Group Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Title: 17.6.2 - Influence of air temperature and evaporation zone length on evaporation and combustion
ALL VOLUMES CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102023/00430
 Material Information
Title: 17.6.2 - Influence of air temperature and evaporation zone length on evaporation and combustion Reactive Multiphase Flows
Series Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Physical Description: Conference Papers
Creator: Ahmadi, W.
Chrigui, M.
Sadiki, A.
Janicka, J.
Publisher: International Conference on Multiphase Flow (ICMF)
Publication Date: June 4, 2010
 Subjects
Subject: spray combustion
evaporation zone length
Eulerian-Lagrangian method
RaNS
 Notes
Abstract: Using an Eulerian-Lagrangian RANS based procedure under a fully two way coupling a detailed numerical simulation of kerosene spray combustion has been carried out in a partially premixed pre-vaporized three dimensional configuration. The investigations were focused on the flame temperature profile dependency with respect to the length variation of the pre-vaporization zone. For the combustion, an approach based on a modified Bray-Moss-Libby model has been adopted to account for the partially premixed combustion. First, the results have been compared to the experimental data for validation purposes. A good agreement has been achieved. Then, a fundamental study has been performed by changing the droplet diameter, the kerosene flammability limits and the location of the combustion initialization. Temperature variations and flame flashback phenomena have been observed and analysed. All these investigations were performed for atmospheric pressure, inlet air temperature of 90°C and a global equivalence ratio of 0.7.
General Note: The International Conference on Multiphase Flow (ICMF) first was held in Tsukuba, Japan in 1991 and the second ICMF took place in Kyoto, Japan in 1995. During this conference, it was decided to establish an International Governing Board which oversees the major aspects of the conference and makes decisions about future conference locations. Due to the great importance of the field, it was furthermore decided to hold the conference every three years successively in Asia including Australia, Europe including Africa, Russia and the Near East and America. Hence, ICMF 1998 was held in Lyon, France, ICMF 2001 in New Orleans, USA, ICMF 2004 in Yokohama, Japan, and ICMF 2007 in Leipzig, Germany. ICMF-2010 is devoted to all aspects of Multiphase Flow. Researchers from all over the world gathered in order to introduce their recent advances in the field and thereby promote the exchange of new ideas, results and techniques. The conference is a key event in Multiphase Flow and supports the advancement of science in this very important field. The major research topics relevant for the conference are as follows: Bio-Fluid Dynamics; Boiling; Bubbly Flows; Cavitation; Colloidal and Suspension Dynamics; Collision, Agglomeration and Breakup; Computational Techniques for Multiphase Flows; Droplet Flows; Environmental and Geophysical Flows; Experimental Methods for Multiphase Flows; Fluidized and Circulating Fluidized Beds; Fluid Structure Interactions; Granular Media; Industrial Applications; Instabilities; Interfacial Flows; Micro and Nano-Scale Multiphase Flows; Microgravity in Two-Phase Flow; Multiphase Flows with Heat and Mass Transfer; Non-Newtonian Multiphase Flows; Particle-Laden Flows; Particle, Bubble and Drop Dynamics; Reactive Multiphase Flows
 Record Information
Bibliographic ID: UF00102023
Volume ID: VID00430
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 1762-Ahmadi-ICMF2010.pdf

Full Text

7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Influence of air temperature and evaporation zone length on evaporation and
combustion


Wahidullah Ahmadi, Mouldi Chrigui, Amsini Sadiki and Johannes Janicka

Technische Universitit Darmstadt, Dept. of Mechanical Engineering, Institute for Energy and Powerplant Technology ,
Petersenstr.30, 64287 Darmstadt, Germany
sadiki@tekt.tu-darmstadt, ahmadi@ tekt.tu-darmstadt.de

Keywords: Spray combustion, Evaporation zone length, Eulerian-Lagrangian method, RANS




Abstract

Using an Eulerian-Lagrangian RANS based procedure under a fully two way coupling a detailed numerical simulation of
kerosene spray combustion has been carried out in a partially premixed pre-vaporized three dimensional configuration. The
investigations were focused on the flame temperature profile dependency with respect to the length variation of the
pre-vaporization zone. For the combustion, an approach based on a modified Bray-Moss-Libby model has been adopted to
account for the partially premixed combustion. First, the results have been compared to the experimental data for validation
purposes. A good agreement has been achieved. Then, a fundamental study has been performed by changing the droplet
diameter, the kerosene flammability limits and the location of the combustion initialization. Temperature variations and flame
flashback phenomena have been observed and analysed. All these investigations were performed for atmospheric pressure,
inlet air temperature of 900C and a global equivalence ratio of 0.7.


