AI Laser Welding Energy Input Prediction for Spatter Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Laser welding process optimization is hindered by the complexity of predicting weld seam quality due to unknown workpiece characteristics and material variations, leading to inefficient experimentation and high manufacturing costs, as existing models fail to accurately predict weld spatter formation.
Innovation Solution
A data-based model is trained to predict energy input into the workpiece using a combination of experimental and simulated data, employing Bayesian optimization and Gaussian processes to optimize process parameters, thereby reducing the number of required experiments and improving precision and productivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If process parameters are optimized using traditional experimental methods, then weld seam quality can be improved, but the number of required experiments and manufacturing costs increase significantly
Solution Approach 1:
The patent applies preliminary action by training a data-based model in advance using simulation data and limited experimental data. This pre-trained model can then predict weld seam characteristics for new process parameters without requiring extensive new experiments, thus reducing the time and cost of process optimization while maintaining prediction accuracy
Solution Approach 2:
The patent uses simulation data as a copy or surrogate for expensive and time-consuming physical experiments. The simulation model replicates the welding process behavior, allowing the data-based model to learn from simulated scenarios and reduce the number of actual experiments needed for optimization
2Manufacturing precision
If many process parameters are varied to find the optimum, then the achievable precision improves, but the complexity of the optimization process increases
Solution Approach 1:
The patent introduces a data-based model as an intermediary between process parameters and weld seam characteristics. This model acts as a mediator that captures the complex relationships between multiple parameters and outcomes, allowing optimization without directly managing the complexity of all parameter interactions through traditional experimental methods
Solution Approach 2:
The patent transforms the optimization problem from directly varying multiple physical process parameters to optimizing the data-based model's predictions. By changing the approach from physical parameter exploration to model-based prediction, the complexity of handling multiple interacting parameters is reduced while maintaining the ability to find optimal settings
3Ease of manufacture
If simplified models are used for prediction, then the ease of manufacture improves, but the reliability of quality prediction deteriorates
Solution Approach 1:
The patent creates a composite modeling approach by combining simulation data and experimental data to train the data-based model. This composite training data strategy allows the model to capture complex physical phenomena from simulations while being grounded in real-world measurements, achieving both simplicity and reliability
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
This approach enables efficient and targeted optimization of laser welding parameters, reducing experimental complexity and costs by accurately predicting weld seam characteristics and minimizing weld spatter, thus enhancing the precision and productivity of the laser welding process.
Implementation Method 1
A focused laser beam is applied to the workpieces to be connected. Due to the very high intensity, the absorbed laser energy results in very rapid local heating of the workpiece materials
Implementation Method 2
the absorbed laser energy results in very rapid local heating of the workpiece materials
Implementation Method 3
the process parameters may be selected in such a way that rapid and local heating of the materials by the laser radiation results in vaporization in the melt bath
Implementation Method 4
The molten material is expelled from the melt bath by the process-related explosively generated vapor pressure
Data Source
AI summary
A method for training a data-based model to ascertain an energy input of a laser welding machine into a workpiece as a function of operating parameters of the laser welding machine. The training is carried out as a function of an ascertained number of spatters.