Electron beam welding for the C1 Wedge Connection

Abstract

The demand for offshore wind turbines (OWT) is rising due to the increasing need for renewable energy. This growth is reflected not only in amounts of OWTs but also in their power rating capacity and size. Currently, the L-flange is the most commonly used connection to join the tower to the foundation. However, this connection type has trouble to handle the unprecedentedly high overturning moments that must be transferred between OWT segments. The C1 Wedge Connection (C1-WC) offers a solution by using multiple fasteners to connect a fork-shaped upper flange (UF) to a lower flange (LF), which are both welded to 100+ mm thick tubular steel sections.
The UF is currently manufactured using submerged arc welding to join its three components: the crown (CR), inner web (IW), and outer web (OW). This research investigates the feasibility of electron beam welding (EBW) as an alternative technique. EBW offers increased production speed by utilizing higher power density, allowing the weld to be made in a single pass. However, its full-scale (long-term) performance is unproven. This study explores its feasibility through small-scale experimental tests. The thesis also examines an alternative weld location in the C1-WC UF, comparing the current and alternative weld positions in terms of fatigue and ultimate limit state (ULS) by using finite element analyses.
The main research question is:
To what extent is the welding technique 'Electron Beam Welding' feasible for application in the C1 Wedge Connection™ upper flange manufacturing?
Numerical analyses offer insights into fatigue performance and the use of increased strength in the net cross-section. Although the alternative design, with the weld located in the gross cross-section, demonstrates acceptable fatigue performance under preloaded conditions, it does not outperform the base-case design in service life. Moreover, the increase in net cross-sectional yield strength of the alternative design remains unutilized in the ULS. This suggests that the UF geometry could be designed more efficiently: by reducing the thickness of the inner and outer webs with 19%, an UF mass reduction of 13%, 3.3 tonnes of steel, is achieved.
Experimental tests presented challenges, particularly regarding the precision required for EBW. A slight misalignment in the weld led to incomplete fusion at the weld’s root. Hardness testing revealed a significant increase in hardness in the heat-affected zone (HAZ) and fusion zone (FZ). Charpy impact tests failed to meet the minimum requirement of 31 Joules at -40°C, while tensile tests showed that failure occurred outside the weld. This raised concerns about the weld’s ductility.
Overall, EBW is not yet a feasible option for full-scale production of the C1-WC, mainly due to the needed stringent tolerance requirements and thereby in achieving a consistent welds along the full OWT circumference. Submerged arc welding remains the most reliable method for assembling the UF. Future research should focus on improving EBW’s robustness and reliability, particularly in understanding its fatigue performance and how to account for geometric imperfections.