Maintaining high-quality welded connections is crucial in many industries. One of the challenges is assessing the mechanical properties of a joint during its production phase. Currently, in industrial practice, this occurs through NDT (non-destructive testing) conducted after the production process. This article proposes the use of a virtual sensor, which, based on temperature distributions observed on the joint surface during the welding process, allows for the determination of hardness distribution across the cross-section of a joint. Welding trials were conducted with temperature recording, hardness measurements were taken, and then, neural networks with different hyperparameters were tested and evaluated. As a basis for developing a virtual sensor, LSTM networks were utilized, which can be applied to time series prediction, as in the analyzed case of hardness value sequences across the cross-section of a welded joint. Through the analysis of the obtained results, it was determined that the developed virtual sensor can be applied to predict global temperature changes in the weld area, in terms of both its value and geometry changes, with the mean average error being less than 20 HV (mean for model ~35 HV). However, in its current form, predicting local hardness disturbances resulting from process instabilities and defects is not feasible.

Assessing the temperature of the joint in on-line mode is a vital task that is demanded to characterize the formations of terns formations that are taking place in a joint and result in reaching necessary properties of the joint. Arc welding generates a high amount of heat that is reflected by the metallic surface of the welded object. In the paper, a temperature measurement credibility increase method is described and evaluated. The proposed method is used to reduce the influence of the reflected temperature of the hot torch and the arc on the temperature distribution observed on the surface of the welded joint using an infrared camera. The elaborated approach is based on comparison between infrared observation of the solidifying weld and precisely performed finite element method (FEM) simulation. The FEM simulations were calibrated according to the geometry of the fusion zone. It allows to precisely model heat source properties. The best-reflected temperature correction map was selected and applied to obtain a temperature representation that differs from the FEM baseline by less than 10 °C. Precise temperature values allowed us to cluster welded joints in 3D feature space (temperature, hardness, linear energy). It was found that by using the k-means clustering method it is possible to distinguish between correct and faulty (in terms of too low mechanical properties) joints.