In-process Measurement of Topography of Wheel by Using Hydrodynamic Pressure Katsushi Furutani, Noriyuki Ohguro, Nguyen Trong Hieu 2 and Takashi Nakamura 3 Department of Advanced Science and Technology, Faculty of Engineering, Toyota Technological Institute 2-, Hisakata 2-chome, Tempaku-ku, Nagoya 468-85 Japan Phone: +8-52-809-796, Fax: +8-52-809-72, e-mail: furutani@toyota-ti.ac.jp. Introduction Wet grinding is one of the major ways of highprecision machining. An error factor in grinding is wear of a grinding. In order to reduce machining error, in-process measurement of the grinding wear is required. Loading and dulling on a grinding frequently occur under inappropriate grinding conditions. Eventually, the surface of a workpiece is burned. The working surface of a grinding is dressed at a certain interval to avoid the burning. Therefore, it is very important to measure both wear amount and topography change of the grinding during grinding process. fluid, which causes the difficulty of the inprocess measurement, is often used to cool a workpiece. Some in-process measurement methods of wear of a grinding have been proposed for both dry and wet grinding [][2]. Some methods to detect or measure loading, dulling and shedding of a grinding have been also proposed by using a triangulation displacement [3] for dry grinding. Because the grinding fluid disturbs light, it must be removed in wet grinding. Sensing acoustics [4] or vibration [5] has been also proposed for both dry and wet grinding. However, they are indirect methods. Some measurement methods by measuring static pressure caused by applying fluid to a grinding have been proposed [6]-[8]. Because these methods are similar to an air micrometer, superfluous grinding fluid must be removed for accurate measurement. The authors have proposed a measurement method using hydrodynamic pressure [9][0]. This method actively uses the grinding fluid to generate the pressure. The topography of the working surface of a grinding affects the flow of the grinding fluid. Therefore, it is expected that the topography change can be measured by analyzing the pressure. In this paper, the in-process measurement method of a grinding in wet condition is introduced. Then the relationship between the topography and the pressure during grinding is discussed. 2. Principle of measurement Fig. shows the measurement principle in the case of applying to a surface grinding process [9][0]. A pressure is set beside a grinding. Additional grinding fluid is poured near the gap between the grinding and a pressure. fluid is dragged by the rotation of the grinding into the gap. Hydrodynamic pressure is generated in the gap. Then the pressure change caused by the hydrodynamic pressure is measured with the. The hydrodynamic pressure decreases with an increase of the gap distance between the and the working surface of the grinding. After a relationship between the gap and the pressure is calibrated, the gap can be predicted by measuring the pressure. The electromagnetic properties of the grinding, a workpiece and grinding fluid do not affect the pressure. Viscosity of the grinding fluid seldom changes during grinding even if the temperature of the working fluid is slightly changed or debris is mixed. The diameter Nozzle for measurement Workpiece Nozzle for grinding Pressure Gap Flow meter Regulator valve Daikin Industries, Ltd. at present 2 Nagoya Institute of Technology at present 3 Nagoya Institute of Technology Fig. fluid tank Pumps Measurement principle by using pressure.
