Application of Taguchi Method to Improve Laser Repair Yield in a TFT-LCD Module Factory

碩士 === 中原大學 === 工業與系統工程研究所 === 106 === Nowadays, the manufacturing process and technology of liquid crystal displays (LCD) are getting more advanced, and investment of production facility and equipment is also much higher. In these advanced manufacturing processes, many defects are caused by process...

Full description

Bibliographic Details
Main Authors: Jhen-Cyuan,Jhuang, 莊振權
Other Authors: WEI-JHONG,SIANG
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/ezck76
Description
Summary:碩士 === 中原大學 === 工業與系統工程研究所 === 106 === Nowadays, the manufacturing process and technology of liquid crystal displays (LCD) are getting more advanced, and investment of production facility and equipment is also much higher. In these advanced manufacturing processes, many defects are caused by processing or environmental abnormal factors. Instead of scraping the finished goods, it is necessary to use laser technology to repair the panel defects, due to the high unit price and production cost of LCD products. Laser repair can reduce scrap cost; rework time, material handling cost, and production costs. If quality of the laser repair process can be assured, the goals of improving product yield and reducing costs could be achieved. This study first introduced types of defects in TFT-LCD panels and how to conduct laser repair process. The objective of this study is to improve the yield of laser repair process, and firing frequency, firing energy, and numbers of firing are selected as main factors. The orthogonal array of the Taguchi method was applied to conduct experiments, and the selected main factors were significant by the result of analysis of variances. The process parameters were set based on the S/N ratios and experiments were conducted again to verify the parameter settings. The verification experimental data and the model-predicted values were analyzed and compared, and it showed that the laser repair yields were improved with the parameter settings.