Scuffing is a prominent surface failure mode of loaded, lubricated contacts of gears and rolling element bearings experiencing excessive relative sliding and high speeds. This temperature-induced failure occurs suddenly when the contact temperatures reach a critical level due to the frictional heat generated at the contact interface. Material properties and geometry of contacting surfaces, operating conditions (normal load, relative sliding and speed), surface texture (roughness amplitude and direction) as well as physical and chemical properties of the lubricant all influence the scuffing behavior of such components.
In this study, a physics-based methodology is proposed for predicting thermal conditions of lubricated contacts under combined sliding and rolling, and for relating these thermal conditions to the likelihood of scuffing. The methodology combines (i) a mixed thermal elastohydrodynamic lubrication (EHL) model to predict temperatures of the contacting surfaces and in the lubricant film in between, (ii) a convective heat transfer model to predict the time-varying temperature distributions of the contacting bodies, and (iii) a scuffing criterion to predict the onset of scuffing.
Considering a generic two-dimensional point (elliptical) lubricated contact problem formed by two rough roller surfaces that are in relative sliding, a new thermal EHL model is proposed first to predict the instantaneous pressure, film thickness, and temperature distributions under considerable metal-to-metal contact conditions in a robust and computationally efficient way. A novel iterative process is devised to combine a convective heat transfer model of the contacting bodies with the thermal EHL model. The bulk temperatures required by the thermal EHL model are predicted in this process by the heat transfer model while the heat flux into the contacting bodies and the heat partitioning coefficients required by the heat transfer model are provided by the thermal EHL model. An extensive set of two-disk experiments are performed to (i) establishing a link between the bulk temperatures and relative sliding during traction tests, and (ii) the limits of scuffing under various speed, load and sliding conditions. Both types of experiments are simulated by using the proposed model to demonstrate its accuracy.
The proposed general methodology is applied to a spur gear problem by considering variations of contact parameters along the tooth surfaces and incorporating a gear load distribution to predict contact loads. This spur gear scuffing model uses a one-dimensional (line contact) thermal EHL model and a convective heat transfer model of a gear pair in an iterative manner to predict the maximum instantaneous contact temperatures, which are used with the scuffing temperature limits established by the experiments to determine the likelihood of scuffing to occur. At the end, the proposed methodology is compared to the conventional gear scuffing criteria to highlight its capabilities to overcome the major shortcoming these criteria.