Research Papers

Friction Compensation of Geared Actuators With High Presliding Stiffness

[+] Author and Article Information
Myo Thant Sin Aung

Department of Mechanical Engineering,
Kyushu University,
Fukuoka 819-0395, Japan
e-mail: aung@ctrl.mech.kyushu-u.ac.jp

Ryo Kikuuwe

Department of Mechanical Engineering,
Kyushu University,
Fukuoka 819-0395, Japan
e-mail: kikuuwe@mech.kyushu-u.ac.jp

Motoji Yamamoto

Department of Mechanical Engineering,
Kyushu University,
Fukuoka 819-0395, Japan
e-mail: yama@mech.kyushu-u.ac.jp

Contributed by the Dynamic Systems Division of ASME for publication in the JOURNAL OF DYNAMIC SYSTEMS, MEASUREMENT, AND CONTROL. Manuscript received August 5, 2013; final manuscript received April 21, 2014; published online August 28, 2014. Assoc. Editor: Luis Alvarez.

J. Dyn. Sys., Meas., Control 137(1), 011007 (Aug 28, 2014) (8 pages) Paper No: DS-13-1299; doi: 10.1115/1.4027503 History: Received August 05, 2013; Revised April 21, 2014

Most of existing friction compensation techniques are based on friction models that uses the velocity as its input. These methods are difficult to apply to inexpensive encoder-based actuator systems that do not exhibit sufficiently large presliding displacement. This paper presents a new method of friction compensation that can be applied to geared actuators with high presliding stiffness. The compensator consists of three components that compensate: (a) static friction, (b) rate-dependent kinetic friction, and (c) dynamic friction involving presliding viscoelasticity. The first component employs dither-like torque command, and the other two are based on friction models involving precalibrated parameters. The proposed method is validated through experiments employing a harmonic drive transmission. In particular, it is suggested that the dither-like static friction compensation and the viscosity in the presliding model significantly improve the performance of the compensator.

Copyright © 2015 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Friction reduction by a feedback-based friction compensator

Grahic Jump Location
Fig. 2

Experimental setup

Grahic Jump Location
Fig. 3

Experimental results obtained by ramp torque input (r = 0.5 Nm/s). The torque τ is set zero once the predefined level τmax is achieved.

Grahic Jump Location
Fig. 4

Measured friction-displacement characteristics under the actuator torque τ applied according to Eq. (1) with τmax = 3.0–3.6 Nm: (a) friction-displacement characteristics. (b) Zoomed view of “A” in (a). In the results, pure sliding occurs only after the torque τ became larger than approximately 2.83 Nm.

Grahic Jump Location
Fig. 5

Example of data obtained from a sinusoidal motion

Grahic Jump Location
Fig. 6

Friction-velocity characteristics: obtained data (markers) and fitted curve (solid). There are 44 markers, which were obtained from 22 different amplitudes. Two of the markers (one in the first quadrant and the other in the third quadrant) correspond to the data in Fig. 5.

Grahic Jump Location
Fig. 7

System model represented by Eqs. (3), (5), and (7). In the model, the blocks move only in the horizontal direction. The vertical spring connected to a ball illustrates a ball-plunger-like mechanism, which is to constrain the relative horizontal motion between the blocks A and B when the ball is pushed into the dent on the bottom surface of the block A.

Grahic Jump Location
Fig. 8

Experimental data obtained during the dither-like actuation (Z = 3 counts, r = 30 Nm/s)

Grahic Jump Location
Fig. 9

Structure of Algorithm 1, i.e., the proposed friction compensator: the blocks (a)–(c) correspond to those in Algorithm 1

Grahic Jump Location
Fig. 10

Results of experiment 1: (a) An example of a set of experimental results. (b) Zoomed view of “A” in (a): measurement of Te and Fe. (c) Averages and standard deviations of Te. (d) Averages and standard deviations of Fe.

Grahic Jump Location
Fig. 11

Results of experiment 2: averages and standard deviations of VAMP obtained with low actuation level. The data of VAMP obtained with NO are not plotted because they are all below 10−30deg/s. Horizontal grid lines show the τAMP values multiplied by 1, 10, and 100 deg/(s·Nm).




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In