Optimization of experimental temperature measurement in GTAW process by means of DoE technique and computational modeling

The prediction of temperature distribution is a critical process in the study of welding thermal field. Therefore, the accuracy in temperature measurement is quite important to provide meaningful results and to establish the coupling among thermal field results and other studies in welding as mechanical and microstructural analyses. This work deals with statistical analysis and design of experiments (DoE) in order to establish an optimal experimental design for thermal history measurements in welding processes with thermocouples independent of base material. Calculations from theoretical correlations and computational modeling of heat flow were used to accomplish the present study. The Gas Tungsten Arc Welding process (GTAW) was used in the experiments along with plates of a structural steel (ASTM A 36) and a martensitic stainless steel (AISI CrMo 12-1). Four different parameters were studied related with the disposition, separation, depth, thermocouple arrangement and application of thermal paste. An optimal design was able to measure the maximum temperatures in adjacent zones to the weld bead, which are important for the study of welding thermal behavior.