Accelerative expansion and DDT of stoichiometric ethylene/oxygen flame rings in micro-gaps

Acceleration and transition to detonation of expanding flame rings ignited at the center of 260 μm and 120 μm gaps between parallel flat pates were experimentally studied. The micro-spacing was initially filled with stoichiometric ethylene/oxygen mixtures at ambient pressure and temperature. Visualizations showed that the outward propagating reaction wave was initially smooth and circular, but petal-like wrinkles quickly developed on the flame ring. Flame wrinkles appeared earlier and closer to the ignition point as the gap width became smaller. The flame underwent fast acceleration during the onset of flame wrinkling, but the acceleration was relatively mild as the wrinkled flame ring continued to expand. Time exponents for the accelerative growth of corrugated flame rings were identical in the two highly confined gaps. The flame ring underwent deflagration-to-detonation transition as the propagation velocities abruptly surged from 1000 m/s to over 2000 m/s. The arc-shaped detonation waves initiated from local explosion spots on the flame ring were propagating at near Chapman–Jouguet velocities. The induction distance and time for detonation transition were both shorter in the smaller gap. Detonation cell patterns and the initiation locations were also clearly recorded through soot film visualizations.     

Near limit behavior of the detonation velocity

The behavior of the detonation velocity near the limits is investigated. Circular tubes of diameters 65, 44 and 13 mm are used. To simulate a quasi two-dimensional rectangular geometry thin annular channels are also used. The annular channels are formed by a 1.5 m long insert of a smaller diameter tube into the larger outer diameter detonation tube. Premixed mixtures of C₂H₂ + 2.5O₂ + 70%Ar, CH₄ + 2O₂ and C₂H₂ + 5N₂O + 50%Ar are used in the present study. The high argon dilution stoichiometric C₂H₂ + 2.5O₂ mixture has a regular cell size and piecewise laminar reaction zone and thus referred to as “stable”. The other two mixtures give highly irregular cell pattern and a turbulent reaction zone and are hence, referred to as “unstable” mixtures. Pressure transducers and optical fibers spaced 10 cm apart along the tube are used for pressure and velocity measurements. Cell size of the three mixtures studied is also determined using smoked foils in both the circular tubes and annular channels. The ratio d/λ (representing the number of cells across the tube diameter) is found to be an appropriate sensitivity parameter to characterize the mixture. The present results indicate that well within the limit, the detonation velocity is generally a few percent below the theoretical Chapman–Jouguet (CJ) value. As the limit is approached, the velocity decreases rapidly to a minimum value before the detonation fails. The narrow range of values of d/λ of the mixture where the velocity drops rapidly is found to correspond to the range of values for the onset of single headed spinning detonations. Thus we may conclude that the onset of single headed spin can be used as a criterion for defining the limits. Spinning detonations are also observed near the limits in annular channels.     

Flame acceleration and the transition to detonation of stoichiometric ethylene/oxygen in microscale tubes

Flame propagation in capillary tubes with smooth circular cross-sections and diameters of 0.5, 1.0, and 2.0 mm are investigated using high-speed photography. Flames were found to propagate and accelerate to detonation speed in stoichiometric ethylene and oxygen mixtures initially at room temperature in all three tube diameters. Ignition occurs at the midpoint along the length of the tube. We observe for the first time transition to detonation in micro-tubes. Detonation was observed with both spark and hot-wire ignition. Tubes with larger diameters take longer to transition to detonation. In fact, transition distance scales with the diameter in our 1.0 and 2.0 mm cases with spark ignition. Flame structures are observed for various stages of the process. Three types of flame propagation modes were observed in the 0.5 mm tube with spark ignition: (a) acceleration to Chapman–Jouguet (CJ) detonation speed followed by constant CJ wave propagation, (b) acceleration to CJ speed, followed by the detonation wave failure, and (c) flame acceleration to a constant speed below the CJ speed of approximately 1600 m/s. The current detonation mechanism observed in capillary tubes is applicable to predetonators for pulsed detonation, micro propulsion devices, safety issues, and addresses fundamental issues raised by recent theoretical and numerical analyses.     

