本文選題：場(chǎng)地效應(yīng)　切入點(diǎn)：數(shù)值分析　出處：《西南交通大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：地震動(dòng)由震源效應(yīng)、路徑效應(yīng)和場(chǎng)地效應(yīng)三者共同決定。大量現(xiàn)場(chǎng)觀測(cè)、理論分析、數(shù)值模擬和物理試驗(yàn)的研究成果都證實(shí)場(chǎng)地效應(yīng)對(duì)地震動(dòng)有顯著影響。其中,地形效應(yīng)作為場(chǎng)地效應(yīng)的重要方面,自然起著不可忽視的作用。但是,大量對(duì)邊坡地形效應(yīng)的研究是頻域內(nèi)的,這些研究成果無(wú)法直接應(yīng)用于工程抗震設(shè)計(jì)。本文即是利用數(shù)值模擬和統(tǒng)計(jì)回歸的方法,求出一系列反應(yīng)譜譜放大的預(yù)測(cè)模型評(píng)估邊坡的地震動(dòng)地形效應(yīng)。本文的預(yù)測(cè)模型能夠得到坡頂面任意場(chǎng)地的設(shè)計(jì)反應(yīng)譜,便于實(shí)際工程的抗震設(shè)計(jì)。具體步驟如下：(1)首先建立80個(gè)單面坡數(shù)值模型,分別具有不同的模型參數(shù),相應(yīng)地建立一系列一維自由場(chǎng)模型；(2)通過(guò)輸入一條基巖地震記錄,得到每個(gè)邊坡模型和對(duì)應(yīng)一維自由場(chǎng)模型上數(shù)個(gè)場(chǎng)地的地震反應(yīng)；(3)通過(guò)對(duì)時(shí)域的地震記錄進(jìn)行傅里葉變換,求出各場(chǎng)地地震反應(yīng)和輸入地震動(dòng)間的頻域內(nèi)傳遞函數(shù)；(4)利用上一步得到的傳遞函數(shù),計(jì)算出每個(gè)邊坡模型和對(duì)應(yīng)一維自由場(chǎng)模型上各監(jiān)測(cè)點(diǎn)在輸入大量地震記錄時(shí)的地震反應(yīng)。這一方法很大地提高了計(jì)算效率；(5)通過(guò)對(duì)比邊坡模型和對(duì)應(yīng)一維自由場(chǎng)的反應(yīng)譜,求出反映地震動(dòng)地形效應(yīng)的譜放大；(6)利用系統(tǒng)的統(tǒng)計(jì)分析和隨機(jī)效應(yīng)方法建立起相對(duì)簡(jiǎn)單的地形效應(yīng)預(yù)測(cè)模型,模型中只包含四個(gè)邊坡模型參數(shù)：坡高、坡度、場(chǎng)地自坡頂點(diǎn)的距離和坡體材料的剪切波速。利用殘差分析法選擇模型參數(shù)并驗(yàn)證參數(shù)在統(tǒng)計(jì)意義上是否顯著。殘差分解為模型內(nèi)殘差和模型間兩個(gè)部分,模型內(nèi)殘差由不同基巖記錄的不同頻率成分造成,預(yù)測(cè)模型對(duì)放大效應(yīng)描述的不精確產(chǎn)生了模型間殘差。預(yù)測(cè)模型的兩類殘差在線性尺度上相對(duì)于模型參數(shù)是無(wú)偏的：用來(lái)擬合殘差的直線是水平的。兩類殘差的標(biāo)準(zhǔn)差都很小,地震參數(shù)主要在短周期對(duì)模型有影響。兩個(gè)數(shù)值驗(yàn)證實(shí)例的結(jié)果表明,側(cè)面吸收邊界產(chǎn)生的反射波會(huì)對(duì)譜放大計(jì)算的準(zhǔn)確性造成影響。但是,這種影響的誤差水平在地震動(dòng)預(yù)測(cè)模型中是可以接受的。殘差分析和誤差分析表明：在選定的模型參數(shù)范圍內(nèi),最終的預(yù)測(cè)模型可以在一階描述邊坡的地震動(dòng)地形效應(yīng)。
[Abstract]:Ground motion is determined by focal effect, path effect and site effect. A large number of research results of field observation, theoretical analysis, numerical simulation and physical test all prove that site effect has significant influence on ground motion. Terrain effect, as an important aspect of site effect, naturally plays an important role. However, a large number of researches on slope topographic effect are in the frequency domain. These results can not be directly applied to seismic design of engineering. In this paper, numerical simulation and statistical regression methods are used. A series of prediction models with amplified response spectrum are obtained to evaluate the topographic effect of the slope ground motion. The prediction model in this paper can obtain the design response spectrum of any site on the top of the slope. The concrete steps are as follows: 1) first, 80 single slope numerical models with different model parameters are established, and a series of one-dimensional free field models are established accordingly) by inputting a bedrock seismic record. The seismic response of several sites on each slope model and corresponding one-dimensional free field model is obtained by Fourier transform of seismic records in time domain. The transfer function in frequency domain between the seismic response of each site and the input ground motion is calculated. (4) using the transfer function obtained from the previous step, The seismic response of each monitoring point on each slope model and the corresponding one-dimensional free field model is calculated when a large number of seismic records are inputted. This method greatly improves the calculation efficiency by comparing the slope model with the corresponding one-dimensional free field response spectrum. Based on the statistical analysis of the system and the random effect method, a relatively simple terrain effect prediction model is established. The model contains only four slope model parameters: slope height, slope gradient, slope height, slope gradient, slope height, slope gradient, slope height, slope height, slope gradient, slope height, slope height, slope gradient, slope height, slope height, slope height, slope gradient, The distance of the site from the top of the slope and the shear wave velocity of the material of the slope body. The model parameters are selected by residual analysis and the statistical significance of the parameters is verified. The residual is decomposed into two parts: the residual and the inter-model. The residuals in the model are caused by different frequency components recorded in different bedrock. The two types of residuals of the prediction model are unbiased relative to the parameters of the model on a linear scale: the lines used to fit the residual errors are horizontal, and the standard deviations of the two types of residuals are very small. The seismic parameters mainly affect the model in short periods. The results of two numerical examples show that the reflection generated by the side absorbing boundary will affect the accuracy of spectral amplification calculation. The residual error analysis and error analysis show that the final prediction model can describe the topographic effect of ground motion in the first order within the selected parameters of the model.
【學(xué)位授予單位】：西南交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TU435

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 劉殿魁;王國(guó)慶;;淺埋圓形孔洞附近的半圓形凸起對(duì)SH波的散射[J];力學(xué)學(xué)報(bào);2006年02期

，

本文編號(hào)：1625519