Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
-s105-
Micro-hole Inspection System Using Low-Frequency Sound
Yoshinori NAGASU
1, 2, Kakumasa EGUCHI
1, Kazunori ITOH
2, Makoto OTANI
2 and Noboru NAKAYAMA
2
1 Nagano Prefecture General Industrial Technology Center, Okaya 394-0084, Japan
2 Faculty of Engineering, Shinshu University, Nagano 380-8553, Japan
(Received 22 November 2013; received in revised form 14 April 2014; accepted 19 April 2014)
Abstract
This paper presents a method for inspecting micro-holes
with diameters of less than 100 µm by using audible sound.
Fluid flow through micro-holes affects the performance of
a product such as CO gas-detection sensor components,
microphone filters, and fuel-injection plates. Therefore, if
the flow rate of fluid passing through micro-holes could be
accurately assessed by means of an inspection system, it
would be possible to guarantee the performance of the
product.
The prototype inspection system used in this study
consists of an AC voltmeter, a function generator and a
cylindrical metal casing containing a small built-in
loudspeaker and a microphone. When a loudspeaker in a
sealed casing is driven by a low-frequency sine wave, an
excited sound vibrates the air inside the metal casing. The
developed system measures the change in pressure at the
microphone as the sound wave passes through a micro-
hole. It is therefore the same as measuring the flow rate of
a fluid passing through a micro-hole. This system is able to
detect a 5-µm diameter difference in a micro-hole.
As the flow rate through a micro-hole can be easily
measured, this inspection method is considered to be
applicable to micro-hole components intended for use in
flow control. Measurement time takes about 1 s after a
standard has been set, which is faster than existing
measurement methods.
Key words
Micro-hole, Diameter, Aspect Ratio, Low-frequency
Sound, Pressure
1. Introduction
Automotive fuel-injection plates, gas-detection sensors,
and high-performance microphones all have components
containing micro-holes that are used to control flow rate.
In order to guarantee their performance, the shape of these
micro-holes needs to be precisely realized in such
components. Scanning electron microscope (SEM)
photomicrographs of a micro-hole are shown in Fig. 1.
Micro-hole shapes can be inspected using conventional
inspection methods such as an optical microscope and
contact 3D scanning [1-3]. However, these methods cannot
accurately predict the flow rate of a fluid nor are they fast
enough for the 100%-inspection that is often required on a
line producing these components. Image-processing
technology can realize high-speed inspection of micro-
holes [4-6], however cannot inspect their shape. For
products that require flow assurance of a gas passing
through holes, such as automotive fuel-injection plates and
gas-detection sensors, the shape of the hole affects the
performance of the product. Therefore, a new inspection
method that is simpler and faster is required. A possible
solution is measurement using sound. For example,
acoustic impedance measurement utilizing audible-range
sound is used for volume and surface area measurements
of an object in a closed volume [7-9]. In this study, we
propose a fast, highly accurate inspection method using
audible sound for micro-holes that are mass-produced by
metal stamping. This paper presents details of the proposed
method and validates its performance.
2. Methodology Flow through micro-holes affects the performance of a
product. Therefore, if the flow rate of fluid passing through
micro-holes could be accurately assessed by means of an
inspection system, it would be possible to guarantee the
performance of the product.
Flow through a circular tube of radius r and length L is as
shown in Eq. (1) by the Hagen-Poiseuille equation. In this
equation, Q is the flow through micro-holes, ν is the
coefficient of viscosity, P1 is the pressure on the inlet side
of the circular tube, and P2 is the pressure on the outlet side.
Q = πr4
( P1 - P2 ) / 8νL (1)
AC voltmeter
cylindrical metal casing
(built-in loudspeaker and a microphone)
Function
generator
Fig. 2 Prototype inspection system Fig. 1 Micro-hole stamped in metal
Journal of JSEM, Vol.14, Special Issue (2014) s105-s109Copyright Ⓒ 2014 JSEM
-s106-
Flow through a circular tube as determined by Eq. (1) is
seen to change according to hole radius r, hole depth L and
pressure gradient P1 - P2. In the prototype system, a
microphone is used to detect the change in pressure of an
sound wave as it passes through a hole. The external
appearance of a prototype inspection system is shown in
Fig. 2, and a schematic is shown in Fig. 3.
The inspection system consists of an AC voltmeter (VT-
181, TEXIO Technology Co.), a function generator
(DF1906, NF Co.) and a cylindrical metal casing with a
small built-in loudspeaker (SW070WA02-01, Wavecor
Ltd.) and a microphone (ATH-C500M, Audio Technica
Co.). A micro-hole component is placed on top of the
metal casing as shown in Fig. 4. The outer diameter of
the test container is 120 mm and the height is 150 mm.
