# Combustion instability of a lean premixed prevaporized gas turbine combustor studied using phase-averaged PIV (2023)

ScienceDirect

ViewPDF

## Article preview

• Abstract
• Introduction
• Section snippets
• References (59)
• Cited by (119)
• Recommended articles (6)

## Combustion and Flame

Volume 161, Issue 4,

April 2014

, Pages 958-970

Author links open overlay panelJacob E.TemmePersonEnvelopePatton M.AllisonJames F.Driscoll

## Abstract

A strong, naturally-occurring “growl” combustion instability was studied for the case of a lean premixed prevaporized (LPP) combustor that shows great promise in reducing pollutant emissions. Phase-averaged particle image velocimetry (PIV) was applied for the first time to an LPP device to measure phase lags and spatial correlations. The extensive data set includes spatial and temporal correlations between six parameters: combustor pressure, plenum pressure, injection velocity, heat release rate (Rayleigh index), flame liftoff distance and flame centroid. Measured phase angles and time lags are consistent with the MIT model of Ghoniem et al., along with the concept of “equivalence-ratio oscillation” discussed by Lieuwen et al. Frequency and phase data prove that a dual-mode Helmholtz resonance is driven by an equivalence ratio oscillation. One common modeling assumption is shown to be not valid; the length of an attached flame is not what is oscillating; instead the flame base oscillates violently due to periodic liftoff and flashback and this presents modeling challenges. Growl boundaries and the effects of varying some geometric lengths were recorded.

(Video) Reacting Swirling Flow Physics; Speaker: Paul Palies

## Introduction

The goal of this work has been to better understand the physics of a strong, low-frequency “growl” combustion instability that occurs when a lean premixed prevaporized (LPP) fuel injector [1], [2] is operated in a pressurized chamber with preheated air, at the high flow rate conditions typical of engine idle. The LPP fuel injector is called TAPS (twin-annular premixed swirler) and it achieves low NOx emission levels, but LPP devices are prone to combustion instabilities [1], [2], [3]. This is because a lifted premixed flame exists which has been shown to be not anchored to the burner [4]. Another goal has been to explain our results using a reduced-order model. The best explanation is provided by the MIT/Ghoniem model of Hathout et al. [5] and colleagues [6], [7] along with the concept of an “equivalence-ratio oscillation” that is described by Lieuwen and co-workers [8], [9], [10], [11], Santavicca and co-authors [12], [13], Dowling and Hubbard [14] and Richards et al. [15]. Most of these previous efforts employed laboratory-scale swirl flames and results cannot always be extrapolated to explain instabilities in realistic large-scale devices, such as those described in Refs. [16], [17], [18].

Three common types of gas turbine combustion instabilities are called the Helmholtz (bulk) mode, organ tones and “convective-acoustic” types. Good reviews of these different modes are provided by Ducruix et al. [19], McManus et al. [20], Lieuwen and Zinn [8] and Mongia et al. [1]. The Helmholtz mode occurs if wall pressures at various locations in the combustor are in-phase and are equal in magnitude. Also, the frequency of a bulk mode instability should be close to that predicted by the Helmholtz resonator formula, using the known geometric volume, neck area and neck length. Organ tones occur if the spatial correlation of wall pressures has a sinusoidal shape. Yu et al. [21] report that their spatial correlation corresponds to a longitudinal standing wave. The frequency of a quarter-wave longitudinal mode should correspond to c/(4L) where c is the speed of sound and L is the combustor length. Convective-acoustic modes exist if the frequency of the instability is proportional to the mean air velocity. Lieuwen and Zinn [8] explain that a disturbance (a vortex structure, a thermal hot spot or a “puff” of air) can convect downstream at the air velocity. When it reaches some constriction a pressure wave rapidly propagates upstream to trigger the next flow structure.

Several types of flame transfer functions have been proposed [19], [20] that relate fluctuations in the heat release rate to the fluctuations in gas velocity and pressure. If the pressure and heat release rate fluctuations are correlated then the pressure oscillations can be amplified. However, it is not known which predicted flame transfer function is most appropriate for a given experiment. Some transfer functions rely on the assumption that the flame remains attached and only the flame length oscillates. If the flame base location oscillates or the flame width oscillates, a different transfer function may be required. The transfer function for an equivalence ratio oscillation has been modeled by Hathout et al. [5] and by Lieuwen et al. [9]. A different set of transfer functions assume that other parameters are important, including fuel droplet size, fuel flow rate oscillations, precession of the vortex core and vortex shedding. For this reason the fuel type and air velocity were varied in the present work in order to rule out certain transfer functions that are not appropriate.

The tradeoff between reduced nitric oxide levels and combustion instability has been described by Cohen et al. [22] who show that as the combustion in made leaner the NOx emissions will decrease but eventually the pressure fluctuations rapidly increase. For their design 60ppm NOx was the minimum achievable without encountering strong engine growl. Shehata [23] also reduced NOx levels to less than 120ppm in an LPP device by operating sufficiently lean.

Table 1 lists some previous studies that are most relevant [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [13], [40], [41]. The only studies that are listed are those that report acoustic measurements of low-frequency combustion dynamics (<350Hz) of a swirl flame in a gas turbine model combustor. The first six experiments listed were operated using liquid fuel. The one other study that considered an LPP device was that of Bernier et al. [24]. They reported that unsteady flame flashback was important and this also was a conclusion of the present work. There are no previous studies that report phase-averaged PIV velocity results in an LPP device, although Giezendanner et al. [34] at DLR performed phase-averaging in a small gaseous-fueled swirl flame. Phase-averaging is needed to determine the phase lag between the injector velocity and the combustor pressure to understand which one of several possible mechanisms is dominant. Table 1 also shows that the present burner is almost three times larger than the others that were operated on liquid fuels.

