| dc.description.abstract | The successful performance of helical piles under many loading conditions, specifically earthquake motions, make them very popular. Nevertheless, only a few, limited, realistic and quantitative studies on their seismic behavior are available in the literature. The main objective of this research was to evaluate the seismic performance of helical piles by analyzing data from a full-scale experimental shake table test, as well as creating a calibrated model using the commercially available computer program DynaPile. In addition, the seismic performance of grouped helical piles supporting a superstructure was assessed through risk analysis.
A full-scale experimental testing program using a large shake table located at the University of California-San Diego was performed on ten steel piles including nine helical piles with varying geometry and one driven pile embedded in dense sand over five days. The test program was subjected to pulse, white noise excitation and two replicated earthquake motions with different frequency content (1994 Northridge California earthquake and 1995 Kobe earthquake). Different intensities of shakes were generated by scaling the original earthquake amplitudes (50%, 75%, and 100%). On Day 1, test sand was compacted in 30 cm-layers in a 6.7 m long, 3.0 m wide and 4.6 m high laminar box to achieve approximately 100% relative density. Several accelerometers were placed at various elevations and locations within the sand bed, as well as on the exterior side of the laminar box to record sand behavior and its dynamic properties. On Day 2, piles were installed in the soil and free-head piles were tested. On Day3 concrete blocks were placed on top of each individual pile to simulate inertial loads. Strain gauges attached to the exterior pile walls provided data for analyzing the behavior of piles. In order to evaluate group pile behavior, two sets of four same-diameter piles were tied together to form two 2x2 grouped helical piles. Each pile group was connected by a steel skid placed atop the piles. Two accelerometers were placed on opposite sides of the skids near its center of mass and connected to the data acquisition system. On Day 4 each skid was connected to every pile head with two bolts to form a fixed connection. On Day 5, the top bolt of each connection was removed to simulate a pinned pile head connection with the same skids.
The dynamic responses were recorded and measured using both strain gauges on individual piles and accelerometers placed throughout the sand bed as well as on the laminar box and center-of-mass of the skids. Load-displacement and p-y curves were developed to evaluate the response of single and grouped helical piles mainly in terms of resistance and dynamic properties including natural frequency and damping ratio. Damping ratios were attained using four different methods including half-power bandwidth, logarithmic decrement, energy and modal analysis methods. The damping responses under small strain vibration (white noise) and large deformation motion (shakes) as well as the effect of deflection on damping were provided. A special emphasis was placed on the effect of pile slenderness ratio and type of pile-structure connections (i.e. pinned or fixed) on damping. The logarithmic decrement method for small strains and the half-power bandwidth method for both small and large strains were found to yield reasonable damping values for the piles and skids. However, when calculating damping for the soil medium itself, the half-power bandwidth method may not be preferable because the soil exhibits nonlinear behavior. When using the energy method to calculate damping ratio, developing hysteresis load-displacement curves using collected data from one single data recorder (e.g. accelerometer) attached to an appropriate place within the pile-soil-structure system, results in less errors than using data from strain gauges attached to various levels of each pile to develop p-y curves. Moreover, the experimental observations were analyzed to evaluate the soil-pile group-skid stiffness. To that end, the slope of the line that connected two ends of the maximum loop was considered as experimental stiffness. Generally, piles with a pinned head connection showed higher energy dissipation, but lower stiffness. It was also found that piles with greater slenderness ratio (length/diameter) demonstrated lower value of stiffness. In addition to damping and stiffness, natural frequency of the system was found and compared using both Fast Fourier Transform (FFT) and Frequency Response Function (FRF).
A numerical model in a commercially available software program, DynaPile, was calibrated based on the experimental results. The model was then used to conduct a parametric study to gain a broader understanding of seismic behavior of helical pile groups under varying conditions. The experimental and numerical results were compared and the effects of varying different properties in a soil-pile-structure system’s seismic response were discussed. Properties of soil (e.g. shear wave velocity) and structural characteristics including stiffness, damping and slenderness ratio were evaluated in detail. Understanding the parameters that effect the dynamic characteristics of soil-pile systems and quantifying the possible range that can occur under real earthquakes will allow engineers to choose appropriate pile geometry, group configuration and connection type to achieve a desired level of performance.
In addition, estimation of vulnerability of structures and foundations is necessary in seismic zones to have better judgment on the structural performance. Seismic fragility analysis is considered the main method in risk assessment, since it represents a measure for defining the safety margin of the structural system. Since high rise buildings are mostly supported by deep foundations, assessing vulnerability of structures supported by piles is essential even though risk assessment of helical piles has not been addressed in the literature yet. Most structural fragility curves do not take into consideration the contribution of pile foundation systems in the structural vulnerability. Therefore, this study aims to modify existing fragility curves of a six-story fixed-base steel frame hospital building with buckling-restrained braces (BRBs), to incorporate the effect of helical pile group behavior on the fragility of the structure. To that end, a finite element model of the investigated structure was modified with results from a full-scale shake table test performed on two groups of helical piles embedded in dense sand supporting a superstructure. The primary results show that fixed base design may not be conservative for all conditions and soil-foundation interaction should be considered when creating fragility curves, especially for a stiff structure on soft soils where a high-intensity earthquake is anticipated. | en_US |
