Analysis of turbojet combustion chamber performances based on flow field simplified

mathematical model
Constantin Rotaru

Citation: AIP Conference Proceedings 1836, 020047 (2017); doi: 10.1063/1.4981987

View online: https://doi.org/10.1063/1.4981987
View Table of Contents: http://aip.scitation.org/toc/apc/1836/1
Published by the American Institute of Physics

Articles you may be interested in

Simple thermodynamics of jet engines
American Journal of Physics 78, 809 (2010); 10.1119/1.3373924

Performance analysis and dynamic modeling of a single-spool turbojet engine

AIP Conference Proceedings 1798, 020005 (2017); 10.1063/1.4972597

Diagnosis of students’ ability in a statistical course based on Rasch probabilistic outcome

AIP Conference Proceedings 1836, 020048 (2017); 10.1063/1.4981988

Aircraft dual-shaft jet engine with indirect action fuel flow controller
AIP Conference Proceedings 1836, 020064 (2017); 10.1063/1.4982004

Cycle of the needs satisfaction, and information support of the society development simulation system
AIP Conference Proceedings 1836, 020039 (2017); 10.1063/1.4981979

Aircraft engine mathematical model – linear system approach

AIP Conference Proceedings 1738, 290012 (2016); 10.1063/1.4952084
 Analysis of Turbojet Combustion Chamber Performances
Based on Flow Field Simplified Mathematical Model
Constantin Rotaru

”Henri Coanda” Air Force Academy, 160 Mihai Viteazul Street, 500183, Brașov, Romania
rotaru.constantin@afahc.ro

Abstract. In this paper are presented some results about the study of combustion chamber geometrical configurations that
are found in aircraft gas turbine engines. The main focus of this paper consists in a study of a new configuration of the
aircraft engine combustion chamber with an optimal distribution of gas velocity in front of the turbine. This constructive
solution could allow a lower engine rotational speed, a lower temperature in front of the first stage of the turbine and the
possibility to increase the turbine pressure ratio. The Arrhenius relationship, which describes the basic dependencies of the
reaction rate on pressure, temperature and concentration has been used. and the CFD simulations were made with jet A fuel
(which is presented in the Fluent software database) for an annular flame tube with 24 injectors. The temperature profile at
the turbine inlet exhibits nonuniformity due to the number of fuel injectors used in the circumferential direction, the spatial
nonuniformity in dilution air cooling and mixing characteristics as well as other secondary flow patterns and instabilities
that are set up in the flame tube.

INTRODUCTION
The pressure and shear stresses in the blades of a compressor or turbine produce a moment about the axis of
rotation that could be evaluated and integrated but instead it is usual to consider the momentum for the flow entering
and leaving. Because the tangential stresses at the casting and the hub are usually very small, the total work of the
blade row can be inferred from measurements of the velocity components upstream and downstream. A small change
in radius can produce changes in static enthalpy of the same order of magnitude as those produced by the deflection
and deceleration of the flux in the blades [1]. Turbine inlet temperature is important because increasing its value makes
the pressure ratio across the core turbine smaller in relation to the pressure rise of the core compressor and thereby
increases the power available from the low pressure turbine. Increasing this temperature also increases the thermal
efficiency, provided that the pressure ratio increases by an appropriate amount. The ratio of the turbine inlet
temperature to the compressor inlet temperature is a very important parameter. Reducing the compressor inlet
temperature, as occurs at high altitude, has a similar effect to increasing the turbine inlet temperature on the ground.
The ability to operate at higher temperature has been an important factor in improving the performance of jet engines,
it being possible because of better materials. The life of hot blades is limited by creep (turbine blade life is halved for
each ten degrees rise in temperature of the metal, at a given level of cooling technology and material, being the
continuing and gradual extension of materials under stress at high temperature), by oxidation or by thermal fatigue. A
better blade could be made by arranging for the crystals to form elongated in the direction of the span, namely,
directionally solidified blades.
The turbine blades operate at high temperatures because the cool air is taken from the compressor and fed to the
inside of the blades. Between 15% and 20% of the core air may be taken from the compressor for film cooling. Thermal
fatigue is not a function of the length of time the engine runs but depends on how many times the engine is started,
accelerated and stopped with the attendant raising and lowering of the turbine temperature. There is a balance between
turbine inlet temperature and cooling air requirements, since cooling air reduces the efficiency and thrust of engine.
Turbine cooling is an expensive technology and its technology is one of the areas where competition is very intense.
One way of assessing the performance of blade cooling is to use the cooling effectiveness defined by
  
