Project Title/ID Number Life Sciences Testbed Simulation—3242002
Start/End Dates 10/1/02—9/30/03
Project Leader Khalid Mosalam (UCB/Faculty)
Team Members Tae-Hyung Lee (UCB/Grad Student)
Project goals and objectives

Develop a state-of-the-art computational model and perform nonlinear dynamic analysis of the UC-Berkeley Life Sciences building using the OpenSees simulation platform to demonstrate PEER performance-based earthquake engineering assessment and design methodologies. The project has dual objectives:

  1. Testbed for simulation in OpenSees using actual buildings, and

  2. Developing floor displacement and acceleration time histories to be utilized in testing building contents

Figure 1: Structural plan view of LSA building
Larger View
Role of this project in supporting PEER’s vision
The project is a realistic demonstration for the computational platform of PEER, namely OpenSees. Moreover, through the simulations, assessment of the importance of a number of uncertain variables will be conducted. This includes not only ground shaking details but also mass, damping, modeling assumptions (e.g. 2-D versus 3-D) and component force-deformation behavior of the selected building.

Figure 2: OpenSees 2-D model of LSA building
Larger View
Methodology employed
  1. Collect information about the structural system of the testbed, including member sizes, foundation type, material properties, and configuration and layout of any non-structural components.

  2. Survey available modeling options in OpenSees for representing structural elements including current nonlinear elements for geometric and material effects. This survey considered modeling issues related to soil conditions and interfaces between structural and non-structural elements.

  3. Develop a 2-D model where emphasis is on the structural system and boundary conditions using the available options in OpenSees.

  4. Conduct time history analysis of the several available ground motions developed for the building site to address uncertainty in the ground motion details.

  5. Generate floor time histories of relative and absolute displacements and accelerations for use by the researchers (at UCB and UCI) conducting the shake table experiments on the LSA building contents.

  6. Extend the structural model to 3-D to account for multi-directional ground motion effects and have more realistic distribution of mass and lateral stiffness.

  7. Simulate the 3-D model of the building with different properties addressing variability in mass, damping, and force-deformation relations of the different structural components comprising the 3-D model.

  8. Explore the application of the finite element response sensitivity analysis for LSA using OpenSees-Reliability developed by T. Haukaas & A. Der Kiureghian.

Figure 3: Description of ENT material to simulate soil behavior in OpenSees
Larger View

Figure 4: Spring forces in soil springs due to 50% in 50 years Coyote Lake earthquake
Larger View

Figure 5: Relative displacements of Roof due to 50% in 50 years, Coyote earthquake
Larger View

Brief description of past year’s accomplishments and more detail on expected Year 6 accomplishments
  1. 2-D working nonlinear model in OpenSees of the coupled shear wall lateral load-resisting system in the transverse direction (normal to the Hayward fault) as shown in Figures 1 and 2 (Year 5).

  2. Account for modeling foundation and soil conditions of the 2-D model, refer to Figures 3 and 4 (Year 5).

  3. Develop floor accelerations and deformation to be used for the shake table tests of the building equipments using several ground motions developed by other researchers for the building site, sample results (one simulation out of thirty) are given in Figures 5, 6, and 7 (years 5 and 6).

  4. Perform deterministic sensitivity study (Year 6) where probabilistic hazard curve is used to determine the intensity measures in terms of spectral accelerations. This is followed by determining representative ground motion profiles. Finally, several sources of uncertainties were accounted for using their swings in the form of Tornado diagrams as shown in Figure 8.

  5. Simulate the 3-D model of the building with different properties addressing variability in mass, damping, and force-deformation relations of the different structural components comprising the 3-D model, Figure 9 (Year 6).

Figure 6: Relative accelerations of Roof due to 50% in 50 years, Coyote earthquake
Larger View

Figure 7: Absolute displacement of Roof due to 50% in 50 years, Coyote earthquake
Larger View


Figure 8: Tornado diagrams for different engineering demand parameters to demonstrate the importance of different sources of uncertainties
Larger View
Other similar work being conducted within and outside PEER and how this project differs
Other testbeds includes the study of the Van Nuys building and the I880 bridges. This project is unique in the fact it represents shear wall building designed according to relatively modern codes and standards. Moreover, 3-D models and deterministic sensitivity analysis are pursued in the current project focusing on performance of building content and non-structural elements, in addition to the structural ones.
Plans for Year 7 if this project is expected to be continued
Further account of uncertainties in ground motion and modeling and developing means of reducing them should be pursued. One major issue is the development of EDP’s for the performance of building content. This should be pursued in light of the tests conducted at UC-Berkeley and UC-Irvine. With this set of test results, possible development of representative limit states of the performance of the building content can be formulated and utilized with the developed computational model of the building to investigate losses due to damage to the building content. One specific issue is to develop tornado diagrams in terms of design variables (related to building content and non-structural components).

Figure 9: 3-D model of the Life Science Addition
Larger View
Describe any instances where you are aware that your results have been used in industry
Our generated floor responses have been extensively used in the tests of the building content at UC-Berkeley and UC-Irvine. This effort is utilized by Berkeley campus in collaboration with Rutherford and Chekene to develop retrofitting schemes of the building and particularly means of attaching its sensitive and valuable contents.
Expected milestones

The project main milestones are:

  1. May 2002: Complete the 2-D model of the building and its associated simulations
  2. Sept 2002: Identify the important uncertain variables based on the 2-D working model
  3. March 2002: Finalize the deterministic sensitivity analysis using the 2-D model
  4. April 2003: Complete the simulations using the 3-D model and investigate the important uncertainties
  5. June 2003: Address the modeling issues related to any non-structural components, such as partitions or building contents
  6. Sept 2003: Summarize results and produce the final report

The project main deliverables include:

  1. A working 2-D model of the building
  2. A working 3-D model of the building
  3. A complete suite of time histories of displacements and accelerations for the building floors under the effect of the developed ground motions for the site.

Results of the deterministic sensitivity analyses from the 2-D and 3-D models with ordering of importance of the different sources of the uncertainties affecting the structural performance of the building and its contents