Massive Transformation - experiments and simulations

Despite numerous efforts in the past there are still some open questions concerning massive transformations in general and the massive a m transformation in titanium aluminides in particular. The growth mechanism has to be clarified and it is not yet clear if a concentration (mole fraction) spike in front of the moving interface is a typical feature of a massive transformation. In addition, a detailed knowledge about the overall kinetics during massive transformations is missing. Diffusion processes in massive transformations occur only in the interface separating the parent and the product phase and / or in its nearest surroundings. Thus, in metallic systems the length scale of interest is in the range of nanometers and below. Diffusion in massive transformations may include phenomena like trans-interface diffusion, solute segregation in the moving interface and diffusion in a mole fraction spike occurring in front of the interface. Highly sophisticated experimental techniques (e.g. energy dispersive X-ray spectroscopy (EDS) and electron energy-loss spectroscopy (EELS) within a transmission electron microscope (TEM)) are required to investigate the composition across the interface and the interface-near region. Although the interface velocity can be estimated from TEM-studies, these experiments are limited to a small observation range. Dilatometer studies have to be performed in order to obtain the volume fraction of the new phase as a function of time and temperature. In case that the evolution of the parent / product phase arrangement can be estimated and that the driving force for the transformation is known the interface velocity can be calculated. It is far from being state of the art, but certainly a challenging task to describe the above mentioned diffusion processes and the simultaneously occurring migration of an incoherent interface within the framework of a micromechanical model. The driving force for the massive transformation has to be calculated from the temperature- and composition-dependent Gibbs energies of the phases. The calculation of these Gibbs energies is non-trivial in many cases and atomistic modeling is required to improve the existing knowledge. Then, the time- dependent mole fraction profiles as well as the interface velocity can be calculated with the micromechanical model and compared to the experimental findings. The study of massive transformations includes several length scales (e.g. inter-atomic distances and a macroscopically detectable progress distance of the transformation front) and time scales (e.g fast motion of the interface and slow diffusion of substitutional components). Only the combination of theoretical models on an atomistic level and on a continuum scale with various experiments is expected to provide additional insight into the underlying physics of massive transformations. Massive transformations are not only of academic but also of industrial interest since they can determine the properties of some important engineering materials as e.g. special low alloyed steels or titanium aluminides.

Massive transformation (MaT) of a certain material from a parent phase into a product phase (e.g. from a hexagonal lattice into a tetragonal lattice) is a nearly sudden process with a moving interface separating both phases. The local rearrangement of atoms is performed by a near-field diffusion process and not by cooperative motion of the atoms (martensitic transformation). Although MaT occurs during processing of a wide class of industrially relevant materials (e.g. steels, ceramics or intermetallics), no clear physical picture exists about the MaT process in the current literature. Therefore, an upcoming new high tech material (Ti-Al-Nb), which shows such a phenomenon as MaT during the cooling process, was investigated both experimentally and theoretically. State of the science electron microscope measurements were performed in the nanoscale range in order to investigate the interfacial region. The goal was to find so-called spikes in the composition in front of the moving interface as an indicator for a MaT. Furthermore, a sophisticated non-equilibrium thermodynamic model has been developed which predicts the local compositional changes in the interfacial region and its surroundings. This model is based on the thermodynamic extremal principle and must be fed by thermodynamic data. These data can either be based on experimental investigations and/or ab-initio calculations. On the one hand the theoretically found spikes are at or even below the limit of experimental resolution due to the unexpectedly large fluctuations in chemical composition within the individual grains. On the other hand the sluggish bulk diffusion process of the substitutional components could be partly suppressed as has being frequently the case during the austenite-to-ferrite transformation in low alloyed steels. Finally, the project shows that MaT-specific spikes in the composition can only be detected by the combined use of modern experimental techniques and numerical calculations based on non-equilibrium thermodynamics. The concept for the detection of spikes and consequently the diagnosis of massive transformation, as developed in the project, is not restricted to the investigated Ti-Al-Nb alloy and applicable for a wide class of materials.