x y z -mm hole Center line fluid Pressure (6 mm in diameter) 25 Block O-ring 80 Table Specifications of grinding machine Revolution of Cylindrical 26.8 s grinding - grinding Maximum machine 29.9 m/s peripheral speed Type JIS-WA60J7V White fused Grain alumina Average grain size 250 µm Bond Vitrified Diameter 355 mm Width 50 mm Fig. 2 Gap change by the wear seldom affects the peripheral speed of the grinding because the ratio of the wear to the diameter is negligible. Therefore, the measurement process by this method becomes easier and more accurate than the measurement process with other noncontact s. Because the grinding fluid is actively used, no other additional device is needed. A roughness of the working surface of a grinding changes because of loading, dulling and shedding during grinding, and affects flow in the gap. As a surface becomes rougher, flow is separated earlier in general and becomes more turbulent. It is expected that its frequency components change. The influence of the roughness is investigated by frequency analysis. 3. Experimental setup (25 mm in width) Arrangement of pressure. An external cylindrical grinding machine was used in the following experiments. Table shows specifications of the grinding machine and a grinding. The grinding was composed of grains made from white fused alumina with an average diameter of 250 µm. Fig. 2 shows an arrangement of a pressure. The was covered with a plate with a hole of mm to avoid the direct contact with the grinding. The surface roughness of the plate is.7 µmr z. Table 2 shows specifications of the pressure. The with a diaphragm of 6 mm in diameter was a strain gauge type. The sensitivity of the pressure was 0 kpa/v at the output of the strain amplifier. The unit was mounted on the table of the grinding machine with a combination of an xyz and a rotational stages, which can be mounted and dismounted with a magnet chuck. The maximum pressure was at 5 Table 2 Specifications of pressure Strain gauge Type (diaphragm) Nominal pressure 00 kpa Pressure Diameter of 6 mm diaphragm Natural frequency 22 khz Linearity % Band width DC-200 khz Strain Accuracy ± % of FS amplifier Linearity ± % of FS FS: full scale mm above the contact point of the grinding with a workpiece. Because the unit was removed during grinding, the pressure must be exactly put back after grinding process by adjusting the position with the stage. The gap was measured with an eddy current displacement with a measurement range of mm and a resolution of 0.4 µm. Soluble-type grinding fluid containing surfactant was diluted 70 times with water. The grinding fluid was supplied at the maximum flow rate of.3 0-4 m 3 /s through a regulator. 4. Experiment 4. Adjustment of gap distance The gap distance between the grinding and the plate mounting the must be adjusted before the experiments. Fig. 3 shows the adjustment process of the gap distance. The was arranged in parallel to the working surface of the grinding. When the grinding contacts with the plate detecting by a spark on the plate, this distance of the gap was defined as zero. Then, the spindle was moved backward to change the gap. The gap distance is measured as the displacement of the spindle. The removed depth in the zero detection procedure is measured with a
Spindle Spindle Displacement Leave Displacement Approach Leave Spark Gap (a) Setting zero-gap distance. (b) Measurement of gap distance. Fig. 3 Gap adjustment process. profilometer to compensate the gap distance after the experiments. 4.2 Influence of supplying direction of grinding fluid for measurement A flow rate of the grinding fluid and a nozzle position affect conditions in the gap []. Their influence on the pressure is investigated. Fig. 4 shows two arrangements of a nozzle. The grinding fluid was supplied from the 20 or tangential directions. The pressure sensitively changed below a flow rate of 6 0-5 m 3 /s with the inclined nozzle. The grinding fluid was frothed with the rough surface of the grinding. This trend grew as the flow rate was increased with the inclined nozzle. However, the pressure was almost flat with the tangential nozzle. The supply from the tangential direction makes the pressure robuster against the ripple of the grinding fluid. The grinding fluid was applied from the tangential direction in the following experiments. 4.3 Relationship between pressure and gap distance The relationship between the pressure and the gap distance must be calibrated for each grinding before actual measurement. Fig. 5 shows a relationship between the pressure and the gap distance in the case of using the grinding shown in Table. The spindle was driven to change the gap. The grinding fluid was applied at a flow rate of 4.2 0-5 m 3 /s. The peripheral speed of the grinding was set to 26 m/s. The pressure was linearly decreased with a sensitivity of 50 Pa/µm from 5 to 30 µm in gap distance. The pressure changes in the same way in both the cases of changing the diameter of the grinding by dressing and of adjusting the spindle position because the diameter change affects its peripheral speed little. 4.4 Pressure change caused by loading Table 3 shows grinding conditions to accelerate loading of the working surface. Mild material, aluminum, was used for a workpiece. Against the general grinding conditions for aluminum, the structure of the grinding was porous and the peripheral speed of the workpiece was slow to deposit debris on the grinding by the accumulated heat.
