Professional Documents
Culture Documents
Chapter-2 Review of Literature
Chapter-2 Review of Literature
REVIEW OF LITERATURE
In general, there exists a range of complex, interacting physical and chemical phenomena in a
combustor flow field. Fuel spray atomization and vaporization, turbulent transport, finite rate
chemistry of combustion and pollutant formation, radiation and particle behavior, and
recirculation zones involving multiple flow streams are included. However when taken
together in the context of multi-dimensional flows rigorous description of these phenomena
are either not available or require mathematical models which are too complex for
computation. For these reasons, models of varying degrees of sophistication have been used
depending on the particular application. Over the last three decades, the level of
sophistication of the predictive models has been continuously increased with improvements
in numerical methods, computer capabilities and physical understanding.
Development of Gas, or combustion, turbines was originally started in the 18th century. The
first patent was issued to England’s John Barber in 1791 for a combustion turbine. Patents for
modern versions of combustion turbines were awarded in the late nineteenth century to Franz
Stolze and Charles Curtis, however early versions of gas turbines were all impractical
because the power necessary to operate the compressors outweighed the amount of power
generated by the turbine. To achieve positive efficiencies, engineers would have to increase
combustion and inlet temperatures beyond the maximum allowable turbine material
temperatures of the day. It was not until the middle of this century that gas turbines evolved
55
into practical machines, primarily as jet engines. Although some prototype combustion
turbine units were designed, the developments that led to their practical use were a result of
World War II military programs. The actual race for jet engines was prompted by World War
II and therefore government started subsidization on R&D. Later gas turbines for power
generation were to emerge to from these military advances in technology. Only Germany’s
Junkers and Great Britain’s Rolls- Royce were successful to enter general production with
their engines during the war. [1]
Technology transfers began to take place as early as 1941, when Great Britain began working
with the US on turbine engines. The sharing of engineering drawings between England’s
Power Jets Ltd. and America’s GE Company was done during this period. American
companies such as GE and Westinghouse began development of gas turbines for land, sea
and air use which would not prove deployable until the end of the war [5]. Solar Turbines (the
“Solar” refers only to the name of the company, not the source of energy) also emerged later
during the war by fabricating high temperature materials, such as steel for airplane engine
exhaust manifolds. The knowledge gained by manufacturers during this time would help
them manufacture other gas turbine products in the post war period.
After World War II, gas turbine R&D was spurred in some areas and stunted in others.
In an example of R&D expansion, the transfer of detailed turbine plans from Rolls-Royce to
Pratt & Whitney was made as a repayment to the US for its assistance to Great Britain under
the Lend- Lease agreement.[2],[3] This allowed Pratt & Whitney, previously specialists in
reciprocating engines, to emerge as a strong developer of combustion turbines. In contrast,
German and Japanese companies were expressly barred from manufacturing gas turbines.
These companies were able to emerge later. For example, Siemens began recruiting engineers
and designers from the jet engine industry as soon as it were allowed, beginning in 1952. [4]
For aircraft use most of developments in gas turbines took place during 1950s and 1960s. GE
and Pratt & Whitney engines were used turbofan engines in early Boeing and Douglas
commercial planes which promote the R&D in this field. This advance of combustion
turbines into the commercial aviation market, and in some cases the boat propulsion market,
allowed manufacturers to sustain their development efforts even though entrance into the
base load electric power generation market was not yet even on the horizon. Gas turbines also
began to emerge slowly in the peaking power generation market. The independent power
generation design group was formed by GE and Westinghouse for their aircraft engine
designers. Westinghouse would later exit the jet engine business in 1960 while keeping its
stationary gas turbine division. Among US turbine manufacturers, only GE was especially
56
able to transfer knowledge between its ongoing aircraft engine and power generation turbine
businesses. The beginning of gas turbine power generation “packages” occurred early 1960s
when GE and Westinghouse engineers were able to standardize (within their own companies)
designs for gas turbines [5]. There were two main regions for marketing innovation took place.
First, providing better power generation option in place of steam engine or reciprocating
engine that will attract the more customers. Second, this standardization helped for multiple
sales with little redesign for each order, easing the engineering burden and lowering the costs
of gas turbines. This introduction of cooling technologies advance was the single most
important breakthrough in gas turbine development since their practical advent during World
War II. The cooling involved the circulation of fluids through and around turbine blades and
vanes. These cooling advances were originally part of the military turbojet R&D program,
but began to diffuse into the power generation turbine programs about five years later.
