Font Size: a A A

Methods On Settlement Calculation And Prediction Of Soft Soil Roadbed

Posted on:2009-11-28Degree:MasterType:Thesis
Country:ChinaCandidate:X JinFull Text:PDF
GTID:2132360242476515Subject:Geotechnical engineering
Abstract/Summary:PDF Full Text Request
The unfavorable characteristics of soft soil foundation are as follows: low bearing capacity, Large settlement and long-time consolidation. Due to these characteristics of soft soil foundation mentioned above, highway construction on the soft ground is a technical tickler. It is very important to forecast the settlement by the theorical and actual model and methodology from the actual settlement data. On the calculation of settlement, the current norm in China is layerwise summation method,which is choosen as the model of settlement calculation of the shallow foundation on natural foundation soils. The layerwise summation method doesn't consider the lateral expansion of the subbed soil. There isn't strict theoretical support on choosing deformation indicator and making sure the thickness of the compression layer either, which leads to the result deviates the actual result a lot.According to the problem mentioned aboved, main research of this essay is described as follows:(1) Methods on settlement calculation and prediction of the soft soil foundation are summarized.(2) For the one-stage loadings of embankment, the influences of calculation multistage starting time and space of time on the accuracy are studied for hyperbola method, exponent curve method, Logistic curve method and the back-propagation artificial neural networks method. Then the four methods with their optimum starting time and space of time are compared, and Asaoka method is used to check the accuracy of final settlement prediction of all the methods. The least squared method is to used to confirm the parameters of Logistic curve method with Matlab. Compared to the traditional three-step method, it can gain better accuracy. (3) For the multistage loadings of embankment, influence of different loading stage of hyperbola method, exponent curve method, Logistic curve method back-propagation artificial neural networks method are discussed. Asaoka method is also used to check the accuracy of final settlement prediction of the methods mentioned above.(4) The relationship between lateral deformation sh and settlement of embankment center sv is preliminary discussed.(5) According to calculating with Plaxis v8.2 finite element method, settlement calculation method which can consider lateral deformation of soil is put forward on layerwise summation method. The relationship among the modification coefficient K, height of embankment H, the porosity e and buried depth Hm of soil is established.
Keywords/Search Tags:Soft soil roadbed, Settlment prediction, Least squared method, Back-propagation artificial neural networks, Settlement modification coefficient
PDF Full Text Request
Related items