skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Non-premixed acoustically perturbed swirling flame dynamics

Abstract

An investigation into the response of non-premixed swirling flames to acoustic perturbations at various frequencies (f{sub p}=0-315 Hz) and swirl intensities (S=0.09 and 0.34) is carried out. Perturbations are generated using a loudspeaker at the base of an atmospheric co-flow burner with resulting velocity oscillation amplitudes vertical stroke u'/U{sub avg} vertical stroke in the 0.03-0.30 range. The dependence of flame dynamics on the relative richness of the flame is investigated by studying various constant fuel flow rate flame configurations. Flame heat release rate is quantitatively measured using a photomultiplier with a 430 nm bandpass filter for observing CH* chemiluminescence which is simultaneously imaged with a phase-locked CCD camera. The flame response is observed to exhibit a low-pass filter characteristic with minimal flame response beyond pulsing frequencies of 200 Hz. Flames at lower fuel flow rates are observed to remain attached to the central fuel pipe at all acoustic pulsing frequencies. PIV imaging of the associated isothermal fields show the amplification in flame aspect ratio is caused by the narrowing of the inner recirculation zone (IRZ). Good correlation is observed between the estimated flame surface area and the heat release rate signature at higher swirl intensity flame configurations. A flame responsemore » index analogous to the Rayleigh criterion in non-forced flames is used to assess the potential for a strong flame response at specific perturbation configurations and is found to be a good predictor of highly responsive modes. Phase conditioned analysis of the flame dynamics yield additional criteria in highly responsive modes to include the effective amplitude of velocity oscillations induced by the acoustic pulsing. In addition, highly responsive modes were characterized by velocity to heat release rate phase differences in the {+-}{pi}/2 range. A final observed characteristic in highly responsive flames is a Strouhal number between 1 and 3.5 based on the burner co-flow annulus diameter (St = f{sub p}U{sub avg}/d{sub m}). Finally, wavelet analyses of heat release rate perturbations indicate highly responsive modes are characterized by sustained low frequency oscillations which accompany the high amplitude velocity perturbations at these modes. Higher intensity low frequency heat release rate oscillations are observed for lean flame/low pulsing frequency conditions. (author)« less

Authors:
; ; ;  [1]
  1. Department of Mechanical, Materials and Aerospace Engineering, University of Central Florida, Orlando, FL (United States)
Publication Date:
OSTI Identifier:
21337874
Resource Type:
Journal Article
Resource Relation:
Journal Name: Combustion and Flame; Journal Volume: 157; Journal Issue: 9; Other Information: Elsevier Ltd. All rights reserved
Country of Publication:
United States
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL AND ANALYTICAL CHEMISTRY; FLAMES; HEAT; COMBUSTION KINETICS; DISTURBANCES; OSCILLATIONS; VELOCITY; CONFIGURATION; ASPECT RATIO; FLOW RATE; SURFACE AREA; AMPLIFICATION; CORRELATIONS; ZONES; VORTICES; HZ RANGE; SOUND WAVES; Swirl stabilized flames; Non-premixed flames; Rayleigh criterion; Forced flame response; Wavelet analysis; Acoustic flame perturbation

