Fine Real-time Monitoring Method Of Dewatering Based On Finite Difference

Foundation pit dewatering is an important part of foundation pit engineering and the timeliness of foundation pit dewatering monitoring is an important factor affecting the safety of foundation pit engineering.Current engineering of a precipitation monitoring method compared with the design is more extensive,engineering of underground water level directly from observation Wells,however local water level observation can only represent difficult to response the whole situation of ground water level in an all-round way,especially the key foundation pit internal often cannot arrange observation Wells,this causes the water in the process of precipitation monitoring data is very subjective,Become a potential safety hazard in the project.This is exacerbated by the fact that many projects have difficulty placing enough observation Wells due to site constraints.In order to make up for the shortage of observation Wells,some projects use the method of shutting down some precipitation Wells irregularly as observation Wells,which not only reduces the efficiency of precipitation but also makes the situation of groundwater level more complicated and difficult to measure.At present,in order to deal with the complex formation situation,freezing method is widely used to consolidate the formation by effectively changing the rock and soil properties.However,there are relatively few relevant studies on the influence of freezing temperature field on seepage field,which leads to many situations that cannot be considered and may also bring some groundwater hazards.In order to make up for the deficiency of traditional precipitation methods,this paper adopts the finite-difference method to realize the refinement of monitoring data based on traditional monitoring data,introduces the multi-scale method to improve the accuracy of calculation,and establishes a real-time numerical method based on GPU-CPU heterogeneous computing platform to make up for the deficiency.The following work has been carried out in this project:(1)The finite difference scheme was established,the measured data of observation Wells was taken as the near-field condition,the far field boundary condition was calculated by multiscale method,the general fine monitoring method of precipitation data was studied based on the groundwater seepage control equation,the fine monitoring data of foundation pit was given,and verified with the field measured data;(2)On the basis of geotechnical test data,a multi-scale method was introduced to establish the far field boundary conditions of the seepage control equation,including the influence radius and other key parameters,and verified with the measured data to improve the value accuracy of the influence radius and other parameters;(3)According to the characteristics of the control equation and boundary conditions of foundation pit dewatering problem,the general parallel method of finite difference scheme for foundation pit dewatering problem was studied based on GPU-CPU heterogeneous computing platform to improve the efficiency of numerical calculation and realize real-time processing of monitoring data;(4)According to the equation of underground seepage field and temperature field,the two fields are coupled to determine the influence of freezing temperature field on seepage field.The above research results can provide a feasible theoretical approach for real-time fine monitoring of precipitation engineering,and also provide a method reference for subsequent research and precipitation engineering design.