Studies on the flame propagation characteristic and thermal hazard of the premixed N2O/fuel mixtures

Yu-yn Li , Rong-pei Jing , Zhi-peng Li , Sen Xu , Feng Pn ,*, Li-feng Xie

a School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing, Jiangsu 210094, PR China

b Beijing Insitute of Aerospace Testing Technology, Beijing Key Laboratory of Research and Application for Aerospace Green Propellants,Beijing 100074,PR China

c National Quality Supervision and Inspection Center for Industrial Explosive Materials, Nanjing, Jiangsu 210094, PR China

Keywords:Flame propagation Flame acceleration rate Quenching diameter Thermal hazard

ABSTRACT An experimental study was carried out to investigate the flame propagation and thermal hazard of the premixed N2O/fuel mixtures, including NH3, C3H8 and C2H4. The study provided the high speed video images and data about the flame locations, propagation patterns, overpressures and the quenching diameters during the course of combustion in different channels to elucidate the dynamics of various combustion processes. The onset decomposition temperature was determined using high-performance adiabatic calorimetry. It was shown that the order of the flame acceleration rate and thermal hazard was N2O/C2H4>N2O/C3H8>N2O/NH3.

1. Introduction

Due to the high toxicity of hydrazine and the request for safer and cheaper propellants with comparable performance, one promising class of green propellants, so-called nitrous oxide fuel blends (NOFB) has attracted a lot of attention [1,2]. The blends generally consist of nitrous oxide and different hydrocarbons (e.g.,C2H2,C2H4,C3H8and NH3).Nitrous oxide is a non-toxic chemical,it is stable at normal conditions and compatible with common structure materials.It can be stored as a liquid or a compressed gas through a wide temperature range[3,4].

There are a few experimental studies on the N2O/fuel flames,including examinations of the ignition delay time of CH4/N2O/Ar mixture and small hydrocarbon/N2O mixtures with and without O2[5-7], the flame structure of NH3/H2/N2O/Ar [8], and the laminar burning velocity of H2/N2O/N2, H2/N2O/Ar and hydrocarbon/N2O mixtures [9-12]. Powell et al. [13,14] have complied the chemical mechanism for modeling the combustion of H2/N2O, CH4/N2O,C3H6/N2O and C3H8/N2O mixtures. Venkatesh et al. [15]investigated the high pressure combustion characteristics of a stoichiometric mixture of C2H4and N2O in an alloy steel vessel.Grubelich et al. [16] studied the flame acceleration (FA) and deflagration to detonation transition(DDT)behavior of a stoichiometric mixture of C2H4/N2O at various initial pressures. Zhang et al. [17]carried out a detailed experimental investigation on the detonation propagation behavior of the stoichiometric mixture of C2H4/N2O in narrow channels.The pressure,temperature and high-speed video data of N2O/C2H4during the flame propagation process were collected by DLR(German Aerospace Center)[18].N2O/C3H8(NOP)rocket engine was tested over a range of mixture ratios by Tiliakos et al. [19], experimental results match well with theoretical predictions,with proper modeling of heat losses.Movileanu et al.[20]measured maximum explosion pressures, explosion times and maximum rates of pressure rise for lean and stoichiometric ethylene-nitrous oxide mixtures diluted with 60% N2, at initial pressures within 50-150 kPa, in cylindrical vessels with different aspect ratio. However, the fundamental combustion and ignition data of N2O/fuel, such as the flame acceleration, pressure rise,quenching tube diameter, and so on, are still limited.

This study aims to investigate the flame propagation and thermal hazard of the premixed N2O/NH3, N2O/C3H8and N2O/C2H4mixtures experimentally,and to obtain a general flame dynamics in various scale channels with different concentrations. The deflagration characteristics including the flame propagation velocity,overpressure,quenching diameter and onset decomposition temperature are studied. Comparisons of flame acceleration process and thermal hazard between the mixtures containing NH3,C3H8and C2H4were made.

2. Experiment

2.1. Flame propagation experiment

The experimental set-up is schematically shown in the previous study [21]. Its main components are the explosion channel, the ignition controller,transducers,high-speed camera and the system for data acquisition. Experiments are performed in the cylindrical PMMA channels and stainless steel channels with different values of length and diameter.Each channel is equipped with several ports for the gas feed/evacuation valve,pressure transducers and ignition wire.