Two-dimensional numerical simulation on galloping detonation in a narrow channel

The numerical simulations of the two-dimensional galloping detonation performed by using two-dimensional full Navier–Stokes simulations with a detailed chemistry model are presented. The detonation in a narrow channel with d = 5 mm, which is approximately twice the half-reaction length of hydrogen, shows a feature of galloping detonation with two initiations during its propagation under the laminar flow assumption. The distance between these two initiations is approximately 1300 mm, which causes the induction time behind the leading shock wave. As the channel width increases, the galloping feature diminishes. The detonation propagates approximately 4% lower than D_CJ for d = 10 and 15 mm. By increasing the channel width, the strength of the detonation increases, as shown in the maximum pressure histories. The effects of turbulence behind the detonation show that the galloping feature disappears, although its propagation velocity becomes 0.9 D_CJ. The strength of the detonation becomes significantly weak compared with the detonation propagating in the wide channel widths, and this feature is similar to the laminar assumption. The trend of the velocity deficits in the NS simulations agrees fairly well with the trend of the modified ZND calculations with η = 0.25.     

Critical conditions of hydrogen-air detonation in partially confined geometry

This work presents the results of the large scale experiments with detonation propagating in hydrogen–air mixtures in partially confined geometries. The main aim of the work was to find the critical conditions for detonation propagation in semi-confined geometries with uniform and non-uniform hydrogen–air mixtures. The experimental facility consisted of rectangular 9 × 3 × 0.6 m channel open from the bottom, acceleration section and test section, safety vessel, gas injection and data acquisition system. Sooted plates technique was used as a witness of the detonation. The rectangular channel was placed in a 100 m³ safety vessel. For uniform hydrogen–air mixtures experiments with four different channel heights h were performed: 8, 5, 3 and 2 cm. The critical hydrogen–air mixture height h* for which the detonation may propagate in a layer is close to the 3 cm which corresponds to approximately three detonation cell sizes. For non-uniform hydrogen–air mixture with hydrogen concentration slope equal approximately ?1.1%H₂/cm the critical hydrogen concentration at the top of the layer is approximately equal 26% and the mean detonation layer height is close to the 8.5 cm corresponding to the hydrogen concentration at the bottom of the layer approximately equal 16–17%.     

Pressure loading of detonation waves through 90-degree bend in high pressure H2–O2–N2 mixtures

Detonation experiments are conducted to investigate the detonation wave behavior in steam pipelines of boiling water reactors. Accumulated gases in BWRs are stoichiometric hydrogen/oxygen mixtures diluted with steam at 7 MPa. In the experiment, flammable gas mixture diluted with nitrogen at room temperature and up to 5 MPa is used to achieve equivalent detonation condition. Two test pieces are used, one is straight tube and the other is 90-degree bend. No initial pressure dependency in detonation wave behavior is observed in the experiments. However, in the straight tube tests, detonation velocities higher than theoretical values are measured when the initial pressures are greater than 2.3 MPa. This result is considered as attribution of real gas effect. In the 90-degree bend experiments, pressure time histories reveal pressure loads greater than the straight tube portion at two locations. One is a high pressure peak at the extrados of the bend and the other is a double pressure peak just downstream of the bend outlet. Second pressure peak just downstream of the bend outlet is due to transverse wave propagation. Additionally, the largest impulse is observed not at the extrados of the bend but at the intrados of bend outlet. These results show the importance of more investigations on transverse wave behaviors in failure potential evaluation.     

Initiation and sustaining mechanisms of stabilized Oblique Detonation Waves around projectiles

