Effects of local soil conditions on dynamic response of trapezoidal valleys

Geoteknik deprem mühendisliğinde karşılaşılan en önemli problemlerden biri, zemin tabakalarının deprem sırasında gösterdiği davranışın belirlenmesidir. Bir sahada oluşan deprem hareketinin özellikleri tektonik yapı, kırılma mekanizması, doğrultu etkisi, merkez üstü uzaklığı, jeolojik yapının ve yerel zemin koşullarının etkisi gibi birçok faktöre bağlıdır. Zemin tabakalarının dinamik analizi için geliştirilen hesap yöntemleri bir, iki ve üç boyutlu olarak tanımlanmaktadır. İki ve üç boyutlu analizlerde, zemin kesitindeki tabakaların iki veya üç boyutlu geometrisi gerektiği için, bir boyutlu yaklaşımın kullanımı daha fazla tercih edilmektedir. Ancak zemin tabakalarının bir boyutlu dinamik analizinde; yüzey topografyası, tabakaların eğimi ve tabakaların sınırlı enine genişliğinin etkisi ihmal edilmektedir. Zemin tabakalarının yatay yönde sınırlı genişliğe sahip olması, vadi kenarlarında dalga hareketi dönüşümlerine sebep olmakta, dolayısıyla yer hareketinin frekans içeriği ve yüzeydeki etkisi vadilerin ortasından kenarlarına doğru değişebilmektedir. Bu çalışmada, seçilen trapez kesitli simetrik vadi modellerinde yerel zemin koşullarının zemin büyütmesine etkisi incelenmiştir. Bu amaçla, kenarlarda anakaya eğimi sabit olan, derinlik ve genişliği farklı vadi modelleri kullanılarak, farklı anakaya ivme kayıtları için bir ve iki boyutlu dinamik analizler yapılmış, elde edilen sonuçlar karşılaştırılmıştır. Her modelde zemin kesitinde en üstte kil tabakasının yer aldığı kabul edilmiştir. Yapılan analizler sonucunda, vadi yüzeyindeki ivme spektrumları, Afet Bölgelerinde Yapılacak Yapılar Hakkında Yönetmelik’te tanımlanan yerel zemin sınıfları için elde edilmiştir. Yüzeyde hesaplanan en büyük ivme değerleri, anakaya ivmelerine oranlanarak zaman ortamındaki zemin büyütmeleri elde edilmiş ve uzaklığa bağlı değişimleri incelenmiştir. Anahtar Kelimeler: Yerel zemin sınıfı, zemin büyütmesi, vadi modeli, dinamik analiz.One of the most commonly encountered problems in geotechnical earthquake engineering is the evaluation of ground response. The characteristics of ground motion at a particular site depend on many factors such as tectonics of the region, epicentral distance, geological formations, bedrock depth, geotechnical site conditions, local surface and subsurface topography. The topographical characteristics include the effects of surface formations, the two and three dimensional geometry of the subsurface soil layers and bedrock to the local site ground response. The calculation methods developed for the dynamic analysis of the soil layers are usually defined as two and three dimensional according to the necessity of the problem met. The two and three dimensional geometry of the soil layers and bedrock is required to perform two and three dimensional ground response analysis, so because of the application convenience one dimensional dynamic analysis is mostly preferred. However in one dimensional ground response analysis the surface topography, two or three dimensional geometry of the subsurface and the effects of the limited width of the soil layers are being neglected. In fact because of the limited lateral width of the soil layers wave transformations at the basin edges occur, surface waves are being focused to the valley center, two dimensional resonance models may occur and consequently the amplitude and frequency content of the ground motion may change from the center of the valley to the edges. In this paper, in order to study the effects of local site conditions, the depth and width of soil layers, the frequency content and amplitude of strong ground motion to the site amplification and frequency content of surface wave motion, one and two dimensional dynamic analyses were performed for five different bedrock acceleration records by using idealized trapezoidal symmetrical valley models. The results of the one and two dimensional analyses were compared, and the variations of the amplifications with the distance from the valley edges were investigated. These models have different depths and widths. In the valley models, the soil layers were assumed to extend horizontally limited with valley edges having a constant slope angle of 45°. The topsoil layer was selected as high plasticity clay for each model. The parameters such as the thickness and initial shear wave velocities of the soil layers above bedrock, which play the main role in the determination of dynamic response, were selected in accordance with the soil groups and soil classes defined in Turkish Earthquake Design Code (1998). With this aim, the valley models were subjected to 1D and 2D dynamic analyses by using five different bedrock strong ground motion records and the results were compared .The results that will be obtained from the 1D and 2D analyses were aimed to reflect the seismotectonical structure of the North Anatolian fault in Turkey. Therefore two bedrock acceleration time histories were selected among the Turkey earthquakes. The vertical and horizontal boundary conditions become important especially in the dynamic analyses of 2D models. In this study viscous dashpots, which are, calculated proportional to the shear and pressure waves of the relevant layers were put at the vertical and horizontal layers. High amplification values were calculated at the surface of rigid valley models having local site class of Z2. In the models having local site class of Z3 or Z4, the amplifications decreased at the high bedrock acceleration values. At the models with sudden rigidity change in the soil profile, the amplifications relatively increased for the surface sections, which are located at the beginning of valley edge (X/H=1). For the deeper valley models having a graded rigidity decrement from bottom to upper layers, the amplifications decreased noticeably and the increase in the peak bedrock acceleration values made this situation more remarkable. The bedrock topography must be carefully investigated to obtain the 2D dynamic behaviour of laterally limited soil layers under earthquake excitation. So the geological and geotechnical investigations should be done carefully to satisfy this condition. When the data about the topography of soil layers and bedrock is insufficient, it will be obligatory to prefer 1D dynamic analysis based upon the assumption of horizontal soil layers extending to infinity. However 1D and 2D dynamic analyses give similar results only for the sections near the middle part of very wide valleys. Keywords: Local site condition, soil amplification, valley model, dynamic analysis