| Abstract [eng] |
The aim of this thesis was to determine and enhance the key variables of a laser ablation scanning method with the goal of minimizing the surface roughness produced. It overviews roughness indicators such as Ra and Sa, spectral analysis of surface topography using Fourier/PSD methods, and the impact of Fresnel reflection losses during laser-material interactions. The outcomes were achieved through the utilization of a numerical model based on MATLAB. Within this model, the process of crater development is explained by a connection established through experimentation between the laser fluence distribution and the resulting crater shape. Furthermore, the model takes into consideration adjustments in the actual fluence due to specific local conditions of incidence and Fresnel reflections. A sample of soda-lime glass with a thickness of 1 mm and no initial unevenness was employed. A wavelength of 257 nm was chosen, and the parameters of the laser were set based on a laser source known as PHAROS Yb:KGW. In the first part, with N = 3 repetitions and a 0° rotation angle, the pulse-to-pulse spacings Δx and Δy was varied. The 2D Fourier spectra showed that the dominant spatial-frequency maxima follow the relation f ≈ 1/Δ. As the spacing decreases, stronger crater overlap reduces the modulation contrast in one direction, leading to a quasi-one-dimensional spectral pattern. The observed harmonic components are associated with the nonlinear dependence of crater depth on absorbed fluence. In the second part, the spacing was fixed at Δx = Δy = 18 µm and random position errors were introduced by shifting each point in x and y following a Gaussian distribution. The results showed that increasing positional disorder weakens the periodic structure: discrete lattice maxima broaden, lose intensity, and gradually transform into a diffuse background. Higher frequency components disappear first, indicating that positioning accuracy is important for maintaining a predictable surface texture. In the third part, the dependence of surface roughness on the number of repetitions and pulse to-pulse spacing was investigated. The Sa value changed more rapidly at low repetition numbers but began to stabilize at approximately N = 15. Therefore, this number of layers was selected for the following optimization steps. The lowest roughness was obtained at Δx = Δy = 18 µm, where Sa = 104.87 nm, indicating an optimal crater overlap under the selected conditions In the fourth part, with Δx = Δy = 18 µm, random shifts limited to 1 µm, and N = 15 repetitions, Sa was evaluated as a function of the rotation angle between repetitions. The dependence was non-monotonic, and the minimum roughness was reached at 82°, where Sa = 90.19 nm. This confirms that the rotation angle is an important optimization parameter, as it changes crater overlap geometry and the superposition of surface relief components. In the fifth part, the influence of pulse lattice geometry on surface roughness was assessed. The square lattice used in previous simulations was replaced with a hexagonal arrangement, where every second row was shifted in the x direction by one beam radius. In this case, Sa = 150.82 nm was obtained, showing that the hexagonal lattice did not improve surface quality for this parameter set. Therefore, lattice geometry should be optimized together with pulse spacing, number of layers, positional shift, and rotation angle between repetitions. |
