PSI - Issue 54

Liese Vandewalle et al. / Procedia Structural Integrity 54 (2024) 180–187 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

182

3

Vanadium carbides have the same NaCl crystal structure as titanium carbides (Baker (2013), Takahashi et al. (2012)) and studies by electrochemical charging indicated relatively strong trapping by these carbides as well. Atom probe tomography (APT) analysis of Takahashi et al. (2012) showed enrichment of deuterium atoms at the broad interface of fine vanadium carbide platelets, indicating that H trapping occurred at the vanadium carbide/matrix interface. Depover and Verbeken (2016b) studied the trapping behavior of vanadium carbides in a quenched and tempered (Q&T) martensitic steel and reported strong trapping of H by the small precipitated carbides, where the E a appeared to depend on the precipitate size. The smallest precipitates (< 5 nm) were related to the strongest trapping sites, with E a values of above 60 kJ/mol, while precipitates larger than 20 nm did not contribute to trapping. Moreover, this experimental work was also further evaluated by Drexler et al. (2020) using a trap-diffusion integrated finite element model. Similarly to the TiC, they found four different types of trapping sites related to the vanadium carbides with E B values of 32 kJ/mol, 58 kJ/mol, 73 kJ/mol and 92 kJ/mol, on average. Additionally, the use of the model-based evaluation of the spectra allowed determination of the respective trapping densities. Combining these findings together with information obtained by experimental characterization of the underlying microstructures they identified the different trapping sites as the coherent interface, C-vacancies at the semi coherent interface and incoherent interface as well as the bulk C-vacancies in carbides smaller than 5 nm thickness. Turk et al. (2018) studied the trapping of vanadium carbides in a ferritic matrix and observed stronger trapping for smaller carbides. A systematic study performed by Takahashi et al. (2018) combined TDS and APT together with high resolution transmission electron microscopy (HRTEM) to study the H trapping at fine vanadium carbides in under-, peak- and over-aged steels. They showed that while little H was introduced in the under aged samples, strong trapping was present in the peak- and over-aged samples, with a E a value of around 60 kJ/mol. Additionally, APT showed no enrichment around the perfectly coherent VC in the under-aged samples while strong segregation at the broad (001) interface of the VC platelets in the peak-aged samples was observed. Based on HRTEM investigation together with comparison of dislocation and trapping density, the misfit dislocations were ruled out as trapping site. Instead chemical analysis showed that the change from VC to V 4 C 3 coincided with increased trapping capacity and hence the C-vacancies at the carbide/matrix interface were considered to be the responsible trapping sites. Lee et al. (2016) also observed strong trapping by fine precipitated vanadium carbides, which could be related to the trapping at the carbon-vacancies at the broad (001) interface but reported a significantly lower E a of only 27 kJ/mol. Additionally, they indicated that large undissolved cuboidal V 6 C 5 , originating from hot rolling, could also trap H, as these carbides also are characterized by a broad (100) interface containing many C-vacancies. Additionally, vanadium carbides are reported to have a higher concentration of bulk C-vacancies than titanium carbides. A DFT study performed by Kawakami and Matsumiya (2012) indicated a very strong binding of H with a C-vacancy inside V 4 C 3 carbides, even of the same range as the C-vacancies inside the TiC. However, DFT calculations performed by Echeverri Restrepo et al. (2020) indicated an E B of H to the C-vacancies in vanadium carbide to be around 56 kJ/mol, which is considerably lower than the values for C-vacancies in TiC. Consequently, uncertainty remains regarding the trapping ability of these carbides at elevated temperatures. 2. Materials and Methods A generic Fe-C-V material was used in this study, for which the composition is given in Table 1. The steel was laboratory cast in a Pfeiffer VSG100 incremental vacuum melting and casting unit under an argon gas atmosphere and subsequently hot and cold rolled to a final thickness of 1.5 mm. Austenitization was performed in an air resistance furnace at two different temperatures, i.e. at 1250 °C and 950°C, in order to introduce different amounts of undissolved and precipitated carbides. Afterwards the material was quenched in brine. According to the VC and V 4 C 3 solubility products in steel (Gorni (2011)), all C and V is dissolved in the austenitic matrix at 1250°C. Austenitization of the Fe-C-V alloy at 950°C, led to the presence of undissolved vanadium carbides, as well. According to the solubility product of V 4 C 3 , which is reported as the most common form of vanadium carbides, only 0.252 wt% C and 1.425 wt% V was dissolved at this temperature. Tempering was performed at 600°C for 1 h, in a tube furnace, either in air or in a H containing atmosphere at ambient pressure. Consequently, gaseous H charging could be performed simultaneous with tempering. A H containing atmosphere was created by applying a constant flux of around 50 l/h of a gas mixture consisting of 95% N 2 and 5% H 2 , known as formier gas, after initial flushing of the atmosphere with the same formier gas. When tempering in air, a flux of 50 l/h air was also blown through the