Beck, Robert Teran (1991) Fundamental problems in the application of structural identification procedures to damage detection. California Institute of Technology . (Unpublished) http://resolver.caltech.edu/CaltechEERL:1991.EERL-91-03
PDF (Adobe PDF (13 MB))
See Usage Policy.
Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechEERL:1991.EERL-91-03
There are fundamental problems in the application of structural identification procedures to damage detection which still need to be resolved. The present study investigates the underlying issues and then provides a number of techniques which solve a series of unresolved problems. Tile techniques developed range from extensions and refinements of previous methods to the adaptation of novel homotopy methods. The results from simulated data show that ill-conditionling, non-uniqueness and temporal synchronization of the data are the most serious problems encountered. Criteria to resolve these are then put forth. From the experimental studies, however, it becomes evident that modeling error is the most serious issue. The experimental results show, nonetheless, that even with large model errors, it is possible to localize the area of damage to within a sub-structure. The techniques are then applied to data obtained from a ten-story steel frame building. Previous studies on such structures have indicated large changes in the natural frequencies, especially during the San Fernando earthquake of February 9, 1971. The present study shows how changes in the natural frequencies and in the modeshapes are related to the degradation of the inter-story stiffness along the height of the building. Low amplitude forced vibration and ambient vibration test data yield one set of results: at these levels of motion the structure seems to retain much of its original uniform stiffness. This is true even after strong motion, leading to the notion that the building "has healed" with time. It is clear from the studies how this apparent stiffness is lost immediately once the strong motion of even moderate earthquakes has begun and it is thought that this is due to a combinations of effects. Results show that for the 1971 San Fernando earthquake, stiffness losses in the order of 50% occurred in the middle stories towards the end of the strong motion part of the seismic motion.
|Item Type:||Report or Paper (Technical Report)|
|Additional Information:||PhD, 1991|
|Group:||Earthquake Engineering Research Laboratory|
|Usage Policy:||You are granted permission for individual, educational, research and non-commercial reproduction, distribution, display and performance of this work in any format.|
|Deposited By:||Imported from CaltechEERL|
|Deposited On:||15 Oct 2001|
|Last Modified:||26 Dec 2012 13:57|
Repository Staff Only: item control page