Experimental statistical analysis of laser micropolishing process

Abstract

Laser micropolishing (LµP) is a new advanced m aterial microprocessing technology that attempts to smooth the original surface geometry through laser-material interactions such as melting or material ablation. Despite the significant advantages of LµP micro features, surfaces, parts, moulds and dies with complex 3D geometries from a wide range of materials, LµP is a complicated dynamic process that requires very fine tuning of a number of process parameters related to laser, optics, laser beam motions, and material properties. This paper describes a new approach for statistical analysis of LµP, where LµP is considered as a single-input (original surface) / singleoutput (polished surface) dynamic system. Original and polished cross-sections were obtained experimentally and their statistical characteristics, such as, surface roughness, material ratio function and autospectrums were calculated and analysed. In addition, LµP process was experimentally investigated as a dy namic operator represented by a transfer function and it was analysed using a coherence function. Analysis of these ch aracteristics allowed finding specific characteristics of the LµP process when surface roughness was improved by 21.3 %, lo wering averaged Ra value from 577 nm to 452 nm, and significantly reducing Ra non-uniformity from 132 nm to 44 nm for a Ti6Al4V sample.