The mechanical response of alloys consisting of multiple phases is governed by microscopic strain and stress partitioning among the various phases, crystals and subgrains. Yet, due to the limitations that are inherent in the experimental characterization of the stress-strain partitioning that takes place at the micro-scale, microstructure optimization of such alloys is typically based on the averaged macro-scale response (e.g. engineering stress-strain curve). To strengthen the connection between microstructure and mechanical properties, a novel methodology is introduced in this work, that enables the joint experimental and simulation based analysis of the deformation-induced evolution of heterogeneous materials with multiphase microstructures. This is achieved through a combined experimental-numerical approach, i.e. relying on in-situ deformation experiments and crystal plasticity (CP) simulations both initiated from the same electron backscatter diffraction (EBSD) mapped representative microstructural area of interest. For the experiments, deformation-induced microstructure evolution is tracked by EBSD to increasing strain levels. To map local strains free of surface roughening effects, a recently developed, digital image correlation-based, high-resolution, 2D strain mapping methodology is employed, in which 3D effects are considered by a post-mortem serial sectioning procedure. For the simulations, the model is created from the EBSD based crystal structure, phase, and orientation maps. The individual phase properties are obtained by additional inverse CP simulations of nanoindentation experiments carried out on the original microstructure. To map local strains, as well as stresses, 2D full-field crystal plasticity simulations are run employing a recently developed spectral solver suitable for high-phase contrast materials. The methodology is demonstrated here on the example of martensite-ferrite dual phase (DP) steel, for which promising correlation between the experiments and simulations is achieved, despite the complex micro-mechanics of this material. Obtained strain maps reveal significant strain heterogeneity arising from martensite dispersion, ferrite grain size, and defect densities effects; and early damage nucleation at notch-like irregularities in martensitic zones that cause high stress triaxiality. Deviations between experiments and simulations can be explained in terms of known limitations of the involved techniques. The presented integrated engineering approach provides a high dimensional set of micro-mechanical output information that can enhance the understanding and further development of complex bulk multiphase alloys.

We present a continuum model of dislocation dynamics that predicts cell structure formation and strain hardening in FCC crystals from the properties of single dislocations. Founded on the powerful concepts of statistical mechanics, the model features a set of kinetic equations of the curl type that govern the space and time evolution of the dislocation density in the crystal, one partial density per slip system. The kinetic equations are coupled to crystal mechanics, stress equilibrium and deformation kinematics, using the eigenstrain approach. A custom FEM method has been used to solve the overall problem. Our investigation show that dislocations self-organize in various patterns due to the line tension and long-range interactions. However, the cell structure was found to form when cross slip is implemented in the model. The slip patterns, dislocation structures and crystal distortions are analyzed in our simulations and compared to corresponding experimental data. The internal elastic lattice rotation and stress fields are compared to X-ray measurements. This work was supported by the U.S. DOE Office of Basic Energy Sciences, Division of Materials Science & Engineering via contract # DE-FG02-08ER46494 at Florida State University and by funding from the School of Nuclear Engineering at Purdue University. The authors thank Ladislas Kubin, Hael Mughrabi, Benoit Devincre and Ben Larson for very fruitful discussions related to this work.

A computationally efficient modeling method called “quasi-coarse-grained dynamics” (QCGD) is developed to expand the capabilities of molecular dynamics (MD) simulations to model behavior of metallic materials at the mesoscales. This mesoscale method is based on solving the equations of motion for a chosen set of representative atoms from an atomistic microstructure and retaining the energetics of these atoms as would be predicted in MD simulations. The energetics of the representative atoms is defined based on scaling relationships for the atomic scale interatomic potentials in MD simulations. The energetics and the degrees of freedom of these representative atoms are scaled to account for the missing atoms in the microstructure so as to define the total energetics of the system. In addition, the scaling relationships render improved time-steps for the QCGD simulations. The success of the mesoscale method is demonstrated by the prediction of the high temperature thermodynamics, deformation behavior of interfaces, phase-transformation behavior, heat generation during plastic deformation as well as the wave-propagation behavior in metallic systems under various conditions, as would be predicted using MD simulations. The reduced number of atoms and the improved time-steps allow the modeling of metallic materials at the mesoscale in extreme environments. The applicability of the QCGD simulations to predict the evolution of defect structures and the microstructure during deformation and failure in FCC and HCP materials systems at the mesoscales will be discussed.

It is well appreciated that the mechanical behavior of a nanoscale object must be affected by its surface. Large stresses, typically compressive, are required for compensating the surface stress. These stresses, which are well documented, superimpose to the stress generated by external load. Indeed, computer simulation testifies to significant asymmetry in compressive versus tensile strength, pointing towards capillary forces favoring compression and impeding tension at small size. Since the relative amount of surface scales inversely with the object size, the associated effects are expected to be most important at the nanoscale. Yet, surface effects also have readily measurable implications for mechanical behavior in the macroscopic world. Many decades ago, the work of Rehbinder advertized strong surface effects on crystal plasticity even for millimeter-size objects, and the zero creep experiments by Udin explored the role of surface tension in compensating external load. Even though modern nanomaterials have orders of magnitude more surface per volume, the surface effects on their behavior are not orders of magnitude larger. It is also puzzling that discussions of capillarity effects in nanomaterials focus on the role of surface stress whereas the 2oth century studies offer good arguments for surface tension being the relevant capillary parameter. Experiments investigating high surface area nanomaterials in during in situ mechanical testing in electrolyte afford a separate variation of surface tension in surface stress that may help to clarify the distinction. The talk will address the role of surface tension, surface stress, and surface excess elasticity on nanoscale mechanical behavior, using experiments on nanoporous gold with different surface states as a testbed.

Magnesium alloys are the lightest known structural metals but their wider commercial use is hindered by their limited room-temperature ductility. Interestingly, alloying with yttrium and rare-earth elements significantly increase their room temperature ductility. The underlying mechanisms for this ductilization effect have not been identified. We have intensively studied this phenomenon by a multi-disciplinary combination of quantum-mechanical calculations, advanced synthesis techniques, state-of-the-art electron microscopy characterization and mechanical testing. We propose a multi-scale mechanism increasing the ductility (Acta Materialia 59 (2011) 429) that is based on the stacking fault energy changes induced by solutes (Acta Materialia 60 (2012) 3011). More specifically, we suggest that a reduction of the I₁ stacking fault energy causes a higher density of I₁ stacking faults which then act as nucleation sources for dislocations. The proposed scale-bridging connection is verified by studying a dense set of Mg alloys theoretically as well as experimentally (Acta Materialia 70 (2014) 92). Having this wealth of data, we mathematically construct a “key parameter” that quantifies for each solute its ability to ductilize Mg alloys. This integral figure of merit is based on selected materials parameters of individual solutes and can be used for future theory-guided materials design of new ductile Mg alloys.

Mechanically-driven chemical mixing of dual-phase alloys during extensive plastic straining requires that the constituent phases co-deform, and is therefore a sensitive function of the constituent phases&’ mechanical properties, particularly differences in their strengths. Accounting for such a phase strength mismatch in a mechanically-driven alloy is crucial when modeling the evolution in microstructure using driven alloy theory. To demonstrate this, we incorporate phase strength effects on strain partitioning into kinetic Monte Carlo simulations of a mechanically-driven, binary alloy. Such simulations provide quantitative insight into the combination of processing and material parameters that dictate the steady state chemical mixity during processing. We use these simulations to generate dynamical phase diagrams that predict temperatures and compositions at which a couple with a given phase strength mismatch should chemically homogenize during mechanical alloying. Several of these dynamical phase diagrams are validated using mechanical alloying experiments in which tungsten-transition metal couples with various phase strength mismatches are mechanically alloyed in a high energy ball mill and their chemical mixing tracked as a function of milling time. The implications of our computational and experimental results with regard to alloy design for powder processing will be discussed.

Nanoindentation and micropillar compression tests have documented large (>1 GPa) macroscopic strengths in numerous types of nanolayered composites, as individual layer thickness approaches ~10-30 nm. The large plastic strength is attributed to the abundant interfaces and their role in moderating dislocation nucleation and propagation in crystalline systems. Despite the wealth of nanoindentation data, little is known about the strength of the individual layers (phases) within such composites and the magnitude of mismatch in flow strength between phases.
This work presents a new method to determine the individual layer flow strengths in nanolayered composites, by coupling finite element simulations of nanoindentation with experimental nanohardness and micropillar compression data. For Cu/Ni nanolayered composites with [001] interfaces, the method reveals that 20 nm thick Ni layers have ~3 times the flow strength of neighboring 20 nm Cu layers. Remarkably, the Cu and Ni flow strengths in this 20-nm system are comparable to those for pure nanocrystalline (d = 20 nm) Cu and Ni, respectively. The new method is validated using more elaborate in-situ diffraction studies of deforming Cu/Ni nanolayered composites.
Overall, combined nanohardness and micropillar compression data can furnish the constituent flow strengths within nanolayered composites—not just the bulk composite strength. The underlying method is also more versatile than in-situ diffraction-based methods that are limited to crystalline phases. Moreover, the work reveals complications in converting between nanohardness and micropillar compression data—a factor that is often overlooked in the literature.

