Prediction of Weld Tensile-Shear Strength using ANN Based on the Weld Shape in Aluminum Alloy GMAW

Abstract

Weld shape and size generally determine the quality of gas metal arc welding. Auto parts manufacturers prescribe the size and shape of the weld because they can indicate the mechanical properties of the weld. It is impossible to evaluate the quality of all welds through destruction inspection. Therefore, research on welding quality inspection using laser vision sensors as a non-destructive inspection method is underway. Although the external profile of the weld can be measured using a laser vision sensor, studies to predict the weld strength are insufficient. In this study, an artificial neural network (ANN) model was developed to predict the welding strength of the lap-fillet weld of an aluminum alloy. Input date for weld size was obtained in two ways. In the first method, a bead profile was acquired using a laser vision sensor, whereas the size of the weld was obtained through the acquired bead profile. In the second method, the size of the weld was obtained directly from cross-section analysis. The output data on the strength of the weld was obtained through a tensile shear test. Two models for predicting the tensile shear strength based on ANN were developed. By predicting the tensile strength of both models, the average error rate was within 10%, but the prediction accuracy using the laser vision sensor was better than that of the cross-sectional method.

Application of Convolutional Neural Networks for Classifying Penetration Conditions in GMAW Processes Using STFT of Welding Data

“…Prior studies have primarily focused on predicting weld bead shape [13][14][15][16] and mechanical properties [17][18][19] using statistical analysis considering various welding conditions. Additionally, artificial neural networks (ANNs) have been employed to predict the morphology [13][14][15] and mechanical properties [18][19][20] of weld joints in response to changes in welding conditions. Additionally, various methods such as fuzzy logic, neurofuzzy, image processing, D-S evidence, and physical models have been used to predict welding quality [21].…”

“…Additionally, various methods such as fuzzy logic, neurofuzzy, image processing, D-S evidence, and physical models have been used to predict welding quality [21]. Due to the challenges associated with incorporating disturbances into GMAW processes, recent studies have explored ANN-based prediction of welding quality using diverse sensor-based data [20][21][22][23].…”