Femtosecond Laser Polishing of AlN Ceramics and Numerical Simulation of Ablated Morphology.

Abstract

To meet the surface polishing requirements of aluminum nitride (AlN) ceramics, this study developed a multi-objective optimization experimental model based on response surface methodology (RSM), with surface roughness as the key optimization target. A systematic series of femtosecond laser polishing experiments were conducted. Polishing effectiveness and the evolution of material properties under different process parameters were comprehensively evaluated through surface morphology characterization, microhardness testing, friction and wear experiments, and energy-dispersive X-ray spectroscopy (EDS) analysis. The experimental results indicated that the optimal combination of process parameters, as determined by RSM optimization, was identified as a laser power of 17.43 W, pulse frequency of 292.29 kHz, and scanning speed of 1004.82 mm/s. Under these parameters, femtosecond laser polishing significantly reduced the surface roughness of the AlN ceramic, with the initial Ra value decreasing from 2.513 μm to 0.538 μm, a reduction of 78.57%. Compared to CO₂ laser polishing (<i>R</i>_a = 0.817 μm), femtosecond laser polishing demonstrated superior performance in enhancing surface quality. Analysis of the microstructural mechanisms revealed that the femtosecond laser, due to its ultra-short pulse characteristics, effectively suppressed the expansion of the heat-affected zone. It passivated surface microcracks through a photothermal ablation effect and reduced the thickness of the subsurface damage layer. Furthermore, the friction coefficient and wear rate of the polished samples decreased, indicating a significant improvement in wear resistance. On the numerical simulation front, a multi-physics model describing the interaction between the femtosecond laser and AlN ceramic was established based on the non-equilibrium two-temperature model (NTTM) coupled with solid mechanics. The key innovation of our model is the full coupling of heat transfer and solid mechanics, which allows for an accurate revelation of the material morphology evolution mechanism during femtosecond laser ablation. The model's accuracy is confirmed by the excellent agreement with experimental results, showing relative errors of only 3.23% and 12.5% for the melt pool width and depth, respectively.