Professional Documents
Culture Documents
Modeling of n-Hexane and n-Octane liquid fuel jets in gaseous crossflow for MARK
evaporation, combustion and breakup evaluation
Abdullah A. AlZahrania,b,⁎, Ibrahim Dincera
a
Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario L1H 7K4, Canada
b
Department of Mechanical Engineering, College of Engineering and Islamic Architecture, Umm Al-Qura University, Al Abdeyah, Makkah 5555, Saudi Arabia
A R T I C L E I N F O A B S T R A C T
Keywords: This paper investigates the phenomena of liquid fuel jets in gaseous crossflow for two types of fuels, namely n-
Heat transfer Hexane and n-Octane. In this regard, a numerical model is developed to predict a droplet behavior including
Droplet trajectory, velocity, evaporation and combustion, size degradation, breakup time and radius of produced child
Fuel jet droplets. Therefore, the mass, concentration, energy and momentum conversation equations are derived to
Combustion
evaluate the droplet acceleration from initial conditions. The velocity distribution is then obtained through a
Crossflow
numerical integration of the acceleration over time. A further integration is made to determine the droplet
Atomization
Secondary breakup position. In addition, evaporation and combustion and Taylor Analogy Breakup (TAB) models are integrated to
assess droplet evaporation, combustion rate, and breakup behavior during the injection process. The professional
version of the Engineering Equation Solver (EES) software is used to solve the model which has the advantage of
providing the thermodynamic properties of the different fluids involved through predefined functions. The
behaviors of droplets are investigated for two injection cases: evaporation only and evaporation and combustion.
The results obtained are presented in the variations in trajectories, velocities, droplet size and surface tem-
perature corresponding to each case and type of fuel.
1. Introduction utilization.
The problem of liquid jet injection into crossflow has been ex-
The vast majority of our energy systems at present depends to some tensively investigated in the open literature, including jet penetration,
extent on a combustion process which involves a liquid fuel being in- trajectories, droplets size, velocities, and breakup. The comprehensive
jected into a combustion chamber especially within the transportation literature reviews have been reported in references [2–6]. Recently,
industry. Therefore, the development of these energy systems mandates Broumand and Birouk [7] have reviewed the liquid jet injection in a
deeper understanding of this problem in order to reduce emissions and subsonic gaseous crossflow. They considered literature on the primary
fuel consumption and thereby improve efficiency and sustainability. breakup, jet trajectory and penetration, breakup length, and droplets
Furthermore, the progressive advances in the computing capabilities features. In addition, numerous experimental and numerical studies
allow the relatively complex numerical solution to become visible. This have considered droplet trajectory and liquid get in crossflow [8–10].
opens the horizon for scientists to explore further and validate experi- Pai et al. conducted investigated primary breakup of turbulent liquid
mental and simplified theoretical results of multiphase flow problems, jets in a crossflow [11]. Balasubramanyam and Chen [12] introduced
in general, and for liquid fuel jets in crossflow, in particular. the finite-conductivity evaporation in their CFD modeling study of li-
The research studies conducted on the liquid jet atomization in a quid jet breakup in high-speed crossflow. The use of Taylor Analogy
crossflow substantially motivated by the various industrial applications. Breakup (TAB) model to simulate and predict droplet breakup has been
For example, in airbreathing jet engines such as ramjet and scramjet, a reported by O'Rourke and Amsden [13]. Turner et al. [14] studied the
liquid fuel is being injected into a turbulent gaseous crossflow for breakup processes of diesel fuel sprays under transient condition.
combustion. Moreover, a similar process also encountered in internal Rachner et al. [1] investigated the atomization of a plain jet of kerosene
combustion engines. In these processes, an efficient atomization of the fuel in air crossflow to evaluate the combustion characteristics in gas
injected fuel considerably promotes efficient combustion process that turbines. Furthermore, Prakash et al. [15] considered the breakup
yields less pollutant such as NOx [1] and allows for a higher fuel processes of spray in gaseous cross-flow involving pressure swirl. The
⁎
Corresponding author at: Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario L1H 7K4, Canada.
