Thrust Area IV – Simulation & Information Technologies

Goals

A central requirement of PEER's research mission on performance-based earthquake engineering methodology is the need to simulate the performance of structural and geotechnical systems. The simulation models must represent the modes of behavior and types of damage that are ultimately important in framing decisions for stakeholders. There are substantial problems and open questions on how to model the highly nonlinear behavior of structural systems with degrading components, or soil undergoing large deformation because of liquefaction, and the interaction between foundations and soils during large deformation. To address these challenges, the rapid advances in information technology can be used in developing the next generation of earthquake engineering simulation applications and also in educating the next generation of earthquake engineers. These advances include high-end computers for solving large-scale problems; databases for searching for new information from experimental data, simulation data, or observed data such as ground motion and field data; and visualization technology for providing engineers, design professionals, and stakeholders understanding about the performance of their systems.

The goal of Thrust Area IV is to develop new simulation models and new methods for performance-based earthquake engineering assessment and design methodologies, to develop modern simulation software tools taking advantage of information technology advances, to deliver the software tools to the community, and to educate students in simulation methods and information technology applications in earthquake engineering. The goal of this thrust area continues through the re-organization of the research program in Year 7 with the application focus spanning building systems (TA I) and bridge systems (TA II). Lifeline systems are considered to a lesser extent, but provide a fertile future area, particularly as lifeline systems research moves toward consideration of lifeline networks. The incorporation of uncertainty in the simulations is essential, and the research in this thrust area has resulted in important developments in the methods and software for reliability computation.

The principal software technology to support all of these activities is the Open System for Earthquake Engineering Simulation, "OpenSees," which has enabled research on simulation and provided a platform for PEER participants and others to conduct advanced simulations. The OpenSees software framework uses object-oriented methodologies to maximize modularity and extensibility for implementing models for behavior, solution methods, and data processing and communication procedures. The framework is a set of inter-related classes, such as domains (data structures), models, elements (which are hierarchical), solution algorithms, integrators, equation solvers, and databases. The classes are as independent as possible, which allows great flexibility in combining modules to solve simulation problems for buildings and bridges, including soil and soil-structure-foundation interaction, and most recently including reliability computational modules. The open-source software is managed and made available to users and developers through the OpenSees website at http://opensees.berkeley.edu.

As an advanced platform for computational simulation, OpenSees provides an important resource for the National Science Foundation-sponsored George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES), and it has now been adopted by NEES Inc. and the NEES information technology services (NEESit) as the NEES simulation component. PEER will be providing the maintenance and operations for use of OpenSees in NEES through a subaward with UCSD's San Diego Supercomputer Center. The NEES decision to utilize OpenSees and incorporate it in the NEESit suite of services for earthquake engineering research will increase the user base and the range of simulation applications for the software. The modular design of OpenSees means that it can be customized for integrating physical and computation simulation through data repositories, visualization, and hybrid control for advanced experimental methods, all of which meet important NEES objectives. OpenSees has proven to be an excellent platform for a new generation of hybrid simulations—combination of physical testing and computational simulation—which will significantly enable new types of experimentation and collect valuable data about the seismic performance of systems. With the broader community support through NEES, OpenSees provides long-term opportunities that include: (1) improvement of model- based simulation using data from advanced experimental facilities, (2) extensions to include grid-based and other high-end computing for earthquake engineering, and (3) integration with structural health-monitoring systems using widely distributed MEMs sensors and processors.

Strategic Research Plan, Milestones and Deliverables

Figure 2.17 shows the strategic research plan for TA IV, emphasizing Years 7–10 and identifying the system-level integration milestones. The first six years of research in the thrust area were largely devoted to the development of new models and computational methods needed for structural and geotechnical simulation and implementation in the OpenSees software framework. The testbed projects in Years 5–7 provided an opportunity to expand the usage of OpenSees, identify problems as it was used for simulation in the building and bridge testbeds, incorporate improvements, and identify future research and development needs. OpenSees is currently in version 1.6.2, which was released in April 2005. As a result of much user experience within PEER and by the broader community, improvements have been made in solution robustness, testing combinations of element models and solution algorithms.

NEESit efforts are addressed by high-end computing and hybrid experimental methods using the simulation technology, and visualization, all of which are important for NEES. Additional capabilities will be released early in summer 2006 with the latest version of parallel solvers.

For Years 8–10, the strategic plan for TA IV is divided into three categories: Modeling, Simulation System and Platform, and Integrated Applications. These areas are described below.

