SURFACE ROUGHNESS OPTIMIZATION IN EDM FOR CIRCULAR COPPER ELECTRODE BY RSM-GA APPROACH

Surface Roughness is important to the quality and performance of the finished products. Therefore, the minimization of roughness in manufacturing sectors, such as: automotive and aerospace, is of utmost importance. It is also economical and desirable if the finished parts do not need further grindin...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Annals of Faculty Engineering Hunedoara 2012-07, Vol.10 (3), p.299
Hauptverfasser: Patwari, Md Anayet U, Chowdhury, N A, Arif, M D, Chowdhury, Md S I
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Surface Roughness is important to the quality and performance of the finished products. Therefore, the minimization of roughness in manufacturing sectors, such as: automotive and aerospace, is of utmost importance. It is also economical and desirable if the finished parts do not need further grinding or polishing operations to meet the required finish tolerances. For achieving the required optimum level of surface quality, the proper selection of machining parameters in EDM is very essential. In this study, a 1 inch diameter round copper electrode was used to achieve minimum surface roughness, in the EDM of mild steel specimens, by controlling three important machining parameters: T^sub on^ (On Time), T^sub off^ (Off Time), and V (Gap Voltage). Machining was performed on a CNC JS EDM machine. A statistical prediction model for average surface roughness, previously developed by the authors using Design Expert Software, was coupled with genetic algorithms to predict the minimum possible R^sub a^. The minimum roughness predicted by GA shows good agreement with the physical measurement. The predicted surface roughness was also validated using a novel Digital Image Processing (DIP) technique developed by the authors and determined to have an accuracy of 90.62%. [PUBLICATION ABSTRACT]
ISSN:1584-2665
2601-2332