Strength and Brittleness of Interfaces in Fe-Al Superalloy Nanocomposites under Multiaxial Loading: An ab initio and Atomistic Study

nanomaterials Article Strength and Brittleness of Interfaces in Fe-Al Superalloy Nanocomposites under Multiaxial Loading: An ab initio and Atomistic Study Petr Šesták 1,2 , Martin Friák 1,*, David Holec 3, Monika Všianská 4,1,5 and Mojmír Šob 5,1,4 1Institute of Physics of Materials, Academy of Sciences of the Czech Republic, Žižkova 22, CZ-616 62 Brno, Czech Republic ; [email protected].cz (P.Š.); [email protected] (M.V.); [email protected] (M.S.) 2Central European Institute of Technology, CEITEC BUT, Brno University of Technology, Purkyˇnova 123, CZ-612 00 Brno, Czech Republic 3 Department of Physical Metallurgy and Materials Testing, Montanuniversität Leoben, Franz-Josef-Strasse 18, A-8700 Leoben, Austria; [email protected] 4Central European Institute of Technology, CEITEC MU, Masaryk University, Kamenice 5, CZ-625 00 Brno, Czech Republic 5Department of Chemistry, Faculty of Science, Masaryk University, Kotláˇrská 2, CZ-611 37 Brno, Czech Republic *Correspondence: [email protected] Received: 9 September 2018; Accepted: 18 October 2018; Published: 24 October 2018   Abstract: We present an ab initio and atomistic study of the stress-strain response and elastic stability of the ordered Fe 3 Al compound with the D0 3 structure and a disordered Fe-Al solid solution with 18.75 at.% Al as well as of a nanocomposite consisting of an equal molar amount of both phases under uniaxial loading along the [001] direction. The tensile tests were performed under complex conditions including the effect of the lateral stress on the tensile strength and temperature effect. By comparing the behavior of individual phases with that of the nanocomposite we find that the disordered Fe-Al phase represents the weakest point of the studied nanocomposite in terms of tensile loading. The cleavage plane of the whole nanocomposite is identical to that identified when loading is applied solely to the disordered Fe-Al phase. It also turns out that the mechanical stability is strongly affected by softening of elastic constants C0 and/or C66 and by corresponding elastic instabilities. Interestingly, we found that uniaxial straining of the ordered Fe 3 Al with the D0 3 structure leads almost to hydrostatic loading. Furthermore, increasing lateral stress linearly increases the tensile strength. This was also confirmed by molecular dynamics simulations employing Embedded Atom Method (EAM) potential. The molecular dynamics simulations also revealed that the thermal vibrations significantly decrease the tensile strength. Keywords: Fe-Al; superalloys; order; tensile strength; elasticity; ab initio; stability; nanocomposite 1. Introduction Iron-aluminium-based materials containing either Fe 3 Al and/or FeAl intermetallic compounds represent one of the most promising classes of metallic alloys intended for high-temperature structural applications. They are known for many excellent properties, e.g., (i) resistance to oxidation [ 1 ] or with respect to various molten salts [ 2 ]; (ii) relatively low density; (iii) electrical resistivity and (iv) low cost of raw materials [ 3 – 5 ]. On the other hand, their wider use is hindered by their low ductility at ambient temperatures and a drop of the strength at elevated temperatures [ 5 ]. Regarding the former , it has been shown that it is caused by an extrinsic effect, in particular hydrogen atoms produced by the reaction of water vapor with aluminum at the surface of the specimen [ 6 , 7 ]. If it is not Nanomaterials 2018,8, 873; doi:10.3390/nano8110873 www.mdpi.com/journal/nanomaterials Nanomaterials 2018,8, 873 2 of 20 for this environmental embrittlement, Fe 3 Al is seen in experiments to have decent ductility [ 8 , 9 ]. Recently, there is a renewed interest in Fe-Al-based materials containing higher number of chemical species and/or phases [1,10–19]. A special sub-class of Fe-Al-based nanocomposites [ 20 – 33 ] are those consisting of two phases, ordered Fe 3 Al with the D0 3 structure and a disordered Fe-Al solid solution with about 18.75 at.% Al (see, e.g., Refs. [ 22 , 24 , 34 ]). These phases co-exist in the concentration range from about 19 to about 25 at.