An Incompressible Smoothed Particle Hydrodynamics (ISPH) Model of Direct Laser Interference Patterning

Abstract

Functional surfaces characterised by periodic microstructures are sought in numerous technological applications. Direct laser interference patterning (DLIP) is a technique that allows the fabrication of microscopic periodic features on different materials, e.g., metals. The mechanisms effective during nanosecond pulsed DLIP of metal surfaces are not yet fully understood. In the present investigation, the heat transfer and fluid flow occurring in the metal substrate during the DLIP process are simulated using a smoothed particle hydrodynamics (SPH) methodology. The melt pool convection, driven by surface tension gradients constituting shear stresses according to the Marangoni boundary condition, is solved by an incompressible SPH (ISPH) method. The DLIP simulations reveal a distinct behaviour of the considered substrate materials stainless steel and high-purity aluminium. In particular, the aluminium substrate exhibits a considerably deeper melt pool and remarkable velocity magnitudes of the thermocapillary flow during the patterning process. On the other hand, convection is less pronounced in the processing of stainless steel, whereas the surface temperature is consistently higher. Marangoni convection is therefore a conceivable effective mechanism in the structuring of aluminium at moderate fluences. The different character of the melt pool flow during DLIP of stainless steel and aluminium is confirmed by experimental observations.

Influence of Sulphur Content on Structuring Dynamics during Nanosecond Pulsed Direct Laser Interference Patterning

“…The evolution of the maximum surface temperature predicted in the simulation for the moderate laser fluence 2Φ 0 = 0.532 J/cm 2 is presented in Figure 9 along with the temporal variation of the laser pulse intensity. The maximum surface temperature trend in Figure 9 is in line with earlier simulation results in [44], although T max ∼ 2570 K is lower here, for DLIP of stainless steel at a laser wavelength of λ = 355 nm and a lower fluence of 2Φ 0 = 0.4 J/cm 2 , which is attributed to the higher reflectivity at the present wavelength. It is evident from Figure 9 that the substrate surface is heated up to a temperature significantly above the liquidus point at the interference maximum due to the action of the laser pulse.…”

“…Concerning the trends in Figure 10, the discrete values of the melt pool dimensions are employed only once at a central point in time to avoid a stair-step appearance of the graphs. The melt pool depth and the duration of the melt presence are compatible with the aforementioned numerical results in [44], whereas the melt pool width in Figure 10 exceeds the earlier calculations owing to the larger periodicity Λ of the interference pattern considered here. Furthermore, simulations were performed for an elevated laser fluence 2Φ 0 = 0.665 J/cm 2 to investigate the influence of the laser fluence on the material behaviour, in particular the nonlinear effect on the melt pool dimensions and the velocity field, and the structure formation reported in Section 3.1.…”

“…Consequently, the evolution of the maximum surface temperature computed by the SPH model is also shown in Figure 9. The temporal maximum of the surface temperature T max ∼ 3150 K at the interference maximum is below the vapourisation point, unlike the simulation results in [44] for the comparable fluence 2Φ 0 = 0.5 J/cm 2 and the laser wavelength λ = 355 nm with a lower reflectivity of the stainless steel substrate. These differences from the earlier results are attributed to the slightly larger width of the laser pulse considered in the present investigation.…”

“…In addition to the velocity magnitudes shown in Figure 11a for the moderate fluence 2Φ 0 = 0.532 J/cm 2 , Figure 12a illustrates the strong outward convection for a low sulphur content of 30 ppm, which gradually decreases and is dominated by the inward convection for higher sulphur concentrations of 100 ppm and 300 ppm. Compared with the results in [44], lower horizontal velocity magnitudes are obtained in this work, which is attributed to the generally smaller magnitude of the surface tension temperature coefficient in the presence of a surfactant and the lower surface temperature gradients due to the larger periodicity Λ and longer pulse duration τ p .…”

“…The mathematical model of the material behaviour during single pulse DLIP was already described in detail by Demuth and Lasagni [ 44 ]. Here, the nonlinear temperature dependence of surface tension considered is presented along with the dimensionless form of the governing equations and the Marangoni boundary condition.…”