Resistance spot welding is an established joining process for the production of safety-relevant components in the automotive industry. Therefore, consecutive process monitoring is essential to meet the high quality requirements. Artificial neural networks can be used to evaluate the process parameters and signals, to ensure individual spot weld quality. The predictive accuracy of such algorithms depends on the provided training data set, and the prediction of untrained data is challenging. The aim of this paper was to investigate the extrapolation capability of a multi-layer perceptron model. That means, the predictive performance of the model was tested with data that clearly differed from the training data in terms of material and coating composition. Therefore, three multi-layer perceptron regression models were implemented to predict the nugget diameter from process data. The three models were able to predict the training datasets very well. The models, which were provided with features from the dynamic resistance curve predicted the new dataset better than the model with only process parameters. This study shows the beneficial influence of process signals on the predictive accuracy and robustness of artificial neural network algorithms. Especially, when predicting a data set from outside of the training space.

No abstract

The application of anti-corrosion coated, high-strength steels in the automotive industry has increased in recent years. In combination with various zinc-based surface coatings, liquid metal embrittlement cracking can be observed in some of these materials. A high-quality, crack-free spot-welded joint is essential to realize the lightweight potential of the materials. In this work, the LME susceptibility of different coatings, which will be determined by the crack length and the occurrence rate, will be investigated using a welding under external load setup. The uncoated specimens did not show any LME. EG, GI and GA showed significantly less LME than ZM coatings. The latter coatings showed much larger crack lengths than the EG, GI and GA coatings. Furthermore, two mechanisms regarding the LME occurrence rate were observed: the occurrence of LME in zinc–magnesium coatings was theorized to be driven by the material properties of the coatings, whereas the occurrence of LME at EG, GI and GA samples was forced mainly by the application of the external tensile load. In the experimental setup of this work, the materials were exposed to unusually high mechanical loads (up to 80% of their yield strength) to evoke LME cracks.

Resistance spot welding is an established joining process in the production of safety-relevant components in the automotive industry. Therefore, a consecutive process monitoring is essential to meet the high-quality requirements. Artificial neural networks can be used to evaluate the process parameters and signals to ensure the individual spot weld quality. The predictive accuracy of such algorithms depends on the provided training data set and the prediction of untrained data is challenging. The aim of this paper is to investigate the extrapolation capability of the multi-layer perceptron model. That means, that the predictive performance of the model will be tested with data that clearly differs from the training data in terms of material and coating composition. Therefore, three multi-layer perceptron regression models were implemented to predict the nugget diameter from process data. The three models were able to predict the trained datasets very well. The models, which were provided with features from the dynamic resistance curve predicted the new dataset better than the model with only process parameters. This study shows the beneficial influence of the process signals on the predictive accuracy and robustness of artificial neural network algorithms. Especially, when predicting a data set from outside of the training space.

No abstract