Knowledge Based Wear Particle Analysis

The paper describes a knowledge based system for analysis of microscopic wear particles. Wear particles contained in lubricating oil carry important information concerning machine condition, in particular the state of wear. Experts (Tribologists) in the field extract this information to monitor the operation of the machine and ensure safety, efficiency, quality, productivity, and economy of operation. This procedure is not always objective and it can also be expensive. The aim is to classify these particles according to their morphological attributes of size, shape, edge detail, thickness ratio, color, and texture, and by using this classification thereby predict wear failure modes in engines and other machinery. The attribute knowledge links human expertise to the devised Knowledge Based Wear Particle Analysis System (KBWPAS). The system provides an automated and systematic approach to wear particle identification which is linked directly to wear processes and modes that occur in machinery. This brings consistency in wear judgment prediction which leads to standardization and also less dependence on Tribologists.

Authors:

• Mohammad S. Laghari
• Qurban A. Memon
• Gulzar A. Khuwaja

Keywords:

• Computer vision
• knowledge based systems
• morphology
• wear particles.

References:

[1] H. P. Jost, "Tribology - Origin and Future," in Wear, vol. 139, 1990, pp.
1-17.
[2] B. J. Roylance and T. M. Hunt, Wear Debris Analysis, Coxmoor
Publishing, Oxford, 1999.
[3] W. W. Seifert and V. C. Westcott, "A method for the study of wear
particles in lubricated oil," in Wear, vol. 21, 1972, pp. 27-42.
[4] T. M. Hunt, Handbook of Wear Debris Analysis and Particle Detection
in Fluids, Elsevier Science, London, New York, 1993.
[5] A. C. Cumming, "Condition monitoring today and tomorrow - an airline
perspective," presented at the 1989 Int. Conf. COMADEN 89,
Birmingham, U.K.
[6] T. P. Sperring and B. J. Roylance, "Some recent development in the use
of quantitative procedures for performing wear debris analysis," JOAP
International Condition Monitoring Conference, Mobile, Al., 2000, pp.
205-210.
[7] G. A. Khuwaja and M. S. Laghari, "Computer vision techniques for
wear debris analysis," in Int. J. Computer Applications in Technology,
vol. 15, no. 1/2/3, 2002, pp. 70-78.
[8] M. S. Laghari and A. Boujarwah, "Wear particle identification using
image processing techniques," in ISCA 5th Int. Conf. on Intelligent
Systems, Reno, Nevada, 1996, pp. 26-30.
[9] Leica Cambridge Ltd., Leica Q500MC Qwin User Manual, Leica
Cambridge Ltd., U.K., 1994.
[10] H. Xu, A. R. Luxmoore and F. Deravi, "Comparison of shape features
for the classification of wear particles," in Engineering Applications of
Artificial Intelligence, vol. 10, no. 5, 1997, pp. 485-493.
[11] S. Raadnui, Wear Particle Characterisation Utilising Computer Image
Analysis, Ph.D. Thesis, Dept. Mech. Eng., University of Wales,
Swansea, 1996.
[12] B. J. Roylance, "Wear debris analysis for condition monitoring," in
INSIGHT 36, vol. 8, 1994, pp. 606-610.
[13] M. S. Laghari, "Shape and edge detail analysis for wear debris
identification," in Int. J. of Computers and their Applications, vol. 10,
no. 4, 2003, pp. 271-279.
[14] A. K. Muhamad and F. Deravi, "Neural networks for texture
classification," in IEE 4th Int. Conf. on Image Processing and its
Applications - IPA'92, Maastricht, The Netherlands, 1992, pp. 201-204.
[15] J. Garcia-Consuegra and G. Cisneros, "Integration of gabor functions
with coocurrence matrices: Application to woody crop location in
remote sensing," in IEEE Int. Conf. on Image Processing, vol. II, Kobe,
1999, pp. 330-333.
[16] M. S. Laghari, "Recognition of texture types of wear particles," in Int. J.
of Neural Comp. & Applications, vol. 12, 2003, pp. 18-25.