Novel Health Monitoring Procedure for Reinforced Concrete Slabs


Article Preview

This research considers the problem for vibration based damage diagnosis in reinforced concrete slabs. It suggests the analysis of the time domain measured accelerations for the purposes of fault detection and quantification. The measured accelerations from different damage states of the slab are first subjected to initial transforms to bring them to vectors with reasonable component number. These vectors are further transformed using principal components analysis (PCA), which brings their coordinates down to two. In addition to reducing the number of measured acceleration points, PCA clusters the new vectors making distinguishable the different damage states. The results from the application of the suggested method convincingly demonstrate that the method can be applied for fault detection as well as for estimating the damage extent.



Edited by:

M. Lucas




I. Trendafilova and M. Imbabi, "Novel Health Monitoring Procedure for Reinforced Concrete Slabs", Applied Mechanics and Materials, Vols. 1-2, pp. 65-70, 2004

Online since:

September 2004




[1] H.C. Tan, O.R. Famiyeisin, M.S. Imbabi, Computers and Structures vol. 79 (2001), p.2413.

[2] I.T. Jolliffee, Principal Component Analysis, 2 nd Edition, (Springer-Verlag, 2002).

[3] C.M. Bishop, Neural Networks for Pattern Reognition, (Oxford University Press, 1995).

[4] K. Worden. and G., Manson, Proceedings of the 17th International Modal Analysis Conference. Orlando, Florida, (1999).

[5] I. Trendafilova, W. Heylen, Mechanical Systems and Signal Processing vol. 17 No4 (2003) p.825.

[6] S.W. Doebling C.R. Farrar, M.B. Prime and D.W. Shevitz, Shock and vibration digest, 30 (2) (1998) p.91.

[7] N. Stubbs,S. Park, C. Sikorsky and S. Choi, International J. of System Science, vol. 31 No11 (2000) p.1361.

[8] J Tou, R Gozales Pattern recognition principles (Addison Wesley Publishing Company, 1984).