Energy Conversion and Management 124 (2016) 104–115

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Effects of structure parameters on flow and cavitation characteristics

within control valve of fuel injector for modern diesel engine
Chao Wang a, Guo-Xiu Li a,⇑, Zuo-Yu Sun a, Lan Wang a, Shu-Ping Sun b, Jiao-Jiao Gu b, Xiao-Jun Wu b
a
School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing 100044, China
b
China North Engine Research Institute, Tianjin 300400, China

a r t i c l e i n f o a b s t r a c t

Article history: Cavitation is a common phenomenon in diesel injector and has a strong influence on the internal flow.
Received 21 January 2016 However, studies so far have focused on cavitation characteristics inside the nozzle. Its influence on
Received in revised form 17 May 2016 the flow during control valve opening remains still unclear. In the paper, a computational study focused
Accepted 3 July 2016
on the flow and cavitation phenomena within control valve has been reported and the effects of control
Available online 9 July 2016
valve’s structure parameters (including rounded edge, seal cone angle and outflowing control-orifice
structure) on the flow and cavitation characteristics have been investigated in detail.
Keywords:
Firstly the 3D model has been validated in terms of single injection quantity and fuel injection duration,
The effect of structure parameters
The flow and cavitation characteristics
showing a good consistency. And then, the development from sheet cavitation to cloud cavitation and the
Control valve relationship between cavitation, pressure and velocity has been discussed. Based on the numerical results
Numerical simulation obtained, it is shown that not only the variation of pressure but also the velocity is the important factor
which affects cavitation. The increase of the flow velocity reduces the pressure within the flow field
which can aggravate the development of cavitation. As cavitation region increases, the fuel flow is hin-
dered and the flow velocity decreases. However, the decrease of flow velocity has suppressed the devel-
opment of cavitation. All of those variations form a cyclical process.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Due to high thermal efficiency, wide power scope, low fuel con-
sumption and reliability, diesel engine is widely applied in various
fields, such as machinery, marine technology and automobile
industry [1–3]. Among the components of diesel engine, fuel injec-
tion system is the core [4–6]. Fuel injector is the most complex
core part of the fuel injection system, which plays important role
in fuel injection and its control process. The structure and flow
characteristics of fuel injector can affect the performance of the
whole fuel injection system, thereby affecting the impact perfor-
mance of diesel engine [7–9].
The flow and cavitation characteristics of fuel are the two main
factors affecting the performance of fuel injector. Therefore so
many investigations pay attention to this field [10–14]. Salvador
et al. [7] study the flow and cavitation characteristics of fuel inside
the nozzle under different degrees of needle lift using numerical
methods. It has been reported that, for different degree of needle
lift, cavitation domain inside the nozzle changes and mass flow
⇑ Corresponding author.
E-mail address: Li_guoxiu@yahoo.com (G.-X. Li).
rate around the outlet of nozzle will be influenced. Molina et al.
[9] compared the flow and cavitation characteristics inside the
round and the oval nozzles employing numerical modeling
method. Its results show that, inside the oval nozzle, cavitation
happens rarely and discharge coefficient is lower, whereas inside
the round nozzle the phenomenon is completely opposite that cav-
itation happens easily and discharge coefficient is higher. Wang
and Su [10] studied the unstable cavitation process caused by pres-
sure fluctuation inside the nozzle and showed that the occurrence
and development of cavitation will be delayed since process
depends not only on the local pressure but also on the flow rate
caused by pressure difference. Sun et al. [15] employed Quasi-
steady method to study the flow and cavitation characteristics of
fuel inside the nozzle and pointed out the factors affecting the rate
of fuel injection and mass flow rate. Salvador et al. [16,17] com-
pared the flow and cavitation characteristics of fuel around the
nozzle of microsac and VCO (different types of diesel engines)
and asserted that the influence of various nozzle and nozzle cham-
ber on those characteristics is different.
Studies so far discussed above have focused on the flow and
cavitation characteristics around the nozzle. However, few studies
have investigated the flow and cavitation characteristics within the
control valve. In fact these characteristics inside control valve have

http://dx.doi.org/10.1016/j.enconman.2016.07.004
0196-8904/Ó 2016 Elsevier Ltd. All rights reserved.
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115 105

Nomenclature

RB bubble radius am volume fraction of vapor

PB bubble surface pressure R net mass source term
P local far-field pressure ml kinematic viscosity of the surrounding liquid
Pv saturation vapor pressure S surface tension of the bubble
n bubble number per unit volume of liquid qm vapor density
Rc mass source terms connected to the collapse of the ql liquid density
vapor qm mixture density
Re mass source terms connected to the growth of the vapor