In this paper, we present the experimental characterization of a dual phase steel with respect to its mechanical and microstructural properties. The work is embedded in the framework of a research training group (GRK 1483 funded by the Deutsche Forschungsgemeinschaft), which aims to simulate a dual phase steel from the beginning of the production process to the final product using multi scale modelling. The steel consists of a soft ferrite matrix and a hard martensitic phase. SEM in combination with EBSD and FIB are used to monitor the microstructure of the material and to prepare micro pillars within selected sample regions. The pillars of different size, orientation and with different phases are then compressed with a flat punch diamond tip in the nanoindenter. The compression experiments of the ferrite matrix show a rather weak size dependence as well as no strong influence of the pillar orientation. The results obtained on pillars containing both phases demonstrate the enhancement of the strength due to martensite.
To elucidate the role of both phases on the evolution of dislocation density and microstructure during deformation, a selected area is repeatedly investigated by EBSD at different values of plastic strain. Using the combination of small scale compression tests and iterative tensile experiments it is attempted to describe the influence of microstructural components on the macroscopic material behavior. The research training group also uses the results obtained in this work as input and validation of the simulations, describing the structure and strength evolution of dual phase steels during manufacturing.

Magnesium (Mg) becomes one of the most promising light-weight elements to replace existing structural materials due to the low density and high specific strength. However, for practical applications its strong anisotropic deformation behaviors as affected by the crystal orientation and loading condition need to be understood. In conventional structural metals with hexagonal closed-packed (HCP), including Mg and its alloys, twinning plays a critical role in deformation process due to their limited number of slip systems. Although the well-known easy deformation modes such as basal slip and tension twinning were investigated extensively by far, each deformation mode was studied for a particular orientation favoring one specific deformation mode in the light of the Schmid factor. Although competition and interplay between slip and twinning are critically important to control a given deformation process, they have not been thoroughly studied yet.
In the present study, we performed compression test of nano-sized Mg pillars in situ inside transmission electron microscope (TEM). In order to understand how twinning and dislocation slip compete and interplay each other, the loading axis which favors both tension twinning and prismatic slip was chosen, which corresponds to the [11-20] direction of Mg. Mg pillars for in situ TEM experiment were fabricated by focused ion beam (FIB). Before compression test, in situ heating was performed inside TEM to remove the defects (i.e. dislocation loops and debris) induced by Ga⁺ ion beam during fabrication. In situ TEM compression tests were performed by using a field emission TEM (JEOL JEM-2100F) operated at 200 kV and a single-tilt nanoindentation holder (Nanofactory^TM).
At the early stage of deformation, tension twinning initiated at the top end of nanopillar due to the high stress concentration at the contact between indenter tip and pillar. Upon nucleation, the twin propagated continuously downward along the pillar, it changed the crystal orientation of the pillar; the loading axis was changed from [11-20] to [2-1-1-3] by tension twinning. While the twin propagates, lots of dislocation activities were observed within the twin domain. After the twin propagates to yield the (engineering) strain around 1.9 %, massive basal slip was followed in the twin domain because of its high Schmid factor in the twinned orientation. Once the basal slip is activated, the twin propagation was delayed considerably and most of the rest deformation (up to 7 % strain) was accommodated by the basal slip, leading to a catastrophic failure of the pillar. This deformation behavior will be discussed in more detail with supporting molecular dynamic simulations.

Dislocation velocities have been measured as a function of local stress and temperature, in pure Fe and various FeX binary alloys, during in situ straining experiments in a transmission electron microscope. The results show that the less mobile dislocations are always of screw character, and that their velocity is controlled by a combination of Peierls stresses and local pinning at two kinds of obstacles: i) the oxide layer at the two foil surfaces and ii) super-jogs formed by double cross-slip at the vicinity of solute atoms.
Experiments in pure Fe show that the local stress necessary to strain a micro-sample at a given strain-rate is higher in the very thin parts than in the thicker parts. Indeed, the plasticity is controlled by Peierls forces in the thick parts (as in the bulk material), and by the bowing of dislocation sources between the two surfaces in the thinner parts. This induces two markedly different kinetics of dislocation motion, respectively steady and continuous in the first case, irregular with long waiting times followed by fast motion in the second case.
Experiments in binary alloys with several percent of solute elements (where the sample thickness is large enough to make anchoring at oxide layers negligible) show the same different kinetics of moving dislocations, but as a function of temperature.
At low temperatures where Peierls forces are dominant, screw dislocations move steadily and homogeneously. Since the distance between super-jogs is inversely proportional to the velocity of screws, a weak hardening is obtained when this distance decreases, i.e. when the solute concentration increases.
At room temperature where the bowing of screw dislocations between super-jogs becomes dominant, dislocations move by series of rapid avalanches separated by waiting times. Since the distance between super-jogs is inversely proportional to the local deformation stress, a much stronger hardening is obtained when this distance decreases, i.e. when the solute concentration increases.
These two different behaviors account fairly well for the solute hardening effects measured in bulk samples deformed in conventional mechanical tests.

Understanding the evolution of strain during plastic deformation is of great importance for correlating defect structure with material properties. While many advances have been made in measuring the global plastic strain more accurately using high precision transducers, there exists a need for more accurate determination of the local elastic strain-field during deformation. In the present work we show for the first time that strain mapping can be carried out during in-situ deformation in a TEM with the precision of a few nanometers without stopping the experiment or obscuring the image. Our method of local strain mapping consists of recording nanodiffraction patterns at each pixel position during deformation. Then, a strain map can be calculated by measuring changes in the lattice constants from the diffraction patterns. The main practical limitation for this method is the speed of the detector. To overcome this limitation we used a Gatan K2-IS direct detection camera operating at a frame rate of 400 f/s and recorded nanodiffraction maps to calculate a time dependent local strain-map from this data. Using a Hysitron PI 95 picoindenter to pull the sample in tension, strain fields around individual dislocations can be observed. In this talk we will discuss the in-situ strain mapping method and compare the local and transient strains occurring around moving dislocations with the global strains measured from the quantitative in-situ deformation experiment.

When the mean free path of dislocations is reduced below the micron scale, either by geometrical constraints such as in whiskers, wires and pillars, or by grain boundaries as in small-grain polycrystals, a very large increase of mechanical resistance is observed.
Because of their complex and out of equilibrium structures, metallic nanocrystals may deform through dislocations nucleated and absorbed at grain boundaries (GBs)[1], but many other alternate plasticity mechanisms have been foreseen.
Shear-migration coupling is one of them and is the focus of many theoretical and experimental studies. At variance from dislocation-based plasticity, the shear produced by a moving GB can result in different values, depending on a parameter called the coupling factor Beta. Recent results obtained by in-situ Transmission Electron Microscopy (TEM) in ultra fine-grained Aluminum, show that many deformation modes are activated, including shear migration coupling. The coupling factor can be measured experimentally using image correlation analysis and therefore confronted to what has been predicted by models such as the one from Cahn and Mishin [2]. Although solid statistical data are still missing, beta appears smaller than what has been predicted. A reason could lie in the atomic-scale mechanisms that guide the migration of GBs. The Cahn and Mishin model assumes collective motion of GB dislocations, while Rae and coworkers insist on the role of steps propagation [3]. High resolution imaging of bicrystals shows that steps decorate GBs and that the motion of imperfect steps could result in the migration of the GB associated with a shear. To take in account these observations we also proposed a geometrical model for the shear migration coupling of grain boundaries [4], based of the shuffling of atoms within extended cells around the GB. Finally, recent simulations show that step dislocations (disconnections) are probably the basic mechanism leading to grain boundary migration [5]. Those disconnections are found in non-ideal GBs and can be created from interactions between lattice dislocations and GBs [6].
[1] F. Mompiou, D. Caillard, M. Legros, H. Mughrabi, Acta Materialia 60/8 (2012) 3402.
[2] J.W. Cahn, Y. Mishin, A. Suzuki, Acta Materialia 54/19 (2006) 4953.
[3] C.M.F. Rae, D.A. Smith, Philosophical Magazine 41/4 (1980) 477.
[4] F. Mompiou, D. Caillard, M. Legros, Acta Materialia 57/7 (2009) 2198.
[5] A.Rajabzadeh, F. Mompiou, M. Legros, N. Combe, Phys. Rev. Lett., (2013) 110,265507
[6] A.Rajabzadeh, F. Mompiou, N. Combe, M. Legros, D. M. molodov, S. Lartigue-Korinek, Acta Materialia 2014

Although nanoscale twinning is an effective means to enhance yield strength and tensile ductility in metals, nanotwinned metals generally fail well below their theoretical strength limit due to heterogeneous dislocation nucleation from boundaries or surface imperfections. Here we show that Au nanowires containing angstrom-scaled twins (0.7 nm in thickness) exhibit tensile strengths up to 3.12 GPa, near the ideal limit, with a remarkable ductile-to-brittle transition with decreasing twin size. This is opposite to the behavior of metallic nanowires with lower-density twins reported thus far. Ultrahigh-density twins (twin thickness<2.8 nm) are shown to give rise to homogeneous dislocation nucleation and plastic shear localization, contrasting with the heterogeneous slip mechanism observed in single crystalline or low-density-twinned nanowires. The twin size dependent dislocation nucleation and deformation represent a new type of size effect distinct from the sample size effects described previously.

