Fundamentals and Practices of Sensing Technologies
Keiji Taniguchi, Hon. Professor of
Dr. Masahiro Ueda, Honorary Professor, Faculty of Education and Regional Studies
Dr. Ningfeng Zeng, an Engineer of Sysmex Corporation
(A Global Medical
Dr. Kazuhiko Ishikawa, Assistant Professor
Faculty of Education and
[Editor’s Note: This paper is presented as Part XIV of a series from the new book “Fundamentals and Practices of Sensing Technologies”]
Chapter Six – Part II
6.4 Charge-to-Mass Ratio Sensor for Toner Particles
The electric charge (q) to the mass (m) ratio q/m of toner particles plays an important role for developing the quality of printers or copy-machine systems. For this reason, the small size of apparatus measuring q/m is desired for on- line use.
As already reported in the literature (10) (11), we have developed a transparent-electrode method for observing the movement of a toner particle using an optical microscope for off-line use.
Furthermore, as a new method (12), we present here the method for measuring Σ(q/m) at the end of the flow path in a toner transport system driven by a traveling wave for on-line use.
6.4.2 Principle and Method
A. Principle (13)
Figure 6.15 illustrates the relationship between toner particles above a sheet of the printed periodic array-conductors driven by the four phase rectangular pulses and parallel electrodes for sensing charged toner particles.
Let us consider the situation that a toner particle is moved toward parallel (sensing) electrodes placed at the end of left side of the transportation system.
In the air gap between parallel electrodes shown in Fig.6.16, the following two forces act on the toner particle. The image force is neglected here.
(1) The Coulomb force due to the electric field caused by the power source is given
as follows: (6.7)
Where, , and are the charge of a toner particle, the voltage applied across parallel electrodes and the air gap between parallel electrodes, respectively.
(2) The gravity force is given as follows: (6.8)
Where, and are the mass of the toner particle and the acceleration of gravity, respectively.
From the relationships between and , there are three major cases as shown in Fig.6.16.
Case 1: (6.9)
In this case, the charged toner particle moves from the air gap between parallel electrodes toward the lower electrode.
Case 2: (6.10)
In this case, the charged toner particle passes through the air gap between parallel electrodes. From Eq.(6.10), q/m can be expressed in the following form:
Case 3: (6.12)
In this case, the charged toner particle moves from the air gap between parallel electrodes toward the upper electrode. First of all, let us consider the case when a charged toner particle moves from the air gap between parallel electrodes toward the upper electrode. As a result, the signal induced in parallel electrodes by the charged toner particle may produce a voltage across the resister R and this signal is amplified by using the differential amplifier as shown in Fig. 6.17.
From , we can obtain the charge of the toner particle (See Fig.6.17):
If is satisfied, the following relation is obtained from Equations (6.11) and (6.13):
The model of the signal sensed due to a toner particle is shown in Fig.6.17 (a).
Secondly, let us consider the case when massive charged toner particles move from the air gap between parallel electrodes toward the upper electrode.
As shown in Fig.6.17 (b), is expressed in the following form by summing up :
From Eqs.(6.14) and (6.15),we can obtain the following relations
Where,, and ,, and are the number of charged toner particles, time delay of charged toner particles, and total mass of charged toner particles, respectively.
Thirdly, furthermore, let us consider in the case where the voltage applied across parallel electrodes is increased to the magnitude of , where is the small increase of . In this case, we can express Σ(q/m) by the following form:
From Eqs. (6.16) and (6.17), the difference ofΣ(q/m) is calculated as follows:
B. Experimental System
The schematic diagram of the experimental system for sensing Σ(q/m) is shown in Fig.6.18. This system consists of the following parts: the four-phase rectangular pulse generator, the sheet printed periodic array-conductors for transporting charged toner particles, the DC power source for supplying the voltage to parallel electrodes and the personal computer for analyzing the output signal.
The air gap between parallel electrodes is about 1000mm. By shifting the electric curtain generated above the periodic array-conductors, the charged toner particles are moved toward the air gap between the parallel electrodes placed as a sensor.
