| Resistance Spot Welding(RSW) is the major sheet-metal joining method for car body, but the quality uniformity of RSW is influenced by many factors in production site, which makes quality inspection an important task. Sampling inspection based on destructive or nondestructive techniques could not guaranteeing the quality of all welds on car body, thus makes the technic of RSW online monitoring and control, which is still to be realized, generally believed the most effective way to improve efficiency and quality of products. To realize 100% inspection of welds on car body, this dissertation takes process monitoring and quality evaluation technologies of Middle Frequency Direct Current(MFDC) RSW as the leading investigated subject. By improving the traditional measurement methods, new monitoring technologies for RSW are put forward, which have both high accuracy and good practicability. Different evaluation methods of quality uniformity is established according to production datas, which are collected by a monitor based on proposed technologies. On this basis, the unstable factor in RSW process is properly recognized, which provides new research ideas for quality monitor in RSW.Firstly, measurement methods of electrical signals in RSW process were investigated. By means of simulation and experiments, it was proved that the voltage signals collected from the transformer secondary include inductive component. The typical error was 35% when calculating the dynamic resistance based on RMS value algorithm in constant current mode because the influence of inductive component is not eliminated. This error could be much smaller when calculating the dynamic resistance based on mean value algorithm in constant current mode, but the typical error was about 300% in variable current mode. Accordingly, an iterative algorithm to calculate the value of resistance was presented: use Jacobi iterative method to solve the voltage balance equation and energy balance equation. The results showed that the iterative algorithm is valid in both the constant current mode and variable current mode. The calculated value of the resistance form the iterative algorithm is much closer to the real value, and the theoretical calculation error is no more than 0.075%.Next, measurement methods of mechanical signals in RSW process were investigated. Displacement of moving electorde was measured measurement with an optical encoder from the distal end of the servo cylinder, which totally eliminate the transmission error caused by pulley, and the measuremtn error was no more than ± 4 μm. Displacement of statice electorde was indirect measured by means of measuring the surface strain of statice arm, and the measuremtn error was no more than ± 4 μm. According to modelling and experiments, it was proved that displacement of static electrode makes up great proportion of the total electrode displacement. Then precision measurement method of electrode displacement from the distal end of the servo gun was put forward: measuring and integrating the displacements of both moving and static electrodes, and the theoretical measurement error of the electrode displacement was less than ± 8.8μm.Based on the research of measurement, a monitor of MFDC RSW with functions of data recording and network communication was fabricated and installed in the production filed. According to the monitoring data, it was found that although the electrical signals of welds with the same position in different parts have the same change tendency, there still exists some differences, of which the maximun is about 110%. On the top of this phenomenum, this thesis proposed an original quality monitoring technology of RSW, that is, to evaluate the quality uniformity through process stability. The off-line evaluation model of the overall welding process stability of welds with the same position was presented: calculate the coefficient of variation(COV) of the instantaneous value of the electrical signals, and weighted average value wase used as the stability index. The experiments showed this method could not noly accurately evaluate the qulality consistency of welds, but also be used to analyze the causes of the differences in stability. It was observed the major factor that lead unstability of weld process in procduction field was poor contact, and poor fitup and angular misalignment were direct causes.On this basis, this paper put forward an on-line stability evaluation method which aimed at the single welding process: the instantaneous average value of electric signals of welds with the same position was calculated and defined as "sample" curve, and the difference between the electric signal and the "sample" curve during a certain welding process was utilized as the stability index. Research results showed that stability index of a good weld follows a lognormal distribution., and a threshold value could be set as 20%(eμ+3ζ) according to 3-sigma principle. Thus, by comparing the on-line evaluation index and the threshold value of each weld, 100% on-line inspection of welds in car body could be realized. Moreover, the evaluation result was displayed in the form of ratio of suspicious welds, which is helpful to intuitively relate to the quality uniformity and recognize the unstable factor in RSW process, and has a strong guiding significance. |
PDF Full Text Request