The second half of Table 1 lists relevant studies of gaseous fueled burners. Meier and colleagues [30], [31], [32], [33], [34] made extensive laser imaging measurements of a gaseous swirl flame; they concluded that a Helmholtz resonance strongly couples with the precession of the vortex core.

Many previous small burners were operated with one end that was open to the atmosphere; some had unrealistic levels of swirl and most were operated with gaseous fuels. It was decided to operate a large-scale single fuel injector burner at realistic pressure, preheat temperature, Jet-A liquid fuel flow rates, and large air flow rates that are near the idle point of a realistic gas turbine engine. A choked downstream boundary condition was imposed, which more closely represents a real jet engine than a combustor that is open to the laboratory. Mongia et al. [1] have reported that low-frequency growl instabilities are known to occur near idle operation if some parameters are sufficiently “off-design”.

An initial investigation was conducted to determine which of the known instability mechanisms that have been observed in small-scale devices also apply to a large-scale LPP device. That is, which modes dominate: Helmholtz (bulk) modes, organ tones (such as longitudinal standing waves), or convective-acoustic “entropy” waves? Additionally, the fundamental assumption of many flame transfer functions was studied. In some models the flame is assumed to be attached to the burner but the flame length (and the heat release rate) is assumed to vary as the air velocity oscillates. Real swirl flames are not attached to a burner, so a goal was determine what flame properties are oscillating. Our movies show that the flame base location oscillates violently, but this introduces modeling challenges because no good model exists that predicts the liftoff distance of a swirl flame.

This work includes several new aspects. There have been previous PIV velocity field measurements in gas turbine combustors undergoing instabilities [27], [30], [35], [36], [40], but previous work was mostly with gaseous fuel (not LPP) and often was not phase-averaged. Phase-averaged PIV velocity measurements in an LPP combustor were made to measure the phase lag between the combustor pressure and the injector velocity. This phase lag was compared to predictions of the MIT model [5] and analysis of Lieuwen et al. [9]. Spatial correlations of wall pressures were made at several locations in the combustor and the supply plenum in order to determine which type of instability occurs (a dual-Helmholtz mode was identified to dominate over four weak organ tones). Phase-averaged measurements provided the magnitudes and phase angles associated with several important variables in the MIT model: the fluctuations in the combustor and plenum pressures, the inlet velocity, the heat release rate, and the flame liftoff distance. Images of Rayleigh index were recorded using simultaneous flame chemiluminescence and pressure measurements. Violent motions of the flame base were correlated with the velocity fluctuations. Unlike many previous studies, some geometric lengths were systematically varied to determine which modes dominate (Helmholtz versus organ tones). Mean velocity was varied to determine the role of convective-acoustic “entropy waves” (which was found to be negligible). The liquid fuel type also was varied (by adding gasoline to the Jet-A) to see if altering the spray properties affect the instability (it did not). Finally, the phase-averaged measurements were compared to the MIT model and to previous work by Lieuwen et al. [9] to examine the role of “equivalence ratio oscillations”. Few previous instability studies have considered LPP operation because it requires the use of liquid fuel; in order to achieve a realistic properly atomized spray it was required to use larger fuel flow rates, elevated pressures and higher preheat air temperatures than were considered in most previous studies.

In the present work two low-frequency modes (at 80 and 160Hz) were found to dominate; intense low frequency noise is commonly called engine growl. Refs. [1], [18] note that most gas turbine engines experience these low-frequency “cold tones” if the engine is not properly maintained or if it is run under sufficiently “off-design” conditions. There has not been agreement about the source of growl. It is unlikely that standing waves are the source if the necessary length scale required to produce such low frequencies exceeds the engine length. If a convective-acoustic mechanism is the source, the growl frequency should be proportional to the air mass flow rate; that was not observed in the present work.

## Experimental arrangement

The LPP fuel injector used is a non-commercial version of the TAPS (Twin Annular Premixed Swirler) described by Mongia [2]. Figure 1a and b shows that it is mounted at the upstream end of a quartz combustor that has a square cross-section that is 112.2mm by 112.2mm (4.4in. by 4.4in.). The walls neck down to an exit size of 88mm by 88mm and this forces the end of the central recirculation zone to remain inside the combustor. The combustor is mounted within the University of Michigan

## Pressure spectra and Helmholtz acoustics

Figure 3 shows that the two largest peaks in the spectrum of the combustor pressure occur at 80Hz and 160Hz; they represent the low-frequency “growl”. Operating conditions correspond to case 1 in Table 1; the transducer is located at x=13.2cm downstream from the injector. The combustor pressure fluctuations are large; ${p}_{\text{rms}}^{\prime }=0.05\phantom{\rule{0ex}{0ex}}\text{atm}$ (0.75psia) and the ratio ${p}_{\text{rms}}^{\prime }/\overline{p}$ is 0.017. Power/frequency in Fig. 3 has the dimensions of decibels per Hertz, as given by:$\text{Power}/\text{frequency range}\phantom{\rule{0ex}{0ex}}\left(\text{dB}/\text{Hz}\right)=\left[10\mathrm{log}\left({p}^{\prime }/\overline{p}\right)$

## Phase-angles that correspond to an equivalence ratio oscillation

Figure 14 summarizes our measured phase differences between the following quantities: combustor pressure, plenum pressure, injection gas velocity, flame liftoff distance and heat release rate of the flame base region. Several of these curves look similar to the predictions of Lieuwen et al. [9] for an equivalence ratio oscillation. The upper two curves in Fig. 14 show that the plenum pressure lags the combustor pressure by ϕ1 equal to 63°. This is consistent with Giezendanner et al. [34] and

## Conclusions

1.