The signal picked up by the parallel electrodes is amplified using the differential amplifier. The amplified signal is sent to the analog input of the A-D conversion board in the personal computer as shown in Fig.6.18.
Figures6.19 shows the structure of the sheet of periodical-array-conductors for the transportation of toner particles. The periodical-array-conductors are connected to the four-phase rectangular pulse generator.
Figure6.20 shows typical examples of toner particles transported by electric fields due to the four-phase rectangular pulses in different times.
Spherical toner particles are used for this experiment. The values of voltage and frequency generated by four-phase rectangular pulse generator are 100(V) and 100(Hz), respectively. Regarding Σ(q/m) of charged toner particles, we assume here that some samples in the flow of transportation may have uniformly distribution.
Figure 6.21 shows the output signal of the sense amplifier in the case when the voltage applied across the sense electrodes is 1.9 (V). In this figure, the induced
noise pulses caused by four- phase rectangular pulses are only seen in the output of the amplifier. These are reduced at the last stage of the sensing circuit using a low pass filter.
Figure 6.22 shows the relationship between voltage applied across parallel electrodes and the normalized values of Σ(q/m) for charged toner particles obtained by applying Eq. (6.16). From Fig.6.22, the distribution of Σ(q/m) for the charged toner particles is calculated by the following equation:
Figure 6.23 shows the relationship between Σ(q/m) and the frequency distribution. From this figure, the distribution ofΣ(q/m) for charged toner particles of this method has a good coincidence with that measured by E-Spart method near the peak.
6.4.4 Improvement of Toner Transportation system (14)
Vibrating the printed sheet of periodical-array-conductors set up on the ultrasonic vibrator is very effective for decreasing the amplitude of four phase rectangular pulses.
The connection between the sheet of array conductors and the ultrasonic vibrator is executed by the use of a sheet of either a glass plate or an expanded polystyrene plate as an interface.
Table 6.8 shows the relationship between the amplitude of four-phase rectangular pulses and the vibration effects.
From this table, the expanded polystyrene plate put between the sheet of array conductors and the ultrasonic vibrator shows better transfer effect than the glass one.
The sheet of these conductors driven by the four-phase rectangular pulses is swung to the vertical direction by the ceramic vibrator driven by the sinusoidal wave generator. Consequently, for the toner particles are floated over the sheet of printed conductors, the adhesive forces acting between toner particles and the sheet of them are greatly decreased. As a result, the toner particles can be easily transported to the direction of the other end of the sheet with the low level amplitude of four-phase rectangular pulse voltage. From the result of this experiment, the best operational condition is that the drive frequency of the ceramic vibrator (Produced by Murata Manufacturing Co. Ltd) is 19KHz and the amplitude of the four-phase rectangular pulse voltage in this situation is 18 volt.
6.5 Active Sensor for Bearing Wear
A power transmission shaft is usually provided with a bearing in order to limit vibrations which are a direct consequence of revolution. As such, the bearing is one of the most important components of a power transmission shaft. Many studies and publications concerning oil contamination caused by bearing wear have been developed15),16) since too much bearing wear contributes to the risk of failure. We have developed a simple method for estimating bearing wear by investigating oil contamination using colorimetric analysis17),18). The method is not, however, direct estimation of the bearing wear. In the past, bearing wear could be measured directly only by overhauling the pump system, which is both laborious and costly.
Another direct detection system for bearing wear without overhauling the pump, in other words, a system for monitoring vibration in real-time, is therefore highly desirable for many applications such as in power plants and manufacturing19). These sensor systems would not only prevent serious failure from occurring, but can also predict the life span of a bearing. Recently, some sensors to detect vibration in a transmission shaft and finally to diagnose bearing wear have been developed by means of an eddy current displacement sensor20),21), including ours22),23),24), and put into practical use. These electromagnetic methods are usually superior to the optical method in regard to a drain pump because the measurement accuracy of this method is not affected by water contamination, which usually has a great influence on the accuracy of the optical method. The electromagnetic method can, however, be used only for a metal shaft because eddy currents are induced only in electrical conducting materials. Furthermore, this method can measure bearing wear only indirectly; it shows Lissajous figures for a vibration of the shaft by means of two sets of displacement sensors mounted close to the shaft at right angles to each other.
