Project Title/ID Number Development of EDP-DM-DV Relationships for RC Building Components—3422003
Start/End Dates 10/1/03—9/30/04
Project Leader Laura Lowes (U Wash/F)
Team Members Nilanjan Mitra (U Wash/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:

The proposed research will develop EDP-DM-DV models for reinforced concrete beam-column joints. One set of models will enable the use of engineering demand parameters (EDPs), which are the traditional output of structural analysis, to predict component-specific damage measures (DMs). A second set of models will link component damage measures to repair techniques and provide a method for computing total repair effort. Total repair effort is considered a decision variable (DV) that is meaningful to building owners.

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

The results of this project will enable researchers and consultants to evaluate the economic impact of earthquake loading of a structure, using damage scenarios that include component-specific damage predictions.

3. Methodology Employed:

A year 4 project resulted in the development of probabilistic models linking joint deformation as observed in the laboratory with repair-specific damage states and then developing a method for predicting repair cost and time given joint damage. A year 2-3 project resulted in the development, and implementation within the OpenSees environment, of a model that can be used to predict the earthquake response of beam-column joints. This year-7 project extends the previous work in three critical ways:

  1. Analyses will be conducted using OpenSees and the previously developed beam-column models to simulate the response of the test assembles used to generate the probabilistic damage model. The results of these analyses will be used to quantify model inaccuracy. Then, the probabilistic model linking observed deformation with damage will be extended to provide a link between predicted deformation and the previously identified repair-specific damage states. In this way, modeling uncertainty will be included in damage prediction.
  2. The results of the year-4 project included a proposed method for predicting repair cost / time using information about component damage. This issue will be investigated further to improve the accuracy with which repair costs are predicted and explore different approaches to predicting economic impact. PEER BIPs will be used as a resource for this phase of the project; other PEER researchers will be consulted as well.
A year-5/6 project focused on simulation of the response of the Van Nuys Holiday Inn building using the component models available within OpenSees. This included explicit simulation of the inelastic response of beam-column joints using numerical models developed by the PI (year 2-3 project) as well as others (e.g., Deierlein). The results of phase 1 and 2 of the project will be demonstrated through prediction of the economic impact of damage resulting from earthquake loading of the Van Nuys Holiday Inn.

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

A year-6 project focused on simulation of the response of the Van Nuys Holiday Inn building using the simulation tools available within OpenSees. The results of this project include the following:

  1. Quantification of the impact of modeling uncertainty on prediction of engineering demand parameters for the Van Nuys Holiday Inn. Examples of numerical data generated as part of this study are provided in Figures 1a and 1b, which show the impact of modeling decisions on the predicted inter-story drift of the Van Nuys building for the “50in50” and “2in50” earthquake hazard levels. These data can be compared with the data in Figures 2a and 2b that show the impact of ground motion variability on predicted inter-story drift. (Click on image to enlarge)

    Figure 1a figure 1b
    Figure 1a: 50in50 eq. hazard level         Figure 1b: 2in50 eq. hazard level

    figure 2a figure 2b
    Figure 2a: 50in50 eq. hazard level         Figure 2b: 2in50 eq. hazard level

Evaluation of exiting models for predicting the response of older reinforced concrete buildings. An example of the numerical data generated as part of the study is provided in Figure 3, which shows predicted curvature ductility demands for beams and columns as well as the locations of column flexure-shear and splice failures in the Van Nuys Holiday Inn building under static lateral loading.

figure 3
Figure 3: Distribution of Damage and Yielding at Failure under Static Lateral Loading

Year 7 accomplishments will extend the results of the year-4 project and provide models for use in predicting beam-column joint damage. Expected outcomes include models linking predicted deformation with component damage. Such models will be similar in nature to that shown in Figure 4 that links observed inter-story drift to joint damage.

 figure 4
Figure 4: Probability of an Older Joint Exceeding a Repair-Specific
Damage State as a Function of Maximum Inter-Story Drift and Number of Load Cycles

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

I am unaware of similar work being done be researchers outside of PEER. Similar work is being done by researchers within PEER and includes the following:

  1. Modeling of beam-column joint response and correlation with damage: Lehman and Stanton at UW are developing simplified joint models for use with OpenSees and correlating predicted joint deformation with damage observed in the laboratory.
  2. Prediction of component damage: Miranda has developed probabilistic models linking structural deformation, as well as other engineering demand parameters that can be predicted using typical nonlinear analysis software, to component damage and loss.
The work proposed here differs from the Lehman/Stanton project in that 1) the numerical model used to predict joint response and the probabilistic models use to link joint deformation to observed damage are appropriate for use for a broad range of joint designs and 2) the work being done here focuses on linking damage with repair to developing models for predicting repair cost / time. The work proposed here complements the Miranda project in that Miranda has not, thus far, addressed damage and loss prediction for beam-column joints.

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

The project is not expected to continue in Year 8.

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

None known.

8. Expected Milestones & Deliverables:

The proposed research objectives translate into a series of research tasks. These tasks will be accomplished by the PI, one new graduate student researcher and one continuing graduate student researcher during a one-year period. Completion of each task is considered a milestone; deliverables are identified.

Task I: Working with an existing set of experimental data and OpenSees component models to develop models that link predicted EDPs with observed DMs, and thereby include modeling uncertainty. Deliverable: EDP – DM models.

Task II: Work with PEER BIPs and other PEER researchers to answer the following question:
How best can repair effort be modeled so that analysis results can be tied to total building repair effort? Deliverable: Method for predicting repair effort given prediction of building damage.

Task III: Interview structural engineers, contractors and consultants and review cost estimating resources to develop a model for computing the cost and time required to repair beam-column joint damage.

Task IV: Propose a method for computing total repair effort on the basis of component damage, as characterized by component-specific damage measures. Deliverable: EDP – DM - DV model.

back to top