Search

The need for an efficient welding process which is fast, safe, cost-effective and will produce a welded structure that combines high strength with integrity during railcar manufacturing is necessary due to the increasing demand and complexity of the manufacturing process. In this study, the Taguchi method and response surface methodology (RSM) were used for the numerical experimentation design of a railcar bogie frame. The modelling and simulation of the railcar bogie frame was conducted using the Commercial Abaqus software 2018. Using the Taguchi method, the designed experiment consists of an orthogonal array of four factors varied over three levels with the welding parameters in the following ranges: voltage (22–25 V), current (200–250 A), speed (0.01 m/s) and arc length (0.03–0.07 m) as input variables. The matrix generated nine experimental runs whose weld distortions were determined via the physical experimentations. Also, the RSM was used to study the cross effect of the input welding parameters on the welded structure. The statistical analysis of the results obtained brought about the development of a predictive model for the determination of weld distortion as function of the independent welding input parameters. In addition, the results indicate that the input welding parameters are critical factors that affect welding processes significantly, hence the need for control. This work will assist manufacturers in the quest for effective process control during the welding operation. [ABSTRACT FROM AUTHOR]

Arc welding can be used to weld the components of rail car brackets during assembly operation. In this study, the welding parameters, namely voltage, current, speed, and arc length, were optimized using the Taguchi method and response surface methodology (RSM). A predictive model was thereafter obtained that correlates the weld distortion and hardness as a function of the welding parameters. The experimental design which consists of Taguchi orthogonal array of L9, three level-four factor matrix was used as input parameters to determine the weld distortion and hardness while the RSM was used to study the cross effect (interaction) of various process parameters on the weld distortion. Analysis of variance (ANOVA) was used to further test and validate the developed model for adequacy and the regression model was found to be highly significant at 95% confidence level as correlation coefficients were very close to 1. The results obtained also show high degree of convergence of the welded components with changes in welding parameters indicating that weld distortion is kept within the permissible range. This will reduce the production cost and increase the quality as well as the reliability of weld during welding processes. [ABSTRACT FROM AUTHOR]