Introduction

The need of fundamental knowledge of spray formation
characteristics is vital for the development and improvement
of many engine applications. Thereby the droplet
vaporization process and the subsequent fuel-air mixture
preparation are of great importance. Poor mixing of the fuel
and oxidant streams leads to mixture inhomogeneities that
affect combustion efficiency and result in enhanced
pollutants formation. In Lean Premixed Prevaporized (LPP)
combustor, the diminishing of the maximum generated
temperature and thereby the reduction of NOx production
are well known challenges. This paper deals with the lean
spray combustion in a chamber where chemical reaction
takes place far away from the injection nozzle. Several
approaches have been published in the literature and aim at
overcoming the flashback and reducing emissions [1]. These
approaches strive low NOx production by designing devices
that permit rapid spray phase transition, mixing and
combustion. Baessler et al. [2] studied the NOx emission of
premixed partially vaporized kerosene spray flame at
atmospheric conditions and found out that dealing with lean
conditions a reduction of NOx requires a prevaporization of
the kerosene spray and should pass over 50% upstream of
the flame. He realized that increasing the prevaporization
zone length reduce the emissions. Shaefer et al. [3] studied
flashback in Lean Premixed Prevaporized (LPP) combustion
of a turbulent kerosene flame. Rocke et al. [4] investigated
venturi LPP on mixing, atomization and evaporation
behavior as well as emission. Nomura et al. [5] studied
experimentally a partially prevaporized spray burner with
monodispersed ethanol droplets to investigate the


interaction between fuel droplets and a flame. They
investigated the effect of mean droplet diameter, and the
entry length of droplets into a flame on the laminar burning
velocity of partially pre-vaporized sprays.
It is worth mentioning that the most data produced in the
literature for studying the NOx reduction and flame flashback
in a LPP context have been achieved for single phase
combustion. Numerical researches that involve two phase
flow, especially sprays under consideration of combustion are
still limited. Therefore we focus in this paper on the analysis
of the flame characteristics with respect to different boundary
conditions for the spray and the carrier phase.

The paper is organized as follows. In the subsequent section
an overview of the mathematical models used and numerical
procedures is provided. To evaluate the prediction capability
of the models proposed, the application configuration under
study will be introduced. In the next section, results obtained
are presented. Different parameter studies are then compared
and discussed, before concluding.


Mathematical models and numerical procedures

Carrier phase description
The carrier phase is considered as continuum phase and is
described using the Reynolds averaging method. For this
purpose, the governing transport equations have been solved
for mass, momentum and concentration. For the turbulence
description, the RNG model which was adjusted for
two-phase flows has been considered. Indeed the presence
of droplets in the carrier phase may be a source for








turbulence dissipation or production. The additional source
terms which characterize the direct interaction of mass,
momentum and species between particle and carrier gas
are explained in detail in [6]. The volume variation of the
carrier phase as consequence of the presence of kerosene
droplets is neglected. This assumption is acceptable, since
the droplet volume fraction is here below 10-4 [7]. Besides
the source terms due to the presence of particles, specific
turbulence modulation model has been introduced and
tested (see in [6] and therein quoted references). Thus a
full two-way coupling is ensured.

Evaporation and dispersion models
In the frame of this work, the droplets are captured
using the Lagrangian procedure, in which all numerical
droplets are tracked by solving their equations of motion
that include only the drag and gravitation forces. In order
to quantify the instantaneous fluid velocity seen by the
droplets and its effect on the droplet distribution one
should model the Root Mean Square (RMS) values of the
fluid parcel velocity at the droplet location. This can be
adequately done using a stochastic Lagrangian process.
The model used in the frame of this work is the so called
Markov-sequence dispersion model [8]. It is based on the
computation of the fluid element instantaneous fluctuation
along the particle trajectory using two correlation factors,
namely the Lagrangian and Eulerian correlation factors
denoting the time and spatial correlation functions,
respectively. To avoid the phenomena of droplet
immigration to locations having low pressure, a drift
correction term has been considered [9]. With regard to
evaporation, the so-called Uniform Temperature (UT)
model by Abramson and Sirignano [10] has been applied.
This model is based on the film thickness theory. It does
not consider any temperature variation in the interior of the
droplet (homogenous temperature). The UT model
describes the evolution of the droplet temperature and
diameter, i.e. evaporation rate and energy flux through the
liquid/gas interface.

