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.

Download full-text PDF

Source
http://dx.doi.org/10.1017/S1431927615015706DOI Listing

Publication Analysis

Top Keywords

spheroidization grade
16
pearlite spheroidization
12
grade steel
12
optical metallography
12
steel optical
8
co-occurrence matrices
8
steel astm
8
proposed method
8
steel
6
spheroidization
5

Similar Publications

Cast austenitic stainless steel (CASS) has been widely used for long-term service periods in high-temperature applications. Nonetheless, the behavior of CASS after high-temperature and long-term service is insufficiently researched alongside the microstructural analysis for sigma phase and carbide. Here, intermediate pipe support made from CASS grade HH was investigated after 24 years of service at 700 °C and compared with the solution-treated specimen at 1100 °C for 2 h.

View Article and Find Full Text PDF

Influence of Sintering Atmosphere, Temperature and the Solution-Annealing Treatment on the Properties of Precipitation-Hardening Sintered 17-4 PH Stainless Steel.

Materials (Basel)

January 2023

Department of Material Engineering, Faculty of Materials Engineering and Physics, Cracow University of Technology, 31-155 Krakow, Poland.

So far, unlike metal injection molding (MIM), conventional powder metallurgy technology (PM) has not been regarded as a method for producing structural elements from 17-4 PH powders, due to the problems of obtaining almost fully compacted shapes after sintering. Nevertheless, recent research demonstrates that it is possible to manufacture sintered parts with high strength by pressing and sintering. The purpose of the study was to determine the degree of densification of 17-4 PH sintered stainless steel during sintering at different temperatures and atmospheres.

View Article and Find Full Text PDF

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.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!