Nanotwinned metals have rare combinations of mechanical strength and ductility. Previous studies have shown that detwinning occurs in plastically deformed nanotwinned metals. Although molecular dynamics simulations have predicted that fine nanotwins can migrate at low stress, there is little in situ evidence to validate such predictions. Also it is unclear if detwinning occurs prior to or succeeding plastic yielding. Here, by using in situ nanoindentation in a transmission electron microscope, we show that a non-elastic detwinning process in nanotwinned Cu occurred at ultra-low indentation stress (0.1 GPa), well before the stress necessary for plastic yielding [Y. Liu et al, APPLIED PHYSICS LETTERS 104, 231910 (2014)]. Furthermore the in situ nanoindentation technique allows us to differentiate dislocation-nucleation dominated microscopic yielding preceding macroscopic yielding manifested by dislocation-transmission through twin boundaries. This study thus provides further insights for understanding plasticity in nanotwinned metals at microscopic levels. This research is funded by DOE-OBES.

Nanoindentation measurements were performed to investigate the impact of a CVD grown graphene layer on top of a Cu substrate on the elastic and plastic behaviors.
Analysis of the elastic deformation region in the load-displacement curve reveals that the typical non-linear load-displacement relationship is much more linear with the presence of graphene on the surface which indicates an effect similar to a stiff plate on a more compliant substrate subjected to a point load [1].
In the plastic regime of the load-displacement curves, the response is made of pop-ins which correspond to the nucleation and glide of dislocation loops in the Cu crystal [2]. The pop-in excursion length is directly proportional to the number of dislocations escaping the free surface to accommodate the change of the geometry imposed by the tip into the sample. The load at which dislocation nucleation starts has been shown to be linearly proportional to the first excursion [3]. However, in the presence of graphene, we show that the pop-in length is almost constant with respect to the load at which the plasticity starts. The graphene layer thus significantly constraints the dislocations which are blocked at the Cu/graphene interface.
In order to understand the deformation mechanisms taking place during nanoindentation, simulations are performed for both the elastic and the elasto-plastic deformation regimes. Linear elastic finite elements simulations are carried out to understand the effect of the graphene layer on the indentation response. The load-displacement curve for the first 5 nm indentation depth follows a modified Hertz law from which the graphene elastic stiffness can be estimated. Furthermore, the elastic stress fields in both Cu and Cu/graphene are studied. Discrete Dislocation Dynamics (DDD) simulations are performed to understand the dislocations nucleation and propagation into Cu during nanoindentation, and analyze the effect of the monolayer graphene at the interface.
Further experimental characterizations with AFM and on TEM cross-sections are performed to unravel the physical mechanisms associated to the constraint induced by the graphene surface layer.
References
[1] P. M. Ramsey, H. W. Chandler and T. F. Page, "Modelling the contact response of coated systems," Surface and Coatings Technology, vol. 49, no. 1-3, p. 504, 1991.
[2] S. Suresh, T.-G. Nieh, and B.W. Choi, "Nano-Indentation Of Copper Thin Films On Silicon Substrates," Scripta Materialia, vol. 41, no. 9, p. 951, 1999.
[3] Y. Shibutani, T. Tsuru, and A. Koyama, "Nanoplastic Deformation Of Nanoindentation: Crystallographic Dependence Of Displacement Bursts," Acta Materialia, vol. 55, no. 5, p. 1813, 2007.

Micro-compression experiments are performed during in-situ Laue diffraction at the MicroXAS beamline of the Swiss Light Source to explore the sequence of activated slip systems in bcc single crystals. Diffraction patterns are obtained in transmission with a 5-23 keV X-ray beam with FWHM of 0.5-1 mu;m. Laue scans allow the mapping of the spatial distribution of strain gradients in the deformed pillars, providing information on local crystallographic orientations and on the activated dislocation slip systems. Additional examination by scanning electron microscopy allows identification of slip traces on the surface.
We have investigated dislocation slip in W single crystals with different crystallographic orientations and present the results reported in two publications (PRL 2014 and Scientific Reports 3, 2547 (2013)). When two {110} planes containing the same slip direction experience the same resolved shear stress, sharp slip traces are observed on a {112} plane. When however the {110} planes are slightly differently stressed, macroscopic strain is measured on the individual planes and collective cross-slip is used to fulfill mechanical boundary conditions, resulting in a zig-zag or broad slip trace on the sample surface. For some of the investigated crystallographic orientations slip on the anomalous plane is observed. Coarse crystallographic slip traces occur first on the primary and the conjugate planes. At larger strains slip traces are also observed on the anomalous (0-11) plane and strain production is alternated with the primary and secondary. Anomalous slip has been reported in literature to occur in W only at small strains.
In a second part of the talk, the influence of alloying and critical temperature on slip activity will be discussed using experimental results obtained from similar experiments carried out on W6%Re, Mo and Nb.

Understanding and predicting deformation in ductile materials constitute scientific grand challenges ultimately requiring coarse graining approaches based on the statistical dynamics of dislocation densities. To probe fundamental aspects of deformation associated with mesoscopic length scales ranging from tenths of microns to several tens of microns, we have combined submicron resolution (0.5 µm) 3D x-ray microscopy (3DXM) measurements of local lattice rotations with finite element simulations for spherical-tip (100-µm radius; 100 mN) micro-indentations in Cu. The inherent localization of deformation under indentations provides a “mesoscale deformation laboratory” in which detailed lattice curvature and GND investigations can be performed over the entire deformed volume - experimentally with 3DXM and computationally by continuum deformation simulations. Ranging from strong to weak distortions as a function of depth and distance from a single indent, the spatial distribution of the deformation field provides a host of test conditions and, hence, a detailed assessment of the underlying mechanisms of deformation. Moreover, the statistical fluctuations inherent in plastic deformation are contained within the measurement volume ensuring a representative volume sampling. We present 3D x-ray microscopy measurements of the spatially confined deformation below 100 mN, 100 micron radius spherical indents in (100), (110), and (111) oriented Cu. The measurements have been analyzed in terms of local lattice rotations and geometrically necessary dislocation densities. Direct, spatially-resolved comparisons of the measurements with crystal plasticity simulations show good overall agreement in the magnitude of lattice distortions as a function of crystal orientation. However, both qualitative and quantitative differences are found to exist between measured and simulated local rotational distortion components and geometrically necessary dislocation density patterns. The 3DXM measurements further show dislocation patterning in the form of micron range volumes with distinct orientation variations as a function of depth below the indented surface. *Research supported by the U. S. Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division and by the Center for Defect Physics Energy Frontier Research Center; microbeam research performed at the Advanced Photon Source is supported by the U. S. DOE Division of User Facilities

Extensive research has been carried out on cyclic stress-strain response and final dislocation arrangements in cyclically deformed metals up to the occurrence of saturation. There is, however, a limited knowledge about dislocation interactions during the early stages of cyclic deformation, which is essential to validate novel cyclic-computational models that are currently being developed. Laue x-ray micro-diffraction has proven to be an effective technique to study dislocation ensembles and their evolution during in-situ mechanical testing [1].
We present a novel in-situ micro-mechanical testing system that allows deforming samples cyclically in shear mode. The system is installed at the MicroXAS beamline at the Swiss Light Source (SLS). The samples are designed based on the Miyauchi&’s geometry. At various stages of cyclic deformation, spatial resolved Laue diffraction patterns are recorded in transmission mode. In particular, we study the evolution of dislocation arrangements during the first cycles on annealed polycrystalline copper samples subjected to reversed-shear-fatigue. The samples are deformed at different strain amplitudes up to different number of cycles (maximum of 75). The evolving dislocation microstructures are analyzed in terms of misorientation and spatial distribution. The experimental results aim to validate a low cycle fatigue modeling by crystal plasticity finite element method [2].
[1] H. Van Swygenhoven, S. Van Petegem, JOM 62, 36-43 (2010).
[2] N. Grilli, K. G. F. Janssens, H. Van Swygenhoven, in progress

Complete experimental recovery of the spatially resolved distribution of Geometrically Necessary Dislocation (GND) densities over multiple length scales has been demonstrated for the special case of a FCC or BCC crystal that undergoes plane strain deformation. High Resolution Electron Backscatter Diffraction (HR-EBSD) reduces the noise floor of the GND measurements by up to two orders of magnitude as compared to traditional EBSD measurements, which allows the GND measurements to resolve dislocation cell structures. In this talk we discuss experiments in which FCC nickel and BCC tantalum are deformed plastically with a wedge indenter. The orientations of the crystals and the loading configuration are such that plastic deformation occurs cooperatively on sets of slip systems that induce a plane deformation state, which can be thought of as being due to three effective in-plane slip systems. Furthermore the three effective in-plane slip systems have the same orientations for both the FCC and BCC crystals. We indented both crystals with a wedge indenter with a 90 degree included angle to a depth of about 200 micrometers. In this talk we identify scaling relationships in the GND densities and compare the GND density distributions on individual slip systems in the two crystal classes as a function of position in the indentation deformation fields.

