Quality Prediction and Control in Wire Arc Additive Manufacturing via Novel Machine Learning Framework

Area of Science:

• Materials Science
• Manufacturing Engineering
• Computational Science

Background:

• Wire Arc Additive Manufacturing (WAAM) enables rapid fabrication of large metal components but struggles with precise control over deposited shape due to process variability.
• Existing data-driven models (e.g., ANNs) predict deposition dimensions but cannot inversely determine optimal parameters for desired shapes.
• Intercorrelated process variables and codependent deposition outcomes (width, height, penetration) limit the reliability of current predictive models.

Purpose of the Study:

• To develop a novel machine learning framework for quantitatively analyzing the relationship between WAAM process parameters and deposition geometry.
• To enable systematic and quantitative prediction of deposition shape, moving beyond qualitative assessments.
• To provide a reliable method for selecting optimal process parameters to achieve desired deposition dimensions and improve part functionality.

Main Methods:

• Development of a machine learning framework to analyze correlations between WAAM process parameters and deposition shape.
• Quantitative prediction of bead width, height, and depth of penetration.
• Inverse determination of optimal process parameters for achieving target deposition geometries.

Main Results:

• The proposed framework quantitatively predicts deposition shape with high accuracy.
• It establishes complex process-quality relationships, enabling better control over the WAAM process.
• Experimental validation confirms the framework's effectiveness in achieving desired deposition geometries.

Conclusions:

• The novel machine learning framework offers a significant advancement in controlling WAAM deposition geometry.
• It facilitates the selection of optimal process parameters for near-net-shape fabrication, reducing post-processing needs.
• This approach enhances the predictability and reliability of WAAM for functional component manufacturing.