We have proposed a new sensor which can detect bearing wear itself24) and can be applied for non-metal shafts as well as metal shafts. In this sense, we call it an “active sensor". This method provides a simple improvement on an eddy current displacement sensor.
In this section, we discuss the characteristics of this method when it was applied practically and show that this sensor can be practically applied for non-metal shafts as well as to metal shafts.
6.5.2 Principles and Methods
A. Primary Use of the Sensor
We have previously developed a system for monitoring the vibration in transmission shafts using an eddy current displacement sensor22),23). This method was an indirect method as shown below. Here we briefly summarize the principle and the method of this system, details of which are given in section 6.5.2.B. In metal, eddy currents are induced by a time varying electromagnetic field, and decrease the output of the sensor if the field is sufficiently high in frequency20),21). Thus in a typical eddy current displacement sensor, the sensor output diminishes as the target metal approaches the sensor. Further, the output, P', is linearly proportional to the distance, d, between the sensor and the metal for a certain range in distance,
where k1 is a sensor sensitivity and is a constant determined by the target metal, the rotor shaft in this case. The metal size is implied to be sufficiently large in comparison to the sensor size. The eddy current displacement sensor has usually been used under this condition, and has been used exclusively as a displacement sensor, as the name suggests. The sensor can thus be used as a vibration detector if the data acquisition rate of the sensor is high enough in comparison to the vibration frequency. As an example, in our previous study22),23) two sensors were mounted close to the metal shaft at a right angle to each other in order to observe the two-dimensional motion of the shaft, i.e., the trajectory of the center of revolution, as shown in Fig. 6.24. An increase in bearing wear can, therefore, be expressed indirectly by an increase in the area of trajectory because the shaft moves in all regions bounded by the bearing.
Fig. 6.24 Principle for detecting rotor vibration by means of an eddy current displacement sensor.
B. Principle of the “Active Sensor"
However, strictly speaking, the sensor output, P, will be affected by the effective size of a target metal, S, across which magnetic flux cuts, as is shown in Fig. 6.25. The sensor output
Fig. 6.25 Electromagnetic flux: (a) in the case of a small target metal, (b) in the case of a shifted metal.
in this case, P, should be proportional to the size of the target metal S as well as the distance
of the sensor ,d, and Eq. (6.20) can be written as follows:
where S0 expresses a region across which all magnetic flux cuts, and k2 is a constant chosen so that the function f(S/S0) is 0f(S/S0)1; the output P is normalized. A positive sign in front of the function f indicates that an eddy current decreases the sensor output as described above. We roughly assumed P=kdS in the previous paper22). However, the above Eq. (6.21) will be well fitted to discuss the characteristics of this new method, i.e., an active sensor, as shown below.
When the size of the effective target metal, S, is greater than three times that of the sensor size, it becomes SS0, in which all the magnetic flux cuts across the metal and produces a maximum eddy current at that distance. This, in turn, produces the sensor output minimum. Thus, the function f(S/S0)(= f(1) ) becomes 0, which results in Eq. (6.20). However, f(S/S0) increases as the size of a target metal diminishes, as is shown in Figure 6.25(a), and finally it becomes 1, i.e., f(0)=1. Such change in the metal size can be simulated by a horizontal shift of the target metal, as is shown in Figure 6.25(b). This indicates that metal size can also be measured even if the distance between the sensor and the target metal, d, is kept constant. That is, it is expected that a decrease of metal size due to bearing wear can be measured directly by means of this eddy current displacement sensor. This is the underlying principle of this method.
An ingenious scheme is necessary in order to realize this idea for practical use. Figure 6.26 provides a conceptual diagram of the method. A thin aluminum plate or a foil with a
Fig. 6.26 Conceptual diagram of the equivalence between bearing wear and a target metal.
