Spark plug-in Fiber LDV for turbulent intensity measurement
of practical SI engine
Yuji Ikeda1 , Hiroyasu Nishihara 1 , Tsuyoshi Nakajima 1 ,
Shinzou Mori2 , Takanari Matsuura 2 and Teruyuki Itoh2
1, Kobe University
Rokkodai, Nada, Kobe 657-8501 JAPAN
Tel & Fax: +81-78-805-5140
E-mail: ikeda@mech.kobe-u.ac.jp
2, Nissan Motor Co., Ltd.
Natsushima-cho, Yokosuka, Kanagawa
237-8523 JAPAN
Tel: +81-468-67-5162
Fax: +81-468-66-0472
E-mail: t-itoh@ mail.nissan.co.jp
ABSTRACT
A compact fiber LDV has developed in order to measure turbulent intensity near the spark plug. An
optical parameter of the FLDV was designed to have high data rate and spatial resolution without any
consideration of seeding particles. The developed FLDV (Fig. 1) was installed in the casing which fits
M14 instead of the conventional spark plug (Fig. 2). The measurements were carried out in a pratical
engine under motoring condition. The test engine is the production engine, no special treatment was
added. (Fig. 3)
The measurement results demonstrate that turbulent intensity near the spark plug changed before/after
TDC dramatically. The data rate was over 5kHz and cyclic variation of the engine operations were
demonstrated.
Laser beam
fr
Din
D
Developed FLDV
Da
Recieving-lens
Front -lens
f
M14
Casing
Optical window
Working distance
Measurement volume
Probe
df
dz
δf
Fig. 2 Developed LDV probe
and casing
Fig. 3 LDV probe and test engine
Fig. 1 Structure of the developed
FLDV probe
1 Introduction
It is well known that a high turbulence near spark plug can increase turbulent combustion speed, which
contributes to stable and lean engine combustion over a wide range of engine operation conditions.
Many measurements by LDV (e.g. Durst et al., 1976 and Obokata et al., 1987) and visualization (Melling,
1997 and Fujikawa et al., 1995) are required to understand turbulent characteristics near a spark plug in
both constant-volume combustion chambers (Morikawa et al., 1999) and model engines (Hamamoto et
al., 1993). Flame propagation characteristics have been discussed in research on visualization (Fujikawa
et al., 1995).
On the other hand, LIF can demonstrate the two-dimensional fuel concentration
distribution before ignition (Kakuhou et al., 1999 and Zhao et al., 1994). This is very useful information
in controlling mixture formation, although it is hard to understand the time-varying characteristics.
Direct turbulent measurement near the spark plug is necessary to understand flame propagation and its
relation to turbulence intensity for various engine speeds. The purpose of this present study is to develop
a fiber LDV system that can be used instead of spark plugs in a standard engine without any modification.
The first step is to demonstrate this measurement in a pratical engine and understand the flow behavior
and turbulence intensity measurement under non-combustion conditions.
2 Optics design of LDV
LDV optics should be installed in an M14 casing where there are no spark plug units. The optical
parameter is determined in order to measure at a very high data rate for cycle-resolved analysis. The
developed optical component is shown in Fig. 4. The measurement location is 5mm from the optical
window, which was determined to be the same location as the spark plug gap. The measurement volume
size and optical parameter were determined in consideration of band width of signal processor,
measurable dynamic range, the accuracy, the data rate, ease of cleaning, ease of setting up, pressure,
vibration, and so on. The fringe spacing was determined to be over 2µm for sufficient visibility and a
seeding particle diameter of 1-2µm (Menon et al., 1991 and Ikeda et al., 1992). For high data rate
measurement, the scattered light from the measurement volume can be increased with a smaller
measurement volume size, but the particle concentration and passing percentage becomes low. Then, the
measurement volume diameter was first set up around 50µm and the measurement volume depth was less
than 500µm. The output beam diameter from the collimate lens was then determined. We used W18
(NSG, selfoc lens: W18, 0.248P, AR coating) for 514.5nm.
Here, the SNR parameter (Durst, et al., 1976) was used to evaluate the LDV’s performance. We tried to
achieve more than 4.5 x 10-5, which is much higher than those commercially available (Ikeda et al., 1990)
and our previous experimental results (Ikeda et al., 1992 and Ohira et al., 1995).