a direct influence on response characteristics, fuel injection quan-
tity, fuel injection pulse width and fuel injector rate, eventually
influence on the power and emission performance of the diesel
engine. Until recent years, the researches of flow and cavitation
characteristics inside control valve have got the attention of schol-
ars. It is difficult to perform the experiments on the flow and cav-
itation characteristics inside the control-valve. Therefore,
determining the cavitation and its occurrence area can be done
only by observing the component damage in the control-valve
due to cavitation.
Ferrari et al. [18] have studied the effects of bypass, pressure
fluctuation and different driving electric current on the quantity
of fuel injected. Duan et al. [19,20] have observed the damage
inside the control-valve caused by cavitation through scanning
electron microscopes. As demonstrated in Fig. 1(a), the inner wall
of the control-valve is damaged more seriously with many small
irregular holes. In order to demonstrate the intensity of this kind
of damage, the 1000-h reliable experiment about fuel injector
has been conducted and results are shows in Fig. 1(b). The damage
area is larger than the tangent area of the ball valve and the wall,
and the flow characteristics around the sealing area of control-
valve provide the possibilities of cavitation. Therefore, cavitation
has an important role in the damage of injector.
Previous works principally focused on the simulation study of
control valve of injector or experimental analyses of cavitation
phenomenon. But there are no sufficient studies on the demonstra-
tion of the structure’s effects on flow characteristics and cavitation
process within control valve. Therefore, the flow and cavitation
characteristics are studied with the numerical simulation method
in this paper. Using 3D model, the effect of structure parameters
(including rounded edge, seal cone angle and OA structure) on
the flow and cavitation characteristics within the control valve
are studied systematically.
2. Simulation method and related settings
2.1. General settings
Fig. 2 shows the structure of fuel injector. Based on the func-
tionality, it can be divided into electromagnetic valve component,
control valve component and fuel injector nozzle component.
Among the three sections, the control valve component (composed
of ball valve, inflowing control-orifice (OZ), outflowing control-
orifice (OA), control chamber and control-piston) controls the
opening and closing of the needle valve by the hydraulic effect.
The simulation calculations have been performed by ANSYS Flu-
ent [21–23]. The internal flow is simulated by the dynamic mesh
technique (combination of structured and unstructured grid) dur-
ing the opening of the control valve. Total number of meshes is
943,131 with a maximum size of 0.08 and a minimum size of
0.008. Some of the specific parameters are presented in Table 1.
Duo to the limitation of software, the number of layers of grid
between the ball valve and the valve seat are kept as two. There-
fore, the ball valve starts to move 0.02 mm ahead the valve seat
and lifts upwards 0.06 mm during 0.1 ms (see Fig. 3).
2.2. Cavitation model
In most engineering situations we assume that there are plenty
of nuclei for the inception of cavitation. Thus, our primary focus is
on proper accounting of bubble growth and collapse. In a flowing
liquid with zero velocity slip between the fluid and bubbles, the
bubble dynamics equation can be derived from the generalized
Rayleigh-Plesset equation as:

2

4ml dRB
2
d RB 3 dRB PB  P 2S
RB þ ¼   ð1Þ
dt
2 2 dt ql RB dt ql RB

Diameter of ball valve Cavitation area

(a) (b)
Fig. 1. Photos of cavitation damage on control-valve-seat with scanning electron microscopes (after 1000 h alternating fatigue test) [19,20].
106 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115

Therefore, neglecting the second-order terms and the surface

tension force, Eq. (1) is simplified to
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
dRB 2 Pm  P
¼ ð2Þ
dt 3 ql

Cavitation model had an exact expression for the net mass

transfer from liquid to vapor. The equation for the vapor volume
fraction has the general form:
@ !
ðq am Þ þ r  ðqm am u Þ ¼ R ð3Þ
@t m
The net mass source term R can be expressed in the general
form:

qm ql dam
R¼ ð4Þ
qm dt
Schnerr and Sauer used the following expression to connect the
vapor volume fraction to the number of bubbles per volume of
liquid

n 43 pR3B
am ¼ ð5Þ
1 þ n 43 pR3B

Combining Eqs. (4) and (5), R is expressed as:

3am ð1  am Þ qm ql dRB
R¼  ð6Þ
RB qm dt
Fig. 2. The structure of the high pressure common rail fuel injector: 1. Oil return Substituting Eq. (2) into Eq. (6), the net mass source term has
pipe; 2. Reset spring; 3. Coil; 4. High-pressure connection; 5. Armature; 6. Ball the following form:
valve; 7. OA; 8. Control chamber; 9. OZ; 10. Control-piston; 11. Nozzle needle; 12. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Fuel injector nozzles. 3am ð1  am Þ qm ql 2 Pm  P
P < Pv Re ¼  ð7Þ
RB qm 3 qm
In practical cavitation models, the local far-field pressure P is usu-
ally taken to be the same as the cell center pressure. The bubble sur- sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
face pressure PB is equal to the saturation vapor pressure PV in the 3am ð1  am Þ qm ql 2 P  Pm
P P Pv Rc ¼  ð8Þ
absence of dissolved gases, mass transport and viscous damping. RB qm 3 qm

Table 1
Basic settings of the simulation conducted in the present investigation.

