Advances in Science and Technology Vol. 54 (28) pp 366-371 online at http://www.scientific.net (28) Trans Tech Publications, Switzerland Online available since 28/Sep/2 Friction Drive Simulation of a SAW Motor with Slider Surface Texture Variation Minoru Kuribayashi Kurosawa 1, a and Takashi Shigematsu 1,b 1 Tokyo Institute of Technology, 4259 Nagatsuta, Midori, Yokohama 226-852, Japan a mkur@ip.titech.ac.jp, b sgmt@ieee.org Keywords: actuator, ultrasonic motor, piezoelectric actuator, friction drive Abstract. Potential of a surface acoustic wave (SAW) linear motor had been demonstrated; high speed up to 1.5 m/s, huge output force more than 1 N, quick response and fine stepping motion down to.5 nm by using a 6 x 15 x 1 mm 3 SAW device at 1 MHz operation frequency and 4x4 mm Si slider by experiments. Mechanical output of the SAW motor was 27 times larger that of the weight and 5 W/kg in power. The conversion efficiency from the traveling wave power to the motor mechanical output was 14 %. Now, transformation mechanism research from high frequency vibration of 1 to 1 MHz frequency range to mechanical linear motion through frictional drive is noteworthy. Contact, friction and other physical matters between a stator and slider are being studied for stable operation and much superior performance. The slider friction surface has a lot of projections to control the contact with the stator. From the experimental results, it was shown that the slider surface projection diameter has huge influence on the motor characteristics. If the total projection contacting surface areas were same, the output force, for example, varied with the projection diameter; smaller projection, for example 2 µm diameter, had superior performance than larger one. In this paper, numerical simulation of SAW motor friction drive was successfully carried out taking account the wave amplitude attenuation in propagation beneath slider projections. Introduction For miniaturization of ultrasonic transducers to fabricate piezoelectric motor, a surface acoustic wave device has an advantage in rigid mounting and high-power-density operation. A surface acoustic wave (SAW) motor has been investigated [1]-[4], and its superior performances have been demonstrated: a speed of 1.5 m/s [5], a thrust of 12 N [4], and a stepping motion of.5 nm [6], [7] in the case of using a 6x14x1 mm 3 plate piezoelectric transducer. In addition, to increase the application range, a higher operation frequency of 1 MHz [8] and a two-dimensional design were investigated. In addition, to reduce power consumption, energy circulation driving has been proposed [9] and investigated. From investigations based on experiments, it was found that slider surface texture affects motor performances such as speed and thrust [3]-[5]. Theoretically, however, the effect of the physical property of a slider-textured surface on motor performance has not been investigated sufficiently [1]. Physical modeling of the SAW motor has been attempted, one slider projection was modeled including the compliance of the slider and stator materials [11], and also the stick and slip at the boundary. Using the slider projection modeling, operations of the SAW motor were simulated, and then, the results were compared with the experimental results [12]. In this paper, for preciseness of simulation, the attenuation of the traveling wave under going the slider was being taken account in the simulation. Principle and Example of Motor Using a Rayleigh wave for a surface acoustic wave motor, a basic device construction is shown in Fig. 1. The motor has only a thin plate transducer and a thin friction material. The plate transducer All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 131.112.125.15-5/9/8,4:3:58)
Advances in Science and Technology Vol. 54 367 is a surface acoustic wave device to generate a Rayleigh wave using two interdigital transducers (IDTs) to excite a bidirectional traveling wave. One IDT at a time is driven to excite the wave, and then, to drive the slider in one direction. The active IDT is changed for the alternative linear motion of the slider. For friction force, which generates the linear motion of the actuator, preload on the slider is given, so that the slider is pressing against the stator. To explain the principle of the actuator driving mechanism, a drawing in which the wave motion is enhanced is convenient to show the image. Actual wave motion is too small to present in the drawing; therefore, the displacement of wave motion is enlarged about 1,-fold in Fig. 2. Regarding the actual dimensions of wavelength and vibration displacements, the stator surface is almost flat; in the case of 9.6 MHz driving frequency, for example, the wavelength and vibration displacements are about 4 µm and 2 nm, respectively. Physically, therefore, the contact between the stator and the slider contains elastic deformation, stick, and slip. At the crest, the particles in the stator have the maximum vibration velocity in the tangential direction, which is opposite to the wave traveling direction. The large frictional force and the large tangential vibration velocity generate the linear motion of the actuator. The limit of the actuator speed is the vibration velocity of the wave crest, which depends on vibration amplitude. The limit of the