Related Experiment Video
Updated: Jun 27, 2026
07:18
Generating Lap Joints Via Friction Stir Spot Welding on DP780 Steel
Published on: August 13, 2019
Process Optimization and Microstructure in High-Speed Coaxial Dual-Laser Welding of SUS301 Thin Sheets Using an
Dexi Wang1, Nan Li1, Xiaohong Yan2
1School of Materials Science and Engineering, Dalian University of Technology, Dalian 116024, China.
Materials (Basel, Switzerland)
|June 26, 2026
Summary
A novel SSA-BP neural network accurately predicts weld geometry in SUS301 stainless steel, linking parameters to strength. This method optimizes welding for improved mechanical properties and microstructure control.
Area of Science:
- Materials Science and Engineering
- Manufacturing Processes
- Computational Modeling
Background:
- Predicting weld geometry and understanding structure-property relationships are crucial for optimizing welding processes.
- High-speed coaxial dual-laser butt welding of thin stainless steel sheets presents complex parameter-morphology correlations.
Purpose of the Study:
- To develop a predictive model for weld geometry in SUS301 stainless steel laser welding.
- To establish structure-property relationships by correlating weld morphology with mechanical performance.
- To optimize welding parameters for enhanced tensile strength and stable weld formation.
Main Methods:
- An SSA-BP (Slime Mould Algorithm-Backpropagation) neural network model was developed to correlate welding parameters (laser power, speed, pulse frequency, pulse width) with weld geometry (width, penetration depth).
- Model validation involved comparison with a conventional BP model and five-fold cross-validation.
- Weld geometry was analyzed using the aspect ratio (Φ = h/w), and tensile strength was evaluated. Microstructural analysis included fractography and Electron Backscatter Diffraction (EBSD).
Main Results:
- The SSA-BP model demonstrated high prediction accuracy (R=0.960) with significantly reduced errors compared to the conventional BP model.
- Optimized weld aspect ratio (Φ = 0.82-0.84) yielded higher ultimate tensile strength (1211.4-1264.8 MPa) and stable weld formation.
- Microstructural analysis revealed ductile fracture characteristics and a graded microstructure with varying grain boundary fractions and localized strain concentrations.
Conclusions:
- The SSA-BP model provides an effective approach for predicting weld geometry and guiding parameter selection in SUS301 laser welding.
- A specific weld aspect ratio range is identified as favorable for achieving superior tensile properties and weld stability.
- The study successfully links weld morphology prediction with tensile response and microstructural heterogeneity, offering insights for advanced materials processing.
