Font Size: a A A

Study On Back-analysis And Prediction Method For High Embankment Settlement

Posted on:2005-08-15Degree:MasterType:Thesis
Country:ChinaCandidate:D Q ZouFull Text:PDF
GTID:2132360155954834Subject:Geotechnical engineering
Abstract/Summary:PDF Full Text Request
Along with the process of expressway extending to Midwest area in China, there will be more and more high embankments. At present, the studies on settlement calculation model and settlement prediction method for roadbed under this great force loaded by high embankment are rare. Along with the rapid development of the computer and calculation technique, it is possible to research on back-analysis and settlement prediction by using the information of observation and combining with optimization theory, which has great engineering meaning and theoretical value.Based on the existing achievements on settlement prediction, the dissertation summarizes and researches on the practical calculation methods for high embankment settlement. Firstly, combining with the programme of nonlinear plane FEM based on Biot consolidation, the basic theories of FEM based on Biot consolidation was introduces detailedly. Reference to the sensitivity analysis method of systematology, this paper researches and analyses the sensitivities of parameters of Duncan-Chang nonlinear elastic model, thereby, the parameters needed to back-analysis are selected. In the process of back-analysis, complex-shapes algorithm is selected to optimize the parameters. And based on the principle and calculatation process, a programme is written in Fortran, then integrate it with BCF to compose a back-analysis programme of nonlinear plane FEM based on Biot consolidation. By applying E-v model and E-B model to analyze two examples of engineering, and the result shows that the settlement prediction methods are viable, and reveals that E-B model has less error than E-v model. Then, based on gray prediction model or GM(1,1) and measured settlement data, this paper solves the parameters of GM(1,1) using Genetic Algorithm(GA) and Minimum Two Multiplication, and the result shows that the former has less error and higher precision than the latter. In the end, this paper introduces primarily a new kind of settlement prediction method—Compertz Curve Method, and compares it with FEM and grey prediction method in the same engineering, the result shows that Compertz Curve Method has many merits, such as good adaptability, small dispersing of prediction result, high precision and so on, therefore, we should expand the application of this method in the engineering practice.
Keywords/Search Tags:high embankment, settlement, back-analysis, FEM based on Biot consolidation, gray prediction model, Genetic Algorithm, Compertz Curve Method
PDF Full Text Request
Related items