Yingwen Yan

Unsteady flow and heat-release fluctuations downstream of afterburner flameholders can trigger combustion instability. Experiments were conducted to investigate pressure fluctuations, velocity fluctuations, and flame evolution behind a bluff body. Under non-reacting conditions, the dominant fluctuation frequency increased with inlet Mach number and temperature, while the Strouhal number remained within 0.23–0.26. Velocity fluctuations were mainly concentrated in the shear layers, near the end of the recirculation zone, and in the wake. The dominant mode exhibited alternating vortex shedding with opposite phases on the two sides, whereas a second-harmonic response appeared near the centerline. Under reacting conditions, the time-averaged flame exhibited M-shaped and V-shaped configurations depending on inlet Mach number and fuel–air ratio, reflecting the coupled effects of atomization, evaporation, mixing, and wake entrainment. Flame pulsations were synchronized with wake vortex shedding, but their dominant frequency was slightly lower than that of the corresponding non-reacting flow, indicating that heat-release-induced thermal expansion and density gradients weaken large-scale vortex development. These results identify bluff-body wake vortex shedding as an important hydrodynamic source of unsteady heat release and combustion instability in afterburners.

This study presents a numerical simulation framework that integrates turbulence, combustion chemistry, and soot formation within OpenFOAM. A non-adiabatic, dual-scale flamelet-generated manifold model—hereafter dual-scale FGM (DS-FGM)—was developed and integrated with the method of moments with interpolative closure. The combined approach enables quantitative prediction of polycyclic aromatic hydrocarbons (PAHs) evolution, PAHs-to-soot conversion, and soot formation. To account for radiative heat loss and the disparate chemical time scales among species, a normalized enthalpy-loss parameter dH was introduced into the conventional FGM framework. Two progress variables were employed to describe separately the evolution of major species and PAHs, thereby improving the representation capability of PAHs formation, accumulation, and transport in the Flamelet database. The numerical method was validated against a laboratory-scale ethylene turbulent diffusion flame; the simulations reproduced the spatial distribution characteristics of soot evolution. Results show that the proposed soot formation modeling framework can reasonably capture the soot formation and evolution processes in typical turbulent non-premixed jet flames. Comparison with experimental data indicates satisfactory accuracy for the temperature and soot volume fraction distributions. The root-mean-square error and normalized peak error analyses were performed to further quantify the deviations in the overall distribution and peak prediction of the soot volume fraction. However, the soot volume fraction is slightly underpredicted in some regions. This discrepancy likely arises from a combination of factors, including biases in precursors formation and soot nucleation caused by simplified PAH growth pathways in the gas-phase PAHs mechanism, local temperature prediction deviations, and modeling errors in turbulence–chemistry–soot interactions.

Using a laboratory-scale, swirling non-premixed flame, we combine experiments, large eddy simulation-acoustic simulations, and phase-closure analysis to compare forced and self-sustained regimes. The results show that although the velocity fields upstream and immediately downstream of the swirler remain similar, significant discrepancies emerge within the combustor: in the forced system, velocity perturbations originate from upstream acoustic forcing, whereas in the self-sustained system, the flame itself acts as a volumetric expansion source. In addition, considering the oscillatory fuel jet induced by back-pressure oscillations in the self-sustained regime, acoustic simulations indicate that the jet-velocity pulsation at the porous fuel nozzle reaches about 30% of the mean velocity. Even at this level, simulations show the resulting flame response is relatively modest. The differing combustor acoustic fields, together with the periodic fuel pulsation, lead to pronounced differences between forced and self-sustained oscillations. Consequently, the phase transfer between oscillating velocity, pressure, and heat release rate fluctuations differs markedly between the two regimes. Under weak oscillations and short combustor lengths, the flame responses obtained from forced oscillations reasonably approximate those in self-sustained oscillations. However, at higher oscillation amplitudes or longer combustor lengths, strong velocity fluctuations within the chamber significantly distort flame pulsations, where phase closure error was found to exceed π/2 radian. Even with careful selection of the velocity reference plane in forced oscillation tests, the pronounced differences in the acoustic field and fuel pulsation introduce new challenges for flame transfer function/flame describing function studies.