This presentation will present a coupled experimental - computational model study of the stochastic behavior in the mechanical response of polycrystalline materials consisting of few grains to hundreds of grains in samples in the 1-100 micron length scale. We study the transition from stochastic (at small scale) to deterministic (at large scale) deformation response in polycrystalline samples using both simulation and nanoindentation experiments. First, the variation in experimental measurements of hardness is described in relationship to the overall dislocation density of a sample. Then, a stochastic crystal plasticity model, combining a Monte Carlo method with a polycrystal continuum dislocation dynamics model in a self-consistent viscoplasticity framework, will be described. This tool is used to predict the flow stress and the variation in strength of polycrystals using randomized sampling of structural variations. The numerical results are compared to nanoindentation experimental data from three samples of ultra-fine grain structures in Cu and Ti manufactured by severe plastic deformation. The results suggest that stochastic plasticity at small scales due to scarcity of dislocations, coupled with microstructural features such as grain size distribution and crystallite orientations govern uncertainty in the mechanical response of the polycrystalline materials. The extent of the uncertainty is correlated to the “effective cell size” in the sampling procedure of the simulations and experiments. The simulations and experimental results demonstrate similar quantitative behavior in terms of coefficient of variation within the same effective cell size.

The detailed analysis of plastic flow at small scale, simulations at the atomistic and discrete dislocation dynamics lengths scale can be used to “mimique” experimental situations and to propose new experiments dedicated to contribute to solve open issues e.g. in the deformation behaviour of bcc metals. In this contribution, a discrete dislocation dynamics tool adapted for simulating tungsten, based on atomistic input is briefly presented [1].
The tool has been successfully applied to explain the occurrence of anomalous slip [2] observed in tungsten (see also talk von H. Swygenhoven). The mechanisms deduced from the discrete dislocation dynamics simulations is based on local dislocation interactions, which lead to the formation of cross-kinks. The cross-kinks trigger glide of a mixed dislocation on the anomalous slip after reaching the surface. The resulting sharp slip trace is consistent with the experimental observations [2].
Furthermore the simulations suggests that two screw dislocations which are oriented such that their mutual interaction is repulsive glide together faster due to non-Schmid effects, included in the simulation [1]. This finding is surprising and experimental verification would be helpful.
Another aspect concerns with the still open question of the active slip systems in bcc metals. From atomistic studies in tungsten only slip on {110} glide planes are expected. A pure geometrical analysis [3] shows that junctions will then only form along <111> and <100> directions. Assuming also {112} and {123} slip, junction direction deviating by more than 15 degrees from these directions would be expected, not reported in literature and worthwhile being looked at. Due to the lower symmetry of the latter glide planes, those junctions should even be more likely to be formed [3].
[1] K. Srivastava, R. Gröger, D. Weygand, P. Gumbsch, Dislocation motion in tungsten: Atomistic input to Discrete Dislocation simulations, IJP 47 (2013) 126.
[2] C. Marichal, K. Srivastava, D. Weygand, S. Van Petegem, D. Grolimund, P. Gumbsch, H. Van Swygenhoven, Origin of anomalous slip in tungsten, (accepted) PRL 2014.
[3] K. Srivastava, D. Weygand, P. Gumbsch, Dislocation junctions as indicators of elementary slip planes in body centred cubic metals, (accepted) J Mater Sci 2014.

Classical work by Johnston and Gilman in 1959 (J. of Appl. Phys. 1959, 30, 129) showed that dislocation velocities are strongly dependent on applied stress. Numerous experiments have subsequently shown this to be valid for individual as well as for groups of dislocations in macroscopic fcc and bcc crystals. The majority of reported dislocation velocity measurements used imaging techniques combined with either mechanical data or time resolved topological data. Developments in small scale mechanical testing allow to directly correlate the intermittency of collective dislocation motion with the mechanical response. In such tests, discrete forward surges in displacement can be related to dislocation avalanches, which are triggered by the evolving dislocation sub-structure. We study the spatiotemporal characteristics of intermittent plastic flow in quasi-statically sheared single crystalline Au micro crystals with diameters between 300 nm and 10 mm, whose displacement bursts were recorded at several kHz (Scripta Mater. 2013, 69, 586). Both the crystallographic slip magnitude, as well as the temporal extent of the slip events are exhibiting power-law scaling as predicted by theory and dislocation dynamics (DD) modeling. The obtained slip velocity distribution has a cubic decay at high values, and a saturated flat shoulder at lower velocities. No correlation between the slip velocity and the applied stress is found. Based on these results, we discuss potential influences the coupled machine-sample dynamics may have, demonstrate that contact between the machine and anvil is maintained during slip, and present DD-simulations that are supportive of our findings. The combination of our experimental results and simulations suggest that the dynamics of the internal stress fields dominate the evolving dislocation structure leading to velocities that are insensitive to the applied stress.

Recently, there has been a lot of interest in the statistics of strain bursts in micro-pillar compression/tensile testing for studying the scaling effects. The strain bursts manifest as displacement bursts (pop-in) in load controlled nanomechanical testing systems and could provide vital information about slip sizes and their stress dependence which can be used to understand the scaling behavior at small scales. Unfortunately, the highly dynamic nature of the displacement bursts presents a significant challenge in their experimental determination as the instruments&’ dynamic response can easily obscure the materials&’ response. This work presents a comprehensive dynamic model that incorporates the response of a nanomechanical actuator and the associated load frame in addition to the test specimen. The model also incorporates the time constants of the measurement signals typically encountered in commercially available nanomechanical systems. The modeling results clearly show a significant contribution of the instruments&’ dynamics to the measured response during a strain burst event. The effect of the time constants of the load and displacement signals on the observed material response will be discussed. A critical examination of the quality and validity of recent experimental data in light of the findings of the current model will be presented.

α-Fe single crystal nanopillars with different diameters were fabricated through focused ion beam (FIB) and then in situ tested using uniaxial microcompression methodology inside a SEM. Besides the well-established tenet of “smaller is stronger”, we demonstrated that the strain rate sensitivity of small-scaled single-crystal α-Fe decreased about 12 times as the pillar diameter decreased from 1000 nm to 200 nm and the size strengthen exponent is a function of strain rate. Combined with molecular dynamics simulation as well as the postmortem TEM observations, we propose that the size dependent strain rate sensitivity observed in BCC structured metals so far is mainly stemmed from the ultrahigh strength accompanying the decreasing sample size or grain size.

Since its introduction by Uchic et al. [1] nearly a decade ago, micro-compression testing has become a preferred technique for investigating size-dependent plasticity. During this period, micro-mechanical techniques have been adapted for a variety of test geometries (cantilever bending, tensile etc.). Recently the test methodology has extended to variable strain rates [2], long duration creep tests [3], elevated temperatures up to ~500-750 °C [2, 4, 5], and to testing in-situ in the electron microscope [2].
By utilizing an intrinsically displacement-controlled micro-compression setup, which applies displacement using a piezo-actuator, we&’ve recently extended the attainable range of strain rates to up to~ 10² s^minus;1. By heating the sample and tip we have reached tip temperatures up to 500°C and furthermore, by circulating cryogenic coolant through the system frame, sub-ambient temperatures in the range of minus;100 °C can be attained.
Using these new capabilities, we examine the plasticity of electrodeposited nanocrystalline Nickel and different monocrystalline semiconductor materials such as Si, InSb and GaN. In order to analyse the fundamental deformation mechanisms variable strain rate and variable temperature micro-compression experiments were performed. Activation parameters such as activation energy and activation volume were determined and discussed in view of the most probable deformation mechanism.
References
[1] M.D. Uchic, D.M. Dimiduk, J.N. Florando, W.D. Nix, Science, 305 (2004) 986-989.
[2] J.M. Wheeler, J. Michler, Review of Scientific Instruments, 84 (2013) 064303.
[3] V. Maier, B. Merle, M. Göken, K. Durst, Journal of Materials Research, 28 (2013) 1177-1188.
[4] J. Milhans, D.S. Li, M. Khaleel, X. Sun, M.S. Al-Haik, A. Harris, H. Garmestani, Journal of Power Sources, 196 (2011) 5599-5603.
[5] S. Korte, R.J. Stearn, J.M. Wheeler, W.J. Clegg, Journal of Materials Research, 27 (2011) 167-176.
[6] M.A. Meyers, A. Mishra, D.J. Benson, Progress in Materials Science, 51 (2006) 427-556.

