Scholarlink

<p>To address the need for quantitatively optimizing the input of pulse-like ground motions (PGMs) during the design phase, this study presents a multi-parameter empirical model for the pulse amplification factor (PAF), incorporating two key components: the response spectrum amplification factor (<i>A</i><sub>f</sub>) and the velocity pulse period (<i>T</i><sub>p</sub>). The <i>A</i><sub>f</sub> model is constructed using Gaussian function branches, decoupling peak ground acceleration <i>A</i><sub>fPGA</sub>, the maximum value <i>A</i><sub>fmax</sub> at <i>T</i>/<i>T</i><sub>P</sub> = 0.96, <i>A</i><sub>fmin1</sub> at <i>T</i>/<i>T</i><sub>P</sub> = 0.1, and <i>A</i><sub>fmin2</sub> at <i>T</i>/<i>T</i><sub>P</sub> = 10.0. The <i>A</i><sub>fPGA</sub>, <i>A</i><sub>fmax</sub>, <i>A</i><sub>fmin1</sub>, and <i>A</i><sub>fmin2</sub> are modeled as functions of source, path, and site parameters. Residual and standard deviation analysis reveal that the <i>T</i><sub>p</sub> predictions significantly influence the PAF model’s predictive performance, prompting the development of a new <i>T</i><sub>p</sub> prediction model. Correlation analysis suggests that <i>T</i><sub>p</sub> correlates strongly with magnitude and moderately with site conditions, rupture distance, and exhibits depth dependence only for shallow events (<i>h</i> &lt; 3·0 km). The multi-parameter <i>A</i><sub>f</sub> model reduces the standard deviation of ln(<i>A</i><sub>f</sub>) by an average of 15%, with a maximum reduction of 38% near <i>T</i>/<i>T</i><sub>p</sub>≈1.0. Following adjustments, the model enhances response spectrum prediction accuracy for PGMs, achieving an average total standard deviation reduction of 0.137 (18.7%) across the spectral period range of 0.01–10·0s, with a maximum reduction of 0.233 (26.0%). This model supports the evaluation of pulse effects on critical infrastructures and can serve as a foundation for integrating directional amplification characteristics into seismic hazard analysis.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A multi-parameter empirical prediction model for the spectral amplification factor of pulse-like ground motions

  • 期刊论文

  • 发表日期：

    2026年1月3日

查看全文
  • Lingsheng Zeng

摘要

To address the need for quantitatively optimizing the input of pulse-like ground motions (PGMs) during the design phase, this study presents a multi-parameter empirical model for the pulse amplification factor (PAF), incorporating two key components: the response spectrum amplification factor (Af) and the velocity pulse period (Tp). The Af model is constructed using Gaussian function branches, decoupling peak ground acceleration AfPGA, the maximum value Afmax at T/TP = 0.96, Afmin1 at T/TP = 0.1, and Afmin2 at T/TP = 10.0. The AfPGA, Afmax, Afmin1, and Afmin2 are modeled as functions of source, path, and site parameters. Residual and standard deviation analysis reveal that the Tp predictions significantly influence the PAF model’s predictive performance, prompting the development of a new Tp prediction model. Correlation analysis suggests that Tp correlates strongly with magnitude and moderately with site conditions, rupture distance, and exhibits depth dependence only for shallow events (h < 3·0 km). The multi-parameter Af model reduces the standard deviation of ln(Af) by an average of 15%, with a maximum reduction of 38% near T/Tp≈1.0. Following adjustments, the model enhances response spectrum prediction accuracy for PGMs, achieving an average total standard deviation reduction of 0.137 (18.7%) across the spectral period range of 0.01–10·0s, with a maximum reduction of 0.233 (26.0%). This model supports the evaluation of pulse effects on critical infrastructures and can serve as a foundation for integrating directional amplification characteristics into seismic hazard analysis.