Developing technological foundations spot welding dissimilar materials with guaranteed quality connections
The essence of development is that for the first time in spot welding of dissimilar materials to enhance quality connections and guaranteed their support for a given level of applied external electromagnetic action. Their essence is that during the welding process each point in the working area generated transverse inter-symmetric magnetic field. In their interaction with the welding current in the melt volume arising ponderomotive force that it intensively stirred. Thus there is some positive effects. Since the early stages of forming compounds reduced the formation of the desired area of contact between the welded materials and as a result, significantly increases the intensity energy input. In the process of compound dilution flows hot melt its borders causes reduction in time to the formation of a given volume of drink area. As a result, to decrease to 30% of energy consumption process. In addition, as a result of the mixing of the melt is its homogenization, as evidenced by the reduction of chemical microinhomogenities. As a result of improvements crystallization process improves the mechanical properties of the compounds decreases their microhardness. Thus, the application of the developed technological measures can guarantee the quality of welded joints.
Developed one-parametric system monitoring the process based on the analysis methods of Artificial Intelligence in welding current waveform for each point. Assessment forms pulse welding current conducted by its first derivative. For increased effectiveness estimates used discrete wavelet analysis to examine the signal as a function of time in terms of vibrations, localized in time and frequency. The analysis was performed using Daubechies wavelets. Identifying the differences between the current shape of the curve under the influence of disturbances on the process carried out by comparing the signal components certain level, reduced by detailing coefficients of wavelet decomposition. Visible differences are localized mainly in the area of attenuation pulse. The difference in levels of detailing component wavelet transform depends on the type of perturbation. To automate the monitoring process used forecasting system developed as a connection based on artificial neural networks.
Вложение | Размер |
---|---|
![]() | 53.5 КБ |