Time-frequency representation as a tool of failure detection

Chemical-Mechanical Planarization, or CMP for short, is one of the final major process involved in semiconductor company when making a microchip. In this process the chip is being made perfectly smooth and flat with the help of chemical and mechanical forces. The tool which does the planarizing proc...

Full description

Saved in:
Bibliographic Details
Main Author: Soeryanto, Yehezkiel
Other Authors: Tegoeh Tjahjowidodo
Format: Final Year Project
Language:English
Published: 2019
Subjects:
Online Access:http://hdl.handle.net/10356/78702
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-78702
record_format dspace
spelling sg-ntu-dr.10356-787022023-03-04T18:37:52Z Time-frequency representation as a tool of failure detection Soeryanto, Yehezkiel Tegoeh Tjahjowidodo School of Mechanical and Aerospace Engineering DRNTU::Engineering::Mechanical engineering::Mechatronics Chemical-Mechanical Planarization, or CMP for short, is one of the final major process involved in semiconductor company when making a microchip. In this process the chip is being made perfectly smooth and flat with the help of chemical and mechanical forces. The tool which does the planarizing process is a polishing pad which will control the final thickness of the microchip. After a certain period, the polishing pad will lose its profile and if used will give in undesirable result from the polishing process. In order to be reprofiled, the polishing pad will be conditioned. The pad is conditioned by the help of a diamond disc, where it is put and held on conditioning head and it will be moved by the conditioning arm. After some time as well, however, just like any abrasive tools the diamond disc will also wear, degrade and loosen due to the friction between the surface and the diamond disc and it will reduce the life cycle of the disc. To prevent failure for the conditioning process, the current solution for the diamond disc is used for a certain amount of cycle before it was replaced. This, though, made a new issue where diamond discs are not used to their full life cycle. Producer could implement a method to check the diamond disc regularly, however one need to halt the machine to do it and that time will be counted as loss to the company. In this project, the objective is to view the possibility of automating the process of examining the diamond disc in CMP machine with the assistance of Deep Learning through Time-Frequency Analysis approach. Vibration value that will undergone signal processing technique and Neural Networks model as Deep Learning technique will be the scope of the project where a combination of both will be used to determine the possibility of this objective and for further research. Bachelor of Engineering (Mechanical Engineering) 2019-06-26T01:23:41Z 2019-06-26T01:23:41Z 2019 Final Year Project (FYP) http://hdl.handle.net/10356/78702 en Nanyang Technological University 43 p. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic DRNTU::Engineering::Mechanical engineering::Mechatronics
spellingShingle DRNTU::Engineering::Mechanical engineering::Mechatronics
Soeryanto, Yehezkiel
Time-frequency representation as a tool of failure detection
description Chemical-Mechanical Planarization, or CMP for short, is one of the final major process involved in semiconductor company when making a microchip. In this process the chip is being made perfectly smooth and flat with the help of chemical and mechanical forces. The tool which does the planarizing process is a polishing pad which will control the final thickness of the microchip. After a certain period, the polishing pad will lose its profile and if used will give in undesirable result from the polishing process. In order to be reprofiled, the polishing pad will be conditioned. The pad is conditioned by the help of a diamond disc, where it is put and held on conditioning head and it will be moved by the conditioning arm. After some time as well, however, just like any abrasive tools the diamond disc will also wear, degrade and loosen due to the friction between the surface and the diamond disc and it will reduce the life cycle of the disc. To prevent failure for the conditioning process, the current solution for the diamond disc is used for a certain amount of cycle before it was replaced. This, though, made a new issue where diamond discs are not used to their full life cycle. Producer could implement a method to check the diamond disc regularly, however one need to halt the machine to do it and that time will be counted as loss to the company. In this project, the objective is to view the possibility of automating the process of examining the diamond disc in CMP machine with the assistance of Deep Learning through Time-Frequency Analysis approach. Vibration value that will undergone signal processing technique and Neural Networks model as Deep Learning technique will be the scope of the project where a combination of both will be used to determine the possibility of this objective and for further research.
author2 Tegoeh Tjahjowidodo
author_facet Tegoeh Tjahjowidodo
Soeryanto, Yehezkiel
format Final Year Project
author Soeryanto, Yehezkiel
author_sort Soeryanto, Yehezkiel
title Time-frequency representation as a tool of failure detection
title_short Time-frequency representation as a tool of failure detection
title_full Time-frequency representation as a tool of failure detection
title_fullStr Time-frequency representation as a tool of failure detection
title_full_unstemmed Time-frequency representation as a tool of failure detection
title_sort time-frequency representation as a tool of failure detection
publishDate 2019
url http://hdl.handle.net/10356/78702
_version_ 1759855121806655488