Current manufacturing process controls are principally based only on statistical performance. The next evolution is to make physics based models combined with the state of the art sensors and actuators to control the manufacturing processes. In this paper, metal inert gas welding is used as an example of how the first steps in developing a reliable estimation technique to implement a physics based controller. The weld bead geometry will be the main focus because it is crucial to creating a quality weld. This paper uses an IR camera to generate and evaluate multiple weld bead width estimation techniques and characterizes their corresponding standard deviations. Also a Gaussian Mixture Model (GMM) is used to fit the temperature linescan data to fit an analytical function to the numerical data. The GMM is then used to estimate the weld bead width. Finally, the optimal linescan location is calculated to produce the best possible weld bead estimation. The result is that only one of the estimation techniques actually follows a step input and vi the optimal linescan location is 4 mm from the back of the arc. Furthermore, the GMM provides an excellent fit to the temperature linescan, but does not increase the accuracy of the estimate. / text
| Identifer | oai:union.ndltd.org:UTEXAS/oai:repositories.lib.utexas.edu:2152/ETD-UT-2009-08-257 |
| Date | 2009 August 1900 |
| Creators | Casey, Patrick John |
| Source Sets | University of Texas |
| Language | English |
| Detected Language | English |
| Type | thesis |
| Format | application/pdf |
Page generated in 0.0019 seconds