time profile is not satisfactory (Fig. 7), this nonzero level of the CH3 concentration is much the same as the experimentally observed. The results of a sensitivity analysis for the experimental conditions specified in Fig. 1 are displayed in Figs. 8 and 9. In contrast to Fig. 4, which shows the sensitivity to the reactions involved in the pyrolysis of acetone, Fig. 8 demonstrates the sensitivities to a number of other reactions that determine the kinetic behavior of CH3 radicals, in particular, to reaction (5), the dissociation of the ethyl radical, and the chain branching reaction
After reaching this maximum, the concentration of CH3 slowly decreases to a non-zero value up to the time of sharp rise in the OH* emission. As soon as the OH* emission reaches the maximum, the CH3 concentration rapidly drops to zero, as at relatively high temperatures (Figs. 5 and 6). At the same time, calculations for all the DKMs under consideration [30,32–35] show a decrease in the CH3 concentration immediately after it reaches a maximum after a rapid rise behind the shock wave front. Thus, the patterns of the formation and consumption of methyl radicals predicted by the simulations differ significantly from those observed in experiment. At the same time, all the DKMs used [30,32–35] correctly predict the value of the CH3 concentration from which it begins to decrease rapidly as the OH* emission signal
Fig. 5. Comparison of the measured (solid) and calculated (with the DKMs from Refs. [32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission for the experimental conditions specified in Fig. 1. 5
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 6. Comparison of the measured (solid) and calculated (with DKMs from Refs. [30,32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission for T50 = 1620 K, P50 = 0.14 MPa.
Н+О2 = О+ОН
value of τ as compared to the consumption of the methyl radical itself. Fig. 9 shows that, in addition to (8), all other reactions have approximately equal effects on the duration of the induction period, much smaller than that of reaction (8). Note that the total effect of reactions (1)−(4) on the formation of the ground-state OH radical is even smaller than effects of all other reactions (Fig. 9). This led us to conclude that the uncertainty in the values of the rate constants of the reactions controlling the pyrolysis of acetone has a not so significant effect on the formation of the ground-state OH radical and, accordingly, the ignition delay time τ. Therefore, despite the fact that all the DKMs considered in this work do not accurately describe the yield of CH3 during the pyrolysis of acetone (Fig. 3), nevertheless, up to a factor of 2, they are able
(8)
It should be noted that, throughout the induction period, the sensitivities to reactions (1) and (4) remain high. However, for simulating the self-ignition of acetone, OH* luminescence is more informative. Fig. 9 shows the sensitivity to the most important pathways of its formation and consumption. Note that, for the ground-state OH radical, the effect of reaction (8) is significantly greater (Fig. 9), as are the effects of reactions of the interaction of H atoms with acetone and its pyrolysis products increases. An inspection of Figs. 8 and 9 suggests that the contribution of the reactions that determine the kinetics of CH3 has a less significant effect on the formation of the ground-state OH radical and thereby on the
Fig. 7. Comparison of the measured (solid) and calculated (with the DKMs from Refs. [30,32–35], dashed and dotted lines) time profile of CH3 absorption and OH* emission at T50 = 1410 K, P50 = 0.12 MPa. 6
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 8. Analysis of the local sensitivity of the main pathways of formation and consumption of the CH3 radical during the oxidation of a 0.5%(CH3)2CO−2%O2−Ar mixture. The experimental conditions correspond to Fig. 1.
to reproduce the τ value (Figs. 10 and 11). Figs. 10 and 11 show the temperature dependences of the ignition delay time for a 0.5%(CH3)2CO−2%O2−Ar mixture measured in the present work at a pressure of ~0.1 MPa and at 0.5 MPa in Ref. [12]. In the present work, the value of τ was determined from the time of maximum luminescence of OH* and CO2*, whereas in Ref. [12], τ was determined from the maximum of the OH* luminescence signal and the time interval between the arrival of the reflected shock wave and the point of intersection of the maximum-slope tangent to the ground-state
CO2 concentration time profile and the base line. Figs. 10 and 11 also show the temperature dependences of the values of τ calculated using the DKMs from Refs. [30,33,34,36]. In Ref. [12], the temperature dependence of τ was also described with the DKM from Ref. [36]. Although this DKM does not contain electronically excited species, it includes the reactions involved in the pyrolysis and oxidation of acetone, which makes it possible to describe the formation of ground-state CO2 molecules. Figs. 10 and 11 show that the DKMs from Refs. [30,33,34,36] closely reproduce the temperature dependence of τ. At
Fig. 9. Sensitivity analysis of the main pathways of the formation and consumption of OH radical during the oxidation of a 0.5%(CH3)2CO−2% O2−Ar mixture. The experimental conditions are specified in Fig. 1. 7
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
Fig. 10. Temperature dependence of the ignition delay time. Triangles (OH*) and circles (CO2*) represent experimental data from the present work (~0.1 MPa) for a 0.5%(CH3)2CO−2% O2−Ar mixture.
