Thin Film Roughness Optimization In The Tin Coatings Using Genetic Algorithms
Optimization is important to identify optimal parameters in many disciplines to achieve high quality products including optimization of thin film coating parameters. Manufacturing costs and customization of cutting tool properties are the two main issues in the process of Physical Vapour Deposition...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
JATIT & LLS
2017
|
| Subjects: | |
| Online Access: | http://eprints.utem.edu.my/id/eprint/20923/1/Thin%20Film%20Roughness%20Optimization%20In%20The%20Tin%20Coatings%20Using%20Genetic%20Algorithms.pdf http://eprints.utem.edu.my/id/eprint/20923/ http://www.jatit.org/volumes/Vol95No24/1Vol95No24.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Optimization is important to identify optimal parameters in many disciplines to achieve high quality products including optimization of thin film coating parameters. Manufacturing costs and customization of cutting tool properties are the two main issues in the process of Physical Vapour Deposition (PVD). The aim of this paper is to find the optimal parameters get better thin film roughness using PVD coating process. Three input parameters were selected to represent the solutions in the target data, namely Nitrogen gas pressure (N2), Argon gas pressure (Ar), and Turntable speed (TT), while the surface roughness was selected as an output response for the Titanium nitrite (TiN). Atomic Force Microscopy (AFM) equipment was used to characterize the coating roughness. In this study, an approach in modeling surface roughness of Titanium Nitrite (TiN) coating using Response Surface Method (RSM) has been implemented to obtain a proper output result. In order to represent the process variables and coating roughness, a quadratic polynomial model equation was developed. Genetic algorithms were used in the optimization work of the coating process to optimize the coating roughness parameters. Finally, to validate the developed model, actual data were conducted in different experimental run. In RSM validation phase, the actual surface roughness fell within 90% prediction interval (PI). The absolute range of residual errors (e) was very low less than 10 to indicate that the surface roughness could be accurately predicted by the model. In terms of optimization and reduction the experimental data, GAs could get the best lowest value for roughness compared to experimental data with reduction ratio of 46.75%. |
|---|