2016Microscopy and MicroanalysisRequires access

Automatic Detection of Pearlite Spheroidization Grade of Steel Using Optical Metallography

Naichao Chen, Yingchao Chen, Jun Ai, Jianxin Ren, Rui Zhu, Xingchi Ma, Jun Han, Qingqian Ma

Open publisher page 10 citations

Abstract

To eliminate the effect of subjective factors during manually determining the pearlite spheroidization grade of steel by analysis of optical metallography images, a novel method combining image mining and artificial neural networks (ANN) is proposed. The four co-occurrence matrices of angular second moment, contrast, correlation, and entropy are adopted to objectively characterize the images. ANN is employed to establish a mathematical model between the four co-occurrence matrices and the corresponding spheroidization grade. Three materials used in coal-fired power plants (ASTM A315-B steel, ASTM A335-P12 steel, and ASTM A355-P11 steel) were selected as the samples to test the validity of our proposed method. The results indicate that the accuracies of the calculated spheroidization grades reach 99.05, 95.46, and 93.63%, respectively. Hence, our newly proposed method is adequate for automatically detecting the pearlite spheroidization grade of steel using optical metallography.

About this research paper

What this paper is about

To eliminate the effect of subjective factors during manually determining the pearlite spheroidization grade of steel by analysis of optical metallography images, a novel method combining image mining and artificial neural networks (ANN) is proposed. The four co-occurrence matrices of angular second moment, contrast, correlation, and entropy are adopted to objectively characterize the images. ANN is employed to establish a mathematical model between the four co-occurrence matrices and the corresponding spheroidization grade. Three materials used in coal-fired power plants (ASTM A315-B steel, ASTM A335-P12 steel, and ASTM A355-P11 steel) were selected as the samples to test the validity of our proposed method. The results indicate that the accuracies of the calculated spheroidization grades reach 99.05, 95.46, and 93.63%, respectively. Hence, our newly proposed method is adequate for automatically detecting the pearlite spheroidization grade of steel using optical metallography.

Why it matters

OpenAlex reports 10 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

To eliminate the effect of subjective factors during manually determining the pearlite spheroidization grade of steel by analysis of optical metallography images, a novel method combining image mining and artificial neural networks (ANN) is proposed. The four co-occurrence matrices of angular second moment, contrast, correlation, and entropy are adopted to objectively characterize the images. ANN is employed to establish a mathematical model between the four co-occurrence matrices and the corresponding spheroidization grade. Three materials used in coal-fired power plants (ASTM A315-B steel, ASTM A335-P12 steel, and ASTM A355-P11 steel) were selected as the samples to test the validity of our proposed method. The results indicate that the accuracies of the calculated spheroidization grades reach 99.05, 95.46, and 93.63%, respectively. Hence, our newly proposed method is adequate for automatically detecting the pearlite spheroidization grade of steel using optical metallography.

Key concepts: Metallography, Pearlite, Materials science, Metallurgy, Microstructure, Austenite

Related papers

Back to paper searchBrowse research topicsOriginal source
Automatic Detection of Pearlite Spheroidization Grade of Steel Using Optical Metallography — Research Paper | ScholarLens