K Tg  Tm / Tg  Tc , where Tg , Tm and Tc are the gas, metal and cooling temperature. It is not only the turbine

Applied Mathematics and Computer Science

AIP Conf. Proc. 1836, 020047-1–020047-8; doi: 10.1063/1.4981987
Published by AIP Publishing. 978-0-7354-1506-5/$30.00

020047-1
which has a temperature limit on it. The actual engines have such high pressure ratios that the temperature of the air
is sufficiently high at the back of the compressor that it is difficult to stand the high stresses in the compressor rotor.
Inside the combustor of a gas turbine there is a very large energy release rate per unit volume and the combustor
must be small to fit between the compressor and turbine without making the shaft to long since that would add to the
weight and could introduce problems of mechanical stiffness. The high energy release in a small volume is made
possible by the high pressure and by the very high level of turbulence. Fuel is injected as a fine sheet or spray and is
broken up by a surrounding air blast. Most of the air from the compressor is diverted to avoid the region where the
fuel is injected so that combustion starts in a relatively rich primary region with a fuel-air ratio of about 0.25 for the
take off and 0.1 at idle. Additional air is then fed in through holes in the combustor lining to complete the combustion
process and to reduce the temperature so that after dilution the effective overall air-fuel ratio is about 0.03. The air
entering the dilution region also puts a layer of relative cooling air on the walls and modifies the exit temperature
radial profile to be suitable for entry into the turbine. The stress in the turbine rotor blades is highest near the root and
to minimize creep it is desirable to have lower temperature near the root than the tip. To prevent the flame from being
blown it is necessary to set up local regions with much smaller velocity in the recirculation zone which is created by
swirling the flow around the fuel spray. Burning liquid droplets requires the liquid fuel to evaporate and then for the
fuel and air to diffuse together to form locally a near stoichiometric mixture.
In order to model the combustion process in a turbojet engine it is necessary to have some important assumptions
and simplifications like these: the rotational speed of the compressor and turbine must me equal on each shaft; the
mass flow through the compressor and turbine must be equal, neglecting the mass flow removed in bleeds and the
mass flow of fuel which is small in comparison with air mass flow; the power output of the turbine must equal the
power input into the compressor on the same shaft; the pressure rise in the compression process must equal the pressure
drop in the expansion process, including the combustor, turbines and nozzle.

COMBUSTION CHAMBER GASDYNAMICS

The mathematical model of the flow field in axial turbomachines is based on the concepts of streamlines, stream
surfaces and stream tubes in the flow path. In order for streamlines and path lines to coincide, the flow must be steady,
this being and additional restrictive assumption in gas turbine study. If the two streams have different molecular
identities, the shear layer is also a mixing layer and it is generated at the interface between the two streams, in which
momentum is transported laterally from the faster to the slower stream. If the flame tube region has a lower static
pressure than the region above, the pressure difference causes a jet to flow through the hole of area Ah Sd h2 / 4 ,
which reduced to A j Sd 2j / 4 . The vena contracta diameter d j is approximately 0.8 y 0.9d h . Because the flow is
assumed to be isentropic, the jet velocity V j is therefore determined by the static pressure difference. Downstream of
the vena contracta all of the kinetic energy in the jet is converted to thermal energy by turbulence and subsequent
viscous dissipation (Fig. 1).

