Influence of uncertain soil models on the construction and design of geotechnical structures

在SPP 1886中的本研究项目中，应调查土壤状态变量、应力状态和孔隙比的变化，及其对静力触探（CPT）结果的影响，以及对岩土结构设计的影响。在汉堡一个岩土工程勘探试验场的CPT数据的帮助下，将进行数值优化的逆参数确定，从而通过CPT的数值模拟推导出垂直应力和孔隙比分布。用这种方法得到的土模型进一步进行了数值模拟，G.将在试验场地进行全位移桩的模拟，这些模拟也包括测量数据。数值和测量结果将用于调查和开发CPT尖端阻力和土壤状态变量之间的相关性。此外，还将评估在桩安装过程中获得的机器参数与土壤特性和估计桩承载力的相关性。总之，数据方差和不确定性的数值结果和设计的影响将进行调查。在拟议的第二个研究期间，最初假设的轴对称调查应该扩大到一个完整的三维数值模型的试验场。在该模型中，将根据CPT探测的可用网格确定土壤状态变量（应力状态和孔隙比）的随机分布。为此，假设状态变量的分形分布。借助三维模型，进一步数值研究了土状态变量的变化对岩土结构数值设计的影响。

In this research project within the SPP 1886 it is supposed to investigate the variation of soil state variables, stress state and void ratio, and its influence on the result of cone penetration tests (CPTs) and beyond on the design of geotechnical structures. With the help of available CPT data from a geotechnically well explored test site in Hamburg an inverse parameter determination with numericaloptimization will be performed, allowing to deduce the vertical stress and void ratio distribution by numerical simulation of the CPTs. With the soil models obtained in this way further numerical simulations, e. g. simulations of full displacement piles at the test site, which are also covered by measured data, will be performed. The numerical and measured results will be used to investigate and develop correlations between CPT tip resistance and soil state variables. Furthermore, the correlation of machine parameters, obtained during pile installation, with soil properties and with estimated pile bearing capacity will be evaluated. All in all, the influence of data variance and uncertainty on the numerical results and on the design will be investigated. In a proposed second research period the initially assumed axisymmetric investigations are supposed to be expanded to a full three-dimensional numerical model of the test site. In this model a random distribution of the soil state variables, stress state and void ratio, will be determined from the available grid of CPT soundings. For this purpose, a fractal distribution of state variables is assumed. With the help of the three-dimensional models further numerical investigations on the influence of variance of soil state variables on the numerical design of geotechnical structures will be performed.