The cylindrical PMMA channel is schematically drawn in Fig.1,which contains Shchelkin spirals placed near the closed end of the channel and spaced 300 mm. The pitch and blockage ratio (BR) of Shchelkin spirals are 10 mm and 0.36, respectively. The piezoelectric transducers are installed in flush-mounted in the channel wall to minimize the flow disturbance and to check the actual gas pressure. The data acquisition rate and the measuring range are 1 MHz and 0-6.895 MPa. The maximum deviation of the pressure sensor in the investigated range is 2.1%. The locations of transducers are 100 mm, 400 mm, 600 mm, 800 mm, 1000 mm,1200 mm, 1400 mm and 1600 mm from the right end of the channel.

The combination channel used in this study is shown schematically in Fig. 2. A combination channel is composed of three channels,i.e.,the right PMMA channel,the left PMMA channel,and the middle stainless steel channel. The right PMMA channel is the main combustion channel, whose inner diameter is 15 mm and lengths are 0.1 m, 0.5 m, 1 m and 2 m, respectively. The middle stainless steel micro-channel with the length of 0.1 m and inner diameters of 0.3 mm, 0.5 mm, 0.7 mm,1.2 mm and 2 mm connects the right and left of PMMA channels by joint assembly. The left PMMA channel with the length of 0.5 m and inner diameter of 15 mm, which is called the exhaust channel, serves to observe whether the flame passing through the stainless steel microchannel.

The metering of the gas components is performed by mass flow controllers, the mass flow range and uncertainty are 0-500 mL/min and±0.5%,respectively.To ensure a homogeneous mixture,the fuels and nitrous oxide streams were brought into a mixing chamber(5 L)prior to entering the explosion channel.Gas mixtures were ignited by a 100-mm-long Ni-Chrome wire with a diameter of 0.6 mm that was placed at the center of geometry of the ignition chamber. The frame rate of the camera was set to 10000 fps. All mixtures used in the tests have an initial temperature of 298 K and an initial pressure of 1 bar.A minimum of three experiments were performed for each initial condition of the explosive mixture. At least three measurements for each parameter variation were carried out.

2.2. High-performance adiabatic calorimetry experiment

The PHI-TEC II adiabatic calorimeter,which is manufactured by HEL Inc., of the UK, can be used to measure the temperature and pressure changes to determine the reaction kinetics of a sample in which an exothermic reaction occurs.The reaction proceeds under nearly adiabatic conditions.The heat-wait search(H-W-S)mode for detecting the self-heating rate is adopted for PHI-TECII[22,23].The equipment can quickly track the thermal decomposition process of the materials,with a maximum tracking rate as high as 200°C/min.

Adiabatic tests were carried out on a PHI-TEC II adiabatic calorimeter.A stainless steel thin-walled sample cell of Model 1A with the volume of 10 mL and heat capacity of 13.21 J/K was employed.In the test,the heating step of 5°C,detection sensitivity of 0.03°C/min, waiting time of 5 min, initial temperature of 60°C and final temperature of 250°C were set. A pressure difference set by the pressure compensation system between the test cell and chamber was less than 0.83 bar.

2.3. Examined system

Ethylene (99.97% purity), nitrous oxide (99.98%), propane(99.97%)and ammonia(99.7%)are produced by Nanjing Electronic Devices Institute. The composition of the N2O/fuel mixtures is shown in Table 1.

3. Theoretical calculation

The calculation of the Chapman-Jouguet detonation parameters for N2O/fuel mixtures was carried out by using the NASA-CEA program [24]. The condition for chemical equilibrium is determined by the minimization of free energy subject to mass-balance constraints. By defining the term G,

Where g the Gibbs energy per kilogram of mixture, λiare Lagrangian multipliers;is the assigned number of element i per kilogram of total reactants,bithe number of element i per kilogram of mixture, According to variational principle, the condition for equilibrium becomes:

Fig.1. Schematic showing of the PMMA channel.

Fig. 2. Schematic showing of the combination channel.

Table 1 The compositions of the mixtures.

Where μjis the chemical potential per kilogram of species j.aijis are the number of element i per kilogram of mixture of species j.Treating the variations δnjand δλias independent gives:

Eq. (3) and Eq. (4) permit the determination of equilibrium compositions for thermodynamic states specified by an assigned temperature and pressure. Table 2 presents parameters of the reactive system in stoichiometric N2O/NH3, N2O/C3H8and N2O/C2H4mixtures at 1 bar and 298 K.

