Term of Award

Fall 2020

Degree Name

Master of Science, Mechanical Engineering

Document Type and Release Option

Thesis (open access)

Copyright Statement / License for Reuse

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Department

Department of Mechanical Engineering

Committee Chair

Bishal Silwal

Committee Member 1

Marcel Ilie

Committee Member 2

Hossein Taheri

Abstract

Wire arc additive manufacturing is a process of making three-dimensional metal parts in a layer-by-layer approach using a feed wire and electric arc as a heat source. Wire arc additive manufacturing (WAAM) is becoming more popular due to its ability to create complex 3D parts, less build time, high deposition rate, and significant cost savings. Out of the many parameters used in WAAM, one of the important parameters is shielding gas which plays a significant role in material quality, properties, and defects. In this study, a controlled amount of Argon (Ar) and Nitrogen (N2) shielding gases are used to see the effect on the weld bead depth and width. In addition, a computational fluid dynamics (CFD) model is used to perform numerical analysis. The data collected from the experiment is used to perform a regression analysis to predict future values. The amount of shielding gas mixture is controlled through a flowmeter to three different total flowrates. The result shows there is an increase in the depth and width of the weld bead with the increase in N2 percentage in the Ar-N2 shielding gas mixture. With the increase in Nitrogen percentage, the tungsten arc is observed unstable and spattering is noticed. The temperature of the surface of the base metal is increased while using the Ar-N2 mixture. The experiment result is further verified by developing and analyzing a three-dimensional computational fluid dynamics model using a volume of fluid (VOF) method. Support vector machine (SVM) regression model with Gaussian kernel function is used to perform the predictive regression analysis. Out of all the regression models, SVM has the lowest model loss for the collected data.

OCLC Number

1239951098

Research Data and Supplementary Material

No

Download

Share

COinS