In high-strength steels, such as FeMn alloys the stacking fault energy is a decisive indicator for the probability of a martensitic phase transition. Therefore, the development of methods for an ab initio determination of its value and temperature dependence is a central research activity of the group. One of the challenges of Fe-Mn alloys is the huge configuration space of possible atomic and magnetic structures. This challenge is tackled by combining and evaluating a large set of advanced methods, including the concept of γ-surfaces, ANNNI models as well as cluster expansion and quasirandom structures. The calculations for various Fe based compounds have revealed a hitherto unknown sensitivity of the stacking fault energy (SFE) to the composition  and the lattice expansion. In particular the dependence of the SFE on the C content  provided highly interesting insights on nano-diffusion during SFE measurements. These results allow a detailed understanding, why certain steels predominantly show the TRIP or TWIP effect, and a meaningful prediction of promising material compositions. The consequences are explored within the collaborative research center Steel ab initio.
The embrittlement experimentally observed in high-strength steels is closely related to another lattice defect investigated in the group, namely the incorporation of interstitial hydrogen. We therefore made intensive and systematic investigations on the solubility and diffusion of hydrogen in steels , including also the effect of superabundant vacancy formation due to hydrogen [4,5]. In order to consider hydrogen diffusion processes, kinetic Monte Carlo simulations in combination with transition state theory are performed. Apart from the hydrogen problem, we have used this tool for several other applications including self-diffusion in Fe-Al alloys and processes related to precipitate formation in steels.