Citation Formats

Idahosa, Uyi, Saha, Abhishek, Xu, Chengying, and Basu, Saptarshi. Non-premixed acoustically perturbed swirling flame dynamics. United States: N. p., 2010. Web. doi:10.1016/J.COMBUSTFLAME.2010.05.008.
Idahosa, Uyi, Saha, Abhishek, Xu, Chengying, & Basu, Saptarshi. Non-premixed acoustically perturbed swirling flame dynamics. United States. doi:10.1016/J.COMBUSTFLAME.2010.05.008.
Idahosa, Uyi, Saha, Abhishek, Xu, Chengying, and Basu, Saptarshi. Wed . "Non-premixed acoustically perturbed swirling flame dynamics". United States. doi:10.1016/J.COMBUSTFLAME.2010.05.008.
@article{osti_21337874,
title = {Non-premixed acoustically perturbed swirling flame dynamics},
author = {Idahosa, Uyi and Saha, Abhishek and Xu, Chengying and Basu, Saptarshi},
abstractNote = {An investigation into the response of non-premixed swirling flames to acoustic perturbations at various frequencies (f{sub p}=0-315 Hz) and swirl intensities (S=0.09 and 0.34) is carried out. Perturbations are generated using a loudspeaker at the base of an atmospheric co-flow burner with resulting velocity oscillation amplitudes vertical stroke u'/U{sub avg} vertical stroke in the 0.03-0.30 range. The dependence of flame dynamics on the relative richness of the flame is investigated by studying various constant fuel flow rate flame configurations. Flame heat release rate is quantitatively measured using a photomultiplier with a 430 nm bandpass filter for observing CH* chemiluminescence which is simultaneously imaged with a phase-locked CCD camera. The flame response is observed to exhibit a low-pass filter characteristic with minimal flame response beyond pulsing frequencies of 200 Hz. Flames at lower fuel flow rates are observed to remain attached to the central fuel pipe at all acoustic pulsing frequencies. PIV imaging of the associated isothermal fields show the amplification in flame aspect ratio is caused by the narrowing of the inner recirculation zone (IRZ). Good correlation is observed between the estimated flame surface area and the heat release rate signature at higher swirl intensity flame configurations. A flame response index analogous to the Rayleigh criterion in non-forced flames is used to assess the potential for a strong flame response at specific perturbation configurations and is found to be a good predictor of highly responsive modes. Phase conditioned analysis of the flame dynamics yield additional criteria in highly responsive modes to include the effective amplitude of velocity oscillations induced by the acoustic pulsing. In addition, highly responsive modes were characterized by velocity to heat release rate phase differences in the {+-}{pi}/2 range. A final observed characteristic in highly responsive flames is a Strouhal number between 1 and 3.5 based on the burner co-flow annulus diameter (St = f{sub p}U{sub avg}/d{sub m}). Finally, wavelet analyses of heat release rate perturbations indicate highly responsive modes are characterized by sustained low frequency oscillations which accompany the high amplitude velocity perturbations at these modes. Higher intensity low frequency heat release rate oscillations are observed for lean flame/low pulsing frequency conditions. (author)},
doi = {10.1016/J.COMBUSTFLAME.2010.05.008},
journal = {Combustion and Flame},
number = 9,
volume = 157,
place = {United States},
year = {Wed Sep 15 00:00:00 EDT 2010},
month = {Wed Sep 15 00:00:00 EDT 2010}
}
  • Two confined lean non-premixed swirl-stabilized flame typologies were investigated in order to achieve detailed information on the thermal and aerodynamic field in the close vicinity of the burner throat and provide correlation with the exhaust emissions. Previous finding indicated the generation of a partially premixed flame with radial fuel injection and a purely diffusive flame with co-axial injection in a swirling co-flow. In the present work, the experimental study is reported which has been conducted on a straight exit laboratory burner with no quarl cone, fuelled by natural gas and air, and fired vertically upwards with the flame stabilized atmore » the end of two concentric pipes with the annulus supplying swirled air and the central pipe delivering the fuel. Two fuel injection typologies, co-axial and radial (i.e., transverse), leading to different mixing mechanisms, have been characterized through different techniques: particle image velocimetry (PIV) and laser Doppler velocimetry (LDV) for a comprehensive analysis of the velocity field, still photography for the detection of flame front and main visible features, and thermocouples for the temperature distribution. Isothermal flow conditions have been included in the experimental investigation to provide a basic picture of the flow field and to comprehend the modifications induced by the combustion process. The results indicated that, although the global mixing process and the main flame structure are governed by the swirl motion imparted to the air stream, the two different fuel injection methodologies play an important role on mixture formation and flame stabilization in the primary mixing zone. Particularly, it has been found that, in case of axial injection, the turbulent interaction between the central fuel jet and the backflow generated by the swirl can induce an intermittent fuel penetration in the recirculated hot products and the formation of a central sooting luminous plume, a phenomenon totally absent in the case of radial injection. (author)« less