The volume of air between the microphone and the
speaker is about 48 mL, and that between the
microphone and the micro-hole is about 0.5 mL.
When a loudspeaker in a sealed casing is driven by a
low-frequency sine wave, an excited sound vibrates the
air inside the metal casing. If the vibration of the speaker
is large, the pressure change inside the metal casing
increases. The microphone detects the pressure change
inside the metal casing. Here, the detected pressure
varies according to the volume of a micro-hole. Thus,
micro-holes can be inspected by analyzing the
differences between the measured values. This system
measures the change in pressure at the microphone when a
sound wave passes through a hole. This is effectively the
same as measuring the flow rate of a fluid passing through
the micro-hole.
3. Validation of the Prototype Inspection System
The prototype inspection system was validated using
micro-hole test samples with micro-holes of various
diameters and depths. By inspecting samples with different
hole diameters however uniform thickness, it is possible to
evaluate whether a difference in the diameter of a hole can
be detected. Likewise, by inspecting samples with a
different thickness however constant hole diameter, it is
possible to evaluate whether a difference in the depth of a
hole can be detected. The appropriate driving voltage and
output frequency of the speaker required to output the
audio signal were experimentally determined and found to
be 5.0 to 10 Vp-p and 10 to 90 Hz, respectively.
3.1 Different diameter samples
The test samples had a thickness of 0.1 mm and the micro-
hole diameter ranged from 50 to 110 µm. Table 1 shows
the front, back, average diameters and hole volumes of the
test samples. There are two basic ways of producing a
hole: electrical discharge machining (EDM) and metal
stamping. The micro-hole diameter was measured using a
non-contact type coordinate measuring machine (NH3-SP,
Mitaka Kohki Co.). The diameter was measured by
loud-
speaker
microphone
micro-hole
sample
AC voltmeter
function
generator
Fig. 3 Inspection system schematic
measures the pressure
change inside the metal
casing
flow rate passing through
the micro-hole
sound vibrates
Fig. 4 Detail plan around microphone
100μm
Front side
Φ=104.8μm
Backside
Φ=100.6μm
Processed by EDM
Fig. 5 Micro-hole formed by EDM
Front side
Φ=99.8μm
100μm
Backside
Φ=104.8μm
Stamped hole
Fig. 6 Micro-hole formed by metal stamping
Table 1 Different diameter samples
Front side
Diameter [µm]
Backside
Diameter [µm]
Average
Diameter
[µm]
Hole
Volume
×10-3
[mm3
]
Metal
Stamping
99.8 104.8 102.3 0.82
Electric
Discharge
Machining
112.4 104.2 108.3 0.92
104.8 100.6 102.7 0.83
101.2 96.0 98.6 0.76
97.6 90.0 93.8 0.69
94.2 88.4 91.3 0.65
79.0 74.2 76.6 0.46
67.8 72.4 70.1 0.39
60.6 56.2 58.4 0.27
48.2 45.6 46.9 0.17
Y. NAGASU, K. EGUCHI, K. ITOH, M. OTANI and N. NAKAYAMA
-s107-
approximating the micro-hole shape to a circle by applying
least square fitting.
The shape of the micro-hole differs depending on the
hole-formation method used. In the case of EDM, the
diameter on the front side is approximately 5 µm larger
than that on the backside (Fig. 5). This is due to stagnant
powder around the EDM electrode. In the metal-stamping
method, the diameter of a micro-hole on the backside is 5
µm larger than that on the front side (Fig. 6). Stamped
holes have a four-layer structure that includes a shear drop,
sheared surface, fractured surface, and burring. The micro-
hole diameter on the backside increases due to the effects
of burring. The cross-sectional shape is not straight either,
so the average of diameters on the front side and backside
was chosen as the representative micro-hole diameter of
the test samples.
3.2 Different thickness samples
The test samples have a diameter of 0.3 mm and the
thickness ranges from 0.3 to 1.9 mm. The aspect ratio of
hole depth and hole diameter is from 1.1 to 6.3. The micro-
holes were drilled by a compound jig borer. Table 2 shows
the thickness, diameter of the front side, diameter of the
backside and aspect ratio of the test samples. Figure 7
shows an image of the front side and backside of the holes
drilled. There is a maximum difference in hole diameter of
about 30 µm for the front side and the backside.
