PBEE Assessment and Design of Enhanced Bridge Piers for Near Field Effects - 2402005

Project goals and objectives

The overall goal of this study is to examine and assess various design concepts for enhancing the seismic performance of new bridge structures of the type being considered by PEER. Promising design details will be assessed though dynamic shaking table and/or quasi-static tests as well as through more extensive nonlinear dynamic analyses using the OpenSees computational framework. In particular, the potential for reducing residual displacements of bridges following severe earthquakes will be examined through the use of special plastic hinge regions containing combinations of unbonded prestressed and mild reinforcement. Other design approaches utilizing high performance concrete will be examined in concert with others working within the PEER Center. The results will be presented in a fashion that will support the development and assessment of the overall PEER PBEE methodology being devised by the Bridge and Transportation Thrust Area. The goal of this project would be to demonstrate physically and through PBEE the value of the PEER methodology for a specific application, and thereby accelerate the adoption of new bridge design technologies into practice.

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

This project will develop and validate column design procedures and analytical models where columns will perform better than those using conventional design methods. This will reduce the amount of residual displacement in a bridge following an earthquake and degree of damage to the plastic hinge region. This will demonstrate the capabilities of OpenSees and of the PEER methodology's to rationally evaluate the desirability of new technologies.

Methodology employed

The primary goals of these continuing investigations are to (1) demonstrate the viability of new design and construction strategies for mitigating the substantial residual displacements that tend to occur in modern bridges when large inelastic deformations occur during severe earthquake shaking, (2) use experimental data from shaking table tests of for conventional and improved column designs to help calibrate the confidence that can be placed in various analytical models and procedures, and (3) support the development of the PEER methodology as applied to bridge structures. Residual displacements have an important impact on post-earthquake operability of bridge structures, and repair costs. In addition, opportunities will be pursued to support the work of others (Lehman, Kunnath, etc.) within PEER regarding other behavioral aspects that adversely affect overall performance, such as spalling, bar buckling and so on.

The initial thrust of this work has been focused on shaking table testing of a lightly reinforced bridge column specimen with unbonded post-tensioning. An identical companion specimen with conventional reinforced concrete was also tested. For the same input motions, both specimens achieved nearly identical peak displacements, but the partially prestressed column nearly re- centered even though the conventional column had substantial residual displacement. However, the tests demonstrated that the partially prestressed column tended to concentrate damage in a short plastic hinge length and as such, the mild longitudinal reinforcement proved to be susceptible to catastrophic failures due to low cycle fatigue. Earlier analytical studies suggested this possibility, and that by debonding the mild reinforcement in the vicinity of the plastic hinge that a more robust connection can be achieved. As part of the Year 8 activities, a set of 4 concrete columns have been constructed and tested. One resembled the partially prestressed column tested previously, but with the ground motion record amplitudes adjusted to better represent design criteria. The second specimen was similar to the first but with the longitudinal mild reinforcement unbonded over a length of about 2D. The third was similar to the second, but the level of post-tensioning was increased. The fourth specimen was similar to the second, but with the extra confinement required for these columns provided by a steel jacket (which makes the column less likely to need repair following an earthquake). These tests were completed towards the end of Year 8 and the results are being analyzed during Year 9.

It was intended that one of the columns utilized an Engineered Cementitous Composite (ECC) instead of concrete. A PEER intern working with Prof. Billington at Stanford worked with another intern at Berkeley to assess the feasibility of using the ECC material. A wide range of concrete cylinders having different degrees of spiral confinement were prepared and cast using conventional concrete, ECC and a hybrid multi-scale concrete. The ECC test results were disappointing, indicating significant constructability problems, and inconsistent test results. During Year 8 we are working to assess improved methods for constructing columns with high performance concrete, with special emphasis on the hybrid concrete composite materials. However, the focus of the work is on improving analytical models of the column behavior, which to date is considered inadequate in terms of assessing residual displacement and local damage, and developing and demonstrating the adequacy of design methods using conventional concrete. Comparison of test results with one another and with numerical predictions will help understand the capabilities of these types of columns, and the accuracy of OpenSees in predicting response.

