A fundamental study of pulsed laser micro polishing (PLµP) was conducted as a method for reducing surface roughness of micro/meso-scale metallic parts. Although PLµP was shown to be an effective process at these scales, the knowledge of the process physics lacked thoroughness. The goal of this work, therefore was to improve upon the existing knowledge of PLµP, with focus on developing physics based models, understanding the effects of various process parameters and developing strategies for selection of process parameters and laser scan trajectories. In PLµP, a small area on a surface is irradiated with laser pulses to melt roughness features. The molten features are smoothed out by surface tension and viscous forces. Two polishing regimes for PLµP, namely thermocapillary regime and capillary regime, were defined based on whether the temperature gradient of surface tension driven thermocapillary flows are dominant or negligible.
Dominant thermocapillary flows, in the thermocapillary regime, result in significant smoothening of both low and high spatial frequency features, but introduce residual low-amplitude high spatial frequency features. In the capillary regime, thermocapillary flows are negligible and the molten features oscillate as stationary capillary waves that damp out due to the viscosity of the molten metal. The capillary regime is only effective at smoothening higher spatial frequency features. A physics based surface finish prediction model was developed and validated for the capillary regime.
Laser pulses are used to create shallow melt pools with a controlled size (e.g., depth) and duration in order to allow surface tension forces to “pull down” asperities with small radius of curvature. There is no ablation occurring in the process being modeled. The melt depth and duration are predicted with a transient, two-dimensional axisymmetric heat transfer model with temperature-dependent material properties. The surface of the melt pool is analytically modeled as oscillations of stationary capillary waves with damping resulting from the forces of surface tension and viscosity. Above a critical spatial frequency, fcr, a significant reduction in the amplitude of the spatial Fourier components is expected. The proposed prediction methodology was validated using line polishing data for stainless steel 316L and area polishing results for pure nickel, Ti6Al4V, and Al-6061-T6. The predicted average surface roughnesses were within 12% of the values measured on the polished surfaces.