Item Classification Setting and/or value

Geometry Quality of ball valve (g) 2.33
Seal diameter of control valve (mm) 0.6
Lift of ball valve (mm) 0.06
Seal cone angle (°) 120°
Aperture of OZ (mm) 0.25
Aperture of OA (mm) 0.27
Physical properties Diesel (main phase) Density (kg/m3) 835
Viscosity (kg/m s) 0.0025
Surface tension (N/m) 0.0200
Saturated vapor pressure (Pa) 360
Diesel steam (second phase) Density (kg/m3) 2.9  102
Viscosity (kg/m s) 3.1  106
Initial boundary Inlet Total pressure (bar) 1200
Static pressure (bar) 5
Outlet Static pressure (bar) 1
Wall boundary No-slip surface conditions
Moving boundary UDF
Model selection Turbulence model Turbulence model Realizable k–e model
Cavitation model Simulation method Multi-phase flow
Cavitation model Schnerr–Sauer model [24–28]
Drag coefficient of two phases Schiller–Naumann method [29–31]
Bubble number density 1.5e+08
Solver options Solver type Pressure solver
Time step 107 s
Solving time 0.11 ms
Maximum iterations 50
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115 107

Sealing-annulus Outlet boundary

Outflowing control-orifice Inflowing control-orifice

Fig. 3. Grid model for the internal channel of the fuel injector control chamber.

The bubble radius RB is given by

14

13
AVERAGED VOLUME FRACTION OF
am 3 1
RB ¼ ð9Þ
1  am 4p n 12

The mass transfer rate in Eqs. (7) and (8) is proportional to 10

am ð1  am Þ and approaches zero when am ¼ 0 and am ¼ 1. The only
VAPOR (%)

8
parameter which must be determined in this model is the number
of vapor bubbles per volume of liquid (n). 6

2.3. Validation of grid independence 4

2
In order to ensure the calculation precision and reduce the cal-
culation time, the grid independence is verified in this section. 0
According to the grid parameter in the Table 2, the simulation 0.01 0.02 0.03 0.04 0.05 0.06 0.07
result of the mass flow rate at outlet of OA and the averaged vol- CONTROL-VALVE LIFT (mm)
ume fraction of vapor are compared (as shown in Figs. 4 and 5). 0.69 million 0.79 million 0.94 million 1.15 million 1.29 million
According to the results of calculation, as the number of grid
Fig. 4. Comparison of the average volume fraction of vapor.
increases from 0.69 million to 0.94 million, the calculation results
have substantial changes, the averaged volume fraction of vapor
is reduced, and the mass flow also has difference. When the num- 17
ber of grid increases from 0.94 million to 1.29 million, the mass
flow rate at outlet of OA and the averaged volume fraction of vapor
are basically consistent, but the computing time is significant
MASS FLOW RATE (g/s)

increased. Therefore, the 0.94 million can take less time to meet
14
the requirements of computational accuracy.

2.4. Validation of the models

11
The calculation results needs to be corrected and testified
relying on the experimental data. Due to the limitation of the mea-
suring and testing equipment, the internal flow and cavitation
within the control valve cannot be directly measured. In conclusion,
under the same boundary conditions, the mass flow rate at outlet of 8
0 0.03 0.06 0.09 0.12
OA of the checked 1D fuel injector model (the model and main TIME (ms)
0.69 million 0.79 million 0.94 million 1.15 million 1.29 million
Table 2
Specific grid parameter. Fig. 5. Comparison of the mass flow rate at outlet of OA.
Total grid number Minimum length scale Minimum length scale
(million) (mm) (mm) parameters have been shown in Table 3) will be compared with that
0.69 0.0048 0.3368
of the 3D model to verify the accuracy of the latter one. The compar-
0.79 0.0045 0.2215 isons on calculation results (1D) and experimental data of injection
0.94 0.0042 0.1851 quantity and the injection duration are shown in Fig. 6.
1.15 0.0041 0.1825 The mass flow rate at the outlets of OA for 1D model and 3D
1.29 0.0040 0.1812
model are compared and results are presented in Fig. 7. It can be
108 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115

Table 3
1D fuel injector model and related parameters.

1D fuel injector model Item Classification Value

Operating condition Common rail pressure 300–1800 bar
Fuel injection pulse width 300–1600 ls
Common rail pipe Pipe length 0.2–0.6 m
Within diameter of pipe 14 mm
Control valve Quality of valve element 2.33 g
Control valve lift 0.06 mm
Seal cone angle 120°
Aperture of OZ 0.25 mm
Aperture of OA 0.27 mm
Fuel injector Initial tension of spring 10.7 + 0.55  0.05 N
Spring stiffness 25 N/mm
Needle valve stroke 0.35 mm
Quality of needle valve 3.07 g
Nozzle accumulator volume 0.01 cm3
Number of injection holes 8  0.2 mm
SINGLE INJECTION QUANTITY (mm³)

200 11 3 10
FUEL INJECTION DURATION (ms)

10 9
9 2.4 8
160
8 7
7

ERROR (%)
ERROR (%)

6
120 1.8
6 5
5
3.40 4
80 3.16 3.37 4 1.2
2.51 2.65 3
3 1.75 1.63
1.35 2
1.00 1.13 2
40 0.6 0.40 1
0.10 1 0.08
-0.21
0 0
0 -1 0 -1
0.7 1.1 1.5 1.9 2.3 0.7 1.1 1.5 1.9 2.3
CONTROL PULSE WIDTH (ms) CONTROL PULSE WIDTH (ms)
Simulation date Experimental data Error Simulation data Experimental data Error

(a) (b)
Fig. 6. Comparison of the 1D simulation data and the experimental data: (a) Comparison of single injection quantity; and (b) comparison of fuel injection duration.

