Ground Motion Selection and Modification: An Overview of Recent Progress for Building Structures and the Implications for Lifeline Structures

Accurately evaluating the seismic response and performance of lifeline structures requires a solid understanding of the characteristics of the ground motions that the lifeline structure will be subjected to in the future. Recent research by the Pacific Earthquake Engineering Research (PEER) Center’s Ground Motion Selection and Modification (GMSM) Program has investigated the seismic response of building structures, and shown that commonly used ground motion selection and scaling methods can lead to an overestimation of the seismic response (e.g. building displacements) by up to 40%. This same research has compared a large number of methods for selecting ground motions, and identified the specific methods that should be utilized to obtain accurate predictions of seismic response. This paper presents an overview of the GMSM Program, and the recent findings related to building structures. The paper then discusses the implication that these findings have on proper selection and scaling of ground motions for lifeline structures, and discusses how the GMSM Program research could be extended to provide specific guidance on this topic. INTRODUCTION AND MOTIVATION Nonlinear dynamic analysis of structures is becoming increasingly prevalent in code and regulatory documents prescribing design and analysis. A recurring challenge for both practicing engineers and developers of these documents is the selection and modification of ground motions for these nonlinear dynamic analyses. Nonlinear structural response is often highly sensitive to the selection and modification of input ground motions and many ground motion selection and modification methods have been proposed. There are virtually no systematic studies that provide impartial guidance to engineers regarding appropriate methods for use in a specific analysis application, leaving the engineers to make an important decision that is virtually uninformed. This is problematic because predicted structural response can vary dramatically depending on the chosen ground motion selection and modification method. The engineer’s choice of ground motion selection and modification method can lead to costly over-design of structures, or worse, dangerous under-design of structures. THE PEER GMSM PROGRAM AND ITS’ OBJECTIVES To address this issue, the Pacific Earthquake Engineering Research (PEER) Center established the Ground Motion Selection and Modification (GMSM) Program. The overall mission of the GMSM Program is to provide practical guidance and tools to the engineering community regarding ground motion selection and modification methods, while at the same time advancing the state of research in this area. In addition to the focus on technical advancements, the GMSM Program is also focused on creating community consensus regarding appropriate GMSM methods. The first study initiated by the PEER GMSM Program resulted in guidelines for the engineering community regarding ground motion selection and scaling when predicting the seismic response of building structures (PEER GMSM 2009). These recently-published guidelines are based on comprehensive research, and include quantitative comparisons of a wide variety of proposed GMSM methods, and identification of the specific methods and method-types that produce accurate and unbiased response predictions. The PEER GMSM Program would like to continue such research to consider the full distribution of response (rather than just median response), investigate other structural responses (e.g. peak floor acceleration), investigate other types of ground motions (e.g. spectrum compatible motions), and investigate other types of systems (e.g. bridges, other lifeline structures, earthen structures, nuclear facilities, and site response). OVERVIEW OF PEER GMSM PROGRAM FINDINGS FOR BUILDING STRUCTURES Generally Applicable Findings. The basic, generally applicable, conclusions developed from the recent PEER GMSM study seems surprisingly simple on the surface. To obtain an accurate and precise prediction of structural response, one must: a) Select ground motions based on record properties important to structural response. b) Use the proper target for these record properties. To illustrate this general finding for the case of predicting the displacement response of building structures, the record property found to be most important was the spectral shape of the ground motion records (in addition to the magnitude of spectral acceleration at a single period, e.g. Sa(T1) where T1 is the fundamental period of the structure). The proper target for this spectral shape was recently developed by Baker and Jayaram (2008), and depends on the ground motion intensity, the site hazard characteristics, and the fundamental period of the building. Therefore, when the goal is to predict the seismic displacement response (e.g. interstory drift response) of a building, the ground motions should be selected and scaled such that they have the proper target spectral shape (in addition to the proper Sa(T1)). Specific Example Results for a 20-Story Reinforced Concrete Frame Building. Figure 1 shows the sample results directly from the recent PEER GMSM study report (2009). This figure shows the predictions of maximum interstory drift response (maximum for any story in the building) for a 20-story reinforced concrete special moment frame (RC SMF) building. These results are for two methods that select and scale ground motions based only on the spectral acceleration at the fundamental period of the building; these methods do not consider the spectral shape of the ground motions. This figure shows the interstory drift responses from individual records (blue dots) and the median interstory drift responses (red crosses) for each set of records (method 100 is based on sets of 7 records, and method 101 is based on sets of 11 records). The blue dots plotted at a drift of 0.08 indicate records that caused sidesway collapse of the structure, and the ratio of collapsed records is given at the top of the figure. The solid black line at 0.019 drift is the High End Prediction (HEP), which is based on structural analysis with many ground motion records and use of advanced statistical analysis methods. The HEP is considered to be the “correct answer” (the best possible answer, given our current level of knowledge in earthquake engineering, which is still incomplete) and is used as a benchmark to measure the accuracy of the other predictions.