PhD Theses
Permanent URI for this collection
Browse
Browsing PhD Theses by Author "Ibrahim, Mwajuma"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
Item Geotechnical uncertainties: quantification and consequences in predicting settlement of shallow foundations on sandy soils(University of Dar es Salaam, 2016) Ibrahim, MwajumaGeotechnical designs will always be accompanied by uncertainties in form of soilΒ΄s inherent variability, measurement errors and transformation uncertainties. The effects of uncertainties has been traditionally be dealt with by application of experience based factors of safety to virtually reduce the capacity or magnify the applied load for a given design case. Reliability Based Design (RBD) is a relatively new design approach which rationally deals with uncertainties by quantifying and incorporating them to the design process. Due to its relative newness and perceived complexity, to date, RBD is not used in routine geotechnical designs. However, geotechnical practitioners can still benefit from RBD concepts by understanding the effects of different magnitudes of uncertainties in relation to soil testing methods and design models. This research focuses on that aspect of RBD. Geotechnical investigations were performed at different locations in Dar es Salaam and Coastal region using three common in-situ methods and laboratory testing. The data were used to quantify uncertainties using random field theory. Relationships were developed between data from the different investigation methods and respective transformation uncertainties were determined. A range of uncertainties in the form of coefficient of variation (πΆππ) was applied to determine the consequences in terms of serviceability limit state probabilities of failure (ππ) for shallow footings. The results indicate that at a desired limiting settlement, a unique relationship exist between πΆππ and ππ for each specific settlement prediction model. It is recommended that practitioners should take advantage of RBD concepts by establishing such relationships for different design models and use them as decision making tools.