Advancing cooling technology, along with improvements in turbine materials, permitted
manufacturers to increase their firing and rotor inlet temperatures and therefore improve
efficiencies. Although manufacturers were making great technological strides in gas turbine
development, it was not until the Great Northeast Blackout of 1965 that the US utility market
truly awoke to need for additional peaking generation capacity. This peaking is exactly what
gas turbines were good for; their fast startup times would allow generators to match periods
of high demand. Even though simple-cycle gas turbines of the day had dismal efficiencies
(only about 25%) compared to those of coal-fired plants, their ability to handle peak loads led
to an increase in demand and renewed R&D from manufacturers.
2.2. Combustor Design
The simultaneous involvement of evaporation, turbulent mixing, ignition, and chemical
reaction in gas turbine combustion is too complex for complete theoretical treatment. Instead,
large engine manufacturers undertake expensive engine development programs to modify
previously established designs through trial-and-error. [26]
They also develop their own proprietary combustor design rules from
57
Figure 2.1: Modern combustor components.
the experimental results of these programs. These design rules provide a means of specifying
the combustor geometry to meet a set of requirements at the given inlet conditions.
Combustor designers without access to proprietary design procedures must derive their own
methods from the literature or from experimentation. Numerous published empirical, semi-
empirical, and analytical tools have been developed to reduce the need for costly
experiments. The two extreme cases, empirical and analytical, differ by the method of
derivation [27-32]. Empirical design tools are correlations derived from experimental datasets
whereas analytical ones are discretized versions of the governing equations. Simple empirical
correlations provide accurate results quickly and are easily implemented into design codes,
yet they are only applicable to cases for which the measured data was based on. Analytical
methods, less accurate in comparison to empirical methods, are much more flexible as they
are only restricted by the simplifying assumptions necessary to reduce their complexity and
computation time. Hybrid semi-empirical tools combine both empirical and analytical
58
methods to provide a reasonable balance between accuracy and computation time.
Basic material on combustion theory and models can be found in the paper, Some
Fundamentals of Combustion, which provides the idea about the modeling of combustor [22]
[23]
.
Numerical procedures used to be based on the SIMPLE algorithm and rectangular grid
systems, as illustrated by the highly popular TEACH code and its derivatives .It may be noted
that other techniques are available however these have not found wide spread use in
combustor simulation because of lack of flexibility, robustness and experience. Only the
broad features of such flow field have been simulated with TEACH-based codes. It is not
trivial, however, to improve the fidelity of these simulations. The models for turbulence and
chemistry and the numerical procedures are together responsible for the problems
encountered. [19, 20, 21]
Performance enhancements and control of heat transfer in high pressure gas turbine vanes
and rotors is dependent on understanding the flow and thermal fields approaching the turbine.
The flow field exiting the combustor has highly non-uniform pressure and temperature
59
variations in both the radial and circumferential directions as well as high turbulence levels
[25]
.
Finite-rate chemistry models for turbulent combustion have been developed and
assessed in jet flames. These models have been used in calculations of simple flow field when
kinetically influenced behavior is of interest. For example, the emission of thermal and
prompt nitric oxides (NOX) is a very prominent issue in the design of stationary gas-turbine
systems [25].The formation of nitric oxides is limited kinetically and so the equilibrium models
are not adequate. The model must recognize the fluctuations in the NOX production rate due
to turbulence. Extensions of the “fast” chemistry model have been proposed (Kent and Bilzer
1976) but have difficulties in accounting for super-equilibrium free radicals, which increase
NOX levels (as demonstrated by Drake et al. 1987). Prompt NOx is more problematic as it
involves the chemistry of hydrocarbon fragments (Iveraiah et al. 1972); the challenge is to
describe the chemical kinetics with a computationally tractable yet realistic scheme. Other
pollutants such as unburned hydrocarbons and CO also detract from the applicability of “fast”
chemistry models. Efforts are being made (e.g., Dasgupta et al. 1993) to predict droplet
evaporation and NOX formation in gas turbine combustor environment.
Several experimental studies have been carried out on the mixing characteristics of
single air jets. Among them the work of Norster is most relevant to gas turbine combustors.
Sridhara investigated the mixing of air jets injected into hot gas streams under conditions that
allowed the temperature and velocity of the hot and cold streams, the injection hole diameter,
the angle of injection and the mixing length to be accurately controlled and varied over a
wide range. Holderman et al. have been reported the results of various NASA studies of a
60
rectangular duct in which jet diameter, jet spacing and jet-to-gas-stream momentum flux
varied.
