Nucleation during phase transformations plays an important role in the crystal structure, the grain size and the texture of the forming product phase, and thus determines the properties of the obtained material. In this study, molecular dynamics simulation is employed to study the heterogeneous nucleation of bcc-phase in fcc iron. It is found that the bcc-phase nucleates at the dislocations in the fcc/fcc grain boundaries in a pseudo-cylindrical morphology. The energy change as a function of the bcc nucleus size conforms to the Cahn's classical model with no energy barrier, and provides interface energies and elastic constants comparable to theoretical calculations and experimental data. Nevertheless, there are aspects that cannot be explained by the classical Cahn nucleation theory, namely the stepwise “fcc→intermediate→bcc” nucleation process, and the aggregation of discrete subnuclei. This noclassical nucleation processes contribute to the decrease of energy barrier and the stabilization of the bcc nucleus.

Producing robust recrystallization models which can assist metallic microstructural design requires effectively understanding recrystallization nucleation. When the nucleation of static recrystallization (SRX) occurs at deformed grain boundaries, strain-induced boundary migration (bulging) is generally accepted as the nucleation mechanism. However, the present study challenges that view, showing, for a Ni-30%Fe alloy, that nucleation at deformed grain boundaries is not solely determined by bulging: results indicate that the number of bulges developed in the deformed microstructure is over four times larger than the number of SRX grains. On the other hand, SRX nucleation is shown to occur only when the low-angle boundary (LAB) between a pre-existing bulge and its parent grain transforms into a high-angle boundary (HAB). Based on this, a novel nucleation criterion is proposed, which may apply to SRX irrespective of the nucleation site (and to dynamic/metadynamic recrystallization): nucleation occurs whenever the misorientation of the LAB surrounding a bulge reaches the minimum HAB misorientation (e.g., 15°). Besides, correlation exists between the dislocation density accumulated around the various triple junction and grain boundary types in the microstructure, and their nucleation efficiency. This has been attributed to the higher fraction of relatively large initial subgrain misorientations measured for higher boundary dislocation density.

Preventive grinding of rails is a recurring maintenance routine to remove damage initiated in the wheel–rail contact. The grinding routine increases the service life of rails and reduces operational costs. Despite these benefits, grinding-related defects are observed. In this work a field test is performed to investigate the contact-surface formation and to better understand its durability. Surfaces after grinding are studied at different stages of the test to characterize wear mechanisms and deformation. The freshly ground surface exhibits a higher roughness and is composed of facets. It is determined that roughness asperities are extruded and fill grinding grooves in the process. High contact stresses at the facet transitions accelerate the extrusion of roughness asperities and the fast formation of the contact surface. The analysis further shows that deeper grinding grooves prevent homogeneous deformation. Strain concentrations arise due to the inhomogeneous deformation leading to damage initiation sites. These grooves are still present in the rail surface after the test. The evolution of the ground surface is captured in a schematic wear model.

This work investigates the formation of the recrystallisation microstructure and texture of various single-phase ferrite low-carbon steels that were rolled at different temperatures and of which the deformation microstructure was characterized by high resolution electron backscatter diffraction (EBSD). Three cases are considered: (i) cold-rolled interstitial-free (IF) steel, warm-rolled IF steel at 550 °C and warm rolled Fe-Si steel at 900 °C (below the austenitization temperature due to Si). It is well-known that the deformation texture after flat rolling of single-ferrite low carbon steels exhibits the characteristic α/γ-fiber texture, i.e. <110>//Rolling Direction (RD) and <111>//Normal Direction (ND), irrespective of the rolling temperature, as long as there is no concurrent phase transformation. However, different recrystallisation textures appear as a function of the rolling temperature. Generally speaking, the γ-fiber recrystallisation texture is obtained after cold rolling, whereas the θ-fiber components ( <100>//ND) intensify at the expense of the γ-fiber orientations with increasing rolling temperature. Although these phenomena are well-known, the reasons for this behavior in terms of preferential orientation selection remain as yet unclear. In the present paper, recrystallisation microstructures and textures are simulated with a full-field cellular-automaton (CA) description, whereby recrystallisation from its incipient stage is considered as a process of sub-grain coarsening controlled by the well-known physical laws of driving force and kinetics. The simulations integrate in one single model the various conditions that give rise to the observed temperature dependence of the evolving static recrystallisation texture and microstructure. The different rolling temperatures will give rise to different initial microstructures at the onset of recrystallisation with noticeable variations in short-range orientation gradients in γ and θ-fiber orientations, respectively. The mere application of local grain-boundary migration laws on the topology of the deformation structure, without imposing any specific nucleation selection criterion, will properly balance the dominance of γ-fiber grains after cold-rolling and θ-fiber orientations after warm rolling. Finally, the well-known nucleation of Goss orientations ({110}<001>) in shear bands occurring in γ-fiber grains is also simulated in this single conceptual framework.

