Modeling the cleanliness level of an ultrasonic cleaning system by using design of experiments and artificial neural networks

C. H. Wu, Y. S. Wong, W. H. Ip, Henry C W Lau, Ka Man Lee, G. T S Ho

Research output: Journal article publicationJournal articleAcademic researchpeer-review

3 Citations (Scopus)

Abstract

The hard disk drive is a reliable and relatively cheap mass storage device used in every computer nowadays. In this study, one major issue affecting the product quality of the fixture inside a hard disk drive is the surface contamination of the arm finger of actuator (AFA). For economical exploitation, a primary concern is to generate a model for optimizing the process parameter settings necessary to sustain the desired cleanliness level in an ultrasonic cleaning process. Two approaches were employed to identify critical process parameters, followed by the determination of the optimal parameter settings. The former approach was a statistical design of experiments (DOE) for developing regression equations for predicting the cleanliness level and finding out the dependence of each parameter and outcome. The latter approach was in using an artificial neural network (ANN) for building prediction models. A comparative study showed that both approaches have advantages over other methods. The results obtained show a reduction in contamination of the AFA; hence it provides an aid in the improvement of product quality.
Original languageEnglish
Pages (from-to)287-300
Number of pages14
JournalInternational Journal of Advanced Manufacturing Technology
Volume41
Issue number3-4
DOIs
Publication statusPublished - 1 Mar 2009

Keywords

  • Artificial neural networks
  • Statistical design of experiments
  • Ultrasonic cleaning process

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering
  • Control and Systems Engineering
  • Computer Science Applications
  • Software
  • Mechanical Engineering

Cite this