Direct initiations and stabilizations of three-dimensional conical detonation waves were attained by launching spheres with 1.06–1.31 times the C–J velocities into detonable mixtures. We conducted high time-resolution Schlieren visualizations of the whole processes over unsteady initiations to stable propagations of the stabilized Oblique Detonation Waves (ODWs) using a high-speed camera. The detonable mixtures were stoichiometric oxygen mixtures with acetylene, ethylene or hydrogen. They were diluted with argon in a 50% volumetric fraction, and a 75% diluted mixture was also tested for the acetylene/oxygen. The direct initiation of detonation by the projectile and the DDT process like the re-initiation appeared in the initiation process of stabilized ODW. This process eventually led to the stabilized ODW supported by the projectile velocity and the ringed shape detonation wave originating in the re-initiation. We modeled the spatial evolution of stabilized ODW after the re-initiation based on its C–J velocity and angle. The model qualitatively reproduced the measured development rate of stabilized ODW. We also discussed about the detonation stability for the curvature effect arising from the three-dimensional nature of stabilized ODW around the projectile. The curvature effect attenuated the detonation wave below its C–J velocity at the vicinity of projectile. The propagation limits of curvature effect will be responsible for the criticality to attain the stabilized ODWs. By accessing the detailed distributions of propagation velocities and curvature radiuses, the critical curvature radiuses normalized by the cell sizes experimentally revealed to be 8–10 or 15–18 for mixtures diluted with each 50% argon or 75% argon/krypton.     

氢氧爆轰波在激波管中的成长机制研究

 针对气相爆轰波成长机制研究,采用压力传感器和高速摄影技术,测试了氢氧混合气体在点火后的火焰波、前驱冲击波以及爆轰波的成长变化过程,计算了冲击波过程参数和气体状态参数,分析了火焰加速机制。实验结果表明,APX-RS型高速摄影系统可用于拍摄气相爆轰波的成长历程;氢氧爆轰波的产生是由于湍流火焰和冲击波的相互正反馈作用,导致反应区内多处发生局部爆炸,爆炸波与冲击波相互耦合,最终成长为定常爆轰波。     

Comprehensive visualization of detonation-diffraction structures and sizes in unstable and stable mixtures

The geometry and characteristic length of diffraction and re-initiation during a two-dimensional detonation propagation were revealed by visualization. C₂H₄ + 3O₂ (unstable), 2C₂H₂ + 5O₂ + 7Ar (stable) and 2C₂H₂ + 5O₂ + 21Ar (stable) were used as the test mixtures. Experiments were performed over the deviation angle range from 30° to 150° and the initial pressure range from 15.8 to 102.3 kPa. By self-emitting photography, we confirmed that the geometry and the characteristic length of diffraction are not different among test gases, with the exception of the fan-like structure of re-initiation that occurred regardless of whether the mixture was unstable or stable. We conducted a compensative experiment by changing the deviation angle and initial pressure, and summarized the detonation diffraction by shadowgraph. At deviation angles larger than 60°, we measured the distances from the vertex of the channel corner to the point where the transverse detonation wave reflected on the under wall (= wall reflection distance) and confirmed that wall reflection distances are approximately in the range of 10–15 times the cell width, whether the mixture is unstable or stable.     

Stabilized flames in hybrid aluminum-methane-air mixtures

A premixed methane–air bunsen-type flame is seeded with micron-sized (d₃₂ = 5.6 μm) atomized aluminum powder over a wide range of solid fuel concentrations. The burning velocities of the resulting two-phase hybrid flame are determined using the total surface area of the inner flame cone and the known volumetric flow rate, and spatially resolved flame spectra are obtained with a spectral scanning system. Flame temperatures are derived through polychromatic fitting of Planck’s law to the continuous part of the spectrum. It is found that an increase in the solid fuel concentration changes the aluminum combustion regime from low temperature oxidation to full-fledged flame front propagation. For stoichiometric methane–air mixtures, the transition occurs in the aluminum concentration range of 140–220 g/m³ and is manifested by the appearance of AlO sub-oxide bands and an increase in the flame temperature to 2500 K. The flame burning velocity is found to decrease only slightly with an increase in aluminum concentration, in contrast to the rapid decrease in flame speed, followed by quenching, that is observed for flames seeded with inert SiC particles. The observed behavior of the burning velocity and flame temperature leads to the conclusion that intense aluminum combustion in a hybrid flame only occurs when the flame front propagating through the aluminum suspension is coupled to the methane–air flame.     