Combustion modelling
Combustion takes place after fuel evaporation, and
occurs only where vapour and air mix. For partially
premixed pre-vaporized combustion, one realizes two main
features, namely inhomogeneity of the equivalence ratio
and the velocity of the flame propagation. In the frame of
this work, these two conditions are satisfied, since the fuel
and the oxidizer necessitate certain time for the mixing, the
mixture forms a spatial variation of the equivalence ratio of
the fresh gas. The laminar burning velocity is also
considered as a function of the equivalence ratio. Various
models based on premixed combustion have been extended
to account for such partially premixed features (e.g. [21])
as it will be presented in the following sub-section.
Once modified, the Bray-Moss-Libby (BML) model is
a suitable combustion model for the simulation of the
multiphase combustion in this kind of configurations. This
achieved by extending the BML theory through the
coupling with the mixing transport according to [22]. The
transport equation for the progress variable, given by
equation (1), has been solved.


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

9 (pc) 9(piic) _i -\ ~
-t + x x H-pcu +w
t x(1)
where c denotes the progress variable defined as:
T -T

Tb -T (2)

In equation (2) the indices u stand for the unburnt, and b for
burnt part of the flame, respectively. Equation (1) contains

also the mean reaction rate term c which is modelled as
following:
=uS (k32 D-3 3/4 D
S0 (3)

where the constant CL is taken to 0.41 in all simulations and
the fractal dimension D is set to 7.7/3. In order to extract the
physical products, results of conditioned combustion
b
products must be multiplied by the probability P of
being behind the flame front.

Y4 ZPPPb (z (4)

In the context of the BML model this probability can be
directly related to the progress variable c, so that

Y(z,Pb)= Y z)c
S(5)

Assuming a bimodal PDF of c it can be shown that [12]


c (,, Z-2)


Sb (,z11)


In addition to the Favre-averaged equations of mass,
momentum, and turbulent quantities, a Favre-averaged
mixture fraction equation z and a Favre-averaged
ff2
equation for its variance z have to be considered. The
Favre-averaged equation for the mixture fraction can be
written as follows [11]:


pz a(pu,zT ) a ( 8z .
ati + pD pu- z +-
9t 9x, 9x, x z


After neglecting the molecular diffusion terms, the final
closed equation for mixture fraction variance yields to:

a(z "2) __ __
+ p +-g "2
at ax, ax S- 8x, k
(8)

When droplets vaporization occurs due to the local sources
of fuel, the mixture fraction z is not a conserved scalar. It
Ss S,
results in two additional source terms ( "' and P)









"2
appearing in the transport equation of z and z The

S- S,
source terms z'P and 'z p for the mixture fraction
(equation 7) and its variance (equation 8) are given in [13]
as follows:

m,, N,
",p --2 -7 --
S vi (9)
V (9)


s" z 1-2z)
(10)


Choice of surrogate for kerosene
Jet fuels are composed of hundreds of aliphatic and
aromatic compounds. Developing a reaction mechanism
under the consideration of all this components is a huge
task that is still under investigation. One should use
therefore a surrogate for kerosene in order to reduce the
size of the reaction mechanism. The major components
of kerosene are alkanes, aromatics and alkenes, thus
surrogates are mainly based on these. Huang et al.[14]
used N-dodecane as a surrogate for jet fuel. Zhou et al.
[15] studied the thermal decomposition of n-dodecane for
comparison with other fuels. Honnet et al. [16] used a
mixture of n-decane 80% and 1,2,4-trimethylbenzene 20%
by weight to model jet-A. In this study we used a detailed
chemical reaction mechanism for N-dodecane as surrogate
for kerosene, describing the combustion process. It
involved 57 species and 281 reactions. The Lewis number
is set to the unity and the strain rate equals 100/s. For the
generation of the look up table a presumed pdf has been
considered. The laminar burning velocity for N-dodecane
at the mixture preheat temperature of 400 K has been
measured by Kumar et al. [17] and is given in Figure 1 as a
function of equivalence ratio. The flammability limits are
about 0.7 and 1.4 respectively for the lean and rich
compositions.

