Boundary Effect On Dewatering-induced Ground Surface Settlement In Deep Excavation

In engineering practice,pumping may significantly provoke ground surface settlement to its surrounding environment.To feasibly estimate dewatering-induced settlement,numerical approach is carried out,following commonly-practiced dewatering schemes.Nonetheless,during ongoing excavation process,as excavation and dewatering are simultaneously conducted,effects of soil-wall and soil-well interactions on dewatering-induced settlement are relatively complex.Therefore,different settings of mechanical boundary conditions(MBCs)are necessitated to simulate such diverse phenomena.In this paper,the finite difference software FLAC3D is employed to model Shanghai soft soils under multi-aquifer-aquitard systems(MAASs)by analyzing results in association with an empirical approach;three core parts,alongside outcomes,are sequentially revealed as following:First and foremost is the validation of numerical and empirical methods under a comprehensive case study of in-situ test in Pudong District,Shanghai.As a result,at the left part of pumping wells,results from numerical method appear much more satisfying to in-situ test than empirical method;the reverse trend is true at the right part.Nevertheless,for empirical approach,despite appearing smaller than others at the left,more satisfying results are clearly seen at the right.Key discrepancies between numerical method and field test are attributed by heterogeneous soil conditions,while those between empirical method and in-situ test are owing to dependence on numerical analysis to simulate drop of pore pressure(?P).More interestingly,despite minor discrepancies,both numerical and empirical approaches are feasible and reliable enough to estimate dewatering-induced settlement.Second is a comparative study of ground surface settlement induced by single and group-well pumping.Numerical analysis is employed to simulate seepage forces and soil deformations caused by drop of pore pressures under axially symmetric and plane strain conditions.For single-well pumping(SWP),axially symmetric boundary is set following radial fluid flow,whereas,for group-well pumping(GWP),plane strain boundary is employed to simulate group well performance.Consequently,under axial symmetry,settlements from empirical method appear overestimated twice to those from numerical analysis because steep trend of hydraulic head near well causes change of effective stress(A?)smaller than change of pore pressure(AP)and significant horizontal strain(?x),contradicting empirical approach based on one-dimensional settlement under ??=?P.Nevertheless,discrepancies appear interestingly marginal while in symmetric plane.Upon increasing distance of group performance,GWP can generate steady-state pore pressures more effectively,causing satisfying ??v=AP and relatively smaller?x.Last but not least,the thesis is finalized by pumping under retaining wall.Three boundary conditions at soil-wall interface:soil-wall friction,wall rotation and wall displacement are considered.As a consequence,during PEP,under influence of shear strength(fs)at soil-wall interface,??v is correspondingly diminished,while sx is significantly restrained through wall stiffness.Concerning wall rotation,though rotational angle is extended up to 44.250,dewatering-induced settlement is unlikely to be affected.With respect to wall displacement,besides feasibility to reduce settlement,fs is significantly diminished upon gradual force reduction.Another is the effect of wall displacement on dewatering-induced settlement,forming a failure plane of soils along slip curve,with weaker shear-strength soils at rD=0.4 and stronger shear-strength soils between rD=0.4 and rD=0.65,where rD is dimensionless distance from the wall.During dewatering,stronger soils tend to drag weaker soils upward to reduce large differential settlements caused by additional compression.Consequently,settlement curve appears larger at rD-0.4,followed by upward shift at rD-0.65 and its trend improves analogously to that of empirical method at rD=2.3 to signify termination of unloading effect on dewatering-induced settlement.