An enhancement of an existing tribometer device developed by Philippon et al. (Wear 257:777-784, 2004) is presented in this work. This experimental device is made up of a dynamometer ring and a specific load sensor allowing to apply an apparent normal force on specimens and to measure frictional forces respectively. A set of strain gauges are added to the upgraded dynamometer ring in this new configuration. The apparent normal force can be recorded accurately during the sliding process. The setup is adapted on a hydraulic testing machine to carry out steel-on-steel dry sliding tests. The first set of standard Steel on standard Steel specimens (XC 38 French standard steel) with two apparent normal pressures are imposed (8 and 80 MPa) as the range of sliding velocities varies from 0.12 to 3.72 m/s for the same contact conditions. The main set of experiments with low sliding velocities (varying from 0 to 3 m/s) for the Steel 1080 on Steel VascoMax are performed in the same tested setup. The recordings of normal and tangential forces leading to the friction coefficient determination are discussed. The values of dry friction coefficient μ according to the experimental parameters are in good agreement with those observed in the literature. Using this new configuration, the effects of the sliding velocity on the surface roughness changes and on the dry fiction coefficient are also investigated. Additionally the surface roughness changes are also investigated. Performing the scans with use of the scanning electron microscope in particular locations of the specimens show the roughness decrease and reveal the occurrence of the wear phenomenon. Moreover, very interesting relations between wear and sliding velocity are observed.

No abstract

An enhancement of an existing tribometer device developed by Philippon et al. (2004, Wear 257, 777-784) is presented in this work. This experimental device is made up of a dynamometer ring and a specific load sensor allowing to apply an apparent normal force on specimens and to measure frictional forces respectively. Problems such as static calibration of both distinct parts supported by conducted numerical simulations is given and described in detail. Further discussion, again supported by numerical analysis, validates the placing of the strain gauges and explains the problems of sensitivity of loading surface and cross-sensitivity. The entire experimental setup is then discussed explaining the mounting of the strain gauges located on the tribometer device, showing the configuration of the Wheatstone quarter-bridges, presenting the signal conditioning amplifier system and the digital oscilloscopes which all together create one uniform data acquisition system. The set up is adapted on a gas gun to carry the Steel 1080 on Steel VascoMax 300 experiments. The main set of experiments with sliding velocities varying from 10 to 60 m s )1 are performed in the same tested setup. The recordings of normal and tangential forces leading to the friction coefficient determination are discussed. The values of dry friction coefficient l according to the experimental parameters are in good agreement with those observed in the literature. Using this new configuration, the effects of the sliding velocity on the surface roughness changes and on the dry fiction coefficient are also investigated. Moreover, very interesting relations between wear and sliding velocity are observed.

Many physical systems require the description of mechanical interactions across interfaces if they are to be successfully analyzed. One of the well-known examples of such a system in the engineering world is the metal to metal friction. This is a complex process that needs to be adequately identified by a constitutive relation in order to better facilitate the design components in severe contact stress applications. In this paper, the formulation of Molinari et al.'s work (J Tribology Trans ASME 35-41, 1999) is revisited in order to investigate the coefficient of dry friction for steel on steel in the high velocity range using physically based viscoplastic constitutive relations. First some of the errors in the work are corrected, and their results are regenerated. The phenomenological constitutive relation used in Molinari et al. (J Tribology Trans ASME 35-41, 1999) is then replaced by the physically based viscoplastic model used in this paper. This constitutive model is implemented into ABAQUS (Analysis User's Manual, 2008) as user-defined subroutine as VUMAT in order to obtain the stress-strain curves at different strain rates and various temperatures. It is shown that the material responses obtained from the simulation using the physically based constitutive viscoplastic model agree well with the real behavior of the metals. Comparing this proposed work with that of Molinari et al. (J Tribology Trans ASME 35-41, 1999) one observes that the proposed theory and constitutive model are superior to the one presented by Molinari et al. This is specifically the case for the artificial shape of softening in the curve.

No abstract