| CMT(Cold Metal Transfer) is a new method of melting polar gas shielded welding developed by Austria Fronius Company based on the SPAZ and micro connection technology. CMT can be used to weld thin sheet with 0.3mm thickness due to the characteristics of alternating hot and cold. The birth of CMT welding process has broken the restricted zone of sheet metal welding.In this paper, the classification and characteristics of general heat source model are introduced at first according to the theory of finite element and the welding heat transfer. The finite element model of CMT welding is established by using the FEA software ANSYS 12.1. The Gaussian heat source model is selected as the numerical simulation model and the model is processed for CMT welding "cold-hot-cold" alternate characteristics. The heat load is assumed by 0 in the schedule of a certain time using APDL language. After lots of simulation comparisons with experimental data, it is specified that the more accurate results can be got when the intermittent period is 30 percent of every cycle time.Then the three-dimensional transient temperature field of CMT welding and molten pool is studied and the distribution of temperature field and the variation of the molten pool shape with time are analyzed. And according to the characteristics of the CMT combination expert system, welding current and welding speed are selected as welding parameters of the numerical simulation and the influence of different parameters on the CMT welding temperature field distribution and the shape of molten pool is studied. By using orthogonal experiment method to test, the welding current, welding speed, welding wire stem elongation, nozzle height and gas flow are selected as the influencing factors, and the five factors three levels orthogonal table is identified through the comparison of orthogonal experiment results with the result of numerical simulation. The results show that the CMT temperature field distribution can relatively accurately be described by numerical simulation and the correct welding joint size can be got.Finally, the welding joint size is processed by multiple linear regressions using statistical software Stata 12.0 with five welding parameters as independent variables and molten width, depth, surplus height as dependent variables, which is regarded the orthogonal test data as samples. The results show that the prediction precision of stepwise regression model is relatively better. Eventually the multiple linear regression equations of molten width and depth are listed and the mathematical model of welding joint shape of CMT is built. It has important significance for improving the accuracy of welding joint geometry prediction. |
