桩侧摩阻力工作机理与预测的研究

MECHANISM OF PILE SHAFT RESISTANCE AND PREDICTING APPROACH

采用主成分和主分量分析 ,研究了粘性土各常规指标综合反映其性质的能力。提出了采用单一指标液性指数IL,综合反映粘土存在误差 ;双指标孔隙比e和IL,综合反映亚粘土相对更加合理 ;相对指标IL、指标e综合反映亚砂土更加合理 ,且亚砂土隶属于粘性土范畴不尽合理的结论。基于主成分和主分量分析的结论 ,采用BP神经网络模型 ,对指标e和IL 预测大直径钻孔灌注桩桩侧摩阻力的方法进行研究。根据实测数据 ,对大直径钻孔灌注桩端承载力的发挥机制进行了讨论。

In the light of principal component analysis,the capacity of the indexes of cohesive soil,which reflects its behaviour is researched.These are errors in that the liquidity index I \-L,regarded as an unique index,account for the characteristic of clay synthetically.Comparatively,it is more rational that the two indexes of the void ratio e and the liquidity index I \-L are combined to characterize the behaviour of silt\|clay.The index e depicts the nature of sandy loam more rational than I \-L.It is discussed in a systematic way that the methodology of BP artificial neural networks combined with principal component analysis is applied to prediction of the LDBPs′ skin friction,for example the indexes of e and I \-L are used to predict the piles′ skin friction.The mobilizing mechanism of base resistance of large diameter bored piles is discussed based on testing data from instrumented piles.