The Relationship between Ground Improvement and Wall Deformation in Deep Excavation Using 3-D Analysis

碩士 === 國立臺北科技大學 === 土木與防災研究所 === 102 === PLAXIS 3D, a 3-dimensional numeric analysis program, was used to investigate the influence of piled ground improvement on the retaining structure deformation during deep excavation. The deep excavation of a factory at Neihu, Taipei City was selected as the...

Full description

Bibliographic Details
Main Authors: Hui-Ying Hsieh, 謝蕙霙
Other Authors: 陳水龍
Format: Others
Language:zh-TW
Published: 2014
Online Access:http://ndltd.ncl.edu.tw/handle/rf9gqs
Description
Summary:碩士 === 國立臺北科技大學 === 土木與防災研究所 === 102 === PLAXIS 3D, a 3-dimensional numeric analysis program, was used to investigate the influence of piled ground improvement on the retaining structure deformation during deep excavation. The deep excavation of a factory at Neihu, Taipei City was selected as the study subject. The simulation was carried out using Model (I), which was a composite soil mass model, and Model (II), which was a simulation of piled ground improvement using solid pile elements. For Model (I), only the undrained shear strength or effective cohesion and modulus of elasticity were changed, whereas only the increase of bearing capacity from ground improvement was considered for Model (II), and the rest remained the same. For clay, undrained parameters (A) and (B) were substituted to the hardening soil with small-strain model (HSS) and undrained (A) and undrained total stress (C) to the Mohr-Coulomb model (MC) for simulation. Sand layers were combined with drained MC model or drained HSS model to develop 10 different combinations for simulation. The analysis was conducted with Model (II). The percentage of inclinometer displacement closet to the monitoring results was the HSS (A) of effective stress or MC (C) of total stress based on clay with a displacement percentage within 110%. If HSS (A) model was used for clay, the analysis result of retaining structure deformation or settlement trends were both close to the monitoring data.