E-mail address: abdullah.alzahrani@uoit.ca (A.A. AlZahrani).
http://dx.doi.org/10.1016/j.icheatmasstransfer.2017.08.015
experimental setup was illustrated, and their results were reported the properties associated with temperature changes.
along with visual photographs categorized based on the breakup re- The numerical model is built based on three main sub-models. The
gimes and Weber number. Another extensive study carried out by first is a dynamic sub-model which determines the droplet acceleration,
Kaushal [16] which has been dedicated to turbulent fuel injection in velocity, and instantaneous position. The second is an evaporation and
internal combustion engines. The study included a simulation in KIVA4- combustion sub-model which evaluates the droplet evaporation/com-
mpi where both primary breakup model and secondary, model (TAB), bustion rate, droplet size degradation, and droplet surface temperature.
model are used. Furthermore, an experimental and numerical in- The third is the TAB breakup sub-model which predicts whether a
vestigation of droplet breakup and atomization of diesel and bio-diesel droplet breakup or not. It also predicts the size of the produced dro-
fuels in a cross flow air was conducted by Kim et al. [17] where TAB plets.
model was considered for secondary diesel droplet breakup. Behzad
et al. [18] reported simulation model developed to investigate the 2.1. Dynamic sub-model
surface breakup of a non-turbulent liquid jet injected into high-pressure
gas crossflow. They reported comparisons between the two fuels in The droplet trajectory in X- and Z-directions can be determined
terms of deformation and breakup regimes at different air flow rates. using the Lagrangian reference frame as
In the current study, a circular liquid fuel jet injected into a gaseous t
crossflow is investigated for two fuels n-Octane and n-Hexane which
have not been considered in earlier studies. The objective of this study
x = x0 + ∫ Ud dt
0 (1)
is to predict the trajectory of these fuels droplet as a function of the
t
position and the initial droplet diameter. In addition, the droplet ve-
locity, surface temperature, and evaporation and combustion rates will z = z0 + ∫ Wd dt
0 (2)
be evaluated. Lastly, the number of droplets and the breakup time will
also be predicted utilizing TAB model. where x0 and z0 are the initial droplet position components and Ud and
Wd are the injection velocity components. The instantaneous droplet
2. Problem formulation acceleration can be derived from the momentum equation (droplet
motion equation) as follows
The circular liquid fuel jet injecting normal to an air crossflow is d (md v )
shown in Fig. 1 with respect to the two Cartesian coordinates and the dt
= ∑ F = Fsurface + Fbody + Fcollisional (3)
two corresponding velocity components. The schematic representation
of the jet in crossflow problem illustrates how a gaseous flow moving at where the body forces are Fbody are only due to gravitational
certain speed flows over a liquid jet causing atomization and deflection Fbody = md g = ∀p ρf g (4)
in the produced droplets' path.
The mathematical models are developed and numerically to be The surface forces Fsurface is given by
solved using Engineering Equation Solver (EES) [19]. The use of EES Fsurface = Fdrag + Flift + Fadded mass + Fhistory (5)
software enables an accurate assessment of the properties of the fluid
due to the built-in functions that consider the instantaneous changes in In the analysis, we make some assumptions as follows:
where
Fdrag = −3π D μf f Vrel (7)
with
f = 1 for Re ≪ 1 (8)
Re
4 Re
f=1+ +
1+ Re b (9)
While the shear force Fshear is equivalent to bouncy force and given
as
Fshear = −ρf ∀p g (10)
dv −3 π D μ f fVrel + ∀p ( ρp − ρf ) g
=
dt m (11)
204
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
d 2 dYO
(r ρv ) = 0 YO − 4π ρg δg r 2 = mϕ
dt (16) dr (26)
Conservation of species: dT
m Cp, g (T − Ts ) − 4 πλ g r 2 = −m (L − q)
dr (27)
d ⎡ 2⎛ dY
r ρv Yi − ρ D i ⎞ ⎤ = 0
⎢ ⎝
dt ⎣ ⎥
dr ⎠ ⎦ (17) Then, the flowing equation can be obtained by equating the two
zones I]II
Conservation of energy:
d ⎡ 2⎛ dT ⎛4 π ρg δg r 2 dT ⎞ − ⎛4 π ρg δg r 2 dT ⎞ = mq
r ρv Cp (T − Tr ) − k ⎞ ⎤ = 0 ⎝ dr ⎠ I ⎝ dr ⎠ II (28)
⎢ ⎝
dr ⎣ dr ⎠ ⎥
⎦ (18)
The Clausius-Clapeyron equation is commonly used to give the
where
following:
n
∑ Yi = 1 YF = YFs (Ts ) (29)
i=1 (19)
Then, B is called the transfer number as follows:
Integration of the mass conservation gives
cp (T∞ − Ts ) + q∅Yo ∞
ṁ B≡
r2 ρ v = = const L (30)
4π (20)
Moreover, the droplet burning rate obtained as
where ṁ is the net mass transfer rate of the drop and for n number of
species which represent fuel and oxidizer species. The mass transfer rate 8k
β= ln (1 + B )
from droplet surface (zone I) outward ambient (zone II), as shown in ρl cp (31)
Fig. 2, is defined as species mass flux as
Finally, the diameter squared law is written as
ṁ F = x f ṁ (21)
D2 = Do2 − βt (32)
ṁ O = x O ṁ (22)
where in the case of no combustion (q = 0), the transfer number be-
The heat transfer coefficient is defined as
comes
cp (T∞ − Ts )
B≡
L (33)
The Taylor Analogy Breakup (TAB) breakup model has been utilized
to predict droplets breakup and child droplets diameters. TAB model
was developed by O'Rourke and Amsden [13] considered among the
most used models for analyzing droplet breakup.