4. Results
Figure 8 shows the effect of the loudspeaker frequency
on the measured effective voltage for different hole
volumes. Here, the input voltage was 10 Vp-p. The average
hole diameters are indicated on the curves. It can be seen
that the measured voltage increased with frequency.
However, the lowest frequency of 10 Hz yielded the
largest dependence of the voltage on the hole volume.
Figure 9 shows the relation between input voltage to
loudspeaker and the measured effective voltages. The input
voltage values were 10, 7.5, and 5.0 Vp-p 5.0 Vp-p and the
frequency was set to 10 Hz. It can be seen that the
measured effective voltage changed in proportion to the
Table 2 Different thickness sample
Thickness
[mm]
Front side
Diameter
[µm]
Backside
Diameter
[µm]
Average
Diameter
[µm]
Aspect
Ratio
0.34 299 308 303 1.1
0.66 315 299 306 2.2
0.91 296 309 302 3.0
1.27 297 308 302 4.2
1.56 301 310 305 5.1
1.88 325 296 311 6.0
Table 3 Difference in measured voltage
Input
Voltage
[Vp-p]
Measured Voltage
at φ108.3µm
[mV]
Measured Voltage
at φ46.9µm
[mV]
Difference
[mV]
5.0 0.24 0.15 0.09
7.5 0.31 0.19 0.12
10 0.38 0.22 0.16
Φ=47μm
5870
77
91
94
99
103
108
Fig. 10 Measured voltages for various micro-
hole diameters
Φ=47μm58 70 77 91 94 99 103 108
Fig. 8 Effects of speaker frequency
Φ=47μm5870
77
9194
99103
108
Fig. 9 Effects of input voltage
Front side
Φ=297μm
Backside
Φ=308μm
100μm
Fig. 7 Drilled micro-hole
Journal of JSEM, Vol.14, Special Issue (2014)
-s108-
input voltages. However, the difference in the measured
voltage between micro-holes of 47 µm and 108 µm in
diameter is maximal for 10 Vp-p input voltage (Table 3).
This result shows that the inspection resolution can be
improved by increasing the input voltage to the
loudspeaker.
Figure 10 shows the dependence of the effective voltage
on the micro-hole volume for an input voltage of 10 Vp-p
and a frequency of 10 Hz. The average hole diameters are
indicated on the curve. The measured voltage clearly
increases with micro-hole diameter; a 5 µm difference
yields a 20 µV difference in voltage, although the change
in voltage becomes smaller as the diameter decreases to
below 80 µm or so. When three measurements per micro-
hole were made, the difference between the measured
voltage was found to be less than ±2 µV.
The two test samples have the same average micro-hole
diameter (approximately 102-103µm). Each was processed
by the metal-stamping method and the EDM method. The
two samples have different front side and backside
diameters, however neither show a significant difference in
measured effective voltage.
Figure 11 shows the relation between the aspect ratio of
a micro-hole and the measured effective voltages. The
input voltage is set to 10 Vp-p and the frequency is set to 10
Hz. The thicknesses are presented along the curve. The
results demonstrate that the measured voltage decreases as
the micro-hole depth increases; a 0.3-mm difference in
micro-hole depth yields a 0.2-mV difference in voltage in
this case. The influence of micro-hole diameters that are different
on the front side and the backside was then examined. The
same measurements were performed with the same test
samples installed the other way up. The results showed no
significant difference in measured voltage between the
original and inverted installations, indicating that the
proposed method is capable of guaranteeing the flow rate
through a micro-hole. However, if the difference in
diameters between the front side and backside is larger, the
effect on measured voltage is unknown.
5. Discussion
In Eq. 1, the flow rate passing through the hole was
shown to vary due to the pressure difference between the
inlet and outlet, hole depth and hole diameter. The results
in Fig. 9 show that the flow rate passing through the holes
is increased by a high input voltage. This is believed to be
because the difference between the atmospheric pressure is
increased because the pressure in the cylindrical casing is
higher. This makes it possible to detect a smaller hole
diameter difference due to the measured voltages being
increased in the case of high flow.
It was also found, according to the results in Fig. 8, Fig.
10 and Fig. 11, that the measurement resolution is
increased at a low frequency. This is thought to be because
lower-frequency sound waves can more easily pass
through a micro-hole due to diffraction. Therefore, in order
to improve the measurement resolution, it is desirable to
set a higher input voltage and a lower frequency.