Many workers have attempted to trace the paths of round jets injected at right angles
into airstreams. For example, Callaghan and Rugger injected heated air into a cold air stream
and defined the penetration by the maximum pressure total-head traverse. Norster injected
cold air jets into a hot gas stream and took temperature traverse in line with the jet at various
distances downstream of its origin. The position of lowest temperature in the traverse defined
the center of the jet, and the maximum penetration was equated to the depth at which the
centerline of the jet became asymptotic to the mainstream flow. [8]
Bharani have shown that the bulk of the flow remains close to the outer liner wall
between the rows of primary and dilution holes while it shifts towards the liner mid plane
after the row of dilution holes. Ahmed and Nejad have carried out experimental investigation
of turbulent swirling flow in a combustor model for coaxial swirling jets with dump diffusers.
Green and Whitelaw have suggested that the standard k- model gives better results than the
other turbulence models in turbulence combustion prediction. Mongia has shown difficulty in
the specification of boundary conditions, inferior resolution of which hampers the ability of
computational models to predict combustor characteristics. Mohan el al. have numerically
investigated annuli flow and effect of inlet swirl on the flow split through the liner holes of
annular reverse flow combustor model. [9]
61
Transverse oscillations in a cylindrical chamber were also treated. It was established
that in addition to its role in energy transfer between modes, nonlinear gas dynamics also
provides a means of shifting the frequencies of oscillations to integral multiples of the
fundamental. This additional role can reduce the efficiency of energy transfer, thus increasing
the acoustic amplitudes. An example of a low amplitude transverse oscillation was produced
suggesting a means by which the amplitudes of transverse modes, as well as non-integral
longitudinal modes, may be reduced.
The coupling between combustion processes and acoustic oscillations was studied as a
possible explanation of the phenomenon known as triggering. Using several ad hoc models,
the effects of nonlinear pressure coupling and velocity coupling on the behavior of the system
were investigated. Substantial regions of possible triggering were produced when using a
model of velocity coupling with a threshold, but only if nonlinear gas dynamics was also
included.
The interaction between combustion noise and acoustic instabilities has received
relatively little attention. The sources of noise in a combustion chamber are associated with
vorticity and entropy waves. By including these contributions in the approximate analysis, the
general forms of the stochastic excitations were obtained. Subsequently, the effects of these
excitations on the amplitudes of acoustic modes were studied. When only nonlinear gas
dynamics was included, no cases of bimodal probability density functions, characteristic of
triggering, were found. However, when the model of velocity coupling with a threshold is
added, bimodal probability densities can occur. [10]
62
results. Addressing that problem is the principal purpose of this paper. Our analysis shows
that, in case of longitudinal modes, a first mode instability problem requires a minimum of
four modes in the modal truncation whereas, for a second mode instability, one needs to
retain at least the first eight modes. A second important problem concerns the conditions
under which a linearly stable system becomes unstable to sufficiently large disturbances.
Previous work has given a partial answer, suggesting that nonlinear gas dynamics alone
cannot produce pulsed or 'triggered' true nonlinear instabilities; that suggestion is now
theoretically established. Also, a certain form of the nonlinear energy addition by combustion
processes is known to lead to stable limit cycles in a linearly stable system. A second form of
nonlinear combustion dynamics with a new velocity coupling function that naturally displays
a threshold character is shown here also to produce triggered limit cycle behavior. [11]
A mathematical model for calculating the three-dimensional flow, gas composition and
temperature fields in gas-turbine combustion chambers is described. The model makes use of
the density-weighted average forms of the governing conservation equations with the two
equation k-ε model for turbulent transport. The chemical reactions associated with heat
release are assumed to be fast and fluctuations in scalar properties accounted for by use of a
β-probability density function. For liquid spray fuelled combustors the droplet concentration
field is described via an equation for droplet size. For a propane fuelled model gas turbine
combustor the method yields calculated fields of gas composition and pollutants (NO, CO
and UHC) which are in good overall agreement with measured values. [13]
63
3.2. Gas Turbine Combustor Modeling with CFD
Understanding of the flow dynamics, chemical kinetics and heat transfer mechanism within
micro-combustors is essential for the development of combustion-based power MEMS
devices. In Part I, CFD based numerical simulation has been proven to be an effective
approach to analyze the performance of the micro-combustor under various conditions. In
this paper, numerical simulations are performed to analyze the combustion behavior in a
three-dimensional micro-combustor based on the prototype used in the MIT micro-gas
turbine engine. The CFD model of the micro-combustor includes fuel/air flow path,
combustion chamber as well as solid walls used to construct the combustor. The simulation
analysis includes not only the detailed chemical reactions occurred in the combustion
chamber, but also the fluid flow dynamics, heat transfer within the combustor and heat loss to
the ambient. The performance of the combustor is evaluated under various fuel/air ratio, flow
rate and heat loss conditions. Through such systematic numerical analysis, a proper operation
space for the micro-combustor is suggested, which may be used as the guideline for micro-
combustor design. In addition, the results reported in this paper illustrate that the numerical
simulation can be one of the most powerful and beneficial tools for the micro-combustor
design, optimization and performance analysis. [14]
Gas turbine combustors are often susceptible to self-excited oscillations, which lead to
unacceptable levels of pressure, velocity, and heat release fluctuations. Although instabilities
can occur in systems with locally constant equivalence ratio, it is very important to take into
account the influence of equivalence ratio fluctuations, which are generated in the fuel air
mixer in the unstable case. These fluctuations are convicted into the flame and lead to an
additional mechanism for the generation of heat release fluctuations. Moreover, entropy
waves are produced in the flame, which travel through the combustor and generate additional
pressure waves during the acceleration of the flow at the combustor exit. To date, available
theories use the physically unrealistic assumption that the equivalence ratio waves as well as
the entropy waves are convected downstream without any spatial dispersion due to the
combustor aerodynamics. An analytical approach is presented, which allows us to take the
spatial dispersion into consideration. For that purpose, the response of the burner and the
combustor to an equivalence ratio impulse or an entropy impulse is calculated using the
Laplace transformation and a more general transfer function for harmonic waves is derived.