actuator output force is the friction force; at the maximum, the product of the preload and the frictional coefficient between the stator and the slider. For stable friction driving conditions, many projections are fabricated on the surface of the slider [4]. The diameter of the projections was 2 µm or smaller to obtain a large thrust and a high speed close to vibration velocity in the case of 9.6 MHz driving frequency. It has been found from experiments that the larger the projection diameters, such as 3 or 5 µm, the lower the performance in terms of speed and thrust [4]. For the experiments, several types of silicon slider were fabricated by the dry etching process. The slider dimensions were 5x5 mm 2 ; 2 µm high projections were fabricated in a 4x4 mm 2 area. For experiments [4], the material of the SAW device was 128 degree y-rotated x-propagation LiNbO3. The dimensions of the device were 6 mm long, 14 mm wide, and 1 mm thick. The electrodes were twenty pairs of IDTs that were 1-µm-wide chromium/aluminum line had a 1 µm gap, and were 9 mm in aperture. The resonance frequency, namely, the operation frequency, was 9.61 MHz. When the driving voltage was 125 V -p, the vibration amplitude and velocity were 24 nm in the direction normal to the surface and 1.3 m/s in tangential direction, respectively; the input power was 7 W. The SAW device was 3.9 g in weight and then total weight of the actuator device including Si slider was less than 5 g. On the other hand, the maximum thrust and the maximum mechanical output power of the actuator were 13 N and 2.5 W. From these values obtained by experiments, mechanical output of the SAW motor was 27 times larger that of the weight and 5 W/kg in power [13]. The conversion efficiency from the traveling wave power to the motor mechanical output was 14 % [13]; the flowing wave power beneath the slider was quarter of the input power to the IDT. RF power preload Slider Electrode (IDT) slider motion slider Rayleigh wave SAW device (Stator transducer) traveling wave stator Fig. 1 Schematic of surface acoustic wave motor. Fig. 2 Contact between traveling wave in stator and projecttions on slider surface.
368 Smart Materials & Micro/Nanosystems Elastic Contact Model A physical modeling of frictional drive of a SAW motor has been carried out [11] on the basis of contact mechanics [14]. For the first step of the modeling, a slider elastic body, a rigid projection, and a stator elastic body were expressed using four springs, one rigid body connected to the elastic slider part, and frictional boundary surfaces, as shown in Fig. 3. In the modeling drawing shown in Fig. 3, a v, a h, P, and Q are the vertical wave amplitude, the horizontal wave amplitude, the vertical direction force acting on the projection, and the thrust force acting on the projection surface, respectively. The equivalent spring constant of the slider and the stator are indicated by k pn, k pt, k sn, and k st in the normal and tangential directions, respectively. Using the physical model of one projection, we carried out a numerical simulation in time domain including preload, friction coefficient, vibration amplitude, and so on [15]. From the simulation, the thrust between the stator and the slider were obtained, then, the mean speed and thrust of the slider were estimated [16]. It is understood that the friction drive has two parts: sticking with elastic deformation in the nanometer range and slipping at the boundaries. Simulation Based on Physical Modeling Simulations of a SAW motor operation were carried out using the projection contact model and results were compared with experimental results at 9.6 MHz motor operation. The driving voltage was 125 V -p, so that the vibration amplitude and velocity were 24 nm in the direction normal to the surface and 1.3 m/s in tangential direction. The slider projection number was changed from 189 to 1 in the case of a 2-µm-diameter slider. Then, performance differences depending on the slider projection diameters ranging from 2 to 5 µm with same contact surface area of about 3 mm 2 in 4 by 4 mm 2 were compared with the experimental results [4], [17], [18]. For the first step in the simulation, the projection distribution in the wave traveling direction was ignored. Namely, the attenuation of the wave by the friction drive was neglected to simplify the calculation. For the second step, attenuation of the wave by the friction drive was introduced. The material constants are introduced in a literature [12]. It is difficult to maintain a uniform contact condition for each of the projections distributed in a 4x4 mm 2 area, owing to the small vibration amplitude and elastic deformation in the nanometer range. Hence, the effective factor [12] of projection contact was investigated and then fixed to be.6 in this paper. In addition, the friction coefficient between the slider and the stator was fixed to be.14 to make the incline of the thrust curves under low preloaded condition agree with the experimental results. The friction coefficient of.14 agrees with the value measured using a friction tester [19]. In addition, detail of the modeling is discussed in other papers [2]-[23]. k pn kpt Slider Projection Stator P k st k sn Q a v a h Stator Motion Fig. 3 Contact modeling of projection on slider and stator surface.