observed that the trend of mass flow, calculated by the 1D model

and the 3D model, is consistent. The maximum error is 10%
20 30
approximately. It can be proved the 3D model is accurate.
The error is due to the effect of cavitation on the flow is
MASS FLOW RATE (g/s)

25
16 excluded in the 1D model. Before 0.03 ms, cavitation is compara-
tively weak in the control valve and the error between 1D and
ERROR (%)

20
3D models is within 1.5%. From 0.03 ms to 0.072 ms, cavitation
12 15 gets stronger, leading to the decrease of mass flow rate. After
11.78 12.19
9.80
10.70
9.78 0.072 ms, cavitation weakens again and the mass flow rate of 3D
6.32
10 model is almost equal to that of the 1D model.
8
4.04 4.28
2.00
2.87 5
1.13 0.86
0.28 3. Results and discussions
4 0
0 0.03 0.06 0.09 0.12
3.1. Effects of rounded edge within control valve on flow
TIME(MS)
characteristics
Calculated data (1D) Calculated data (3D) Error

Fig. 7. Comparison of the mass flow rate at outlet of OA between 1D simulation and From the discussions in the previous section, the flow charac-
3D simulation. teristics inside control valve have a great influence on the fuel
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115 109

injection function. In order to get the rules of structure parameter 20

affecting the flow and cavitation characteristics, the section study

MASS FLOW RATE (g/s)

how the rounded edge within control valve affects them.
16
3.1.1. Effects of rounded edge within control valve on the mass flow
rate at outlet of OA
Since the edges and corners within the OA and around the seal-
12
ing cone of the ball valve easily induce cavitation, the effects of
rounded edge on the flow and cavitation characteristics within
the control valve should be considered during the investigation.
Fig. 8 gives the comparison on the structure of control valve. 8
The mass flow rates at outlet of OA with and without rounded 0 0.03 0.06 0.09 0.12
edge are compared when the ball valve is opened and results are TIME (ms)
presented in Fig. 9. It can be seen that the mass flow rate increases Without rounded edge With rounded edge
gradually with the ball valve opens and the mass flow rate of OA is
Fig. 9. Comparison of the mass flow rate at outlet of OA with different rounded
higher with the rounded edge case than the other case. When the
edge.
valve opens completely, the mass flow rate at outlet of OA with
rounded edge case is 18.22 g/s, which is 13.14% higher than that
of the other case. It can be concluded that the flow characteristics 200 320
of the control valve with the rounded edge is superior.
270
170

VELOCITY (m/s)
PRESSURE (MPa)
220
3.1.2. Effects of rounded edge within control valve on pressure and
140
velocity 170
Fig. 10 shows that the pressure decreases while the velocity 110 120
increases at the outlets of the two control chambers. The pressure
70
at outlet of the control chamber with rounded edge is always larger 80
than that of the other case. The maximum pressure difference is 20
50
20 MPa. On the contrary, the velocity at the outlet without rounded -30
edge is higher. The maximum velocity difference is 124.3 m/s. 20 -80
According to the pressure contours at the center section of con- 0 0.02 0.04 0.06 0.08 0.1 0.12
trol valve component in Table 4, the pressure within the control TIIME (ms)
chamber decreases sharply when the ball valve is opened. Because Pressure(with rounded edge) Pressure(without rounded edge)
of the better flow characteristics and faster decompression within Velocity(with rounded edge) Velocity(without rounded edge)

the control valve with the rounded edge, at the same time, pres-
sure is lower than the other case and distributes evenly. In addi-
tion, the better flow characteristics are observed by the quickly
decrease of pressure within the control valve with rounded edge.
When the ball valve begins to lift, the pressure within the control
valve with rounded edge is slightly larger than that of the other
case. After 0.06 ms, it becomes converse. With the ball valve lifts
completely, the pressure decreases to 1 MPa within the OA in both
types of control valve.
There are 3 stages of the velocity contour within the control
valve in Table 3:
In the first stage, the better flow characteristics within the con-
trol valve with rounded edge make the pressure drop quickly, and
higher pressure difference between the control chamber and the
OA leads to higher flow velocity within OA.
In the second stage, compared to the case with rounded edge,
the other case need longer time to make the pressure stable. But
Fig. 10. Comparison of pressure and velocity at the outlets of control chamber.
the pressure difference between control valve and OA is weaker
in the case with rounded edge. Thus the flow velocity within OA
is less than the other case.
In the third stage, after the pressure becomes stable within both
the control valves, cavitation within the OA of the control valve
with the rounded edge case is greater than that of other case,
which hinders fuel flow to some extent and the velocity is a little
lower than the other case.
Moreover, the increased flow velocity reduces the pressure
within the flow field, which enables cavitation to emerge. The
increase of cavitation witnesses the decline of flow velocity. How-
ever, the decrease of flow velocity exerts an adverse impact on the
development of cavitation, but increases the fuel flow. All the vari-
ations form a cyclical process.