Some relative parameters and dimensionless numbers to be defined
before undertaking the formulation of TAB model are surface tension,
Weber number, and Ohnesorge number. The surface tension is defined
as a force per length acts parallel to the liquid surface
205
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
ρg Vd2 Dd Table 1
We = Crossflow properties and droplet initial state assumptions list.
γlg (35)
Crossflow properties
and the Ohnesorge number as
Gas type Real air
μd Wed Velocity in X direction, Ug (m/s) 0
Oh = =
ρd σ Dd Red (36) Velocity in Z direction, Wg (m/s) − 38
Gas molar weight, Mg (kg/kmol) 28.96
where Dd is the droplet diameter, Vdis the impact velocity, ρl is liquid Viscosity, μg (kg/kmol) 27.33 × 10− 6
density, and γlg liquid–gas surface tension. The critical Weber number is Density, ρg (kg/m3) 13.8
Specific heat, Cp , g (kJ/kg K) 1.039
calculated as
Thermal conductivity, kg (W/m K) 0.04
Cf Droplet initial state
Wec = We Position in X direction, X0 (m) 0
Ck CD (37)
Position in Z direction, Z0 (m) 0
The TAB model is formulated analogically to spring-mass system as Velocity in X direction, Ud , 0 (m/s) − 2.4
Velocity in Z direction, Wd , 0 (m/s) 0
mx¨ = F − kx − dx ̇ (38)
where the external force
3. Results and discussion
F ρg u2
= CF The results of the present study are based on the model developed
m ρd r (39)
earlier which is composed of three sub-models: 1) dynamic sub-model,
The spring 2) TAB breakup sub-model, and 3) evaporation and combustion sub-
k C σ model. The three sub-models are integrated into one model and solved
= k 3 simultaneously for two types of fuels n-Hexane and n-Octane, which
m ρd r (40)
consider two cases with and without combustion for each fuel. The
The damper becomes droplet dynamic sub-model is concerned with droplet acceleration,
d μ velocity, and trajectory, while the TAB breakup sub-model assesses the
= Cd d 2 breakup criteria of droplets and evaluates child droplets size if a
m ρd r (41)
breakup occurs. The evaporation and combustion sub-model de-
The constants CF, Ck, and Cd are dimensionless numbers. Further termines the droplet diameter and temperature changes over time in
introducing a breakup criterion of x > Cbr with y = x/Cbr. Thus, the according to the case of evaporation and/or combustion.
droplet breakup is assumed to occur when y > 1. Now, the equation The thermodynamic properties of the gaseous crossflow and the
can be written as droplet initial state are presented in Table 1. The crossflow gas is real
CF ρg u2 C σ Cd μd air which properties are provided by EES through built-in functions.
y¨ = − k 3y − ẏ The crossflow has a velocity in a direction perpendicular to the initial
Cb ρl r 2 ρd r ρd r 2 (42)
jet velocity. The jet position is taken and the reference point with co-
The solution results in ordinates x and z equal zeros.
