Symposium MB7 : Shear Transformation Mechanisms and Their Effect on Mechanical Behavior of Crystalline Materials

Twinning induced plasticity (TWIP) enables designing austenitic steels with >70% elongation and >1 GPa ultimate tensile strength. They are characterized by high strain hardening due to the formation of twins and dense dislocation substructures. Twins impede dislocation glide. They act as barriers to dislocations, promoting their storage and decreasing their mean free path. Twinning depends on the stacking fault energy (SFE). In face-centered cubic (fcc) alloys the SFE can be tuned by alloying. This opens compositional access to ‘optimal strain hardening design’. This term refers to a state where twins and the associated substructures are not forming instantly upon initial yielding but over a wide loading range to enable permanent dynamic reduction of the dislocation free path and hence continuous strain hardening. Here we apply this design concept to twinning induced plasticity (TWIP) steels. For understanding how strain-hardening responds to a change in SFE via alloy tuning, the first step lies in identifying its dependence on composition. With this knowledge the description of strain hardening can be rendered chemistry sensitive. In a second step we have to understand the underlying strain hardening mechanisms, i.e. twin formation, cross slip, twin–slip, slip–slip and twin–twin interactions. While the first step is accessible to thermodynamic calculations on the basis of ab initio derived interface energies, the second step can be rendered formal by mapping the individual strain hardening effects into a mean field model which uses internal variables and couples the twin nucleation rate to the dislocation substructure and to the SFE. Validation measures are dislocation densities and twin volume fractions determined by electron channeling contrast imaging (ECCI) and transmission electron microscopy (TEM). Texture and grain size effects are mapped by using electron backscatter diffraction (EBSD). This connection renders these steels ideal model materials for theory-based alloy design: Ab-initio guided composition adjustment is used to tune the stacking fault energy and thus the strain hardening. We extend this concept also to the case of weight reduced austenitic steels.

Shear transformations have been extensively studied, particularly related to advanced alloys such as high-strength transformation-induced-plasticity steels (TWIP steels), shape memory alloys and super-elastic beta-titanium alloys, the so called gum alloys. Here, through a DFT-guided alloy design approach which enables compositional tuning of the stacking fault energy, we introduce shear transformation into a novel five-component non-equiatomic CoCrFeMnNi high-entropy alloy (HEA). The transformation from a stable face-centered cubic (FCC) matrix into a hexagonal close-packed (HCP) martensitic phase upon tensile deformation was examined using a set of probing techniques such as X-ray diffraction (XRD), electron backscatter diffraction (EBSD), electron channeling contrast imaging (ECCI), energy dispersive spectroscopy (EDS) and atom probe tomography (APT). We found that shear transformation is the primary deformation mechanism in the homogenized coarse-grained non-equiatomic HEA. Moreover, through the modification of the alloy’s grain size, the transformed HCP phase fraction could be tuned, thus the mechanical behavior of the alloy can be further tailored in this way, rendering the transformation effect a size dependent constitutive response in this material. Compared to the corresponding equiatomic CoCrFeMnNi alloy, the non-equiatomic HEA exhibits significantly improved strength and ductility due to the coupling effects of shear transformations and other deformation mechanisms.

TWIP steels are materials with high strength and exceptional strain hardening capability, parameters leading to large energy absorption before failure. However, TWIP steels also exhibit reduced (often negative) strain rate sensitivity (SRS) which limits the post necking deformation. In this study we demonstrate for an austenitic TWIP steel with 18% Mn a strong dependence of the twinning rate on the strain rate, which results in negative strain hardening rate sensitivity (SHRS) [1]. The SRS is observed to decrease with strain, becoming negative at larger strains. Further, we probe the SRS at various scales with nano- and micro-indentation and observe that the effective strain rate sensitivity is scale dependent [2]. The SRS evaluated based on the nano-scale hardness is positive, while that evaluated based on micro-scale hardness is smaller and becomes negative as the size of the indenter and hence the volume of material probed increase. The effect is linked to the dislocation-twin interaction mechanism. The mechanical tests are supported by detailed microstructural observations with electron microscopy, atomic force microscopy and electron backscatter diffraction.
[1] A. Bintu, G. Vincze, R.C. Picu, A. Lopes, J. Gracio, F. Barlat, Mat. Sci. Eng. A629, 54, 2015.
[2] A. Bintu, G. Vincze, R.C. Picu, A. Lopes, I. Bdikin, Mat. Sci. Eng. A, in press, 2016.

The displacive phase transition of β to ω has been actively investigated for several decades in titanium base alloys because of its complex transition mechanism and its significant influence on mechanical properties. The displacive β to ω transition occurs on rapid cooling from the β phase field or under loading in Ti-V alloys when the content of V is lower than 30 at.%. Thus, the Ti-V alloys have β+ω microstructures where the ω phase precipitates in the parent β phase. However, theoretical calculations had already indicated that the ω phase is thermodynamically stable than the β phase in Ti-V alloys with the content of V as high as above 30 at. %., which contradicts experimental results. This implies that there is energy barrier for the displacive phase transition of β to ω. In this paper, ab initio calculations were performed to probe the energy landscape of the transition pathway, i.e:, the free energies of the intermediate transition state between the β and ω structures of Ti-(15-30) at.% V alloys at temperatures of 0 and 300 K, based on shuffle mechanism of purely displacive collapse of the {111} planes of the β phase. The calculated results show that energy barrier appears for the displacive phase transition of β to ω in Ti-(15-30) at.% V alloys at 300 K, but does not at 0 K. The existence of energy barrier suggests that the β to ω transition is not spontaneous and requires contribution of external stimuli acting as the driving force for the transition. The energy barrier increases from 3.96 to 21.78 meV/atom with an increase of V content of from 15 to 30 at.% in Ti-V alloys. This result agrees well with experimental observations, in which the maximum mass fraction of ω phase appears with V content between 10 and 15 at.%, and the athermal ω phase transition occurs less significantly with increase of V content over 15 at.% in Ti-V alloys.

