MCEER/NCEER Bulletin Articles: Research
An Evaluation of Progressive Damage in Reinforced Concrete Circular Bridge Columns
by Sashi K. Kunnath and Ashraf El-Bahy
This article presents research resulting from NCEERs Highway Project, task 106-E-5.3. It summarizes research reported in an upcoming NCEER technical report entitled Cumulative Seismic Damage of Reinforced Concrete Bridge Piers, NCEER-97-0006 (see report review on page 17). The work was conducted by researchers at the University of Central Florida and the National Institute of Standards and Technology (NIST). Comments and questions should be directed to Sashi Kunnath, NIST, at (301) 975-6078; email: firstname.lastname@example.org.
A comprehensive experimental study was undertaken to investigate the mechanics of damage accumulation in reinforced concrete circular bridge piers subjected to a series of simulated earthquake excitations. Twelve identical quarter-scale bridge columns were designed and fabricated in accordance with current AASHTO specifications. A unique setup to expedite the testing process was designed and built in the structural test facility at the National Institute of Standards and Technology (NIST).
The testing was divided into two phases. Phase I testing consisted of benchmark tests to establish the monotonic force-deformation envelope, the energy capacity under standard cyclic loads, and constant amplitude tests to determine the low-cycle fatigue characteristics of the bridge column. Phase II testing was composed of a series of analytically predicted displacement amplitudes representing the bridge response to typical earthquakes. The results of Phase I testing provided information on the fatigue behavior of reinforced concrete and Phase II provided data on the effects of load path on cumulative damage.
A major departure from past practice of laboratory testing that was pursued in this investigation was development and use of random displacement histories rather than "standard" displacement cycles with increasing amplitudes. Given the complexity of the cumulative damage process and the innumerable parameters affecting the response, every effort was made to keep system variables to a minimum. Consequently, the imposed displacement history was the only variable introduced in the experimental testing.
Test observations indicate two potential failure modes: low cycle fatigue of the longitudinal reinforcing bars; and confinement failure due to rupture of the confining spirals. The former failure mode is associated with relatively large displacement amplitudes in excess of 4% lateral drift while the latter is associated with a larger number of smaller amplitude cycles. Analytical studies indicate that most earthquakes induce few large amplitude cycles pointing to the need for proper confinement to prevent catastrophic failure.
Figure 1 shows a typical displacement history and the resulting shear force versus drift response of the bridge column.
The results of the testing were also used in an analytical study of cumulative damage. A simple fatigue-based model, derived from existing theories in the literature, was used to predict cumulative damage in flexural bridge piers. Critical damage measures, such as stiffness degradation, dissipated hysteretic energy, ductility, and fatigue were evaluated against observed behavior. It was found that none of these damage measures consistently predict observed damage limit states though fatigue-based models demonstrated better reliability.
Fatigue-life expressions, using a Coffin-Manson rule in combination with Miners hypothesis, account only for low-cycle fatigue damage of steel. It appears that a model which combines low-cycle fatigue failure in combination with confinement deterioration will yield excellent results. A simple fatigue life relationship was proposed, derived from the original work of Mander and Cheng (1995), based on the experimental data generated from constant-amplitude testing of specimens in Phase I, and was shown to produce improved damage prediction characteristics. The following expression was obtained to define the number of half-cycles to failure:
and the resulting damage, expressed in terms of a damage index D.I., is computed from the following expression:
where ndi is the number of half cycles at a particular drift "di" obtained from the analysis indicated in step (a), and Nfi is the number of half cycles to failure at the same drift "di" obtained from Equation (1). Figure 2 shows the predicted damage to all columns tested in Phase II using a damage index based on the above formulation.
It was further observed that the energy-dissipation capacity of members is path-dependent, hence, models of seismic damage that rely only on measures of energy dissipation cannot predict failure if it is not related to ductility. Findings from this study will provide additional input into the development of performance-based design specifications wherein design is linked to damage limit states.
NCEER Bulletin, October 1997, Vol. 11, No. 4