Fig. 11. Temperature dependence of the ignition delay time for a 0.5%(CH3)2CO−2%O2−Ar mixture. Squares represent experimental data from Ref. [12] (0.5 MPa).
CH + O2 → OH* + CO,
T > 1600 K, the DKM proposed in Ref. [33] (Fig. 10) describes the ignition delay time determined from the maximum of OH* luminescence somewhat worse, which is caused not by defects of the DKM itself, but by an overestimated rate constant of the О + Н = ОН* reaction, which determines the steady-state level of OH* after the emission signal drops [39]. Since OH* and CO2 * are formed in reactions of CH with molecular oxygen [29,30],
the characteristics of the maximum and the first quasistationary level, as well as the rates of decay of the OH* and CO2* emission signals, are determined by the formation of CH radicals. The concentration of CH radicals is determined by that of CH3 radicals. Thus, it can be argued that the accuracy of the simulation of the OH* and CO2* luminescence signals is determined by the kinetics of CH3 radicals. Therefore, a further improvement in the description of the chemiluminescence of OH* and CO2* is directly related to the quality of the simulation of the
CH + O2 → CO2* + H, 8
Acta Astronautica xxx (xxxx) xxx–xxx
A.M. Tereza, et al.
behavior of methyl radicals. This requires additional experimental data on the time history of methyl radicals not only in the induction period, but also in the region of decrease of the CH3 concentration to zero (Fig. 1). These data are expected to provide new quantitative information on the rate constants of reactions (1)−(3), (6), which determine the OH* and СО2* emission profiles at all stages of the selfignition, combustion, and burnout of various hydrocarbon fuels.
[12] T. Tsuboi, K. Ishii, S. Tamura, Thermal oxidation of acetone behind reflected shock wave, Nihon Kikai Gakkai Ronbunshu B Hen Part B Trans. Jpn. Soc. Mech. Eng. 67 (2001) 2797–2804. [13] K. Sato, Y. Hidaka, Shock-tube and modeling study of acetone pyrolysis and oxidation, Combust. Flame 122 (2000) 291–311. [14] D. Yua, Zh.-Y. Tian, Z. Wang, Yu-X. Liu, L. Zhou, Experimental and theoretical study on acetone pyrolysis in a jet-stirred reactor, Fuel 234 (2018) 1380–1387. [15] S. Pichon, G. Black, N. Chaumeix, M. Yahyaoui, J.M. Simmie, J.H. Curran, R. Donohue, The combustion chemistry of a fuel tracer: measured flame speeds and ignition delays and a detailed chemical kinetic model for the oxidation of acetone, Combust. Flame 156 (2009) 494–504. [16] Ch.T. Chong, S. Hochgreb, Measurements of laminar flame speeds of acetone/methane/air mixtures, Combust. Flame 158 (2011) 490–500. [17] S. Saxena, J.H. Kiefer, S.J. Klippenstein, A shock-tube and theory study of the dissociation of acetone and subsequent recombination of methyl radicals, Proc. Combust. Inst. 32 (2009) 123–130. [18] A.M. Tereza, V.N. Smirnov, P.A. Vlasov, V.V. Shumova, A.A. Garmash, Emission of OH* and CO2* during the high-temperature oxidation of acetone in reflected shock waves, J. Phys.: Conf. Ser. 946 (2018) 012071. [19] K.-Y. Lam, Shock Tube Measurements of Oxygenated Fuel Combustion Using Laser Absorption Spectroscopy, Dissertation. Stanford University., 2013. [20] D.F. Davidson, S.C. Ranganath, K.