英文篇名
Site Response Analysis Approach via Generic Profile Correction
作者
張毓文、許尚逸、劉勛仁、張哲瑜、郭俊翔、林哲民、張志偉、簡文郁、吳俊霖、蔡明欣、溫國樑、孫一鴻
關鍵字
地盤反應分析、Vs30、通用地盤剖面、均佈危害度反應譜、地震動評估模型
摘要
傳統大地工程地盤反應分析係將岩盤設計地震輸入於符合相同地盤條件(如Vs30 = 760 m/s)之基盤露頭,以一維動力分析求得地震波經由土層剖面傳遞至地表或結構基礎面之振動,作為考量現地場址放大效應之設計地震,可供後續結構動力分析的輸入運動使用。
上述岩盤設計地震透過地震危害度分析獲得,是為在特定年超越機率下之場址均佈危害度反應譜(uniform hazard response spectrum, UHRS),其過程中對於地震動之評估係基於由地表測站實測地震動資料所建立的地震動評估模型(ground motion prediction equation, GMPE)。近期發展之地震動評估模型,除考慮震源特性對地震動之影響外,亦包括採用以地表下30公尺平均剪力波速(Vs30)為指標的場址項,即可包括測站淺層土壤特性受到地震動影響之地盤放大效應評估。故,UHRS在經由地震動評估模型時已存在一般震源下特定Vs30的地盤放大效應。因此,若仍直接將其置於特定Vs30基盤露頭並考慮場址之土層剖面,進行一維波傳分析,將重複考慮在基盤以下土層之地盤放大效應。
美國DCPP(Diablo Canyon Power Plant)核電廠在獲得Vs30為760m/s的均佈危害度反應譜後,擬以地盤反應分析程序求得地表處之地震動時,鑒於地震動評估模型所評估得之均佈危害度反應譜已存在地盤放大效應之事實,且Vs30僅能代表淺層土壤特性,並無法完全實際反映深層土壤經不同地震動所引致之地盤放大。因此,提出以相同輸入地震進行兩次地盤反應分析求得真正場址地盤放大倍率之程序。以符合場址控制地震之點震源經通用地盤剖面(generic rock profile)求得與建立均佈危害度相符之輸入地震,同樣置於深部岩盤半無限域處分別經由通用地盤剖面(generic rock profile)與場址整體地盤剖面傳遞至地表面,此二組土層剖面於地表輸出之比值,即為該場址之地盤放大函數,再經此放大函數乘上均佈危害度反應譜後,即可得出整體場址土層剖面在此地震情境下考量現地場址地盤放大效應(amplification effect)之地表振動。本文進一步以案例分析,探討傳統地盤反應分析與DCPP電廠之地盤反應方法對分析結果之差異,並說明其影響。
英文摘要
For the traditional site response analysis method, the design earthquake, also known as UHRS (Uniform Hazard Response Spectrum), is assigned at the rock outcrop, such as Vs30 = 760m/s. In this analysis, a one-dimensional wave propagation program, like SHAKE, was used to calculate the response at ground surface and foundation level for further structural dynamic analysis and seismic design.
The UHRS from seismic hazard analysis is based on Ground Motion Prediction Equation (referred to as GMPE). Since the GMPE is based on the strong motion measured at the ground surface, it has implications of the site effects when earthquake waves propagated through the layers. Therefore, when we use the traditional method to calculate the site response, some of the amplification effects would be repeatedly considered, leading to conservative results.
This paper refers to the practice of the DCPP (Diablo Canyon Power Plant) in the United States. First, the point source model is used to generate the input motion at the hard bedrock (Vs ≥ 2800 m/s) of the generic profile. Next, the linear site response analysis of the whole generic profile is conducted by using this input motion and if the resulting surface spectrum at the generic profile is similar to UHRS then the input motion at hard bedrock is consistent with seismic hazard. Then the nonlinear site response analysis of the site-specific profile merged with generic profile is calculated with the same input motion to obtain the site ground motion. By taking the ratio of these two sites responses, the site-specific amplification function would be determined. Hence the site Ground Motion Response Spectrum (GMRS) and the Foundation Input Response Spectrum (FIRS) could be got by multiplied UHRS with amplification function.
The comparison of the ground motions between traditional and modified procedures of site response analysis will be discussed and revealed in this paper.