3.1.3. Effects of rounded edge within control valve on cavitation

characteristics
The curves in Fig. 11 show the average volume fraction of vapor
within these two types of control valves during the operation of
ball valve. It shows the average volume fractions of vapor first
increase and then decreases tending to be stable. When the ball
valve is raised to 0.3 mm, the average volume fraction of vapor
R 0.06
reaches to its peak. The peak value is 17% within control valve with
R 0.08 rounded edge; while it is 12% within the other case.
R 0.05
The main reason for this phenomenon is the design of rounded
edge reduces the throttle effect of OA and OZ as well as the reduc-
Without rounded edge With rounded edge tion of local pressure creates the condition for the inception of cav-
itation. Therefore, it suggests that the existence of rounded edge
Fig. 8. Comparison of the structure of control valve. can actually improve the whole flow characteristics of control
110 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115

Table 4
Contours of velocity and pressure within two forms of control valve.

Time Velocity Pressure

(ms) Without rounded edge With rounded edge Without rounded edge With rounded edge
0.012

0.06

0.11

valve lifts to 0.048 mm, average volume fractions of vapor within

AVERAGED VOLUME FRACTION OF

18
these different control valves are almost equal to each other.
16 As the ball valve lifts from 0.048 mm to 0.06 mm, cavitation
14 emerges on the OA’s surface, which eventually results in the
12 increase of average volume fraction of vapor. It is obvious that,
VAPOR (%)

compared with the other format, both the area and the degree of
10
cavitation within the OA with rounded edge are greater.
8 Similar to the case discussed above, the cavitation within con-
6 trol valve can be divided into three stages:
In the first stage, fuel temperature remains stable. As the ball
4
valve is lifted, cavitation occurs around the ball valve and develops
2 continuously since the pressure drops.
0 In the second stage, when the ball valve lifts to a certain posi-
0.01 0.02 0.03 0.04 0.05 0.06 0.07 tion, pressure tends to become stable. With the development of
CONTROL VALVE LIFT (mm) cavitation, bubble takes lesser time to collapse than the formation
Without rounded edge With rounded edge of cavitation and accompanying cloud cavitation collapses. The
average volume fraction of vapor reduces and the cavitation area
Fig. 11. Comparison of the average volume fraction of vapor between two forms of
decreases.
control valve.
In the third stage, when ball valve lifts completely, the flow
velocity within the OA increases. On the OA’s surface, pressure
valve component. However, it also aggravates the cavitation and reduces which makes the cavitation develop continuously.
influences the discharge coefficient of OA in a manner. Moreover, Based on the above discussion, it can be concluded that besides
Fig. 11 also shows the process of bubble collapse that takes lesser the pressure, both flow characteristics and structure also affect the
time than that of cavity formation. Therefore, bubble collapse does inception and development of cavitation within control valve.
more harm to control valve’s surface.
Following are some of the additional conclusions that are drawn 3.2. Effects of seal cone angle on flow characteristics
from Fig. 11.
As ball valve lifts from 0.02 mm to 0.035 mm, cavitation mainly Seal cone angle seriously affects the flow characteristics. This
occurs around the ball valve; bubble stick to the ball valve’s surface section studies three kinds of seal cone angle’s (as shown in
to form the sheet cavitation and expand along the surface. The Fig. 12) effect on flow and cavitation characteristics within control
average volume fraction of vapor increases continuously and cavi- valve.
tation within the control valve with rounded edge is more serious
than the other case. 3.2.1. Effects of seal cone angle on the mass flow rate at outlet of OA
When ball valve lifts from 0.035 mm to 0.048 mm, because of Fig. 13 shows the comparison of the mass flow rate with differ-
the unstable bubbles at the end of sheet cavitation, bubbles begin ent seal cone angles as the ball valve opens. With the ball valve
to break and move downwards to form cloud cavitation. Finally, lifts, the mass flow rate at outlet of OA increases constantly.
bubbles break continuously and cloud cavitation collapses, the Because the flow Characteristics in case of 110 are better than
average volume fraction of vapor decreases as well. As the ball the others, the mass flow rate attains the peak value of 17.42 g/s
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115 111

110° 120° 130°

Seal cone angle of 110° Seal cone angle of 120° Seal cone angle of 130°

Fig. 12. Comparison of different formats of seal cone angles.

20 higher than that of the other two cases. After ball valve lifts
completely, the pressure at outlets of the three types reaches sta-
MASS FLOW RATE (g/s)

bility and the value is about 35 MPa.

16
Fig. 14 also shows that, before 0.06 ms, the velocity at outlet of
control chamber with the 110° cone angle is the highest; after
12 0.06 ms, both the pressure and velocity becomes stable. Mean-
while, the velocity at outlet of control chamber (in the case of
8 the 120° and the 130° cone angles) increases continuously. When
the ball valve opens completely, the flow velocity in case of 120°
cone angle is 303.18 m/s, and the value in both 110° and the
4
0 0.03 0.06 0.09 0.12 130° cases are 289 m/s.
TIME (ms) Following are some of the inferences from the pressure con-
120° 110° 130° tours at Table 5. At 0.012 ms, pressure within control valve with
the 110° case is lower than that of 120° and 130° cases. At
Fig. 13. Comparison of mass flow rate of outlet of control valve. 0.06 ms, the pressure in the 110° case reaches stability, but the
pressure is maintained at a high level in the 120° and 130° cases.
After ball valve completely opens, the pressure in the three cases
(21% higher than that of the 120° cone angle case) at 0.09 ms after is stable. Therefore, when seal cone angle is 110°, pressure drops
the ball valve lifts. After that, pressure within the OA and the mass more quickly and flow characteristics are better, which benefits
flow rate are all stable. Meanwhile, the cavitation hinders the fuel the response characteristics of the injector.
flow and the mass flow rate decreases slightly. Contrast to the Furthermore, from the velocity contours, it is also obvious that
others case, the peak value of mass flow rate with 130° cone angle the velocity within OA increases quickly because of better flow
is only 14.97 g/s (7.47% lower than that of 120° case). In conclu- characteristics. From 0.012 ms to 0.06 ms, velocity within the OA
sion, the mass flow rate increases with the seal cone angle with the 110° cone angle is apparently higher than that of 120°
decreases. and 130° cases.