Advances in Science and Technology Vol. 54 369 No attenuation by friction drive. Using five different slides with projection numbers of 1, 64, 4356, 25 and 189, thrust and speed that were estimated by the simulation were compared with experimental results; the projection diameter was 2 µm. The thrust at null speed, namely, maximum thrust, depends on the preload and the number of the projections, as indicated in Fig. 4 [12]. There are five curves obtained for simulation and plots for experimental results, as indicated in Fig. 4 as follows: n=1, 64, and so on. From the simulation, it is understood that the stiffness at the slider contact surface is important for a large thrust. This is because the stiffness is in proportion to the number of projections. The stiff surface is suitable for high-speed operation, as shown in Fig. 5 [12]. In the case of the soft slider surface, the projections contact the bottom of the wave with a large preload. With attenuation by friction drive. Frictional driving simulation was carried out taking account the wave attenuation by friction drive. The traveling Rayleigh wave motion drives the slider projections at the conditions of contact/take-off and stick/slip. The formerly carried out simulation did not include the wave attenuation by slider projection drive. It means that the wave amplitude was maintained constant in spite of the mechanical output power and frictional loss for a first order approximation. Actually, however, at the first row, the flowing wave power is consumed, so that the wave amplitude will be decreased at the second slider row. After the several tens of projection row friction drive, in some case, the amplitude becomes half or less, due to the power consumption for the friction drive. From the actual dimensions of the slider pitch and the wavelength, continuously distributed modeling of the wave and the projections are required for preciseness. However, if the projection pitch is long enough compared with the wavelength, wave attenuation is estimated by one projection row and then the attenuated wave amplitude will be applied to the next projection row friction drive. By introducing the wave attenuation at each projection row, influence of the slider projection arrange can be taken account. Friction drive simulation was carried out using the elastic contact model of projection for each projection row. From the given vibration amplitude fixed by the driving voltage and IDT design the preload value, simulation was carried out step by step as follows; 12 1 n=1 1.2 1 Thrust s [N] 8 64 6 4356 4 25 2 189 Speed f [m/s].8.6.4.2 n=1 64 189 25 4356 Fig. 4 Thrust of 2 micron diameter projection slider with change in number of projections from 1 by 1 to 33 by 33. Fig. 5 Speed of 2 micron diameter projection slider with change in number of projections from 1 by 1 to 33 by 33.
37 Smart Materials & Micro/Nanosystems Step 1: at the first row of the slider projection, giving the contact area of the wave and slider speed then calculated the preload, output force, and consumed power. Step 2: from the flowing power, the vibration amplitude is obtained. Then the contact area of the wave is calculated to meet the preload value at Step1. From the slider speed, output force and the consumed power are estimated. Step 3: following projections driving conditions are simulated one by one to end row Step 4: summation of the preload, thrust and power give the total output and loss in the slider. 1 1.4 Thrust s [N] 8 6 4 2 189 25 4356 64 n=1 Speed f [m/s] 1.2 1.8.6.4.2 189 25 4356 64 n=1 Fig. 6 Thrust characteristics of 2 micron diameter projection slider with change in number of projections from 1 by 1 to 33 by 33; in case of wave attenuation in propagation projection rows. Fig. 7 Speed characteristics of 2 micron diameter projection slider with change in number of projections from 1 by 1 to 33 by 33; in case of wave attenuation in propagation projection rows. Thrust s [N] 1 8 6 4 2 ø5 µm ø3 µm ø2 µm Speed t [m/s] 1.4 1.2 1.8.6.4.2 ø5 µm ø2 µm ø3 µm Fig. 8 Differences in thrust characteristics between projections of 2-micron-, 3-micron-, and 5-micron-diameter sliders. Fig. 9 Differences in speed characteristics between projections of 2-micron-, 3-micron-, and 5-micron-diameter sliders.
