Statistical Analysis of Geotechnical Parameters in Starnmeer and Alkmaardermeer

Statistical Analysis of Geotechnical Parameters in Starnmeer and Alkmaardermeer: Dike Stability analysis

Bitar, Zeid (TU Delft Civil Engineering and Geosciences)

Hicks, Michael (mentor)
de Gast, Tom (mentor)

Delft University of Technology

Applied Earth Sciences

2018-08-23

The aim of this thesis report is to analyze slope stability parameters given by HHNK for the dike around a polder located in the province of North-Holland, “Starnmeer”, and a lake located to the west of it called “Alkmaardermeer”. Most of the dike has been reported to be too low in some places and plans of restoration will be implemented in 2019. A RFEM (Random Finite Elemental Method) analysis is to be done in collaboration with TU Delft and HHNK on these dikes to identify slope failure probabilities, where the input data is to be addressed in this report. Starnmeer contains thick layers of peat and clay until a depth of approximately -5 meters NAP, where a large sand aquifer begins to much larger depths. For geotechnical analysis, the first few meters are of primary concern, so only the thick peat and clay layers will be assessed. Rising water levels proves problematic for the polder, but fortunately drainage opportunities to adjacent water bodies is possible. The addition or reduction of water heavily influences soil behavior by affecting underlying pore water pressures and stress states. Parameters to be analyzed include wet bulk density, dry bulk density, water content, cohesion, and the friction angle. A statistical analysis and distribution of each parameter provide meaningful insight on probabilities and disseminations of measurement data. The clay layer was found to have a significant amount of sand and silt, affecting the porosity, sorting, cohesion, and bulk densities resulting in an uncompacted porous clay layer. The peat layer is also highly saturated, highly organic, and rather loose. The layers are hence prone to heaving and rapid primary consolidation. By classifying each of the five parameters for both regions to be normally distributed or lognormally distributed, the data can be then standardized and correlated to each other to determine linear dependencies. The relations heavily vary depending on the environmental setting and other physical aspects. For instance, the wet and dry bulk densities are found to have a strong positive correlation with each other in Starnmeer but not in Alkmaardermeer. The layers were found to be heavily saturated, which negatively influences the correlation between water content and bulk densities. Cohesion and friction angle are primarily dependent on the consistency, shape, and packing of the soil grains, so low to no correlation was established.

http://resolver.tudelft.nl/uuid:9fbd867a-fc1f-45ce-ac78-a5e7ad1921d4

Student theses

bachelor thesis

© 2018 Zeid Bitar