Because more than 90% of the domestic manufacturing processes for the Liquefied Natural Gas (LNG) storage tanks rely on welding and processing technologies, the advancement of welding and processing technologies is directly connected to the productivity and therefore the advancement is critical to be competitive in the domestic shipbuilding industry. The welding technology using a laser light source is a more advanced technology than conventional arc welding in terms of workability, precision, and productivity. Although its application area is currently limited, this technology has been emerging as an important assembly tool in the manufacturing process of shipbuilding and offshore structures in the future. Because the LNG storage tank is a cryogenic structure, 9% nickel steel is widely used to manufacture the tank for both room temperature and low-temperature environments due to its excellent mechanical properties and fatigue strength. In terms of strength, 9% Ni steel is equivalent to 680 MPa-level high-tensile strength steel, and is usually used in applications where the operating temperature is below-150℃, such as LNG tanks with QT treatment. The 9% Ni steel has higher strength and better weldability than A5083-O aluminum alloy, has better impact toughness at cryogenic temperatures than SUS304L, and is economic. Therefore, 9% Ni steel is widely used to manufacture LNG tanks. Previous studies on the 9% Ni steel are based on butt welding, and research has been conducted according to the welding process. However, because 30-40% of LNG storage tanks are formed in a curved shape, research on the fillet welding process to overcome the limitations of butt welding has not been actively conducted to date. More specifically, research on the development of an algorithm for setting process variables, which is the core technology of fillet welding, needs to be conducted. Therefore, in this study, fiber laser welding, which is a fillet shape, is studied and performed using 9% Ni steel. The main objective of this study is to optimize the welding process variables by predicting weld properties. To derive the optimal process variables, the GBO algorithm was developed based on mathematical models. Finally, the developed algorithm showed an average error rate of 0.01831%, which ensures the high reliability of the optimal process variables.

No abstract

No abstract

No abstract

No abstract