Advances in Science and Technology Vol. 54 371 Simulation was carried out at the same conditions mentioned above in the case without wave attenuation. Simulation results are shown in figs. 6 and 7. They are similar to the results shown in Figs 4 and 5. However, the decrease of the thrust at large preload condition seems to be deferent from the previous simulation results. Now at the large preload condition, the calculation became unstable, so that the drawing area of the simulation is limited. A difference in slider projection diameter as a function of the output force of the SAW motor indicated in Fig. 8. The simulation results are indicated by a dotted line, a dashed line, and a solid line for 2, 3, and 5 µm projection diameter sliders, respectively. The higher surface stiffness due to small-diameter projections provides superior performance in terms of speed also, as shown in Fig. 9; about three different-diameter sliders, similar to that shown in Fig. 8. Three sliders had the same contact area of 3 mm 2 but different projection diameters; the 2-µm-diameter projection slider had 1 projections. The small diameter projections have lower stiffness than the large projections. However, the total stiffness becomes higher if the total projection contact surface area is the same as the large projection slider. It is clear that the stiffness is proportional to the radius of projections, not to the square of projection radius. Thus, if the total projection areas are the same, the smaller projection slider has a higher surface stiffness. The high surface stiffness of slider induces a large thrust. Conclusion Numerical simulation of SAW motor friction drive was successfully carried out taking account the wave amplitude attenuation in propagation beneath slider projections. The decrease of thrust of motor at large preload condition became similar to experimental results. Calculation process for the simulation, however, required improvement for much larger preload condition estimation. References [1] M. Kurosawa, M. Takahashi, and T. Higuchi: IEEE Trans. UFFC, 43 (1996) 91. [2] M. K. Kurosawa, M. Takahashi, and T. Higuchi: IEEE Trans. UFFC, 45 (1998) p. 1229. [3] K. Asai, M. K. Kurosawa, and T. Higuchi: Proc. IEEE Ultrasonics Symp., 2, p. 675. [4] M. K. Kurosawa, H. Itoh, K. Asai, M. Takasaki, and T. Higuchi: Proc. MEMS, 21, p. 252. [5] Y. Nakamura, M. K. Kurosawa, T. Shigematsu and K. Asai: Proc. IEEE Ultrasonics Symp., 23, p. 1766. [6] T. Shigematsu, M. K. Kurosawa, and K. Asai: IEEE Trans. UFFC, 5 (23) 376. [7] T. Shigematsu, M. K. Kurosawa, and K. Asai: Proc. IEEE-NANO, 23, p. 299. [8] T. Shigematsu and M. K. Kurosawa: IEEJ Trans. Sens. Micromach., 126-E (26) 166. [9] K. Asai and M. K. Kurosawa: IEICE Trans. Fundam., J86-A (23) 345 [in Japanese]. [1] K. Asai and M. K. Kurosawa: IEICE Trans. Fundam., J86-A (23) 1442 [in Japanese]. [11] T. Shigematsu and M. K. Kurosawa: Proc. Actuators 26, p. 444. [12] M. K. Kurosawa and T. Shigematsu: Jpn. J. Appl. Phys., 47-5B (28) in printing. [13] M. K. Kurosawa: Jpn. Soc. Mech. Eng., 111 (28.5) in printing [in Japanese]. [14] K. L. Johnson: Contact Mechanics (Cambridge University Press, Cambridge, U.K. 1985). [15] K. Asai and M. K. Kurosawa: IEICE Trans. Fundam., J85-A (22) 1428 [in Japanese]. [16] T. Maeno, T. Tsukimoto, and A. Miyake: IEEE Trans. UFFC, 39 (1992) 668. [17] M. K. Kurosawa, H. Itoh, and K. Asai: Proc. Transducers, 21, p. 726. [18] M. K. Kurosawa, H. Itoh, and K. Asai: Ultrasonics 41 (23) 271. [19] T. Shigematsu and M. K. Kurosawa: to be published in IEEE Trans. UFFC. [2] T. Shigematsu and M. K. Kurosawa: to be published in IEEE Trans. UFFC. [21] T. Shigematsu and M. K. Kurosawa: to be published in IEEE Trans. UFFC. [22] T. Shigematsu and M. K. Kurosawa: to be published in IEEE Trans. UFFC. [23] T. Shigematsu and M. K. Kurosawa: to be published in IEEE Trans. UFFC.