Using high-resolution transmission electron microscopy (HRTEM) and atomistic simulations (density function theory and molecular dynamics), we explored zero-shear transformation mechanisms for (i) twinninmg in face-centered cubic metals, (ii) nucleation of a face-centered cubic titanium in hexagonal close packed titanium, (ii) transformation for zinc-blende InAs in wurtzite InAs nanowires, and (iv) nucleation of deformation twin {10-12} in hexagonal metals. The growth of these structures can be accomplished through either shear shuffle or pure shuffle mechanisms, depending on whether marco-scale shear strains will be resulted.

Among the different families of titanium alloys, β-stabilized Ti alloys have attracted a lot attention during the last years. However, these alloys generally suffer from lack of work hardening, which limits their ductility. Recently, several β-metastable Ti grades have been designed to exhibit simultaneous TRIP (TRansformation Induced Plasticity) and TWIP (TWinning Induced Plasticity) effects [1,2]. It has been demonstrated that the work hardening capacity of the as-quenched β phase is improved when stress-induced α” martensitic laths and mechanical twins appear in the course of deformation, forming obstacles to dislocation glide [3].

Twinning and martensitic transformation imply local strains which must be accommodated by neighboring grains. In this work, a three-dimensional crystal plasticity finite element simulation is carried out to investigate the local stress fields around deformation twins and martensites. Effects of various factors on the local stress fields, including grain orientations, morphology of the twin/martensite, are considered. The numerical predictions are assessed against experimental observations after a tensile test on a Ti-12 wt.% Mo alloy, for which the local elastic stress field and geometrically necessary dislocation (GND) content was determined experimentally using the high-resolution electron backscattered diffraction (HR-EBSD) technique [4].

[1] M. Marteleur et al., “On the design of new β-metastable titanium alloys with improved work hardening rate thanks to simultaneous TRIP and TWIP effects”, Scripta Materialia, 66 (10) (2012), 749-752.
[2] F. Sun et al., “A new titanium alloy with a combination of high strength, high strain hardening and improved ductility”, Scripta Materialia, 94 (1) (2015), 17-20.
[3] F. Sun et al., “Investigation of early stage deformation mechanisms in a metastable β titanium alloy showing combined twinning-induced plasticity and transformation-induced plasticity effects”, Acta Materialia, 61 (17) (2013), 6406-6417.
[4] N. Isasti et al., “Analysis of Complex Steel Microstructures by High-Resolution EBSD”, JOM, 68 (1) (2016), 215-223.

The deformation mechanisms and mechanical properties of Fe-Mn-C-Al twinning-induced plasticity (TWIP) steels of a chemical composition range of 12-18% Mn and 0-3% Al, are reviewed. The in-depth microstructural analysis by means of electron-backscatter diffraction (EBSD) and transmission electron microscopy (TEM) revealed that all the investigated TWIP steels exhibit deformation twinning as the main deformation mechanism in addition to dislocation glide. The nucleation and the growth mechanisms of deformation twin developed from emitted stacking faults from grain boundaries will be discussed. Twin thickness shows a tendency to decrease with decreasing stacking fault energy (SFE). The Al-free TWIP steels show a much more complex deformation behaviour than the Al-added TWIP steels. Deformation of Fe15Mn0.6C steel is accompanied by a very small amount of strain-induced ε martensite, in addition to deformation twinning, and deformation of Fe12Mn0.6C steel is accompanied by several deformation mechanisms which are simultaneously activated, such as strain-induced ε martensite, formation of shear bands and strain-induced α′ martensite, in addition to deformation twinning. Based on the in-depth microstructural characterization, the SFE range for twinning and the simultaneous occurrence of twinning and ε or α′ martensitic transformation in TWIP steel was determined. The upper limit for the value of SFE for strain-induced martensitic transformation was determined to be approximately 13 mJ/m². The results confirm that the SFE is the key parameters affecting the strength and the ductility of TWIP steel. A linear relation between the ultimate tensile strength (UTS) and the SFE is found, i.e. with the UTS increasing with decreasing SFE. The Fe18Mn0.6C steel, which has a SFE slightly above the upper limit SFE value for strain-induced martensitic transformation achieves the highest combination of tensile strength and elongation.