-Y. Lam, M. Liaw, Z. Hong, R.K. Hanson, Ignition delay time measurements of normal alkanes and simple oxygenates, J. Propul. Power 26 (2010) 280–287. [21] H. Minwegen, U. Burke, K.A. Heufer, An experimental and theoretical comparison of C3-C5 linear ketones, Proc. Combust. Inst. 36 (2017) 561–568. [22] H. Matsui, M. Koshi, M. Oya, K. Tsuchiya, Improvement of chemical kinetic data at high temperatures by piston actuated shock tube, excimer laser photolysis, and atomic resonance absorption spectrometry, Shock Waves 3 (1994) 287–292. [23] B. Rotavera, E.L. Petersen, Model predictions of higher-order normal alkane ignition from dilute shock-tube experiments, Shock Waves 23 (2013) 345–359. [24] V.V. Leschevich, · V.V. Martynenko, O.G. Penyazkov, K.L. Sevrouk, S.I. Shabunya, · Auto-ignitions of a methane/air mixture at high and inter mediate temperatures, Shock Waves 26 (2016) 657–672. [25] A.M. Tulgestke, S.E. Johnson, D.F. Davidson, R.K. Hanson, High-speed imaging of inhomogeneous ignition in a shock tube, Shock Waves 28 (2018) 1089–1095. [26] P.A. Vlasov, V.N. Smirnov, A.M. Tereza, Reactions of initiation of the self-ignition of H2–O2 mixtures in shock waves, Russ. J. Phys. Chem. B. 10 (3) (2016) 456–468 https://DOI:10.1134/S1990793116030283. [27] B.V. L'vov, Atomic Absorption Spectrochemical Analysis, Hilger, London, 1970. [28] Yu.P. Petrov, Yu.K. Karasevich, S.V. Тuretskii, Decomposition of azomethane in shock tube as the example of “concerted” decomposition mechanism, 22nd ICDERS. Minsk. Belarus. 2009 Paper ICDERS2009-0133. [29] V.N. Smirnov, A.M. Tereza, P.A. Vlasov, I.V. Zhiltsova, Luminescent characteristics of the shock−wave ignition of an ethylene−oxygen mixture, Combust. Sci. Technol. 189 (2017) 854–868 http://old.chph.ras.ru/lab_0114.html. [30] A.M. Tereza, S.P. Medvedev, V.N. Smirnov, Experimental study and numerical simulation of chemiluminescence emission during the self-ignition of hydrocarbon fuels, Acta Astronaut. 163 (2019) 18–24, https://doi.org/10.1016/j.actaastro.2019. 03.001. [31] R.J. Kee, F.M. Rupley, E. Meeks, J.A. Miller, CHEMKIN III: Tech. Report No. SAND96–8216, (1996). [32] C.K. Westbrook, W.J. Pitz, O. Herbinet, H.J. Curran, E.J. Silke, A detailed chemical kinetic reaction mechanism for n-alkane hydrocarbons from n-octane to n-hexadecane, Combust. Flame 156 (2008) 181–199. [33] H. Wang, S.J. Warner, M.A. Oehlschlaeger, R. Bounaceur, J. Biet, P.A. Glaude, An experimental and kinetic modeling study of the self-ignition of α-methylnaphthalene/air and α-methylnaphthalene/n-decane/air mixtures at elevated pressures, Combust. Flame 157 (2010) 1976–1978 http://ignis.usc.edu/USC_Mech_II.htm. [34] W.K. Metcalfe, S.M. Burke, S.S. Ahmed, H.J. Curran, A hierarchical and comparative kinetic modeling study of C1−C2 hydrocarbon and oxygenated fuels, Int. J. Chem. Kinet. 45 (2013) 638–675 http://www.nuigalway.ie/c3/Mechanism_ release/frontmatter.html. [35] H. Wang, A.B. Dempsey, M. Yao, M. Jia, R.D. Reitz, Kinetic and numerical study on the effects of di-tert-butyl peroxide additive on the reactivity of methanol and ethanol, Energy Fuels 28 (8) (2014) 5480–5488. [36] E. Ranzi, A. Frassoldati, R. Grana, A. Cuoci, T. Faravelli, A.P. Kelley, C.K. Law, Hierarchical and comparative kinetic modeling of laminar flame speeds of hydrocarbon and oxygenated fuels, Prog. Energy Combust. Sci. 38 (4) (2012) 468–501. [37] M. Frenklach, W.C. Gardiner (Ed.), Modeling, Combustion Chemistry, SpringerVerlag., 1984, pp. 423–453. [38] W.C. Gardiner Jr., J. Troe, W.C. Gardiner (Ed.), Rate Coefficients of Thermal Dissociation, Isomerization, and Recombination Reactions, Combustion Chemistry, Springer-Verlag., 1984, pp. 173–196. [39] V.N. Smirnov, A.M. Tereza, P.A. Vlasov, I.V. Zhiltsova, A study of the chemiluminescence of CH*, OH*, C2*, and CO2* during the ignition of C2H2/O2/Ar mixture behind reflected shock waves, Proc. Int. Symp. Shock Waves. 31 (1) (2017) 159–167, https://doi.org/10.1007/978-3-319-91020-8_17.