Liner x

(a) (b)

FIG. 1. Parallel streams mixing (a) and secondary zone trajectories (b) in a combustion chamber

020047-2
 The length of the potential core varies from about 5d j to 7d j as the jet Reynolds number varies from 104 to 105
and the mixing transition point is about 10d j downstream of the vena contracta [6, 7]. The micromixed region
occupies about one half the jet width, so that G m # 0.1 y (Fig. 2).
The first rotating blade row is exposed to lower temperatures because of circumferential averaging, dilution of
turbine gases with first stator vane cooling air and relative velocity effects. The second stator vane is exposed to a
lower temperature because of cooling air dilution, work extraction from the turbine gases and mixing that dilutes the
hot spots. The first stator vane is fed by compressor discharge air that bypasses the combustor, because it requires a
very high supply pressure. The first rotor blade is also fed by compressor discharge air. This amount of air is
accelerated through a row of nozzles pointed in the direction of rotation [2]. The effect of this is to reduce the amount
of work required to pump the cooling air and to reduce the cooling air temperature within the blade. In gas turbines,
the heat transfer coefficients to the blades are very high and the conductivities of the materials are fairly low.
x
y

Potential core
macromixing
transition zone
vena macromixing
contracta zone micromixed
zone
potential core

Flame micromixing
tube zone

(a) (b)
FIG. 2. The Model flow in a combustor hole (a) and turbulent round jet (b)

This combination makes it imperative that the heat transfer coefficient distribution over a hole airfoil must be
known in detail. At the stagnation point this coefficient can be correlated as that of the stagnation point of a cylinder
in a cross flow that is affected by the freestream turbulence.
On the concave or pressure side of the blade airfoil, the boundary layer trips to turbulent flow and attains a heat
transfer coefficient level corresponding to a low Reynolds number turbulent flow. On the convex or suction side of
the blade airfoil, the boundary layer is often at first laminar, but has heat transfer coefficients higher than those
corresponding to laminar flow because of the effects of freestream turbulence. The heat transfer coefficient distribution
to a blade airfoil is calculated utilizing two-dimensional boundary layer theory. The end wall regions are exposed to
a lower gas temperature than the corresponding airfoils. In developing the governing integral equations which describe
the flow in the turbine, the local density appears under the integral sign. Excepting the case of incompressible flow
with small temperature differences, it becomes necessary to determine the variation of the density across the boundary
layer. Since the static pressure is constant across the boundary layer, the gas law gives the product of local density and
local temperature as constant, thus the problem becomes one of specifying the local temperature [3].
In the usual steady turbulent boundary layer equations, if the laminar transport term is not neglected with respect
to the turbulent transport, the governing equations are everywhere parabolic. In certain cases, the neglect of the laminar
transport term can result in a hyperbolic set of equations, but this depends on the precise nature of the model of the
turbulence transport. The boundary layer equations are nonlinear when they are expressed in term of velocity variables
but implicit methods make a formal linearization at some points. In the axial turbojet engine the hub and casing
boundary layers can be thought of as a swirling axially symmetric boundary layers. The hub and casing surfaces are
often not continuous and frequently stationary surfaces lie adjacent to the rotating components. Leakage and coolant
flows often occur through gaps in the surface, but all these difficulties shrink to insignificance when compared to the
effect of the blade of stator rows (Fig. 3). In the pitch averaging approach, the azimuthal variations are expressed as a
mean value plus a small perturbation and the equations of motion averaged in the azimuthal direction. Within
turbomachinery, three-dimensional flows are encountered with unfortunate regularity in the boundary layers and the

020047-3
freestream. An additional feature of the extended boundary layer equations concerns the absence of zones of influence
which are the regions bounded by the body surfaces, the boundary layer edges and a pair of characteristic surfaces,
one of the surface normal envelope passing through the external streamline and the other, the surface normal envelope
passing through the surface streamline.

