Application of Principal Component Analysis and Response Surface Methodology in the Process of Steel Wire Tempering

Cristie Diego Pimenta,Messias Borges Silva,Fabrício Maciel Gomes,Rose Lima De Morais Campos

doi:10.5539/ijsp.v9n1p9

Cristie Diego Pimenta, Messias Borges Silva + Show 2 more

Open Access

https://doi.org/10.5539/ijsp.v9n1p9

Copy DOI

Abstract

The purpose of this work was the creation of a statistical modeling capable of replacing the process used to set up of the ovens of the quenching and tempering who is traditionally accomplished through adjustments made based on the results of mechanical properties tested in laboratory and required in customer specifications. This study seeks to understand the influence of input variables (factors) on the tensile strength limit, in SAE 9254 draw steel wires, with diameters 2.00 mm and 6.50 mm, used in the manufacture of valve springs and clutch springs for automobiles. The process input variables were investigated: steel wire diameter, processing speed, temper temperature and liquid polymer concentration (which is the tempering medium). Methodologies were used where design of experiments, multiple regression and quadratic regression, principal components analysis (multivariate statistical) and response surface methodology. Results revealed which variables are significant in the process. Also models obtained were validated using appropriate statistical methods. If this study is used, it can provide the automation of this process. It&rsquo;s important to point out that it could impact the increase in productivity and quality of product.

Highlights

This research, statistical methods were used to assist in the development of a statistical modeling to come to replace the traditional way as adjustment of the input variables of the heat treatment oven
The purpose of this work was the creation of a statistical modeling capable of replacing the process used to set up of the ovens of the quenching and tempering who is traditionally accomplished through adjustments made based on the results of mechanical properties tested in laboratory and required in customer specifications
To produce a mathematical model that could be used for diameters from 2.00 mm to 6.50 mm it was necessary to use the methods of design of experiments, multiple regression with terms of interaction and significance test (Tstudent)

Summary

Introduction

This research, statistical methods were used to assist in the development of a statistical modeling to come to replace the traditional way (trial and error) as adjustment of the input variables of the heat treatment oven. In this specific case, the initial setting was performed by means of tests of mechanical property (tensile strength limit) in a sample pilot. After going through all steps of a heat treatment, will be forwarded to the physical laboratory analysis This implies considerably operating routine analysis and waiting time, reducing the productivity of the process due to low income, since the oven remains inoperable until they are carried out in laboratory tests that serve as the basis for setting the temper oven. Steel mills have sought to develop these mathematical models to reduce the amount of laboratory tests and the time to set up the ovens, which can mean reducing costs for the company

Objectives

Methods

Results

Conclusion