Much of the model development and implementation research will be completed by the end of Year 9, leaving model validation to be completed in Year 10, including structural models for degrading cyclic behavior of RC components (including shear interaction in columns and joint behavior); improving models for low-cycle fatigue, bar buckling, and fracture; and understanding how these behaviors are affected by loading history. Year 10 modeling research will be on evaluating RC systems at incipient collapse and the validation of system models using experimental data such as from shake table tests. The other modeling thrust is to develop improved models for nonlinear response and soil liquefaction suitable for large-scale simulation, with substantial challenges in modeling SFSI for large-diameter shafts and bridge abutments to address needs in TA II. These two areas, among others, remain a topic for further experimental research and computational validation, and include major 3D response mechanisms that must be accounted for. The results of this research will provide insights that can translate into design revisions, with most significant economical outcomes (in view of the involved large expenditures on these two bridge components). Overall, the modeling research contributes to the milestones SFSI, EDP-DM-DV relations for building and bridge systems, and enhanced performance models.

The second category is Simulation System and Platform. Through the collaboration between PEER and NEESit, we will integrate OpenSees with the NEESit data repositories, which are currently being revised from the NEESgrid versions. This will provide OpenSees users the ability to access NEES data on experiments and simulation data, and to upload simulation results into the repository. In addition, we will address what has become an important need: providing integrated PBEE tools based on advanced simulation. To meet this strategic need, new projects are being initiated to focus on the application of advanced simulation using OpenSees for geotechnical systems, high-rise buildings, and bridge structures. These application studies will involve PEER's industry partners, who are leading efforts in improved simulation for engineering practice, and in developing projects for application of OpenSees to demonstrate its capabilities for difficult seismic projects. The goal of these is to both further develop the simulation tools while supporting a cadre of early adopters of these technologies from the practicing engineering community.

Strategic plan: Thrust area IV — Simulation and information technologies

Finally, PEER will convene the first international symposium on OpenSees in summer 2007. This follows the annual OpenSees workshops and the successful developer symposium in 2005. The objective of the international symposium is to communicate and exchange the large amount of development and application of OpenSees over the past several years. Participants will submit papers and the proceedings will be published as a PEER report.

Critical Mass and Level of Effort

The research team for TA IV includes experts on modeling for reinforced concrete components and systems and geotechnical systems. For development of the software framework, several of the thrust area researchers have computer science backgrounds, and in many cases collaborate with computer scientists on research related to the simulation framework. As the simulation methods are being used in the bridge and building testbeds, PEER researchers and industry partners are providing feedback on the effectiveness of the research products in simulation and the usefulness of the databases. Many of the graduate students conducting research in the thrust area are taking courses in computer science, generally as a minor program of study. This breadth of graduate education in computer science is unusual in earthquake engineering, and it has brought new technology and computer science methods into the PEER research program.

Over the past three years, we have developed important collaborations with the George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES). The NEES system integration project has selected OpenSees as the simulation component for the NEESgrid. In addition to the core simulation capability, PEER is contributing to the development of data models for simulation data for use in the NEESgrid data repositories, a web-based portal for simulation services, and porting of OpenSees to grid-based computing resources. In collaboration with the NEESit group at the UCSD Super Computing Center, the OpenSees development team at UC Berkeley has a contract with the NEES Consortium to provide ongoing maintenance and operation of the simulation component. This support, along with PEER's continuing commitment to simulation and information technology, will expand the users and development opportunities for OpenSees.

Research Advances and Deliverables

Highlights of accomplishments in Year 9 include:

  • - Soil-foundation-structure interaction in bridge systems, including deep foundations in liquefiable soil and new research on shallow foundations.
  • - Component models for reinforced concrete with an initial examination of damage measures for performance evaluation. A new plastic hinge model that provides objective response for degrading behavior.
  • - Simulation for reliability computation, including exact computation of response gradients for highly nonlinear systems.
  • - Database applications to support simulations for the bridge and building application (testbed and benchmarking) projects.
  • - Completion of a collaboration with seismologists and computer scientists to develop an integrated methodology for understanding the Seismic Performance of an Urban Region (SPUR).
  • - Application of OpenSees to hybrid experimental-computational simulation, including use of grid-based communication, and demonstration of a hybrid test at the University of Kyoto as controlled by OpenSees running at UC Berkeley.

Over the past three years, significant effort in the thrust area has been devoted to the support of the simulations in other PEER projects using OpenSees. The support entailed the following activities: (a) training of students and researchers on OpenSees; (b) improvement of OpenSees user documentation; (c) assistance with development of models and scripts; (d) responding to bug reports and technical assistance; and (e) review and feedback of experience with OpenSees models, facilities, and computational efficiency.