The Portevin-Le Chatelier effect is well documented in the iron-carbon system - manifesting itself as unstable plastic flow with a characteristic “serrated” nature in conventional tensile tests. It is known to be a function of both the applied strain rate and testing temperature and can be suppressed by prior deformation. It is of importance for the processing of many sheet metal products due to the uneven surface finish it generates, although precise details of the mechanisms involved, particularly at the micro-scale, are not fully understood.
Recent advances in nanoindentation have resulted in the ability to perform tests at temperatures up to 1000K in vacuum. In this work nanoindentations have been performed on high purity iron (3appm Carbon and Nitrogen) with varying temperature, strain rate and prior deformation, to allow a comparison of the PLC effect at the sub-micro scale with bulk mechanical tests.
Nanoindentations were performed with a loading rate of 1mN/s at temperatures from 298K to 598K in high purity iron. Serrated flow is seen to occur at temperatures above 398K but is supressed above 540K. This is in direct agreement with macro-scale tensile experiments. At a constant temperature of 403K nanoindentations were performed at indentation strain rates of 0.01s^-1 ,0.05s^-1, 0.1 s^-1, 0.5 s^-1, and 1 s^-1. The serrated flow is observed to be supressed at strain rates above 0.1 s^-1 which is again in line with bulk experimental data.
To study the effect of pre-straining at 403K a large, 3000nm indent was made into a large grain to provide a region of plastic deformation. Smaller 1000nm Indents were immediately made into the plastic zone around the large indent, as well as far field from the region of deformation (asymp;50µm from the 3000nm indent centre). While serrated flow was seen in the far field indents, the indents within the plastic zone did not show serrated flow. The sample was then aged at 403K for 1200 minutes. 1000nm were then again made within the plastic zone of the 3000nm indent and asymp;50µm from the 3000nm indent centre. While serrated flow was still observed in the far field indent no serrated flow was seen in the indent made in the pre-deformed region.
The sample was then aged at 573K for 1200minutes, before returning to 403K. After this ageing serrated flow was observed in indents close to and far from the initial plastic deformation showing that the serrated flow is controlled by locations of the carbon and nitrogen atoms at dislocation cores. The initial plastic deformation removes the dislocations from their interstitial atmospheres. After aging at 403K there was not sufficient diffusion for serrated flow to resume. After the second aging at 573K sufficient interstitial atoms had diffused to the dislocation cores for serrated flow to resume.
This work has shown that the Portevin-Le Chatelier can be studied on the sub-micron scale by using high temperature nanoindentation, showing direct agreement with macro-scale tensile tests.

Capturing the effects of strain rate at practical loading rates is difficult to access using Discrete Dislocation Dynamics (DDD) due to the very short times steps required to track the trajectory of the moving dislocation. This issue is substantially alleviated by the Quasi-static Discrete Dislocation (QSDD) model presented in this work. The essential idea is to avoid tracking the dislocation path, instead using energy minimization to determine the equilibrium configuration under given boundary conditions. The fundamental assumption that dislocation glide occurs on time scales that are much shorter than any other short range events, such as nucleation and escaping from obstacles. Therefore, the time scale associated with dislocation glide is eliminated. Short range effects are accommodated at a given applied load, by repeatedly minimizing the total strain energy until all the dislocations are at rest. Thermal activation from obstacles is performed after the system reaches equilibrium. We regard the escape frequency as the average frequency escaping from a series of obstacles along the dislocation line. The waiting time of dislocation in our 2D model is calculated analytically in terms of the free energy barrier at current load, temperature and loading rate. When the residence time (accumulated time at an obstacle) of a dislocation is greater than its waiting time, we allow a thermally activated release from that obstacle. We show that this strategy is statistically equivalent to a Kinetic Monte Carlo simulation of a 3D model. Given the generality of our model, the free energy barrier can be chosen to accommodate a variety of strengthening mechanisms. Finally we include dislocation climb within our model by solving the coupled vacancy diffusion problem along with the mechanical boundary value problem. The climb rate for dislocation at rest is determined by net flow of the concentration around the dislocation core and the dislocation is allowed to climb, until the driving force for dislocation force for glide exceeds that for climb. Once this occurs, the new equilibrium positions of the dislocations are determined. We present several examples of simulations in uniaxial tension showing that (a) the new model is equivalent to DDD in the athermal case, (b) strain-rate sensitivity and temperature sensitivity of plastic deformation can be captured over a wide range of temperatures and strain rates and lastly, (c) the effects diffusion controlled deformation is demonstrated. We also show that the computational effort is not significantly increased by strain rate and is completely controlled by the dislocation density.

An experimentally-validated polycrystalline plasticity model for deformation in body centered cubic (BCC) metals is under development at Sandia National Laboratories for the exploration of microstructural effects on stochastic material reliability. Dislocation plasticity in body centered cubic (BCC) metals is more complex than face centered cubic metals due to many confounding factors including anomalous slip, inconsistent reports of slip plane selection, strong sensitivity to impurities, quantum effects, and non-Schmid effects resulting in tension-compression asymmetry. Moreover, dislocation activity in BCC metals is known to be more strongly temperature and strain-rate dependent. As originally described by Seeger (Z. Metallkd., 1981), dislocation kink-pair theory provides a mechanistic basis for understanding this thermally-activated motion of screw dislocations. The thermal component of the flow stress is described in two regimes: a regime at low temperatures and high stresses where the line tension of a partially-formed kink dominates, and a regime at high temperatures and low stresses where the elastic interaction between the two sides of the kink pair dominates. This model can be directly calibrated against existing single crystal deformation experiments for molybdenum, tantalum, tungsten, and niobium. The model has been incorporated directly into a finite element polycrystal plasticity formulation, and permits the analysis of temperature- and rate-dependent deformation in any single crystal or polycrystalline ensemble. The polycrystalline model is consistent with published data for polycrystals of the same four BCC metals. Most importantly, this mechanistic model faithfully reproduces the temperature- and strain-rate dependence of the Los Alamos Mechanical Threshold Stress (MTS) continuum-scale constitutive law, successfully bridging from atomistic theory to continuum behavior. By contrast, the model deviates from Johnson-Cook and Zerilli-Armstrong phenomenological continuum constitutive descriptions of temperature- and rate-dependence. We will briefly describe ongoing efforts to experimentally validate this model using grain-scale digitial image correlation strain field mapping and electron backscatter diffraction during in-situ deformation.
Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.

In body-centered cubic (bcc) metals, plasticity arises from the motion of 1/2<111> screw dislocations. These dislocations are subjected to a lattice resistance that can be described through the two-dimensional energy landscape of the dislocation in the {111} plane, the so-called 2D Peierls potential. Here, we employ ab initio calculations based on Density Functional Theory (DFT) to determine the 2D Peierls potentials in bcc transition metals: V, Nb, Ta, Mo, W, Fe [1]. We use these potentials to deduce several properties of dislocation glide and in particular, the kink-pair formation enthalpy, as well as the dependence of the Peierls stress on crystal orientation, well-known in bcc crystals for its deviation from Schmid law.
Dislocations move between Peierls valleys by the nucleation and propagation of kink-pairs. In this work, we calculated the kink-pair formation energy in bcc transition metals using a line tension model parameterized on the 2D ab initio Peierls potentials [2], allowing to predict kink properties from ab initio calculations performed in cells containing only a few hundred atoms. The formation enthalpies thus obtained show large deviations from isotropic elasticity, and are compared to experimental data.
We then used the 2D ab initio Peierls potentials calculated under no applied stress to predict the behavior of dislocations under stress. We thus calculated the dependence of the Peierls stress on the orientation of the applied stress with respect to the crystal orientation, allowing us to analyze deviations from Schmid law in various bcc metals. These results are compared to direct ab initio calculations of Schmid law deviation, in order to test the validity of using zero-stress Peierls potentials. The calculations show large variations between metals that are discussed with respect to available experimental data.
[1] L. Dezerald, L. Ventelon, E. Clouet, C. Denoual, D. Rodney and F. Willaime, Phys. Rev. B 89, 024104 (2014).
[2] L. Proville, L. Ventelon and D. Rodney, Phys. Rev. B 87, 144106 (2013).

Nowadays there is an increasing interest in developing new materials that could work as sensor and actuators in MEMS, allowing a new generation of smart MEMS.
Shape memory alloys (SMA) are considered as smart materials because of their properties of shape memory and superelasticity associated to a reversible thermoelastic martensitic transformation, and are firm candidates to be incorporated into MEMS. New properties as ultra-high mechanical damping at nano-scale and a very fast response along thousand of cycles have been reported in Cu-Al-Ni SMA [1, 2]. However there is a lack of a quantitative characterization of the observed size effects in SMA.
The goal of this work I to offer such quantitative characterization and analysis by presenting the evolution of the superelastic behaviour of a series of pillars covering a broad range of size diameters from few hundred of nanometers to some microns. The pillars have been milled by FIB in three different laboratories in order to verify the reproducibility of the results independently of the milling procedure. Then they have been tested by nano-compression tests by using an instrumented nano-indenter. The obtained results show a clear size effect on the critical stress for the stress-induced transformation during superelastic behaviour.
Finally, this size effect has been also studied by in-situ nano-compression tests inside a FEG scanning electron microscope, by using a pico-indenter, allowing visualizing the conditions of martensite nucleation and growing at different scales.
[1] J. San Juan et al., Nature Nanotechnology 4, 415 (2009).
[2] J. San Juan et al., Applied Physics Letters 104, 011901 (2014).

