Bearing Surface Optimisation on Hydrodynamic Lubrication Film with Vibration-Assisted Machining and Ultrasound Reflectometry

The collapse of full-film hydrodynamic lubrication in a journal bearing often leads to a significant failure of the whole machine. In order to delay such full-film lubrication collapse, surface optimisation by implementing bespoke surface texture has spurred industrial interest. This thesis aims to optimise the journal bearing surface for an automotive transmission, addressing the deficiency of understanding of the effects of bespoke surface texture on hydrodynamic lubricant film formation under real operation by developing an accurate, inexpensive and fast texturing technique to create bespoke surface textures and a non-invasive technique to measure the film thickness. A novel variant of non-resonant vibration-assisted machining was developed using an off-the-shelf piezoelectric actuator to create bespoke surface textures. Surface textures consisting of a repeating radial striation pattern of sine waves were reproducibly generated on the face of the disc work piece (an aluminium alloy AlSi1MgMn and a low-alloyed steel 16MnCr5) in a conventional milling machine when the frequency of the superposed vibration was in phase with the rotational speed of the work piece. The developed device was successfully implemented on shaft work pieces (alloyed steels C40 and SCM420H) in a conventional turning machine also. A journal bearing test platform, which represents operational conditions of an automotive transmission, was developed incorporating a non-invasive ultrasound reflection technique. On the test platform, film thickness measurement was examined under operational conditions in steady-state. The measured film thickness and the attitude angle on the plain journal bearing (98 mm of the diameter with 0.255 of L/D) agreed well with theoretical curves deduced by a classical Reynolds equation with the short-bearing approximation. Film thickness measurements on two textured journal bearings (a rough-meshed by a conventional milling and a fine-meshed by the vibration-assisted machining device) were performed on the journal bearing test platform. The load capacity ratio, i.e. a ratio of the load capacity experimentally measured on the textured bearing and on the un-textured bearing, were 0.40 to 0.59 in the rough-meshed bearing and 1.12 to 1.27 in the fine-meshed bearing. Despite several differences in the geometrical model and test conditions between the present experimental work and the literature CFD works compared, the load capacity ratios were quantitatively similar.