(a) Shift of aluminum plate, (b) bearing wear.
width of W was pasted onto a wedge-shaped acrylic plate having an angle of , which substituted for a target metal. Bearing wear is simulated by a shift of the acrylic plate away from the metal shaft in the direction, i.e., from position to position in this case, as shown in Fig. 6.26(a). In practice, the aluminum foil is shaved off by bearing wear as shown in Fig. 6.26(b); the bearing wear can be measured by means of this eddy current displacement sensor fixed on a bearing. This method can thus measure the bearing wear itself and be applied for non-metal shafts as well as metal shafts; in this sense, we call it an “active sensor".
In this case, the output of the sensor is affected by both the distance, d, and the size of the target metal, S, and Eq. (6.21) is reformed as follows:
where dmax is the maximum value between them, and k3 is a constant so determined that 0g(dmax, S/S0)}1 for the normalization of g as mentioned above. The effective length, L, of the target metal, i.e., effective metal size, S, decreases linearly with a decrease of dmax, since they are related by S=Wdmax/sin, where W is a width perpendicular to this surface, i.e., width in depth. Eq. (6.22a) can then be expressed concisely by making these factors, d and S, into one factor, D-dmax as follows;
where D is the initial maximum distance between an aluminum plate and a sensor, as is shown in Fig. 6.26(b). Thus, D-dmax, expressing the amount of bearing wear, can be estimated by the sensor output, P. The function, g(D-dmax), should be pre-obtained for the practical application of this method, and it can be used as a calibration curve for the bearing wear. Eq. (6.22b) expresses a basis for this active sensor.
In practice, the shaft vibrates at approximately the revolution frequency of the shaft as was shown previousl23); as such, it approaches the sensor and then moves away from the sensor periodically. The close proximity of the shaft to the sensor will decrease the sensor output. However, an alternative component caused by this vibration of the metal shaft was not taken into in Eq. (6.22b); that is, the output in Eq. (6.22b) has only direct value. This is reasonable in an early phase of the bearing wear because the greater part of the electromagnetic flux reaches the aluminum plate, and the metal shaft has no effect on the output. However, when the bearing was greatly worn, some part of the electromagnetic flux reaches the metal shaft. Thus the output P in Eq. (6.22b) decreases abruptly due to the approach of the metal shaft to the sensor and increases again abruptly up to the direct value expressed in Eq. (6.22b), due to increasing distance of the shaft from the sensor. Then the output P becomes periodic with this revolution frequency, as mentioned above in this section and further shown later (see Figs. 32 and 33 in section 6.5.3). This alternative component with this frequency is, however, too high to be cut out from the direct component by means of low-pass filter, as shown in the previous study23). In contrast to this alternative component, the direct component expressed by Eq. (6.22b) increases gradually because the electromagnetic flux reaching the aluminum plate decreases due to the decrease of the metal size.
The data processing procedures are as follows. The output of the eddy current displacement sensor was amplified to a proper level for digital data processing. Secondly, the amplified signals were digitized by means of a 14-bit A/D converter. Finally, the digitized signals were fed into a personal computer and were normalized and averaged. The computer has other functions such as a motor drive and data acquisition. The data acquisition rate, i.e., maximum sampling frequency, of this system was 20kHz. However, the sampling frequency was 1 Hz in the experiments in this study, since the shift velocity of the acrylic plate was as small as 50ms-1, and the wear rate in this experiment was below 1 nms-1. Further, the bearing wear in a practical plant will be far slower than this value.
C. Output and Processing System
Figure 6.27 shows the measuring system. The principle of the eddy current displacement sensor is based on a high frequency electro-magnetic field. The eddy currents are induced within a metal placed in this electro-magnetic field, which changes the oscillation condition, i.e., amplitude and phase, in the electronic circuit of the sensor. The amplitude decreases as the sensor approaches a metal, which determines the output of the sensor. The output is roughly proportional to the distance between them as shown in Eq. (6.20).
The output from the sensor is amplified to a proper level for digitization by A/D converter. These digitized signals are led to a personal computer, processed, and the results are monitored on a display as shown in Fig. 6.27(b).
