Modelling of elliptical dimples generated by five-axis milling for surface texturing

Surface texturing processes that improve workpiece surface properties such as the friction of textured surfaces, the lubrication of sliding surfaces and the adhesion of workpiece surfaces are of increasing interest in industry. The most frequently employed processes for surface texturing are based on laser methods and on chemical etching processes. A widespread application of surface texturing is the improvement of tribological properties in terms of friction and load capacity. In these applications, surface texturing usually consists of generating dimples that are uniformly distributed on the workpiece surface. Five-axis milling with a ball-end mill can be an effective and productive option for these processes in these applications, as it is able to generate surface texturing that consists of periodic elliptical dimples if the tool geometry and the cutting conditions (feed, depth of cut and yaw and tilt angles defining tool axis orientation) are appropriately selected. In this paper, a model that predicts the geometry (shape, dimensions and orientation) of elliptical dimples machined by five-axis milling on flat surfaces for given cutting conditions is developed. In order to achieve this, equations expressing the trajectory followed by the cutting edges in five-axis milling are derived. The model takes into account the effect of tool parallel axis offset on dimple geometry. Next, the influence of cutting conditions on dimple geometry is analysed in order to define the cutting conditions that generate a given dimple geometry. From this analysis, a significant influence of the tools tilt angle on the elliptical shape of the dimples and a linear dependence of the yaw angle on dimple orientation are observed. In order to mill elliptical-shaped dimples, tilt angles larger than 30 degrees and feed, step over and depth of cut values that avoid interference between the dimples generated by different cutting edges of the ball-end mill are required. Finally, a series of five-axis milling tests was carried out in order to validate model predictions. A low dispersion in the dimensions and area of milled dimples was obtained. The shape and dimensions of predicted and measured dimples are compared and good correspondence between them is observed, the largest error being 7%.