Running-in of gears - surface and efficiency transformation

With ever shorter development times and market demands on overall system performance such as efficiency, reliability and low maintenance, accurate predictive tools are necessary and gear drives prove to be no exception. All these characteristics have an impact on a process which has remained a riddle: running-in. Even though no consensus on a definition of this phenomena is readily available, this thesis examines efficiency, surface roughness and simulation through the optics of running-in. Geared transmissions are known for their formidable efficiency and their extreme reliability. However, with an ever increasing power density, the ability to accurately predict mesh losses becomes of utmost importance. The accurate quantification of bearing losses as well as efficiency of ground and superfinished gears under dip lubrication are examined with respect to running-in. Results show a considerable influence on the calculation of gear mesh losses originating from which bearing loss model is chosen. Furthermore, when a larger running-in load is used on ground gears, an increase in efficiency can be observed during working operation, while for superfinished no significant changes are found. These efficiency/frictional changes are also shown to occur in the initial cycles of the running-in phase. From a surface transformation point of view running-in is shown to be a reduction of asperity tips in case hardened ground gears, while in superfinished gears no changes were seen. These gear surface changes were measured with a novel method with a surface profilometer in-situ before, after running-in and after efficiency testing. Results also show that such changes in ground gear roughness profile occur during the very initial cycles. In order to predict running-in, a simulation method was developed. Such method utilizes a 2D surface integral method to simulate contact between rough surfaces, but requires the use of surface hardness and an accurate lower cutoff wavelength. This cutoff wavelength proved to play a pivotal role in determining an accurate contact pressure at the proper level of granularity, hence a well defined real contact area. The predicted and measured run-in surfaces are compared and are found to be in accordance with each other.