The Application of Regression Analysis on the Core Size Changes of Furnace Welding of Water Tank

碩士 === 南台科技大學 === 工業管理研究所 === 101 === In recent years, besides price, the quality of a radiator has become an influential factor to a customer’s purchase intention as competition increases and maintenance market rises. So the melioration of quality and techniques is one of the strategies for improvi...

Full description

Bibliographic Details
Main Authors: FANG,JIN-KUN, 方勁焜
Other Authors: 方正中
Format: Others
Language:zh-TW
Published: 102
Online Access:http://ndltd.ncl.edu.tw/handle/58404512031237500747
Description
Summary:碩士 === 南台科技大學 === 工業管理研究所 === 101 === In recent years, besides price, the quality of a radiator has become an influential factor to a customer’s purchase intention as competition increases and maintenance market rises. So the melioration of quality and techniques is one of the strategies for improving competitiveness. When the production of radiator cores is in progress, manufacturers may encounter the complexity of process and the diversification of products. To ensure international standard-compliant quality and driving safety, skilled staff and stable process are often required. The purpose of this research is to create a predictor model of variation of size of radiator core, and this model meets both quality control and production management requirements. Among all processes of radiator production, furnace-welding is the most vital one. So in this research, we took size variation of radiator core as dependent variable, and selected content percentage of iron, aluminum, and the size of core before welding as independent variables through cause-and-effect diagram. And then, we found an ideal prediction model for radiator core size changes through stepwise regression procedure. After that, we found out an optimal combination of parameters correspondent with quality characteristics, with Solver in Excel. In the end of the research, an actual manufacture process was carried out to verify the effectiveness under the conditions in this mode. The result indicated that the quality of radiators would be improved and the defective rate would be decreased. After the optimal parameters were applied with manufacture process, the qualified rate has increased from 87% to 95%.