In sparsely populated areas single-track railway lines are still common. Despite the low-density traffic and low axle loads, rail damage is observed to initiate in the rails of these lines. Not only large cracks requiring repair are found, but also newly initiated, post-grinding damage is observed. Rail specimens from the track containing representative damages are extracted to identify the reasons for the damage initiation in the R260Mn steel rail. At the rail surface, three years after the preventive grinding maintenance, the characteristic grinding facets and roughness patterns are still present. White etching layers are observed to surround the residual grinding grooves, maintaining the roughness as a result of the high wear resistance of these layers. The strain orientation at the gauge side of the rail is uni-directional due to lateral creep in the wheel–rail contact while in the center of the contact surface the strain patterns evidence shear stress reversal associated with the bi-directional traffic. The damage initiation mechanism after preventive grinding is associated with low-wear conditions and stresses concentrated around long-lasting grinding grooves. The findings show that preventive grinding maintenance specifications for regional single-stock railway lines must be improved. Specific points of interest are stringent requirements on the number of facets and surface roughness, as well as directions for the removal of corrugation.

Recycling of rare-earth elements (REEs) from NdFeB magnets is an important strategy to mitigate the risks associated with the REE supply chain. In this article, we propose an electrochemical process to recover REEs wherein all the reagents required for both leaching of REEs as well as the precipitation are generated in situ electrochemically. A three compartment electrochemical reactor was used in which a rare-earth containing salt along with an additive salt, ammonium sulfamate, was fed into the middle compartment. Upon electrolysis, the salts were split into acid and rare-earth hydroxides. The acid generated in the anolyte compartment was used to leach the NdFeB magnet waste. The rare-earth hydroxides were collected in the catholyte compartment and calcined to obtain rare-earth oxides. More than 95% of REEs and cobalt were extracted into the solution, and more than 85% of iron was removed as Fe(OH)₃precipitate in the same step. Subsequently, the leachate was oxidized and neutralized to remove more than 99% of iron. By using electrons as green reagents, this process combines leaching and precipitation in a single reactor enabling process intensification. The leachate produced at the end is rich in REEs and can be fed again into the middle compartment, forming a completely closed-loop process. Overall, the process consumes no acid, only electricity, ammonium hydroxide for neutralization, and an additive salt, ammonium sulfamate.

Increasing train speeds and the reduction of maintenance slots places high demands on the railway rails. To meet the challenging demands, producers regularly introduce new steel types. In this experimental investigation is the mechanical behavior of an air-cooled vanadium-alloyed hypereutectoid rail steel presented. The rail is produced applying conventional hot rolling of a reheated bloom and is then cooled on a cooling bed. The mechanical behavior is determined by performing standardized linear elastic fracture mechanics tests. The necessary specimens are extracted from new rails that are made in series production. Monotonic tensile test results have shown that the strain-hardenability of the steel is comparable to standard grade eutectoid rail steel and is higher than that of an accelerated-cooled eutectoid rail grade. The fracture toughness test results showed, statistically, no difference when compared with the fracture toughness values of the accelerated-cooled eutectoid rail grade. The tests were performed at room temperature. The fatigue crack growth rates are, in the linear Paris-regime, slightly higher than in the previously mentioned steels. The results are explained considering the distinct microstructural characteristics of the air-cooled vanadium-alloyed hypereutectoid steel and the fractured surface of the specimens. This experimental investigation contributes to selecting railway steels and predicting the actual in-service behavior.