n-Dodecane/Air Mixtures
100 ...............=tK .
90
S80
70
6I
50 .
S40
30
20
0.6 0.7 0.8 0.9 I 1.1 1.2 1.3 1.4 1.5
Equivalcncc ratio I-]

Figure 1. Laminar flame speed of n-dodecane/air
mixtures with unburned mixture temperature of
360K [17]


Configuration and boundary conditions


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

The geometry of the configuration for the premixing
and partially premixed pre-vaporization of kerosene droplets
is shown in Figure 2. The burner is composed of two parts:
the pre-vaporization zone and the combustion tube. Within
the first one, the kerosene is fed to an ultrasonic nozzle. The
carrier phase (air) is heated by a set of sinter metal plates. It
enters the pre-vaporization after being accelerated by the
nozzle and it does entrain the dispersed phase [2]. The
droplets are being injected with a Sauter mean diameter of
50 m. They are subjected to heated environment; thereby
they change their physical state and evaporate. The kerosene
droplets are initialized with 90% slip velocity of the fluid
element. The mixing between the air and the vapor takes
place along the distance L (Figure 2). Downstream the
mixing zone, a second nozzle (54mm) is used to increase
the flow velocity and prevent flame flash back. The main air
volume flow rate was 300 In/min (normalized 1/min) and its
temperature equals 900C. The kerosene nozzle uses 20
In/min additionally air for the amelioration of the spray
dispersion during the injection. The mixture is ignited by a
hot wire ring. In the wake of this ignition source a stable
flame develops and spreads over the cross section further
downstream.
A three-dimensional CFD-code in which the equations
for the gas phase are solved by finite volume method has
been used. The diffusion terms are discretized with flux
blending schemes on a non orthogonal block-structured grid.
The velocity-pressure coupling is accomplished by a
SIMPLE algorithm. The whole system is solved by the
SIP-solver. The Lagrangian equations for droplets are
discretized using first order scheme and solved explicitly.
The source terms for the gas phase are computed in each
cell with the contributions of all relevant droplets. The
interaction between the continuous and the dispersed phase
consists in couplings between two codes. After the
convergence of gas phase, the gas variables are kept frozen
and all the droplets representing the entire spray are injected
in the computational domain. Due to the presence of the
droplets source terms, the conventional residuals are
characterized by a jump after each coupling. To avoid
oscillations, an additional under-relaxation technique should
also be employed for droplet source terms.
The droplet injection is based on a stochastic approach
by considering the droplet mass flux and the droplet size
distributions obtained from the experimental measurements
at the inlet near the nozzle exit. In this work the simulations
were performed using monodispersed particles. The overall
mesh for the single annular combustor is about 522 000
control volumes (Figure 3). The inlet conditions for the
turbulent kinetic energy are calculated using a turbulence
intensity of 10% of the resultant velocity through the inlet.
The distribution of the dissipation rate is estimated using the
expression


4 k3/2
e 0.41-Ar


Here the turbulent length scale was assumed to be
equal to the hole's diameter or inlet's opening. The mixture
fraction boundary conditions are set to zero at all inlets
(2 = 0), since the injected air does not contain any fuel and
the variation of mixture fraction is originated only by the









produced vapor.
Since there was no temperature variation at the inflow,
the progressive variable was set to zero at all inlet
boundaries except at the position of the hot wire ring for
the ignition or stabilization and insurance of continuous
combustion. The global equivalence ratio equals 0.7.


Figure 2. Geometry of the
partially premixed
pre-vaporized combustor
[2]


Figure 3.Numerical Grid


Results and discussion

As part of the validation procedure, the results for the
evaporation, dispersion and entertainment of the dispersed
phase were compared to experimental data and published
in [18]. As operating condition for the reference simulation,
the carrier phase temperature is set to 900C, and the length
of the pre-vaporization zone equals 0.8 m. The number of
the numerical droplets are 160 000 within one coupling.
When plotted, the properties of the dispersed phase
featured a smooth profile, i.e. the statistics were
completely reliable. Increasing the droplet number did not
change the results but enhanced the computing time.
Figure 4 shows the temperature distribution along the
axis of the combustion tube, thereby "0.0 m" represents the
position of the ignition source. The flame lift-off agrees
very well with the experimental measurements, whereas
the temperature maximum values show a A T of ca. 250 K.
This deviation is not originated from the negligence of
radiation, since the experimental data were corrected and
the effect of heat losses due to radiation were accounted for.
Nevertheless, one should mention at this stage, that the
temperature measurements were performed using
thermocouple elements and the data plotting showed very
high fluctuation (of ca. 450 K) with high frequency. This is
of course a strict limitation of the thermocouple, and
therefore the temperature measurements were effected with
a considerable error. This fact was also confirmed by the
experimentalists.

