Combustion

instability

Combustion instabilities are physical phenomena occurring in a reacting flow (e.g., a

flame) in which some perturbations, even very small ones, grow and then become
large enough to alter the features of the flow in some particular way.[1][2][3]

Stability map of a hypothetical

combustor. This combustor operates
at conditions in which no dangerous
combustion-instabilities will happen.

In many practical cases, the appearance of combustion instabilities is undesirable.

For instance, thermoacoustic instabilities are a major hazard to gas turbines and
rocket engines.[1] Moreover, flame blowoff of an aero-gas-turbine engine in mid-flight
is clearly dangerous (see flameout).

Because of these hazards, the engineering design process of engines involves the
determination of a stability map (see figure). This process identifies a combustion-
instability region and attempts to either eliminate this region or moved the operating
region away from it. This is a very costly iterative process. For example, the numerous
tests required to develop rocket engines [4] are largely in part due to the need to
eliminate or reduce the impact of thermoacoustic combustion instabilities.

Classification of
combustion instabilities
In applications directed towards engines, combustion instability has been classified
into three categories, not entirely distinct. This classification was first introduced by
Marcel Barrère and Forman A. Williams in 1969.[5] The three categories are[6]

Chamber instabilities - instabilities

arising due to the occurrence of
combustion inside a chamber
(acoustic instabilities, shock
instabilities, fluid-dynamic
instabilities associated with the
chamber, etc.,)
Intrinsic instabilities - instabilities
arising irrespective of whether
combustion occurs inside a
chamber or not (chemical-kinetic
instabilities, diffusive-thermal
instabilities, hydrodynamic
instabilities such as Darrieus–
Landau instability, Rayleigh–Taylor
instability etc.,)
System instabilities - instabilities
arising due to the interaction
between combustion processes in
the chamber and anywhere else in
the system (feed-system
interactions, exhaust-system
interactions, etc.,)
Thermoacoustic
combustion instabilities
In this type of instabilities the perturbations that grow and alter the features of the
flow are of an acoustics nature. Their associated pressure oscillations can have well
defined frequencies with amplitudes high enough to pose a serious hazard to
combustion systems.[1] For example, in rocket engines, such as the Rocketdyne F-1
rocket engine [7] in the Saturn V program, instabilities can lead to massive damage of
the combustion chamber and surrounding components (see rocket engines).
Furthermore, instabilities are known to destroy gas-turbine-engine components during
testing.[8] They represent a hazard to any type of combustion system.

Thermoacoustic combustion instabilities can be explained by distinguishing the

following physical processes:

the feedback between heat-release

fluctuations (or flame fluctuations)
with the combustor or combustion
chamber acoustics
the coupling of these two
processes in space-time
 the strength of this coupling in
comparison with acoustic losses
the physical mechanisms behind
the heat-release fluctuations
The simplest example of a thermoacoustic combustion instability is perhaps that
happening in a horizontal Rijke tube (see also thermoacoustics): Consider the flow
through a horizontal tube open at both ends, in which a flat flame sits at a distance of
one-quarter the tube length from the leftmost end. In a similar way to an organ pipe,
acoustic waves travel up and down the tube producing a particular pattern of
standing waves. Such a pattern also forms in actual combustors, but takes a more
complex form.[9] The acoustic waves perturb the flame. In turn, the flame affects the
acoustics. This feedback between the acoustic waves in the combustor and the heat-
release fluctuations from the flame is a hallmark of thermoacoustic combustion
instabilities. It is typically represented with a block diagram (see figure). Under some
conditions, the perturbations will grow and then saturate, producing a particular
noise. In fact, it is said that the flame of a Rijke tube sings.

Combustion instabilities represented

with a block diagram as a feedback
amplifier.

The conditions under which perturbations will grow are given by Rayleigh's (John
William Strutt, 3rd Baron Rayleigh) criterion:[10] Thermoacoustic combustion
instabilities will occur if the volume integral of the correlation of pressure and heat-
release fluctuations over the whole tube is larger than zero (see also
thermoacoustics). In other words, instabilities will happen if heat-release fluctuations
are coupled with acoustical pressure fluctuations in space-time (see figure). However,
this condition is not sufficient for the instability to occur.

Thermoacoustic combustion
instabilities happening in a bluff-body-
flame-stabilized combustor. Dark
regions indicated strong release of
heat, and large deformations
indicated high pressure. Notice that
whenever and wherever large
deformations happen, dark regions
are seen. This is the hallmark
coupling of pressure and heat-release
seen in thermoacoustic combustion
instabilities.

Another necessary condition for the establishment of a combustion instability is that

the driving of the instability from the above coupling must be larger than the sum of
the acoustic losses.[11] These losses happen through the tube's boundaries, or are
due to viscous dissipation.