For certain off-design conditions (listed in Table 2) a strong combustion instability at 80Hz and 160Hz naturally occurred during the operation of a lean premixed prevaporized (LPP) TAPS fuel injector. The instability causes violent liftoff and flashback movement of the flame at 80Hz and very large pressure variations of 0.05atm (0.75psi). The Jet-A liquid fuel rates, the large heated air flow rates and the elevated pressures correspond to engine conditions (for a single injector) near idle.

2.

To

## Acknowledgments

This work was supported by the GE-USA program and was monitored by Drs. Michael Foust, Kent Lyle, David Wisler and Hukam Mongia of GE Aviation. The authors thank Dr. Sulabh Dhanuka for his previous thesis work.

## References (59)

• P.M. Allison et al.

### Proc. Combust. Inst.

(2013)

• J. Diederichsen et al.

### Combust. Flame

(1965)

• B.O. Ayoola et al.

### Combust. Flame

(2006)

• S.K. Dhanuka et al.

### Proc. Combust. Inst.

(2011)

• S.K. Dhanuka et al.

### Proc. Combust. Inst.

(2009)

• P.L. Therkelsen et al.

### Combust. Flame

(2013)

• J.M. Samaniego et al.

### Combust. Flame

(1993)

• K. Schildmacher et al.

### Proc. Comb. Institute

(2007)

(Video) Turbulent Combustion: Experiments and Fundamental Models, Driscoll, Day 4, Part 1

• I. Boxx et al.

### Combust. Flame

(2010)

• A.M. Steinberg et al.

### Combust. Flame

(2010)

• J.H. Uhm et al.

### Combust. Flame

(2005)

• J.H. Uhm et al.

### Combust. Flame

(2004)

• K. Zahringer et al.

(2003)

• M. Shehata

### Fuel

(2009)

• K. McManus et al.

### Prog. Energy Combust. Sci.

(1993)

• K.T. Kim et al.

### Combust. Flame

(2010)

• J.G. Lee et al.

### Proc. Combust. Inst.

(2000)

• J.H. Cho et al.

### Combust. Flame

(2005)

• T. Lieuwen et al.

### Proc. Combust. Inst.

(1998)

• J.P. Hathout et al.

### Proc. Combust. Inst.

(2000)

• H.C. Mongia et al.

### J. Propul. Power

(2003)

• H.C. Mongia, AIAA Paper 2003-2657, American Inst. of Aeronautics and Astronautics,...
• T. Lieuwen

### J. Propul. Power

(2003)

• S.K. Dhanuka et al.

### J. Propul. Power

(2011)

• J.P. Hathout et al.

### J. Propul. Power

(2002)

• M. Fleifil, J.P. Hathout, A.M. Annaswamy, A.F. Ghoniem, AIAA Paper 2000-0708, American Inst. of Aeronautics and...
• .T. Lieuwen et al.

### ASME J. Eng. Gas Turb. Power

(2001)

• B.T. Zinn, T.C. Lieuwen, in: T.C. Lieuwen, V. Yang (Eds.), AIAA Progress in Astronautics and Aeronautics, vol. 210,...
• A.P. Dowling et al.

(2000)

• ## Cited by (119)

• Gas-phase velocity estimation in practical sprays by Phase-Doppler technique

2022, International Journal of Multiphase Flow

Practical sprays are characterized by a two-way coupling between droplets and the surrounding gas. The effect of sprays on process performance is critical in numerous applications; hence, the gas-phase velocity is often estimated via two methods, using Phase Doppler measurement data. The first one is using a rule of thumb, i.e., estimating the gas-phase velocity by analyzing droplet sizes below a few micrometers. The second way of estimating the Stokes number, Stk, is using a threshold well below unity, such as 0.1. There are numerous definitions available in the literature for Stk, resulting in several magnitudes difference. These complex problems are resolved in this paper in the following way. A new definition for Stk is provided, which is sufficiently robust for, e.g., pressure and twin-fluid atomizers. According to the results, the Stk < 0.1 threshold means droplet sizes between 2 and 10 micrometers above 10m/s gas-phase velocities. Filtering for too small droplets could lead to biased characteristics, especially in estimating the turbulent properties. Hence, the gas-phase velocity estimation is a function of the measurement setup and has a significant spatial dependence. The reader can find the software code online and the algorithm in Appendix A.

• Experimental investigation of the helical mode in a stratified swirling flame

2022, Combustion and Flame

In a centrally-staged combustor, the stratified swirling flame often manifests a helical mode, which is characterized by the emergence of helical vortices, including the precessing vortex core (PVC) and the outer helical vortex (OHV). To understand the flame-vortex interaction mechanism responsible for the coherent heat release rate (HRR) fluctuation of such a helical mode, this work employs 20kHz simultaneous measurements of PIV/CH2O PLIF and PIV/CH* chemiluminescence to resolve the unsteady flow field and flame motion. Through a combination of time-resolved and phase-averaged analysis, together with the three-dimensional reconstruction of the coherent flow and flame structures, we arrive at the qualitative understanding that in the helical mode the pilot flame dominated by PVC is a major source of coherent HRR fluctuation, while the main flame only slightly modifies the fluctuation intensity. Quantitative analyses further demonstrate the different effects of PVC and OHV on the coherent HRR fluctuation. Specifically, the PVC affects the primary heat-release process by dictating the convective transport of the reactants upstream of the flame front. And the OHV contributes to additional heat release by inducing stretching and roll-up of the pilot flame tip; it happens during the OHV's second growth, which is likely caused by the thermal expansion resulting from the primary heat release. This explains our finding that the OHV-induced heat release occurs at a slightly delayed phase angle and a downstream location compared with the primary PVC-induced heat release.