Carbon segregation to defects in martensite is a phenomenon known for its occurrence and interference with mechanisms such as carbon partitioning in multiphase steels. Especially in martensite–austenite partitioning processes, carbon trapping at/de-trapping from martensite defects plays an important role since it interacts with the austenite enrichment. In this work, we develop a physics-based model in which we incorporate the concurrent evolution of carbon partitioning and trapping at/de-trapping from martensite defects. The model describes the global and local, time-dependent distribution of carbon between three lattice types, namely martensite defects, martensite solid solution, and austenite. We implement the model in mean-field and full-field descriptions, and discuss the interaction between carbon enrichment in austenite and segregation to martensite defects, on the basis of global equilibrium as well as on the carbon kinetics. We apply the model in several martensite — austenite microstructures and discuss the dependence of the interaction between carbon partitioning and trapping at/de-trapping from defects on specific microstructural features, i. e. phase fractions and microstructural banding.

Study of the cleavage behavior of heat treated S690 steel by a microstructure-based approach combined with finite element analysis is present in this paper. Cleavage simulations of steels subjected to heat treatments that cause grain refinement or simulate heat affected zones are performed, and are compared with experiments. It is found that the experimental improvement of toughness from grain refinement is 80% of what would be expected based on the model. The 20% difference is due to the lower number fraction of high-angle misorientation boundaries. It is also found that the resistance to micro-crack propagation is more effective in heat affected zones, which can be explained by the residual compressive stress in martensite-austenite constituents. This research assesses the balance between microstructural parameters for controlling cleavage toughness.

In many commercial steel processing routes, steel microstructures are reverted to an austenitic condition prior to the final processing steps. Understanding the microstructure development during austenitization is crucial for improving the performance and reliability of the microstructure that forms from austenite. In this work, austenite formation in a high-C steel (0.85 wt%) from a microstructure containing martensite/austenite and bainite bands is investigated. It is shown that austenite formation from bainite results in a refined austenite grain structure, and the martensite matrix thus obtained on quenching has a homogeneous distribution of carbides with a relatively low fraction of retained austenite (24%). On the other hand, a coarser austenite microstructure is obtained when austenite forms from a mixture of martensite and retained austenite. The reason for the coarse austenite grains is argued to be a memory effect, which is substantiated by in situ X-ray diffraction analysis. After quenching, an inhomogeneous carbide distribution and a higher retained austenite fraction (30%) are observed in the regions that were initially martensite/austenite. The global microstructure, hence, has a bimodal size distribution of prior austenite grains and carbide-dense bands. The causes for these heterogeneities are discussed with the help of interrupted quench experiments, equilibrium phase calculations, and DICTRA simulations.

In the present study, the nucleation of static recrystallization (SRX) in austenite after hot deformation is experimentally analyzed using a Ni-30 pct Fe model alloy. In agreement with the predictions by current models, nucleation rate exhibits a strong peak, early during SRX. Whereas such an early peak is explained by current models by the saturation of nucleation sites, this condition is far from reached, even after the peak declines. In addition, triple-junction and grain-boundary sites are shown to make a quantitatively similar contribution to nucleation. However, for a given boundary between deformed grains, nucleation predominantly starts at one of the triple junctions. Triple-junction nucleation initiates by strain-induced boundary migration of the nucleus (bulging) along one of the boundaries at the junction. Annealing twin boundaries contribute negligibly to nucleation through their grain-boundary sites. By contrast, their junctions with the boundaries of the parent grains do play a relevant role. The earlier nucleation at the triple junctions is attributed to the higher dislocation density observed around them, and the energy of the boundary consumed by the bulge. Both the maximum and average number of nuclei formed per boundary between deformed grains increase with increasing boundary length.

