2004 AIMETA International Tribology Conference, September 14-17, 2004, Rome, Italy EXPERIMENTAL INVESTIGATION OF CLEARANCE EFFECTS IN A REVOLUTE JOINT Alessandro TASORA, Edzeario PRATI, Marco SILVESTRI Department of Industrial Engineering, University of Parma 43100, PARMA, Italy Phone +39 0521 905903, Fax +39 0521 905705, E-mail: tasora@ied.unipr.it, prati@ied.unipr.it, silvestri@ied.unipr.it ABSTRACT This paper deals with the effects of clearances in revolute joints. Experimental data is acquired from a testbed made of a four-bar mechanism of crank-rocker type, with motorized crank and clearance between rod and rocker. Since the revolute joint pairs can be changed, different values of clearance can be tested. Also the radius of the crank, the length of the truss and the angular speed of the crank can be changed, in order to investigate different settings. During test sessions, rocker and rod accelerations are acquired with accelerometers, and output graphs are compared to the numerical results coming from off-line multibody simulations. Surface wear on the axle and bushing of the revolute joint is measured either with a micrometer tool, either with a three-dimensional surface roughness tester. Wear does not affect the entire surface of the shaft, but mostly happens on specific spots, as predicted by our numerical model. INTRODUCTION Clearances in revolute joints, for instance in four bar linkages and in slider-crank mechanisms, may cause rapid wearing, undesired vibrations, noise and low motion precision. Many articles address the need of numerical models which can realistically reproduce the odd motion of mechanisms with clearances [1][2]. In literature, most experimental works (for example [3],[4],[5]) deal with recording impact events, if any, in joints with clearance. In the present paper we present both the effects of clearances on accelerations and the outcomes of such impacts and vibrations on joint wear. Experimental acceleration data is compared to numerical simulations provided by a known model [6], which can handle separations and impacts in clearance pairs. A plain implementation of such model [7] tends to overestimate the oscillatory effects, therefore we developed a more advanced model [8] which features dissipative effects caused by friction, and which introduces a more advanced impact model. Measures on wearing can be useful to suggest predictive models for surface deterioration: we propose to use the Reye hypothesis to record the amount of work of the friction forces in polar coordinates. In fact measured wear does not affect the entire surface of the axle, but mostly happens on specific spots, as predicted by our numerical model that shows high-speed