In combination with application studies of TA I and II, OpenSees models are being evaluated against test data from large-scale experiments. In one case, soil continuum models for simulating ground deformations are being evaluated against a large-scale test in Japan, where explosives were used to trigger liquefaction in a test field containing pile foundations and a buried pipe (Ashford 2342003). In another case, OpenSees frame models have been validated against a full-scale pseudo-dynamic frame test, results of which are made available through collaboration with the National Center for Research in Earthquake Engineering (NCREE) in Taiwan (http://rcs.ncree.gov.tw/). In several TA II projects, simulation results from OpenSees are being extensively compared and validated against data from previous and ongoing tests of RC beam-columns (Mahin 2402004, Billington 2462004, and Eberhard 2452004); and in TA I, shallow foundation models have been implemented and compared to centrifuge test data (Hutchinson and Kutter 1352004).

Year 9 has seen the completion of a number of efforts for the models and computational features of OpenSees. A range of hierarchical models for beam-column elements is now available, including flexure, axial, and shear effects (Fenves and Filippou) and generalized hinges (Deierlein). The models include material and component behavior for cyclic degradation and large-displacement analysis. To support reliability and other applications, a new efficient algorithm for computing the response sensitivity for force-based elements has been developed and implemented (Fenves and Filippou). In addition, a beam-column element using force-based interpolation has been developed that is objective under degrading behavior, which had been an open problem. To solve large-scale systems with degrading components, a new quasi-Newton solution method based on a Krylov subspace has proven to be very efficient and robust when used in the testbed projects. New models under development include reinforcing bar buckling (Kunnath 4232004) and improved building collapse analysis (Mosalam 4252004).

Continued progress has been made with integrating reliability computation into OpenSees. Der Kiureghian has extended the first-order reliability method, and many of the element and material models now support direct differentiation for computing response sensitivities for reliability computation. The research has also made progress on importance sampling for Monte Carlo simulations and extending a library of distributions and correlation structures for random variables. Conte has used these methods to begin probabilistic evaluation of the Humboldt Bay bridge with the completion of a complete model of the SFSI system. In addition, significant sensitivity analysis procedures have been developed this year for a class of nonlinear plasticity- based soil models for seismic applications. Progress on these projects responds to concerns raised in previous years' site visit reports about the need in OpenSees for reliability tools that facilitate application of the PEER PBEE methodology and are not generally available in other earthquake analysis software.

Future Plans

Support and continued development for OpenSees will continue as a high priority, given the central role OpenSees plays as an enabling technology in PEER. During Year 10, the substantial progress in OpenSees software will be integrated into the framework (Fenves 4102004). Version 2.0 will include advanced capabilities using the parallel computing resources at SDSC.

Model development for RC members will continue with cyclic degradation of RC members including low-cycle fatigue (Kunnath 4232004). There will be increased focus on RC building systems, with new research on simulation for incipient collapse (Mosalam 4252004) and validation of system models using shake table data (Moehle 4282004). For geotechnical models, Elgamal (4242004) will begin research on modeling and simulation of large-diameter pile shafts and abutments for bridge systems, and Jeremic´ (4262004) will develop coupled (solid-fluid) models for liquefiable soils and large-scale simulations. These efforts integrate the structural and geotechnical elements of OpenSees and address topical challenges in seismic SFSI research. Conte (4132004) will conduct such integrated studies (PBEE framework applied to the Humboldt Bay bridge Testbed), and further introduce sensitivity analysis tools for geomechanics applications.

Computational reliability research will continue with Der Kiureghian (4142004) beginning research on non-ductile systems based on the completion of methods for ductile systems, and Conte (4132004) developing reliability methods for large-scale models of SFSI systems.

Finally, we will have news application impact projects (4272005) to demonstrate the OpenSees capabilities for geotechnical systems, buildings, and bridges. BIP members will be identified to carry out these projects. PEER expects that these projects will substantially speed the adoption of OpenSees in earthquake engineering practice. To communicate the developments of OpenSees, an international symposium will be held to disseminate recent work and discuss future directions.

