Effective soil-stiffness validation

Abstract

In an attempt to decrease the modelling uncertainty associated with the soil-structure interaction of large-diameter monopile foundations, a hydraulic shaker was used to excite a real-sized, in-situ monopile foundation in stiff, sandy soil in a near-shore wind farm. The response in terms of natural frequency and damping of a pile-only system is significantly more influenced by the soil than a full offshore wind turbine structure, and therefore ensures a higher degree of certainty regarding the assessment of the soil reaction. Steady-state vibration amplitudes with frequencies between 1 and 9 Hz were retrieved from strain gauges vertically spaced along the embedded pile, and accelerometers attached to the top of the pile and to the shaker. The measured response is used to validate an effective 1D stiffness method, which is applied as a smart initial guess for a model-based identification of the effective soil-structure interaction properties in terms of stiffness, damping and soil inertia. The performance of the stiffness method is compared to the currently employed p-y stiffness design method. While the effective stiffness method seems to overestimate the actual low-frequency stiffness with about 20%, the p-y method appears to underestimate this stiffness with 140%. The assumption of linear soil behaviour for most of the occurring pile displacements is shown to be acceptable. A damping ratio of 20% (critical) is identified as effective soil damping for the monopile, which is estimated to correspond to a 0.14% damping ratio contribution from the soil for the full structure. The unique measurement setup yielded a ‘first-off’ opportunity to validate a soil-structure interaction model for a rigidly behaving pile. We have shown that indeed such a pile reacts stiffer than predicted by the p-y curve method, and that its response can be modeled more accurately with our recently developed effective stiffness method.