Project Title/ID Number Advanced Seismic Assessment Guidelines—Lifelines 507
Start/End Dates 10/1/02—9/30/03
Project Leader Allin Cornell (Stanford/Faculty)
Team Members Charles Menun (Stanford/Faculty), Paolo Bazzurro (AIR/Industrial Collaborator), Maziar Matahari (Stanford/Grad Student), Gee Liek Yeo (Stanford/Grad Student), Joe Maffei (R&C/Industrial Collaborator)
Project goals and objectives

The primary objective of this two-phase project is the development of a new level of guidelines for the assessment of existing buildings. The scope is limited to typical PG&E 1-3 story buildings often of older and mixed construction. The limit states considered are those of direct electrical-system-reliability interest, e.g., collapse and (red or yellow) tagging, the latter because of their operability implications. An advanced state-of-practice engineering is employed, e.g., the structural engineer will provide a nonlinear (NL) static pushover analysis (SPO). In keeping with sponsor-stated needs and recent SAC and PEER developments, the guidelines will incorporate both aleatory and epistemic uncertainty measures, and the guidelines’ output products will include limit-state “fragility curves”, i.e., curves of the probability of limit state given ground motion intensity (IM) level, that reflect these various uncertainties. The first-year objective has been a “beta-version” of these guidelines, using results of only the first year’s research and default methods and parameter values where necessary. Nonetheless this first-year version is “operational”, i.e., up to or beyond current practice in all respects, such that it is being tested by two top-level structural engineering firms in the second year.

The objectives of the second-phase efforts are to guide, assist, monitor, and evaluate the “beta-testing” of the guidelines in order to critique the first version and to prepare ultimately a second, improved, final version of the guidelines. These improvements will also include the results of research into subjects identified in the first year as requiring basic study: improvements in the use of static methods to assess damaged capacity, aftershock-hazard-based tagging criteria/procedures, and joint treatment of the effects of the two horizontal components of ground motion.

Role of this project in supporting PEER’s vision
It is anticipated that PG&E will use these products in a seismic system-level probabilistic performance study of their network, as well as for their own in-house purposes.
Methodology employed

The engineer will use state-of-practice NL static methods. These have been improved to utilize a new “SPO2IDA” spread sheet that predicts, for an “arbitrary” backbone shape, the dynamic response to a complete range of ground motion intensities from static input, and to apply the procedure to various levels of the damaged structure as well. A novel aspect is that the method bases tagging (occupancy control) explicitly on the likelihood of collapse of the intact or damaged structure under aftershocks.

Brief description of past year’s accomplishments and more detail on expected Year 6 accomplishments

In the first year we have delivered a set of guidelines that incorporate a practical six-step procedure. In brief, the engineer prepares a static pushover (SPO) of the intact structure, from which he identifies a set of several potentially interesting, identifiable structural damage states. For each of these possible damage states the nonlinear static model is first deformed to that state, then unloaded and finally re-subjected to a static pushover; together this sequence represents (statically) the main shock followed by an aftershock. The aftershock capacity of each of these (mainshock-induced) damage states is determined, and used as a basis, along with one or more decision rules, to determine which of the damage states should be chosen to represent the onset of a condition that merits yellow tagging or red tagging, should such states be observed in the post-mainshock inspection. (These steps are the subjects of more detailed study in the Year 6; see below) The initial, intact SPO is then processed by a new “SPO2IDA” spreadsheet-based tool, developed in another recent project, which translates this static assessment into a median IDA (Incremental Dynamic Analysis).

From this result one can determine the two values of mainshock spectral acceleration (Sa) necessary to put the structure into the two damage states determined above to be the states associated with the onset of yellow and red tagging. These two Sa levels become the median values of the yellow and red tag fragility curves. See Fig. 1. (In fact this same spreadsheet tool is used to determine - from the “aftershock SPOs” of the damage states discussed above - the capacities of these states in Sa terms, i.e., what aftershock ground motion level would cause collapse and severe life loss should the building be occupied after the mainshock.) To complete the fragility curves one needs “betas”, measures of randomness and uncertainty. The record-to-record randomness “beta” is provided by the same spreadsheet. The remaining uncertainty, that due to the engineer’s limited information (materials properties, appropriate nonlinear model, etc.), must be obtained primarily by subject judgment. Based on the literature and systematic interviews with a panel of local engineers we have prepared tables of such values and detailed guidance to the engineer in assigning these values. (Dr. Joe Maffei, of Rutherford and Chekene, was retained by the project to assist in preparing this guidance.) Finally we prepared two quite detailed application case studies.


Figure 1: Median spectral acceleration capacity, Sa, cap associated with the structural limit states (or damage state, DS) yellow tag, red tag, and collapse, based on the “Static Pushover (SPO) to Incremental Dynamic Analysis (IDA)” tool.
Larger View


Two tests of the guidelines are currently underway by local engineering firms, Rutherford and Chekene and Degenkolb and Associates. We are monitoring and assisting these studies in order to improve and revise the final draft of the guidelines in Year 6. In addition, in Year 6 we have been focusing on several technical research problems related to the guidelines. One is the study of how best to modify the nonlinear static pushover method (NSP) to address the two-stage assessment of damage in a mainshock followed by collapse in an aftershock. This is being addressed by comparing the two-stage NSP assessments described above with the results of conducting dynamic assessments (IDAs) for multiple mainshock records followed by IDAs to collapse for multiple aftershock records of the structure after it has been left a specific damage state under the mainshock. A second study is the use of aftershock ground motion hazard analysis (PASHA) to improve the tagging decision making process. This has led to the need to consider how conventional life safety criteria should be modified to deal with the non-constant hazard faced in the aftershock problem. A third issue is the appropriate consideration of two-component ground motions in the guidelines.

Other similar work being conducted within and outside PEER and how this project differs
We are not aware of similar work elsewhere.
Plans for Year 7 if this project is expected to be continued
Describe any instances where you are aware that your results have been used in industry
The year-one draft guidelines are being used by two engineering firms, but under contract to test the guidelines for future application.
Expected milestones
  • Draft Guidelines (2/02 Done).
  • Final, Revised Guidelines 6/03

A set of guidelines for beta-testing, followed in year two by updated guidelines reflecting the testing and other second year work. Reports and papers.