The material of this study is VDM Alloy 780, which is based on alloy 718. The differences are in the chemical composition, which makes the newly developed superalloy more expensive because of a higher amount of Cobalt. In exchange for a high price, this alloy showshigher serviceability at temperatures up to 750 °C. Longer performance under operation conditions makes this alloy attractive to do more research to use in different applications in the hot structural part of the aerospace industry.

The used process to join small parts of VDM Alloy 780 is keyhole TIG (or GTAW)welding, which is a patent to develop traditional TIG and gives some benefits over it. The reason for using this method to weld is to prove the good penetration and fewer defects of it which is not popular yet. Additionally, comparing the results of this process with the laser process in the same thicknesses of this material (in another master’s thesis).

The final aim is to show the weldability perspective of this alloy and compare the results with standard grade alloy 718. All the welding parameters that can affect the responses should be considered to be optimized for specific thickness. In this case, current and travel speed are factors, and welding geometry and metallurgical defects are considered according to AWS (American welding society) plus one other reference to measure the weld width. The responses are Porosity percentage, Average diameter of pores, Face undercut, Excess face weld, Excess root weld, and Minimum weld width.

Before welding, a full factorial with two levels design of experiment (DOE) is used to specify the parameters in each run in the initial step with a linear regression model by MODDE software. After welding, the metallurgical responses are measured by ImageJ software, and geometry responses are measured by Zeiss Axio microscope software. Then, using MODDE software, statistical data analysis is utilized to show how different variables and their interactions are connected to geometry and metallurgical defects, with the goal of optimizing the parameters to minimize defect density.

The results showed all the runs met the requirements except for the acceptable limits of the average diameter of pores. Moreover, a prediction of process parameters is presented. This prediction window suggests the best range of inputs to achieve the desired output for each response and also a combination of responses.