a) Influence of the evaporation zone length
In order to study the influence of the variation of the


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

pre-vaporization zone length on the combustion process,
two additional simulations were performed, with L=0.6m
and L=0.7m. The number of grid cells and tracked droplets
are plotted in Table 1. Experimentally the distance for the
pre-vaporization was adjusted by adding tube elements that
are interconnected by aluminium rings.


2000

16 00 -" n m r n
1800


6100 -
2400



60O

400 num.rfO-nueq
200 ex *
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 04
L nm]

Figure 4. Temperature distribution along the
axis of the combustion tube

In contrast the mass of kerosene droplets which arrive
to the combustion tube and evaporate there increases. This
phenomena influences the mixing process very strongly and
a new field of mixture is obtained, namely the first part of
the configuration exhibit less richer mixture whereas close
to the ignition source the concentration of vapor increases
and the mixture strides the flammability limit. Thus the
combustion process starts earlier than the reference one. It is
also to important to note that using shorter pre-vaporization
zone the maximum temperature increases of AT=100K.
This is to be explained by the fact that the mixture in the
combustion tube was slightly moved toward the
stochiometric value since the amount of kerosene vapour at
the second part of the configuration increased too.
The experimental data presented similar results of the
temperature enhancement. One observes also that the
combustion process takes place earlier compared to the
reference measurement.

2200 ii
2000 0 ..''- Co o t
1800 ..
1600 Q Ai ---.- O
1400

S1000 num-LO.6 --
num-LO.7 -
800 num-LO.8 ----n----
600 exp-LO.6 ----O-
400 a ; exp-LO.8 ---
200 '
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
L [m]

Figure 5. Temperature profiles for different lengths of the
pre-vaporization zone (num. & exp.)

Figure 5 shows the results of the temperature profile for
different pre-vaporization length "L". One observes a clear
diminishing of the flame lift-off with decreasing value of
"L"; for reduction of A L=25% the flame reduces its
lift-off with A h=0.05m. Due to the reduction of the









pre-vaporization length "L", droplets have less time to
endure there, thus the evaporation degree is in turn
reduced.

Table 1. Grid cells and droplet number for different L

L [m] Grid cells Droplet Nbr.
[10001 [1000]
0.6 475 160
0.7 500 160
0.8 522 160

b) Influence of the droplet diameter
The boundary conditions provided by the
measurements are limited for monodispersed droplet with a
Sauter mean diameter of 50 m. The question that rises is
what would happen if droplets change the size
distribution? The diameter size variations happen by
changing the geometry of the nozzle or even by changing
the operating conditions. In gas turbine the Sauter mean
diameter has often a value of 20 / m [19]. Therefore
simulations with a relatively smaller diameter are worth.
Figure 6 presents the temperature profile for the case of a
Sauter mean diameter at the inlet equal 20 / m. One
observes a very high temperature at the beginning of the
combustion tube, which is originated to a flame flash back.
The small droplets (20 1 m) evaporate much faster than
these with 50 m. They are not able to reach the end of
the evaporation zone, thus the mixing takes place in the
early stage of the pre-vaporization zone exclusively. After
ignition the flame propagates close to the hot wire,
afterwards it come back to the pre-vaporization zone and
stabilizes there. Unfortunately there was no experimental
data for this class of droplet diameter; however it is of
particular interest to study the limit of the droplet diameter
at which flame flash back occur. One should mention here,
that the evaporation model at the same operating
conditions and same geometry has been already validated
[18].