The stoichiometric combustion reactions for n-Hexane and n-Octane
−t 1 y − Wec ⎞
y (t ) = Wec + e td ⎡ (y0 − Wec ) cos ωt + ⎛y0̇ + 0 sin ωt⎤
⎜ ⎟ in the air are given in Table 2. Based on these reactions the stoichio-
⎢ ω ⎝ td ⎠ ⎥ (43)
⎣ ⎦ metric ratios have been selected for the combustion cases. The liquid
fuel thermodynamic properties (droplet properties) are given in Table 3
ρ u2r 1 C μd σ 1
We = , = d , ω2 = Ck − 2 for both fuel n-Hexane and n-Octane. The n-hexane droplet trajectories
σ td 2 ρd r 2 ρd r 3 td (44) are shown in Fig. 3, where trajectories of the droplet with initial dia-
Considering the following boundary conditions: meters of 100, 80, 60, and 40 μm are plotted for two cases one involves
evaporation and another involves evaporation and combustion. The
We = ρg vd2 Dd dashed lines in Fig. 3 represented the case of evaporation and com-
y (0)
γlg (45) bustion while solid lines represent the case of evaporation only. Simi-
larly, n-Octane droplet trajectories are shown in Fig. 4 and for the same
cf
y′ (0) Wec = We set of droplet diameters including evaporation and combustion cases.
ck cD (46) It can be observed from both figures, namely Fig. 3 and Fig. 4 that
The size of child drops produced after a breakup is predicted in TAB the droplets with larger diameters penetrate more in the X direction,
model based on the energy conservation analysis equals the parent drop perpendicular to crossflow direction. Also, it can be clearly seen that
before breakup with the energies of the subsequent product drops after the smaller the droplet size, the sooner it deflects and take the gas flow
a breakup. direction. For the other cases, where combustion is considered to take
place, the droplets tend to penetrate less and completely vanish earlier
π
Eparent = 4 π r 2 σ + K ρ r 5 (y 2̇ + ω2 y 2 ) than the case of evaporation only.
5 d (47)
In the case of evaporation, all droplets considered were able to
r π travel between 0.6 and 0.8 m along Z direction the direction of the gas
Echild = 4 π r 2 σ + ρd r 5y 2̇
r32 6 (49)
Accordingly, the child droplet equation is given as
Table 2
r 8K ρ r3 6K − 5 Stoichiometric combustion reaction of fuels in air.
=1+ + d y 2̇ ⎛ ⎞
r32 20 σ ⎝ 120 ⎠ (50) n-Hexane
C6H14 + 9.5 (O2 + 3.76 N2)→6 CO2 + 7 H2O + 35.72 N2
where r32 is the mean Sauter Mean Radius (SMR) of the parent droplet,
n-Octane
K is a constant experimentally measured (K = 10/3), ρl is the liquid C8H18 + 12.5 (O2 + 3.76 N2)→8 CO2 + 9 H2O + 47 N2
density, and y = x/(Cbr). The constant Cb is given equals 0.5.
206
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
Table 3
Thermodynamic properties of n-Hexane and n-Octane fuels.
Fig. 3. Droplets trajectory in X- and Z-directions for different n-Hexane droplets initial
diameter.
Fig. 4. Droplets trajectory in X- and Z-directions for different n-Octane droplets initial
diameter.
207
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
Fig. 8. Variation in drag coefficient for different n-Octane droplet diameters. Fig. 11. Velocities variation with time for n-Hexane droplets.
208
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
the two figures reveals that n-Hexane and n-Octane have similar
burning rates.
Fig. 15 shows the predicted child droplets SMR radius for four dif-
ferent initial parent droplet sizes of n-Hexane fuel. It can be noticed that
a child droplets with SMR of 22, 17, 12 and 8 μm will be produced from
a parent with a diameters of 100, 80, 60, and 40 μm, respectivly. Fig. 13
is one of the results derived from the application of TAB breakup model.
Fig. 16 shows the variation in the droplet breakup time with the
increasing droplet radius. It can be seen that the droplet breakup time
increases with increasing the droplet radius. In other words, a droplet
with larger radius needs more time to breakup compared with smaller
one.
In Fig. 17, the number of the droplets produced is presented against
time. It can be seen that the predicted number of produced droplets is
declining with time since droplets reach stable conditions and cannot
further split. In addition, the reduction in droplet size due to evapora-
tion may also result in more stabilization of droplet.
Fig. 18 shows the increase in the droplets surface temperature for
Fig. 17. Number of produced child droplets with time.
the n-Hexane fuel under two cases evaporation and combustion as well
as for the n-Octane and also for two cases evaporation and combustion.
The variation in the surface temperature over time for a droplet of
100 μm diameter is assessed. The n-Octane shows slightly higher sur-
face temperature for both evaporation and combustion cases.
Fig. 18. Temperature temporal variation for n-Hexane and n-Octane droplets.