3.2.2. Effects of seal cone angle on pressure and velocity
Form Fig. 14, it can be observed that pressure at outlet of con-
trol chamber in case of 110° is lower than that of 120° and 130°.
As ball valve lifts, the velocity of depressurization in the 130° is
Fig. 14. Comparison of pressure and velocity at outlets of control chamber with
different seal cone angle.
3.2.3. Effects of seal cone angle on cavitation characteristics
Fig. 15 shows the comparison of the average volume fraction of
vapor inside the control valve with different seal cone angles. The
average volume fraction of vapor first increases, and then it
decreases and finally stabilizes. The average volume fraction of
vapor with the 110° cone angle case is always larger than that of
the other two cases and the maximum value is 18.43%. When the
ball valve is raised to 0.033 mm, the average volume fraction of
vapor with 110° cone angle case comes to its peak, about 6.1%,
which occurs a little earlier than that of the other two cases that
occurs when the lift is 0.036 mm. After all the average volume
fractions of vapor get to the peak, bubbles separate from ball
valve surface quickly to form cloud cavitation, and cavitation
moves with fuel flow. Because of the external pressure, the bubbles
begin to break up and the cloud cavitation collapses and
disappears.
The above mentioned phenomena can be explained from the
following aspects. At the beginning, cavitation within the control
valve with 110° cone angle is poorer and the effective flow area
is larger. Therefore, the velocity and the mass flow rate are higher;
pressure decreases more quickly. When velocity reaches to a cer-
tain degree, cavitation becomes severe within the OA; this is the
reason why the average volume fraction of vapor reduces to the
112 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115

Table 5
Contours of velocity and pressure within control valve with different seal cone angles.

Time Velocity Pressure

(ms) 110 120 130 110 120 130
0.012

0.06

0.11

20 3.3. Effects of OA structure on flow characteristics

AVERAGED VOLUME FRACRION

18
In this section, three types of OA structures (show in Fig. 16) are
16
compared to analyze the flow and cavitation characteristics. First
14
OF VAPOR (%)

being a straight hole with outlet’s and inlet’s diameter is

12 0.27 mm. Keeping outlet’s diameter unchanged, the second one
10 is cone hole with inlet’s diameter of 0.029 mm and the third one
8 is inverted cone hole with inlet’s diameter of 0.025 mm.
6
4
3.3.1. Effects of OA structure on the mass flow rate at outlet of OA
2 Fig. 17 shows the comparison on mass flow rate at the outlets of
0 all the three OA. It can be seen that as the ball valve lifts, the mass
0.01 0.02 0.03 0.04 0.05 0.06 0.07
flow rate tends to increase with similar tendency. Before 0.03 ms,
CONTROL VALVE LIFT (mm) the mass flow rate of the inverted cone hole case is higher than
120° 110° 130° that of the other two structures, which is attributed to its better
Fig. 15. Comparison of the average volume fraction of vapor within control valves
flow characteristics. Between 0.03 ms and 0.08 ms, the mass flow
with different seal cone angles. rate of the cone hole is slightly higher, which is caused by the
stronger influence of the quickly decreased pressure rather than
lowest point and then increases. In addition, in the 110° case, cav- cavitation. After 0.08 ms, the mass flow rate of the inverted cone
itation within OA is rather severe than that of the other two types hole increase quickly, which is 5% higher than that of the other
of control valves. two structures at the same time. After pressure within three

0.27mm 0.27mm 0.27mm

0.27mm 0.29mm 0.25mm

The straight hole The cone hole The inverted cone hole
Fig. 16. Comparison of different formats of OA.
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115 113

20 3.3.2. Effects of OA structure on pressure and velocity

MASS FLOW RATE (g/s)
The pressure at outlet of control chamber for all the three struc-
16 tures are compared in Fig. 18. It can be seen that the pressure in the
straight and cone hole cases decreases continuously when ball
valve opens. Within the inverted cone hole, before 0.06 ms the
12
pressure reduces and then gradually stabilizes. This proves that
the inverted cone hole case has better flow characteristics.
8 The pressure contours in Table 6 also testify the above argu-
ments. At 0.012 ms, pressure within the three control valves is
4 almost same. At 0.06 ms, although the pressure inside control
0 0.03 0.06 0.09 0.12
valve with the inverted cone hole is higher than that of the other
TIME (ms)
Straight hole Inverted cone hole Cone hole two types of control valves, it has been stable already. Moreover,
pressure within OA with the inverted cone hole is also higher.
Fig. 17. Comparison of mass flow rate at outlets of control valves with different OA Therefore, its response characteristics are better as well.
structures.
In Fig. 18, with the opening of ball valve, the velocity at outlet of
control chamber with straight and cone hole increases continu-
control valve becomes stable, within the inverted cone hole, little ously. Within the inverted cone hole case, before 0.06 ms, the
cavitation area has an little impact on the flow characteristics, velocity continues to increase. After that, velocity tends to be
which promotes mass flow rate. stable.

