Search

Downward propagating flames ignited at the open end of an open-closed tube exhibit thermo-acoustic instability due to interaction of combustion generated acoustic fluctuations with the flame front. At sufficiently high laminar burning velocity (SL) two regimes of thermo-acoustic instability are observed, namely, primary instability (where initial cellular flame transitions to a vibrating flat flame) and a secondary instability (where vibrating flat flame transitions to vibrating turbulent flame due to parametric instability of flame front). On further increasing SL to a particular value, “complete instability” of flat flames is observed meaning flat flame cannot be stabilized and initial cellular flame transitions directly to parametric instability. This particular SL introduced in this work is termed “critical SL”. In past experimental works, stability of flat flames in the acoustic field had only been studied in terms of acoustic velocity amplitude and a critical acoustic velocity amplitude had been measured at the onset of parametric instability. The novelty of this work is that boundary of unconditional instability of flat flame (flat flame is unstable irrespective of acoustic velocity amplitude) is determined in terms of mixture conditions, e.g., SL. Particularly for propagating flames, this critical SL can be measured more easily and accurately than the critical acoustic velocity. This work presents the effect of Le (Lewis number) on critical SL. Three different fuels, CH4, C2H4 and C3H8 are tested with two different dilution gases (N2 and CO2) for equivalence ratio of 0.8 (lean) and 1.2 (rich). Twelve different Le ranging from 0.7 to 1.9 are generated through these mixture combinations. Generally, larger Le mixtures show higher critical SL than lower Le mixtures for any fuel. Theoretical calculations are performed to predict critical SL by studying instability of planar flame fronts in presence of acoustic forcing. Theoretical calculations successfully captured the effect of Le as predicted stability region of planar flame is narrower for lower Le than that for higher Le. However, accurate quantitative predictions of critical SL couldn't be obtained from existing theory, particularly for non-unity Le. Hence, a correction (a function of Zeldovich number, β and Le) to width of stability region is proposed to obtain better quantitative agreement for critical SL between experiments and theory and performs significantly well. The correction factor acts to compensate for the inaccuracies in Markstein number obtained from an analytical relationship during calculation of stability region width.

Recent investigations suggested that the primary influence of an electric field on a flame is flow modification caused by ionic wind, and that negative ions produced by the electron impact attachment should play a key role in the bi-directional ionic wind. In order to prove this hypothesis in electric fields parallel to the propagating flames, we designed a coflow experiment with laminar lifted flames in vertical electric fields produced by a nozzle and ground electrode installed over the flame. We found that applying DC and AC increased the flame displacement speed, and decreased the unburned velocity even to negative velocity. Velocity measurements revealed the influence of the electric body force on the flow volume, indicating the importance of the electron impact attachment when the nozzle was charged with positive voltage. The flame propagation speeds were estimated by subtracting the unburned velocity from the displacement speed, and were well correlated with those of stationary lifted flames without an applied electric field as a function of flame curvature. This supported our hypothesis that the effect of the electric field is reflected in the flow modification, and that the flame is affected by the modified flow. It also suggested that the propagation direction of premixed or nonpremixed edge flames can be manipulated by coupling the appropriate electric fields.

A vortical structure occurring at the fuel stream in laminar nonpremixed jet flames was recently found and shown to have both a fluid-dynamic impact on the flow field and a possible influence on the flame stability and soot formation. We designed a systematic experiment and numerical simulation to investigate the physical mechanisms of this recirculation phenomenon in a coflow system. We hypothesized that a negative buoyancy, caused by the fuel jet being heavier than the ambient air, may play a significant role in the recirculation. Therefore, we experimentally varied the density of the fuel jet using a binary mixture of methane and n-butane, and tested the density of the coflow oxidizer by replacing nitrogen with carbon dioxide. Several fuel jet velocities, flame temperatures, and nozzle diameters were also studied to thoroughly investigate all parameters that might possibly affect the recirculation. As a result, we found that our modified Richardson number, which is based on the cold density difference between the fuel and the coflow, the flame length, and the jet momentum flux, explained the physical mechanism of the recirculation well, with Ri ∼60 being the critical value for formation of the recirculation. The negative buoyancy was the primary driving force behind the recirculation, while the jet momentum mitigated its formation.

The oscillating lifted flame in a laminar nonpremixed nitrogen-diluted fuel jet is known to be a result of buoyancy, though the detailed physical mechanism of the initiation has not yet been properly addressed. We designed a systematic experiment to test the hypothesis that the oscillation is driven by competition between the positive buoyancy of flame and the negative buoyancy of a fuel stream heavier than the ambient air. The positive buoyancy was examined with various flame temperatures by changing fuel mole fraction, and the negative buoyancy was investigated with various fuel densities. The density of the coflow was also varied within a certain range by adding either helium or carbon dioxide to air, to study how it affected the positive and negative buoyancies at the same time. As a result, we found that the range of oscillation was well-correlated with the positive and the negative buoyancies; the former stabilized the oscillation while the latter triggered instability and became a source of the oscillation. Further measurements of the flow fields and OH radicals evidenced the important role of the negative buoyancy on the oscillation, detailing a periodic variation in the unburned flow velocity that affected the displacement of the flame.