  • A two-dimensional laminar premixed flame is stabilized over a burner in a confined duct and is subjected to external acoustic forcing from the downstream end. The equivalence ratio of the flame is 0.7. The flame is stabilized in the central slot of a three-slotted burner. The strength of the shear layer of the cold reactive mixture through the central slot is controlled by the flow rate of cold nitrogen gas through the side slots. The frequency range of acoustic excitation is 400-1200 Hz, and the amplitude levels are such that the acoustic velocity is less than the mean flow velocitymore » of the reactants. Time-averaged chemiluminescence images of the perturbed flame front display time-mean changes as compared to the unperturbed flame shape at certain excitation frequencies. Prominent changes to the flame front are in the form of stretching or shrinkage, asymmetric development of its shape, increased/preferential lift-off of one or both of the stabilization points of the flame, and nearly random three-dimensional fluctuations over large time scales under some conditions. The oscillations of the shear layer and the response of the confined jet of the hot products to the acoustic forcing, such as asymmetric flow development and jet spreading, are found to be responsible for the observed mean changes in the flame shape. A distinct low-frequency component ({approx}60-90 Hz) relative to the excitation frequency is observed in the fluctuations of the chemiluminescent intensity in the flame under most conditions. It is observed that fluctuations in the flame area predominantly contribute to the origin of the low-frequency component. This is primarily due to the rollup of vortices and the generation of enthalpy waves at the burner lip. Both of these processes are excited at the externally imposed acoustic time scale, but convect/propagate downstream at the flow time scale, which is much larger. (author)« less
  • An open-open organ pipe burner (Rijke tube) with a bluff-body ring was used to create a self-excited, acoustically-driven, premixed methane-air conical flame, with equivalence ratios ranging from 0.85 to 1.05. The feed tube velocities corresponded to Re = 1780-4450. Coupled oscillations in pressure, velocity, and heat release from the flame are naturally encouraged at resonant frequencies in the Rijke tube combustor. This coupling creates sustainable self-excited oscillations in flame front area and shape. The period of the oscillations occur at the resonant frequency of the combustion chamber when the flame is placed {proportional_to}1/4 of the distance from the bottom ofmore » the tube. In this investigation, the shape of these acoustically-driven flames is measured by employing both OH planar laser-induced fluorescence (PLIF) and chemiluminescence imaging and the images are correlated to simultaneously measured pressure in the combustor. Past research on acoustically perturbed flames has focused on qualitative flame area and heat release relationships under imposed velocity perturbations at imposed frequencies. This study reports quantitative empirical fits with respect to pressure or phase angle in a self-generated pressure oscillation. The OH-PLIF images were single temporal shots and the chemiluminescence images were phase averaged on chip, such that 15 exposures were used to create one image. Thus, both measurements were time resolved during the flame oscillation. Phase-resolved area and heat release variations throughout the pressure oscillation were computed. A relation between flame area and the phase angle before the pressure maximum was derived for all flames in order to quantitatively show that the Rayleigh criterion was satisfied in the combustor. Qualitative trends in oscillating flame area were found with respect to feed tube flow rates. A logarithmic relation was found between the RMS pressure and both the normalized average area and heat release rate for all flames. (author)« less
  • A laboratory-scale swirling burner, presenting many similarities with gas turbines combustors, has been studied experimentally using planar laser induced fluorescence (PLIF) on OH radical and acetone vapor in order to characterize the flame stabilization process. These diagnostics show that the stabilization point rotates in the combustion chamber and that air and fuel mixing is not complete at the end of the mixing tube. Fuel mass fraction decays exponentially along the mixing tube axis and transverse profiles show a gaussian shape. However, radial pressure gradients tend to trap the fuel in the core of the vortex that propagates axially in themore » mixing tube. As the mixing tube vortex enters the combustion chamber, vortex breakdown occurs through a precessing vortex core (PVC). The axially propagating vortex shows a helicoidal trajectory in the combustion chamber which trace is observed with transverse acetone PLIF. As a consequence, the stabilizing point of the flame in the combustion chamber rotates with the PVC structure. This phenomenon has been observed in the present study with a high speed camera recording spontaneous emission of the flame. The stabilization point rotation frequency tends to increase with mass flow rates. It was also shown that the coupling between the PVC and the flame stabilization occurs via mixing, explaining one possible coupling mechanism between acoustic waves in the flow and the reaction rate. This path may also be envisaged for flashback, an issue that will be more completely treated in a near future. (author)« less
  • Cited by 1