MATHEMATICAL METHODOLOGY FOR CFD CALCULATION

A basic structure model for separated flow regions interacting with an inviscid outer flow region leads to expect
considerable economy of computation relative to solving the full Navier-Stokes equations. If the external flow can be
assumed irrotational, a single scalar potential equation may be used and the correction perturbation to this inviscid
outer flow may be obtained from linearized theory (in transonic flow some perturbation potential must be solved
throughout the flow). The conventional boundary layer approach leads to a parabolic system of equations but the
interacted boundary layer is not well posed as an initial value condition [4].
The maximum combustion temperature occurs when hydrocarbon fuel molecules are mixed with just enough air
so that all of the oxygen atoms are consumed, all of the hydrogen atoms form water vapor and all of the carbon atoms
form carbon dioxide, this ideal mixture of fuel and air being represented by the atom-balance equation
§ y ·ª 79 º y 79 § y·
C x H y  ¨ x  ¸«O2  N 2 » o xCO2  H 2O  ¨ x  ¸ N 2 (1)
© 4 ¹¬ 21 ¼ 2 21 © 4¹
The stoichiometric mass-basis fuel-air ratio is given by
36 x  3 y
f st (2)
1034 x  y 
A generic molecule representing jet fuels is Cn H 2n  2 , which for n 12 gives f st 0.0669 kg fuel/kg air. To
quantify the off-stoichiometric mixtures of fuel and air, it is defined the fuel/air equivalence ratio, M f / f st .
For the incomplete combustion, the atom-balance equation can be generalized as
NS
§ y ·ª 79 º
Cx H y  ¨ x  ¸«O2  N 2 » o ¦ ni Ai (3)
© 4 ¹¬ 21 ¼ i 1
where NS represents the total number of product species, Ai represents the chemical formula of the ith gas molecule
appearing in the NS product gases and ni represents the mass-specific mole number of the ith species.
The rotating external blade boundary layer are very thin and the effects of the Coriolis and centrifugal forces on
the mean flow are negligible. In partially-premixed combustion, fuel and oxidizer enter the reaction zone in distinct
streams and the mixture fraction f can be written in terms of the atomic mass fraction as [5]:
Z i  Z i.oxidizer (4)
f
Z i. fuel  Z i.oxidizer
where Z i is the elemental mass fraction for element i.
The mean time averaged mixture fraction equation is:
§P ·
w
wt
  
U f  ’ ˜ U vf  ’ ˜ ¨¨ t ’f ¸¸  S m
V
(5)
© t ¹
The source term S m is due to solely to transfer of mass into the gas phase from liquid fuel. The mixture fractions
variance, f c 2 could be find from equation:
w§ §P · H
¨ U f c ·¸  ’ ˜ §¨ U v f c ·¸ ’ ˜ ¨¨ t ’ f c2 ¸¸  Cq ˜ Pt §¨ ’ 2 f c2 ·¸  Cd U f c2
2 2
(6)
wt © ¹ © ¹ © Vt ¹ © ¹ k
where f c f  f and P t is the subgrid-scale viscosity.
The p f  probability density function can be thought of as the fraction of time that the fluid spends at the state
f , so
i
p f 'f lim ¦W i (7)
T of T i

020047-4
where T is the time scale and Wi is the amount of time that f spends in the 'f band. The analytical shape of the
function p f  depends on the turbulent fluctuations in f and it is expressed as a mathematical function that
approximates the shapes. The temporal fluctuations describing by function p f  can be used to compute the time
averaged values of variables that depend of f with the equation:
1
Mi ³ p f Mi  f d f (8)
o
where
 fuel / air actual (9)
M
 fuel / air stoichiometric
The expression of p f  is
f D 1 1  f E 1
p f  (10)
³ f D 1 1  f E 1 df
where

1  f « f 1  f   1»
ª º
D f«

ª f 1 f 
º
 1» ; E (11)
«¬ f c 2 »¼ «¬ f c2 »¼
At each point in the flow field, the p f  can be computed and used as the weighting function to determine the
time-averaged mean values of species mass fraction, density and temperature.

NUMERICAL RESULTS
The effect of turbulence on premixed flame speed is to enhance momentum transfer between the burning front and
the unburned reactants. In addition, turbulence increases the total surface area of the flame and hence increases the
heat transfer between the reaction zone and the unburned gas. A recirculation zone in the burner is needed to stabilize
a premixed flame. To stabilize the flame in the primary zone of a main burner, air is introduced through single or
double rows of swirl vanes and in order to create a stirred reaction zone in the flame tube, part of excess air is injected
through the primary air holes as radial jets.
Some important CFD results and flame tube geometrical shapes are presented in the following pictures.

