Prediction of Surface Roughness in Functional Laser Surface Texturing Utilizing Machine Learning

Functional laser surface texturing (LST) arose in recent years as a very powerful tool for tailoring the properties of parts and components to their later application. As result, self-cleaning surfaces with an improved wettability, efficient engine optimized tribological properties, functional implants increased biocompatibility can be achieved today. However, increasing capabilities LST, prediction resulting becomes more important order reduce development time those functionalities. Consequently, advanced approaches laser-processed surfaces—the so-called predictive modelling—are required. This work introduces concept modelling respect LST by means direct writing (DLW). Fundamental concepts are presented employing machine learning approaches, theoretical concepts, statistical methods. The takes into consideration used parameters, analysis topographical, other process-relevant information predict roughness. For this purpose, two different algorithms, namely artificial neural network random forest, were trained experimental data stainless steel Stavax surfaces. Statistical results indicate that both models desired topography high accuracy, despite use small dataset training process. further optimize process regarding efficiency, overall throughput, outcomes.