WAAM Shape Compensation Using Machine Learning Deformation Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wire and arc additive manufacturing (WAAM) technologies face significant challenges with large shape deviations and high surface roughness due to nonlinear heat stresses and residual stresses, making comprehensive process characterization costly and time-consuming, especially for large parts with frequent design changes.
Innovation Solution
A method involving a computing device that converts point-cloud data from training objects into functional data, calculates shape deviations, constructs an engineering-informed tensor-product basis representation of deformation and roughness patterns, and creates an optimal compensation plan to minimize shape deformation during WAAM fabrication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If comprehensive process characterization is conducted to optimize deposition parameters, then manufacturing precision is improved, but loss of time and loss of substance increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-characterizing materials and geometrical features through experiments before actual production. The system stores deposition parameters, shape deviations, and process data for various materials and geometries in advance, allowing rapid retrieval and application without conducting full characterization during production, thus reducing time loss while maintaining precision.
Solution Approach 2:
The patent uses copying by creating a digital twin or virtual model of the physical WAAM process. The system replicates the complex thermal-stress behavior and shape deviations in a computational environment, allowing virtual experimentation and parameter optimization without physical trials, thereby reducing both time and material consumption while achieving high manufacturing precision.
2Manufacturing precision
If physics-based simulation is used to predict deformation, then manufacturing precision is improved, but device complexity and loss of time increase due to computational requirements
Solution Approach 1:
The patent applies segmentation by dividing the complex physics-based simulation into modular components. The system separates material property modeling, thermal field simulation, stress analysis, and shape deviation prediction into independent modules. Each module can be developed, validated, and optimized separately, reducing overall system complexity while maintaining high deformation prediction accuracy through coordinated integration.
3Productivity
If WAAM is used for large-size part production, then productivity is improved, but manufacturing precision deteriorates due to large shape deviations
Solution Approach 1:
The patent implements feedback by integrating in-situ sensing systems that continuously monitor shape deviations during WAAM fabrication. The measured deviations are fed back to the control system, which automatically adjusts deposition parameters such as heat input, deposition speed, and toolpath to compensate for emerging shape errors, thereby maintaining high manufacturing precision while preserving the high productivity of WAAM for large-size parts.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method significantly reduces shape deviations in WAAM parts by predicting and compensating for deformation patterns, improving geometric accuracy and reducing the need for extensive physical experimentation.
Implementation Method 1
the filler material is melted with an electrical arc between the wire electrode and the top layer surface
Implementation Method 2
the generated heat stresses due to the multiple fusion, solidification and phase change cycles
Data Source
AI summary
A generalized additive modeling approach to separate global geometric shape deformation from surface roughness is provided. Under this statistical framework, tensor product basis expansion is adopted to learn both the low-order shape deformation and high-order roughness patterns. The established predictive model enables the optimal geometric compensation for product redesign to reduce shape deformation from the target geometry without altering process parameters. Experimental validation on WAAM manufactured cylindrical walls of various radi shows the effectiveness of the proposed framework.