Evaluation of a proposed neural network predictive model for grind-hardening

H. A. Youssef, M. Y. Al-Makky, M. M. Abd-Elwahab

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


This work describes the development of neural network model for grind-hardening process which utilizes the grinding heat to induce martensitic phase transformation in annealed or tempered steels. Neural networks have been shown to be versatile for performance prediction involving non-linear processes. Machining performance prediction involving various process variables is a non-linear problem. The developed neural network represents the obtained results from the developed network which are compared with experimental data achieved from surface grind-hardening process for temperd steels. The developed neural network attemps to use numerous variables involved in the grind-hardening process and develops a knowledge based system with the capabilities of utilising both existing and new grind data. The influence of different grinding parameters on the obtained temperature, surface roughness, cutting forces and hardness will be tested and predicted during the process. The comparison between the results shows a significant correlation which assure the benifites of using the proposed neural network in the machining field. It is believed that neural and adaptive systems should be considered other tools in the engineer's toolbox. However, today's measuring facilities and process technologies allow measuring the important process-related signals with high sampling rates. Consequently, it is possible to build up a database from the measured data and to produce appropriate process models by using novel technologies, such as artificial intelligence techniques.

Original languageEnglish (US)
Pages (from-to)411-417
Number of pages7
JournalAEJ - Alexandria Engineering Journal
Issue number4
StatePublished - Jul 1 2003
Externally publishedYes


  • Grind-hardening
  • Grinding
  • Multilayer perceptron
  • Neural networks

ASJC Scopus subject areas

  • Engineering(all)


Dive into the research topics of 'Evaluation of a proposed neural network predictive model for grind-hardening'. Together they form a unique fingerprint.

Cite this