000019655 001__ 19655
000019655 005__ 20170118182330.0
000019655 04107 $$aeng
000019655 046__ $$k2017-01-09
000019655 100__ $$aGiaralis, Agathoklis
000019655 24500 $$aNon-Stationary Frequency Content Characterization of Seismic Accelerograms Via the Slope “Alpha” of the Wavelet-Based Mean Instantaneous Period

000019655 24630 $$n16.$$pProceedings of the 16th World Conference on Earthquake Engineering
000019655 260__ $$b
000019655 506__ $$arestricted
000019655 520__ $$2eng$$aTypical recorded acceleration traces of seismic ground motions (GMs) exhibit a time-varying frequency composition, however, all GM properties currently considered in earthquake engineering to quantify the structural damage potential of GMs do not explicitly characterize the time-evolving trends of GM frequency content. Indeed, the intensity measures (IMs) and record selection criteria adopted in the context of performance-based earthquake engineering (PBEE) do not account for the non-stationary frequency content of GMs in a straightforward manner. In this regard, this paper considers a novel scalar quantity termed alpha, “a”, defined by the average slope (angle) of the wavelet-based mean instantaneous period (MIP) to characterize the temporal evolution of the mean frequency content of recorded GMs. Specifically, the MIP is the time-varying first-order average along the frequency or, equivalently, along the period axis of the wavelet-based GM spectrogram (squared magnitude of the GM wavelet transform) treated as a GM energy distribution on the time-frequency/period plane. Consequently, a captures the rate by which the mean frequency content of GMs changes in time. Linear regression analyses are undertaken involving a dataset of 684 GMs from 30 seismic events of Magnitude 6.5<M<8 and distance to rupture plane 20km<Rrup<120km to quantify statistical/empirical correlation trends between a and well-established GM properties, namely the peak ground acceleration (PGA), peak ground velocity (PGV), and the mean Fourier-based frequency (Tm), oftentimes used as intensity measures (IMs) and record selection criteria in PBEE. Further, regression analyses are undertaken to probe into the relationships between a and important seismological and local site characteristics, namely M, R rup, and the shear wave velocity Vs30. It is found that no significant correlation exist between a and M, Rrup, or PGA. However, it was established that a is well-correlated with the average frequency content of GMs as captured by Tm and by PGV: the lower the average frequency content, the larger the value of a tends to be, that is, the faster the time evolution (transition) of the average frequency content is from higher to lower frequencies. Further, the reported numerical data indicate that the level of the above correlation depends on the intensity of GMs in terms of PGA (conditional on PGA): a is larger for fixed Tm as PGA increases and a increases faster as Tm increases for larger PGA values. Moreover, GMs recorded on softer soils are more likely to have larger a conditional on PGA, a phenomenon that is attributed to the fact that soft soils under strong seismic shaking exhibit strong non-linear behavior that enriches the low frequency content of free field recorded GMs. Lastly, the influence of a in predicting the peak inelastic structural response is assessed within the PBEE framework through a standard sufficiency statistical test on PGA and PGV, treated as non-structure specific IMs in conducting incremental dynamic analysis (IDA) for a hysteretic oscillator with strength and stiffness degradation representing a benchmark 12-storey reinforce concrete frame exposed to the above set of 684 GMs. The considered statistical tests demonstrate that the non-stationary average frequency content of GMs as captured by a influences the peak inelastic structural response at collapse. Overall, the herein furnished results establish the validity and usefulness of a in characterizing the evolutionary frequency content of GMs and suggests that a should be considered as a record selection criterion in undertaking IDA using PGA and PGV as IMs.

000019655 540__ $$aText je chráněný podle autorského zákona č. 121/2000 Sb.
000019655 653__ $$awavelet analysis, non-stationary frequency content, mean instantaneous period, sufficiency of intensity measure

000019655 7112_ $$a16th World Conference on Earthquake Engineering$$cSantiago (CL)$$d2017-01-09 / 2017-01-13$$gWCEE16
000019655 720__ $$aGiaralis, Agathoklis$$iMargnelli, Alessandro
000019655 8560_ $$ffischerc@itam.cas.cz
000019655 8564_ $$s1530205$$uhttps://invenio.itam.cas.cz/record/19655/files/4106.pdf$$yOriginal version of the author's contribution as presented on USB, paper 4106.
000019655 962__ $$r16048
000019655 980__ $$aPAPER