Vision-based localization for robot-CNC hybrid manufacturing

Thumbnail Image
Files
GOODWIN-THESIS-2022.pdf (10.73 MB)
Author(s)
Goodwin, Jesse
Advisor(s)
Editor(s)
Associated Organization(s)
Organizational Unit
Organizational Unit
Series
Collections
Theses and Dissertations
Supplementary to:
Permanent Link
https://hdl.handle.net/1853/72471
Abstract
Wire arc additive manufacturing (WAAM) has shown promise in recent years for producing large scale parts with higher deposition rates than other additive processes. WAAM is often combined with subtractive machining to form a hybrid manufacturing process. This hybrid process can be realized by retrofitting Computer Numerical Control (CNC) machines with deposition heads, adding spindles and deposition heads to robots, or developing part localization methods to transfer parts from an additive cell to a CNC machine. Here, a novel, robot-CNC hybrid configuration is introduced where a maneuverable robot is placed in front of a CNC machine to deposit material within the machine envelop. This method removes the need for part localization and the extensive machine modifications required for retrofitting; however, the problem of robot localization is also added. In this work, the effects of error in vision-based, contactless robot localization on machining parameters in a robot-machine hybrid process were studied. Performance was characterized on an implementation of this system using classical computer vision techniques. In addition, machining simulations were conducted to evaluate the effects of image-induced error on chip thickness, material removal rate, and machining allowance. Initial tests showed that computer vision could adequately locate a robot for the hybrid WAAM process without exceeding machining constraints.
Sponsor
Date
2022-05-04
Extent
Resource Type
Text
Resource Subtype
Thesis
Rights Statement
Rights URI