TRansformation Induced Plasticity (TRIP) is an attractive route to increase the strain hardening capacity of multiphase steels containing a fraction of metastable austenite in order to raise both the strength and uniform elongation [1]. Excellent performances have been already attained, with a recent recrudescence of interest through the development of third generation of high strength steels often involving TRIP effect. Nevertheless, microstructure and composition optimization is complex due to the interplay of coupled effects on the transformation kinetics and work hardening such as phase stability, size of retained austenite grains, temperature and loading path. In particular, recent studies have shown that the TRIP effect can only be quantitatively captured for realistic microstructures if strain gradient plasticity effects are taken into account, although direct experimental validation were missing [2].
An original computational averaging scheme has been developed for predicting the elastoplastic response of TRIP aided multiphase steels based on the Fleck Hutchinson 2001 [3] strain gradient plasticity model. The microstructure is represented by an aggregate of many elementary unit cells involving each a fraction of retained austenite with a specified stability. The model parameters, involving the transformation kinetics, are identified based on experimental results at different temperatures. The model is then validated by comparison with other data involving temperature changes during the deformation [4]. While classical plasticity is not able to capture the TRIP effect on the mechanical response, the strain gradient version properly predicts the variation of the strain hardening with deformation and temperature, hence the uniform elongation. A parametric study is performed to get more insight on the effect of the material length scale as well as to determine the optimum transformation kinetics to reach the highest possible strength ductility balance. It is shown that the uniform elongation can potentially be increased by a factor 1.5 or more, guiding future microstructure engineering efforts.

References
Jacques P.J.: Transformation-induced plasticity for high strength formable steels. Curr. Op. Solid State Mat. Sci 8(3):259-265, 2004.
Mazzoni L. et al.: Strain gradient plasticity analysis of transformation induced plasticity in multiphase steels. Int. J. Sol. Struct 45(20):5397-5418, 2008.
Fleck N. A., Hutchinson J.W: A reformulation of strain gradient plasticity. J. Mech. Phys. Sol. 49(10):2245-2271, 2001.
Delannay L. et al.: Modelling of the plastic flow of trip-aided multiphase steel based on an incremental mean-field approach. Int. J. Sol. Struct, 45(6):1825-1843, 2008.

Dislocation interaction with and accumulation at twin boundaries have been reported to significantly improve the strength and ductility of nanostructured face-centered cubic (fcc) metals and alloys. Here I describe plausible dislocation interactions at twin boundaries and twin-twin interactions in nanocrystalline fcc metals. Depending on the characteristics of the dislocations and the driving stress, possible dislocation reactions at twin boundaries include cross-slip into the twinning plane to cause twin growth or detwinning, formation of a sessile stair-rod dislocation at the twin boundary, and transmission across the twin boundary.

The localization of stress in FCC polycrystals containing random grain boundaries was studied through molecular dynamics (MD) modeling. Large scale MD was used to study the local stress states along these random boundaries. We particularly study the role of stress localization in the interaction of grain boundaries with dislocations. We found that variations in the local grain boundary structure can affect stress localization and be critical to the deformation behavior of the overall polycrystal.

The core structure of an a/2[112] dislocation was determined in Cu, Ni and Ni-15Al . It was found that the core dissociates on successive [111] planes, forming the nucleus of a twin in all three materials. Three a/6[112] type Shockley partials, separated by a SISF and an SESF, form the core. Under glide stresses greater than g/2b, where g is the stacking fault energy and b the Burger’s vector of the Shockley partials, the SESF portion of the a/2[112] edge core extends indefinitely, forming a 2-layer twin. Incidence of a/2[110] type screw dislocations gliding on the cross-slip plane at this nucleus should result in growth of the twin. The twinning onset stress in this mechanism, g/2b, is in fairly good agreement with experimentally measured values for twinning onset stresses in several fcc metals. It is found that the multilayer core for the a/2[112] dislocation is preferred over splitting of the a/2[112] dislocation into two a/2[110] type lattice dislocations at low stacking fault energies and for near 30^o orientations. Such an a/2[112] dislocation in the g matrix entering a g’ precipitate in a superalloy can undergo a low activation energy core transformation and defeat the precipitate easily forming a low energy SESF in the precipitate. This is suggested to be the mechanism of microtwinning in superalloys under intermediate temperature, high stress creep conditions. Such a process allows for the defeat of the precipitate every other [111] plane, as is found in experiments and does not require diffusional rearrangement of atoms within the precipitate.

The formation of twin bundles in face centered cubic (fcc) metals is known to influence both the texture development [5], the anisotropy and the hardening [2]. Indeed, mechanical twins shorten the mean free path of dislocations along impinging slip planes and this promotes dislocation slip parallel to the twinning plane. Twin bundles thus increase the plastic anisotropy of individual grains and this, in turn, favors the occurrence of microscopic shear bands [3, 4] that cut through the bundles [7].

A crystal plasticity model has been adapted in order to account for the interplay of mechanical twins and localized shear bands. The latter mechanical instabilities are predicted as a result of texture softening inside the bands [1] and they have a significant effect on the grain lattice rotation.

Model predictions have been compared to experimental observations of pure copper after high strain rate uniaxial compression at liquid nitrogen temperature. Such dynamic plastic deformation (DPD) produces a nanoscale microstructure in which twin bundles are formed mostly in grains with a <100> crystal direction aligned with the compression direction [6]. Twin lamellae are then inclined at about 30° to the compression plane. When DPD induces larger compressive strain, grains without twins develop a <110> fiber texture, which is typical for compression of fcc metals. In contrast, twin bundles rotate gradually towards the compression plane, corresponding to a <111> fiber texture. Relying on the crystal plasticity model, it is demonstrated that micro shear bands must be considered in order to capture the texture development.