% of Al (see the original Fe-Al phase digram by Kattner and Burton [ 35 ]) reproduced, for example, in an excellent review by Sundman and co-workers [ 36 ]. Importantly, the transformation of phases in the Fe-Al system is particularly complicated and the final state of samples is very sensitive to many factors including thermo-mechanical history [ 37 ]. The co-existence of Fe 3 Al compound and a disordered Fe-Al solid solution is best experimentally confirmed by transition electron microscopy (TEM) technique which is sensitive to anti-phase boundaries (APBs) as these have a specific character in Fe 3 Al and a different one in other, at least partly ordered, Fe-Al phases. In particular, Oguma et al. [ 38 ] developed a time-dependent Ginzburg-Landau (TDGL) formulation for the ordering processes of B2 and D0 3 types in binary alloy systems. Specifically in the case of Fe-Al, numerical simulations of the kinetic equations have been performed for concurrent ordering and phase separation to disordered A2 and ordered D0 3 and helped to explain TEM observations. This combined theoretical and experimental analysis identified round/oval droplets of the disordered Fe-Al phase formed on the expense of diminishing amount of ordered Fe 3 Al phase. The rounded shape of these droplets indicate that the interface energy is not sensitive to crystallographic orientation and, therefore, the (001) interfaces studied in this paper are equally probable as others (when mechanical properties can admittedly be orientation-sensitive). Papers related to first-principles calculations of these two-phase coherent nanocomposites are rather rare as most of the previous studies focused on individual phases appearing in the Fe-Al phase diagram (see a few selected examples listed below). In order to fill in this gap, we study in this work thermodynamic, structural and elastic properties of interfaces between these two phases (see Figure 1) without any external load as well as under extreme uniaxial loading conditions leading to the instability of these composites. In particular, we compare the properties of the two-phase nanocomposite with properties of both constituting phases. 𝛔 𝛔 𝛔 z y x axial stress lateral stresses interface c a b disordered ordered D03 Figure 1. A schematic visualization of a supercell used in our ab initio calculations. The 32-atom supercell contains a disordered Fe-Al phase (left-hand side) and an ordered Fe 3 Al compound with the D03structure (right-hand side). The interface between both phases is highlighted by the blue planes. Let us summarize the main results of the previous studies on these constituting phases first. Watson and Weinert [ 39 ] reported heats of formation for binary and ternary 3d transition-metal Nanomaterials 2018,8, 873 3 of 20 (Ti, V, Fe, and Ni) aluminides using the local density approximation (LDA). They found their predictions in the case of Fe aluminides overestimated by about 0.15 eV/atom when compared with experimental data. As the most likely reason they identified a poor description of bcc Fe by LDA (a fact which lead us to use the generalized gradient approximation (GGA) in our work). Another calculations by Gonzales-Ormeno et al. [ 40 ] were performed employing Perdew-Burke-Ernzerhof (PBE) parametrization [ 41 ] of GGA and Full Potential - Linear Augmented Plane Wave method (FP-LAPW). The computed formation energies of the D0 3 (Fe 3 Al) and B2 (FeAl) compounds show excellent agreement with available calorimetric data on standard enthalpies of formation of Fe-Al alloys up to 50 at.% aluminium. Lechermann et al. [ 42 ] demonstrated that neither LDA nor PBE parametrization of GGA can correctly reproduce the D0 3 structure as the ground state of Fe 3 Al and both above mentioned parametrizations of the exchange-correlation energy prefer the L1 2 structure (at T = 0 K and in the case of defect-free static lattices without any collective excitations). Subsequently, in another paper [ 43 ], Lechermann and co-workers studied electronic correlations and magnetism in Fe 3 Al employing local density approximation with an additional energy term (LDA+U) and correctly obtained the D0 3 structure as the ground-state structure of Fe 3 Al. Similarly, Connetable and Maugis [ 44 ] calculated structural, magnetic, elastic and vibrational properties of Fe 3 Al employing PBE parametrization [ 41 ] of the GGA and found out that Fe 3 Al has a lower energy in the L1 2 structure than in the experimentally observed D03structure. As far as solid solutions of Al and Fe are concerned, Amara and co-workers [ 45 ] performed first-principles calculations to study the electronic structure and energetics of the dissolution of aluminum in α -iron and the interaction between Al atoms and vacancies. It was found that the stability of these complexes is mainly driven by strong Al-vacancy attractions whereas Al-Al interactions are repulsive. Liu et al. [ 46 ] calculated the difference in vibrational entropy between chemically disordered and ordered Fe-Al compounds. Kulikov et al. [ 47 ] have studied the electronic structure of disordered bcc Fe x Al 1−x (0.4 <x< 0.75) alloys around the equiatomic stoichiometry, as well as of the ordered B2-structure FeAl