Research And Application Of The Method To Predict Settlement On CFG Pile Composite Foundation In Mountainous Road

Mountain soft soil not only has low bearing capacity, settlement volume,unfavorable long time consolidation engineering properties, but also has a complex,subtle, unique uneven unfavorable characteristics. When road construction on softground mountainous Settlement after the potential harm to the road structure is easyand safe traffic operations. Not only to the soft ground to be properly handled, butalso on soft ground after treatment were more accurate settlement prediction. Use ofrelevant engineering data, select a reasonable settlement prediction model andprediction methods for the control of post-foundation settlement construction schedule,arrange construction organization, to determine the layered filling height, intermittenttime, and pre-press time to provide some guidance, have certain theoretical andengineering significance.In this paper, the Guizhou Provincial Department of Transportation TechnologyProject "highway soft foundation treatment mountainous Guizhou key technologyresearch and demonstration project" as the basis, Combines the soft ground Hangruicontract section18of the high-speed highway in Guizhou segment completed withincompliance of the mountain road CFG pile composite foundation settlementprediction methods study，On this basis, the forecasting methods used in differentinstances of the project and predicted results were analyzed. This paper studiescompleted several of the following areas:①The CFG pile composite foundation applied to the soft ground mountainousprocessing, characteristics of composite foundation bearing capacity and computing,and compression deformation theory settlement amount calculation methods aredescribed, and demonstration projects conducted CFG pile composite foundationdesign calculations.②Introduces several widely used to forecast the subsidence prediction methodbased on measured data, focusing on the curve fitting, the principle of gray systemmodel and neural network method, the calculation process and range of features. Inthe above methods selected three representative methods: Poisson curve, GM (1,1)and BP neural network, under the guidance of its principles and calculationprocedures, using MATLAB software write settlement data forecasting system,Achieve the integration and application of prediction system and to support projectssubgrade settlement data for example, the use of settlement data forecasting system. ③The use of finite element analysis software Plaxis CFG pile compositefoundation settlement were analyzed to get a better calculation results and thecalculation results with the experimental data and subsidence subsidence predictionsystems are compared.④Indoors centrifugal model test CFG pile composite foundation, observing theentire roadbed deformation and settlement model; At the same time, the results ofnumerical simulation model to predict the results of centrifuge test results and forecastsystem of each method and finite element comparison, further analysis and study ofcomposite foundation settlement.