• Premixed flames subjected to extreme levels of turbulence part II: Surface characteristics and scalar dissipation rates

2022, Combustion and Flame

The current work assesses the impact of turbulence on flame surface characteristics and scalar dissipation rates (${\chi }_{C}$) of piloted premixed methane-air flames. This assessment is facilitated via the application of planar Rayleigh scattering (PRS) to a subset of the flames considered in Part I (13 in total), which possess turbulent Karlovitz, Damköhler, and Reynolds numbers in the range of 1.1$\le K{a}_{T,P}\le$144, 0.24$\le D{a}_{T,P}\le$5.79, and 1,200$\le R{e}_{T,0}\le$35,000, respectively. Instantaneous temperature maps are derived from the PRS images with an accuracy that is conservatively estimated to be within $\sim$3%. Additionally, the fidelity of the collected images was enhanced via the application of a combined wavelet-based and edge-preserving guided filtering scheme that allows scales associated with the peak of the dissipation spectra in the modestly turbulent flames considered here to be fully resolved. After such filtering, the temperature images were converted to reaction progress variable maps (${C}_{T}$). Two-dimensional isocontours were extracted from these maps and used to assess the flame surface density (FSD; $\Sigma$) and the in-plane curvature (${\kappa }_{2D}$) of the flames considered here. Furthermore, scalar dissipation rate (${\chi }_{{C}_{T}}$), its density weighted average ($\stackrel{˜}{{\chi }_{{C}_{T}}}$), and other relevant quantities (e.g., the density-weighted variance of ${C}_{T}$: $\stackrel{˜}{{C}_{T}^{″2}}$ and the generalized FSD: $\overline{|\nabla {C}_{T}|}$) were derived from the ${C}_{T}$-maps. Subsequently, relationships between these quantities were assessed and compared against theoretical models that aim to capture such relationships through simple expressions.

The results indicate that the distribution of ${\kappa }_{2D}$ values broadens with enhanced turbulence, as does the integrated and peak values of FSD. Moreover, each of these parameters exhibit a strong positive correlation with $K{a}_{T,P}$, highlighting the role this non-dimensional parameter plays in dictating flame surface wrinkling/area-generation. Relationships between ${\chi }_{{C}_{T}}$, $\stackrel{˜}{{C}_{T}^{″2}}$, and $\overline{|\nabla {C}_{T}|}$, are compared to flamelet-type models proposed by Vervisch etal. as well as by Bray, Swaminathan, Kolla, Chakraborty, and coworkers. Overall, despite the fact that the considered flames exhibit substantial preheat zone broadening and are subjected to extreme levels of turbulence, the proposed theoretical relationships between ${\chi }_{{C}_{T}}$, $\stackrel{˜}{{C}_{T}^{″2}}$, and $\overline{|\nabla {C}_{T}|}$ show favorable agreement to the measurements. While differences between the measured and theoretical results are observed, the two can be reconciled with minor adjustments to the various constants in these models (i.e., changing them by no more than a factor of $\sim$2). Ultimately, such observations provide support for utilizing flamelet-type models to simulate premixed flames even when they are subjected to extreme turbulence levels. Moreover, the wealth of information provided here can help guide the development of robust yet efficient numerical tools for simulating highly turbulent premixed flames, which will aid in reducing the cost of and time to produce robust, efficient, and clean-burning combustion engines.

• Flame structures and thermoacoustic instabilities of centrally-staged swirl flames operating in different partially-premixed modes

2021, Energy

Flame structures and thermoacoustic instabilities of centrally-staged swirl flames fueled with methane at atmospheric pressure are experimentally investigated by varying the stratification ratio (SR) and covering four different partially-premixed modes, namely fully-premixed main and pilot stages, fully-premixed-pilot/partially-premixed-main stages, partially-premixed-pilot/fully-premixed-main stages, and partially-premixed main and pilot stages. The flame structures in the quiet cases at SR=0–3 are studied, and three basic flame types can be defined with increasing SR: the lifted flame, the attached twin-flame, and the V-shaped flame. The above basic flame structures show deviations in different partially-premixed modes. The links between flame structures and thermoacoustic instabilities are investigated in the unstable cases by exciting thermoacoustic instabilities with higher thermal power at increased global equivalence ratio. The results show that the partial premixing in the main stage excites stronger thermoacoustic instabilities and thus leads to different flame structures. Simplified thermoacoustic network analysis is also conducted to provide an insight into the nature of thermoacoustic instabilities in the three basic flame structures summarized in the unstable cases. The present investigation reveals the effect of partially-premixed modes on flame structures and thermoacoustic instabilities, which can guide the early-stage design and the development of instability control strategies in practical combustion systems.

• Parametric study of the slope confinement for passive control in a centrally-staged swirl burner

2021, Energy

Lean premixed centrally-staged combustion holds the potential of meeting low-emission requirements, but it is known to be susceptible to thermoacoustic instabilities. A passive control method, using slope confinements has been demonstrated previously to be efficacy in suppressing thermoacoustic instabilities. Recognizing that the detailed flow features and the role of slope confinement structures associated with these thermoacoustic instabilities have not been well understood, the present study aims at revealing the suppressive mechanisms and assessing the suppressive efficacy of different slope confinements by measuring flow fields and conducting experiments with varying geometric parameters, including slope step height and slope angle. It is found that the slope confinements investigated have a general suppressive effect on thermoacoustic oscillations, which can be attributed to the limited vortex shedding from the dump plane. The slope confinement with smaller slope step height and slope angle is shown to have better performance on combustion stability. However, oscillations still occur in slope confinements with relatively large slope angle and step height. A qualitative relationship between the oscillation amplitude and the characteristic scale of corner recirculation zone is then established from the flow separation aspects. The present work provides insights into practical confinement design for passive control.

