Data mining for the investigation of unsteady flow field in a hard disk drive

Seiichiro Morizawa, Shinkyu Jeong, Koji Shimoyama, Shigeru Obayashi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This study was performed to examine the mechanism of flow-induced vibration (FIV) in a hard disk drive (HDD). For this purpose, data mining using self-organizing map (SOM) and Bayesian network was applied to unsteady computational fluid dynamics (CFD) simulation data for a hard disk drive. The present data mining started from the extraction of temporal indices from the time series data of fluid properties given at each grid point. Then, a set of grid points was divided into several clusters based on the similarity of the temporal indices by using SOM, and the clustered data were mapped onto a real space of HDD. Through this process, characteristic phenomena latent in the unsteady flow field were classified and identified. Finally, the relations between temporal indices and FIV were constructed by using Bayesian network. The resulting network structure revealed a possible mechanism ofFIV that originates from a temporal sequence of flow energy dissipation and production.

Original languageEnglish
Title of host publicationProceedings - 2010 2nd World Congress on Nature and Biologically Inspired Computing, NaBIC 2010
Pages152-157
Number of pages6
DOIs
Publication statusPublished - 2010
Event2010 2nd World Congress on Nature and Biologically Inspired Computing, NaBIC 2010 - Kitakyushu, Japan
Duration: 15 Dec 201017 Dec 2010

Publication series

NameProceedings - 2010 2nd World Congress on Nature and Biologically Inspired Computing, NaBIC 2010

Conference

Conference2010 2nd World Congress on Nature and Biologically Inspired Computing, NaBIC 2010
Country/TerritoryJapan
CityKitakyushu
Period15/12/1017/12/10

Keywords

  • Bayesian network
  • Flow-induced vibration
  • Hard disk drive
  • Self-organizing map
  • Unsteady flow field

Fingerprint

Dive into the research topics of 'Data mining for the investigation of unsteady flow field in a hard disk drive'. Together they form a unique fingerprint.

Cite this