Publications

  • Barbato, M. and Conte, J. P. 2005. Smooth versus non-smooth material constitutive models in gradient-based optimization for structural reliability analysis. Proc., 9th International Conference on Structural Safety and Reliability, June 19–23, 2005, Rome, Italy.
  • Barbato, M., Gu, Q., and Conte, J. P. 2006. Response sensitivity and probabilistic response analyses of reinforced concrete frame structures. Eighth National Conference on Earthquake Engineering, April 18–22, 2006, San Francisco, California, USA.
  • Barbato, M., Conte, J. P., and Peng, B.-F. 2004. Response mean upcrossing rate for linear MDOF systems subjected to fully nonstationary earthquake ground motion model. Proc. of the 13th World Conference in Earthquake Engineering, Paper No. 2977, Vancouver, Canada, 2004.
  • Barbato, M., Zona, A., and Conte, J. P. 2004. Nonlinear finite element response sensitivity analysis of steel-concrete composite beams. Proc., 9th ASCE EMD/SEI/GI/AD Joint Specialty Conference on Probabilistic Mechanics and Structural Reliability, July 26–28, 2004, Albuquerque, New Mexico, USA.
  • Cheng, Z. and Jeremic, B. 2005. A return mapping algorithm for isotropic and anisotropic large deformations. Third M.I.T. Conference on Computational Fluid and Solid Mechanics. The Massachusetts Institute of Technology, Cambridge, MA, U.S.A., June 14–17, 2005.
  • Der Kiureghian, A., and Fujimura, K. 2005. A non-parametric linearization method for nonlinear random vibrations. Keynote paper, Proc., Structural Engineering Convention, Bangalore, India, December 2005.
  • Der Kiureghian, A., Haukaas, T., and Fujimura, K.. 2005. Structural reliability software developed at the University of California, Berkeley. Proc. 9th Intl. Conf. On Structural Safety and Reliability, Rome, Italy, June 19–23, 2005, Milpress, Rotterdam (CD-ROM, pp. 1807–15).
  • Der Kiureghian, A. 2005. Non-ergodicity and PEER's framework formula. Earthq. Engrg. Struct. Dyn. 34:1643–52.
  • Der Kiureghian, A., Haukaas, T. and K. Fujimura, K. 2006. Structural reliability software at the University of California, Berkeley. Structural Safety 28(1–2):44–67.
  • Elgamal, A., Lu, J., and Yang, Z. 2005. Liquefaction-Induced settlement of shallow foundations and remediation: 3d numerical simulation. J. Earthq. Engineering 9 (Special Issue 1): 17–45.
  • Elgamal, A., Lu, J., and Yang, Z. 2005. Application of numerical methods to the analysis of liquefaction. Proc. 11th International Association of Computer Methods and Advances in Geomechanics, Torino, Italy, June 19–24.
  • Elwood, K. J., and Moehle, J. P. May 2006. Idealized backbone model for existing RC columns and comparisons with FEMA 356 criteria. Proc. Los Angeles Tall Buildings Structural Design Council 2006 Annual Meeting.
  • Elwood, K. J., and Moehle, J. P. 2005. Axial capacity model for shear-damaged columns. Structural J. 102(4)(July): 578–87 .ACI.
  • Fenves, G. L., Park, J. and Stojadinovic´, B. 2006. Modification of design strength for directivity effects. 8th National Conference on Earthquake Engineering, San Francisco, Paper No. 8NCEE-000444, April 2006.
  • Fenves, G. L. Park, J., and Stojadinovic, B. 2006. Parallel simulation methods for regional seismic response analysis. Proc. 8th National Conference on Earthquake Engineering, San Francisco, Paper No. 8NCEE-001349. April 2006.
  • Fujimura, K., and Der Kiureghian, A. 2005. FORM approximation of stationary first-passage probability for nonlinear systems. Proc. 9th Int. Conf. On Structural Safety and Reliability, Rome, Italy, June 19–23, 2005, Milpress, Rotterdam (CD-ROM, pp. 1939–1946).
  • Ghannoum, W., Shin, Y. B., and Moehle, J. P. July 2005. Collapse Tests of lightly confined reinforced concrete columns and frames. NEES-EDEFENSE Workshop on Collapse Simulation of Reinforced Concrete Buildings, Berkeley, California.
  • Haukaas, T., and Der Kiureghian, A. 2005. Parameter sensitivity and importance measures in nonlinear finite element reliability analysis. J. Engrg. Mechanics 131(10): 1013–1026. ASCE.
  • Haukaas, T., and Der Kiureghian, A. 2006. Strategies for finding the design point in nonlinear finite element reliability analysis. Probabilistic Engrg. Mechanics 21(2):133–147.
  • He, L., and Elgamal, A. 2006. Lateral load on piles due to liquefaction-induced lateral spreading during one-g shake table experiments. Proc. 100th Anniversary Earthquake Conference Commemorating the 1906 San Francisco Earthquake, San Francisco, CA, April 18–22.
