Chapter 44. Recognition of Subsurface Defects in Machined Ceramics by Application of Neural Networks to Laser Scatter Patterns

  1. John B. Wachtman Jr.
  1. Michael C. Stinson1,
  2. Owen W. Lee2,
  3. J. Scott Steckenrider1 and
  4. William A. Ellingson2

Published Online: 28 MAR 2008

DOI: 10.1002/9780470314500.ch44

Proceedings of the 18th Annual Conference on Composites and Advanced Ceramic Materials - A: Ceramic Engineering and Science Proceedings, Volume 15, Issue 4

Proceedings of the 18th Annual Conference on Composites and Advanced Ceramic Materials - A: Ceramic Engineering and Science Proceedings, Volume 15, Issue 4

How to Cite

Stinson, M. C., Lee, O. W., Steckenrider, J. S. and Ellingson, W. A. (1994) Recognition of Subsurface Defects in Machined Ceramics by Application of Neural Networks to Laser Scatter Patterns, in Proceedings of the 18th Annual Conference on Composites and Advanced Ceramic Materials - A: Ceramic Engineering and Science Proceedings, Volume 15, Issue 4 (ed J. B. Wachtman), John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9780470314500.ch44

Author Information

  1. 1

    Department of Computer Science Central Michigan University

  2. 2

    Electrical Engineering Department University of California at Los Angeles Energy Technology Division Argonne National Laboratory

Publication History

  1. Published Online: 28 MAR 2008
  2. Published Print: 1 JAN 1994

ISBN Information

Print ISBN: 9780470375327

Online ISBN: 9780470314500

SEARCH

Keywords:

  • backpropagation;
  • porosity;
  • training pair;
  • averaging;
  • delineation

Summary

Laser scatter has shown promise as a method to characterize damage microstructural variations as well as a method to characterize surfaces in optical translucent ceramics. Because large volumes of data need to be handled (and sorted) quickly, automated pattern recognition methods using neural networks have been implemented to recognize differences in patterns. A He-Ne laser (γ=0.632μ) has been used to obtain scatter patterns from hot pressed Si3N4 with various microstructural variations. By use of a backpropagation neural network running on an IBM PC clone 486/33 machine, a correlation was established between subsurface microstructure and position in Si3N4 ball bearings. The data were confirmed by destructive analysis.