5. Conclusions Thus, an experimental study of the self-ignition of a stoichiometric acetone−oxygen−argon mixture behind reflected shock waves in conjunction with numerical simulations showed the following. The published DKMs do not adequately describe the production of methyl radicals during acetone pyrolysis. At the same time, this fact does not substantially affect their ability to predict the temperature dependence of the ignition delay time for acetone−oxygen−argon mixtures. Simulations demonstrated that the reactions that control the pyrolysis of acetone influence only the initial stage of the induction period for the ignition of acetone−oxygen−argon mixtures. Nevertheless, to improve the quality of description of acetone self-ignition, it is necessary to refine the kinetic parameters of the interaction of methyl radicals with each other and with H atoms and OH radicals. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgment This work was performed under financial support from the Russian Science Foundation (Grant no. 19-19-00554). References [1] W. Viezee, J. Oblanas, Lidar-observed haze layers associated with thermal structure in the lower atmosphere, J. Appl. Meteorol. 8 (3) (1969) 369–375. [2] G. Silva, K. Iha, Hypergolic systems: a review in patents, J. Aerosp. Technol. Manag. 4 (4) (2012) 407–412, https://doi.org/10.5028/jatm.2012.04043812. [3] S. Matsumura, J.J. Rusek, Thermodynamic Design of an Alternative Monopropellant for Emergency Power Units, 23-rd Intern. Congress of Aeronautical Sciences, (2002) Paper ICAS 2002-8.3.4. [4] H.B. Singh, D. O'Hara, D. Herlth, W. Sachse, D.R. Blake, J.D. Bradshaw, M. Kanakidou, P.J. Crutzen, Acetone in the atmosphere: distribution, sources, and sinks, J. Geophys. Res. Atmos. 99 (1994) 1805–1819. [5] M. Yujing, A. Mellouki, The near-UV absorption cross sections for several ketones, J. Photoch. Photobiol. A Chem. 134 (2000) 31–36. [6] G.Yu. Grigoriev, A.S. Lagutin, Sh.Sh. Nabiev, A.A. Vasiliev, O.I. Orlov, L.N. Mukhamedieva, A.A. Pakhomova, A.V. Rodin, V.M. Semenov, D.B. Stavrovskii, M.G. Golubkov, Acta Astronaut. 163 (2019) 112–119. [7] J.-S. Chern, B. Wu, Y.-S. Chen, A.-M. Wu, Suborbital and low-thermospheric experiments using sounding rockets in Taiwan, Acta Astronaut. 70 (2012) 159–164, https://doi.org/10.1016/j.actaastro.2011.07.030. [8] N.N. Smirnov, V.B. Betelin, V.F. Nikitin, Yu.G. Phylippov, Koo Jaye, Detonation engine fed by acetylene–oxygen mixture, Acta Astronaut. 104 (2014) 34–146, https://doi.org/10.1016/j.actaastro.2014.07.019i. [9] N.N. Smirnov, O.G. Penyazkov, K.L. Sevrouk, V.F. Nikitin, L.I. Stamov, V.V. Tyurenkova, Onset of detonation in hydrogen-air mixtures due to shock wave reflection inside a combustion chamber, Acta Astronaut. 149 (2018) 77–92. [10] S.P. Medvedev, A.N. Polenov, B.E. Gelfand, Simulation of non-ideal explosions in a conical shock tube, in: R. Brun, L.Z. Dumitrescu (Eds.), Shock Waves @ Marseille IV, Proc. 19th Int. Symp. Shock Waves, Springer, Berlin, Heidelberg, 1995, pp. 381–386. [11] S.P. Medvedev, A.N. Polenov, S.V. Khomik, B.E. Gelfand, Initiation of upstreamdirected detonation induced by the venting of gaseous explosion, Proc. Int. Symp. Combust. 25 (1994) 73–78, https://doi.org/10.1016/S0082-0784(06)80630-2.
9