References
[1] R.J. Asaro and J.R. Rice, Strain localization in ductile single crystals. J. Mech. Phys. Solids 25:309-338, 1977.
[2] S. Dancette, L. Delannay, K. Renard, M.A. Melchior and P.J. Jacques, Crystal plasticity modeling of texture development and hardening in TWIP steels. Acta Mater. 60:2135-2145, 2012.
[3] N. Jia, F. Roters, P. Eisenlohr, C. Kords and D. Raabe, Non-crystallographic shear banding in crystal plasticity FEM simulations: Example of texture evolution in -brass. Acta Mater. 60:1099-1115, 2012.
[4] E. El-Danaf, S.R. Kalidindi, R.D. Doherty and C. Necker, Deformation texture transition in brass: critical role of micro-scale shear bands. Acta Mater. 48:2665-2673, 2000.
[5] T. Leffers and R.K. Ray, The brass-type texture and its deviation from the copper-type texture. Prog Mater Sci. 54:351-396, 2009.
[6] F.X. Lin, Y.B. Zhang, N.R. Tao, W. Pantleon and D. Juul Jensen, Effects of heterogeneity on recrystallization kinetics of nanocrystalline copper prepared by dynamic plastic deformation. Acta Mater. 72:252-261, 2014.
[7] H. Paul, A. Morawiec, J.H. Driver and E. Bouzy, On twinning and shear banding in a Cu8 at.% Al alloy plane strain compressed at 77 K. Int. J. Plast. 25:1608-2145, 2012.

Nanotwinned metals have been extensively studied as they typically exhibit high strength and ductility. Recent studies show that growth twins or stacking faults can be introduced into Al with high stacking fault energy. Here we will summarize several methods that have been adopted to introduce growth twins into epitaxial and nanocrystalline Al and Al alloys. Furthermore in epitaxial nanotwinned Al, the predominant growth defects are incoherent twin boundaries. The deformation mechanisms of Al investigated by ex situ and in situ nanoindentation (in TEM) technique will be presented. In situ nanoindentation studies show that Al with high density growth twins has high strength and prominent work hardening.

Recent experiments on small-volume metals have shown that deformation twinning (DT) initiates at an ultra-high stress and on a very short timescale. This indicates strongly correlated dislocation dynamics, which remain poorly understood. Using atomistic calculations, we demonstrate that under a wide range of laboratory experimental conditions DT can be accomplished by surface "rebound" of strongly overdriven dislocations. A dislocation nucleated under high stresses can accelerate to a significant fraction of the sound speed within a distance of 10¹ nanometers. The resulting high-speed dislocation rebounds back when hitting a free surface. The ensuing rebounds, back and forth from opposing surfaces, lead to self-sustained breeding of twinning partial dislocations, directly initiating DT. Due to its stronger temporal correlation, surface rebound sustained (SRS) relay of twinning dislocations is shown to be dominant in initiating DT over thermally activated nucleation (TAN). This dislocation breeding mechanism is also expected to play a role in other high-stress and high-strain-rate deformation processes.

To enhance the knowledge about deformation twinning in bcc materials, a study on the correlation between compression loading and twinning was carried out. To obtain the formation of twins in bcc iron, samples were stressed in quasi-static (strain rate 10⁰ s^-1) and impact (strain rate 10⁴ s^-1) compression experiments. Besides the pure iron also iron solid solution alloys were used for the experiments. Thereby the influences of the temperature in the range of -196°C to 20°C and the grain size in the range of 40 µm to 200 µm were determined. By means of EBSD, the orientation conditions and texture were characterized before and after compression. Using image analyzing software, the amount of twinned area fraction was quantified and correlated with the experimental parameters. The grain size influence could be shown as the major effect on twinning during high strain deformation, whereas temperature influence was just a minor effect.
A further aim in this study was the identification of the nucleation mechanism referring to the experimental set-ups. According to the literature, different mechanisms for twin nucleation in bcc materials are known. Particularly the pole mechanism (Cottrell and Bilby, 1951), the slip dislocation interactions (Priestner and Leslie, 1965), the dislocation core dissociation (Sleeswyk, 1963) and the α→ε→α phase transition (Minshall, 1955) are different approaches to explain twin nucleation in the bcc lattice. In this study, the evolution of twins and the impact of dislocations on twinning were analyzed based on molecular dynamics simulations. The results indicate that two of these mechanisms mainly occur. For impact scenario with low grades of compression, the pole mechanism is favored. For high grades of compression, the α→ε→α phase transformation becomes the dominating twin nucleation mechanism.

