The bridge unconnected piles-caisson foundation,consisting of gravity caisson,mattress,and pile group,is a new type of aseismic foundation.The aseismic performance of the foundation is due to the mattress between the caisson and the pile group can dissipate seismic energy.The foundation has been applied to several bridges in strong earthquake hazard area.However,there is no application of this type of foundation in China.Dynamic behaviors of the unconnected piles-caisson foundation,including failure mode,bearing capacity,influencing factors of aseismic performance,are analysed through model tests,numerical simulations and theoretical calculation in this thesis.For the easily application of this type of foundation,the aseismic design method of the unconnected piles-caisson foundation is preliminary proposed based on the analysis.The main content and research results are as follows:(1)Six slow cyclic tests were conducted on model foundations with the variables of mattress material,mattress thickness,vertical safety factor,and ground improvement.The failure mode,bearing capacity,deformability and aseismic performance are analysed in detail through experimental results,including the experimental phenomenon,the hysteresis loops,the backbone curves,the stiffness degradation,the damping ratio,the accumulation of residual displacement,the accumulation of settlement,and the pile bending moment.The test results show that a large amount of energy was dissipated through the mattress during dynamic loading.Pile group can improve the bearing capacity and the shear stiffness of the foundation effectively.With increasing loading displacement,the shear stiffness of these models decreased to a stable value.An empirical equation for the stiffness and loading displacement is developed.Below the loading displacement of 5mm,there is no evident difference in the dynamic behavior of the foundation.After that,the aseismic behavior of the foundation depends on the mattress material,the mattress thickness,and the vertical pressure and is affected by the earthquake intensity and frequency.(2)Shake table tests were conducted on five model foundations with the variables of mattress material,mattress thickness,and ground improvement.The experimental results,including the experimental phenomenon,the model natural frequency,the acceleration amplification effect of foundation soil,the mattress aseismic performance,the caisson displacement,and the pile bending moment,are analysed in detail through.The test results show that pile group can improve the aseismic capacity of the foundation effectively.The dynamic behavior of the different models is similar when the input peak acceleration below0.4g.And the peak acceleration of the caisson of the mattress is slightly larger than the input peak acceleration.Also there is no significant difference in the displacement recovery characteristics of the caisson.When the input peak acceleration above 0.4g,the peak acceleration of the caisson are different: the aseismic ratio of the model without piles decreases;the aseismic ratio of the model with 10 cm mattress is similar as that under small earthquake;the aseismic ratio of the model with 5cm mattress increases,and the model with sand mattress shows the best aseismic performance.However,the model with sand mattress shows the worst displacement recovery characteristics.(3)The discrete-continuous coupling method is applied to simulate the slow cyclic tests and shake table tests of the unconnected piles-caisson foundation.The simulation results are compared with the test results to verify the correctness of the coupled numerical simulation method from macro perspective.The aeismic mechanism of unconnected piles-caisson foundation are analysed from macro and micro perspective.The results show that: when solving static or low-frequency dynamic problems,the discrete-continuous interface coupling method should be used.When solving dynamic problems,the discrete-continuous bridge domain coupling method should be used.The density of the mattress under the caisson increases first and then decreases under dynamic load.The loading displacement corresponding to the mattress changes from dense to loose,is close to the inflection point of the backbone curve,equivalent viscous damping,and residual displacement.Under dynamic load,part of the mattress was gradually pushed out of the range of the caisson.Fragmentation of the mattress grains was concentrated mainly on the contact surface between the mattress and the caisson and that between the mattress and the piles.The rotation of the mattress particles under the caisson cannot exceed 180°,and the rotation of the outer particles is almost0.(4)Discrete element method is used to simulate the mattress and the caisson under seismic load in actual engineering size.In order to design better mattress,the influencing factors of mattress aseismic performance are studied from the the aseisimic ratio,the displacement recovery ratio,the settlement,and the porosity.The results show that: when the mattress density decreases,the aseisimic ratio increases,but the displacement recovery ratio decreases and the settlement of the caisson increases.When the mattress thickness increases,the aseisimic ratio and displacement recovery ratio increase,but the caisson settlement increases.In this thesis,the mattress can reduce the acceleration when the mattress thickness is greater than 1.5m.The mattress shows good aseismic performance when the mattress thickness is greater than 2.0m.And the mattress would amplify the acceleration when the mattress thickness is greater than 6.0m.With the increase of the particle size,the mattress aseisimic ratio decreases,the displacement recovery ratio changes little,and the caisson settlement decreases.The mattress with better gradation shows a lower aseisimic ratio than the mattress with uniform particle size,but the displacement recovery ratio and settlement are better than the uniform particle.The round particle mattress with unlimited particle rotation has higher aseisimic ratio and higher displacement recovery ratio,but the settlement is large.The particles of the mattress are best not easily broken.The caisson displacement recovery ability is poor and the caisson settlement is large when fragmentation of the mattress particles occur.(5)The macro-element method is proposed with the consideration the possible slip and tilt of the caisson,regarding piles and soil as a whole.The macro-element method and beam on nonlinear winkler foundation method(BNWF)are applied in Open Sees3.2.2 to calculate the slow cyclic tests and shake table tests.The results show that the macro-element method can consider the coupling effect between different degrees of freedom.The calculation of energy consumption is more accurate with macro-element method.The BNWF method has a clear concept and easily available parameters.And the calculation of bearing capacity is more accurate with BNWF method.However,the coupling effect between different degrees of freedom cannot be considered in BNWF method.Finally,based on previous studies and the results of this thesis,a direct displacement-based seismic design method for the unconnected piles-caisson foundation is proposed. |