Data-driven white layer control in EDM using image processing and machine learning
Emerging Materials Research · 19.11.2026 · 15 (4) · 605-620
DOI: 10.1680/jemmr.25.00148 ↗Kurz gefasst
Fachartikel in der Zeitschrift Emerging Materials Research, online erschienen am 02.10.2026. Verfasst von Masoud Azadi Moghaddam und 5 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben.
Abstract (Originalsprache)
Electrical discharge machining (EDM) of hard-to-cut materials often suffers from excessive white layer thickness (WLT), high surface roughness (SR), and low material removal rate (MRR). This study presents an integrated data-driven modeling and multi-objective optimization framework to address these challenges in die-sinking EDM of AISI2312 steel. An L36 Taguchi design was used to investigate the effects of pulse-on time (Ton), pulse-off time (Toff), discharge current (I), voltage (V), and duty factor (η) on WLT, SR, and MRR. Four machine learning models were developed, with extreme gradient boosting achieving superior predictive accuracy (R2 = 0.972, MAPE = 4.1%). A hybrid simulated annealing–grey relational analysis (SA–GRA) algorithm was then used for multi-objective optimization. The SA–GRA framework outperformed conventional Taguchi and response surface methodology methods by directly exploring nonlinear multi-response interactions and efficiently identifying robust process windows. Optimal parameters (Ton = 25 μs, Toff = 10 μs, I = 2.5 A, η = 1.6 s, V = 60 V) yielded a predicted WLT of 0.88 μm while improving SR and MRR. This integrated approach demonstrates both high accuracy and computational efficiency, providing a scalable tool for smart EDM systems and Industry 4.0-enabled digital twin platforms.
Zitieren
APA
Moghaddam, M. A., Arteshyar, K., Tavallaei, M. E., Ghazijahani, S. M., Nourani, A., & Kolahan, F. (2026). Data-driven white layer control in EDM using image processing and machine learning. Emerging Materials Research, 15(4), 605-620. https://doi.org/10.1680/jemmr.25.00148
Vancouver
Moghaddam MA, Arteshyar K, Tavallaei ME, Ghazijahani SM, Nourani A, Kolahan F. Data-driven white layer control in EDM using image processing and machine learning. Emerging Materials Research. 2026;15(4):605-620. doi:10.1680/jemmr.25.00148
BibTeX
@article{Moghaddam2026,
author = {Moghaddam, Masoud Azadi and Arteshyar, Kiana and Tavallaei, Mohammad Ebadi and Ghazijahani, Shahrzad Mohammadzadeh and Nourani, Ahmad and Kolahan, Farhad},
title = {Data-driven white layer control in EDM using image processing and machine learning},
journal = {Emerging Materials Research},
year = {2026},
volume = {15},
number = {4},
pages = {605--620},
doi = {10.1680/jemmr.25.00148}
}
EUVIMED