Fig. 6.27 Measuring system. (a) Electronic part of the eddy current displacement sensor and
(b) data processing system.
6.5.3 Results and Discussion25,26)
First, to assure Eq. (6.21), that is, explore a possible practical application of this method, the sensor output was measured at intervals of 100m horizontal shifts of the aluminum by maintaining a constant distance between the sensor and the target metal. This simulates a decrease of effective target size, S, in Fig. 6.26 and Eq. 6.22. Figure 6.28 shows the results. They resembled each other in shape except regarding the strength of the absolute output in the area of the usable operating range. Figure 6.28(b) expresses a normalized sensor output, i.e., the function f(S/S0) in Eq. 6.22. In this figure, a slight shift of the curve upward (or leftward) as an increase in distance can be explained as follows. When the displacement sensor is used at a greater distance from the aluminum plate (e.g. at d=3.0 mm rather than d=1.0 mm) the spatial resolution of the sensor deteriorates, causing the edge of the plate to be detected further from the edge. This deterioration is plainly evident in the curves in Fig. 6.28(a) and, when normalized, is likely to account for the upward (or leftward) shift in Fig. 6.28(b).
Fig. 6.28 Relation between the sensor output and the amount of horizontal shift of the acrylic plate (a). Normalization of this output (b).
In this experiment, a sensor size of =10 mm was used, which determines the extent of the measurable distance, i.e., a working distance. The working distance is usually half the sensor size, and was about 5 mm for this sensor. This value is sufficient since the amount of bearing wear to be measured in a manufacturing plant is approximately 2 mm.
Second, in order to test the system in terms of its practical use, an aluminum foil pasted on a wedge-shaped acrylic plate having an angle of was worn away with a grinding wheel. Figure 6.29 shows the experimental setup. An eddy current displacement sensor was fixed on
a wedge-shaped acrylic plate, and this united body was pressed against a grinding wheel at a constant pressure, which simulates practical bearing wear. Three wedge-shaped plates with =18.6, 15.6, and 12.6were used to examine the sensitivity for bearing wear, i.e., the ratio of output increment to wear increment.
Fig. 6.29 Experimental setup for simulated bearing wear, i.e., for a non-metal shaft.
Figure 6.30 shows the result obtained by means of this setup, which is a function of g(D-dmax) in Eq. (6.22) for a non-metal shaft; this expresses a correction curve for the bearing wear. The wear amount, D-dmax, was measured by means of a laser displacement sensor with a precision of 1m. The extent of the measurement was rather small in this case, i.e., 0~1.7 mm, in that the extent of measurement in a practical context is 2mm at maximum and excluding corrosion. All three curves have similar characteristics except for a small change in sensitivity, and further resembled those in Fig. 6.28. Similar experiments were performed using wedge-shaped plates at angles above 20. They all share similar characteristics. This can be predicted from the result shown in Figure 6.28(b) showing that the shape of the normalized output was hardly affected by the distance, d, but by the effective metal size, S. That is, an increase in output depends largely on a decrease in the effective target size, S, due to bearing wear because the angle of the wedge-shaped plate is rather small; a small change in distance results in a large decrease in effective target size. The angle should, rather, be determined by such factors as sensor size, i.e., the working distance of the sensor, and the amount of bearing wear to be measured. The relation between sensor output and wear amount shown in Fig. 6.30 expresses bearing wear itself for the non-metal shaft.
Fig. 6.30 Relation between the sensor output and the amount of bearing wear obtained by the
experimental setup shown in Figures 6 and 8.
Finally, the main experiments in this section, which simulate approximately a practical use for the metal shaft, were carried out by means of the experimental setup shown in Figure 6.31. In order to simulate a practical metal shaft, a circular stainless steel disk of thickness
Fig. 6.31 Experimental setup for practical bearing wear, i.e., for a metal shaft.
