HYDRO-REAL: Scientific Background
Hydrogen embrittlement (HE) is one of the most important detrimental processes in metals that limits their serviceability. Despite the fact that this phenomenon is known for more than a century, the exact mechanism behind and subsequently a definitive remedy for it are not yet fully understood. This ambiguity is reflected in the presence of many competing theories of HE in the literature which are enlisted in [1]. In this project, the interplay of retained austenite (RA) in high strength steels and H under realistic cyclic loading will be addressed. Both theoretical and experimental investigations indicate that RA acts as a trap for H in steels [2]-[3]. At the first look, this feature yields a reduction of the HE susceptibility of steels by preventing H form approaching the critical regions such as grain boundaries and nano-cracks. However, it can also be envisioned that during the lifetime of the steel the RA phases undergo a martensitic transformation and subsequently release the previously trapped H. These excess H atoms can damage the material in the vicinity of the previously present RA.
There is experimental evidence supporting both sides of the aforementioned dual role. On the one hand, Fan et al. [3] showed that the increase of RA content can decrease the HE susceptibility of martensitic stainless steel. On the other hand, Zhu et al. [4] observed an opposite behavior. Moreover, there is experimental evidence that H atoms are mainly concentrated at the interface between the martensite and austenite [5]. This observation is particularly important for applications with cyclic loading, in which the interfaces move. The kinetics of the martensitic transformation and trapping/releasing H atoms is crucial in these cases.
Alongside the experiments the simulations in different scales can be used to clarify the underlying processes of HE. At the very fine atomistic scale, ab initio simulations can unravel the trapping potency of different phases and interfaces [6]. Moreover, these simulations can clarify the effect of H on the different phase stabilities and the H diffusion barriers in each phase [7]. At a different scale, the larger MD/MS simulations can capture the collective behavior of H atoms at the critical regions such as GBs and cracks and clarify the nucleation and propagation of the cracks under different loading conditions [8]. Atomistically informed continuum models transfer the findings from ab intio calculations to finite element models, where long range diffusion and stress state are taken into account in multiphase microstructures. The theoretical framework describing diffusion and trapping of interstitial species in metals was derived from the thermodynamic extremum principle by Svoboda and Fischer [9] and implemented in a finite element framework by Drexler and coworkers [10]-[11]. Turc et al. used a multi-trap diffusion model to study the effect of austenite on H diffusion in ferritic-austenitic steels [12]-[13]. In this project we opt to perform a comprehensive study of the processes of HE in the presence of the fcc phase and interfaces between austenite and martensite across different scales.
References
Acta Mater. 60, 1211-1220, (2012)
Journal of Alloys and Compounds 826, 154057 (2020)
Materials Science and Engineering: A 800,140276 (2021)