• Transient aerodynamic performances and pressure oscillations of a core engine combustor during start up

2021, Aerospace Science and Technology

This paper focuses on transient aerodynamic performances and pressure oscillations of a practical core-engine annular combustor during start up core-engine test article and its measurement systems are introduced at first. Then the test results. The transient evolutions of pressure drops which represent basic aerodynamic characteristics of combustor are discussed. A new and special aerodynamic “swing” phenomenon of static pressure (or pressure drops across combustor liner) and low frequency pressure oscillations (50 Hz∼180 Hz) were observed in the tests and are discussed in detail. The variations of pressure drops and pressure oscillation magnitudes are affected by compressor outlet flow conditions with different core-engine acceleration rates. With elaborate analysis on “waterfall” maps and Campbell maps of transient pressure oscillations, the driving of low frequency pressure oscillations in combustor is confirmed to be unsteady flow originating from pre-diffuser flow passage. Combined with analysis on dynamic pressures monitored in compressor passage, the aerodynamic “swing” phenomenon is confirmed to be a resonance process between unsteady flow produced by upstream stages rotation of compressor and acoustic field of compressor (with natural frequency bands 97 Hz ± 10 Hz and 152 Hz ± 15 Hz). Results also infer that low frequency pressure oscillations will amplify magnitude of swing under certain conditions.

View all citing articles on Scopus

## Recommended articles (6)

• Research article

On the phase between pressure and heat release fluctuations for propane/hydrogen flames and its role in mode transitions

Combustion and Flame, Volume 160, Issue 12, 2013, pp. 2827-2842

This paper presents an experimental investigation into mode-transitions observed in a 50-kW, atmospheric pressure, backward-facing step combustor burning lean premixed C3H8/H2 fuel mixtures over a range of equivalence ratios, fuel compositions and preheat temperatures. The combustor exhibits distinct acoustic response and dynamic flame shape (collectively referred to as “dynamic modes”) depending on the operating conditions. We simultaneously measure the dynamic pressure and flame chemiluminescence to examine the phase between pressure (p′) and heat release fluctuations (q′) in the observed dynamic modes. Results show that the heat release is in phase with the pressure oscillations (θqp≈0) at the onset of a dynamic mode, while as the operating conditions change within the mode, the phase grows until it reaches a critical value θqp=θc, at which the combustor switches to another dynamic mode. According to the classical Rayleigh criterion, this critical phase (θc) should be π/2, whereas our data show that the transition occurs well below this value. A linear acoustic energy balance shows that this critical phase marks the point where acoustic losses across the system boundaries equal the energy addition from the combustion process to the acoustic field. Based on the extended Rayleigh criterion in which the acoustic energy fluxes through the system boundaries as well as the typical Rayleigh source term (pq′) are included, we derive an extended Rayleigh index defined as Re=θqp/θc, which varies between 0 and 1. This index, plotted against a density-weighted strained consumption speed, indicates that the impact of the operating parameters on the dynamic mode selection of the combustor collapses onto a family of curves, which quantify the state of the combustor within a dynamic mode. At Re=0, the combustor enters a mode, and switches to another as Re approaches 1. The results provide a metric for quantifying the instability margins of fuel-flexible combustors operating at a wide range of conditions.

• Research article

Structure conditioned velocity statistics in a high pressure swirl flame

Proceedings of the Combustion Institute, Volume 37, Issue 4, 2019, pp. 5031-5038

A piloted, partially premixed, liquid-fueled swirl burner is operated at high pressure (1 MPa). High-speed (6 KHz) stereoscopic PIV is used to investigate the characteristics of the stagnation line separating the pilot jet and the central recirculation zone (CRZ) with varying pilot-main ratio and global equivalence ratio. The mean curvature of the stagnation line displayed a large spatial scale pattern that was present for all operating conditions. All three components of velocity, in-plane shear, and swirling strength are conditioned upon the instantaneous stagnation line. Mean distributions of the velocity normal to the stagnation line show that velocity is oriented towards the CRZ when the stagnation line is found nearer the centerline of the combustor. The conditioned out-of-plane velocity (w) shows a distinct concentration of large mean and fluctuation RMS values towards the center of the measurement domain. Varying fuel flow does not significantly change this spatial structure, only the magnitudes of the w statistics. The in-plane shear stress was the largest for the pilot biased condition as a stronger shear layer develops. For the leanest flame, large fluctuation RMS values of shear stress were confined to a region where the pilot jet begins to interact more heavily with the main jet. Operating with less pilot fuel flow enhanced the mean conditional swirling strength indicating that the pilot shear layer was shedding more intense eddies. Disregarding spatial relations, a scatter plot of w, shear stress, and swirling strength displayed trends between the variables. The largest swirling strength values coincide with highest magnitude shear stresses and the widest range of w. These conditioned statistics highlight how certain aspects of the combustor flow field are invariant with fuel distribution. This is desirable for aeropropulsive combustors that must maintain stable ignition from a range of conditions from landing/take-off to cruise.

• Research article

Interference mechanisms of acoustic/convective disturbances in a swirl-stabilized lean-premixed combustor

Combustion and Flame, Volume 160, Issue 8, 2013, pp. 1441-1457

Interference mechanisms of acoustic/convective disturbances were experimentally investigated in a swirl-stabilized lean-premixed gas turbine combustor operated with natural gas fuel and air at atmospheric pressure and elevated temperature. Interference between azimuthal and acoustic velocity disturbances at high-amplitude limit cycle oscillations is characterized in detail as a function of axial swirler location, oscillation frequency, and mean nozzle velocity. We show that both the frequency and the intensity of self-excited instabilities in a model gas turbine combustor are correlated with axial swirler position, which indicates that a vorticity wave generated at the swirl vanes is a primary source of convective disturbances in the absence of equivalence ratio nonuniformities. Flame transfer function measurements confirm that the linear/nonlinear heat release response is a strong function of axial swirler location, even when unforced flame structures remain unchanged. The key parameter controlling this phenomenon is the phase difference between the azimuthal and acoustic velocity perturbations at the combustor dump plane; the phase difference is affected by swirler location, frequency, mean velocity, and the speed of sound. It was found that out-of-phase interference between azimuthal and acoustic velocity disturbances at the combustor inlet yields large flame angle fluctuations in relation to swirl number fluctuations, and therefore the formation of a coherent structure is hindered due to high kinematic viscosity within the vortex formation region. In-phase interference mechanisms, on the other hand, lead to high-amplitude limit cycle oscillations. This interference mechanism is then explored in the presence of temporal equivalence ratio nonuniformities, in which two different sources of convective mechanisms should be considered simultaneously in connection with acoustic velocity perturbations and the vortex dynamics. Results reveal that equivalence ratio oscillation has a significant effect on the strength of combustion-acoustic interactions. Strong self-excited instabilities of partially premixed flames are produced by in-phase interactions between acoustic velocity and equivalence ratio oscillations, which are governed by fuel injection location, frequency, mean nozzle velocity, and fuel injector impedance. At this phase condition, unburned reactants with high equivalence ratio impinge on the flame front with high inlet velocity, potentially causing large fluctuations of heat release rate.

