Minimization of Kerf and Spatial Variation Analysis for Laser Cutting Process

Course: 2.830 Control of Manufacturing Process [MIT]

Collaborators: Ben Schilling, Yuwen Zhang

If you are someone who uses the laser cutter a lot, especially for prototyping mechanical designs, you have probably had an unpleasant experience because of the laser kerf. The kerf is the material that is removed by the laser beam during the cutting/engraving process. It is not possible to remove a 'hairline' width of material and so the kerf results in outer edges/boundaries being shorter than specified while holes and slots are oversized. This can be frustrating when your features for your prototype require certain tolerances. As the thickness or color of the material varies, so does the kerf. Moreover, given certain wear in the machine, there might be variation in the kerf along different directions and locations on the cutting bed. This aim of this project is to quantify and the amount of kerf that can be expected for any combination of process settings within an suitable operating range. Upon understanding the relation between process settings and the kerf, it is then possible to determine the combination of machine settings that can minimise the kerf. Knowing the expected (minimized) kerf, the designer, in the future, can decide how to suitable modify the 2D drawing files to compensate for this kerf, if needed. Alternatively, the designer can choose to operate the machine at predicted settings such that a desired kerf is achieved which provides the desired tolerance. 

Abstract— The effects of speed, power, location, and color of acrylic were investigated to determine the optimal settings to minimize kerf and variance in the x and y direction of an Epilog Fusion laser cutter. A 2^4 full factorial with a face centered composite design was conducted to determine the significant factors  in  the  laser  cutting  process  and  capture  the curvature which was determined to be significant in the initial experiments. From a multivariate analysis of variance (MANOVA), the significant factors are speed, location, color, color*color, and speed*location. According to the regression model that was derived from the experiments, the optimal settings to minimize kerf and the variability in the x and y direction is to use clear acrylic at 93% power and 9% speed. A spatial analysis was also conducted. It was determined that the optimal settings to have uniformity across the whole laser cutter is to use white acrylic at 85% power and 10% speed.