Sie sind auf Seite 1von 6

ILASS Europe 2013, 25 th European Conference on Liquid Atomization and Spray Systems, Chania, Greece, 1-4 September 2013

Modelling of Unsteady Effects on Jet Breakup

E.O. Diemuodeke 1 and I. Sher 2 1: School of Engineering, Cranfield University, Cranfield, Bedfordshire, UK

Abstract

The theory of jet instabilities has been developed under several assumptions, which include the assumption that the jets are steady. The common-rail diesel fuel injection systems utilizing pulsed injection have drastically improved the ability to lower emissions, noise, and fuel consumption. However, with the application of the injection techniques in modern engines, the unsteady effects introduced by the pulsed injection are yet to be fully investigated. These unsteady effects may be attributed to some of the deviations observed in the literature between theoretical predictions and experimental data. The present paper has been able to demonstrate through analytical means that unsteady effects of liquid jet have significant impacts on the instability and breakup of liquid fuel. The model was developed from the Navier-Stokes equation with linear perturbations, which considers transient base flow parameters. Results obtained have good agreement with experimentally obtained data for penetration length and spray angle.

Introduction

Fuel injector performance has a direct effect on the combustion efficiency, pollutant emissions and combustion instability of combustion systems[1]. Government legislations regarding emissions are becoming stringent. Therefore, there is a need for more effective design of fuel combustion systems. The effective design of combustion systems is a function of good understanding of fuel breakup processes. The aim is to introduce the liquid fuel in a high surface area to volume ratio for a stable combustion process. The process surrounding the injection of liquid through a small aperture and the subsequent breakup of the bulk liquid has proven to be very complex [2][3].

The atomization processes are mainly attributed to cavitation, turbulent and aerodynamic forces, but it is generally acknowledged that the aerodynamic forces are the major controlling factor of the atomization process[4][5]. The Kelvin-Helmholtz (KH) instability model for liquid breakup was suggested by Reitz and Bracco [6] on the basis of surface waves formation as a result of hydrodynamic and aerodynamic forces interactions. The model is constructed on the basis of a first order linear analysis of a KH instability growing on the surface of a cylindrical liquid jet that is penetrating into a quiescent incompressible gas with a relative velocity. Both the liquid and the gas are assumed to be incompressible, and the gas is assumed to be inviscid. The dynamic of the liquid jet and gas interaction are described by the linearization of Navier-Stokes equations for two-phase flow based on steady base flow parameters. The liquid phase is assumed as the disperse phase and the gas phase as the continuous phase. The detailed analysis, which can be found in [6], yields a dispersion equation relating the growth rate of a perturbation to its wavelength. The dispersion equation obtained is not amenable to analytical solution; however, numerical curve fit was sorted, which shows that there exists a single maximum in the wave growth rate curve and assumed that the maximum growth rate controls the liquid breakup.

The theory of jet instabilities has been developed under several assumptions, including an assumption that the jets are steady. However, in most practical engineering applications these jets are highly unsteady, and the transient effects are attributed to some of the inconsistencies between experimental data and theoretical data [4] [5][7][8][9]. The acceleration of the liquid during start-up of the convectional injectors is in the order of 10 6 [m/s 2 ] at the orifice exit for high Reynolds numbers. This transient effect on the liquid breakup mechanisms would be more severe in the modern fuel injection techniques, e.g pulsed injection (illustrated in Figure. 1), which are intended to mitigate exhaust emission and to improve fuel economy. Sazhin et al [4] have stressed the importance of transient effects on jet instability and breakups by incorporating the jet acceleration into the classical KH instability model. The solution method was further advanced in Turner et al. [5]. In this paper we present results of a reconstructed K-H model [10, 11], predicting the characteristic breakup and penetration lengths of an unsteady jet.