20 o Nozzle Nozzle 7 62 26 5 (a) 20 -direction. (b) Tangential direction. Fig. 4 Arrangement of nozzle. Pressure kpa 20 5 0 5 0 Fig. 5 0 0 20 30 Gap µm Relationship between pressure and gap distance. Table 3 conditions to accelerate loading Type JIS-WA60J7V Peripheral speed 26 m/s Dressing Depth of dressing 0 µm conditions Feed speed 5 µm/s Method Plunge grinding conditions Depth of grinding 2.5 µm Material Aluminum (JIS- A204, ASTM- Workpiece 204), φ30 mm Peripheral speed 3 m/s Rotational Reverse of direction grinding Total amount of grinding was 208 mm 2 after grinding to a thickness of 50 µm. The wear of the grinding was too small to measure it. Frequency components were observed with a Fast Fourier Transform (FFT) analyzer. Fig. 6 shows an example of spectra of the pressure, which is the average of 50 measurements to avoid random errors. The initial gap was set to 20 µm. As the gap between the and the grinding increased with the progress of grinding, total pressure becomes smaller. However, some components, a frequency of 350 Hz mainly in this case, were increased after grinding. The roughness of the working surface of the grinding is decreased from 285 to 9 µmr z and from 3 to 9 µmr a because of loading. No diameter change was measured. This trend was clearly observed at higher flow rate. Fig. 7 shows a change of the spectra of the pressure with a progress of grinding. The horizontal lines Pressure after dressing kpa 0 Fig. 6 After loading After dressing 0 00 000 0.00 Pressure after grinding to 50- m kpa Example of pressure spectra before and after loading.
Pressure X0 kpa/div amount = 20 µm 5 µm 0 µm 5 µm Initial wear Dressing Freqency range to be observed Table 4 Experimental conditions for progress of loading JIS-WA60J7V Peripheral speed 26 m/s Dressing Depth of dressing 0 µm conditions Feed of table 5.2 mm/s Bearing steel Workpiece (JIS-SUJ2, Initial wear ASTM-5200) Peripheral speed 0.27 m/s Depth of grinding 2.5 µm Amount 40 mm 2 Aluminum (JIS- Workpiece A204, ASTM- Loading 204), φ30 mm Peripheral speed 3 m/s Depth of grinding 2.5 µm Amount 4 mm 2 Pressure Initial gap 0 µm measurement Flow rate 8.3 0-5 m 3 /s Fig. 7 0 00 000 Change of pressure spectra with progress of loading. indicate kpa in each set of spectra. Table 4 shows grinding conditions. Bearing steel (JIS-SUJ2, ASTM- 5200) is ground to a thickness of 0 µm to remove cracked part of grains just after dressing. Then the pressure was observed during grinding aluminum. Though the spectra from 400 to 500 Hz were observed very little after dressing and the initial wear by grinding bearing steel, they were increased with a progress of grinding. This band depends on the structure of the grinding. Fig. 8 shows a dependency of frequency components of the pressure on the gap distance. Higher components decreased with an increase of the gap distance. Smaller gap is preferable to detect the topography change. 4.5 Pressure change caused by dulling The working surface becomes flatter also by dulling. In-process measurement of dulling was also tried. After grinding a tungsten carbide rod with a diameter of 40 mm to a thickness of 50 µm with a depth of cut less than µm, the working surface of the grinding was dulled. Fig. 9 shows examples of the spectra of the pressure before and after dulling under the grinding conditions shown in Table 5. The spectra in a higher frequency Pressure X0 kpa/div Fig. 8 After loading at gap of 20 µm After loading at gap of 0 µm After dressing at gap of 0 µm 0 00 000 Comparison of spectra with respect to the gap distance. range are increased as well as loading. Therefore, dulling can be detected by this method. The discrimination among loading, dulling and shedding should be considered for detail analysis of the topography of the grinding. However, it is enough to detect the necessity of dressing. 5. Conclusions In this paper, the topography change of a grinding was observed using hydrodynamic pressure. The conclusions can be drawn as follows.
