Eurocode 8 Compliant Real Record Sets for Seismic Analysis of Structures

Abstract

Among all the possible options to define the seismic input for structural analysis, natural recordings are emerging as the most attractive. Easily accessible waveform databases are available and evidence shows that only a relatively limited number of criteria has to be considered in selection and scaling to get an unbiased estimation of seismic demand. Like many codes worldwide, Eurocode 8 (EC8) allows the use of real ground-motion records for the seismic assessment of structures. The main condition to be satisfied by the chosen set is that the average elastic spectrum does not underestimate the code spectrum, with a 10% tolerance, in a broad range of periods depending on the structure's dynamic properties. The EC8 prescriptions seem to favour the use of spectrum-matching records, obtained either by simulation or manipulation of real records. The study presented herein investigates the European Strong-Motion Database with the purpose of assessing whether it is possible to find real accelerogram sets complying with the EC8 spectra, while accounting for additional constraints believed to matter in the seismic assessment of buildings, as suggested by the current best practice. Original (un-scaled) accelerogram sets matching EC8 criteria were found, for the case of one-component (P-type) and spatial sets (S-type), for the spectra anchored to the Italian peak acceleration values. The average spectra for these sets tend to be as close as possible to the code spectrum. Other sets, requiring scaling, have been found to match the non dimensional (country-independent) EC8 spectral shape. These sets have also the benefit of reducing, in respect to the un-scaled sets, the record-to-record variability of spectra. Combinations referring to soft soil, stiff soil, and rock are presented here and are available on the internet at http://www.reluis.it/

Keywords:

• Record Selection
• Dynamic Analysis
• Eurocode 8

Acknowledgments

The study presented in this article was developed within the activities of Rete dei Laboratori Universitari di Ingegneria Sismica – ReLUIS for the research program founded by the Dipartimento della Protezione Civile. The authors thank Dr. Fatemeh Jalayer for her precious comments and Mrs. Racquel K. Hagen for her help in proofreading the article. Finally, the authors want to express their appreciation to Prof. Julian Bommer and two anonymous reviewers; their comments certainly improved the quality and readability of the article.

Notes

¹The a_g values for the Zones 3, 2, and 1 are 0.15 g, 0.25 g, and 0.35 g, respectively. These values are related to the probabilistic seismic hazard analysis (PSHA) [McGuire, 1995] for the site of interest. In fact, if the PGA (on rock) with a 10% exceeding probability in 50 years falls in one of the intervals ]0.25g, 0.35g], ]0.15g, 0.25g], or ]0.05g, 0.15g], then the site is classified as Zone 1, 2, or 3, respectively [OPCM 3519, 2006].

²Many authors have recently questioned the use of UHS as target spectrum, see for example Baker and Cornell [2006a].

³Sufficiency of an intensity measure is discussed in detail in the next section.

⁴Epsilon (ϵ) is defined as the difference between the log of the spectral acceleration, at a given period, of a record and that predicted by an ordinary ground-motion prediction equation divided by the standard deviation of the residuals.

⁵In the rest of the article, all calls and verbatim citations of Eurocode 8 will be simply indicated in italic.

⁶Many national codes in Europe have the EC8 as a main reference. The recent Italian seismic code [OPCM 3274, 2003], for example, has very similar prescription for record selection. However, this (b) criterion is not present.

⁷The upper limit accounts for the lengthening of period due to the nonlinear structural behavior, while the lower considers the contribution of higher modes to structural response. The recent Italian seismic code prescribes the lower period range limit as 0.15 s.

⁸In the Italian code the condition on the a_vg does not exist while the rest is the same.

⁹The Italian code has a coincident spectral shape for Type A site class, whereas for other soil conditions it changes.

¹⁰Other than those listed, two more special ground types, S1 and S2, exist. For such cases, special studies for the definition of the seismic action are required, and they are not considered in the investigation.

¹¹EC8 defines two types of spectral shapes to be selected depending on the magnitude of the earthquakes contributing most to the seismic hazard, this also sets an indirect connection between the design spectrum and the hazard. In the following only Type 1 spectra, which apply for surface-wave magnitude larger than 5.5, are given.

¹²This particular database was chosen because the European origin of the considered code. However, as shown in the following, it was only possible to find record sets each of those consisting of ground-motions from different countries and seismic areas. This weakens the motives behind the choice of that specific database; therefore, other waveform depositories could have been considered.

¹³For many of the records in the ESD the X and Y correspond to the east-west and north-south components of motion.

¹⁴Condition (b), checked with the actual PGA values, was always verified for randomly sampled resulting sets, indicating that this approximation seems acceptable.

¹⁵Another more refined option to carry out this job via genetic algorithms is that proposed by Naeim et al. [2004].

¹⁶The lower bounds for Zone 1 of A, B, C site classes are given in Table 9; 30, 20, and 35%, are the minimum lower bounds to find results. These levels have been obtained iteratively increasing (with a 5% step) the lower bound in the analyses. In other words it was not possible to find suitable results using 15, 25, and 30%, respectively. The upper bounds have been chosen arbitrarily to obtain a significant number of resulting record sets.