Steel weight reduction in offshore wind jacket structures by wrapped FRP joints

Abstract

This report investigates the potential reduction of steel weight for offshore wind turbine supporting jacket structures, if conventional welded joints are replaced by innovative wrapped FRP joints. This new type of connection is under development by Dr. Marko Pavlovic at the Delft University of Technology, and shows outstanding fatigue performance compared to welded counterparts. As jacket structures suffer highly cyclic load, member thickness of current jackets is governed by the fatigue performance of welds. Due to the superior fatigue performance of wrapped FRP joints, substantial weight benefit is expected to be made. The study examines a jacket supported 5 MW wind turbine located in 50-meter water depth in the North Sea. The structure and model are based on the UpWind project. The model includes soil-structure interaction by non-linear depth-dependent springs along with foundation piles. Fatigue limit state (FLS) and ultimate limit state (ULS) are simulated by respectively five and three scenarios. The scenarios consider different combinations of wind (speed and direction), waves (height, period and direction) and current (speed and direction). Six 10-minute simulations are performed for each scenario with different wind turbulence and wave irregularity seeds. Wind and waves are applied in a single simulation, and normal force N and bending moments Mip and Mop time series are recorded at a selection of elements. The time series are post-processed in a self-written MATLAB procedure. For FLS, detailed fatigue analyses of welded joints are performed by evaluating crown and saddle hotspot stress, according to DNVGL-RP-C203. For every time step, the hot spot stress is calculated by applying geometry and load-dependent stress concentration factors (SCFs). Rainflow counting is applied, and the resulting stress range is projected on the details’ S-N curve to evaluate the damage. Linear Palmgren-Miner is applied to accumulate damage. A similar procedure, including stress concentration at thickness transition, is applied to calculate fatigue of elements. For ULS, welded joints are checked for chord face and punching shear failure. Members are checked for tension yielding, local buckling and global buckling. ULS calculations are performed for all time steps and according to Eurocode manuals. The unity check of both FLS and ULS is calculated for each individual member. Next, the member thickness is manually optimised to obtain the most optimal use of material. This optimisation is performed for three different cases with both mild S355 steel and high strength S690 steel. The welded steel structure, case 1, acts as a reference. The unwelded structure, case 2, is the lightest structure if joint fatigue does govern design. Case 3 gives the wrapped FRP structure and includes fatigue results obtained from small scale lab tests. Additionally, due to limited production length of steel tubular elements, it includes circumferential welds in the legs. The potential jacket weight reduction if wrapped FRP joints are applied is large, and the governing unity check shifts from fatigue to global buckling. For mild steel, the reduction of steel weight is more than 50%. The additional reduction of mass for high strength steel is low and not economical. The eigenfrequency of the wrapped FRP structure is viable, as it is outside operating frequencies. The results for the wrapped FRP structure are based on two major assumptions. Firstly, satisfactory joint performance can be obtained, and secondly, this can be accomplished by increasing wrapping thickness only. These assumptions should be verified by future experiments to support the weight reduction statement. In conclusion, the potential benefit of wrapped FRP joints to offshore wind turbine supporting jacket structures is large, and future experiments will show if, or to what extent, the full potential can be exploited.