1

25 th ILASS Europe 2013

Modelling of Unsteady Effects on Jet Breakup

Pulsed Convectiona injection l injection Start-up Shut-down Injection velocity, u z [m/s]
Pulsed
Convectiona
injection
l injection
Start-up
Shut-down
Injection velocity, u z [m/s]

Time after start of injection t[s]

Figure 1. Velocity Evolution after Start of Iinjection

Mathematical modelling

It is assumed that the interaction between aerodynamic forces and the jet hydrodynamic forces induce surface waves, as shear flow, because of relative velocity between the liquid and gas. Considering the assumptions made by Reitz and Bracco [6] and Egger and Villermaux [12], one proceeds with the reconstruction of the classical KH instability model for liquid breakups by modifying the linearization process of the Navier-Stokes equation. The free-surface shear of a liquid jet indicating the linearization parameters is shown in Figure 2.

r Free- surface velocity, u zg =0

Nozzle wall z (0,R 0 ) Shear layer thickness u(z,r,t)
Nozzle
wall
z
(0,R 0 )
Shear layer
thickness
u(z,r,t)
r Free- surface velocity, u z g =0 Nozzle wall z (0,R 0 ) Shear layer
r Free- surface velocity, u z g =0 Nozzle wall z (0,R 0 ) Shear layer

Figure 2. Cylindrical material coordinate system model for a liquid jet

The Navier-Stokes equations for the 2D cylindrical material coordinate system shown in Figure 2 can be written, as follows:

The continuity equation:

(1)

(1)

The momentum equation:

r-direction momentum

(2)

(2)

z-direction momentum

(3)

(3)

The transient effects of the fuel injection are incorporated in the linearization process by decomposing velocity and pressure as follows:

process by decomposing velocity and pressure as follows: and (4) It is postulated that the growth

and

process by decomposing velocity and pressure as follows: and (4) It is postulated that the growth

(4)

It is postulated that the growth of perturbed parameters varies with time [10, 11], thus, the perturbed parameters have the following forms:

; ;
;
;

2

(5a,b,c,d)

25 th ILASS Europe 2013

The governing equations are subject to the following boundary conditions:

Modelling of Unsteady Effects on Jet Breakup

(1)

The kinematic boundary condition

 
(6)

(6)

(2)

The dynamic boundary condition at the interface of liquid and gas

 
With
With

where

With where
 

(7)

denotes the surface for which

denotes the surface for which has to be determined, it

has to be determined, it

follows thus

  (8)
 

(8)

(3)

The axisymmetric assumption

 
(9)

(9)

Solving this system with its boundary conditions for the dispersion relation, for a case of a quiescent gas phase gives ([11]):

It is assumed that the disturbance growth rate is approximately constant,
It is assumed that the disturbance growth rate is approximately constant,

velocity rate change is the jet acceleration, hence,

andvelocity rate change is the jet acceleration, hence, (10 ) , and that the base .

velocity rate change is the jet acceleration, hence, and (10 ) , and that the base

(10)

, and that the base

. With these assumptions and

(11)

(10 ) , and that the base . With these assumptions and (11) considering the real

considering the real part, , in the frame of temporal analysis, Eq.(11) is obtained:

The plot of Eq.(11), for constant axial velocity, shows that there exists a maximum (Figure 3). The dominant or

the liquid breakup controlling wave number,

is expected to occur at

breakup controlling wave number, is expected to occur at , which corresponds to the maximum growth

, which corresponds to the maximum growth rate,

,
,
occur at , which corresponds to the maximum growth rate, , [6]. 350 Uz=50 300 250

[6].

350 Uz=50 300 250 a=200 a=180 a=160 200 a=140 a=120 a=100 150 a=80 a=60 100
350
Uz=50
300
250
a=200
a=180
a=160
200
a=140
a=120
a=100
150
a=80
a=60
100
a=40
a=20
50
a=0
0
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
k[1/m]
w[1/s]

Figure 3. Variation of growth rate,

5000 k[1/m] w[1/s] Figure 3. Variation of growth rate, , with wavenumber, , and jet acceleration,
5000 k[1/m] w[1/s] Figure 3. Variation of growth rate, , with wavenumber, , and jet acceleration,

, with wavenumber,

, and jet acceleration, a[m/s 2 ] at constant velocity The breakup controlling wavenumber and the corresponding growth rate are, therefore, respectively given as

([11]):

growth rate are, therefore, respectively given as ([11]): (12) and (13) The unsteady jet breakup macro