Low-frequency sound propagation in porous media: Glass spheres and pea gravel

The sound propagation properties of two air-filled granular materials: large sifted pea gravel and 10 mm diameter glass spheres have been measured in an impedance tube. The experimental method was essentially the same as reported earlier [Swenson et al. Low-frequency sound wave parameter measurement in gravels. Appl Acoust 2010; 71: 45–51] for two other kinds of gravel: crushed limestone and undifferentiated pea gravel. Additional sampling and processing steps were applied to the microphone signals such that instead of tones, band-limited random noise was used as the input signal, and spectral domain complex pressures are now offered as input to the estimation algorithm. The estimation process extracts the best-fit attenuation coefficient, phase velocity, and characteristic impedance for the material over the signal frequencies, all with better precision than we previously obtained. Quadratic approximations for the acoustical parameters are given over the frequency range 25–160 Hz. The media are both slightly attenuating and dispersive, having attenuation coefficients within 0.13–0.34 Np/m, phase velocities smaller than those in air (180–240 m/s), and characteristic impedance approximately 3–5 times that for air. Pea gravel was more attenuating, and had slightly higher characteristic impedance, but lower phase velocities than the glass spheres.     

Experimental study of detonation in a cryogenic two-phase H2–O2 flow

Direct initiation and propagation of detonation through a cryogenic two-phase flow constituted by liquid oxygen droplets in gaseous hydrogen at 100 K are experimentally investigated. The influence of droplet size distribution is characterized in a cryogenic gaseous helium and liquid oxygen two-phase flow. Droplet sizing and detonation experiments are conducted by varying different parameters: distance from the injector, helium and hydrogen mass flow rates, global equivalence ratio and addition of gaseous nitrogen. Droplet size distributions reveal quick vaporization of the smallest droplets of the cryogenic jet. Results in terms of wave velocity, pressure, and detonation cells show that a detonation wave can be directly initiated, with a propagation wave velocity of 20% higher than the Chapman–Jouguet value. Cell size measurements show that the mixture sensitivity is not affected by the presence of droplets. Addition of gaseous nitrogen reduces only slightly the peak pressure, but the detonation velocity is reduced by about 30%.     

Fundamental burning velocities of meso-scale propagating spherical flames with H2, CH4 and C3H8 mixtures

This study is performed to experimentally examine the fundamental burning velocity characteristics of meso-scale outwardly propagating spherical laminar flames in the range of flame radius r_f approximately from 1 to 5 mm for hydrogen, methane and propane mixtures, in order to make clear a method for improving combustion of micro–meso scale flames. Macro-scale laminar flames with r_f > 7 mm are also examined for comparison. The mixtures have nearly the same laminar burning velocity (S_L0 = 25 cm/s) for unstretched flames and different equivalence ratios ?. The radius r_f and the burning velocity S_Ll of meso-scale flames are estimated by using sequential schlieren images recorded under appropriate ignition conditions. It is found that S_Ll of hydrogen and methane premixed meso-scale flames at the same r_f or the Karlovitz number Ka shows a tendency to increase with decreasing ?, whereas S_Ll of propane flames increases with ?. However, S_Ll tends to decrease with the Lewis number Le and the Markstein number Ma, irrespective of the type of fuel and ?. It also becomes clear that the optimum flame size and Ka to improve the burning velocity exist for some mixtures depending on Le and fuel types.     

An experimental investigation of the onset of detonation

An experimental configuration is devised in the present investigation whereby the condition at the final phase of the deflagration to detonation transition (DDT) process can be generated reproducibly by reflecting a CJ detonation from a perforated plate. The detonation products are transmitted downstream through the plate, generating a turbulent reaction front that mixes with the unburned mixture and that drives a precursor shock ahead of it at a strength of about M = 3. The gasdynamic condition that is generated downstream of the perforated plate closely corresponds to that just prior to the onset of detonation in the DDT process. The turbulence parameters can be controlled by varying the geometry of the perforated plate; thus, the condition leading to the onset of detonation can be experimentally investigated. A one-dimensional theoretical analysis of the steady wave processes was first performed, and the experimental results show good agreement, indicating that the present experimental condition can be theoretically described. Two different detonation tube geometries (one with a square cross-section of 300 mm by 300 mm and the other with a circular cross-section of 150 mm) are used to demonstrate the independence of the tube diameter at the critical condition for DDT. Perforated plates with different hole diameters (d = 8, 15, and 25 mm) were tested, and the hole spacing to hole diameter ratio was maintained at 0.5. Different hydrogen–air mixtures were tested at normal temperature and pressure. For the plate with 8 mm holes, the onset of detonation is never observed. For the plate with 15 mm holes, successful initiation of a detonation is achieved for 0.8 < < 1.75 in both detonation tubes. For the plate with 25 mm holes, detonation initiation is observed for 0.7 < < 2.1 in the square detonation tube and for 0.8 < < 1.6 in the smaller circular detonation tube.     