According to previous theory, pulsating propagation in a premixed flame only appears when the reduced Lewis number, β ( Le -1), is larger than a critical value (Sivashinsky criterion: 4(1 + 3 ) ≈ 11), where β represents the Zel'dovich number (for general premixed flames, β ≈ 10), which requires Lewis number Le > 2.1. However, few experimental observation have been reported because the critical reduced Lewis number for the onset of pulsating instability is beyond what can be reached in experiments. Furthermore, the coupling with the unavoidable hydrodynamic instability limits the observation of pure pulsating instabilities in flames. Here, we describe a novel method to observe the pulsating instability. We utilize a thermoacoustic field caused by interaction between heat release and acoustic pressure fluctuations of the downward-propagating premixed flames in a tube to enhance conductive heat loss at the tube wall and radiative heat loss at the open end of the tube due to extended flame residence time by diminished flame surface area, i.e., flat flame. The thermoacoustic field allowed pure observation of the pulsating motion since the primary acoustic force suppressed the intrinsic hydrodynamic instability resulting from thermal expansion. By employing this method, we have provided new experimental observations of the pulsating instability for premixed flames. The Lewis number (i.e., Le ≈ 1.86) was less than the critical value suggested previously.

The generation mechanism of primary acoustic instability of downward-propagating flames in a combustion tube is investigated experimentally. The discussion first treats the effects of the coupling constant βM , where β and M represent the Zel'dovich and Mach numbers, respectively, and the effects of the flame surface area variation in equidiffusive flames on primary acoustic instability. A higher coupling constant tends to generate stronger acoustic sound under the same acoustic losses, whereas the variation in the flame surface area does not seriously affect the coupling between pressure and heat release rate fluctuations. We then describe the effect of the Lewis number on primary acoustic instability in non-equidiffusive flames, where the diffusive thermal effects largely determine the internal flame structure. For a Lewis number below (above) unity, where the reaction rate increases (decreases) with reducing (increasing) reaction zone thickness, relatively strong (weak) acoustic sound is produced under the same coupling constant, because the chemical reaction rate becomes very sensitive (insensitive) to gas temperature fluctuations in the acoustic field. Finally, we obtain a linear relationship between the coupling constant and the average acoustic intensity, and results show that relatively large coupling constant is required to generate primary acoustic instability as the Lewis number increases.

The capabilities of CFDShip-Iowa V4.5 and V6.2 simulations for wave run-up for single/multiple surface-piercing cylinders in regular waves are assessed in conjunction with the ITTC OEC Workshop on VIV and Wave Run-up held in Nantes, France October 17⿿18, 2013. Statistical convergence and grid/domain/turbulence model sensitivity studies show uncertainty UI values are less than 1% for both the experiment/simulation data and grid/domain/turbulence model sensitivities are less than the facility bias UFB. Simulations are qualitatively validated using UFB and error E between experiment and simulation. Errors for single/four cylinder(s) are comparable with UFB and average error E ¯ for the other CFD data of the workshop participants. Results from both V4.5 and V6.2 agree well with experiment and also show capabilities to apply for complex/large-scale problem and to describe detailed wave breaking/spraying for V4.5 and V6.2, respectively. Study on run-up profile shows mean wave elevations are similar to free surface elevations for a cylinder in a steady current. For wave steepness/wave length effect, 2nd to 1st harmonic ratio on the cylinder weather face decreases with steepness for short wave, but increases for long wave. Wave diffraction analysis shows that 2nd/3rd to 1st harmonic ratio in diffracted wave contour reach up to 18⿿41%/9⿿16% respectively. Due to interaction between cylinders, crest/trough heights increase 5⿿17%/4⿿10% for 4C, respectively. Study on vortex shedding shows short wave has insufficient time to develop periodic vortex shedding compared to long wave.

This paper deals with the onset mechanism of primary acoustic instability of downward-propagating flames in a combustion tube. We focus on the effects of a coupling constant, βM, where β and M represent the Zel'dovich and Mach numbers, respectively, and the variation in the flame surface area. To change the coupling constant, various gas compositions for lean ethylene flames diluted with carbon dioxide or nitrogen are used. We obtain a linear relationship between the coupling constant and the average acoustic intensity, and the critical values of the coupling constants are acquired through linear approximation regarding the onset of the primary acoustic instability. Furthermore, we adopt the CO2 laser irradiation method to alter the shape of the flame front, and experimental results show that the variation in the flame surface area does not always cause spontaneous generation of the primary acoustic instability in initially nonvibrating flames. Furthermore, even in initially vibrating flames, the growth rate of the primary acoustic instability is not associated with the growth or decline in the flame surface area in the present experiments. Finally, we also estimate the effects of acoustic losses on acoustic instability. Larger total acoustic losses tend to suppress acoustic vibration even at the same coupling constant. In addition, a remarkable transition from a secondary acoustic instability to the primary acoustic instability is observed at strong laser irradiation and is briefly discussed.