"00 " -. .... .
1600 D -
1400
I o
200D
1000 ..
sn
aoo o

400 ......
200
0 Ms O a015 2 0a25 03 035 04
L Iml
Figure 6. Comparison of the temperature profiles
for droplet diameter 50 u m (refO), 20 u m (dp20)
and exp. (50 P m)



c) Influence of the flammability limit
The flammability limits of fuels depend generally on
different parameters e.g. temperature, pressure etc. The
flammability limits are not absolute, they depend also on
the type and strength of the ignition source, therefore a


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

slight variation on the flammability limits of dodecane,
which feature different volatility than kerosene is worth to
investigate. The upper and lower limits of the kerosene
surrogate were set to 0.65 and 1.45, respectively.
Figure 7 illustrate the influence of enlarging of the
flammability limits. It is to remark that the flame lift-off has
been reduced with Ah =0.02m, whereas the temperature
profile remained unchanged. Since the new laminar velocity
has values different than zero for mixture starting from an
equivalence ratio equal 0.65, the chemical reaction was able
to arise for a leaner composition and than the combustion
started earlier than the reference case.
The upper flammability limit did not manifest any
influence on the results, because we are dealing with a lean
combustion.

2200
2000
1800
1600 -
1400
1200 -

000 -
WO m f
400 Cmff- exp. -

200 I I
S 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
L [m]
Figure 7. Effects of the flammability limit
enlarging (refO- 1 ) on the temperature distribution



d) Influence of the ignition location
Statistically steady turbulent flames require
stabilization mechanisms [11]. Instead of providing
continuous ignition using a heat source, which was
numerically performed by setting the boundary condition
for the progress variable to 1 at the inlet close to the hot
wire ring, one has changed the aspect using spark ignition.
Here it is very important to study where the spark should be
initialized. For that the boundary condition of the progress
variable is set to zero in all inlets and three locations for the
initialization of the progress variable were chosen (see
Figure 8). It is of particular importance to note that the
spark ignition, i.e. progress variable initialization, could not
be placed arbitrarily, namely two observations have been
made. 1) Flame flash back occurs when the progress
variable was initialized behind the hot wire i.e. within
pre-vaporization zone. 2) The combustion process did not
take place when the progress was initialized at a position
higher than 0.2m the hot wire forwards.
Figure 9 shows the temperature profile for the two
cases of initializations. When flash back happens the
temperature rises up to 1400K at 0.0 m. Figure 10 displays
the progress variable field in a center plane cross-section.
Due to the high value of the progress variable, one clearly
observes the flame flash back and thereby the temperature
jump in the pre-vaporization zone. In case of initialization
between the hot wire and a distance of 0.2m forwards, one
finds no difference to the reference simulations.


-










































































Conclusions


This work studied the influence of the evaporation
degree on the temperature evolution of partially premixed
pre-vaporized kerosene spray combustion. The numerical
simulations were performed in a RANS-spray context


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

module embedded in an Eulerian Lagrangian approach. The
Bray-Moss-Libby (BML) combustion model was extended
for the application of the LPP concept. The numerical
simulation of the flame lift-off showed a very good
agreement with the experimental measurements, whereas
the temperature maximum values showed a A T of ca. 250
K.
The decreasing of pre-vaporization zone length
involved diminishing of the flame lift-off. This was
originated by the reduction of the evaporation degree, which
generated different mixture fields, so that the combustion
process started earlier. It was also observed that using
shorter pre-vaporization zone the maximum temperature
increased of AT=100K. The experimental data presented
similar results for the temperature enhancement.
By changing the droplet Sauter mean diameter from 50
p m to 20 p m, one remarked a flame flashback. On the
other hand the enlarging of the flammability limits provoked
a flame lift-off reduction of A h =0.02m, whereas the
temperature profile remained unchanged. An augmentation
of the upper flammability limit did not manifest any
influence on the results as dealing with a lean combustion.
The fractal dimension "D", which is a model parameter in
the BML combustion model may demonstrate an important
influence on the prediction of the temperature profile. This
aspect and further investigations are left for future work.

Acknowledments

For financial support we gratefully acknowledge the ESA
consortium, the Deutsche Forschungsgemeinschaft (DFG)
through the Sonderforschungsbereich 568 (project A4) and
the FAUDI Stiftung through the project Nr. 75.