Twinning - a mechanism by which a lattice reorients itself in order to accommodate deformation - plays an important role in many different classes of materials. Although twins are observed in all the common metallic crystal structures, twinning plays an especially important role in the deformation of hexagonal close-packed (HCP) materials such as magnesium, a promising basis for the engineering of the next generation of lightweight alloys to replace aluminum and steel. Though the existence of twinning in HCP is known, a comprehensive understanding of details - in particular, which modes are possible and likely to form in a given material - is still an area of active research and necessary for the complete understanding of material deformation mechanisms. To provide a comprehensive catalog for all twins worth consideration given an arbitrary material, we propose a novel framework, extending upon ideas developed by Ericksen, Pitteri, Ball, and James and systematically predicting all possible twin modes given only basic lattice information. We then consider the energetics of these twin configurations, first identifying twin modes of interest by making use of molecular statics simulations and then refining our results through ab initio simulations with a density functional theory code developed within our group, MacroDFT. The energetics picture is then completed by examining the path to twin formation through nudged elastic band simulations, resulting in the prediction of a set of twin modes which are likely to be visualized that differs from what previous works of literature have considered. We describe how we use these findings, with particular attention to HCP magnesium, to supplement experimental identification of twin modes. We then discuss future directions, such as integration with models at higher length scales, in order to take steps towards the computational design of a new generation of lightweight, strong materials.

Nanoindentation provides the opportunity to study single twinning events. In the present work, a nanoindenter equipped with spherical tips of radii R = 5, 10 and 50 µm was employed to examine a range of magnesium alloys as well as a single crystal of pure Mg. Our objective is to correlate features of the load displacement curve with twinning events and to ascertain twin initiation and growth stresses as well as the general phenomenology of the twinning events. Samples prepared using OPS polishing all displayed a yielding point on the load-penetration depth curve corresponding to departure from Hertzian contact. This corresponded to the appearance of basal slip lines on the sample surface. Pop-in events were then observed at higher loads and these were marked by the simultaneous appearance of twins, evident on the surface following unloading. Twin thickening during continued penetration and shrinkage during unloading were also detected. Crystal plasticity finite element modelling was employed to estimate the stress state prior to the appearance of twinning. This enables us to estimate the critical stress for twin initiation but there is still considerable uncertainty in the values the non-unique nature of the model predictions. Annealing the samples after OPS polishing enables twin initiation stresses to be established in the absence of prior basal slip, using Hertzian contact. These initiation stresses are found to be approximately 588 MPa. Analysis of pop-in loads and depths reveals an effect of alloying addition on friction stresses for twin growth.

Zn is a hexagonal metal with a large c/a ratio under ambient conditions (c/a=1.856). Under ambient testing conditions, deformation is predominantly accommodated by basal slip and by {10-12} <11-23> compression twinning. Remarkably, hydrostatic pressure drastically influences the c/a ratio of Zn with a decrease of c/a with increasing pressure. As a consequence, the compression twin should become a tensile twin at pressures > 9 GPa.

In this work we strain-cycle a wire of pure Zn in the D-DIA deformation press under multiple superimposed hydrostatic pressures ranging between
3 and 17 GPa. Over this pressure range, the c/a ratio of Zn will go over the the compressive-tensile transition. During deformation, the state of sample is monitored in-situ through powder x-ray diffraction, allowing the extraction of texture and sample stress.

The purpose of this work is three-fold: to determine if detwinning is a possible mechanism at 3GPa pressure when Zn is cycled in tension-compression, to find out if at pressure state where c/a ~
sqrt(3) only basal slip is active or whether another slip mechanism operates, and to find out whether {10-12}<11-23> twins tensile are active in Zn when c/a is below sqrt(3).

We expect that elasto-plastic polycrystal simulations of the cyclic process will allow us to interpret the experimental data and to elucidate the type and strength of the crystallographic deformation mechanisms under pressure conditions.

As the most widely reported twin mode, {10-12} twins are easily activated in magnesium alloys by compression perpendicular to the c-axis of the crystal. The intersection of two {10-12} twins was extensively discovered at grain boundaries with low misorientation angles, which is similar to the case of two {11-22} contraction twins in pure titanium. Primary {11-22} twins accommodated by adjacent {10-12} extension twins were found in deformed pure titanium. These twin pairs composed of {11-22} twins and {10-12} twins are observed at high angle grain boundaries with misorientation angles close to 90°. The mechanism accounting for the variant selection of the adjoined {10-12} twins is examined in terms of Schmid factor and accommodation capacity of potential twin variants. The selection of the twin variants is dependent on the accommodation capacity of the potential twin variants while it is not controlled by the applied stress. {11-24} and {11-21} twins are also generated in the sample with large grain size. A small amount of double twins can be produced within {11-24} twins.

Recent years have seen the emergence of mechanical twinning as the source of outstanding mechanical properties in FCC austenitic steels. Similarly to the TRIP (Transformation-induced plasticity) effect, the TWIP (Twinning Induced Plasticity) effect is now considered as an effective mechanism for increasing the work hardening rate close to the theoretical limit. Recently, several β-metastable Ti grades have also been designed to exhibit simultaneous TRIP and TWIP effects bringing work hardening rates never reached before. The alloy design method, based on Morinaga’s approach involving the calculation of two electronic parameters, aimed at the synergetic activation of α” stress-induced martensitic phase and mechanical twinning with ω precipitation.

This work investigated, at the grain-scale, the activation of the TRIP and TWIP effects in both steels and Ti alloys. The mechanisms of twinning and martensitic transformation during straining were highlighted owing to in situ TEM and EBSD characterization. Analysis of the interactions of dislocations with the interfaces of these features seems to play a key role in the increase of the work hardening. In-situ synchrotron x-ray diffraction helped in characterizing the mechanical consequences on these plasticity phenomena.

