Seitz, Karlotta-FranziskaKarlotta-FranziskaSeitzHeins, EvelynEvelynHeinsCarstensen, AstridAstridCarstensenGrabe, JürgenJürgenGrabe2020-05-062020-05-062016Proceedings of the 5th International Conference on Geotechnical and Geophysical Site Characterisation, ISC 2016 (2): 1271-1274 (2016)http://hdl.handle.net/11420/6056Cone penetration tests are often used for geotechnical site investigation before pile installation. Usually the measured data for cone and shaft resistance are roughly interpreted and applied for the design of the pile resistance. This paper presents an approach for soil modelling included in numerical simulations of pile loading behavior which employs the CPT data directly. Considering the material properties to be known, e.g. from laboratory tests which have been performed on bore hole excavations, state variables such as void ratio distribution and in situ stress state need to be determined to model an accurate initial state. This paper proposes an inverse determination of these unknown variables with mathematical optimization using an evolutionary algorithm to minimize the difference between measured CPT data and simulation results obtained from FEM. The obtained initial state variable distribution for void ratio for a given stress state may then be used to simulate the pile loading behavior without the need of prior interpretation of CPT data. The solution of the proposed inverse problem will be compared to the determination of state variables for a fictitious problem, in which the initial state is known.enGeowissenschaftenTechnikConference PaperOther