The obtained expression has three parameters, which represent the influence of the burner or
the combustor aerodynamics, respectively. This equation can be used in numerical codes,
64
which represent the combustion system through a network of acoustic multiports, if the
equivalence ratio and the entropy are added to the vector of variables considered. The
parameters required for the dynamic combustor model can be deduced from a detailed CFD
analysis of the combustor flow in case of the application of the theory to a particular
combustor design. As an example, a simple model combustor is used to demonstrate the
application of the theory. It is highlighted how the spatial dispersion of the equivalence ratio
and entropy fluctuations can be included in the stability analysis. The calculated examples
reveal that the influence of both variables on the generation of instabilities is highly over
predicted if the spatial dispersion is not taken into account. Furthermore, it can be deduced
from the study that burner and combustor designs with a wide range of convective time scales
have advantages with respect to the stability of the combustor. [15]
65
induced, leading to a finite-amplitude limit cycle oscillation. The feasibility of the approach
is demonstrated with three-dimensional simulations of a simple model annular combustor.
The effect of the model parameters and of axial mean flow on the stability and the shape of
the excited modes is shown. [12]
66
most difficult mode of heat transfer to simulate in the combustion environment. A discrete
transfer radiation model is developed and validated for use within the network solver. The
effects of soot concentration on radiation are evaluated with the introduction of radial
properties profiles. The accuracy of the heat transfer models are evaluated with comparisons
to experimental thermal paint temperature data on a reverse flow and annular combustors.
The resulting network analysis code represents a powerful design tool for the combustion
engineer incorporating a novel and unique strategy. [14]
The results showed that the position of the eye of the primary vortex varied in the cross-
stream direction. In the gaps between the primary holes, it was located near the midplane of
the combustor and moved toward the upper wall in regions close to the primary jets. The
usual in-line arrangement of the primary holes with vaporizer exits resulted, therefore, in hot
regions in the primary zone which were in line with vaporizer exits and positioned close to
the upper wall. Moving the primary holes by (1/2 - pitch) caused the downward movement of
the hot regions in the primary zone and with this arrangement excess fuel was transported out
of the primary zone along a path further from the center plane so that in the exit plane hot
regions were located away from the vaporizer exists contrary to the usual arrangement where
the general temperature pattern of the primary zone was maintained at the exit. The effect of
air---fuel ratio was found to be small in the primary zone and modest in the exit plane and the
exit plane pattern factors were higher than those of engine practice mainly due to the
atmospheric conditions of the present experiments. Results also showed that in the primary
zone, reaction was controlled more by physical than chemical processes and that in the
67
intermediate and dilution zone fuel breakdown and CO oxidation were chemical kinetic rate
limited. [15]
The chamber had to be designed so that the flame is self-sustaining and the temperature of the
products of combustion is sufficiently below the maximum working temperature of the
turbine. The inner shell, the flame tube, is made of 22-gauge perforated steel. Air flow from
the compressor is fed through the flame tube as well as around its sides. The air in the flame
tube is burned with the propane, while the air on the outside of the flame tube is used to cool
the products of combustion before they enter the turbine. The molar flow rate of propane is
0.028 C3H8 mol./sec. This corresponds to a mass flow rate of propane, mC3 H8, of 0.001232
Kg/Sec. to stoichiometrically balance the air with the propane, the mass of burned air per
second must be 0.019 kg/ sec. Therefore the percentage of air flow from the compressor that
68
must be burned is 15.3% of that air flow should be directed through the flame tube section of
the combustion chamber [17].
69