Process Optimization to Control Welding Distortion of High Speed Train Carriage by High Efficiency Numerical Simulation

ABSTRACT The residual stresses and welding distortion of large welded structure are more serious due to the long welded-lines. Numerical simulation is a powerful tool in residual stresses and distortion control. But long computational time for thermo-mechanical simulation of welding process hinders the application of this technique in large welded structure. During manufacturing aluminum alloy train carriage, the welding distortion of carriage roof, which is 12m long and with 12 welds, is an important factor resulting in assembly mismatch. A high efficiency numerical simulation method, temperature function method, was applied in the present study to shorten the computing time. The quasi-steady state character of welding process could be exploited in temperature function method. Also the temperature and temperature history dependent material property models were established and applied. The constraint of fixture on structure was also considered. The numerical simulation results corresponded well with the measured distortions. The errors between computed results and the experimental results were no more than 15%. This indicates that the numerical simulation model is not only high efficiency, but also accurate for the large and complex structure. By means of experiments and FEA study, it was found that the main welding distortion pattern is the transverse distortion across the cross-section of carriage roof structure. The cross-section of carriage roof was optimized based on numerical simulation. The simulation results showed that the distortion of structure after optimization was reduced to 20% of that of original structure. Structure with the same cross-section as optimized has been produced and its distortion was measured. The measurement results proved the effectiveness of this high efficiency numerical simulation.

KEY WORDS: welding distortion；high-efficiency numerical simulation；constitutive model; carriage roof; process optimization