4. Results and discussion

4.1. Flame propagation process

Fig.3 shows the front flame propagation velocity as a function of time in the PMMA channel. Results show that the flame propagation process of the mixtures with C3H8and NH3can be divided into four typical stages, including slow acceleration, deflagration to detonation transition,steady detonation and decay stage.However,a difference occurs for the mixture of N2O/C2H4, which undergoes three stages during the flame propagation down the axis of the channel, deflagration to detonation transition, steady detonation and decay stage. After ignition, the flame speed begins to increase sharply due to the production of burned gas in the channel,which reaches 1877 m/s at 0.5 ms,while the premixed N2O/C3H8and N2O/C2H4reach their maximum values of 2227 m/s and 2240 m/s at 0.7 ms and 0.9 ms respectively. In particular, the flame speed of N2O/NH3increases slightly within the range of 0-0.6 ms at the initial ignition stage.The steady detonation velocities are 2198 m/s,2166 m/s and 2208 m/s for N2O/NH3, N2O/C3H8and N2O/C2H4,respectively, the theoretical values are 2240 m/s, 2180 m/s and 2200 m/s respectively,as shown in Table 2.The deviations betweenthe experimental and the theoretical values are 1.9%,0.6%and 0.4%respectively. When the flame approaches the vent of the channel,the propagation velocities decrease sharply mainly due to the unburned gas began to escape from the vent.

Table 2 C-J detonation parameters of N2O/NH3, N2O/C2H4 and N2O/C3H8.

Fig. 3. The front flame propagation velocity as a function of time in the PMMA channel.

The flame accelerations rate of N2O/C2H4and N2O/C3H8are clearly higher as compared to that of N2O/NH3. NH3is a potential energy carrier because it has a great advantage as a carbon-free fuel. However, the activity of NH3is much lower than that of hydrocarbons[25].Therefore,it is observed that N2O/NH3mixture has the lowest value of flame acceleration rate.For N2O/C2H4and N2O/C3H8, the flame acceleration rate of the former is obviously higher than that of the latter, the same order (C2H4>C3H8), has been observed for the normal burning velocity [26].

4.2. Overpressures

Fig.4 indicates the pressure histories of the experimental results taken from the eight gauges in the channel. It is seen that the maximum overpressure occurs at a distance of 400 mm from the ignition position and is 56.5 bar and 59.2 bar for N2O/NH3and N2O/C3H8,respectively,as shown in Fig.4(a)-(b).In this instance,spiral wires enhance turbulence due to their vortex creation,and a strong interaction between the fast flame and turbulence is promoted by the reflective wave,which increases the flame velocity,followed by a pressure rise.Similar observation has been reported on the flame acceleration [27]. Then, the pressure abruptly drops to around 40 bar.However,the highest overpressure for N2O/C2H4is observed at x=100 mm, as shown in Fig. 4(c), and it takes a shorter time to reach the maximum value. As mentioned earlier, the flames are susceptible to flame wrinkling for various mixtures,particularly for the more reactive fuels, due to the diffusional-thermal instability effect.Consequently,this wrinkled flame causes an increase in the mass burning rate, giving a higher overall flame velocity and overpressure[28,29].A more rapid combustion rate thus leads to a faster flame propagation velocity while generating a larger maximum overpressure.

Fig. 4. Overpressure history of N2O/fuel taken from eight gauges in the channel.

Fig. 5. The plots of maximum overpressure obtained for the mixtures.

Fig. 5 shows the plots of maximum overpressure obtained for the mixtures. The average detonation pressures are 37.1 bar,45.1 bar and 40.3 bar between P4 and P8 for NH3, C3H8and C2H4,respectively, the calculated values are 32.4 bar, 40.5 bar and 39.0 bar,respectively,as shown in Table 2.The deviations between the experimental and the calculated values are 14.5%, 11.4% and 3.3%, respectively. The observation of the average detonation pressures higher than theoretical CJ values is also observed for all the mixture. During the deflagration process leading to the detonation the gases act like a piston and compresses the combustible mixture ahead of it,this results in a measured detonation pressure much greater than the theoretical value without of precompression [9]. On the other hand, as mentioned above, it is observed that the shortest time required for the flame propagation in the combustion channel is the mixture of N2O/C2H4, which results in the lowest heat loss duo to the channel wall. Hence, the deviation of the N2O/C2H4mixture between the experimental and the calculated value also presents the lowest, followed by the mixtures of N2O/C3H8and N2O/NH3.