A longitudinal self-sustained combustion oscillation in a bluff-body-stabilized propane diffusion flame was experimentally suppressed using quarter-wave resonators. The baseline combustor exhibited a limit-cycle oscillation at 207.28 Hz, with a pressure amplitude of approximately 972.56 Pa and nearly in-phase pressure and heat release rate fluctuations. A nonlinear acoustic-impedance model of the QW was developed and validated by impedance-tube measurements; within the frequency range most relevant to the resonator function, most amplitude and phase errors remained within 5%. The experiments show that a single QW provides narrowband control and may either suppress or amplify the oscillation depending on its resonance frequency relative to the preexisting combustion oscillations frequency. To quantify this behavior, three nondimensional parameters were introduced: the normalized QW resonance frequency, normalized oscillation pressure amplitude, and normalized oscillation frequency. The optimal suppression cases were mainly located within −0.15 < ηfa < −0.07, whereas amplification became more likely when ηfa > −0.07. The strongest amplification occurred in case 3, where the oscillation pressure amplitude increased by 17.7%. Under the optimal 12-QW configuration, the pressure amplitude was reduced by 77.8% (13.1 dB), and the original single-mode limit cycle collapsed into weaker low-frequency components around 140 and 155 Hz. CH* dynamic mode decomposition further showed that the suppressed state was accompanied by an asymmetric projected flame-pulsation structure, which was qualitatively consistent with the simulated redistribution of transverse acoustic velocity near the resonator sidewall. These results indicate that QW suppression is governed primarily by resonator-induced acoustic-pressure-response redistribution and the associated modification of the flame–acoustic feedback loop, rather than by the sound absorption coefficient alone.

This study investigates the effects of inlet rotation angle on the lean ignition and blowout characteristics of an integrated afterburner under high-temperature rotating inflow conditions. Ignition and blowout experiments are conducted in an integrated afterburner with a single strut-cavity structure. The lean ignition and blowout fuel–air ratios are measured at strut trailing-edge Mach numbers ranging from 0.20 to 0.35 and rotation angles from 0° to 15°. Numerical simulations are also performed to analyze the effects of inlet rotation angle on the flow field structure, fuel atomization and mixing, and combustion characteristics. The results show that: 1) as the inlet rotation angle increases, the lean ignition and blowout fuel–air ratios follow a nonmonotonic trend, initially decreasing and then increasing, with optimal performance occurring at a rotation angle of 5°; 2) a small rotation angle enhances flame retention within the recirculation zone, promotes secondary fuel atomization and mixing, and improves local combustion conditions; 3) a larger rotation angle weakens the recirculation zone downstream of the strut, shifts the shear layer outward, and increases local heat dissipation, raising the risk of flame blowout. This study provides a valuable reference for optimizing combustion organization and designing stable lean-blowout combustion in integrated afterburners.

Bluff body non-premixed flames are widely used in afterburners of aviation engines, where combustion oscillations pose significant challenges to operational stability. A self-sustained combustion oscillation system and a flame dynamic response measurement system were employed to analyze the effects of fuel staging on oscillation amplitude, frequency, and flame dynamic response. The results demonstrated that increasing the primary fuel ratio from 0.6 to 0.88 effectively suppressed combustion oscillations, reducing pressure oscillation amplitude by 65.6%, while the dominant oscillation frequency remained nearly unchanged in the range of 176–180 Hz. Mechanistically, adjustments in fuel staging significantly modified the flame length, fuel injection depth, and recirculation zone size, resulting in pronounced spatial variations in heat release rate fluctuations. These alterations introduced damping regions within the flame, reducing the coherence of heat release oscillations. Additionally, flame response analysis via the flame transfer function and flame describing function showed that increasing the primary fuel ratio shifted the phase delay of heat release rate fluctuation relative to velocity fluctuation, fundamentally altering the thermoacoustic coupling. The results indicate that the primary fuel ratio is the key control parameter for oscillation suppression in the present study, whereas the oscillation frequency is determined mainly by the combined effect of combustor acoustic characteristics and flame-response phase behavior. The study provides essential insights into optimizing fuel staging strategies and offers guidance to suppress combustion oscillation in afterburners.