The developed Fiber LDV is shown in Fig. 5. As shown, the size was for M14 and is easy to install in a
practical engine. Figure 6 shows the practical installation in the practical SI engine used in the
experiment.
All optical parameters were determined in these processes. The final dimensions of
the measurement volume are also shown in Fig. 4.
Optical parameter
Laser beam
Din
D
fr
Da
Recieving-lens
Front-lens
Optical window
f
Working distance
Measurement volume
Wave-length
•Fƒ É=514.5 nm
Focal length of front-lens
•
F f =31mm
Working distance
•FW =5 mm
Effective diameter of recieving-lens •
FDa =15.74mm
Space of laser beams
•FD =7.5mm
Diameter of laser beam
•FDin = 0.4037mm
Measurement volume dimension
•Fdf =50.3 µm
•Fdz =415.8 µm
Fringe•s
@pacing
•Fδ f =2.142 µm
Fringes number
•FNf =23.65
AOM
•F40MHz
SNR parameter =
df
dz
δf
(
Da •EDin
f
SNR parameter
TSI(9273)
KANOMAX(8835)
DANTEC(60x17)
2
2
)
= 45.24 x 10
-6
: 0.64 x 10 -6
: 11.11 x 10 -6
: 2.28 x 10 -6
Fig. 4 Optical parameter
Cap for casing
Probe
Casing
Rear pipe of LDV probe
Probe body
Casing
M14
Casing
O-ring
Catch for optical window
Probe
Optical window
Assembly of LDV probe
Developed LDV probe & casing
Fig. 5 LDV probe
Developed FLDV
Fig. 6 LDV probe & test engine
3 Experimental Apparatus
The engine used in this experiment was a practical 1000cc four-cylinder engine [bore x stroke = 71 x
80.5mm:NISSAN]. The measurement system is illustrated in Fig. 7. An Ar-ion laser and a Burst Digital
Correlator [BDC (Ikeda et al., 1990)] were used for the measurement. The test engine was operated in
normal motoring conditions, and the developed FLDV was installed into one of the cylinders instead of a
standard spark plug, as shown in Figs 6 and 7. The measurement location was at the center of the
cylinder shown.
The seeding particle used in the measurement was MSF (Nishigaki et al., 1992) and a constant feeding
system was used.
Manipulator
Beam
Separator
Topview
Ar-ion Laser
Frequency
Shifter
INT
INT
EXH
EXH
FLDV Plobe
V
Photomultiplier
Burst
Digital
Correlator
U
Measurement
Point
Encorder
Engine
5
Multi-channel A/D Converter
WS
FLDV Probe
ƒ 7
Ó1
Measurement
Point
LAN
Measurement system
Fig. 7
Measurement point
Measurement system & Measurement point
4 Results and discussion
The measurement results are shown in Fig. 8, which shows LDV data in three consecutive cycles. The
figure indicates two data rates calculated from the short period and wide period. A high data rate of over
5kHz can be measured in each cycle. The direction U means from intake to exhaust and V was
perpendicular to U. The results show that both velocity traces indicate large velocity fluctuations, but the
fluctuating ranges are similar. The fluctuating range near TDC is about –4m/s to 7m/s. Since this
developed FLDV was designed to measure turbulence intensity, it is found that the measured data can
provide sufficient turbulence information near TDC. It is understood that a constant and high level of
turbulence can increase turbulence combustion speed just after ignition, so that formation of constant
turbulence around ignition time and TDC should be clearly understood for optimum engine operation
control and reduction of UHC in exhaust gas. In this paper, we do not discuss the detailed flow
characteristics, because the purpose of this study is to develop a fiber LDV that can be installed in a
practical engine. Although the present measurement data can only provide the evaluation results of the
present system, it was found that the velocity fluctuation in the expansion stroke is much larger than the
compression stroke near the spark plug. A large reverse flow was found at BDC on the compression
stroke.
The ensemble average velocity was compared to one of an instantaneous cycle, as shown in Fig. 9. The
rms value is very small in this experiment. The window size used was 0.5°CA and the sample number
distribution indicated that the particles can travel the measurement volume sufficiently around TDC.