• Research article

Examining flow-flame interaction and the characteristic stretch rate in vortex-driven combustion dynamics using PIV and numerical simulation

Combustion and Flame, Volume 160, Issue 8, 2013, pp. 1381-1397

In this paper, we experimentally investigate the combustion dynamics in lean premixed flames in a laboratory scale backward-facing step combustor in which flame-vortex driven dynamics are observed. A series of tests was conducted using propane/hydrogen/air mixtures for various mixture compositions at the inlet temperature ranging from 300K to 500K and at atmospheric pressure. Pressure measurements and high speed particle image velocimetry (PIV) are used to generate pressure response curves and phase-averaged vorticity and streamlines as well as the instantaneous flame front, respectively, which describe unsteady flame and flow dynamics in each operating regime. This work was motivated in part by our earlier study where we showed that the strained flame consumption speed Sc can be used to collapse the pressure response curves over a wide range of operating conditions. In previous studies, the stretch rate at which Sc was computed was determined by trial and error. In this study, flame stretch is estimated using the instantaneous flame front and velocity field from the PIV measurement. Independently, we also use computed strained flame speed and the experimental data to determine the characteristic values of stretch rate near the mode transition points at which the flame configuration changes. We show that a common value of the characteristic stretch rate exists across all the flame configurations. The consumption speed computed at the characteristic stretch rate captures the impact of different operating parameters on the combustor dynamics. These results suggest that the unsteady interactions between the turbulent flow and the flame dynamics can be encapsulated in the characteristic stretch rate, which governs the critical flame speed at the mode transitions and thereby plays an important role in determining the stability characteristics of the combustor.

• Research article

Effect of fuel–air mixture velocity on combustion instability of a model gas turbine combustor

Applied Thermal Engineering, Volume 54, Issue 1, 2013, pp. 92-101

Nowadays, it is easy for unstable combustion phenomenon to develop in a gas turbine that is working in a lean premixed condition. To eliminate the onset of these instabilities and develop effective approaches for control, the mechanisms responsible for their occurrence must be understood. The flame recirculation zone is very important, as it can modulate the fuel flow rate and may be the source of instability, plus its flame structure has a major impact on heat release rate oscillation and flame stabilization. In this study, we conducted experiments under various operating conditions with a model gas turbine combustor to examine the relation of combustion instability and flame structure by OH chemiluminescence. Swirling CH4 - air flame was investigated with an overall equivalence ratio of 1.2 to lean blowout limit and dump plane velocity of 30–70m/s. Phase-locking analysis was performed to identify structural changes at each phase of the reference dynamic pressure sensor under conditions of instability. At an unstable condition, flame root size varies a lot compared to stable condition which is because of air and fuel mixture flow rate changes due to combustor pressure modulation. After this structural change, local extinction and re-ignition occur and it can generate a feedback loop for combustion instability. This analysis suggests that pressure fluctuation of combustion causes deformation of flame structure and variation of flame has a strong effect on combustion instability. In this study, we observed two types of combustion instability characteristics related to the instability of both the thermo-acoustic and flame vortex type.

• Research article

A new pattern of instability observed in an annular combustor: The slanted mode

Proceedings of the Combustion Institute, Volume 35, Issue 3, 2015, pp. 3237-3244

In annular combustion chambers of aero-engines and gas turbines, acoustic coupling may arise from azimuthal modes which are less well damped than axial modes. Also, since the circumference is the largest length in the combustor, the azimuthal modes have the lowest resonance frequencies and are most prone to instability. Such a coupling raises many scientific issues which are considered in a small number of fundamental experiments. The present investigation focuses on this problem and provides experimental data on a special type of combustion instability in which the thermo-acoustic resonant coupling involves a combination of modes. This produces an unusual pattern of flame responses in which the distribution of heat release rate is slanted. Data are provided in the form of free radical light intensity patterns (interpreted as heat release rate distributions) and microphone signals detected in the plenum and chamber. It is shown that the slanted pattern is the signature of a combination of two modes with coinciding frequencies, the first being a standing azimuthal mode while the second is an axial mode. Measurements of the flame describing function on a single matrix burner at the fundamental frequency are used to explain the observed phase shift and amplitude in the flame responses of the different injectors in the annular combustor.

View full text

## FAQs

### What is lean premixed combustion? ›

The Gas Turbine Association defines Lean Premix Stationary Combustion Turbine as “Lean premixed stationary combustion turbine means any stationary combustion turbine designed to operate at base load with the air and fuel thoroughly mixed to form a lean mixture before delivery to the combustor.

How much is the percentage of compressed air that passes directly to combustor for combustion process for a gas turbine? ›

About 15-20% of air is introduced around the jet of fuel in primary zone to provide necessary high temperature for rapid combustion. Some 20 % of the total air is then introduced through holes in the flame-tube in the secondary zone to complete the combustion.

