Performance Assessment of Reinforcing Bar Anchorage and Buckling Effects in RC Piers - 2472005

Project goals and objectives

The primary objective of this research project is to quantitatively study parameters that influence the occurrence of bar buckling in bridge components and to develop data to be used by other PEER researchers to develop and calibrate appropriate simulation (Kunnath) and damage (Eberhard) models. The specific Year-8 research project objectives include:

• - Generation of experimental data for validation and calibration of performance and analytical models.
• - Improvement of a practical performance model to predict bar buckling
• - Meet the needs of those conducting the analytical (Kunnath) and Performance (Eberhard) simulation including coordination and exchange meetings

Role of this project in supporting PEER's mission (vision)

Study of experimental and analytical research results suggests that bar buckling is influenced by the response of the longitudinal steel, which is in turn affected by the strain history and the type of steel, and the lateral restraint on the bar, typically provided by the transverse reinforcement and the cover. Most previous research studies have focused on the longitudinal steel response and have resulted in models which modify the steel constitutive law to account for bar buckling. However, experimental research shown that neglecting the influence of the lateral restraint limits the applicability of these models. In this study, the restraint and other salient parameters will be considered explicitly using experimental research methods. The projects will results in reliable experimental data for supporting the development and validation of bar buckling damage and simulation models (in development by others).

Methodology employed

The experimental program has two goals, which require two separate test series. The first is to generate experimental data from which the stiffness of the circular hoops can be obtained. Modeling is still necessary, if for no other reason than the fact that a finite number of tests will have to be conducted and at model scale. However, such tests will provide physical data against which to calibrate analytical or numerical models of an isolated bar restrained against lateral movement by a flexible support. This model represents a small part of the entire column system, but it is the most critical part.

The tie stiffness tests were conducted on specially designed specimens, shown in Figure 1. A short segment of a circular column is cast with longitudinal bars and individual, welded, hoops. The #5 longitudinal bars were placed evenly around the perimeter. The hoops were spaced at 8" along the column length to allow room for the test equipment. A hook was placed under the spiral and pulled radially by a center-hole ram supported by a steel bridge over the hoop being tested, as shown in Figure 1 to simulate a single longitudinal bar pushing radially outwards against the hoop. The load and displacement of the hook were measured to assess the spiral stiffness. The three primary variables tested were (a) the bar size, (b) the presence of cover, and (c) the bond along the hoop.

A pilot test series was conducted. Observations during the pilot tests already reveal some useful findings. First, the hoop experiences bending as well as tension, and the curvature is much larger at the hook than elsewhere. The local curvature is, in fact, defined by the diameter of the hook, which should be the same as that of a typical longitudinal bar if the conditions in a complete column test are to be replicated faithfully. The curvature diminishes with distance from the hook, and changes sign between the hook and the point of contact with the concrete core. A plastic hinge occurs at the hook while the average strain in the hoop is still relatively low. This suggests that the criterion for hoop failure in a complete column should be based on combined tension and bending at a buckling longitudinal bar rather than on pure tension fracture of the hoop due to uniform lateral expansion of the concrete core. In this sense, initiating of longitudinal buckling is indeed the event that sets in motion the process of column failure.

The second goal is to produce test data on complete columns subjected to cyclic lateral loads. These tests represent the whole column system, and will differ from others conducted in the past in that bar buckling will be the primary focus of the study. Special instrumentation was developed to ensure detection of the initiation of bar buckling, and load histories will be selected to determine as precisely as possible the effect of load history.

The experimental study on complete columns includes eight specimens at 1/3 scale. In an effort to capture buckling of the longitudinal reinforcement, two methods of measuring displacements will be used in conjunction with strain gages placed on both the longitudinal bars and the spiral. First, a set of string potentiometers will be attached to the longitudinal bars and another to the body of the column at three locations above the top of the footing as shown in Fig. 2. The differences between the readings of the pairs of instruments will give the displacement of the bar relative to the column at each level. Second, digital photogrammetry will be used to capture column displacement, concrete bulging and, eventually, buckling of longitudinal bars as shown. Three cameras set up with overlapping fields of view (FOV) will record digital images simultaneously to piece together 3-dimensional displacement information about the column as concrete damage (cracking, spalling, bar buckling) progresses.

One of the primary test variables was the displacement history imposed on the column. Figure 3 shows two a-typical displacement histories. The first, subjected the column to a large strain demand (to 8% drift) followed by a ratcheted history to determine the drift range needed to cause buckling. The second imposed a ratcheted history with a constant drift range (4%). Together they can be used to determine the combination of drift range and maximum drift needed to cause buckling.

Other similar work being conducted within and outside PEER and how this project differs

Previous researchers, within and outside of PEER, have developed models to predict the occurrence of bar buckling (e.g., Eberhard, Pantazopoulou, Bayrak and Sheikh). Experimental programs have identified approximately the onset of bar buckling (e.g., Lehman and Moehle, Henry and Mahin, Eberhard and Stanton). However the experimental research programs have not been developed to specifically study the parameters that influence bar buckling, in particular the restraint provided by the cover concrete and the transverse reinforcement. This study will provide a much needed understanding, including reliable experimental data, of the influence of salient column parameters on the onset and occurrence of bar buckling. The research results will be useful to PEER researchers to calibrate and validate performance and simulation models, and to bridge engineers to predict bar buckling.

Expected milestones & deliverables

The expected outcomes of the project include:

• - Practical guidelines outlining the range of application, and the limits, of existing bar buckling models
• - Improved understanding of the mechanism of bar buckling
• - Local data to support the development of advanced bar buckling models for performance- based assessment and design
• - Improvements to existing damage models for bar buckling and cross section fatigue
• - Generation of experimental data for validation and calibration of performance and analytical models.
• - Improvement of a practical performance model to predict bar buckling