Laminar lifted butane flames diluted with nitrogen have been investigated experimentally to determine distinctive self-excitation regimes in the flame stability maps and also to elucidate the individual self-excitation characteristics. Self-excitations of lift-off height are classified into five regimes in laminar free-jet lift-off butane flames diluted with nitrogen: a stationary lifted regime (regime I), a heat-loss-induced self-excitation (regime II), a buoyancy-induced self-excitation due to flame flicker as well as a heat-loss-induced self-excitation (III), a combined form of an oscillation prior to blow-out and a heat-loss-induced oscillation (regime IV), and a combined form of an buoyancy-induced self-excitation and a heat-loss-induced oscillation as well as an additional buoyancy-driven self-excitation due to flame flicker (regime V). Extremely low-frequency (

We experimentally investigated lifted propane jet flames diluted with nitrogen to obtain flame-stability maps based on heat-loss-induced self-excitation. We found that heat-loss-induced self-excitations are caused by conductive heat loss from premixed flame branches to trailing diffusion flames as well as soot radiation. The conductive-heat-loss-induced self-excitation at frequencies less than 0.1 Hz is explained well by a suggested mechanism, whereas the oscillation of the soot region induces a self-excitation of lift-off height of the order of 0.1 Hz. The suggested mechanism is also verified from additive experiments in a room at constant temperature and humidity. The heat-loss-induced self-excitation is explained by the Strouhal numbers as a function of the relevant parameters.

The characteristics of lifted laminar propane flames diluted with nitrogen have been investigated experimentally to elucidate self-excitation and the effects of flame curvature. Flame oscillation modes are classified as follows: oscillation induced by heat loss, a combination of oscillations induced by heat loss and buoyancy, and a combination of the oscillations induced by heat loss and diffusive thermal instability. It is shown that the oscillation induced only by heat loss is not relevant to the diffusive thermal instability and hydrodynamic instability caused by buoyancy; this oscillation is observed under all lift-off flame conditions irrespective of the fuel Lewis number. These experimental evidences are displayed through the analysis of the power spectrum for the temporal variation of lift-off height. The possible mechanism of the oscillation induced by heat loss is also discussed.

This paper presents the centrifugation of the nano-particles using the rotating flow in the microchamber. The feasibility of the microfluidic centrifuge was demonstrated numerically and experimentally. The centrifugation of the nano-particles of 300 and 500 nm was visualized by using a high-speed CCD camera and a micro-PIV. In addition, its performance was characterized quantitatively. The experimental results of the mechanical contact-free microfluidic centrifuge showed the same physical phenomena with the conventional one even in its small size. Also, a new finding was obtained that the transient motion influenced the stable operation of the microfluidic centrifuge.

This paper presents the centrifugation using the self-rotating secondary flow in the microchamber. The feasibility of the microfluidic centrifuge was demonstrated numerically an experimentally. The centrifugation of the nano-particles was visualized and its performance was characterized quantitatively. Also, a new finding of the transient motion of the rotating flow influenced the stable operation of the centrifuge.

A new treatment of cell-interface convective fluxes in AUSM-type methods is introduced to substantially reduce the numerical dissipation in a smooth region, including a contact discontinuity, without compromising accuracy in shock regions. The core idea of the new scheme is the modification of the transferred property at a cell-interface under the consideration of physical and multi-dimensional phenomena. The transferred property of the M-AUSMPW+ is revealed in two aspects. One is that the newly defined cellinterface value is closer to the real physical value. The other is that it can eliminate numerical dissipation effectively in a non-flow aligned grid system.

*† ‡ On the basis of the Eu’s generalized hydrodynamic (GH) theory for diatomic gas and multi-species gas, computational models are developed for the numerical simulation of hypersonic rarefied gas flows. The rotational non-equilibrium effect of diatomic molecules is taken into account by introducing excess normal stress associated with the bulk viscosity. Starting from the development of the diatomic GH computational model, the multi-species GH theory is applied to a multi-species gas including 5 species; O2, N2, NO, O, N. Two kinds of GH theories (diatomic GH model for single species gas and multi-species GH model for monatomic gas) are combined to derive the general formulation of the multi-species GH theory considering collision between monatomic and diatomic molecules. The multi-species GH model includes diffusion relation due to the molecular collision and thermal phenomena. Two kinds of GH models are developed for an axi-symmetric flow solver. By comparing the computed results of diatomic and multi-species GH theory with those of the Navier-Stokes equations and the DSMC results, the accuracy and physical consistency of the GH computational models are examined.