A predictive model for characterizing hardness of D2 tool steel by eddy current method: A statistical optimization approach. Issue 4 (2nd October 2018)

• Record Type: Journal Article
• Authors: Hejazi, Taha-Hossein; Kahrobaee, Saeed
• Is Part Of: Research in nondestructive evaluation. Volume 29:Issue 4(2018)
• Page Start: 237
• Page End: 254
• Publication Date: 2018-10-02
• Subjects: Response surface modeling -- heat treatment -- eddy current method -- optimal frequency -- hardness variations; Nondestructive testing -- Periodicals; 620.112705
• DOI: 10.1080/09349847.2018.1439206
• Languages: English
• ISSNs: 0934-9847

Abstract

ABSTRACT: Response surface modeling is a well-grounded method to identify empirical relationships between factors involved in engineering experiments. This paper presents a novel mathematical/statistical approach to design an optimal predictive model to evaluate hardness by eddy current method. A new approach has been presented to find an optimal frequency in eddy current testing which maximizes detectability and accuracy of the measured values. In order to determine the hardness values of AISI D2 tool steel during tempering, the eddy current tests were conducted by applying different values of operating frequency in the range 50–1000 Hz. Unknown heat treatment conditions estimated from eddy current tests creates a link that provides hardness estimation. The results showed that coupling the eddy current method to the proposed model acts as a high accurate hardness measuring system for the samples with unknown heat treatment conditions.