Furthermore, specific patterning of the twins or martensite laths depending on grain size is observed in both steels and Ti alloys, with some proportionality in the characteristic lengths. It is proposed that this patterning also influences in a large way the mechanical response of these materials.

Due to their very low activation stress, deformation twinning, the main plastic carrier of magnesium based alloys has been recognized as the origin of the several major unfavorable mechanical properties, such as low strength, tensile-compression asymmetry and texture evolution. Interestingly, it has been reported recently that compared with our traditional understanding on deformation twinning, {10-12} deformation twinning in magnesium can have the following distinct features: first of all, it is the basal-prismatic (BP) transformation at the BP interface that dominates the boundary migration instead of the gliding of twinning dislocations on twinning planes; second, the boundary that separates the reoriented crystal from its parent crystal is not a crystallographic mirror plane. Third, the boundary created by BP transformation have a terrace-like morphology in 3-dimensional (3D) space. In other words, compared with dislocations, the deformation front of the {10-12} deformation twinning in magnesium has the characteristics of quicksand. This prompt us to speculate that the strategies adopted to strengthen the materials dominated by dislocation mediated plasticity may not work effectively for magnesium based alloys and new strategies that can withstand quicksand like deformation should be used. This surmise was confirmed by performing in-situ transmission electron microscope study on several magnesium alloys. It was found that regardless of the density of the precipitates, {10-12} deformation twinning could swept across particle-shaped precipitates without obvious hardening. In contrast, in samples with nanoscale lamellar structure along basal plane, deformation twinning could be suppressed significantly accompanied by dramatically improved yield stress and flow stress. We further proposed that the mechanical performance of magnesium based alloys can be designed by tuning the twin behavior through changing the 3D morphology of the strengthening phases.

Deformation of a grain is highly influenced by neighboring grain deformation, as all grains impose boundary conditions on each other that vary spatially. To investigate this effect, detailed experimental measurements of heterogeneous deformation that include local stress and strain measurements are used to assist model development and validation, and reflexively, models assist in interpreting measured quantities. Three examples of this kind of interactive analysis will be described: in-situ tensile deformation of polycrystalline (1) pure titanium and (2) solder using High Energy X-ray Diffraction, with both near-field and far-field measurements that provide grain shape and average strain tensors, and (3) assessment of bent polycrystalline samples using Differential Aperture X-ray Microscopy, which provides 1 micron voxels of orientation with most of the strain tensor. Origins of localized heterogeneous deformation, formation of subgrains, and nucleation of mechanical twins are examined locally. Supported by NSF, DOE/BES, and the Advanced Photon Source beamlines 1 and 34.

Binary Cu-Fe alloys containing 1-2wt%Fe can be aged to form coherent (or semi-coherent) fcc Fe precipitates in a Cu matrix. In the past, such aged alloys have been used as model materials for studying the nucleation of the deformation induced fcc to bcc transformation in pure Fe. In contrast to the deformation induced martensitic transformation in bulk fcc ferrous alloys the complicating effects of autocatalytic nucleation and impingement of neighbouring martensitic plates are avoided. Analysis of these experiments has, however, been limited to post-mortem analysis of the morphology and crystallography of the product martensite phase. In this study we have sought to provide further insight into these observations by studying the role of the imposed elastic stress and dislocations via molecular dynamics/statics simulations of single fcc Fe precipitates in a Cu matrix. Loss of static and dynamic mechanical stability of the fcc Fe crystal is analyzed. The simulation results are compared to the post-mortem experimental observations and the path the crystal takes between the initial fcc and final bcc state is discussed.

Due to the scarcity of “easy slip” systems in hexagonal materials, deformation twinning plays a crucial role in determining mechanical properties and texture evolution. Molecular dynamics (MD) simulations are of a power tool in studying twinning-related deformation events. However, introducing a twin domain in atomistic simulation models remains challenge due to the complicated processes and the uncontrollability of nucleating twins in hexagonal metals. We proposed two methods for creating both {102}á10ñ growth and deformation twins in hexagonal materials for molecular dynamics simulations. Corresponding to twin propagation in the three-dimensional ‘normal’, ‘forward’ and ‘lateral’ motion of twin interfaces, we studied defect structures, energies and migration mechanisms of these twin boundaries, and the stress fields of a finite-size twin domain embedded in the parent, with respect to twin types. We found that the three twin boundaries are distinct from defect contents that account for the difference in migration mechanisms of twin boundaries.

TEM studies on shock-recovered pure polycrystalline Ta and Ta-W alloys have been conducted in order to elucidate the effects of strain rate and tungsten addition on dislocation arrangement and shock-induced twinning and alpha (bcc) --> omega (hexagonal) shear transformation. Pure Ta and Ta-W alloys shocked at 30 GPa (strain rate: ~1x 10⁴ sec^-1) responded differently from pure Ta and Ta-W alloys shocked at 45 GPa (strain rate: ~1 x 10⁵ s^-1) in which both twinning and omega phase were observed to form in association with an evenly-distributed dislocation structure with a stored dislocation density > 5 x 10¹² cm^-2. While neither twin nor omega phase was found to form in pure Ta and Ta-5wt.%W shocked at 30 GPa which contained mainly a cell-wall structure with a stored dislocation density < 5 x 10¹¹ cm^-2. The omega phase was found to form in Ta-10wt.%W shocked at 30 GPa which contained mainly an evenly-distributed dislocation structure with a stored dislocation density > 2 x 10¹² cm^-2. The results clearly reveal that dislocation generation rates, rather than dislocation velocities, are the rate-limiting step under shock staining conditions; shock-induced twinning and/or shear transformation become a predominant deformation mechanism when the arrangement of cell-wall structure resulting from dynamic recovery processes is largely suppressed. It is accordingly suggested that there exists a density limit for the evenly-distributed dislocation structure above which further dislocation generation and glide become very limited. Twinning and the alpha --> omega transition take place as an alternative mechanism to accommodate high-strain-rate straining when the stress required for dislocation multiplication exceeds the threshold stresses required for twinning and alpha --> omega shear transformation. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