Fig. 18. Comparison of pressure and velocity at outlets of control chamber with different OA structure.

Table 6
Contours of velocity and pressure within control valve with different OA structures.

Time Velocity Pressure

(ms) Straight hole Cone hole Inverted cone hole Straight hole Cone hole Inverted cone hole
0.012

0.06

0.11
114 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115
AVERAGED VOLUME FRACTION
16
14
12
OF VAPOR (%)
10
8 (4) The cavitation is related not only to pressure and fuel flow
6
4 makes the fuel flow more inordinate.
2 (5) According to the comparison of mass flow rate, it can be con-
0 cone angle of 110° and inverted cone hole OA has better flow
0.01 0.02 0.03 0.04 0.05
CONTROL VALVE LIFT (mm)
Straight Hole Cone Hole Inverted Cone Hole
Fig. 19. Comparison of the average volume fraction of vapor within control valve
with different OA structures.
Velocity contours in Table 6 show that, the variations of veloc-
ity, within the three type of control valve, are similar when ball
valve just moves. After 0.06 ms, it begins to change. Velocity is
highest within OA with the inverted structure. This is caused by
the huge pressure difference at both ends of OA. Moreover, under
larger pressure difference, velocity within the OA with the inverted
cone hole is always higher.
3.3.3. Effects of OA structure on cavitation characteristics
Fig. 19 shows the comparison of the average volume fraction of
vapor within the control valve with three types of OA. The average
volume fraction of vapor in the cone hole case is the highest among
the three cases, which is 14.38%. However, the highest value of
average volume fraction of vapor for the straight hole is only
11.23%. The time taken for the average volume fraction of vapor
in the cone hole case to reach the peak value is smaller compared
to the other two types of OA.
Cavitation contours show the cavitation area within the three
control valves is almost similar. When ball valve lifts to
0.048 mm, velocity within OA with inverted cone hole is the high-
est, which causes negative pressure zone close to the wall. Conse-
quently, cavitation occurs in less time. When ball valve moves to
0.06 mm, cavitation begins to form within OA with straight hole
and cone hole. Meanwhile, cavitation within OA with the inverted
cone hole is more serious but around ball valve it is weak.
4. Conclusion
This paper has studied fuel flow and cavitation characteristics
systematically, and the main results are as follows:
(1) This paper compares the 1D fuel injector model with the
experimental data (fuel-injection quantity, control pulse
width) to testify the reliability of it, and then certifies the
accuracy of the 3D model by the comparison of the mass
flow rate at outlet of OA in 1D model. The results prove that
the 3D model is reasonable and scientific.
(2) When the ball valve is opened, the mass flow rate at outlet of
OA tends to increase in a nonlinear form due to the variation
of section area and cavitation. The mass flow rate increases
continuously with the increase of the section area, and cav-
itation within control valve impedes the increase of mass
flow rate.
(3) The cavitation can be formed around the ball valve because
of the geometric features of ball valve. It should be noticed
that the top area of ball valve is the flow around circular
cylinder phenomenon, not the cavitation area. After the ball
valve completely opens, the pressure becomes stable caus-
ing the cavitation collapsing. However, inside OA, due to
the existence of the low pressure area makes the cavitation
enhance gradually.
rate, but also to interactions with turbulence. The fuel flow
affects the shape of cavitation. And the collapsed bubble
cluded that the type of control valve with rounded edge, seal
0.06 0.07
characteristics.
References
[1] Zhang J, Zhang HG, Yang K, et al. Performance analysis of regenerative organic
Rankine cycle (RORC) using the pure working fluid and the zeotropic mixture
over the whole operating range of a diesel engine. Energy Convers Manage
2014;84:282–94.
[2] Lamaris VT, Hountalas DT. A general purpose diagnostic technique for marine
diesel engines-application on the main propulsion and auxiliary diesel units of
a marine vessel. Energy Convers Manage 2010;51:740–53.
[3] Labeckas G, Slavinskas S. Performance and emission characteristics of a direct
injection diesel engine operating on KDV synthetic diesel fuel. Energy Convers
Manage 2013;66:173–88.
[4] Lešnik L, Iljaž J, Hribernik A, et al. Numerical and experimental study of
combustion, performance and emission characteristics of a heavy-duty DI
diesel engine running on diesel, biodiesel and their blends. Energy Convers
Manage 2014;81:534–46.
[5] Li XR, Sun ZY, Du W, et al. Research and development of double swirl
combustion system for a DI diesel engine. Combust Sci Technol
2010;182:1029–49.
[6] Su LW, Li XR, Zhang Z, et al. Numerical analysis on the combustion and