What do you mean by combustion stability? ›

An index determining to a certain extent the combustion stability is the heat load of the burner belt defined as the ratio of heat supplied to the furnace with fuel and hot air and the wall surface in the burners area – formula (8.66). If its values are relatively high, it is easier to maintain the flame stability.

What is the best type of combustor? ›

The annular type of combustor is usually used in MTs and integrated design. Advantages of the annular type include shorter size, more uniform combustion, less surface area, permitting better mixing of the fuel and air as well as simple structure.

Why is lean combustion more efficient? ›

First, excess air in the mixture reduces combustion process temperature, which reduces the formation of NOx with respect to conventional stoichiometric engines. Second, due to excess Oxygen, the combustion process is more efficient and more energy is generated from the same amount of fuel.

What measurement is taken when performing a combustion efficiency test? ›

O2 reading: The oxygen (O2) reading is by far the most important reading a combustion analyzer measures. The oxygen level in the atmosphere remains constant (20.9 percent); it's the only true constant in the combustion process.

Which type of combustion chamber is the most efficient in regards to thermal efficiency and weight? ›

So, the pentroof combustion chamber has excellent airflow, maximum intake and exhaust capacity, and minimal heat loss. This combination makes it the most efficient of the five types of combustion chambers.

What percentage of air passing through the combustion section is burned? ›

6.3.

Generally, 50% to 70% of the total air is underfire air and the remaining portion is overfire air. Most mass burn furnaces operate with between 50% and 100% excess air.

What causes combustion instability? ›

In the simplest view, combustion instability can be thought of as pressure swings in the engine caused by the multiple streams of liquid oxygen and rocket fuel combining and igniting at extremely high pressures in such a way that causes violent vibrations.

Why is combustion instability a problem? ›

Thermoacoustic or combustion instabilities may be amplified to extremely high-pressure levels that reduce the engine efficiency, induce mechanical vibration, and may damage or in some extreme cases destroy the burner.

### What does heat of combustion tell you about stability? ›

Just remember the lower the energy, the more stable the compound. So the compound that had the highest heat of combustion was the least stable. And the compound with the lowest heat of combustion was the most stable. So branched alkanes are lower in energy, or more stable, than straight chain alkanes.

What are the 3 types of combustion chambers in a turbine engine? ›

4.3 Combustion Chamber Types

Even though three main types of combustion chambers can be identified (the multiple-chamber, the tubo-annular and the annular combustion chamber), it is worth noting that the majority of modern jet engines including high-bypass engines have annular combustion chambers.

What are the three primary types of combustion chamber found in turbine engines? ›

Types of Combustion Chamber. There are three main types of combustion chamber in use for gas turbine engines. These are the multiple chamber, the tubo-annular chamber, and the annular chamber.

How do you increase the thrust of a gas turbine engine? ›

Afterburner. The most widely recognized method of boosting thrust is the afterburner, also known as tailpipe burning. Fuel is injected into the hot exhaust gas flowing between the turbine and nozzle. The combustion of the gas expands the airflow as it enters the nozzle, which increases thrust.

How do you tell if an engine is lean-burn or rich burn? ›

Rich-burn engines generally operate with (λ) equal to 0.995. Lean Burn: Lean-burn engines operate with an AFR that has a lower concentration of fuel to air, making it a “fuel-lean” mixture. Lean-burn engines operate with (λ) anywhere between 1.5 and 2.2.

How lean mixture affects combustion? ›

A lean mixture reduces fuel film or fuel droplets, improves air-fuel mixture, reduces rich-fuel phenomenon, and improves thermal efficiency. An excess of air, conversely, leads to an increase in oxygen content and increases particulate oxidation [13], [32].

Is it better to run rich or lean? ›

Q: Is it better to run the too lean or rich engines? An engine running slightly rich will give more power, but running lean will cause catastrophic engine damage. Running too rich can also cause severe damage to your engine.

What are the 4 stages of combustion? ›

The cycle includes four distinct processes: intake, compression, combustion and power stroke, and exhaust.

What are the 4 stages of an engine? ›

1:fuel injection, 2:ignition, 3:expansion(work is done), 4:exhaust. The four-stroke engine is the most common types of internal combustion engines and is used in various automobiles (that specifically use gasoline as fuel) like cars, trucks, and some motorbikes (many motorbikes use a two stroke engine).

What is the first phase of combustion? ›

The first phase of combustion is called as ignition delay (ID), in which the tiny fuel droplets evaporates and mixes with high temperature (or high pressure) air.

### What is considered a good combustion efficiency? ›

Combustion efficiency is the calculation / measurement, in percentage, of how well your equipment is burning a specific fuel.
...
Combustion Processes and Their Combustion Efficiency Ranges.
Combustion ProcessEfficiency Range
Boiler with Gas-powered Burner75-83%
Condensing Furnace (Gas & Oil) “High Efficiency”75-90+%
6 more rows
Oct 15, 2019

What are good combustion analyzer readings? ›

A typical reading would be 2% to 6% (see Figure 2). I like adjusting the air for the middle of the acceptable range in case the blower wheel gets dirty and delivers less air for combustion.

How do you calculate combustion analysis? ›

Thus we need to perform these general steps. First, convert from the data given to grams of carbon, hydrogen, and oxygen. Second, determine the empirical formula from the grams of carbon, hydrogen, and oxygen. Third, determine the molecular formula from the empirical formula and the given molecular mass.

What is the most efficient combustion engine? ›

Wärtsilä's new 31 four-stroke engine is the most fuel-efficient engine currently on the market. The diesel variant of the engine uses 8–10 g/kWh less fuel on average than the nearest rival over the whole load range. This value may be as low as 165g/kWh when operating at peak efficiency.

How can combustion efficiency be improved? ›

An often stated rule of thumb is that boiler efficiency can be increased by 1% for each 15% reduction in excess air or 40°F reduction in stack gas temperature. Operating your boiler with an optimum amount of excess air will minimize heat loss up the stack and improve combustion efficiency.