Multi-barrier cleavage models consider cleavage fracture which is characterized by a series of microscale events. One of the challenges for multi-barrier cleavage models is the strong variations of cleavage parameters across different types of steels. The source and magnitude of the variations have not been studied systematically. In the current paper, cleavage parameters corresponding to fracture initiation at a hard particle and crack propagation overcoming grain boundaries are determined for three bainitic steels, a martensitic steel, and a ferritic steel, using a recently proposed model. It is found that the particle fracture parameter depends on particle morphology and composition, while the grain boundary cleavage parameter depends on the hierarchical grain structure. The determined values of cleavage parameters present a high degree of consistency among the five different steels, which allows the further application on microstructure design to control macroscopic toughness.

The influence of carbon concentration variations on pearlite formation (20 h at 600 °C) in a case-carburized steel is investigated. The resultant microstructure shows three distinct regions: carburized case, a transition region, and the original core. The microstructural transition from the case to the core regions is observed to be relatively sharp. The investigated region of the carburized case (0.9 wt.% C) contains two types of pearlite: ferrite + cementite and ferrite + M₂₃C₆, where the pearlitic aggregate with M₂₃C₆ shows faster formation kinetics. The kinetics of pearlite formation in the transition region (0.3 wt.% C) is very slow and is observed with only M₂₃C₆ carbide. Only around 40% austenite decomposes into pearlite in the transition region, which, in comparison to the carburized case region of 0.9 wt.% C is a fraction that is lower by a factor of two. Pearlite is absent in the investigated core region (0.16 wt.% C). The microstructure in this region is predominantly martensite and pro-eutectoid ferrite, with a fraction of ferrite well below the equilibrium fraction. Ferrite formation in this region is limited by the redistribution of mainly Ni, Mn, and Cr, and their resulting solute drag effect on the austenite/ferrite interface. A thermodynamic and kinetic argumentation of these observations is provided with the help of thermodynamic data, precipitation simulations, and a general mixed-mode Gibbs energy balance model.

Modelling dislocation glide over the initial part of a stress–strain curve of metals received little attention up to now. However, dislocation glide is essential to ones understanding of the fundamental relationship between inelastic deformation and the evolution of the dislocation network structure. Therefore, we present a model of dislocation-driven deformation under static loading conditions. We reproduce repeated cyclic uniaxial tensile tests on Interstitial-Free and Low-Alloy steels. The elastic mechanical behaviour is described by isotropic linear elasticity, pre-yield anelastic mechanical behaviour by a dislocation bow-out model with dissipation, and the post-yield evolution of dislocation network structure by a statistical storage model. We hypothesise that when the local anelastic compliance is lower than the global plastic compliance, deformation is mechanically recoverable, and vice versa. This hypothesis is corroborated with the classical Taylor relation. We report the relation between stable and unstable dislocation glide using this prototypical modelling framework. We find four structural variables, that are based on dislocation physics, to describe the stress–strain curve: total dislocation density, average dislocation segment length, dislocation junction formation rate, and average dislocation junction length. Firstly, we quantify the dislocation network evolution during uniaxial monotonic loading, and verify work-hardening by dislocation junction formation and a Taylor-type equation for flow. Finally, we present a semi-empirical relation for the evolution of the dislocation network structure. Which allows us to: refine the physical interpretation of the Taylor relationship, and rationalise experimental observations on apparent modulus degradation by thermomechanical processing. Both these findings circumvent the limitations of current, physics-based hardening models.

