Surface morphology of low-carbon-steel oxygen-assisted laser cutting

Surface morphology of oxygen assisted laser cutting of low carbon steel sheets exhibits characteristic drag lines regularly spaced and more or less pronounced depending on process parameters and depth of observation. Examination of roughness profiles does not correspond directly to the periodicity of the drag lines, especially with cutting speed near to oxidation reaction front speed (about 2 m min -1 ). In these conditions the two phenomena (laser cutting proper as opposed to oxygen cutting) overlap and the two types of accidents, superimposed over the cutting surface with different periodicity and amplitude, become confused. In this case the roughness profile can be considered as a sum of two populations with different mean and standard deviation, therefore, the height density probability distribution is bimodal and it can be represented as a mixture of two univariate distributions. In this paper a new algorithm is illustrated in order to establish the two parameters of each unimodal distribution, which is well modeled by a Beta-distribution, and the degree of mixing, useful for the evaluation of the relative influence of the two phenomena previously described.