Statistical Analysis and Machine Learning-based Modelling of Kerf width in CO<sub>2</sub> Laser Cutting of PMMA

Recently, engineering polymers like PMMA have increasingly replaced traditional materials in industry where feasible, with CO2 laser cutting gaining attention for its high quality and speed in processing these materials. Achieving precise cuts is crucial for product accuracy, with kerf width serving as a key quality attribute to ensure quality and functionality of the final product. This study focuses on the im-pact of three critical process variables: stand-off distance, laser power, and cutting speed, on the kerf width in CO2 laser cutting of PMMA. Through a full-factorial experiment, the process parameters are systematically varied to understand their individual and interaction effects on the cutting process. The kerf width is measured as an indicator of precision using an optical microscope to evaluate the quality of the laser cuts. To address the non-linear relationships between these process parameters and kerf width, several machine learning models were utilized. Performance comparisons indicated that the Artificial Neural Network (ANN) model provided the highest accuracy, with R² values of 0.98 for the validation dataset and 0.95 for the testing dataset. The optimized ANN model offers a robust tool for parameter optimization, facilitating the determination of optimal settings to achieve the desired kerf width while ensuring productivity.