Plastic localization in small scale systems is a major issue for a variety of modern technologies based on freestanding metallic films or nanowires that rely on sufficient ductility during operation. Plastic localization in the form of diffuse necking or localized shear bands constitute a class of well understood phenomena from the viewpoint of continuum “macroscopic” mechanics. All this existing knowledge can be used to address plastic localization problems in micro or nano-objects but with the additional complexity resulting from the existence of different size effects and from the interplay of a variety of competing or cooperating deformation mechanisms.
Recently, a standard 1D imperfection based localization analysis has been revisited [1] to address the problem of necking in nanocrystalline thin films, taking into account the grain size dependent high strength, low strain hardening and high rate sensitivity typical of these systems combined with possible strain gradient effects and creep relaxation. The model shows that the ductility defined as the strain at the onset of the instability can either drop to small values for very small grain sizes and/or film thickness due to the high strength and to the presence of imperfections, or can remain constant or even increase owing to an increased rate sensitivity resulting from thermally activated mechanisms. This last stabilization effect can be reinforced by gradient plasticity effects if allowed by the dominant deformation mechanism.
The predictions of the model are used to analyze recent experimental data obtained on a variety of thin freestanding metallic film systems involving AlSi, Cu, Pd, Ni as well as amorphous ZrNi with thickness ranging between 50 and 900 nm. These films have been deformed in uniaxial tension up to failure using an on chip nanomechanical testing method [2-4]. All these systems show, in some cases, uniform elongation larger than 10% combined with a very high strength. The good ductility is shown to be a consequence of the combination of the moderate to high rate sensitivity and statistical distribution of imperfections, and, potentially, from strain gradient plasticity effects.
References
Pardoen, T., Size and rate dependent necking in thin metallic films, Journal of the Mechanics and Physics of Solids 62 (2014) 81-98
Gravier, S., Coulombier, M., Safi, A., Andre, N., Boe, A., Raskin, J.P., Pardoen, T., 2009. New MEMS - Based micromechanical testing laboratory - Application to aluminium, polysilicon and silicon nitride. Journal of Microelectromechanical Systems 18, 555-565
Coulombier, M., Boé, A., Brugger, C., Raskin, J.P., Pardoen, T., 2010. Imperfection-sensitive ductility of aluminium thin films. Scripta Mater. 62, 742-745
Idrissi, H., Wang, B., Colla, M.S., Raskin, J.P., Schryvers, D., Pardoen, T., 2011. Ultrahigh strain hardening in thin palladium films with nanoscale twins. Advanced Materials 23, 2119-2122

Linear defects in crystalline materials, known as dislocations, are central to the understanding of plastic deformation, mechanical strength, and damage tolerance. In addition to being the fundamental vehicle for plasticity, dislocations: mediate crystal growth kinetics, degrade optoelectronic response in epitaxially layered systems such as heterojunction solar cells and light emitting diodes, augment phase change behavior in electronic memory devices, limit electron mobility enhancements in strained field effect transistors, and provide a relaxation mechanism for materials subjected to intense radiation fluxes. Despite nearly a century of research on dislocation structure, interactions, and response, the mechanisms, energetics, and kinetics of dislocation nucleation in otherwise pristine crystals remain poorly understood. Nanostructures with large fractions of surface atoms are particularly prone to nucleation at free surfaces. We report on experiments that directly measure the temperature-dependent (90 to 500K) surface dislocation nucleation barriers in defect-free fcc metallic nanowhiskers subjected to uniform tension. Our measurements show that, whereas nucleation strengths are weakly size- and strain-rate-dependent, a strong temperature dependence is uncovered, corroborating atomistic simulations predicting that nucleation is strongly thermally activated. We measure activation volumes as small as singular atomic volumes, which explain both the ultrahigh athermal strength (~8 GPa) as well as the temperature dependent scatter, evident in our experiments and well captured by a thermal activation model. Our experiments highlight the pronounced probabilistic nature of surface dislocation nucleation, which is crucial input to device design using nanoscale building blocks.

We report direct observations on the incipient plasticity of dislocation-free single crystal Au [110] nanowires by in situ transmission electron microscopy compression tests. The diameter of the tested nanowires ranged from ~ 80 nm to 350 nm and the length-to-diameter ratio was larger than 5. The top end of all [110]-oriented Au nanowires is bound by two inclined {111} faces in a wedge shape, on the other hand the side faces consist of four large {111} and two small {100} planes, resulting in a truncated rhombic cross-section. In our deformation setup where the wedge-shaped growth end of nanowire was compressed with a flat diamond punch, the strain becomes localized to the region under the contact. Under such a strong strain gradient condition, the initial compressive deformation began with the emission of small prismatic loops from the top corner, which have a radius ranging from ~20 nm to 100 nm. The Burgers vector of these loops was determined to be 1/2[-1-10], which generates the vertical downward displacement of the inner area encompassed by the prismatic loops. Right after the nucleation, these prismatic loops glided immediately down to reach a certain position where it remained stationary until newly generated loops force to glide downward in jerky manner.
After a certain number of closed loops being punched out (typically less than ten), there was a clear transition in the nucleation mechanism of the loops; open loops started to bulge out and then released from the contact area. Very different from the closed prismatic loops, the freely moving ends of the open loops intersected and swept across the top faces before being released, thereby relaxing the strain accumulated beyond the contact area. The stress field accumulated inside the nanowire is also released by escape of open loops through the free surface. More importantly, these loops can act as sources for ordinary dislocations which slip along the inclined {111} planes. Based on the direct observation of prismatic loops and supporting molecular dynamics simulations, detailed characteristics of the loops and their behaviors at the initial stage of deformation of nanowire will be presented.

The oxide scale that forms on metal surfaces at ambient conditions may not influence their bulk mechanical properties, but when the surface-to-volume ratio increases, as in nano-wires, the oxidizing environment can significantly alter the mechanical properties of materials that have a high affinity to oxygen such as Al. Here, reactive molecular dynamics (MD) simulations with ReaxFF were used to simulate tensile deformation of single crystal Al nanowires (NWs) in an oxygen environment. In oxygen, Al NWs with 3.2-5.6 nm diameters formed ~1 nm thick, oxygen deficient amorphous oxide scale with a low density. The oxide shell had a low Young&’s modulus of 25.5 GPa, and thus reduced the overall stiffness of the Al NW. In vacuum, pure Al NW yielded by nucleation of partial dislocations from the surface, and did not show substantial ductility. On the other hand, when Al had an oxide shell, the dislocation nucleation processes occurred at the Al/oxide interface, and by increasing the number of nucleation sites, and the activation volume for dislocation nucleation, oxide shell lowered the yield stress of Al. In oxygen, the oxide shell showed an interesting superplastic deformation behaviour due to viscous flow, in which oxygen diffusion repaired the broken Al-O bonds, below a critical strain rate. The interplay between the strain rate and oxidation rate was shown to be essential for designing nano-devices in ambient environments and also for improving the efficiency in large-scale forming processes of Al sheets, during which the formation of nano oxide fibers on the Al surface enhanced the Al adhesion.

Physical properties of nano-objects have attracted considerable attention this last decade, from both experimental and theoretical points of view. Thanks to the development of the nano-objects deformation tests, striking modifications of their mechanical properties compared to their bulk counterparts have been revealed. Besides their technological interest, these objects, with sizes comparable to those conceivable with classical atomistic simulations, are model materials for the study of the elementary mechanisms of plasticity, in particular those related to the size-scale effects usually observed when going towards smaller dimensions.
Notably, the dislocation nucleation mechanisms have most probably a key role in some features of these systems, such as the increase of the elasticity limit compared to bulk materials. Different scenarios have been proposed to explain such feature, but most studies suggest that when the dimensions go down below the micrometer, the usual dislocation bulk sources cannot operate like in bulk materials. They are too few and/or hindered. Therefore new sources, such as surface ones, must come into play, and the yield stress must then be very high for activating these new dislocation sources.
Another open question is the implication of those processes in the so-called brittle to ductile transition at low scale for semiconductor materials. Indeed, while most semiconductors are brittle in their bulk form at room temperature, they become ductile below few hundreds of nanometers.
In that context, we have examined the deformation response of silicon pristine nano-objects, using both classical and ab initio atomic scale simulations. We chose silicon as a model for semiconductor materials, and because there exists a large variety of well-tested classical interatomic potentials available to describe its mechanical properties. Different geometries and deformation conditions have been studied, for which different plasticity modes have been activated. Among them, an interesting feature shown by the simulations is the unexpected activation of {011} planes when nanowires or thin films are under compression along [001]. Such a mechanism has been identified as an indirect consequence of the low dimensions of the systems.
We have also considered the case of core-shell silicon nanowires with an amorphous shell around a crystalline core, representative of most realistic nanowires that are often naturally covered by a thin amorphous oxide shell as soon as they are exposed to the air, or that present an amorphous layer of silicon when they are milled by focused ion beam (FIB) techniques. In that case, native defects at the amorphous-crystalline interface behave as “seeds” producing dislocation cores. During deformation, it is one of these dislocation cores that initiates plasticity, when the local stress is high enough. We have shown that this behavior has a strong effect on the variation of the elasticity limit.

Knowing the mechanical properties of nanoparticles is critical to design the devices composed of nanoparticles for high durability and reliability, while mechanical characterization of individual nanoparticles is very challenging. The only feasible way of characterizing the mechanical properties of individual nanoparticles may be nanoparticle compression/indentation for which, however, there is no quantitative data analysis available in the literature. The previous researchers speculated that the contact mechanics for nanoparticle indentation is complex, so they derived their analysis models based on oversimplified assumptions, e.g., rigid substrate, flat indenter (even for spherical or sharp indenters), and small indentation depth compared to the nanoparticle diameter. These oversimplifications make the previous models to be inaccurate or with limitations. In this study, based on large-deformation contact mechanics, a first-ever analytical model with very simple closed-form expression is derived for elastic indentation of an individual nanosphere on an elastic substrate using a spherical indenter, and the model perfectly matches finite element analysis for indentation depth even over 60% of the sphere diameter. A methodology based on this model is proposed to be able to determine the modulus, hardness, yield strength, and fracture strength of the nanospheres. This methodology is applied to analyze the experimental nanoindentation results of SiO₂, TiO₂, and Au nanoparticles of various sizes in the range of 100~600nm. Thus, the size dependence of mechanical properties of nanoparticles is systematically studied. We believe that this first-ever quantitative methodology can not only be used to accurately characterize the mechanical properties of individual nanoparticles but also impact the understanding of size dependence of mechanical properties at small scales.