20 mm and diameter 100 mm was fixed to one end of the motor shaft of a drilling machine, and was immersed in a brass cylinder filled with water. The cylinder has an inner diameter of about 101 mm, thickness of 15 mm and a height of 200 mm, which simulates the bearing. The “active sensor" developed in this study was embedded into this vessel wall. A rough side-face of the disk was used to increase bearing wear. The bearing wear was induced by the vibration of the disk, i.e., run-out of the disc center to the revolution center, due to centrifugal force and a slight eccentricity of the disk to the revolution center, as discussed in the previous paper22).
Figure 6.32 expresses the result showing the relation between sensor output and bearing wear, which simulates practical use of the sensor. In this figure, curve (a) was obtained for a
Fig. 6.32 Relation between the sensor output and the amount of bearing wear obtained by the
experimental setup shown in Figure 6.27. Curve (a) was obtained for a grinding
wheel, i.e., a non-metal shaft, (b) for a metal shaft distant from the sensor within a
vibration amplitude, (c) for a metal shaft on the revolution center, and (d) for a
metal shaft in close proximity to the sensor.
grinding wheel, i.e., for a non-metal shaft, (b) for a metal shaft distant from the sensor within a vibration amplitude, (c) for a metal shaft on the revolution center and (d) for a metal shaft in close proximity to the sensor. As discussed in 6.5.2B, the output, (b), for the metal shaft is always lower than that, (a), for the non-metal shaft, even when the metal shaft is distant from the sensor. This decrease depends on the construction of the “active sensor" in relation to the bearing, i.e., the vibration amplitude of the metal shaft. The output, (d), is lowest of all and is also dependent on the construction, because a great part of the magnetic flux which reduces the output reaches the metal shaft. Therefore a practical output signal will be sinusoidal with a peak-to-peak amplitude between the curves (b) and (d) and with a revolution frequency, as predicted in this figure, s1~s4.
Figure 6.33 shows some examples of this practical signal on the display. The center of this output lies on the curve (c) in figure 6.32. The result was obtained in a period of about a
Fig. 6.33 Real time signal expressing bearing wear: (a) after 12 h, (b) after 24 h, (c) after 3
days, (d) after 1 week, (e) after 2 weeks and (f) after 1 month. (The sampling
frequency is 1 Hz)
month of continuous operation because the bearing wear was extremely small in water, about a few tenths that in air. In water, rolling friction between shaft and bearing decreases greatly since water acts as lubricant, which decreases rolling friction and heating due to this friction. The sinusoidal signal can easily be rectified by a low-pass filter. These results shown in Figs .632 and 6.33 can thus be used to estimate the bearing wear in a practical installation.
The measurement error of this sensor can be estimated to be about 20 m based on a small ripple in Figure 6.30. In this experiment, the amount of bearing wear was measured by means of a laser displacement sensor with a measurement error of 1 m. The error due to the electronic circuit including an error caused by quantification of the 14 bit A/D converter was about 2 m as was shown in the previous paper22), which was obtained when no signal was passed from the sensor. The main measurement error of this sensor was caused by the abrupt wearing away of the acryl plate, which seemed to be caused by local heating due to sliding friction. The measurement error of about 20 m was, however, small enough with regard to our practical purposes.
In this section, an aluminum plate and a piece of aluminum foil were used exclusively since they were not only easy to obtain but could also be easily worn away. However, in terms of specific practical applications, other metal foils such as gold foil and titanium steel can be used, i.e., metals that would not corrode in seawater, since eddy currents are induced in all metals. In all cases, a calibration curve for the corresponding metal should, of course, be pre-obtained for the practical use of this method.
In conclusions, the following results were obtained.
(1) The sensor unites an eddy current displacement sensor and a metal foil sandwiched between wedge-shaped acrylic plates in a body.
(2) The measurement error of this sensor system is about 20 m, which is sufficiently small for practical bearing monitoring applications.
(3) The whole system consists of this sensor, a data processing system including an amplifier, a 14-bit A/D converter, a personal computer, and a display.
References in Chap. 6 Part II:
10) Y.Yamamoto,K.Taniguchi,H.Yamamoto,K.Matsubara: “A New Technique for Measuring the Charge-to-Mass Ratio q/m of a Toner Particle", IS&T's 10th International Congress on Advances in Non-Impact Printing Technologies,pp.165-167(1994)
Yamamoto, K. Taniguchi, a, Y. Nakano, Y. Yamamoto, and Y. Takahara :A Method
for Measuring the Charge to Mass Ratio of a Spherical Toner Particle, Trans.