The stress-induced phase transformation response of Shape Memory Alloys (SMAs) is strongly dependent on grain-scale microstructure. Specifically, the local transformation strains are dependent on the crystal orientation of the grains and the interaction between grain neighborhoods. The role of the microstructural constraint on SMA mechanics has been extensively studied empirically as well as numerically. However, when the size of the SMA specimens themselves or features in them (e.g. holes, cracks, cavities) is reduced to length scales comparable to the grain size, a structural constraint, consisting of stress concentration near features and relaxation near free surfaces, plays a role in determining local response. The relative role played by the microstructural constraint vs. structural constraint in the vicinity of micron-scale features in SMA specimens has not been systematically studied.

We empirically demonstrate a homogeneous transformation response in planar, monolithic specimens of polycrystalline, superelastic NiTi SMA. As artificial structural features of decreasing size, specifically micron-sized through-holes of varying diameter, are introduced in the specimens, heterogeneities in the response appear around the structural features. When the size of the features approaches the average grain size in the material, microstructural features e.g. grain orientations and neighborhoods play a noticeable role in determining local strains around the features. Using a combination of empirical and numerical techniques - digital image correlation (DIC), electron backscatter diffraction microscopy (EBSD) and microstructural modeling for phase transformation, we estimate the role played by (i) stress concentration around the structural features, (ii) grain orientations, and (iii) interaction between grains, in determining the local strains in planar specimens with varying microstructures and with holes of varying sizes.

The mechanistic understanding of the structural vs. microstructural constraint on the local superelastic response in specimens with small structural features is key in choosing a reliable modeling framework for predicting local phase transformation response in miniature SMA components (e.g. micron-sized sensors and actuators) and in porous SMAs with pore sizes comparable to the grain size.

Steel is one of the most ubiquitously used engineered materials owing to our ability to manipulate its properties by composition and microstructure modifications. We have investigated deformation in 201 austenitic steel, with high and low Nickel, respectively. It is well known that the stability of austenite decreases with decreasing Nickel content. This makes 201 austenitic steel (201 SS) the perfect candidate to understand how changes in microstructure depend on deformation mechanisms such as slip and phase transformation.
Our studies show that only dislocation slip occurs during deformation of 201 SS (high Ni), and grain orientation significantly influences deformation behavior. In addition to characterization of deformed microstructure using electron backscater diffraction (EBSD), we use dislocation dynamics simulations to better understand orientation efects. We further highlight differences in lattice distortion in the neighborhood of grain and twin boundaries and correlate it with dislocation accumulation at the boundaries.
In comparison, deformation of low Ni 201 SS involves significant amount of phase transformation in addition to slip. The influence of strain and strain path on phase transformation is quantified. The contribution of slip to deformation is measured in terms of misorientation developed in grains. We show that strain hardening exponent, n, shows a complex variation with plastic strain due to a contribution from both slip and phase transformation. An empirical relationship between n and some microstructural parameters is proposed.

The stress state and strain paths which metallic components undergo in real-world applications are complex and often include strain path changes. It is thus important to investigate the response of alloys to multiaxial deformation routes.
Deformation-induced martensitic phase transformations has been mostly studied under uniaxial deformation conditions. Few studies address the influence of different stress states. There exists a disagreement in whether transformation is enhanced or retarded when applying multiaxial stress states . The influence of strain path changes on phase transformations has not been addressed.
Using a multiaxial deformation rig [1] installed at the neutron strain scanning beamline POLDI at the Swiss Neutron Spallation Source, we investigate in situ the influence of proportional and non-proportional loading on the microstructural evolution of 304 stainless steel.
The in situ diffraction measurements provide an insight to the microstructural and phase fraction evolution, the distribution of internal stresses in various phases and residual stresses during uniaxial, equi-biaxial deformation as well as changing strain paths. Post-mortem microscopy studies, on the morphological changes induced by the phase transformation, (e.g. the shear-band formation) are correlated to the results obtained from in-situ techniques and discussed with respect to the nature of the deformation and the strain path changes.
[1] S. Van Petegem, J. Wagner, T. Panzner, M.V. Upadhyay, T.T.T. Trang, H. Van Swygenhoven, Acta Materialia, 105, 2016, p. 404-416.

