Simulasi Dan Eksperimental Pengelasan dengan Pengembangan Subrutin dan Variasi Model Sumber Panas Pada Pengelasan Multi Pasa Gas Metal ARC Welding Material S235

Abstract

Welding is one of the five main metal working processes. This process is widely used to produce joints on certain metals that are used for household needs, construction to various needs in industries. The main problem that occurs in welding is the change in the microstructure between the weld area, the heat affected zone and the base metal due to the heat distribution during welding which will result in changes in mechanical properties. If this change in properties is too significant, then this welded joint will be dangerous during service. The purpose of this research is to obtain good welding results by carrying out simulations. The simulation was carried out using three heat source models and the development of a cooling rate subroutine is a novelty in this research. The cooling rate is needed in calculating the hardness of the welding results which is also carried out using a subroutine that has been used by previous researchers. The research was conducted using the same geometry in both simulation and experiment on structural steel material S235 with two types of welding, namely butt joint with 3 welding passes on material with dimensions of length 250 mm, width of each plate 100 mm, thickness 10 mm. The next material with bead on plate welding on material with dimensions of length 150 mm, width 100 mm, thickness 10 mm. The heat source models to be compared in the simulation is Goldak (double ellipsoid), Rectangular and Cylindrical Heat Source Model (HSM). Simulations are enhanced using subroutines to predict hardness in the HAZ and base metal areas as an impact of changes in mechanical properties due to the microstructural changes. After the two research methods were compared, the results showed that the Goldak Heat Source Model provided the best approach to the material hardness number approach with a difference of 0.31% ~ 8.30 % compared to the experiment. The best temperature distribution was obtained from Rectangular HSM with a difference of 4,27%.