(a) (b)
FIG. 3. The flame tube (a) and the actual combustion chamber shape (b)

020047-5
 (a) (b)
FIG. 4. The geometrical shape of the flame tube, 3D view (a) and quarter view (b)

(a) (b)
FIG. 5. The flow domain mesh (a) and flame tube wall position (b)

(a) (b)
FIG. 6. Stream lines (a) and velocity distribution (b)

020047-6
 (a) (b)
FIG. 7. O2 mass fraction (a) and jet fuel mass fraction (b)

(a) (b)
FIG. 8. Turbulent kinetic energy (a) and temperature distribution (b)

CONCLUSIONS
The model used in this paper to describe the combustion process in an annular flame tube has the advantage of
cleanly separating aerodynamic and chemical features of the process. The ignition time was calculated from a global
model of the chemical reaction rate and the chemical concentrations and temperature used in the combustion chamber
are taken from a mathematical model of the mixing zone that is based on time-averaged measurements. Concentration
profiles for the products of combustion are similar to the temperature profiles and the vortex structures lie in the region
with a strong average velocity gradient. The interaction of the traveling waves in this process will depend on the
geometry and velocity distribution in the flame tube of the combustion chamber. A low turbulence intensity
environment in a fuel-air mixture flow of 50-100 m/s results in a root mean square of the turbulence fluctuation speed
of 2.5-5 m/s. In a high turbulence intensity in a flow with a mean speed of 100 m/s, the turbulence contribution to
flame propagation speed may be as high as 30 m/s.

REFERENCES
1. C. J. Jaw, J. D. Mattingly, “Aircraft Engine Controls. Design. System Analysis and Health Monitoring”, AIAA,
Reston, Virginia, USA, 2009.
2. J. D. Mattingly, “Elements of Propulsion. Gas Turbines and Rockets”, AIAA, Reston Virginia, USA, 2006.
3. C. Rotaru, M. Mihăilă-Andres, P. G. Matei, “An Extended Combustion Model for the Aircraft Turbojet Engine”,
International Journal of Turbo&Jet Engines, ISSN 0334-0082, Vol. 31, Issue 3, Pages 229-237, doi: 10.1515/tjj-
2013-0048, 2014.

020047-7
4. T. V. Chelaru, A. Chelaru, “Reaction Control System using Hybrid Micro-Thrusters for Guides Sounding
Rocket”, Proceedings of the 6 th International Conference on Recent Advances in Space Technologies – RAST
2013, pp. 837-842, ISBN:978-1- 4673-6394-5, Istambul, Turkey, 12-14 June, 2013.
5. S. Farokhi, “Aircraft Propulsion”, John Wiley & Sons Inc., USA, 2009.
6. C. Rotaru, I. Sprintu, “State variable modeling of the integrated engine and aircraft dynamics”, 10th International
Conference on Mathematical Problems in Engineering, Aerospace and Sciences ICNPAA 2014, Volume 1637,
pp. 889-898, 2014.
7. H. Richter, “Advanced Control of Turbofan Engines”, Springer Science, USA, 2012.
8. C. Rotaru, A. Arghiropol and C. Barbu, “Some aspects regarding possible improvements in the performances of
the aircraft engines”, Proceedings of the 6th IASME/WSEAS International Conference on Fluid Dynamics and
Aerodynamics, Greece, (2008).
9. C. E. Moldoveanu, A. Giovannini, H. C. Boisson, “Turbulence receptivity of longitudinal vortex dominated
flow”, Lecture Notes in Computational Science and Engineering, Vol. 69: BAIL 2008- Boundary and Interior
Layer Proceedings of the International Conference on Boundary and Interior Layers-Computational and
Asymptotic Methods, Limerick, Springer Verlag, Berlin, pag. 227-236, ISBN: 978-3-642-00604-3, July, 2008;
10. A. N. Tudosie, “Supersonic Air Inlet Control System Based On The Inner Channel Minimum Cross-Section
Position Control”, International Conference on Military Technologies ICMT 2009, Brno, Czech Republic, 2009;
11. C. Rotaru, I. Cîrciu, C. Aramă, C. G. Constantinescu, “Aspects Regarding Velocity Distribution in the Secondary
Zone of a Gas Turbine Combustor”, Review of the Air Force Academy, No. 3(30), DOI: 10.19062/1842-
9238.2015.13.3.5, pp. 33-37, 2015;
12. A. N. Tudosie, “Aircraft Jet Engine Exhaust Nozzle Controller Based on Turbine Pressure Ratio Sensor with
Micro-Jet System”, Proceedings of the 11th International Conference on Applied and Theoretical Electricity
(ICATE), Craiova, 2012

020047-8