سال انتشار: ۱۳۹۰
محل انتشار: ششمین کنفرانس بین المللی زلزله شناسی و مهندسی زلزله
تعداد صفحات: ۸
Hadi Bahadori – Department of Civil Engineering, Urmia University, Urmia, Iran
Hooman Motamedi – Department of Civil Engineering, Urmia University, Urmia, Iran
When an earthquake occurs, shallow saturated sandy soils will be vulnerable to liquefaction. Considering the existence of this kind of soil in Iran, it seems necessary to present a solution for reducing the settlement due to liquefaction. This paper presents the results of a series of 1g shaking table tests carried out on a model of a four story building with a scale factor of 1/25 resting on geosynthetic reinforced saturated sand beds to find out the effect of this element on the reduction of settlement. A transparent tank of size 200*50*70cm and a rectangular footing of size 24.5*45cm are used for conducting model tests. 161 Firoozkooh sand and two types of geosynthetics, namely geogrid and geogrid-geomembrane geocomposite each with different tensile strength and permeability properties are used in the tests. Meshiran fence is used for modeling geogrid and for modeling geogrid-geomembrane geocomposite, a plastic sheet is used because it possesses tensile strength and it is not permeable so it can model the use of geogrid and geomembrane simultaneously. A complete instrumentation is done using pore water pressure, acceleration transdusers and LVDT to measure and record data continually. It is observed that settlement is reduced in models reinforced with geocomposites. It is also observed that the excess pore water pressure ratio is reduced as much as 70 percent in the models reinforced with geogrid-geomembrane geocoposite layers. Also the appearance of models reinforced with geocomposite layers show improvement afer shaking compared to virgin model and model reinforced with geogrid layer. Profile of geosynthetic layers after shaking show little displacement from their initial shape except for the areas beneath the footing which show settlement comparing to their initial profile.