phases with point defects employing the coherent potential approximation within the Korringa-Kohn-Rostoker (KKR) method for the disordered case and the tight-binding linear muffin-tin orbital (TB-LMTO) method for the intermetallic compounds. Studying in particular the onset of magnetism in Fe-Al they found the appearance of large local magnetic moments associated with the transition metal antisite defect in FeAl, in agreement with the experimental findings. Furthermore, Friák and Neugebauer [ 48 ] performed an ab initio study of a dense set of Fe-Al compositions and local atomic arrangements in order to explain the anomalous volume-composition dependence in Fe-Al alloys. They found that the spin-polarized calculations for Fe-rich compounds reproduce very well the anomalous lattice-constant behavior in contrast to both the nonmagnetic and fixed-spin-moment calculations that result in nearly linear trends without any anomaly. The change in magnetism of iron atoms caused by an increasing number of Al atoms in the first coordination spheres was thus identified as the decisive driving force of the anomalous behavior. Regarding other published papers, Fähnle et al. [ 49 ] applied cluster-expansion method to predict the phase diagram for the system Ni-Fe-Al, Friák et al. [ 50 ] studied an impact of solutes (in particular Ti additions) on the elastic properties of Fe 3 Al both theoretically by first-principles calculations and experimentally by ultrasonic measurements, Kirklin et al. [ 51 ] performed a high-throughput computational search for strengthening precipitates in alloys including Fe matrix and Fe-Al-based compounds, Airiskallio et al. [ 52 ] studied corrosion resistance of Fe–Al and Fe-Al-Cr alloys in oxidizing environment using the Exact Muffin-Tin Orbitals (EMTO) method as an alternative of screened Korringa-Kohn-Rostoker (KKR) method, Medvedeva et al. [ 53 ] calculated impact of Al and C on the stacking fault energies in fcc Fe using generalized gradient approximation and the projector-augmented waves (PAW) potentials [ 54 ] implemented in the same code VASP [ 55 ] as we use (see below), ˇ Cížek et al. [ 56 ] used quantum-mechanical calculations when characterizing quenched-in vacancies in Fe-Al alloys, Ipser et al. [ 57 ] developed a statistical-thermodynamic model for intermetallic Nanomaterials 2018,8, 873 4 of 20 phases with D0 3 -structure and Kellou et al. [ 58 ] used DFT-GGA calculations to study the magnetic properties Fe3Al and Fe3AlX (X = H, B, C, N, O) compounds. As an evidence of how intensive have been first principles calculations applied in the case of Fe-Al-based materials it should be noted that the papers listed above represent only minor part of all publications focused on individual phases within the Fe-Al binary system so far. This paper is organized as follows. After the Introduction, Section 2describes computational details. The results obtained are discussed in Sections 3and 4then presents the conclusions and summarizes the whole paper. 2. Materials and Methods The present simulations were performed with the help of ab initio total-energy and molecular-dynamics program VASP (Vienna ab initio simulation package) developed at the Fakultät für Physik, Universität Wien [ 55 ]. In the presented study the electron interactions were described with the projector-augmented waves (PAW) potentials as supplied with the VASP code [ 54 ] and the exchange correlation energy was evaluated by means of the generalized gradient approximation (GGA) with parametrization of Perdew-Wang [ 59 ]. A Methfessel-Paxton method of the first order was adopted with a smearing width of 0.1eV. Importantly, our setting prefers the experimentally observed D03structure of Fe3Al over the L12structure by 5 meV/atom. The sampling of the Brillouin zone was done using Monkhorst-Pack [ 60 ] grids 5 × 5 × 1, 5 × 5 × 3 and 5 × 5 × 5 for the simulation cells containing 64 (double cell - composite), 32 (composite) and 16 (D0 3 and disordered phases) atoms, respectively. The convergence steps in the DFT cycle were considered as self-consistent when the differences in energy between two consequent steps was below 10 −6 eV/(sim. cell) and the plane wave basis set was expanded with the cut off energy 350eV. During the simulations it was necessary to optimize all atomic positions and the cell shape. The atomic positions were optimized using the internal VASP procedure until all forces between atoms were lower than 10 meV/Åwhile for the optimization of the cell shape we used our own external program that cooperated with the VASP code via reading its output files and writing new structure input files. In stress control calculations this program allowed us to relax the stress tensor components to their targeted values within the selected tolerance. In this work this tolerance was set to be ± 0.10 GPa. In all present calculations, magnetism was included via spin polarization and all simulations always started in ferromagnetic state. The simulation supercell used in the present work is depicted in Figure 1together with the cell dimensions and orientation of the coordinate system. This cell is assembled from two parts where the first one corresponds to the ordered Fe 3 Al phase with the D0 3 structure and the second one to the disordered Fe-Al, i.e., a solid solution of Al atoms in bcc Fe with 18.75 at.