(12)

and

rate are, therefore, respectively given as ([11]): (12) and (13) The unsteady jet breakup macro structure

(13)

are, therefore, respectively given as ([11]): (12) and (13) The unsteady jet breakup macro structure is

The unsteady jet breakup macro structure is obtained accordingly. The relation for the half spray angle, , in Trinh and Chen (2006) as proposed by Reitz and Bracco (1982) [6] is used:

where
where

(14)

is a constant that accounts for the nozzle entrance shape, which must be fitted experimentally [6], and

is the optimum wavelength.used: where (14) is a constant that accounts for the nozzle entrance shape, which must be

3

25 th ILASS Europe 2013

Modelling of Unsteady Effects on Jet Breakup

The spray angle can be obtained by geometrical considerations - approximating the half of the spray as an isosceles triangle with the height being half the penetration length[13][14], as follows:

where
where

is the upstream half image projected spray area and

The penetration length is obtained as:

spray area and The penetration length is obtained as: where is the spray penetration length. (15)

where

spray area and The penetration length is obtained as: where is the spray penetration length. (15)
spray area and The penetration length is obtained as: where is the spray penetration length. (15)

is the spray penetration length.

(15)

(16)

Results and Discussion

spray penetration length. (15) (16) Results and Discussion Transient experimental injection velocity, , presented in

Transient experimental injection velocity, , presented in [8][9] for a 3 hole 0.2 mm diameter valve covered orifice (VCO) diesel injector at different injection pressures, were used for the validation of the developed transient KH instability model.

Figure 4 shows the model predicted spray angle and experimental spray angle data [8] for transient fuel injection. Figure 4 shows that there is a good agreement between model predicted results and the experimental data. The spray angle is observed to be very high at the initial start of injection and later settled for a value around 17 o , which corresponds to the steady fuel injection region. It is observed that the 60 [MPa] injection pressure has a wider spray angle compared to higher injection pressures, this may be attributed to initial low penetration velocity into dense ambient and more time to adjust to the surrounding gases, which makes it less constrained to expand. The significant difference in the spray angle in the accelerating region, 0-0.5 [ms] after start of injection and the steady region shows that the unsteady effects have significant effects on the liquid fuel breakups.

have significant effects on the liquid fuel breakups. Figure 4. Spray Angle Variation with the Time

Figure 4. Spray Angle Variation with the Time after Start of Injection in an Ambient of Density 47 [kg/m 3 ] at various Injection Pressures.

Figure 5 shows the penetration length variation with injection time at various injection pressures in an ambient density of 47 [kg/m 3 ]. Figure 5 shows that there exists an agreement between unsteady KH instability model and the experimental penetration length [8]. Careful observation of Figure 5 reveals that there exists a maximum in the predicted penetration length by the KH instability model. This observation is in conformity with majority of experimental data presented in the literature [15]. In terms of the injection pressure, it is observed that the liquid penetration reached the maximum value faster with elevated injection pressure due to the higher jet velocity, which induces instability and breakup more rapidly.

4

25 th ILASS Europe 2013

Modelling of Unsteady Effects on Jet Breakup

Europe 2013 Modelling of Unsteady Effects on Jet Breakup Figure 5. Evolution of Penetration Length after

Figure 5. Evolution of Penetration Length after time of injection at various Injection Pressures in an ambient Density of 47 [kg/m 3 ]

Conclusions

The theory of jet instabilities has been developed under several assumptions, which include the assumption that the jets are steady. These assumptions may be attributed to some of the deviations observed in the literature between theoretical predictions and experimental data. The present work has been able to demonstrate through analytical means that unsteady effects of liquid jet have significant impacts on the instability and breakup of liquid fuel. Results obtained have good agreement with experimentally obtained data for penetration length and spray angle. Unsteady jets are particularly prevalent in modern fuel injection strategies, like pulsed injection.

