5282003 - Database and Simulation Performance Criteria for Reinforced Concrete Columns

Project Title/ID Number	Database and Simulation Performance Criteria for Reinforced Concrete Columns—5282003
Start/End Dates	10/1/03—9/30/04
Project Leader	Marc Eberhard (UW/F)
Team Members	Mike Berry (UW/GS)

F=faculty; GS=graduate student; US=undergraduate student; PD=post-doc; I=industrial collaborator; O=other

Click on images to enlarge in a new window

1. Project Goals/Objectives:

As part of this project, we are developing and calibrating tools for assessing RC column seismic performance, including data and models for both seismically conforming (ductile) and non-conforming (brittle) columns. The project will lead to the evaluation, development and application of a variety of OPENSEES models of column performance. For year 7, our specific goals are:

To evaluate and improve the accuracy of OpenSees models of 1D column, hysteretic force-deformation response for both ductile and degrading members.
To include the effects of cumulative deformation demands in EDP-DM relationships for flexural damage accumulation.
To extend the 1D force-deformation and damage accumulation research results to columns subjected to biaxial deformations and/or variations in axial load.

2. Role of this project in supporting PEER’s mission (vision):

Accurate analytical models of force-displacement response and damage progression in reinforced concrete columns are needed to successfully implement Performance Based Earthquake Engineering. Damage to reinforced concrete columns can result in large economic losses, and in severe cases, structural collapse. Column models need to be sufficiently versatile to handle ductile and brittle columns, uniaxial and biaxial response, and the effects of repeated cycling on damage accumulation.

3. Methodology Employed:

The project has the following tasks for Year 7:

Although we have developed OpenSees models of columns within the PEER Structural Performance Database, the accuracies of various modeling strategies have not yet been evaluated systematically. In cooperation with Professors Kunnath, Fenves and Mahin, we will choose the details of the particular modeling strategies (e.g., element selection, number of integration points, plastic-hinge lengths) that we wish to evaluate. These selected strategies will then be applied to the columns of the database (starting with columns tested by Lehman), and we will evaluate the accuracies of the calculated force-displacement responses (monotonic and hysteretic).
Based on the results of the accuracy evaluation, we will recommend modeling strategies. This activity will be helped greatly the recently developed capabilities within OpenSees to evaluate model sensitivity
Once the key local deformation parameters (e.g., plastic curvature, strain growth, nominal strain-energy) are identified in OpenSees, EDP-DM relationships will be developed for these EDPs. These parameters will need to be more versatile than the simple parameters used so far in our research.
The EDP-DM relationships will be evaluated, and perhaps modified, to reflect the effects of cumulative deformations, variations in axial loads, and biaxial loading. Only a few test series are available to evaluate these capabilities (far fewer than the 450 columns in the database), but it is important that the final models incorporate these important effects, which are present in many columns in actual structures.
As we have done in the past, we will continue to update the PEER structural performance database. For example, we recently received data from PEER-sponsored tests performed at UCLA.

4. Brief Description of past year’s accomplishments (Year 6) & more detail on expected Year 7 accomplishments:

Structural Performance Database Documents Force Deflection Response
and Damage Development in Reinforced Concrete Columns

Accomplishments during Year 6:

We continued to update the PEER structural performance database. The PEER structural-performance database is now available on the PEER website at http://nisee.berkeley.edu/spd/. We also added a users' manual, which is available online.
We finished our work on linking column damage to simple Engineering Damage Parameters (EDP), such as the drift ratio, displacement ductility and plastic rotation. The developed EDP-DM relationships were then applied to the PEER I-880 testbed.
We documented our column research results in a PEER technical report, an ASCE Structural Journal manuscript and a series of conference papers. We are also prepared a chapter for the I-880 testbed report.
We developed OpenSees models to analyze the columns within the structural performance database. We are currently evaluating the accuracy of a variety of modeling strategies, developing modeling recommendations and plan to develop guidelines for estimating analysis accuracy. We are also studying the details of these models to identify local deformation measures that can be used to predict the likelihood of various levels of column damage, such as the onset of spalling and the onset of bar buckling.

5. Other Similar Work Being Conducted Within and Outside PEER and How This Project Differs:

Many researchers are using the database to calibrate their own models of column performance. However, with the exception of Prof. Pantazopolou (U. Athens), these researchers seem to focusing on strength loss, rather than observable damage. In any case, I do not know of similar efforts outside of PEER to calibrate column models implemented in Opensees.

This work will be coordinated closely with Prof. Kunnath, who is developing new material and member models, and Profs. Lehman and Stanton, who will be performing detailed tests to better the details of progression and damage accumulation. Any development of OpenSees capabilities will need to be coordinated with Prof. Fenves.

6. Plans for Year 8 if project is expected to be continued:

The XML capability of the database provides the opportunity to develop new applications. For example, OpenSEES could automatically identify columns similar to that being modeled analytically. The force-deformation response for that particular analytical model for that particular column element could then be compared with the observed experimental response. If the amount of images were increased, it would also be possible to automatically visualize the level of damage corresponding to a type of column and a level of deformation demand.

The structure of the database makes it possible to store a wide range of information, including drawings, text and photographs. To take advantage of this capability, such data would be collected for columns. The scope of the database could also be increased by adding tests of other elements, such as reinforced concrete joints, reinforced concrete walls and steel beams. Further developments should be coordinated with the development of the NEES data repository.

In the long run, this research effort will lead to coupling of the force-deformation response models and damage accumulation models. Damage progression would vary with the deformation history, and changes in force-deformation response would be tied to specific levels of damage. For example, the splice deformation and flexural capacity would vary with the level of axial load at a particular instant.

7. Describe any actual instances where you are aware your results have been used in industry:

The ACI-ASCE Shear and Torsion Committee (ACI-ASCE 445) is using the data to evaluate existing and proposed code provisions for shear design of reinforced concrete columns.

8. Expected Milestones & Deliverables:

The deliverables include:

A PEER report that documents the accuracy of OpenSees available column-modeling strategies, and provides modeling recommendations (December, 2004).
Performance models and recommendations for RC columns (implemented in OpenSees) (end of Year 7). These models will consider the effects of cumulative deformations, axial-load variation and biaxial deformations.
Although database development will no longer be a focus of our research, we will continue to update the website as new data become available. In particular, we will add data from the PEER-sponsored tests at UC San Diego and UCLA. The UCLA data has already been collected.