Spot Welding Qualitative Analysis Based On Dynamic Resistance

Resistance spot welding which has lots of advantages has a important stage in industry segment of aerospace manufacture , automobile production and light industry. So it is meaningfull to monitor welding process. It has more important content to quality detect and control spot welding and research procedure analysis, production control.Recently, conventional quality detection and analysis technique is direct at single characteristic quantity generally, currently the common resistance spot welding quality monitoring technology contain constant current control method , inter electrode voltage method , energy monitor method and dynamic resistance method, and so on.But because of resistance spot welding is a process of electricity, hot, force, any factors such as welding current change, heat production and dissipate, magnitude of pressures and different kinds of non-controllable factor all can impact the welding quality of resistance spot welding. Because of the intricacy extent of spot welding process ,the monitor method can only provide a reliable quality information in a little condition, industrial production practice has proved that detection and analyses can not reflect much influencing factor in spot welding process for signal processing parameter, so the technology adopting single parameter to monitor and analyses is short sight and inaccuracy.First this dissertation analysis the parameter which affect the quality of resistance welding, include welding current, welding time, welding pressure,then, using the experimental data fit nugget size chart based on various processing parameter,at the same time,analysis the characteristic of dynamic resistance curve.at last it was impact stainless steel resistance electric welding important factor that ascertain welding current, weld time, assert welding current, welding time, dynamic resistance which all impact the quality of stainless steel resistance welding.Then aimed at the feature of spot welding process of existing non-linearity, multivariable action and much uncertain factors,Using Matlab consider much more important factors of stainless steel resistance spot welding and avoid only detecting signal parameter which result in incompletion and inaccuracy synthetically to build BP ANN(back propergation algorithm)artificial neural net ANN model. Training the BP ANN model using certain swatch, and then using the proven swatch to verify the model, to definite a ideal model based on multi- welding parameter finally.