12) H. Yamamoto, K. Taniguchi, Y. Takahara, Y. Nakano, Y. Yamamoto,
and S.Watanabe: Measurement of the tribo-charge values on a surface of a Spherical Toner Particle, Trans. IEE Japan,,120-E,12, pp. 582-587(2000)
13) K. Taniguchi, H. Yamamoto , Y. Nakano, T. Sakai, S Morikuni, S. Watanabe and Y.Yamamoto: A New Technique for Measuring the distribution of Charge -to- Mass Ratio for Toner Particles with On-Line Use, Journal of Imaging Science and Technology, Vol.47, No.3, pp.224-228 (2003)
14) K.Taniguchi, T.Yagi, Y.Nakano, T.Sakai, H.Yamamoto, and S.Watanabe“An Improved Technique for Driven Characteristics for Toner Transportation", IS&T's NIP17: International Conference on Digital Printing Technologies, pp.856-859 (2001)
15) L.B.Schein:Electrophotography and Development Physics, 2nd edition,
16) E. Ioannides, & B. Jacobson: Dirty lubricants-reduced bearing life, Ball Bearing Journal Special 89(1989) p. 22.
17) A. Sasaki, S. Uchiyama, S. Kawasaki, S. Sjoeberg, K. Leola, & Reed T: Criticism on oil cleanliness standard and suggestion of a new method, 1997 Proc. Fifth Scandinavian Int'l. Conf. Fluid Power, SICFP 97(1997) p.243.
18) T. Yamaguchi, S. Kawaura, T. Honda, M. Ueda, A. Sasaki, & Y. Iwai: Investigation of oil contamination by colorimetric analysis, 2002 J. Soc. Tribol. Lubr. Engr. 58-1(2002) p. 12
19) T. Yamaguchi, T. Honda, Y. Iwai, M. Ueda, & A. Sasaki: Investigation of oil contamination by colorimetric analysis, Mem. Fac. Eng. Fukui Univ. 51-1(2003) p. 81.
20) Ed. Soc. Tech. Diagnosis: A report on a technical development for a detection and a diagnosis of a light water reactor - techniques for a diagnosis of fatigue B1-B28(2001) .
21) B. Nicola, & H. Yongqiang: Electrical conductivity measurement of metal plates using broadband eddy-current and four-point methods, Meas. Sci. Technol. 16(2005) p. 2193.
22) W. Yin, S. J. Dickinson, & A. J. Peyton: A multi-frequency impedance analyzing instrument for eddy current testing, Meas. Sci. Technol. 17(2006) p.393.
23) N. Ichinose: 1998 Industrial Materials 47-8 17 Jpn.
24) Y. Sotoyama, S. Ishizuka, T. Hirata, T. Sekino, T. Ogawa, T. Kuronuma, H. Ninomiya, Z. Dai, & T. Yamazaki: EBARA Topical Report 188 34(2000) Jpn.
25) T. Yamaguchi, Y. Iwai, S. Inagaki, & M. Ueda: Vibration monitoring system for a transmission shaft in real time using an eddy current displacement sensor, Trans. Soc. Instrum. Control Engr. 38-12(2002) p. 1129.
26) T. Yamaguchi, Y. Iwai, S. Inagaki, & M. Ueda: A method for detecting bearing wear in a drain pump ustilizing an eddy-current displacement sensor, Measurement 33-3(2003) p. 205.
27) T. Yamaguchi, & M. Ueda: An active sensor for monitoring bearing wear by means of an eddy current displacement sensor, Meas. Sci. Technol. 18(2007) p. 311.
[Editor’s Note: This concludes Chapter 6; the final chapter of this book,
Chapter 7, will appear in the upcoming July-August 2011 issue of this Journal.]
[ BWW Society Home Page ]
© 2011 The Bibliotheque: World Wide Society