Abstract This paper presented a dynamic fluid-solid coupled finite element (FE) method incorporating the soil cap yield hardening model to analyze the improvement on saturated foundation under dynamic compaction (DC). Biot’s dynamic u–U–p formulation was employed to describe the coupling of pore fluid and solid phases, which was discretized into finite elements by application of the Galerkin method and viscous Cartesian connectors. The proposed numerical model showed reasonably good agreement with the existing analytical solutions of one-dimensional transient loading problems and field measurement of dynamic compaction. The effects of groundwater table and soil permeability were examined via the development of excess pore water pressure, effective soil stress and void ratio. Results showed that groundwater table had a significant effect on the foundation improvement by DC. Higher groundwater table resulted in larger excess pore water pressure, and effective soil stress was not found to develop under the groundwater table during the compaction process. The decrement in the void ratio was only limited to soils located above the groundwater table, and a critical depth of groundwater table existed for the DC reinforcement in fine soil foundation, which was suggested to be 6 m for the DC reinforcement with the tamping energy of 2500 kN·m per blow. Four soil permeability coefficients of k = 10−2 m/s, 10−3 m/s, 10−5 m/s and 10−7 m/s were chosen in the parametric analysis, which represented gravel, coarse sand, silt and silty clay, respectively. Higher permeability resulted in lower excess pore water pressure and larger decrease in void ratio. However, the saturated foundation was not suitable to reinforce by dynamic compaction, no matter how much the soil permeability was. Compared to the increase in soil permeability, it was more crucial to lower the groundwater table before implementing dynamic compaction. Moreover, the gravel columns were preferred to be used to accelerate the drainage of groundwater than the sand columns.

中文翻译：

---

基于土盖模型的流固耦合法强夯改善饱和地基评价

摘要 本文提出了一种结合土盖屈服硬化模型的动态流固耦合有限元(FE)方法来分析强夯(DC)下饱和地基的改进。Biot 的动态 u-U-p 公式用于描述孔隙流体和固相的耦合，通过应用 Galerkin 方法和粘性笛卡尔连接器将其离散为有限元。所提出的数值模型与现有的一维瞬态加载问题和强夯现场测量的解析解显示出相当好的一致性。通过超孔隙水压力、有效土壤应力和空隙率的发展来检验地下水位和土壤渗透性的影响。结果表明，地下水位对直流地基改善效果显着。较高的地下水位导致较大的超孔隙水压力，在压实过程中未发现地下水位下产生有效的土壤应力。孔隙比的降低仅限于地下水位以上的土壤，细土地基中直流加筋存在地下水位临界深度，建议直流加筋为6 m，夯实能量为每次冲击 2500 kN·m。在参数分析中选择了四个土壤渗透系数 k = 10-2 m/s、10-3 m/s、10-5 m/s 和 10-7 m/s，分别代表砾石、粗砂、淤泥和粉质粘土，分别。较高的渗透率导致较低的超孔隙水压力和较大的孔隙比下降。然而，饱和地基无论土壤渗透性如何，都不适合强夯加固。与提高土壤渗透性相比，实施强夯前降低地下水位更为关键。此外，与沙柱相比，砾石柱更适合用于加速地下水的排出。