Tort and Hajjar 2004
This paper presents the development of an experimental database of rectangular CFT (RCFT) member and connection tests as well as an investigation of the evolution of damage of RCFTs. This work represents the first stage of research intended to generate performance-based design guidelines for RCFT beam-columns and connections.
Analytical Study and Discussion
In order to investigate the evolution of damage the authors created a database of experimental results which documents various damage states. The data was divided into tests on columns, beam-columns, panel zones, pinned connections, moment connections, and frames, and was further divided into those subjected monotonic and cyclic loading. Included in the database were descriptions of the experimental setup, material properties, geometric properties, and experimental results. When available, both measured and nominal values were included. In addition to the peak strength and deformation, typical of an experimental database, information on the occurrence of local damage states was also included. The information regarding the local damage states consisted of the point, called the damage point, on the load-deformation curve at which the local damage state was first observed. These damage states include concrete cracking, concrete crushing, steel yielding, and steel fracture. The damage point was obtained from either experimental reporting or analytical assessment.
Points along the load-deformation curve were identified with non-dimensional measures of the deformation and energy called damage functions. The deformation-based damage function was defined as the ratio of the deflection to the deflection attained at the peak load. The energy-based damage function was defined as the ratio of the energy absorbed to the energy absorbed at the end of the test.
For the different types of specimens, i.e. monotonically loaded beam-columns, the damage points were curve fit to simple equations of structural parameters. This resulted in a suite of equations which predict the value of the damage function at which each damage state will occur. A similar method was used to create formulas which predict the ductility of the various types of specimens. Trends were identified based on the curve fit equations. For instance, it was shown that for low D/t ratios the deformation damage function predicting local buckling is larger than one indicating that local buckling takes place after the peak load is attained.
Parametric studies were performed to investigate the effect of specific damage states on the overall response of structural members and to compare the different damage states. For each type of specimen, a series of RCFT members was generated with structural properties chosen to represent the full range of the respective databases. The order of occurrence of the various damage states was inferred from the predicted values of the damage functions, lower values will occur first. The relative occurrence time of the local damage states were compared and observations were made on the overall structural behavior.
The objective of the research is to develop performance-based design guidelines. The equations which were developed play a key role in the performance-based design methodology which is described in the paper.
Tort, C. and Hajjar, J. F. (2004). “Damage Assessment of Rectangular Concrete-Filled Steel Tubes for Performance-Based Design,” Earthquake Spectra, Vol. 20, No. 4, November, pp. 1317-1348.