High strength steels are widely used for structural applications, where a combination of excellent strength and ductile-to-brittle transition (DBT) properties are required. However, such a combination of high strength and toughness can be deteriorated in the heat affected zone (HAZ) after welding. This work aims to develop a relationship between microstructure and cleavage fracture in the most brittle areas of welded S690 high strength structures: coarse-grained and intercritically reheated coarse-grained HAZ (CGHAZ and ICCGHAZ). Gleeble thermal simulations were performed to generate three microstructures: CGHAZ and ICCGHAZ at 750 and 800 °C intercritical peak temperatures. Their microstructures were characterised, and the tensile and fracture properties were investigated at − 40 °C, where cleavage is dominant. Results show that despite the larger area fraction of martensite-austenite (M-A) constituents in ICCGHAZ 750 °C, the CGHAZ is the zone with the lowest fracture toughness. Although M-A constituents are responsible for triggering fracture, their small size (less than 1 μm) results in local stress that is insufficient for fracture. Crack propagation is found to be the crucial fracture step. Consequently, the harder auto-tempered matrix of CGHAZ leads to the lowest fracture toughness. The main crack propagates transgranularly, along {100} and {110} planes, and neither the necklace structure at prior austenite grain boundaries of ICCGHAZs nor M-A constituents are observed as preferential sites for crack growth. The fracture profile shows that prior austenite grain boundaries and other high-angle grain boundaries (e.g., packet and block) with different neighbouring Bain axes can effectively divert the cleavage crack. Moreover, M − A constituents with internal sub-structures, which have high kernel average misorientation and high-angle boundaries, are observed to deflect and arrest the secondary cracks. As a result, multiple pop-ins in load-displacement curves during bending tests are observed for the investigated HAZs.

The austenitization of an initial pearlitic microstructure is simulated using the phase field model to achieve insight into White Etching Layer (WEL) formation in pearlitic railway steels. The simulations take into account the resolution of the cementite lamellae within a pearlite colony as well as the presence of pro-eutectoid ferrite. The austenite growth kinetics and morphology obtained via simulations are compared with dilatometry and microscopy observations. The influence of γ/θ and γ/α mobilities on the austenite growth morphology are studied. The simulations reproduce the microstructural features as well as the experimentally observed kinetics behavior of austenite formation, involving the correlation between mobilities and nucleation behavior.

The reversible behaviour of metals at low applied stresses is more complex than the generally assumed linear behaviour. This is primarily because of the reversible nature of dislocation motion leading to a strain contribution known as anelasticity. This work aims to investigate (a) quantification of dislocation structures in industrial grade stainless steels, (b) unloading behaviour, and (c) the fundamentals of reversible and mechanical behaviour occurring below the yield stress. Mechanical testing of martensitic stainless steel (Stavax ESR) was performed in two different modes: incremental plastic deformation and cyclic loading-unloading below the yield stress with a focus on the measurement of small strains and corresponding stresses, occurring in the pre-yield regime. The non-linear reversible behaviour was quantitatively analysed, as opposed to the common approximation of an empirical determination of apparent Young's modulus. The recently proposed pre-yield model has been refined and, for the first time, successfully applied to a complex microstructure such as stainless steel. The quantification of dislocation structure parameters is shown to be an efficient alternative to the conventional experimental methods of quantifying dislocation structure. Further, a unique representation and quantification of the unloading and hysteresis behaviour provides more insight into the material behaviour. Lastly, the little studied microplasticity occurring below the yield stress upon cyclic loading-unloading was determined. Importantly, the physical basis of the model will allow quantification of plastic deformations in the pre-yield region, large enough to be significant in industrial processes.

High-resolution three-dimensional crystal plasticity simulations are used to investigate deformation heterogeneity and microstructure evolution during cold rolling of interstitial free (IF-) steel. A Fast Fourier Transform (FFT)-based spectral solver is used to conduct crystal plasticity simulations using a dislocation-density-based crystal plasticity model. The in-grain texture evolution and misorientation spread are consistent with experimental results obtained using electron backscatter diffraction (EBSD) experiments. The crystal plasticity simulations show that two types of strain localization features develop during the large strain deformation of IF-steel. The first type forms band-like areas with large strain accumulation that appear as river patterns extending across the specimen. In addition to these river-like patterns, a second type of strain localization with rather sharp and highly localized in-grain shear bands is identified. These localized features are dependent on the crystallographic orientation of the grain and extend within a single grain. In addition to the strain localization, the evolution of in-grain orientation gradients, misorientation features, dislocation density, kernel average misorientation, and stress in major texture components are discussed.