Top view
IN
V
IN
EX
Measurement
point
EX
Data rate
Motoring 1500r/min Throttle 100%
U
40
6.9kHz
6.6kHz
5.9kHz
7.5kHz
7.1kHz
5.6kHz
9.7kHz
10.1 kHz
10.6kHz
7.0kHz
8.9kHz
9.4kHz
Direction : U
Instantaneous Velocity[m/s]
20
0
-20
40
Direction : V
20
0
-20
TDC
I.V.O.
BDC
I.V.C.
TDC
BDC
E.V.O.
TDC
E.V.C.
I.V.O.
BDC
TDC
I.V.C.
BDC
E.V.O.
TDC
BDC
E.V.C.
I.V.O.
Crankangle
Fig. 8
Cycle-resolved LDV measurement
I.V.C.
TDC
BDC
E.V.O.
TDC
E.V.C.
Fig. 9
Comparison of ensemble averaged velocity to instantaneous velocity
5 Conclusions
We have developed a fiber LDV that can be installed in a practical engine instead of a standard spark
plug. The evaluation was done under motoring conditions. The design process and the results can be
proved to achieve high data rate measurement in practical engine conditions. The measurement results
show that the data rates near ignition timing and around TDC were sufficient to determine the turbulent
flow characteristics in real time and cycle-resolved measurement can be demonstrated.
References
1)
F. Durst, et al.,”Principles and Practice of Laser-Doppler Anemometry”, Academic Press, (1976)
2) ”Applications of Laser Techniques to Fluid Mechanics”, 5th Int. Symp. on Application of Laser
Techniques to Fluid Mechanics, (1990)
3) T. Obokata, et al., “Velocity and Turbulence Measurements in a Combustion Chamber of S.I. Engine
under Motored and Firing Operations by L. D. A. with Fiber-Optic Pick-Up”, SAE Paper No.870166,
(1987)
4) Alex Taylor, “Instrumentation for Flows with Combustion”, Academic Press, (1993)
5) A. Melling, “Tracer Particles and Seeding for Particle Image Velocimetry”, Measurement Science and
Technology, Vol.8, No.12, pp.1406-1416, (1997)
6) Fujikawa, T., Kataoka, M., Katsumi, N., and Kobayashi, T., JSME Transactions Vol.61, No.590B, 1995
(in Japanese).
7) H. Morikawa, et al., “Flame Behavior and Unburnt Gas Motion in Stratified Charge Conditions”, 15th
internal combustion engine symposium, 9935455, (1999)
8) Y. Hamamoto, et al., “Tumbling Flow in a Pentroof-Type Combustion Chamber of an Internal
Combustion Engine”, JSME Transactions Vol.59, No.559B, 1993 (in Japanese).
9) A.Kakuhou, T. Urushibara, et al., “Characteristics of Mixture Formation in a Direct Injection SI Engine
with Optimized Incylinder Swirl Air Motion”, SAE Paper 1999-01-0505
10) Zhao, F. Q., Taketomi, M., Nishida, K., and Hiroyasu, H., SAE Paper No. 940988, 1994.
11) R. Menon and W. T. Lai, “Key Considerations in the Selection of Seed Particles for LDV
Measurements”, Laser Anemometry Advanced and Applications, ASME, pp.719-730, (1991)
12) Y. Ikeda, et al., “Optimum Seeding Particles for Successful LDV Measurements”, 6th Int. Symp. on
Application of Laser Techniques to Fluid Mechanics, pp.32.2.1-322.5, (1992)
13) Yuji Ikeda, et al., “A Compact Fiber LDV with a Perforated Beam Expander, Measurement Science
and Technology, Vol.1, No.3, pp.260-264, (1990)
14) Yuji Ikeda, et al., “Flow Vector Measurement at the Scavenging Ports in a Fired Two-Stroke Engine”,
SAE Paper No.920420, 1992
15) Tetsuya Ohira, et al., “In-Cylinder Flow Measurement and its Application for Cyclic Variation
Analysis in a Two-Stroke Engine”, SAE Paper No.950224, 1995
16) Y. Ikeda, et al., “Performance of Prototype Hardware of Burst Digital Correlator”, 5th Int. Symp. on
Application of Laser Techniques to Fluid Mechanics, pp.25.1 (1990)
17) M. Nishigaki, et al., “New High-Performance Tracer Particles for Optical Gas Flow Diagnostics”,
Measurement Science and Technology, Vol.3, pp.619-621, (1992)