References

[1] M. J. Ramotowski, R. J. Roby, PE., Leo D.
Eskin., M. S. Klassen, PE. "Fuel Flexibility for Dry Low
Emission Gas Turbines Cleanly Burning Biofuels, Coal
Liquids and Petroleum Fuels" Power Gen International
2007
[2] S. Baessler, K. Moesl and T Sattelmayer: "NOx
Emissions of a premixed partially vaporized kerosene spray
flame" J. of engineering for gas turbine and power, July
2007, Vol. 129 /695-702
[3] O. Schafer, R. Koch, and S. Wittig: Flashback in
Lean Prevaporized Premixed Combustion: Non swirling
Turbulent Pipe Flow Study J. Eng. Gas Turbines Power,
July 2003 Volume 125
[4] N. A. Rokke and A. J. W. Wilson Experimental
and Theoretical Studies of a Novel Venturi Lean Premixed
Prevaporized (LPP) Combustor J. Eng. Gas Turbines Power
July 2001 Volume 123
[5] H. Nomura, M. Hayasaki and Y. Ujiie Effects of
fine fuel droplets on a laminar flame stabilized in a partially
prevaporized spray stream Proceedings of the Combustion
Institute Volume 31, Issue 2, January 2007, Pages
2265-2272
[6] Chrigui, M., Ahmadi, G, Sadiki, A., Progress in
Computational Fluid Dynamics, Special issue, 2004.
pp162-174


Figure 8. Three positions for the comb.
initialization


2200 i i i i
200
1860

1400
1200 -
1000 -
S80 num.re --
n00 num.refs-posi
num.refo-pos2 ---------
400 --- exp. -----
200 ---I ------ I-------- ---;, I--
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
L [m]

Figure 9. Influence of the initialization location on the
combustion process. RefO = continuous ignition, posi
= in the pre-vaporization zone, pos 2 = between the
hot wire and a distance of 0.2 m


x


Figure 10. Progress variable field






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

[7] Crowe CT, Troutt TR, Chung JN (1996): Annu
Rev Fluid Mech 28:11-43
[8] Walklate, P. J.: 1986, 'A Markov-Chain Particle
Dispersion Model Based on Air Flow Data: Extension to
Large Water Droplets', Boundary-Layer Meteorol. 37,
313-318.
[9] Kohnen G und Sommerfeld M. Chemische
Technik 50 (1998) Heft 5 pp:225-234
[10] Abramson B. and Sirignano W.A.:, Int. J. Heat
Mass Transfer, Vol 32, pp. 1605-1618, 1989
[11] Poinsot T. and Veynante D. "Theoretical and
numerical combustion" ISBN 1-930217-05-6 R.T.
Edwards, Inc. Philadelphia USA
[12] K.N.C Bray and PA. Libby: In PA. Libby and
F.A. Williams, editors Turbulent Reacting FLow, pp.
115-151. Academic Press, London, San Diego, New York,
1994
[13] Hollmann, C., Gutheil, E., Flamelet-Modeling
of Turbulent Spray Diffusion Flames based on Laminar
Spray Flame Library" Combust. Sci. and Tech., Vol. 135,
pp. 175-192, 1998
[14] H. Huang, D. R. Sobel, L.J. Spadaccini, AIAA
(2002), 3871.
[15] P. Zhou, O.L. Hollis, B.L. Crynes, Ind. Eng.
Chem. Res. 26 (1987) 852.
[16] S. Honnet, K. Seshadri, U. Niemann and N.
Peters: ,,A surrogate fuel for kerosene Proceedings of the
Combustion Institute Volume 32, Issue 1, 2009, Pages
485-492
[17] K. Kumar and C. Sung, "Laminar flame speeds
and extinction limits of preheated n-decane/02/N2 and
n-dodecane/02/N2 mixtures". Combustion and Flame
Volume 151, Issues 1-2, October 2007, Pages 209-224
[18] M. Chrigui, A. Zghal and J. Janicka Droplet
behavior with a LPP ambiance Proceeding of the 4th
International Conference on Advances in Mechanical
Engineering and Mechanics 2008
[19] Smiljanovski V and Brehm N. "CFD liquid
spray combustion analysis of a single annular gas turbine
combustor" ASME paper 99-GT-300 1999
[20] F.C. Gouldin, K.N.C. Bray and J.-Y. Chen:
Chemical Closure Model for Fractal Flamelets.
Combustion and flame. Vol. 77, pp. 241-259, 1989
[21] E. Schneider, A. Sadiki and J. Janicka:
Modelling and 3D-Simulation of the kinetic effects in the
post-flame region of turbulent premixed flames based on
the G-equation approach, Flow, Turbulence and
Combustion, Vol. 75, p. 191-216, 2005
[22] A. Maltsev, A. Sadiki and J. Janicka: A new
BML-based RANS modelling for the description of gas
turbine typical combustion processes, Progress in
Computational Fluid Dynamics, Vol. 4, (3-4) p. 229-236,
2004.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs