Pulsating Flame Research Articles

Influence of low air pressure on combustion characteristics and flame pulsation frequency of pool fires

To investigate the effect of low air pressure on the combustion characteristics and puffing flame frequency of pool fires, ethanol and n-heptane pool fires were performed using 15 square burners of various size in both Lhasa (altitude: 3650m; air pressure: 65kPa) and Hefei (altitude: 24m, air pressure: 100.8kPa) fire laboratories. Comparison of the experimental results for pool fires of the same size in the two places shows that, firstly, the maximum rise in the centerline temperature in flames in Lhasa is generally larger than that in Hefei. Secondly, the dependence of the burning rate exponent n (m˙″∝pn) on the air pressure varies with the equivalent diameter D of the burner, with n<0 (D<7cm), (0–1) (7cm<D<10cm), (1–1.45) (10cm<D<19cm) and 1 (D>19cm). Thirdly, radiation fraction of the pool fire flames is smaller at low air pressure. Finally, the puffing frequency of the pool fire flames is higher at low air pressure. Compared with Hefei, in Lhasa, for the same size burner, a higher maximum temperature together with less radiation from the flame is observed. This shows that in low air pressure, the pool fire is more buoyant, which leads to stronger periodic oscillation and a higher flame puffing frequency.

Propagation, dynamics and control of laminar premixed flames

Current state of knowledge of laminar premixed flame propagation, dynamics and control is reviewed. A distinction is made between planar and multi-dimensional flames. Based on the well-established theory of flame propagation of free adiabatic flames, different mechanisms to control planar flames, mainly based on heat transfer processes in porous media, but also using electrical and gravitational fields, are considered. Flame instabilities, leading to flame pulsations or oscillations, for instance in an acoustic field, are also considered. A simple physical model for the interaction between a premixed flame and a burner is presented to explain the main physical mechanisms of planar instabilities of burner-stabilized flames. Knowledge on flame propagation of planar flames is well-established and physical models to describe the behavior of these flames are mature, although it is expected that more sophisticated methods will be used in future research to control their behavior.Multi-dimensional flames are also well understood, although the state of knowledge is less mature. Models for the interaction of flames with flow distortions, leading to flame stretch and curvature are derived. The influence of such distortions on the burning velocity and flame dynamics in terms of the Markstein number is studied. Weak-stretch models are available but multiple species chemistry and transport effects are less well understood, especially for rich and stoichiometric mixtures. Some ways to extend or improve current models are indicated. Current knowledge on the interaction of premixed (Bunsen-type) flames with flow distortions, such as acoustic waves, is also reviewed. Experimentally observed phenomena can be modeled satisfactory using detailed numerical models. However, for this research area, it is clear that current physical models or analytical descriptions are less mature because the origin of some phenomena is still largely unknown. Extensions of current (mostly G-equation like) models to account for flame–flow interaction, flame–flame interaction and flame–burner interaction are needed. To conclude: a multiple set of measures to adapt the flames to the combustion system of interest is at hand and different ways to influence the flame by heat transfer, acoustic waves and other less often used control mechanisms like electric fields are available. This is needed to modify and control premixed laminar flames in practical devices to guarantee safe and proper operation.

Flame dynamics of a meso-scale heat recirculating combustor

The dynamics of premixed propane–air flame in a meso-scale ceramic combustor has been examined here. The flame characteristics in the combustor were examined by measuring the acoustic emissions and preheat temperatures together with high-speed cinematography. For the small-scale combustor, the volume to surface area ratio is small and hence the walls have significant effect on the global flame structure, flame location and flame dynamics. In addition to the flame–wall thermal coupling there is a coupling between flame and acoustics in the case of confined flames. Flame–wall thermal interactions lead to low frequency flame fluctuations (∼100 Hz) depending upon the thermal response of the wall. However, the flame–acoustic interactions can result in a wide range of flame fluctuations ranging from few hundred Hz to few kHz. Wall temperature distribution is one of the factors that control the amount of reactant preheating which in turn effects the location of flame stabilization. Acoustic emission signals and high-speed flame imaging confirmed that for the present case flame–acoustic interactions have more significant effect on flame dynamics. Based on the acoustic emissions, five different flame regimes have been identified; whistling/harmonic mode, rich instability mode, lean instability mode, silent mode and pulsating flame mode.

Stabilized three-stage oxidation of DME/air mixture in a micro flow reactor with a controlled temperature profile

Ignition and combustion characteristics of a stoichiometric dimethyl ether (DME)/air mixture in a micro flow reactor with a controlled temperature profile which was smoothly ramped from room temperature to ignition temperature were investigated. Special attention was paid to the multi-stage oxidation in low temperature condition. Normal stable flames in a mixture flow in the high velocity region, and non-stationary pulsating flames and/or repetitive extinction and ignition (FREI) in the medium velocity region were experimentally confirmed as expected from our previous study on a methane/air mixture. In addition, stable double weak flames were observed in the low velocity region for the present DME/air mixture case. It is the first observation of stable double flames by the present methodology. Gas sampling was conducted to obtain major species distributions in the flow reactor. The results indicated that existence of low-temperature oxidation was conjectured by the production of CH 2O occured in the upstream side of the experimental first luminous flame, while no chemiluminescence from it was seen. One-dimensional computation with detailed chemistry and transport was conducted. At low mixture velocities, three-stage oxidation was confirmed from profiles of the heat release rate and major chemical species, which was broadly in agreement with the experimental results. Since the present micro flow reactor with a controlled temperature profile successfully presented the multi-stage oxidations as spatially separated flames, it is shown that this flow reactor can be utilized as a methodology to separate sets of reactions, even for other practical fuels, at different temperature.

Development of Optical Measurement Techniques for Thermo-Acoustic Diagnostics: Fibre-Optic Microphone, Rayleigh-Scattering, and Acoustic PIV

Thermo-acoustic investigations require reliable measurement techniques in hot environments for pressure, density fluctuations with a high dynamic range and acoustic particle velocity. This paper presents recent developments of optical measurement techniques in combustion diagnostics. A fibre-optic microphone based on the interferometric detection of membrane deflections was designed to measure acoustic pressure oscillations. Due to the heat resistant design, the sensor has an upper temperature limitation of approximately 970 K. Rayleigh-Scattering measurements, using the density dependent intensity of scattered light were performed in an unconfined flame with approximately 1600 K to study amplitude and phase distribution of the flame pulsation. Acoustic particle velocity can be determined applying acoustic PIV (particle image velocimetry) technique. This paper shows a way to measure simultaneously the acoustic particle velocity and the locally resolved mean flow velocity of a turbulent flow. Together these non-invasive techniques are applicable to study thermo-acoustic processes and sound generation in combustion chambers or turbines.

CO/NOx Emissions of Strongly Pulsed Jet Diffusion Flames

The CO and NOX emissions of strongly pulsed, turbulent diffusion flames were examined experimentally in a co-flow combustor. Video imaging was performed and time-averaged emissions were measured at the combustor exit and near the visible flame tip. Both the case of a fixed fuel injection velocity during the injection interval and a constant fuelling rate were studied for jet Reynolds numbers ranging from 5,000 to 15,000. For fixed injection velocity, maximum emission indices of CO occurred for compact, isolated flame puffs. CO decreased substantially with decreasing jet-off time as the flame puff interaction increased. The pulsed flames had lower NOX than the steady flames, particularly for the case of isolated flame structures. Similar trends in NO formation were seen for constant fueling rate. The CO emissions were, however, considerably different, largely due to a significant impact of the Reynolds number. Radial emissions profiles suggest an improved fuel/air mixing for shorter jet-off times and longer jet-on times. The correlation of CO/NO emissions with a global flame residence time is discussed.

Flame Spread across Shallow Pools in Modulated Opposed Air Flow in Narrow Tube

The effect of forced, opposed air flow on flame spread across shallow n-butanol pools in a narrow tube was studied experimentally. The flame spread was accompanied by pulsations, whose amplitude depended on the opposed air velocity. Dependences of the average flame speed on temperature and the opposed air velocity were obtained. The average flame speed was found to be independent of the liquid depth ahead of the flame within the experimental error. Fourier analysis has shown there are natural frequencies of pulsating of the flame. The effects of modulating frequency of forced, opposed air flow on average flame speed was studied experimentally. Response of the flame to the modulation of the opposed air velocity depends on the modulating frequency. If the modulating frequency is more than the natural frequency of pulsations of the flame then the average flame speed depends on time; at first the average flame speed increases and then gradually the value of speed comes back to the value without modulation. If frequency is close to natural frequencies of flame pulsations then the average flame speed increases.

Scaling Analysis on Pulsating Flame Spread over Liquids

Scaling analyses based on subsurface layer instability were performed to explore the role of three independent (surface tension, gravity, and viscosity) influences on the mechanism of pulsating flame spread under normal and microgravity conditions. These three influences form two independent pi-numbers: the Marangoni (Ma) number and Grashof (Gr) number, which include the characteristic length scale ratio (depth of subsurface circulation)/(horizontal length of preheated liquid surface). The Prandtl (Pr) number was introduced to compensate for the different thermal diffusivity and kinematic viscosity of different liquids. Also a nondimensional flame spread rate,V/VD(=Vδ/D, whereδis the quenching distance andDis the diffusivity of fuel vapor) was introduced. Using these nondimensional parameters, the flame spread mechanism was divided into two separate regimes: for the shallow liquid pool the nondimensional flame spread rate was correlated with{Gr0.2/(Ma·Pr)}1.0, while for the deep liquid pool it was correlated with{Gr0.2/(Ma·Pr)}1.5.

Detailed characterization of the dynamics of thermoacoustic pulsations in a lean premixed swirl flame

A nozzle configuration for technically premixed gas turbine flames was operated with CH 4 and air at atmospheric pressure. The flames were confined by a combustion chamber with large quartz windows, allowing the application of optical and laser diagnostics. In a distinct range of operating conditions the flames exhibited strong self-excited thermoacoustic pulsations at a frequency around 290 Hz. A flame with P = 25 kW thermal power and an equivalence ratio of Φ = 0.7 was chosen as a target flame in order to analyze the dynamics and the feedback mechanism of the periodic instability in detail. The velocity field was measured by three-component laser Doppler velocimetry, the flame structures were measured by chemiluminescence imaging and planar laser-induced fluorescence of OH, and the joint probability density functions of major species concentrations, mixture fraction, and temperature were measured by laser Raman scattering. All measuring techniques were applied in a phase-locked mode with respect to the phase angle of the periodic pulsation. In addition to the pulsating flame, a nonpulsating flame with increased fuel flow rate ( P = 30 kW , Φ = 0.83 ) was studied for comparison. The measurements revealed significant differences between the structures of the pulsating and the nonpulsating (or “quiet”) flame. Effects of finite-rate chemistry and unmixedness were observed in both flames but were more pronounced in the pulsating flame. The phase-locked measurements revealed large variations of all measured quantities during an oscillation cycle. This yielded a clear picture of the sequence of events and allowed the feedback mechanism of the instability to be identified and described quantitatively. The data set presents a very good basis for the verification of numerical combustion simulations because the boundary conditions of the experiment were well-defined and the most important quantities were measured with a high accuracy.

Flame–sound interaction in tubes with nonslip walls

Flame interaction with sound is studied for a premixed flame propagating to the closed end of a tube with nonslip walls. The flow geometry is similar to that in the classical Searby experiments on flame–acoustic interaction [Combust. Sci. Technol. 81 (1992) 221]. The problem is solved by direct numerical simulations of the combustion equations. The flame–sound interaction strongly influences oscillations of the flame front. Particularly, sound noticeably increases the oscillation amplitude in comparison with that in an open tube with nonreflecting boundary conditions at the ends studied previously. Oscillations become especially strong in the second part of the tube, where flame pulsations are in resonance with the acoustic wave. Parameters of the flame oscillations are investigated for different values of the tube width and length. It is demonstrated that the oscillations are stronger in wider tubes, though the investigated tube width is limited by the computational facilities. In sufficiently wide tubes, violent folding of a flame front is observed because of the flame–acoustic resonance. By increasing the Lewis number, one also increases the oscillation amplitude.

Effects of thin-layer boilover on flame geometry and dynamics in large hydrocarbon pool fires

This work aims to estimate the effects of thin-layer boilover on flame geometry and dynamics. A series of large scale experiments (in pools ranging from 1.5 to 6 m in diameter) were performed using gasoline and diesel as fuel. As expected, only diesel showed evidence of this phenomenon. This article presents a summary of the results obtained for flame height, tilt and pulsation. Flame height increases during water ebullition, though the increase is no longer detectable when wind speed exceeds certain values. Correlations previously presented in the literature to predict flame length and tilt were modified in order to fit the results obtained during thin-layer boilover. However, the influence on flame tilt is not as great and the equations for the stationary period seem suitable for the entire fire. Results of flame pulsation during the stationary period fill the gap in the literature for fires between 1.5 and 6 m and fit previous correlations. On the other hand, during ebullition, the flame pulsates faster, as air entrainment is greater and, as one would expect, this effect decreases with pool size. A new equation for estimating pulsation frequency during boilover is proposed.

Mixing characteristics in strongly forced non-premixed methane jet flames

Previous and current work show that flame length and soot luminosity of laminar and turbulent non-premixed jet flames can be significantly reduced by high-frequency, high-amplitude forcing of the fuel flow rate. The current work focuses on understanding the physical mechanisms responsible for these changes to the flow/flame structure for the case where the forcing is sufficiently strong that significant reverse flow exists at the exit of the fuel delivery tube. The high-amplitude forcing is achieved by pulsing the flow at the organ-pipe resonant frequency of the fuel delivery tube. Quantitative mixture fraction imaging in non-reacting jets indicates that the strongly pulsed jets exhibit dramatically enhanced mixing as compared to unpulsed jets. In the pulsed jets, the mean mixture fraction falls below 0.08 in as little as 3 D downstream of the nozzle exit and no pure jet fluid exists outside of the nozzle due to in-tube mixing. Simultaneous OH and acetone PLIF performed in methane jet flames shows that the OH zones are much broader and more wrinkled in the resonantly pulsed jet as compared to the unpulsed jet, and are greatly contorted by the vortical structures. A unique feature of the pulsed flames is that the reaction zones appear to close downstream of a vortical structure, just a few diameters downstream of the nozzle exit, in a region where the mixture fraction imaging (for non-reacting flows) shows reduced mixture fraction. Furthermore, the flame anchors on the outer-upstream edge of the vortical structures where the fuel mixture fraction is reduced due to enhanced entrainment. The significant in-tube premixing and the enhanced entrainment appear to be the predominant mechanisms that cause the reduction in length and luminosity of high-amplitude, high-frequency pulsed non-premixed jet flames.

Phase-Resolved Laser Diagnostic Measurements of a Downscaled, Fuel-Staged Gas Turbine Combustor at Elevated Pressure and Comparison to LES Predictions

A technical gas turbine combustor has been studied in detail with optical diagnostics for validation of large-eddy simulations (LES). OH* chemiluminescence, OH laser-induced fluorescence (LIF) and particle image velocimetry (PIV) have been applied to stable and pulsating flames up to 8 bar. The combination of all results yielded good insight into the combustion process with this type of burner and forms a database that was used for the validation of complex numerical combustion simulations. LES, including radiation, convective cooling, and air cooling, were combined with a reduced chemical scheme that predicts NOx emissions. Good agreement of the calculated flame position and shape with experimental data was found.

Diffusive-thermal instabilities of diffusion flames: onset of cells and oscillations

A comprehensive stability analysis of planar diffusion flames is presented within the context of a constant-density model. The analysis provides a complete characterization of the possible patterns that are likely to be observed as a result of differential and preferential diffusion when a planar flame becomes unstable. A whole range of physical parameters is considered, including the Lewis numbers associated with the fuel and the oxidizer, the initial mixture fraction, and the flow conditions. The two main forms of instability are cellular flames, obtained primarily in fuel-lean systems when the Lewis numbers are generally less than one, and planar pulsations, obtained in fuel-rich systems when the Lewis numbers are generally larger than one. The cellular instability is predominantly characterized by stationary cells of characteristic dimension comparable to the diffusion length, but smaller cells that scale on the reaction zone thickness are also possible near extinction conditions. The pulsating instability is characterized by planar oscillations normal to the flame sheet with a well-defined frequency comparable to the reciprocal of the diffusion time; high-frequency modes are also possible just prior to extinction. The analysis also alludes to other possible patterns, such as oscillating cellular structures, which result from competing modes of instability of comparable and/or disparate scales. The expected pattern depends of course on the underlying physical parameters. Consequently, stability boundaries have been identified for the onset of one or another form of the instability. The conditions for the onset of cellular and pulsating flames, as well as the predicted cell size and the frequency of oscillations, compare well with the experimental record.

