Enhancing CO2 Laser Cutting Efficiency for Diverse Wood Species Using Artificial Neural Networks

Modern manufacturing technologies include wood processing using laser technologies. The most used laser for wood cutting is the CO2 laser, which offers many advantages such as processing speed, efficiency, and minimal impact on the material’s structure after cutting. To achieve a high-quality cut, characterized by the cutting kerf parameters, it is necessary to know the appropriate combination of cutting parameters, primarily laser power (P) and the cutting speed (v). Therefore, this article investigates the effect of P, v, and cutting direction on the cutting kerf widths on the upper surface WKU, lower surface (WKL), and their ratio (WKR). The analysis was performed on samples of spruce, beech, and oak wood, while also evaluating the influence of the anatomical cutting direction. The correlation coefficient between the predicted values and the measured values is at least 0.94, with the mean square error not exceeding 4%. Consequently, the employed models demonstrate validity in predicting cutting kerf widths and optimizing the cutting process based on the type of timber, cutting direction, and the specified laser parameters.