Abstract
The paper presents the results of a thermal analysis of a set of disk experiments carried out by Patching et al. to investigate scuffing. The experiments used crowned steel disks at 76-mm centers with maximum Hertzian contact pressures of up to 1.7 GPa. Experimental measurements of contact friction were used as the basis for a thermal analysis of the disks and their associated support shafts. Temperatures measured by embedded thermocouples 3.2 mm below the running tracks of the disks were used to determine the heat partition between the faster and slower running disks in order to match the experimental with calculated temperatures. This partition was found to vary approximately as a function of the product of sliding speed and surface temperature difference. A transient (flash) temperature analysis of one of the experiments was also carried out. This shows large differences between the disk transient surface temperatures. These surface temperature distributions were compared with those obtained from corresponding elastohydrodynamic lubrication (EHL) analyses using two different non-Newtonian lubricant formulations. The EHL analyses show that the heat partition obtained depends on the form of non-Newtonian behavior assumed, and that to achieve the same partition as is evident in the experiment a limiting shear stress formulation is necessary. It is suggested that the combination of heat transfer and EHL analysis presented in the paper could be used as a sensitive tool for distinguishing between different non-Newtonian lubricant models under realistic engineering loads and with high sliding speeds.