Combining the above two conditions, and for simplicity assuming here small
fluctuations and an inviscid flow, leads to the extended Rayleigh's criterion.
Mathematically, this criterion is given by the next inequality:

Here p' represents pressure fluctuations, q' heat release fluctuations, velocity
fluctuations, T is a long enough time interval, V denotes volume, S surface, and is a
normal to the surface boundaries. The left hand side denotes the coupling between
heat-release fluctuations and acoustic pressure fluctuations, and the right hand side
represents the loss of acoustic energy at the tube boundaries.
 Graphical representation of the
extended Rayleigh's criterion for some
combustor showing a region where
gains exceeds losses and the
combustor response is strong. This
suggests a strong likelihood of having
a combustion instability. This figure is
adapted from.[1]

Graphically, for a particular combustor, the extended Rayleigh's criterion is

represented in the figure on the right as a function of frequency. The left hand side of
the above inequality is called gains, and the right hand side losses. Notice that there
is a region where the gains exceeds the losses. In other words, the above inequality is
satisfied. Furthermore, note that in this region the response of the combustor to
acoustic fluctuations peaks. Thus, the likelihood of a combustion instability in this
region is high, making it a region to avoid in the operation of the combustor. This
graphical representation of a hypothetical combustor allows to group three methods
to prevent combustion instabilities:[1] increase the losses; reduce the gains; or move
the combustor's peak response away from the region where gains exceed losses.

To clarify further the role of the coupling between heat-release fluctuations and
pressure fluctuations in producing and driving an instability, it is useful to make a
comparison with the operation of an internal combustion engine (ICE). In an ICE, a
higher thermal efficiency is achieved by releasing the heat via combustion at a higher
pressure. Likewise, a stronger driving of a combustion instability happens when the
heat is released at a higher pressure. But while high heat release and high pressure
coincide (roughly) throughout the combustion chamber in an ICE, they coincide at a
particular region or regions during a combustion instability. Furthermore, whereas in
an ICE the high pressure is achieved through mechanical compression with a piston
or a compressor, in a combustion instability high pressure regions form when a
standing acoustic wave is formed.
The physical mechanisms producing the above heat-release fluctuations are
numerous.[1][8] Nonetheless, they can be roughly divided into three groups: heat-
release fluctuations due to mixture inhomogeneities; those due to hydrodynamic
instabilities; and, those due to static combustion instabilities. To picture heat-release
fluctuations due to mixture inhomogeneities, consider a pulsating stream of gaseous
fuel upstream of a flame-holder. Such a pulsating stream may well be produced by
acoustic oscillations in the combustion chamber that are coupled with the fuel-feed
system. Many other causes are possible. The fuel mixes with the ambient air in a way
that an inhomogeneous mixture reaches the flame, e.g., the blobs of fuel-and-air that
reach the flame could alternate between rich and lean. As a result, heat-release
fluctuations occur. Heat-release fluctuations produced by hydrodynamic instabilities
happen, for example, in bluff-body-stabilized combustors when vortices interact with
the flame (see previous figure).[12] Lastly, heat-release fluctuations due to static
instabilities are related to the mechanisms explained in the next section.

Static instability or flame

blow-off

Flame from a swirl-stabilized,

premixed, academic combustor
undergoing blow-off. The flow is from
right to left. The fuel-air ratio is
decreased. This makes the flame to
change its shape, then become
unstable, and eventually blow-off.

Static instability [2] or flame blow-off refer to phenomena involving the interaction
between the chemical composition of the fuel-oxidizer mixture and the flow
environment of the flame.[13] To explain these phenomena, consider a flame that is
stabilized with swirl, as in a gas-turbine combustor, or with a bluff body. Moreover,
say that the chemical composition and flow conditions are such that the flame is
burning vigorously, and that the former is set by the fuel-oxidizer ratio (see air-fuel
ratio) and the latter by the oncoming velocity. For a fixed oncoming velocity,
decreasing the fuel-oxidizer ratio makes the flame change its shape, and by
decreasing it further the flame oscillates or moves intermittently. In practice, these
are undesirable conditions. Further decreasing the fuel-oxidizer ratio blows-off the
flame. This is clearly an operational failure. For a fixed fuel-oxidizer ratio, increasing
the oncoming velocity makes the flame behave in a similar way to the one just
described.

S-shape curve resulting from the

solution of an homogeneous reactor
model representing a flame.

