Date of Degree
PhD (Doctor of Philosophy)
Civil and Environmental Engineering
M. Asghar Bhatti
In urban areas, vibrations generated by pile driving often affect the neighboring properties vulnerable to ground shaking. These vibrations may cause damage to surrounding structures either by shaking the ground or by causing settlement of the soil beneath foundations in the proximity of pile driving. It is important to distinguish between the conditions under which the vibrations will cause damage and those under which vibrations are tolerable. The numerical studies on the analysis of pile driving have mostly focused on assessing the driving efficiency and the bearing capacity of dynamically loaded piles. A limited number of studies included the study of ground vibrations due to pile driving and its effects on adjacent structures. However, the factors affecting the ground vibrations in soils such as the nonlinear constitutive behavior of soil, soil-pile interaction and penetration depth of the pile have not been clearly identified.
The objective of this research is to implement a numerical method to simulate dynamic loading of a single pile, and study the factors influencing the stress wave propagation in the soil surrounding the pile. The thesis is comprised of two main analyses: (1) the static analysis of a pile in which the phenomenon of static consolidation is studied, and (2) the dynamic analysis of a pile in which pile driving and ground vibrations are studied.
In the static analysis, the load capacity of a single pile is investigated. The results from the finite element method are compared with widely recognized theoretical methods. The theoretical methods that are used to estimate the end bearing capacities are: (1) General Formula, (2) Vesic's Method, (3) Janbu's Method, (4) Meyerhof's Method, and (5) Coyle & Castello's Method. The estimation of skin friction resistance (shaft capacity) of single piles is performed using the (1) Alpha method, (2) Beta method, and (3) Lambda method. Two numerical applications are performed to predict the load capacity of single piles in normally consolidated clays. It is observed that the model with no slippage at the interface predicts almost twice as much load capacity as the model with interface. In regards with the end bearing capacities, Coyle & Castello's method is found to be most conservative followed by the finite element method, the Janbu's method, the Meyerhof's method, and finally the Vesic's method. In respect to skin friction resistance, the finite element is found to be the most conservative method, followed by the Beta, the Lambda, and the Alpha method.
In the dynamic analysis, the amplitudes of ground vibrations are investigated based on the variation of: (1) the soil type, (2) the pile embedment length and (3) the released hammer energy. In the first analysis, five types of soils - loose and dense sands and, soft, medium stiff, and stiff clays - are modeled. The highest vibration amplitude is calculated for the loose sand with a peak particle velocity (PPV) of 10.0 mm/s followed by the dense sand with a PPV of around 4.0 mm/s. Among the clay types, the vibrations are higher for the stiffer clay in the near field, which is 9 m (half a pile length) or less away from the pile. In the second analysis, three different embedment lengths - full, half, and quarter pile length - are modeled. It is found that the quarter embedded piles produce greater vibration amplitudes as compared to the half and fully embedded piles. Larger amplitudes of vibrations are encountered on the ground surface for shorter pile embedment lengths. In the third analysis, three different impact forces consisting of 2,000 kN (F), 6,000 kN (3F) and 10,000 kN (5F) are applied on the pile head. It is concluded that increase in hammer energy causes increase in the peak particle velocities.
cohesive soil, finite element, ground vibrations, numerical model, pile driving
Copyright 2010 Mehmet Serdar Serdaroglu