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
The objective of this project is to establish guidelines for inertial and kinematic load combinations to be used in simplified pushover analyses of pile foundations in liquefied and laterally spreading ground.
This project supports PEER's mission of providing performance-based methodologies for bridge systems. It addresses an important component of emerging design methodologies that has not been addressed.
OpenSees-based simulations will be validated against recent centrifuge model tests and then used in expanded parametric studies to develop practical guidelines on inertial/kinematic load combinations to use with simplified pushover design methodologies.
The dynamic centrifuge model test data are from a Caltrans funded project at UCD. The most recent archived test results involved superstructures with fixed-base fundamental periods of 0.3 and 0.8 seconds, supported on six-pile groups embedded in a soil profile of non-liquefiable clay over loose liquefying sand over dense sand. The soil profile sloped toward an adjacent open channel, and the ground subsequent spread laterally toward this channel during earthquake shaking. Each model was shaken with a series of earthquake motions that included weak and strong-scaled versions of a motion from Port Island in 1995 and Santa Cruz in 1989. The structural models were highly instrumented to enable detailed back-calculation of load-transfer mechanisms.
The first task will be to summarize findings and design implications from the back-calculated load-transfer time histories from the centrifuge test data. The back-calculation work was largely completed as part of the Caltrans project, but some additional work is required to fully interpret the results from all ten earthquake events and summarize the design implications.
The second task will be to validate the OpenSees simulations against the centrifuge models with superstructures during lateral spreading. The ability of the OpenSees simulations to capture the relative phasing of the inertial and lateral spreading loads during earthquake shaking is of particular interest in these validation studies.
The third task will be to parameterize the OpenSees simulations to cover a broader range of structural parameters, soil profile parameters, and earthquake ground motions. These parametric studies will focus on developing guidelines for inertial and kinematic load combinations for determining peak loads and displacements experienced by foundations and superstructures in areas subjected to liquefaction and lateral spreading.
The continuing interpretation of the back-calculated time histories of various load components from the dynamic centrifuge model tests has clarified some important features of behavior. These results are showing that the peak lateral loading conditions occur mainly during strong shaking and are very close (within 10%) to the sum of the peak kinematic and peak inertial loads alone. Some researchers have argued that lateral spreading loads could be assumed to occur late in, or after, shaking when ground displacements are reaching their maximum values, and therefore the peak inertial and lateral-spreading loads could be analyzed as occurring at different times. However, the back-analyses of the centrifuge tests are showing that this is not the case. The occurrence of liquefaction causes the surface crust layer motions to have much longer period components, with the consequence that the inertial and lateral crust loads are nearly in phase during peak loading cycles. This finding is consistent with recent large shaking table studies in Japan.
The OpenSees work will build upon related efforts for lateral spreading load analyses. Chang has been making progress in the development of the FE models and parameter calibrations as well.
This project utilizes data obtained in a largely-completed Caltrans-funded project involving dynamic centrifuge model tests of piles in liquefied and laterally spreading ground, and the PEER core project by Boulanger on simulation of soil-foundation-structure interaction for deep foundations (2312003). The Caltrans project provided essential experimental data on inertial/kinematic load combinations during lateral spreading and the project 2312003 provided OpenSees tools to model this behavior. This project differs in that it focuses on developing guidelines for inertial and kinematic load combinations based on detailed parametric studies via OpenSees modeling.
This project has a very specific short-term deliverable and is not expected to continue itself in Year 8.
There are other closely related projects that are expected to start/continue in Year 8. These projects include additional dynamic centrifuge modeling, OpenSees modeling, and design methodology development related to the performance of bridge abutments subject to lateral spreading.
The first milestone will be a summary document on the findings from the back-analyses of the centrifuge test data. This task should take about 1 person-month to complete.
The second milestone will be a comparison of the dynamic time-history analyses against the centrifuge models with superstructures during lateral spreading. This task is expected to take about 3 person-months to complete.
The third milestone will be a report summarizing the results of the parametric study and the resulting guidelines for inertial and kinematic load combinations for determining peak loads and displacements experienced by foundations and superstructures in areas subjected to liquefaction and lateral spreading.