Phase-locked two-line OH planar laser-induced fluorescence thermometry in a pulsating gas turbine model combustor at atmospheric pressure

Two-line OH planar laser-induced fluorescence (PLIF) thermometry was applied to a swirling CH4/air flame in a gas turbine (GT) model combustor at atmospheric pressure, which exhibited self-excited combustion instability. The potential and limitations of the method are discussed with respect to applications in GT-like flames. A major drawback of using OH as a temperature indicator is that no temperature information can be obtained from regions where OH radicals are missing or present in insufficient concentration. The resulting bias in the average temperature is addressed and quantified for one operating condition by a comparison with results from laser Raman measurements applied in the same flame. Care was taken to minimize saturation effects by decreasing the spectral laser power density to a minimum while keeping an acceptable spatial resolution and signal-to-noise ratio. In order to correct for the influence of laser light attenuation, absorption measurements were performed on a single-shot basis and a correction procedure was applied. The accuracy was determined to 4%-7% depending on the location within the flame and on the temperature level. A GT model combustor with an optical combustion chamber is described, and phase-locked 2D temperature distributions from a pulsating flame are presented. The temperature variations during an oscillation cycle are specified, and the general flame behavior is described. Our main goals are the evaluation of the OH PLIF thermometry and the characterization of a pulsating GT-like flame.

Pulsating and hydrodynamic instabilities at large Lewis numbers

The dynamic behavior of freely propagating premixed flames with large Lewis numbers was computationally simulated using a sixth-order central difference scheme and nonreflective boundary conditions. Results in the linear stage of the instability growth show that the growth rate dramatically decreases with increasing Lewis number and that the large activation energy excites the pulsating instability and increases the growth rate of the hydrodynamic instability. In the nonlinear growth stage, there exist regimes of stable cell propagation, periodic pulsating cellular flames, and irregular pulsating cellular flames as the activation energy is increased. Characteristics of these regimes were further studied for the effects of Lewis number on the flame front structure in the regime of stable cell propagation, the effects of flame pulsation on the flow and flame cell structure in the regime of periodic pulsating cellular flame, and the complex pattern formation in the regime of irregular pulsating cellular flame. It is further demonstrated that unsteady pulsating flames can propagate faster than the adiabatic flame when the local stretch rate is positive, implying that models based on quasi-steady flame propagation may not correctly predict the behavior of unsteady flames with large Lewis numbers.

Adaptive numerical simulation of pulsating planar flames for large Lewis and Zeldovich ranges

We study numerically the behaviour of pulsating planar flames in the thermo-diffusive approximation. The numerical scheme is based on a finite volume discretization with an adaptive multi-resolution technique for automatic grid adaption. This allows an accurate and efficient computation of pulsating flames even for very large activation energies. Depending on the Lewis number and the Zeldovich number, we observe different behaviours, like stable or pulsating flames, the latter being either damped, periodic, or a-periodic. A bifurcation diagram in the Lewis–Zeldovich plane is computed and our results are compared with previous computations [Rogg B. The effect of Lewis number greater than unity on an unsteady propagating flame with one-step chemistry. In: Peters N, Warnatz J, editors, Numerical methods in laminar flame propagation, Notes on numerical fluid mechanics, vol. 6. Vieweg; 1982. p. 38–48.] and theoretical predictions [Joulin G, Clavin P. Linear stability analysis of nonadiabatic flames: diffusional-thermal model. Combust Flame 1979;35:139–53]. For Lewis numbers larger than 6 we find that the stability limit is again increasing towards larger Zeldovich numbers and not monotonically decreasing as predicted by the asymptotic theory. A study of the flame velocities for different Zeldovich numbers shows that the amplitude of the pulsations strongly varies with the Lewis number. A Fourier analysis yields information on their frequency.