Flame propagation of nano/micron-sized aluminum particles and ice (ALICE) mixtures

Flame propagation of aluminum–ice (ALICE) mixtures is studied theoretically and experimentally. Both a mono distribution of nano aluminum particles and a bimodal distribution of nano- and micron-sized aluminum particles are considered over a pressure range of 1–10 MPa. A multi-zone theoretical framework is established to predict the burning rate and temperature distribution by solving the energy equation in each zone and matching the temperature and heat flux at the interfacial boundaries. The burning rates are measured experimentally by burning aluminum–ice strands in a constant-volume vessel. For stoichiometric ALICE mixtures with 80 nm particles, the burning rate shows a pressure dependence of r_b = aPⁿ, with an exponent of 0.33. If a portion of 80 nm particles is replaced with 5 and 20 μm particles, the burning rate is not significantly affected for a loading density up to 15–25% and decreases significantly beyond this value. The flame thickness of a bimodal-particle mixture is greater than its counterpart of a mono-dispersed particle mixture. The theoretical and experimental results support the hypothesis that the combustion of aluminum–ice mixtures is controlled by diffusion processes across the oxide layers of particles.     

A structural study of premixed hydrogen-air cellular tubular flames

Two dimensionally spatially resolved structural measurements are reported for cellular phenomena in lean laminar premixed hydrogen-air tubular flames. Laser-induced Raman scattering and chemiluminescence imaging are combined to investigate low Lewis number lean hydrogen-air flames. The strong effect of thermal-diffusive imbalance is observed in radial profiles interpolated through the centers of reaction and extinction zones. In the flame cell, the equivalence ratio is ～80% higher than the inlet mixture, resulting in a peak flame temperature of 1600 K that is 550 K above the adiabatic flame temperature of the inlet mixture (1055 K). In the adjacent extinction zone, the temperatures are ～900 K lower than the peak flame temperature and the equivalence ratio is similar to the inlet mixture. Despite doubling the global stretch rate from 200 s^?1 to 400 s^?1, the enhancement of local equivalence ratio and peak temperature in the flame cell remain similar. This enhancement seems dependent on the local cellular flame curvature, that is similar between both cases. With strong preferential diffusion effects, cellular flames offer unique validation data to improve the accuracy of current molecular transport modeling techniques.     

Thermographic visualization of the superficial vein and extravasation using the temperature gradient produced by the injected materials

There are few effective methods to detect or prevent the extravasation of injected materials such as chemotherapeutic agents and radiographic contrast materials. To investigate whether a thermographic camera could visualize the superficial vein and extravasation using the temperature gradient produced by the injected materials, an infrared thermographic camera with a high resolution of 0.04 °C was used. At the room temperature of 26 °C, thermal images and the time course of the temperature changes of a paraffin phantom embedded with rubber tubes (diameter 3.2 mm, wall thickness 0.8 mm) were evaluated after the tubes were filled with water at 15 °C or 25 °C. The rubber tubes were embedded at depths of 0 mm, 1.5 mm, and 3.0 mm from the surface of the phantom. Temperature changes were visualized in the areas of the phantom where the tubes were embedded. In general, changes were more clearly detected when greater temperature differences between the phantom and the water and shallower tube locations were employed. The temperature changes of the surface of a volunteer’s arm were also examined after a bolus injection of physiological saline into the dorsal hand vein or the subcutaneous space. The injection of 5 ml room-temperature (26 °C) saline into the dorsal hand vein enabled the visualization of the vein. When 3 ml of room-temperature saline was injected through the vein into the subcutaneous space, extravasation was detected without any visualization of the vein. The subtraction image before and after the injection clearly showed the temperature changes induced by the saline. Thermography may thus be useful as a monitoring system to detect extravasation of the injected materials.     