4.3. Quenching diameter

Fig. 6 shows the diagram of the flame behaviors in the combination channels with various diameter(ds)of the middle channels and the length (lp) of the combustion channel. Two propagation regimes: (1) passing: the flame runs through the middle channel;(2) quenching: the flame extinguishes in the middle channel.×represents the flame quenching and ■represents passing.The quenching diameters(d)of N2O/C2H4,N2O/C3H8and N2O/NH3are 0.5 mm

Fig.6. Diagram of the flame behaviors in Combination channel with various diameter(ds) and the combustion channel with various length (lp).

It is clear that C2H4is easier to detonate than other fuels used in the experiments due to its high reactivity, it undergoes DDT in shorter run-up distances, i.e., closer to the ignition point. The turbulence and reaction rate increase fast between lp=0.1 m and lp=0.5 m(see Fig.7),leading to the temperature and pressure rise rapidly, which make the flame pass the middle channel easily at constant experimental conditions. Zhou K [30,31] conducted a gas deflagration flame quenching in narrow channels by experiments and obtained the relationship between the flame propagation velocity and the narrow gap.And the larger the flame velocity or gap of channel,the easier the quenching of the deflagration flame.The length of the combustion channel has an influence on the quenching diameter for N2O/C2H4.It is also found that NH3,a lower reactivity fuel, results in longer run-up distances in the experiments since the flame undergoes a relatively slow acceleration after being ignited. Results show that the channel diameter has more significant effect compare with the length (lp) of the combustion channel for N2O/C3H8and N2O/NH3mixtures.

A proper flashback arrester needs to have an effective passageway length to quench the flame definitely. An empirical correlation [32] for designing flashback arresters is given by:

Here L marks the flame arresters passageway length, Stis the turbulent flame velocity and dqthe quenching diameter of the passageway.The unit for L and dqis cm,while the flame speed is in m/s.When dq=0.5 mm,the maximum turbulent flame velocity of premixed N2O/C2H4in combustion channel is 2533 m/s, the flame arresters passageway length should be more than 12.67 cm to make the flame quench completely at 1 bar. Therefore, according to the empirical correlation, the effective flame arresters passageway length can be calculated.

4.4. Thermal hazard analysis

Fig.8 exhibits the variations of temperature and pressure during the process in the test. For N2O/NH3, the initial decomposition temperatures (Tonset) and maximum pressure are 160°C and 1.57 bar,respectively,in Fig.8(a).For N2O/C3H8mixture,the onset decomposition temperature is measured to be 140°C, then the exothermal reaction continues until the temperature is up to 250°C, adiabatic temperature rises of N2O/C3H8decomposition(ΔTad)is more than 90°C and maximum pressure is 1.60 bar during the test, as shown in Fig. 8(b). For N2O/C2H4, the initial decomposition temperatures is 135°C,the pressure increased from 1.36 bar to 1.53 bar during the exothermic process in Fig.8(c).The runaway phenomena of N2O/fuel mixtures could not be completely recorded, since the automatic shutdown occurred before the reaction had gone to completion.Thus the potential maximum pressure and temperature during these uncontrolled runaway reactions could not be determined. Prior to the automatic shutdown, the data recorded for the rates of pressure and temperature rise were increasing. As thermal sensitivity is one of the most important hazards for production and storage,much research has been made[33,34]. According to the initial decomposition temperature, it is observed that the thermal hazard order is N2O/C2H4> N2O/C3H8>N2O/NH3. As mentioned above, the flame acceleration rate also follows the order: N2O/C2H4> N2O/C3H8>N2O/NH3.

Fig. 8. The temperature and pressure history during the process in the test.

5. Conclusions

The dynamics of the flame front propagation and thermal hazard of the premixed N2O/fuel systems using different channels and PHI-TECII were investigated with their flame velocities, overpressures, extinction conditions and initial decomposition temperatures.

(1) The measurements provided the steady detonation velocities and overpressures of the premixed N2O/fuel. The chemical equilibrium calculations were performed to clarify these values. Experimental results confirmed the orders of the flame acceleration rate, the detonation velocities and overpressures were C2H4>C3H8>NH3, NH3>C2H4>C3H8, and C3H8>C2H4>NH3, respectively.

(2) The flame quenching diameters (d) were 0.5 mm

(3) The initial decomposition temperatures of N2O/fuels were obtained. The results showed that the thermal hazard order was N2O/C2H4>N2O/C3H8>N2O/NH3.

Acknowledgements

This research was supported by Open Research Fund Program of Science and Technology on Aerospace Chemical Power Laboratory(STACPLXXXXXXXX).