4. Conclusions
209
A.A. AlZahrani, I. Dincer International Communications in Heat and Mass Transfer 88 (2017) 203–210
Nomenclature References
∀ Volume (m3) [1] M. Rachner, J. Becker, C. Hassa, T. Doerr, Modelling of the atomization of a plain
a Acceleration (m/s2) liquid fuel jet in crossflow at gas turbine conditions, Aerosp. Sci. Technol. 6 (7)
(2002) 495–506.
A Area (m2) [2] N. Ashgriz, Handbook of Atomization and Sprays, Springer, 2010.
B The transfer number [3] M. Birouk, I. Gökalp, Current status of droplet evaporation in turbulent flows, Prog.
Cp Specific heat (kJ/kg-K) Energy Combust. Sci. 32 (4) (2006) 408–423.
[4] G. Faeth, Current status of droplet and liquid combustion, Prog. Energy Combust.
D Diameter (m) Sci. 3 (4) (1977) 191–224.
EES Engineering equation solver [5] G. Faeth, Evaporation and combustion of sprays, Prog. Energy Combust. Sci. 9 (1)
F Force (1983) 1–76.
[6] C. Law, Recent advances in droplet vaporization and combustion, Prog. Energy
g Gravitational force
Combust. Sci. 8 (3) (1982) 171–201.
h Convective heat transfer (W/m2-K) [7] M. Broumand, M. Birouk, Liquid jet in a subsonic gaseous crossflow: recent progress
k Thermal conductivity (W/m-K) and remaining challenges, Prog. Energy Combust. Sci. 57 (2016) 1–29.
[8] A. Flock, D. Guildenbecher, J. Chen, P. Sojka, H.-J. Bauer, Experimental statistics of
L Latent heat (kJ/kg)
droplet trajectory and air flow during aerodynamic fragmentation of liquid drops,
m Mass (kg) Int. J. Multiphase Flow 47 (2012) 37–49.
ṁ Mass flow rate (kg/s) [9] A. Mashayek, Experimental and Numerical Study of Liquid Jets in Crossflow,
Oh Ohnesorge number University of Toronto, 2006.
[10] J. Marchant, Calculation of spray droplet trajectory in a moving airstream, J. Agric.
P Pressure (MPa) Eng. Res. 22 (1) (1977) 93–96.
Re Reynold number [11] M.G. Pai, O. Desjardins, H. Pitsch, Detailed simulations of primary breakup of
s Specific entropy (kJ/kg) turbulent liquid jets in crossflow, Center for Turbulence Research, Annual Research
Briefs, 2008, pp. 451–466.
SMR Sauter Mean Radius [12] M.S. Balasubramanyam, C.P. Chen, Modeling liquid jet breakup in high speed cross-
t Time (s) flow with finite-conductivity evaporation, Int. J. Heat Mass Transf. 51 (15–16)
T Temperature (°C) (2008) 3896–3905.
[13] P.J. O'Rourke, A.A. Amsden, The TAB Method for Numerical Calculation of Spray
TAB Taylor Analogy Breakup Droplet Breakup, SAE Technical Paper (1987).
U X-axis velocity component [14] M. Turner, S. Sazhin, J. Healey, C. Crua, S. Martynov, A breakup model for transient
V Velocity (m/s) Diesel fuel sprays, Fuel 97 (2012) 288–305.
[15] R.S. Prakash, H. Gadgil, B. Raghunandan, Breakup processes of pressure swirl spray
W Z-axis velocity component in gaseous cross-flow, Int. J. Multiphase Flow 66 (2014) 79–91.
We Weber number [16] K.P. Nishad, Modeling and Unsteady Simulation of Turbulent Multi-phase Flow
Y Molar fraction Including Fuel Injection in IC-engines, (2013).
[17] S. Kim, J.W. Hwang, C.S. Lee, Experiments and modeling on droplet motion and
atomization of diesel and bio-diesel fuels in a cross-flowed air stream, Int. J. Heat
Greek letters Fluid Flow 31 (4) (2010) 667–679.
[18] M. Behzad, N. Ashgriz, B.W. Karney, Surface breakup of a non-turbulent liquid jet
η Efficiency injected into a high pressure gaseous crossflow, Int. J. Multiphase Flow 80 (2016)
100–117.
μ Viscosity (kg/m-s) [19] S.A. Klein, Engineering Equation Solver (EES) for Microsoft Widows Operating
ρ Density System; Academic Commercial version, F-Chart Software, Madison, 2002.
δ Change
β The burning rate
210