Effects of carbon nanotubes on flame spread rate over 1-propanol

This paper discusses the potential for thermally conductive nanoparticles to reduce flame spread velocities over fuel dispersion mixtures. Multi-walled carbon nanotubes (MWNTs) were dispersed in 1-propanol at levels above the mixture's percolation limit, 0.5, 1.0, 2.0, 4, 8, and 12 wt%, and tested in a rectangular fuel tray and a shallow pan flame spread system. The MWNTs/1-propanol mixtures had much lower flame spread rates than the original fuel itself at temperatures below 1-propanol's flash point. A 2 wt% MWNT loading reduced the flame spread rate to 9–15% of the neat fuel rate (the corresponding increase in the average flame spreading time over a 16 cm horizontal test length is approximately 6–11 times) and significantly inhibited the pulsating flame spread behavior. Flame spread rate was calculated for both the pure fuel and MWNT mixtures with a surface-tension-driven spread rate equation. The calculations favorably compare with corresponding measured values despite several assumptions made. Thermally conductive high aspect ratio nanoparticles could be effective flame spread suppressants for liquid fuels.

Flame structure of steady and pulsed sooting inverse jet diffusion flames

In turbulent buoyant fire plumes, local inverse diffusion flames of air injected into gaseous fuel or fuel vapors occur, but little is known about the tendency to form soot and produce thermal radiation in these flame structures. To investigate these phenomena, steady and pulsed normal and inverse jet diffusion flames of methane/air and ethylene/air have been stabilized on a slot burner, which has advantages over a coannular flame geometry for performing flame imaging measurements in sooty flames. OH and PAH laser-induced fluorescence (LIF), soot laser-induced incandescence (LII), and soot thermal emission at 850 nm have been measured in the lower flame region of steady and pulsed flames. These measurements reveal that the relative positions of these different structural features are very similar in the normal and inverse steady flames of each fuel. Also, the OH signals are nearly identical in the normal and inverse flames. The inverse flame PAH signals and soot concentrations are somewhat smaller than for the normal flames, and the near-infrared radiation is approximately 25% lower for the inverse flame. When the central slot is pulsed, the primary buoyant vortex roll-up occurs on the fuel-rich side of inverse flames, resulting in strongly enhanced PAH signals and soot concentrations. The near-infrared radiation also increases in the pulsed flames, but not from the soot within the vortex roll-up region. In general, enhancements in peak signals from soot and near-infrared radiation similar to those in pulsed normal diffusion flames are apparent in pulsed inverse diffusion flames. PAH signals are clearly greatest in the pulsed inverse flames.

Scaling and instability analyses on flame spread over liquids

Stability and scaling analyses were applied to experimental data obtained by this group and other researchers on pulsating flame spread over liquids. Data to be analyzed include recent findings of cyclic appearance of a cold temperature valley at the liquid surface-created surface-wave ahead of the spreading flame, and main-pulsation of 0.5–2 Hz and sub-pulsation of 5–10 Hz. Our stability analysis is performed to understand the mechanism of instability on the liquid surface ahead of a flame’s leading edge, which is thought of as the major cause for pulsating flame spread. The scaling analysis is performed to explore the role of four independent (gravity, surface-tension, viscose, and inertia) forces on the mechanisms of flame spread. These four forces form three independent pi-numbers: Marangoni ( Ma) number, Weber ( We) number, and Froude ( Fr) number, all of which include the critical length scale ratio: (height of sub-surface circulation)/(horizontal length of preheated liquid surface). We combined the wave equation obtained from the stability analysis, the three pi-numbers, and the critical length scale ratio, and used them as a universal formula to describe flame spread over liquids. Using this formula, flame spread mechanism over four different types of alcohols was divided into two separate regimes: the thin liquid pool and the thick-liquid pool. For the thin liquid pool, the flame spread rate was correlated with ( Fr/Ma 0.5) −1.0, while for the thick-liquid pool it was correlated with ( Fr/Ma 0.5) −1.5. Change of flame spread pattern from the uniform to the pulsating can be described with temperature difference between the flash point and bulk liquid temperature. For the thin liquid pool this temperature difference is correlated with Ma −0.5, while for the thick-liquid pool it is correlated with Ma −1. The frequency of pulsation is correlated with We −1.0 for the thin liquid pool, while it is correlated with We −1.5 for the thick-liquid pool.