The interaction between deformation twins and grain boundaries plays an important role in the deformation behavior of hexagonal close packed crystal materials such as Mg, Ti, and Zr. In this work, we employ automated electron backscatter diffraction (EBSD) and full-field 3D spatially resolved, Fast-Fourier-Transform (FFT) mechanics modelling to study the influence of twin shears and other material and boundary characteristics (crystallography, active slip modes, etc.) on the stress distributions at twin/grain boundary intersections. Results from our large data set statistical analysis for both Mg and Zr suggest that whether or not twins transmit across grain boundaries depends not only on grain boundary crystallography but also strongly on the anisotropy of crystallographic slip. The modeling investigation reveals that enhancing material plastic anisotropy increases the driving forces for twin transmission and the cut-off angle above which twin transmission is not likely.

Given that thermal activation and climb of dislocation contribute significantly to high temperature plastic behavior, a thorough understanding of dislocation motion is extremely critical. A new discrete dislocation model is presented to determine equilibrium configurations of dislocations as a function of temperature and strain rate. The equilibrium states are achieved by minimizing the total strain energy of the dislocations. The thermal activation mechanisms and diffusion-assisted climb are incorporated at appropriate time scales. This approach allows decoupling of time scales associated with dislocation glide and it gives access to modeling both rate and temperature effects of interest in practical applications. Thermal activation of obstacles is modeled using an energy barrier and an average activation time scale that is applicable to a large variety of obstacle models. A new universal model for diffusion assisted climb is presented that is applicable and accurate for arbitrary choice of time scales. It involves solution of an auxiliary mechanical-diffusion boundary value problem by treating dislocations as continuous source/sink of vacancies. It enables computing the rate of climb of dislocations at arbitrary temporal scales. Several applications are shown for our new model from tension/bending of micron sized samples, to creep and fatigue at high temperature. Numerical results show that the new model can capture rate effects over a wide range of temperatures and strain rates with no appreciable change in computational cost or accuracy. Our results also show that our model is capable of capturing positive rate sensitivity typically shown in precipitate strengthened materials and several attributes of high temperature creep.

Nanostructured metallic multilayers have attracted substantial attention as they often possess high mechanical strength. Here we report on the microstructure and mechanical strength of sputtered Ag/Fe multilayers with individual layer thicknesses (h) varying from 1 to 200 nm. Phase transformation of Fe from body-centered-cubic (bcc) to face-centered-cubic (fcc) structure occurred when h < 5 nm. Nanotwins formed in fine Ag/Fe multilayers. Although modulus mismatch is similar between Ag/Fe and Ag/Ni multilayers, the peak hardnesses of Ag/Fe multilayers is much lower than that of Ag/Ni system. Comparison of mechanical strength of several Ag based multilayers reveals that this drastic difference may arise from chemical stress due to the difference in stacking fault energy of the layer constituents [Acta Materialia, 114, pp 154 (2016)]. This study is supported by NSF-DMR-Metallic Materials and Nanostructures Program under grant no. 1304101.

The formation process of self-accommodation microstructure (SAM) of the Ti-Ni-Pd shape memory alloy was investigated by in-situ optical microscope examination to reveal the origin of the incompatibility in the microstructure. A SAM that is the martensitic microstructure of shape memory alloy is formed by HPVs connecting each other. Our previous study on the SAM of Ti-25Ni-25Pd (at%) revealed that the morphology of the SAM and the place of incompatibility; the SAM consisted of HPVs containing the compatible {111} type I twin as a lattice invariant deformation (LID), whereas all the junction planes between HPVs were incompatible. However, the origin of the place of the incompatibility in the SAM is still not clear. In this study, the formation process of SAM microstructure of Ti-Ni-Pd shape memory alloy was investigated to reveal the origin of the place of the incompatibility. Ti-25Ni-25Pd and Ti-39Ni-11Pd were fabricated by Ar arc melting method. LID twin is vanished in the SAM of Ti-39Ni-11Pd. The ingot was homogenized at 1373K for 21.6ks. For in-situ optical microscopy observation, specimen surface of Ti-25Ni-25Pd and Ti-39Ni-11Pd were flattened by a twin-jet polishing with an electrolyte of 80% CH₃OH + 20% H₂SO₄ and etching with an 25% HF + 25% H₂SO₄ + 25% HNO₃ + 25% H₂O respectively. The crystallographic orientation of the specimen surface was identified by electron backscattering diffraction (EBSD) analysis to identify the HPV by the single trace analysis. The reverse martensitic transformation process of Ti-25Ni-25Pd and martensitic transformation process of Ti-39Ni-11Pd were investigated by observing the formation process of the surface relief caused by the reverse martensitic transformation and martensitic transformation respectively on the optical microscope equipped with high-speed digital video camera. The formation process of the reverse martensitic transformation is considered as a rewind of the formation process of martensitic transformation as reported in Ti-Ni and beta-titanium. In-situ observation in Ti-39Ni-11Pd showed that martensitic transformation started at the junction plane between HPVs having exact {111} type I twin orientation relationship. The incompatibility exists at the junction plane formed by the collision of grown HPVs. In-situ observation in Ti-25Ni-25Pd showed that martensitic transformation started at the LID twin plane in each HPV. All the junction planes between HPVs are formed by the collision of grown HPVs and incompatible. These results show that the martensitic transformation started at the compatible {111} type I twin interface. Incompatibility exists at the interface formed by the collision of HPVs. Therefore, the place of the incompatibility in the SAM is determined by the formation process of the junction plane.