Date of Award

2014

Publication Type

Master Thesis

Degree Name

M.A.Sc.

Department

Mechanical, Automotive, and Materials Engineering

Keywords

Artificial Neural Networks, Design of Experiments, K- Mean Classification, Laser Cladding, Response Surface Methodology, Sensitivity Approach

Supervisor

Urbanic, Ruth

Rights

info:eu-repo/semantics/openAccess

Abstract

Laser cladding is an additive manufacturing technique involving deposition of powdered clad metal in successive 2D layers onto a substrate thereby creating surface coatings with enhanced material properties. Process and shape parameters contribute in defining the geometry of the clad bead; however, due to the highly coupled nature of the process, it is difficult to determine the relationship between parameters. This research predicts such parameters through development of a cognitive artificial intelligence system using artificial neural networks. A robust experimentation design process applying response surface methodology technique is adopted to collect the bead geometry data for various process configurations. Furthermore, the research identifies the extent of contribution of each factor and the impact of their interactions on the model output through ANOVA and sensitivity analysis. Lastly, a K-mean clustering algorithm is incorporated to identify optimal number of clusters present in the collected dataset on the basis of bead shape characteristics.

Share

COinS