Researchers have made significant strides in optimizing the manufacturing process of a β-Titanium alloy produced through direct energy deposition. This innovative technique, which involves layer-by-layer manufacturing using a laser beam to melt metal powder, offers high deposition rates and improved feature resolution compared to other methods. However, the variability in print quality and mechanical properties due to numerous process parameters poses a challenge.
One key parameter that researchers often consider is the global energy density (GED), which is a measure of input energy over a surface area. By breaking down GED into its constituent parameters such as laser power, scanning speed, and spot size, researchers can better understand how each parameter influences critical aspects like layer height and grain size in the alloy.
The deposition rate, layer height, and microstructure of the alloy are all influenced by the process parameters, highlighting the need for optimization strategies. Through a central composite design of experiments, researchers were able to create a process map for the alloy by varying key parameters like laser power, scanning speed, and spot size. Machine learning models, including multi linear regression and artificial neural networks, were then employed to predict the effects of these parameters on the alloy properties.
While machine learning has shown promise in optimizing additive manufacturing processes, the challenge lies in balancing accuracy and efficiency. By combining experimental data with predictive models, researchers aim to streamline the optimization process and gain better control over the solidification process to enhance grain size variation.
Overall, this research offers a material-specific modeling approach that could revolutionize the optimization of additive manufacturing processes. By leveraging cutting-edge technologies and in-depth data analysis, researchers are paving the way for more efficient and cost-effective manufacturing techniques in the future.