Effects of NO2 addition on hydrogen ignition behind reflected shock waves

Ignition delay time measurements of H₂/O₂/NO₂ mixtures diluted in Ar have been measured in a shock tube behind reflected shock waves. Three different NO₂ concentrations have been studied (100, 400 and 1600 ppm) at three pressure conditions (around 1.5, 13, and 30 atm) and for various H₂–O₂ equivalence ratios for the 100 ppm NO₂ case. Results were compared to some recent ignition delay time measurements of H₂/O₂ mixtures. A strong dependence of the ignition delay time on the pressure and the NO₂ concentration was observed, whereas the variation in the equivalence ratio did not exhibit any appreciable effect on the delay time. A mechanism combining recent H₂/O₂ chemistry and a recent high-pressure NOx sub-mechanism with an updated reaction rate for H₂ + NO₂ ? HONO + H was found to represent correctly the experimental trends over the entire range of conditions. A chemical analysis was conducted using this mechanism to interpret the experimental results. Ignition delay time data with NO₂ and other NOx species as additives or impurities are rare, and the present study provides such data over a relatively wide pressure range.     

Front shock behavior of stable curved detonation waves in rectangular-cross-section curved channels

The propagation of curved detonation waves of gaseous explosives stabilized in rectangular-cross-section curved channels is investigated. Three types of stoichiometric test gases, C₂H₄ + 3O₂, 2H₂ + O₂, and 2C₂H₂ + 5O₂ + 7Ar, are evaluated. The ratio of the inner radius of the curved channel (r_i) to the normal detonation cell width (λ) is an important factor in stabilizing curved detonation waves. The lower boundary of stabilization is around r_i/λ = 23, regardless of the test gas. The stabilized curved detonation waves eventually attain a specific curved shape as they propagate through the curved channels. The specific curved shapes of stabilized curved detonation waves are approximately formulated, and the normal detonation velocity (D_n)?curvature (κ) relations are evaluated. The D_n nondimensionalized by the Chapman–Jouguet (CJ) detonation velocity (D_CJ) is a function of the κ nondimensionalized by λ. The D_n/D_CJ?λκ relation does not depend on the type of test gas. The propagation behavior of the stabilized curved detonation waves is controlled by the D_n/D_CJ?λκ relation. Due to this propagation characteristic, the fully-developed, stabilized curved detonation waves propagate through the curved channels while maintaining a specific curved shape with a constant angular velocity. Self-similarity is seen in the front shock shapes of the stabilized curved detonation waves with the same r_i/λ, regardless of the curved channel and test gas.     

Combustion wave propagation in mixtures of JSC-1A lunar regolith simulant with magnesium

Combustion of lunar regolith mixed with energetic additives is a potential method for production of construction materials in future moon missions. Recently, self-sustained combustion in the mixtures of JSC-1A lunar regolith and magnesium has been demonstrated. However, the concentration of magnesium in those mixtures was as high as 26 wt%. Note that magnesium must be either transported from Earth or recovered from lunar minerals or used structures. The present paper focuses on the minimization of magnesium content in JSC-1A/Mg mixtures. The mixtures were compacted into pellets and ignited in argon environment. Initial attempts to decrease magnesium concentration resulted in the observations of a spinning combustion wave at 23 wt% Mg. The observed spin combustion involved periodical motion of two counterpropagating hot spots along a helical path on the sample surface. These observations, including features such as formation of a faster hot spot after collision of the counterpropagating spots, confirm theoretical predictions for spin combustion in solid–solid mixtures. High-energy mechanical milling of JSC-1A in a planetary ball mill significantly increased its reactivity and improved combustion of its mixtures with magnesium. Mixtures of the obtained powder (the median diameter of about 3 μm) with 26 wt% Mg exhibit easy ignition and vigorous combustion. The minimum concentration of magnesium required for self-sustained propagation of a planar combustion front is as low as 13 wt%.