研究目的
To minimize the distortions of a complex frame structure with multiple welds by using a meta-model based on an Artificial Neural Network and a genetic algorithm to efficiently find suitable welding parameters.
研究成果
The method reliably identifies optimized welding parameters for minimizing distortions in complex structures with multiple welds. The combination of metamodeling based on Artificial Neural Networks and optimization by means of a genetic algorithm is effective in handling the complexity resulting from a multitude of welding possibilities.
研究不足
The model does not account for phase transitions and preloading due to the production history. The accuracy could be improved by refining the FE model in terms of temperature-dependent and strain rate-dependent material properties.
1：Experimental Design and Method Selection:
The method involves dividing the global structure into sub-areas and using a meta-model to predict local distortions. A genetic algorithm is then used to find optimized welding parameters.
2：Sample Selection and Data Sources:
The study uses an exemplary frame structure with 7 welds, including butt welds and lap joints.
3：List of Experimental Equipment and Materials:
The study uses laser beam welding on EN AW-6082 aluminum sheets.
4：Experimental Procedures and Operational Workflow:
The process involves thermal FE simulation, automated heat source calibration, thermo-mechanical calculation of component distortions, and optimization using a genetic algorithm.
5：Data Analysis Methods:
The study uses Artificial Neural Networks for metamodeling and a genetic algorithm for optimization.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