Nomenclature

Symbol

Quantity

A

Area

A

Acceleration

I

Modified first-kind Bessel function

K

Modified second-kind Bessel function

L

Length

p

Pressure

R 0

Jet radius

Re

Real part

t

Timescale

u

Velocity

wavenumber

wavenumber

Greek letter

Function parameter Full spray angle Wavelength Smooth function Constant Perturbed boundary Density Surface tension Growth rate

Subscript

0

Zero-order

1

First-order

g

Gas

l

Liquid

max

Maximum parameter

5

Units

m 2

m/s 2

m

N/m 2

m

s

m/s

1/m

m

o

m

-

m

kg/m 3

N/m

1/s

25 th ILASS Europe 2013

opt

Optimum parameter

p

Penetration

r

Radial direction

z

Axial direction

Superscript

Perturbed parameter; derivative

References

Modelling of Unsteady Effects on Jet Breakup

[1]

Stiesch, G. Modelling Engine Spray and Combustion Processes: Heat and Mass Transfer. Berlin

[2]

Heidelberg: Springer Verlag, 2010. Gorokhovski, M.; Herrmann, M. “Modelling Primary Atomization,” Annual Review of Fluid Mechanics,

[3]

vol. 40, pp. 343366, 2008. Jiang, X.; Siamas, G.A.; Jagus, K.; Karayiannis, T. G. “Physical Modelling and Advanced Simulation of

[4]

Gas-Liquid Two-phase Jet Flows in Atomization and Sprays,” Progress in Energy and Combustion Science, vol. 36, pp. 131167, 2010. Sazhin, S.S.; Turner, M.R.; Healey, J.J.; Martynov, S. B. “Transient Diesel Fuel Jets and Spray:

[5]

Mathematical Analysis and Application,” in ILASS-Europe, 24th European Conference on Liquid Atomization and Spray Systems, Sept. 2011, 2011. Turner, M.R.; Healey, J.J.; Sazhin, S.S.; Piazzesi, R. “Wave Packet Analysis and Breakup Length

[6]

Calculations for an Accelerating Plannar Liquid Jet,” Fluid Dynamics Research, vol. 44, no. 1, p. 015503, 2012. Reitz, R.D.; Bracco, F. V. “Mechanism of Atomization of Liquid Jet,” Physics of Fluids, vol. 25, no. 10,

[7]

pp. 17301742, 1982. Bae, C.; Kang, J. “The Structure of a Break-up Zone in the Transient Diesel Spray of a Valve-Covered

[8]

Orifice Nozzle,” International Journal of Engine Research, vol. 7, no. 4, pp. 319334, 2006. Karimi, K. “Characterisation of Multiple-Injection Diesel Sprays at Elevated Pressures and

[9]

Temperatures,” PhD Thesis, School of Engineering, University of Brighton, 2007. Karimi, K.; Sazhina, E M.; Abdelghaffar, W A.; Crua, C.; Cowell, T.; Heikal, M R.; Gold, M. R.

[10]

“Developments in Diesel Spray Characterisation and Modelling,” in THIESEL 2006 Conference on Thermo- and Fluid Dynamic Processes in Diesel Engines, 13-15 September,, 2006. Sher, I. “Stability and Breakup of Transient Jets,” in ILASS-Europe, 6-8 September, 2010.

[11]

Sher, I., E.O. Diemuodeke, "Transient Jet Breakup Modelling", to be published, 2013.

[12]

Eggers and E. Villermaux, J. “Physics of liquid jets,” Reports on Progress in Physics, vol. 71, no. 3, p.

[13]

036601, 2008. Naber, D.; Siebers, D. L. “Effects of Gas Density and Vaporization on Penetration and Dispensation of

[14]

Diesel Spray,” SAE Paper, p. 960034, 2006. Payri, F.; Payri, R.; Salvador, F.J.; Bardi, M. “Effect of Gas Properties on Diesel Spray Penetration and

[15]

Spreading Angle for the ECN Injectors,” in ICLASS-Europe, 2-6 September, 2012. Ming, H.; Shenlun, L.; Haifeng, L.; Chia-fon, F. L. “Investigation on Spray Characteristics of Water Emulsified Diesel with Different Injection Pressure and Ambient Temperature,” in Proc. of ICLASS- Europe, 2-6 September, 2012.

6