The experimental results show that the proposed method
is capable of detecting a difference in micro-hole diameter
from the measured voltages. Specifically, for 100-µm
diameter micro-holes having a thickness of 100-µm, which
were the parameters adopted in this study, the proposed
method is able to detect a 5-µm difference in micro-hole
diameter.
The results in Fig. 10 show that the measured voltage is
correlated with the volume of a micro-hole, not the
diameter or cross-sectional shape. As the micro-hole
volume is proportional to the flow rate through the holes,
the proposed method is suitable for inspection of flow
control parts. The results in Fig. 11 show that differences in hole depth
are also detectable, although the resolution is only about
0.1 mm, which is significantly lower than that for the hole
diameter.
The time required to measure the flow rate through a
micro-hole is about 1 s, after a micro-hole components has
been set which is faster than existing measurement
methods.
The coefficient of the viscosity of a fluid is proportional
to temperature, however, the influence of this on the
measurements in this study were not confirmed. It is
believed that in order to obtain a stable measurement result,
it is necessary to keep the measurement environment as
constant as possible.
Further study is needed with regard to the shape of the
metal casing. There is a possibility that the accuracy of the
measurement can be improved by optimizing the volume
of the container. In particular, the detection performance of
the microphone might be improved by reducing the
volume between the micro-holes and the microphone. In
addition, by selecting speakers and a microphone that
better match low frequencies, the stability of the
measurement results may be increased.
6. Conclusion
This paper proposed a method for inspecting micro-
holes with diameters of less than 100 µm by using audible
sound. This system is able to detect a 5-µm diameter
difference in a micro-hole.
The experimental results indicate that a flow rate
through a micro-hole correlates with a voltage measured
by a microphone installed in the inspection system. The
t=0.3mm
0.7
0.9
1.3
1.6
1.9
Fig. 11 Measured voltages for micro-hole aspect ratio
Y. NAGASU, K. EGUCHI, K. ITOH, M. OTANI and N. NAKAYAMA
-s109-
flow rate passing through the holes is increased by a high
input voltage. The measurement resolution is increased at a
low frequency. Thus, it is expected that the proposed
method can be used as an effective tool for determining the
diameter of micro-holes.
Acknowledgement
This work was supported by A-STEP FS-Stage for
Exploratory Research, Grant Number AS231Z02722B, by
the Japan Science and Technology Agency. We are
indebted for the guidance we received from Hikaru
Yamagishi, Ryouichi Arai, Kazuyuki Kamijo and Takeo
Chigono on how to operate the measuring and processing
equipment used in this study. The advice and comments
given by Seiichi Kudo, Eimatsu Sakagami and Hideaki
Miyasaka were a great help in my reseach. We also
received generous support from Micro-Technology
Division of Misuzu Industries Co. in the form of the
provision of the measurement samples.
References
[1] Youkaichiya, M., Funabashi, T. and Takaya, Y.: The
high precision measurement of the minute shape by the
UA3P series, Proc. Autumn Meet. Jpn. Soc. Precision
Eng. (2011), 878-879.
[2] Murakami, H., Katsuki, A., Onikura, H., Sajima, T. and
Kondo, E.: Development of a system for 3-D micro
metrology using an optical fiber probe, Proc. Spring
Meet. Jpn. Soc. Precision Eng. (2011), 811-812.
[3] Murakami, H., Katsuki, A., Onikura, H., Sajima, T.,
Kawagoishi, N. and Kondo, E.: Development of a
system for measuring micro hole accuracy using an
optical fiber probe, J. Advanced Mechanical Design,
Systems, and Manufacturing 4 (2010), 995-1004.
[4] Hata, S.: Trends and Future of Industrial Application of
Image Processing Technologies, J. Soc. Precision
Engineering, 67-6 (2001), 879-882.
[5] Hata, S.: Trends of Automatic Visual Inspection, J. Soc.
Precision Engineering., 56-8 (1990), 1366-1370.
[6] Okamoto, S. and Yoshimura, K.: Automatic Visual
Inspection of Soldered Parts on Printed Circuit Boards, J.
Welding. Soc., 58-4 (1989), 300-304.
[7] Torigoe, I.: Measurement using audio-frequency sound,
J. Acoust. Soc. Jpn., 61-11 (2005), 653-658.
[8] Ishii, Y.: An acoustic volumeter, J. Instrument and
Control Eng., 36-4 (1997), 288-291.
[9] Komiya, K.: Measurement of volume of old weights
using an acoustic volumeter, Bull. Soc. Historical
Metrology Jpn., 30-1 (2008), 55-58.
Journal of JSEM, Vol.14, Special Issue (2014)