Even though the processes just described are studied with experiments or with
Computational Fluid Dynamics, it is instructive to explain them with a simpler
analysis. In this analysis, the interaction of the flame with the flow environment is
modeled as a perfectly-mixed chemical reactor.[14] With this model, the governing
parameter is the ratio between a flow time-scale (or residence time in the reactor) and
a chemical-time scale, and the key observable is the reactor's maximum temperature.
The relationship between parameter and observable is given by the so-called S-shape
curve (see figure). This curve results from the solution of the governing equations of
the reactor model. It has three branches: an upper branch in which the flame is
burning vigorously, i.e., it is "stable"; a middle branch in which the flame is "unstable"
(the probability for solutions of the reactor-model equations to be in this unstable
branch is small); and a lower branch in which there is no flame but a cold fuel-oxidizer
mixture. The decrease of the fuel-oxidizer ratio or increase of oncoming velocity
mentioned above correspond to a decrease of the ratio of the flow and chemical time
scales This in turn corresponds to a movement towards the left in the S-shape curve
In this way, a flame that is burning vigorously is represented by the upper branch, and
its blow-off is the movement towards the left along this branch towards the
quenching point Q. Once this point is passed, the flame enters the middle branch,
becoming thus "unstable", or blows off. This is how this simple model captures
qualitatively the more complex behavior explained in the above example of a swirl or
bluff-body-stabilized flame.

Intrinsic flame
instabilities
In contrast with thermoacoustic combustion instabilities, where the role of acoustics
is dominant, intrinsic flame instabilities refer to instabilities produced by differential
and preferential diffusion, thermal expansion, buoyancy, and heat losses. Examples of
these instabilities include the Darrieus–Landau instability, the Rayleigh-Taylor
instability, and thermal-diffusive instabilities (see Double diffusive convection).

References

1. Culick, F. E. and Kuentzmann, P.

(2006). Unsteady Motions in
Combustion Chambers for Propulsion
Systems. NATO Research and
Technology Organization.
2. Lieuwen, T. C. (2012). Unsteady
Combustor Physics. Cambridge
University Press.
3. Matalon, M. (2007). "Intrinsic flame
instabilities in premixed and
nonpremixed combustion". Annual
Review of Fluid Mechanics. 39 (1):
163–191.
Bibcode:2007AnRFM..39..163M (http
s://ui.adsabs.harvard.edu/abs/2007A
nRFM..39..163M) .
doi:10.1146/annurev.fluid.38.050304.
092153 (https://doi.org/10.1146%2Fa
nnurev.fluid.38.050304.092153) .
4. Pempie, P. and Vernin, H. "Liquid
rocket engine test plan comparison".
AIAA Paper 2001-3256.
5. Barrere, M., & Williams, F. A. (1969,
January). Comparison of combustion
instabilities found in various types of
combustion chambers. In Symposium
(International) on Combustion (Vol.
12, No. 1, pp. 169-181). Elsevier.
6. Williams, Forman A. Combustion
theory. CRC Press, 2018.
7. Oefelein, J. C.; Yang, V. (1993).
"Comprehensive review of liquid-
propellant combustion instabilities in
F-1 engines". Journal of Propulsion
and Power. 9 (5): 657–677.
Bibcode:1993JPP.....9..657O (https://u
i.adsabs.harvard.edu/abs/1993JPP.....
9..657O) . doi:10.2514/3.23674 (http
s://doi.org/10.2514%2F3.23674) .
 8. Lieuwen, T. C.; Yang, V. (2005).
Combustion instabilities in gas
turbine engines. AIAA.
9. Poinsot, T. and Veynante, D. (2005).
Theoretical and numerical
combustion. RT Edwards.
10. Rayleigh, J. W. S. (1896). The Theory
of Sound Volume 2. Dover
Publications.
11. Nicoud, F. and Poinsot, T. (2005).
"Thermoacoustic instabilities: Should
the Rayleigh criterion be extended to
include entropy changes?" (https://ha
l.archives-ouvertes.fr/hal-00908285/fi
le/chu-CandF-final.pdf) (PDF).
Combustion and Flame. 142 (1–2):
153–159.
doi:10.1016/j.combustflame.2005.02.
013 (https://doi.org/10.1016%2Fj.com
bustflame.2005.02.013) .
12. Schadow, K. C. and Gutmark, E.
(1992). "Combustion instability
related to vortex shedding in dump
combustors and their passive
control". Progress in Energy and
Combustion Science. 18 (2): 117–
132. doi:10.1016/0360-
1285(92)90020-2 (https://doi.org/10.1
016%2F0360-1285%2892%2990020-
2) .
13. Glassman, I. and Yetter, R. A. and
Glumac, N. G. (2014). Combustion.
Academic Press.
14. Peters, N. (2000). Turbulent
Combustion. Cambridge University
Press.
Retrieved from
"https://en.wikipedia.org/w/index.php?
title=Combustion_instability&oldid=10877995
94"

This page was last edited on 14 May 2022, at

16:14 (UTC). •
Content is available under CC BY-SA 4.0
unless otherwise noted.