  • Jeremic, B., and Cheng, Z. 2005. Significance of Equal principal stretches in computational hyperelasticity. Communications in Numerical Methods in Engrg. 21(9)(September): 477–86.
  • Jeremic, B., and Preisig, M. 2005. Seismic Soil-Foundation-structure interaction: numerical modeling issues. ASCE Structures Congress 2005, New York, NY, U.S.A., April 20–24, 2005.
  • Koo, H., Der Kiureghian, A., and Fujimura, K. 2005. Design-point excitation for nonlinear random vibrations. Probabilistic Engrg. Mechanics 20:136–147.
  • Kunnath, S. , Jeremic, B., Eberhard, M., Der Kiureghian A., and Porter, K. 2005. Application of the PEER performance-based methodology for seismic assessment of the I-880 viaduct. Caltrans Bridge Research Conference 2005, October 31st – November 1st 2005, Sacramento, California. (paper # 04-503).
  • Kunnath, S K, Larson, L., and Miranda, E. 2006. Modeling considerations in probabilistic performance-based seismic evaluation: case study of the I-880 viaduct Earthq. Engrg. & Struct. Dyn. 35(1). 2006.
  • Lu, J., Elgamal, A., and Yang, Z. 2005. Pilot 3D Numerical simulation of liquefaction and countermeasures. Proc. ASCE Geo-Frontiers 2005, Austin, TX, January 24–26.
  • Lu, J., and Elgamal, A. 2005. A 3D finite element user-interface for soil-foundation seismic response. Proc. of the Caltrans Bridge Research Conference, Sacramento, CA, October 31– November 1.
  • Lu, J., Elgamal, A., Law, K. H., and Yang, Z. 2006. Parallel computing for seismic geotechnical applications. Proc. ASCE GeoCongress 2006, Atlanta, GA, February 26–March 1.
  • Moehle, J. P. 2005. Nonlinear analysis for performance-based earthquake engineering. The Structural Design of Tall and Special Buildings, V. 14, No. 5, December 2005, pp. 385–400. http://www3.interscience.wiley.com/cgi-bin/jhome/102522232
  • Moehle, J. P., Ghannoum, W., and Bozorgnia, Y. 2005. Collapse of lightly confined reinforced concrete frames during earthquakes. Proc. Intl. Closing Workshop, NATO SfP Project 977231: Seismic Assessment and Rehabilitation of Existing Buildings, Istanbul, Turkey, 30 May – 1 June 2005
  • Moehle, J. P., and Yang, T. Y. 2005. Next-generation performance-based seismic assessment procedures. Proc. ATC 15-10, Applied Technology Council Kobe, Japan, October 2005.
  • Moehle, J. P., Ghannoum, W., and Bozorgnia, Y. 2006. Collapse of lightly confined reinforced concrete frames during earthquakes. Proc. 8th US National Conference on Earthquake Engineering, San Francisco, California, April 2006.
  • Moehle J. P. 2006. Seismic analysis, design, and review for tall buildings. Proc. Los Angeles Tall Buildings Structural Design Council 2006 Annual Meeting, May 2006.
  • Moehle, J. P., and Kunnath, S.K. 2006. Modeling cyclic degrading behavior in RC frame structures. Paper 639. Proc. 8th US National Conference on Earthquake Engrg. 2006. EERI.
  • Neeman, A., Jeremic, B., and Pang, A. 2005. Visualizing tensor fields in geomechanics. IEEE Visualization Conference (Vis-05), October 23–28, 2005 Minneapolis-Saint Paul, Minnesota.
  • Saritas, A., Filippou, F. C. 2005. Failure Analysis of shear critical RC structural members. Proc. 5th International Conference on Computation of Shell and Spatial Structures, June 1–4, 2005 Salzburg, Austria.
  • Takahashi Y. and Fenves, G. L. 2006. Software framework for distributed experimental- computational simulation of structural systems. Earthq. Engrg. & Struct. Dyn. 35: 267–281.
  • Yang, Z., and Jeremic, B. 2005. Study of soil layering effects on lateral loading behavior of piles. J. Geotech. & Geoenv. Engrg.131(6)(June): 762–70. ASCE.
  • Yang, T.Y., Moehle, J. P., Stojadinovic´, B., Der Kiureghian, A. 2006. An application of PEER performance-based earthquake engineering methodology. Proc. 8th US National Conference on Earthquake Engineering, San Francisco, California, April 2006.
  • Yang, T. Y., and Moehle J. P. 2005. Methodology for Building-Specific Performance Assessment, Proc., ANCER 2005, Korea, August 2005.
  • Zona, A., Barbato, M., and Conte, J. P. 2004. Finite element response sensitivity analysis of continuous steel-concrete composite beams. Proc. 2nd International Conference on Steel & Composite Structures, September 2–4, 2004, Seoul, Korea.

For additional details, see the PEER Annual Report - Volume 2 (PDF file - 7.6 MB).