Book Description
Design of offshore foundations can be difficult due to challenging soils that can vary from high plasticity, soft clay to very dense sand, and complex loading conditions from the respective environments (e.g., wind, waves, seismicity), in the form of combinations of monotonic and cyclic load patterns. Understanding the interaction of the soil-foundation-structure system under design loads is critical for reliable operations of offshore structures. This dissertation provides the evaluation of performance and investigation of mechanisms against cyclic loading for: (1) subsea wellhead-pipeline-manifold systems on soft clay; and (2) multi-pile-supported offshore wind turbine structures in dense sand. Commonly, the product from deep gas wells is collected at a central manifold founded on the seabed via jumpers (i.e., pipelines). The connections to the jumpers are relatively stiff, with limited tolerance against shear failure induced from relative displacement. A centrifuge test was conducted on the 9-m centrifuge at the UC Davis Center for Geotechnical Modeling to study the seismic performance of a caisson-supported manifold structure and a deeply-installed wellhead founded on soft clay when subjected to extreme and abnormal level earthquakes. Dynamic response of jumpers connecting the manifold structure and the wellhead was interpreted as the difference between the dynamic displacement time histories between the manifold structure, the wellhead, and the free-field clay surface. Comparison demonstrated that the governing jumper connections lie between the manifold and the wellhead and between the wellhead and the free-field surface, and the wellhead is the more critical component under the specific ground motion. Offshore wind turbine structures (OWTS) are subject to wind and wave loads with varying magnitudes of static and cyclic loads over their design lives. During normal operation, these structures are further loaded by rotor and blade-passing imbalance forces. Cyclic loading can cause significant degradation in the capacity and generate excessive movement, as well as reduction of the soil-pile stiffness and the natural frequency toward resonance with rotor frequencies. A centrifuge program was designed and performed on the 1-m Schaevitz centrifuge at UC Davis to evaluate the performance of tension piles against cyclic loads for multi-pile-supported offshore wind turbines. The potential for obtaining meaningful results using a small centrifuge for this application was demonstrated, and an initial data set from centrifuge testing of piles subjected to one-way and two-way cyclic axial loading was developed. The data set was presented and evaluated within the interaction diagram framework that is commonly used to predict the cyclic stability of piles. Results from the centrifuge tests were generally consistent with predictions from interaction diagrams (e.g., under one-way loading, increase in cyclic load amplitude lowers pile stability). However, inconsistencies were also observed in the comparison, such as a reduction of capacity for combinations of static and cyclic loads where the interaction diagram suggested “stable” behavior, and an increase in capacity for combinations where the diagram suggested “unstable” behavior. Other observations and implications of the centrifuge results are discussed. Inconsistencies between expected and observed response, such as that mentioned above, demonstrated a lack of full understanding on the complex mechanisms concerning the cyclic stability of tension piles. An axisymmetric finite-element model was developed in OpenSees (McKenna et al., 2010) to help understand the mechanisms affecting the evolution of the axial response (i.e., capacity, stiffness, and pullout rate) under different load combinations of static and cyclic loads on tension piles. The 2004 Dafalias and Manzari bounding surface plasticity model was used for the response of the soil. Five loading stages were performed to simulate this axial problem: confinement, installation by cylindrical cavity expansion and downward shear, static tensile shearing, cyclic shearing, and monotonic pullout. Results from the numerical analysis demonstrated the dependence of the evolution of axial response on the magnitudes of the static and cyclic shear stresses, and the number of applied cycles. Specifically, the analysis suggested the possibility for increase in tensile capacity and stiffening of the soil-pile stiffness for some load combinations, which is typically not considered in design. Other mechanisms and observations, as well as practical implication on current design, are presented.