emission characteristics of forced swirl combustion system for DI diesel
engines. Energy Convers Manage 2014;86:20–7.
[7] Salvador FJ, Martínez-Lopez J, Caballer M, et al. Study of the influence of the
needle lift on the internal flow and cavitation phenomenon in diesel injector
nozzles by CFD using RANS methods. Energy Convers Manage
2013;66:246–56.
[8] Monhan B, Yang W, Chou SK. Development of an accurate cavitation coupled
spray model for diesel engine simulation. Energy Convers Manage
2014;77:269–77.
[9] Molina S, Salvador FJ, Carreres M, et al. A computational investigation on the
influence of the use of elliptical orifices on the inner nozzle flow and cavitation
development in diesel injector nozzles. Energy Convers Manage
2014;79:114–27.
[10] Wang X, Su WH. Numerical investigation on relationship between injection
pressure fluctuations and unsteady cavitation processes inside high-pressure
diesel nozzle holes. Fuel 2010;89:2252–9.
[11] Jiang GJ, Zhang YS, Xiao G, et al. Study of the generated density of cavitation
inside diesel nozzle using different fuels and nozzles. Energy Convers Manage
2015;103:208–17.
[12] López J, Salvador FJ, de la Garza OA, et al. A comprehensive study on the effect
of cavitation on injection velocity in diesel nozzles. Energy Convers Manage
2012;64:415–23.
[13] Valdés JR, Rodrígueza JM, Monge R, et al. Numerical simulation and
experimental validation of the cavitating flow through a ball check valve.
Energy Convers Manage 2014;78:776–86.
[14] Sou A, Biçer B, Tomiyama A. Numerical simulation of incipient cavitation flow
in a nozzle of fuel injector. Comput Fluids 2014;103:42–8.
[15] Sun ZY, Li GX, Yu YS, et al. Numerical investigation on transient flow and
cavitation characteristic within nozzle during the oil drainage process for a
high-pressure common-rail DI diesel engine. Energy Convers Manage
2015;98:507–17.
[16] Salvador FJ, Carreres M, et al. Comparison of microsac and VCO diesel injector
nozzles in terms of internal nozzle flow characteristics. Energy Convers
Manage 2015;103:284–99.
[17] Salvador FJ, Carreres M, et al. Analysis of the combined effect of hydrogrinding
process and inclination angle on hydraulic performance of diesel injection
nozzles. Energy Convers Manage 2015;105:1352–65.
[18] Ferrari A, Mittica A, Spessa E. Benefits of hydraulic layout over driving system
in piezo-injectors and proposal of a new-concept CR injector with an
integrated Minirail. Appl Energy 2013;103:243–55.
[19] Duan L, Yuan SQ, et al. Injection performance and cavitation analysis of an
advanced 250 MPa common rail diesel injector. Int J Heat Mass Transf
2016;93:388–97.
 C. Wang et al. / Energy Conversion and Management 124 (2016) 104–115
[20] Duan L, Yuan SQ, et al. Cavitation analysis of control-valve in high-pressure
common rail injector. Agric Mach 2015;46(5):321–7.
[21] Som S, Aggarwal SK, El-Hannouny EM, et al. Investigation of nozzle flow and
cavitation characteristics in a diesel injector. J Eng Gas Turb Power
2010;132:042802.
[22] Lisowski E, Czyzycki W, Rajda J. Three dimensional CFD analysis and
experimental test of flow force acting on the spool of solenoid operated
directional control valve. Energy Convers Manage 2013;70:220–9.
[23] Sun ZY, Li GX, Chen C, et al. Numerical investigation on effects of nozzle’s
geometric parameters on the flow and cavitation characteristics within
injector’s nozzle for a high-pressure common-rail DI diesel engine. Energy
Convers Manage 2015;89:843–61.
[24] Arcoumanis C, Flora H, Gavaises M, et al. Cavitation in real-size multi-hole
diesel injector nozzles. SAE paper no. 2000-01-1249; 2000.
[25] Lee WG, Reitz RD. A numerical investigation of transient flow and cavitation
within minisac and valve-covered orifice diesel injector nozzles. J Eng Gas Turb
Power 2010;132:052802.
115
[26] Salvador FJ, Romero JV, Roselló MD, et al. Validation of a code for modeling
cavitation phenomena in diesel injector nozzles. Math Comput Model
2010;52:1123–32.
[27] Ficarella, Laforgia D, Landriscina V. Evaluation of instability phenomena in a
common rail injection system for high speed diesel engines. SAE technical
paper 1999-01-0192; 1999. p. 1–19.
[28] Zhu, JK, Chen Y, Zhao DF, et al. Extension of the Schnerr–Sauer model for
cryogenic cavitation. Eur J Mech B Fluids 2015;52:1–10.
[29] Martynov SB, Mason DJ, Heikal MR. Numerical simulation of cavitation flows
based on their hydrodynamic similarity. Int J Engine Res 2006;7:283–96.
[30] Martynov S, Mason D, Heikal MR, et al. Modelling of cavitation flow in a nozzle
and its effect on spray development. In: Proceeding of the 13th international
heat transfer conference. Sydney, Australia: Begell House Inc.; 2006.
[31] Roohi E, Zahiri AP, Passandideh-Fard M. Numerical simulation of cavitation
around a two-dimensional hydrofoil using VOF method and LES turbulence
model. Appl Math Model 2013;37:6469–88.