Nano-colloidal crystals (NCCs) have emerging applications in photonics and optoelectronics, but their poor mechanical properties due to the weak interparticle interaction largely hinder these applications. This implies that, increasing the intrinsically-weak interparticle interaction (IPI) is essential for reinforcement. Compared to the sintering method, Atomic Layer Deposition (ALD) is a much more efficient approach to reinforce the NCCs due to the introduced bonding by depositing atomically-controlled layers of reinforcing materials around the nanoparticles (NPs) within NCCs. However, there is a clear knowledge gap in full understanding on the mechanical reinforcement mechanism of ALD in colloidal assemblies. In our study, NCCs composed of monodisperse SiO₂ NPs was characterized using nanoindentation. The results show that ALD-treated NCC is drastically stiffened up to 30 times and hardened up to 200 times. With increasing the ALD-layer-to-NP-radius ratio to ~7%, the deformation mechanism of NCC transits from granular, bonded granular, to particle-reinforced composite behavior, and crack and creep are suppressed. Thus, ALD enable precise tailoring to make NCCs be hardened, stiffened, and toughened. A model is successfully proposed to interpret the mechanical behavior of ALD mechanical behavior. In addition, unstable indentation loading in the form of “pop-in” is observed especially for NCCs without ALD treatment or treated with thin ALD layers. The pop-in events are highly statistical, and the maximum pop-in magnitude is fairly proportional to the indentation displacement. According to our best knowledge, this is the first comprehensive study on the mechanical behavior of as-assembled and ALD treated NCCs. The ALD-modified NCC provides a model system to study the effect of nanoscale modifications on the mechanical behavior and failure mechanism of nonporous structure materials.

The size and strain rate effects on the deformation behaviors of submicron vanadium were studied through in situ compression test inside a transmission electron microscope (TEM). The single crystal vanadium pillars with diameters ranging from 100 nm to 600 nm were fabricated through focused ion beam and tested under displacement control with the strain rates ranging from 0.001 s^-1 to 0.05 s^-1. It was found that for pillars with larger diameter or deformed under lower strain rate, screw dislocations are prone to dominate the mechanical behavior of vanadium and localized slip bands along {110} plane are frequently observed. In contrast, for pillars with smaller diameter or deformed under higher loading rate, mixed dislocation mediated plasticity are dominant. Correspondingly, wavy-slip morphology instead of slip bands is observed in the deformed pillars. In addition, compared with almost all the reported submicron-sized single crystal metal pillars which are characterized with intermittent and discontinuous plasticity, submicron-sized single crystal vanadium pillars exhibit surprisingly smooth and continuous plastic flow.

Tremendous research efforts have revealed an extrinsic sample
size-effect with a strength-size power-law relation of the type $d^{-n}$. At
the same time, small scale testing has strengthened the viewpoint that
plasticity proceeds via scale free intermittent slip activity. Both of these
observations hold true for a large variety of crystals and microstructures,
prompting considerations based on a generic dislocation network. Here we
show how a general expression for the scaling exponent $n$ is found,
assuming that only the statistics of the strain burst magnitude changes with
decreasing sample size. In doing so, the simple expression
$n=(\tau+1)/(\alpha+1)$ is found, where $\alpha$ reflects the leading order
algebraic exponent of the low stress regime of the critical stress
distribution and $\tau$ is the scaling exponent for intermittent plastic
strain activity. This quite general derivation, based on extreme value
statistics and the universal properties of a dislocation network supports
the experimental observation that the size effect paradigm is applicable to
a wide range of materials, differing in crystal structure, internal
microstructure and external sample geometry.

The fatigue resistance of metal films can be improved by decreasing the film thickness into the sub-micron regime, presumably due to the inhibition of dislocation motion and to the accompanying increase in strength. It remains unclear, however, what happens to the fatigue resistance of very thin films (less than 100 nm), where diffusive processes may become important and pose a threat to the film reliability. In this study, we investigate the ultra-high cycle fatigue behavior of supported Cu films with thicknesses between 40 and 360 nm in order to search for a transition from dislocation plasticity to diffusive deformation control. We use a novel AFM-based resonance method to investigate the damage created under strain controlled fatigue loading as a function of applied strain, film thickness and cycle numbers up to 5x10¹⁰. For films thicker than 100 nm, extrusions and boundary cracks limit the fatigue performance but only appear above a threshold in the applied strain amplitude which scales inversely with the square root of the film thickness. This damage formation is attributed to dislocation activation which is controlled by the film thickness and grain size. In films of 100 nm and thinner, the grain boundary cracks are replaced by grain boundary grooves which are believed to form by diffusion mediated creep processes, similar to observations at higher temperatures but here driven by cyclic stresses and capillarity. In contrast to the thicker films, no threshold strain is observed for the formation of damage formed by diffusion mediated processes. These results indicate that films thinner than approximately 100 nm are less resistant to fatigue damage formation than thicker films due to diffusive processes and suffer from an increased reliability threat, particularly at high cycle numbers.

The fracture behavior of ductile nanostructured materials is complicated by the effect of interfaces and size on both crack propagation and dislocation activity. In an attempt to gain basic insights into the mechanisms of crack propagation in nanostructured materials we use in-situ fracture testing in a transmission electron microscope with the ultimate goal of developing tactics for designing tough and recyclable composites. We use multilayer films as a model system and propagate cracks parallel to the interfaces. The initial tests have focused on multilayer systems of polycrystalline titanium and amorphous zirconium oxide layers deposited by pulsed laser deposition. This material system is especially suitable because the interfaces are sharp and smooth and the layer thicknesses can be varied, in this case between 10 and 100 nm. In addition, the mechanical properties of the multilayer films have been determined using microcompression and nanoindentation. In the case of the 100 nm thick layers, the cracks propagate through the middle of the Ti layers, despite the fact that bulk Ti is expected to be tougher than ZrO₂. The crack front is accompanied by dislocation activity in the nearby Ti layers, from which the size of the plastic zone can be estimated, and leads to rough crack surfaces. In the case of the 10 nm thick layers, the cracks propagate cleanly along the interfaces between the Ti and ZrO₂ and are accompanied by less plasticity. We interpret this change in fracture behavior in terms of the interface strength relative to the thickness dependent flow stress of the Ti layers. Based on these observations, we suggest various tactics for designing tough and recyclable nanoscale composites.

The mechanical behaviour of 170 nm-thick freestanding nanocrystalline copper films has been studied with an on-chip suite of MEMS-based mechanical testing structures.
The working principle of the technique relies on the use of internal stress within a long actuator beam to deform a second beam (specimen) made of the same or of another material owing to the etching of an underlying sacrificial layer. The actuator beam has a tapered shape so as to control the strain rate of the specimen beam during the mechanical loading. The backside etching of the substrate allows the observation of the deformed films by Transmission Electron Microscopy (TEM) without any additional post-processing steps.
Directly after the etching of the sacrificial layer, the copper films exhibit a moderate ductility. The nominal fracture strain extracted from different films varies from 2 to 5.5%. All the broken films exhibit a typical ductile fracture process with significant diffuse necking. The true local fracture strain attained in the necking region is very large as a result of a stable necking process and slow damage accumulation. The significant variations in the strain at necking indicate an effect of imperfections and/or local microstructural features.
The creep/relaxation behaviour of the unbroken films is monitored during several days and weeks after release. The relaxation kinetics and amplitude depend on the plastic strain imposed during the initial loading and on the corresponding stress level. The specimens keep deforming very significantly and fail at a nominal strain equal to ~3 times the average fracture strain reached during the first loading. More surprisingly, in some cases, multiple necks are observed. The diffuse necking is sometimes accompanied by a more localized shear type process as well as, finally, by intergranular fracture mechanisms. The details of the failure mechanism are different when comparing specimens broken during the etching of the sacrificial layer or during the relaxation.
The combined effect of the thickness/free surfaces area, imperfections and microstructural features as texture, grain size, twin spacing, dislocation density and grain boundary distributions on the selection of the fundamental mechanisms of deformation driving the plastic/ductile and creep/relaxation behaviours as well as fracture process will be discussed in the light of TEM characterization and basic plastic localization analysis.

