arXiv subjects

Maijia Su

Publications and source records attributed to Maijia Su.

4 recordsLinked to original sources

arXiv2025

Recorded Versus Synthetic Spectral-compatible Ground Motions: A Comparative Analysis of Structural Seismic Responses

This paper presents a comparative analysis of structural seismic responses under two types of ground motion inputs: (i) synthetic motions generated by stochastic spectral-compatible ground motion models and (ii) recorded motions from an earthquake database. Both ground motion datasets are calibrated to a shared target response spectrum to ensure consistent spectral median, variance, and correlation structure. Five key stochastic response metrics-probability distributions, statistical moments, correlations, tail indices, and variance-based global sensitivity indices-are systematically evaluated for two representative structures: a medium-period building and a limiting case of a long-period tower. The comparison accounts for uncertainties both from ground motion and structural parameters. The results reveal that synthetic motions closely replicate recorded motions in terms of global response behavior-including distributions, mean and variance, correlation structure, and dominant uncertainty sources-indicating their suitability for routine seismic design and parametric studies. However, substantial differences emerge in response extremes for long-period structures, particularly in metrics governed by rare events, such as higher-order moments and tail behavior. These differences, which often exceed 50%, can be attributed to the non-Gaussian features and complex characteristics inherent in recorded motions, which are less pronounced in synthetic datasets. The findings support the use of synthetic ground motions for evaluating global seismic response characteristics, while highlighting their limitations in capturing rare-event behavior and long-period structural dynamics.

physics.geo-ph↗

arXiv2024

Review and Validation of Stochastic Ground Motion Models: which one does it better?

Stochastic ground motion models (GMMs) are gaining popularity and momentum among engineers to perform time-history analysis of structures and infrastructures. This paper aims to review and validate hierarchical stochastic GMMs, with a focus on identifying their ''optimal'' configuration. We introduce the word ''hierarchical'' as its formulation contains two steps:(1) selecting a modulated filtered white noise model (MFWNM) to replicate a target record and (2) constructing a joint probability density function (PDF) for the parameters of the selected MFWNM, accounting for the record-to-record variability. In the first step, we review the development of MFWNMs and explore the ''optimal'' modeling of time-varying spectral content. Specifically, we investigate different frequency filters (single- and multi-mode) and various trends (constant, linear, and non-parametric) to describe the filters' time-varying properties. In the second step, the joint PDF is decomposed into a product of marginal distributions and a correlation structure, represented by copula models. We explore two copula models: the Gaus-sian copula and the R-vine copula. The hierarchical GMMs are evaluated by comparing specific statistical metrics, calculated from 1,001 real strong motions, with those derived from their corresponding synthetic dataset. Based on the selected validation metrics, we conclude that (1) Hierarchical stochastic GMMs can generate ground motions with high statistical compatibility to the real datasets, in terms of four key intensity measures and linear- and nonlinear-response spectra; (2) A parsimonious 11-parameter MFWNM, incorporating either the Gaussian copula or the R-vine copula, offers sufficient and similar accuracy.

stat.AP↗

arXiv2024

Surrogate modeling for probability distribution estimation:uniform or adaptive design?

The active learning (AL) technique, one of the state-of-the-art methods for constructing surrogate models, has shown high accuracy and efficiency in forward uncertainty quantification (UQ) analysis. This paper provides a comprehensive study on AL-based global surrogates for computing the full distribution function, i.e., the cumulative distribution function (CDF) and the complementary CDF (CCDF). To this end, we investigate the three essential components for building surrogates, i.e., types of surrogate models, enrichment methods for experimental designs, and stopping criteria. For each component, we choose several representative methods and study their desirable configurations. In addition, we devise a uniform design (i.e., space-filling design) as a baseline for measuring the improvement of using AL. Combining all the representative methods, a total of 1,920 UQ analyses are carried out to solve 16 benchmark examples. The performance of the selected strategies is evaluated based on accuracy and efficiency. In the context of full distribution estimation, this study concludes that (i) AL techniques cannot provide a systematic improvement compared with uniform designs, (ii) the recommended surrogate modeling methods depend on the features of the problems (especially the local nonlinearity), target accuracy, and computational budget.

stat.ME↗

arXiv2024

The Importance of Corner Frequency in Site-Based Stochastic Ground Motion Models

Synthetic ground motions (GMs) play a fundamental role in both deterministic and probabilistic seismic engineering assessments. This paper shows that the family of filtered and modulated white noise stochastic GM models overlooks a key parameter -- the high-pass filter's corner frequency, $f_c$. In the simulated motions, this causes significant distortions in the long-period range of the linear-response spectra and in the linear-response spectral correlations. To address this, we incorporate $f_c$ as an explicitly fitted parameter in a site-based stochastic model. We optimize $f_c$ by individually matching the long-period linear-response spectrum (i.e., $Sa(T)$ for $T \geq 1$s) of synthetic GMs with that of each recorded GM. We show that by fitting $f_c$ the resulting stochastically simulated GMs can precisely capture the spectral amplitudes, variability (i.e., variances of $\log(Sa(T))$), and the correlation structure (i.e., correlation of $\log(Sa(T))$ between distinct periods $T_1$ and $T_2$) of recorded GMs. To quantify the impact of $f_c$, a sensitivity analysis is conducted through linear regression. This regression relates the logarithmic linear-response spectrum ($\log(Sa(T))$) to seven GM parameters, including the optimized $f_c$. The results indicate that the variance of $f_c$ observed in natural GMs, along with its correlation with the other GM parameters, accounts for 26\% of the spectral variability in long periods. Neglecting either the $f_c$ variance or $f_c$ correlation typically results in an important overestimation of the linear-response spectral correlation.

stat.AP↗