Offshore wind turbines are now more essential than ever for transitioning to a greener, sustainable society. Since the foundations of these wind turbines are surrounded by seawater during operation, they must be protected against corrosion. One method of corrosion protection is the impressed current cathodic protection (ICCP) system, which can lead to hydrogen production due to over-protection when the potential falls into the hydrogen evolution reaction (HER) region. S355ML is typically the chosen material for the most common wind turbine foundation type, the monopile, which is bent and welded during its manufacturing process.

In this MSc thesis, the effect of pre-strain and microstructure on the hydrogen embrittlement (HE) behavior of S355ML is studied. Different levels of pre-strain (0%, 2%, 8%, 16%) are applied to S355ML in the thermomechanical rolled (TM) condition. The goal is to study and compare the binding energies of the traps and the hydrogen concentration at each pre-strain level. Additionally, the binding energies and hydrogen concentration of the heat-affected zone (HAZ, replicated by heat treatment) and the TM microstructures pre-strained at 8% are compared.

To investigate how the mechanical properties of S355ML are affected by the presence of hydrogen in the material, slow strain rate tests (SSRT) and profilometry-based indentation plastometry (PIP) analysis have been carried out on uncharged 0% pre-strained TM samples and hydrogen-charged, 8% pre-strained TM samples. The SSRT tests show that the hydrogen-charged samples experience HE, as the elongation at failure of charged samples decreases with strain rate, with a 37% decrease for the slowest crosshead displacement rate. Since strain rate plays an important role in HE phenomena, it is observed that PIP tests are not suitable for studying the HE behavior of S355ML.

It is found that without the effect of pre-strain, S355ML in the TM condition does not retain hydrogen. Moreover, hydrogen concentration increases with pre-strain following the power law H=0.12ϵ0.54.

Therefore, dislocations seem to be the main traps present in S355ML samples in the two studied microstructures. However, the values of binding energies vary with pre-strain level and microstructure and do not match those usually assumed for dislocations: 19.0 kJ/mol for the traps of 2% pre-strained TM samples, 15.6 kJ/mol for the traps of 8% pre-strained TM samples, 34.2 kJ/mol for the traps of 16% pre-strained TM samples, and 8.2 kJ/mol for the HAZ microstructure with 8% pre-strain. The variation of binding energies with pre-strain and microstructure can be attributed to the presence of overlapping peaks in the thermal desorption spectroscopy (TDS) analysis, making it difficult to apply the Kissinger equation to the correct data points. Consequently, a higher number of samples should be tested to determine the accuracy of the Kissinger theory and the Choo-Lee plot for binding energy determination.