Fatigue failure remains a prominent problem in engineering and current management of this issue relies primarily based upon understanding growth of observable cracks until failure. This strategy is too conservative, as it ignores fatigue crack initiation and relies instead on non-destructive testing to reveal observable cracks which can be difficult and costly to manage. A significant step forward will be lifeing based upon the entire life of a part and this will be underpinned by developments and applications of new high fidelity approaches to confirm physical understanding the nucleation of fatigue cracks.
In this talk, we will discuss the role of slip and geometrically necessary dislocation evolution through the life of a cyclically deformed nickel superalloy sample in three point bending. This sample has been cycled and periodically observed to capture snapshots of the deformation using high (angular) resolution electron backscattered diffraction and high (spatial) resolution digital image correlation. HR-EBSD involves direct comparison of high quality electron backscatter diffraction patterns to reveal inhomogeneous residual stress distributions, and geometrically necessary dislocation content through evaluation of lattice curvatures. HR-DIC reveals in plane accumulated plastic strain through direct comparison of high spatial resolution micrographs with surface markers.
Our findings reveal that even in simple deformation states, slip is clearly inhomogeneous and the role of lengthscale effects is of high importance. Use of HR-EBSD and HR-DIC reveals the complementary and differing nature of the quantitative data obtained from these techniques, each of which fundamentally measure different components involved in plastic deformation (stored dislocation content, residual elastic strain and accumulated plastic slip). Our quantitative analysis of the slip, residual stress state and accumulated GND density are linked with physically based crystal plasticity finite element modelling within our groups to drive forward a physically based modelling effort to improve component life prediction.

Hydrogen embrittlement (HE) in Ni-base alloys is an important concern for sour well applications and is very sensitive to microstructure. By coupling electron backscatter diffraction and in situ tensile tests on H-charged alloy 725 in a scanning electron microscope, we determine the relationship between microstructure and hydrogen (H) assisted fracture. Our results show that HE is intergranular and that the grain boundary (GB) character plays a key role in crack initiation and propagation. By designing our experiment to yield copious secondary cracks, we identify the GB types that are most susceptible to initiating and propagating cracks, and investigate the role that plasticity plays in determining this susceptibility. Our findings offer new insights on GB structure/plasticity relations and open the path for the design of HE-resistant alloys by grain boundary engineering.
This work was supported by the BP-MIT Materials and Corrosion Center.

This study investigates the time-dependent and cyclic plastic deformation mechanisms as well as the resulting fatigue crack initiation mechanisms in ultrathin (100 nm) nanocrystalline (average grain size: 75 nm) gold thin films. The experimental approach relies on quantitative in situ TEM, MEMS-based nanomechanical testing to measure the stress relaxation and fatigue properties of small gold tensile specimens (100 nm thick, 1 micron wide, ~5 micron long) at room temperature. This approach allows observation of the microstructure evolution in the TEM while measuring the evolution of stress and strain under stress relaxation or cyclic loading. Stress relaxation experiments were employed to calculate primary (i.e., between 5 and 15 minutes) and steady-state (i.e., after several hours) creep rates for stresses ranging from ~100 to ~1300 MPa, as well as activation volumes associated with the initial stress relaxation (first few minutes). The steady-state creep rates are about two orders of magnitude lower that the primary creep rates, and the activation volumes are in the 10-20b^3 (with b, Burger&’s vector) range. In the primary regime, in situ TEM observations reveal significant dislocation-based activities, while post-mortem SEM and TEM observations (after several hours or days of testing) reveal voids along grain boundaries and triple junctions that are consistent with diffusional creep. Under cyclic loading, a ratcheting behavior is observed, which can be predicted based on the stress relaxation results. However, the fracture surface of the fatigued specimens reveals unique features such as texturing of the microstructure near the crack. These observations will be discussed in light of the proposed plastic deformation mechanisms in nanocrystalline metals.

Using molecular dynamics, we investigate crack tip dislocation emission and decohesion during intergranular fracture in Ni. For some GBs, cracks propagate in a brittle-like manner by bond breaking despite copious dislocation emission. The debonding and dislocation nucleation processes during crack propagation are analyzed on the atomic scale. We find that there are strong correlations between dislocation activity and bond breaking in both time and space. This result will be discussed in reference to the Rice-Thompson criterion, which views dislocation emission and decohesion as competing mechanisms during crack propagation. This work was funded by the BP-MIT Materials and Corrosion Center.

From published studies of A.Misra et al., N.Mara et al. etc.., multilayer Cu/Nb composites have already proved to be one of the promising materials which can be subjected to extreme environments involving high radiation damage, temperature and mechanical loading. This idea of using nanolayers where Frank-Read source does not operate in a single layer, to build a strong solid was proposed by Koehler. The mechanical strengths of these Cu/Nb nanolayer composites have found to reach as high as ~2.5 GPa with a ductility of ~30%. The ability to reach near theoretical strength and large plastic deformation has made these materials a serious contender for the above applications. The strength of these nanoscale materials are strongly derived from the interface structure compared to its bulk counterpart. The underlying strengthening mechanisms at this incoherent (large lattice parameter mismatch) interface of the face centered cubic (FCC) Cu, and body centered cubic (BCC)
Nb structure, has been investigated using experimental techniques (ex/in-situ) and simulations (continuum/atomistic). Due to the recent technological advances in fabricating nanoscale, multi-material films of (few to tens) nanometers thick are possible, which are useful to interpret the interface dominated plasticity phenomenon of these materials.
In this study, pillar compression of nanoscale Cu/Nb single crystal multilayers with individual layer thickness (20 nm) is investigated. The samples were subjected to successive compression experiments with strain ramping up-to 35% respectively. Synchrotron X-ray micro-diffraction experiments, using a monochromatic beam of 10 keV energy were also performed on the pillars after each compression strain, providing us with insights on how plasticity in Cu and Nb nanolayers evolve. We observe a significant increase of peak broadening in both Cu and Nb layers up-to strains of ~4% followed by saturation of the X-ray ring width broadening until large plastic deformation of 35%. This observation indicates that the interfaces of the Cu/Nb nanolayers are very stable and effective in trapping and annihilating dislocation content during mechanical deformation. The nanolayer composite shows a maximum flow strength of ~1.6 GPa at ~24.2% compression strain. Further, these investigations affirm that, the Cu/Nb nanolayers can be deformed to large plastic strains without any onset of plastic instabilities. Beyond the plastic flow regime of these composites, understanding the failure also has attracted some interests recently, which is an additional point of interest.

The applicability of utilizing metallic materials in extreme environment applications is dependent upon understanding the mechanisms of failure at the atomic scale. The failure of these materials is largely determined by the evolution and interaction of defect structures during a deformation process. A principle factor for the evolution of defects is the microstructure of the material. A fundamental understanding of the microstructure effects on the defect evolution at the atomic resolution and the related contribution to plasticity at the macro-scales is needed to obtain a reliable performance of metallic materials in an extreme environment.
As a result, large-scale molecular dynamics (MD) simulations are used to characterize the dynamic evolution of defect structures (dislocations, twins, stacking faults, etc) for various microstructures under extreme loading conditions. The MD simulations are carried out for single crystal and nanocrystalline metals (FCC, BCC) under loading conditions of high strain rates and shock. The defect structures generated in these MD simulations are characterized using various computational tools for different loading conditions and microstructures. The evolution of various types of dislocations, twins, faults, etc. and the related deformation and failure response is investigated. The relationships between the microstructure, defect density, dislocation type, loading conditions and the failure strength will be discussed in this poster.

Today, representative volume element (RVE) or unit cell (RUC) models are routinely employed for the estimation of mechanical properties of complex material microstructures. Numerous advances have appeared in the last 20 years for computational methods seeking to coarsen, homogenize, or statistically estimate properties in such models. Significant challenges now remain at the mesoscale where models are needed for semi-crystalline, percolated, or experimentally-determined structures and configurations. We present numerical evidence that direct application of existing numerical approaches may prove problematic.
In general, RVE discretized models must be refined so that the computed properties are converged in the sense of h-convergence. When the internal material interfaces in the simulated domain are of such random arrangements that small feature sizes exist, such as that may appear near a percolation threshold, the resulting modeling errors can make converged properties difficult to obtain. The microstructures may contain load bridges several orders of magnitude smaller than other more volumetrically-dominant morphological features in the unit cell.
In the talk, we report h-convergence rates, the microstructure models from which they are obtained, and the proposed explanation for why such surprisingly low convergence rates may be a general and possibly commonly-encountered result. The complexity of the microstructure is presently quantified by the number, length, and approximate radii of the load bridges. Through the study of a sequence of increasingly refined models, the convergence of elastic properties with respect to grid resolution of the computer models is determined. The rates are contrasted with those of simpler microstructures, namely polycrystals and unidirectional composites, for which traditional finite element error estimates apply. The numerical results show that microstructure complexity can drastically degrade the convergence of the homogenized properties and therefore may demand prohibitively large numbers of degrees of freedom to yield accurate solutions.

Because of the large surface-area-to-volume ratio that they exhibit, nanoporous metals offer exciting possibilities in various fields such as catalysis, sensing, MEMS and biomedical applications. However, when the cell size of porous metals is decreased to the nanoscale, samples become extremely brittle, in spite of the inherent ductility of metals. This represents a serious disadvantage for direct applications.
Bulk nanoporous gold (np-Au) samples were produced by dealloying of gold-silver alloys. A combination of free and electrochemical dealloying steps was used to obtain a crack-free structure without shrinking the sample. Removal of the sacrificial element (silver) and formation of the nanoporous structure were verified by scanning electron microscopy and energy dispersive x-ray spectroscopy.