It is planned to work closely with other investigators working on this subject. For example, Dr. Billington is assessing DM/DV and EDP/DM relationships for bridge piers containing unbonded post-tensioned reinforcement and high performance, cement-based, ductile concrete. She will be using our results and has been supplying mix designs for construction. Dr. Stojadinovic is carrying our extensive simulations to understand the contribution of various structural and ground motion characteristics to bridge fragilities, and further extending the PEER methodology as it applies to bridges. Marc Eberhard and Sashi Kunnath are working on improving OpenSees modeling capabilities, and we will work closely with them.

Brief Description of previous year's achievements, with emphasis on accomplishments during last year (Year 8)

As noted above, 4 columns using the proposed self-centering technology were constructed and tested on the Berkeley shaking table. These results substantiate the ability of these partially prestressed concrete columns to achieve displacements similar to those developed by conventional bridge columns, but to have residual displacements substantially smaller. It was found that unbonding of the mild reinforcement in the plastic hinge region softens the column and results in slightly larger lateral displacements. By increasing the amount of post-tensioning, stiffness is increased to desirable levels, and residual displacements are very small, even for large earthquakes. However, during the maximum considered earthquake shaking, the amount of spalling and the degree of bar buckling is increased with larger post-tensioning forces. It was found that a partially prestressed concrete column with unbonded post-tensioning and mild reinforcement (in the plastic hinge region) which has the plastic hinge region confined by a steel jacket has similar or smaller displacements compared to a conventional reinforced concrete column, has residual displacements less than 0.5% when the peak displacement ductilities exceeded 10 (this is less than a tenth of what was retained by a similar RC column), and very little visible damage is apparent following even the maximum considered events.

To date, analyses using OpenSees and other programs have been able to predict the peak lateral displacements of the test specimens fairly well. However, prediction of post peak response, and in particular the residual displacements, has been poor. Thus, we are continuing to work on this issue in conjunction with others (Eberhard, Billington, Filippou, etc.).

The disappointing results of the feasibility tests with the ECC material resulted in a decision not to build a full test specimen constructed with ECC. However, additional work on cylinders and in the construction of a small-scale pilot test specimen is planned for the remainder of Year 9. The main thrust of Year 9 is to learn from the results of the 4 tests conducted in Year 8 and design, construct and test a small bridge system that employs the self-centering columns. This is based on a specimen using conventional columns that was tested in Year 7. The new specimen is in construction and will be tested during the summer 2006. The specimen includes columns with steel jackets that confine the plastic hinge region at the top and bottom of the columns. The structural system employed is such that the specimen develops frame action in the longitudinal direction and plastic hinging would be expected at the top and bottom of the specimens, and that the specimen would exhibit primarily cantilever action in the transverse direction. The two columns are different in length (by D/2) so that the specimen will develop a small amount of torsional response as well. This data will help demonstrate the capabilities of this concept for bridge systems, and be extremely helpful in assessing models for use in OpenSees for predicting post peak responses.

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

Prof. Billington is examining in a companion PEER project the effect of self-centering columns on the overall performance of bridges. Some related research is underway at UC Reno using prestressed precise columns (with no mild reinforcement) and use of shape memory alloys for the longitudinal reinforcement to help reduce the residual displacements of columns. At the University of Washington, Profs. Eberhard, Stanton and others are exploring the use of precast, post-tensioned columns. In Japan, there is a significant effort related to shaking table testing of conventional concrete columns being begun by PWRI, E-Defense, and the Tokyo Institute of Technology. We will keep abreast of these and other related efforts.

Describe any instances where you are aware that your results have been used in industry

Several instances of partially prestressed concrete columns have been found in Taiwan and in a few instances in the US.

Expected milestones & deliverables

We expect to begin construction of the two self-centering column test specimen in Spring 2006 and test it during the Summer 2006. We will carryout additional work on the high performance concrete materials during the Summer 2006. The analytical research using OpenSees is continuous. A report on the first two specimens tested is in final stages of preparation and will be released as a PEER report soon. A report on the results of the next four specimens and the two column specimen will be part of the thesis of graduate student H. Jeong, expected to be completed in Year 10. A series of papers on the findings of the research have and are continuing to be prepared and published.