What is the maximum efficiency of an internal combustion engine? ›

Modern gasoline engines have a maximum thermal efficiency of more than 50%, but road legal cars are only about 20% to 35% when used to power a car.

What are the two main factors which affect the performance of gas turbines? ›

Performance of a gas turbine is mainly affected by the various parameters like pressure ratio, turbine inlet temperature and ambient conditions.

How much excess air is in a gas turbine? ›

Natural gas-fired boilers may run as low as 5 percent excess air. Pulverized coal-fired boilers may run with 20 percent excess air. Gas turbines runs very lean with up to 300 percent excess air.

What are the important factors affecting combustor design? ›

What Are The Factors Affecting Combustion Chamber Design? Temperature Of Gases. Temperature Distribution. Combustion.

What do you mean by lean-burn combustion? ›

Lean burn combustion is defined as the reaction of combustion of fuel along with excess mass of air, it means the mixture of fuel, and air is diluted by the excess amount of air as compared to the stoichiometric air required for combustion of unit mass of fuel.

### What do you mean by lean-burn engine? ›

lean-burn engine in British English

(ˈliːnˌbɜːn ˈɛndʒɪn ) automobiles. an engine designed to use a lean mixture of fuel and air in order to reduce petrol consumption and exhaust emissions.

What is premixed flame with example? ›

Premixed flames occur in any homogeneous mixture where the fuel and the oxidant are mixed prior to the reaction. Examples are the Bunsen burner flame and the flame in most spark-ignited engines. Premixed flames can progress either as deflagration or detonation processes.

Which is better lean or rich mixture? ›

However, the key difference between lean and rich fuel mixture is that we use a lean mixture for maximum efficiency while we use a rich mixture for maximum power in an engine. These air-fuel mixtures are useful in internal combustion engines and industrial furnaces.

How do you tell if an engine is rich burn or lean burn? ›

Rich-burn engines generally operate with (λ) equal to 0.995. Lean Burn: Lean-burn engines operate with an AFR that has a lower concentration of fuel to air, making it a “fuel-lean” mixture. Lean-burn engines operate with (λ) anywhere between 1.5 and 2.2.

How lean mixture affects combustion? ›

A lean mixture reduces fuel film or fuel droplets, improves air-fuel mixture, reduces rich-fuel phenomenon, and improves thermal efficiency. An excess of air, conversely, leads to an increase in oxygen content and increases particulate oxidation [13], [32].

Lean-burn engines (both gasoline and diesel) enjoy higher fuel economy and cleaner emissions than conventionally tuned engines. By nature they use less fuel and emit fewer unburned hydrocarbons and greenhouse gases while producing equivalent power of a like-sized “normal” combustion engine.

What is a lean air/fuel ratio? ›

Any mixture greater than 14.7:1 is considered a lean mixture; any less than 14.7:1 is a rich mixture – given perfect (ideal) "test" fuel (gasoline consisting of solely n-heptane and iso-octane).

Why lean burn engines produce small amounts of carbon monoxide? ›

Lean burn operation involves the burning of fuel with an excess of air, in ratios up to 24 parts of air to one part of fuel. Under these conditions nitrogen oxides and carbon monoxide emissions are at a minimum, but hydrocarbons can rise at the onset of unstable combustion, as can be seen in Figure 2.

Which of the gas is produced more when combustion is running on a lean mixture? ›

If you run a bit lean and the engine has good combustion quality, Hydrocarbon and Carbon Monoxide emissions will be low, but NOx (Nitrogen-oxide) emissions will be high. There is enough excess Oxygen in the cylinder to react almost all of the HC and CO during combustion.

What is the difference between premixed and non-premixed combustion? ›

The properties of a laminar premixed flame were explored in a previous post and the properties of turbulent premixed flames will be outlined in a future post. A non-premixed flame occurs when the fuel and oxidizer are not mixed prior to reacting. An example of this is the diffusion flame from a lighter as shown above.

### What is the difference between premix combustion and diffusion combustion? ›

Diffusion flames occur at the interface where fuel vapour and air meet. Unlike premixed flames, the fuel vapour and the oxidant are separate prior to burning. The dominant process in the diffusion flame is the mixing process.

Why is a premixed flame more hazardous than a diffusion flame? ›

Diffusion flame temperatures are lower than premixed flames because incomplete combustion releases less heat.

How do you tell if a carburetor is rich or lean? ›

Q: How Do You Tell if a Carburetor Is Rich or Lean? A: One way to tell for sure is by "reading" the spark plugs. If the plug tip is white, the mixture is lean. If it's brown or black, it's rich.

Does a lean mixture burn faster? ›

Since we know that "rich" air/fuel charges require more time to combust, it stands to reason that smaller (leaner) charges will (1) burn more rapidly and (2) will, consequently, be likely to convert more fuel into heat in the same time period.

Why does a lean mixture burn slower? ›

The ideal ratio of gasoline to air for combustion is 14.7:1, meaning 14.7 parts air to one part gas by mass. A lean mixture contains more air than that, more than can actually be used in combustion.

Top Articles
Latest Posts
Article information

Author: Sen. Emmett Berge

Last Updated: 01/27/2023

Views: 5849

Rating: 5 / 5 (80 voted)

Author information

Name: Sen. Emmett Berge

Birthday: 1993-06-17

Address: 787 Elvis Divide, Port Brice, OH 24507-6802

Phone: +9779049645255

Job: Senior Healthcare Specialist

Hobby: Cycling, Model building, Kitesurfing, Origami, Lapidary, Dance, Basketball

Introduction: My name is Sen. Emmett Berge, I am a funny, vast, charming, courageous, enthusiastic, jolly, famous person who loves writing and wants to share my knowledge and understanding with you.