Pressure after dressing kpa 0 Fig. 9 After dulling After dressing 0 00 000 () Spectra of the pressure in higher frequency were increased with the progress of loading and dulling. (2) Not only the flow rate but also turbulent flow of the grinding fluid and air affected the measured pressure. (3) The pressure was decreased with an increase of the gap distance in the case of a narrow gap. A measured diameter of the grinding by this method changes by the wear and topography changes caused by loading, dulling and shedding. Their influence should be discriminated in the measurement process in future. Acknowledgement The authors wish to thank Mr. Masaru Ohta for his helpful cooperation. This study is financially supported by Osawa Scientific Grant (OSG Fund), Fluid Power Technology Foundation and Grant-in-Aid for High-tech Research Center Space Robotics Research Center and Academic Frontier Center for Future Data Storage Materials by Ministry of Education, Culture, Sports, Science and Technology, Japan. References 0.00 [] W. P. Dong, L. Annecchino and J. A. Webser: On- Line Measurement of Wear Using Acoustic Emission, Proc. Ann. Meet. Am. Soc. Precis. Eng., Monterey, California, USA, pp. 566-57 (996). [2] G. Spur and F. Leonards: Sensoren zur Erfassung von Prozesskenngrössen bei der Drehbearbeitung, Ann. CIRP, 24,, pp. 349-355 (975). [3] E. Brinksmeier and F. Werner: Monitoring of Wheel Wear, Ann. CIRP, 4,, pp. 373- Pressure after dulling kpa Example of pressure spectra before and after dulling. Table 5 conditions for measurement of dulling JIS-WA60J7V Peripheral speed 26 m/s Depth of Dressing 0 µm dressing conditions Feed of table 5.2 mm/s Method Plunge grinding Tungsten Workpiece carbide φ40 mm conditions Peripheral speed 0.33 m/s Depth of grinding less than µm 376 (992). [4] S. Hagiwara, T. Obikawa, M. Gomi, T. Matsumura, T. Iizuna, A. Yui: Prediction of Status Based on Dressing Sounds - Approaches by Linear Prediction Method and Neural Network -, J. Jpn. Soc. Precis. Eng., 67,, pp. 37-4 (200) [in Japanese]. [5] A. Hosokawa, K. Sakuma, and T. Ueda: In situ Characterization of Wheel Surface, Proc. 4th Ann. Meet. Am. Soc. Precis. Eng., Monterey, CA, USA, pp. 99-02 (999). [6] H. J. Schreitmüller and M. Dederichs: Pneumatisches Meßverfahren zur Ermittlung des Schleif-scheibenverschleißes, Industrie-Anzeiger, 93, 68, 733-734 (97). [7] T. M. A. Maksoud, A. A. Mokbel, and J. E. Morgan: In-process Detection of Wheel Truing and Dressing Conditions Using a Flapper Nozzle Arrangement, Proc. Insttn. Mech. Engrs., J. Eng. Manufact., 2, Pt. B(5), pp. 335-343 (997). [8] J. F. Rahman, V. Radhakrishnan: Measurement of Wheel Surface Topography Using Electro- Pneumatic Turbulence Amplifier System, Int. J. Mach. Tool Des. Res., 20, 3/4, pp. 89-96 (980). [9] K. Furutani, T. Katoh and N. Mohri: In-process Measurement of Wear of Wheel by Using Hydrodynamic Pressure in Wet Condition (st Report) - Measurement Principle -, J. Jpn. Soc. Precis. Eng., 66,, pp. 27-3 (2000) [in Japanese]. [0]K. Furutani, T. Katoh and N. Mohri: In-process Measurement of Wear of Wheel by Hydrodynamic Pressure, Proc. 4th Ann. Meet. Am. Soc. Precis. Eng., Monterey, CA, USA, pp. 60-63 (999). []S. Ebbrell, N. H. Woolley, Y. D. Tridimas, D. R. Allanson and W. B. Rowe: The Effects of Cutting Fluid Application Methods on the Process, Int. J. Mach. Tools Manufact., 40, 2, pp. 209-223 (2000).