The study evaluates three hydrogen blending strategies (Substitution, Addition, and Equal-power) with blending ratios up to 60% in a bluff-body stabilized non-premixed swirling flame and examines their effects on soot emissions. High-speed simultaneous measurements of the flow field, reaction zone, and soot distribution were performed using particle image velocimetry, OH*-chemiluminescence and Rayleigh scattering. The results show that the mean flow field remains largely unchanged for three strategies, whereas the reaction zone shifts downstream by ∼18% as the blending ratio increases to 50%. The soot load monotonically increases with hydrogen addition, whereas it first rises and then declines under the substitution strategy, reaching a maximum at a 40% blending ratio. The probability of soot presence follows a similar trend. Whether fuel power increases or remains constant, hydrogen blending increases soot load, likely due to a local temperature rise. Soot residence time isopleths reveal three distinct transport paths: a loop in the recirculation zone, deposition onto the bluff body, and soot leakage, which mainly occurs near the bluff body during flame liftoff, as captured by high-speed simultaneous measurements. Soot leakage occurs heavily at a 20% hydrogen blending ratio for the substitution strategy, and an increasing trend is observed for the other two strategies, which is governed by the combined effects of soot load and flow-field-controlled leakage probability. We make a unique contribution by proposing the mechanism behind it, which is essential for decarbonizing existing combustors that will continue to rely on hydrocarbon-based fuels for the foreseeable future.

Thermoacoustic instabilities pose significant challenges to the safety and efficiency of modern aero-engines, often triggered by complex inlet conditions and flame dynamics. This study explores how the acoustic-phase relationship at the flame location influences the onset and suppression of self-sustained oscillations within a scaled model combustor featuring a bluff-body flameholder. By systematically adjusting the flame’s axial position, we examine its effect on the development of longitudinal acoustic modes. Combining high-speed chemiluminescence imaging with pressure measurements, we reveal how variations in the flame’s acoustic phase modulate the energy transfer mechanisms responsible for oscillation growth. Complementary acoustic simulations, integrating a flame-response model with the Helmholtz equation, identify the precise conditions under which the flame’s acoustic characteristics resonate with the combustor’s natural modes. Our findings demonstrate that repositioning the flame alters the acoustic impedance experienced by the flame—a key factor influencing oscillation amplitude and stability. Specifically, the magnitude of the impedance governs the pressure oscillation strength, while its phase determines the likelihood of instability onset and mode coupling. Detailed acoustic field analyses provide valuable insights into the mechanisms underlying thermoacoustic oscillation initiation and offer pathways for targeted control strategies. Ultimately, this research underscores the critical role of flame acoustic-phase management in advancing stable, high-performance combustion systems.

In this study, a high-speed camera was used to investigate the effects of auxiliary atomization methods on the spray characteristics of a direct-injection nozzle operating under high-temperature and large-Mach-number conditions. Three configurations were examined: (1) adjusting the coflow pressure of circular holes, (2) varying the nozzle insertion depth, and (3) altering the relative position between the circular-corner rectangular coflow holes and the nozzle. The quantitative analysis focused on the breakup length, fuel concentration, droplet size distribution, and oscillation behavior. The results are summarized as follows: (1) The flow boundary layer is thinner than the thermal boundary layer; thus, increasing the nozzle insertion depth decreases the penetration breakup length, while the impinging breakup length remains unchanged with variations in coflow pressure or hole configuration. (2) Because the circular coflow air jet exhibits weak momentum, the coflow pressure hardly affects the breakup length, concentration boundary, droplet distribution, or main oscillation frequency. (3) For the circular-corner rectangular coflow, the relative position of the liquid and low-temperature air jets does not change the breakup length but affects the pulsation frequency and fuel concentration boundary. Moreover, the concentration boundary is maximized when the air jet is upstream and the fuel jet is downstream.

Acoustic liners are key passive dampers for reducing self-sustained combustion oscillations in gas turbine combustors and afterburners. This study experimentally and numerically explores how acoustic liners mitigate longitudinal combustion oscillations in a combustor featuring a bluff-body diffusion flame. Combustion tests measured oscillation suppression using liners with varying porosities, complemented by acoustic simulations clarifying liner operating mechanisms. Results indicate that optimizing the maximum sound absorption coefficient alone does not yield optimal oscillation attenuation. Instead, lowering the acoustic impedance (dimensionless impedance <0.5) significantly enhances suppression effectiveness. Additionally, attenuating combustion oscillations reduces their frequency by up to 40 Hz, explaining why maximizing absorption and minimizing oscillation intensity cannot be simultaneously optimized. Liners with larger aperture ratios exhibit better suppression performance under identical backing cavity conditions, owing to reduced oscillating velocity within the holes, which lowers the nonlinear acoustic resistance. When the peak sound absorption frequency is approximately 70 Hz lower than the self-sustained oscillation frequency, the liner can reduce the pressure oscillation amplitude by 81% (equivalent to 14.5 dB). This work offers new insights into acoustic liner design for passive control of combustion oscillations in afterburners.