Behaviour of piles in clay subjected to cyclic lateral loading
MetadataShow full item record
In order to make offshore wind turbines more cost-efficient and competitive in international energy markets, design methods must be improved to reduce uncertainties in load and soil modelling. The current method of modelling the lateral soil response of piles is performed using API p-y curves. This method has some clear limitations and tends to underestimate soil strength and stiffness parameters, resulting in the overly conservative design of offshore wind turbines (NGI, 2016). NGI and BP America have developed a novel soil testing device that can obtain monotonic and cyclic p-y curves directly from laboratory tests of soil samples. The device, which is denoted p-y apparatus, was originally developed for well conductor fatigue limit state assessment. For that purpose, displacement controlled tests with symmetric cyclic loading were performed. Comparisons with centrifuge tests and field measurements showed reasonablygood agreement (Zakeri et al., 2017). The objective of the present study is to investigate the possibility of using the p-y apparatus for serviceability limit state design of offshore wind turbines and other pile-shaped foundations subjected to lateral cyclic loading. A series of load-controlled tests with varying average and cyclic load amplitudes have been conducted. Two different types of clay were tested, reconstituted kaolin clay and intact samples taken at an offshore field in the North Sea. The testing program for each specimen was comprised of a series of non-symmetric cyclic load parcels with varying average and cyclic load amplitudes followed by a monotonic push-over test at the end of each test. In addition, one test was subjected to a random load series for comparison purposes. In general, the device performs well and is considered to produce plausible results based on the evaluation of the performed tests. The evaluation of the results includes assessment of drainage and comparisons with contour diagrams, direct simple shear test results and centrifuge test results. The observed pile response is generally as expected and can be described with cyclic contour diagrams. An estimate of the soil drainage indicates that the specimens may behave undrained within a cycle, but partially drained over the duration of the test execution. This is supported by the observed response in the test results. In some tests there has been observed a response similar to behaviour typically seen in the case of gapping during testing at larger load amplitudes. Because of the small displacements induced by the applied cyclic loading, this is interpreted as highly disturbed soil in close vicinity to the model pile resulting in behaviour resembling gapping. In conclusion, the p-y apparatus has shown to produce plausible results when used to performload controlled tests with non-symmetric cyclic loading and is therefore considered applicable to assess the serviceability of piles subjected to cyclic lateral loading. Not all observations made in the tests can be explained and there still remain some open questions. In order to use the p-y apparatus in actual design projects for assessing the serviceability of laterally loaded piles, it is recommended to performfurther tests and as well as complementary finite element analyses using appropriate constitutive soil models. The presented work and the conducted tests may serve as a basis for further investigations.