% Al and the atoms distributed according to the special quasi-random structure (SQS) concept developed by Zunger et al. [ 61 ]. The SQS concept is based on the idea that atoms are distributed in a rather small periodically repeated supercell in such a way that their statistical characteristics (average occupations of nearest-neighbor shells, so-called Warren-Cowley short-range order (SRO) parameters) mimic those in an ideal disordered solid solution of atoms with the same chemical composition in an infinitely large system. In order to achieve this goal up to, e.g., second, third or fourth coordination shell, different local atomic environments are typically included. For example, the SQS-part of Figure 1contains Al atoms distributed in way that they mutually form the first and second nearest neighbor pairs. This is in a clear contrast to the Fe 3 Al-part of Figure 1which contains Al atoms forming solely the third nearest neighbor pairs (a characteristic feature of the D0 3 structure). The interface between both parts (the blue plane) is located in the middle and due to application of the periodic boundary conditions in ab initio simulations is also located at the cell edges with respect to the zdirection. The strength characteristics in this work are represented by the stress-strain responses obtained from the tensile loading that was always oriented perpendicular to the interface (or along the equivalent directions for the perfect crystals). The stress and strain acting perpendicular to the interface are Nanomaterials 2018,8, 873 5 of 20 denoted as the axial stress σax and axial strain εax , respectively. The maxima at the stress-strain dependence can be considered as the tensile strength σts . For comparison, the tensile strength characteristics were not only determined for the simulation cell depicted in Figure 1but also for the perfect D0 3 and the disordered phases. It must be pointed out that we have simulated behavior of defect-free systems while some crystal defects or instabilities might lower the tensile strength ( e.g., dislocations , grain boundaries, phonon or elastic instability, etc.) before reaching the maxima at the stress-strain curve. However, quantum-mechanical phonon calculations, which can assess temperature effect, are computationally very demanding and, therefore, we leave it for future studies. To obtain at least partial information related to finite-temperature properties, we complement our ab initio calculations with the molecular dynamics simulations (MD) to see if and how temperature decrease the tensile strength σts . For the MD simulations, we used the code LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) [ 62 ] with the embedded-atom method (EAM) potential type [63]. To check the precision of the MD potential we performed the comparison of the stress-strain dependencies and the related tensile strengths with those obtained from first-principles calculations. This cross-checking of atomistic and quantum-mechanical methods was realized under quasi-static MD simulations which means that the atomic motions were set to absolute zero (no kinetic energy). All the atoms in the simulation were thus kept frozen and the tensile tests were realized in a similar way as in the case of ab initio simulations, e.g., via homogeneous dilatation of the simulation cell followed by the optimization of all atomic positions and the cell shape at each strain increment. The Polak-Ribiere version of the conjugate gradient (CG) algorithm was used for the atomic optimization during the quasi static simulations performed by LAMMPS. For the MD simulations (non quasi-static) we set the time step to 2 fs and the strain rate was chosen 10−4/ps. In quantum-mechanical calculations, the stress values can be computed from the total energy changes between two strain increments according the following formula σax =1 V dEtot dεax (1) where dEtot is the total energy change between two consequent deformation steps, dε is the increment of the axial strain and V corresponds to the volume of the simulation cell (a recent review of strength studies may be found in Ref. [ 64 ]). Another way how to obtain the axial stress is to read its value directly from the VASP or LAMMPS outputs. The ab initio results presented in this work are based on the stress tensor that was obtained directly from the VASP output (OUTCAR). This is due to fact that we read not only the axial stress but also all remaining stress tensor components which cannot be obtained from Equation (1) . Because the axial stress computed from the total energy changes is less sensitive to the settings of first-principles simulations we compare the values from both approaches to check whether the simulation settings are sufficient to obtain reliable results. The deformation of the supercell in the quantum-mechanical and MD (quasi-static) simulations was realized via homogeneous straining along the axis perpendicular to the interface. During each strain increment the cell shape and ionic positions were optimized within the cell according to selected deformation model (for a detailed review, see e.g., Refs. [ 65 , 66 ]). Because the real crystal structures are mostly subjected to complex loading conditions, we systematically studied the mechanical responses under optimized uniaxial deformation (OUD), optimized uniaxial loading (OUL) and also under superimposed lateral stress σlt. The OUD mode is based on the strain increments along only one direction while the dimensions of the crystal along other two directions remain constant during the entire tensile test. Thus, the simulation cell changes its shape only in one direction ( εx= 0, εy= 0, εax 6= 0). Here, we would like to point out, that the OUD model usually leads to the triaxial loading state [ 65 , 66 ]. On the other hand, the OUL model comprises relaxation of the lateral stresses and therefore all three dimensions change during the deformation ( εx6= 0, εy6= 0, εax 6= 0). The last deformation model Nanomaterials 2018,8, 873 6 of 20 is very similar to the OUL model and difference is that both lateral stresses are not relaxed close to zero value. Instead of that, they are relaxed close to predefined certain values which are kept constant for the entire tensile test. Because both lateral stresses are chosen to be equal in all our models ( σx = σy ) we marked them as lateral stress σlt . In this work, the lateral stresses were chosen from a range starting from 0GPa to 20GPa with a step 5GPa. Due the cubic crystal symmetry and the loading conditions in the [001] direction there are no shear stresses and therefore the stress tensor acquires a following simple form: b σ=        σlt 0 0 0σlt 0 0 0 σax        (2) 3. Results 3.1. Mechanical Properties of Individual Phases (D03and Disordered) As the first step, we obtained the stress-strain characteristics of the D0 3 and the disordered phases subjected to the uniaxial deformation (OUD) and the uniaxial loading (OUL) deformation modes. The resulting stress-strain responses for the D0 3 phase are depicted in Figure 2a. Surprisingly, all three curves for the OUD model (the axial ( σax ) and two laterals ( σx and σy )) are almost identical with respect to the axial strain εax values except the location close to maxima where the axial stress σax has a slightly higher values than the lateral ones. Figure 2. The stress-strain dependencies obtained from ab initio simulations for the uniaxial deformation (OUD) and uniaxial loading (OUL) deformation models for perfect Fe 3 Al with the D0 3 structure ( a ), a disordered Fe-Al phase ( b ) and their nanocomposite ( c ). The blue, magenta and orange curves represent axial and two transverse stresses for the OUD model while the red one belongs to the axial load of the OUL model. Elastic instabilities are marked by black dashed lines. Nanomaterials 2018,8, 873 7 of 20 The fact that the uniaxial deformation (OUD) leads to nearly identical values of all three normal stresses, i.e., σx=σy=σax , means that the loading is almost hydrostatic (the stresses are nearly equal). We would like to point out that the material response to the uniaxial deformation is usually represented by smaller values of the lateral stresses with respect to the axial one (see, e.g., results for perfect Ni crystal in direction [001] and [120] in Ref. [65]). It is worth noting that the expected maxima of the axial strain for the OUD model in Figure 2is located in rather large values of the axial strain. The strain in the direction [001] is, in fact, comparable with deformations related to the Bain’s transformation path when the bcc lattice transforms into the fcc one. Indeed, in the case of bcc-based D0 3 structure these conditions occur for the OUD model when the axial strain reaches value ( √2− 1) [ 67 – 69 ]. Beyond this value the structure cannot be considered as D0 3 and hence the axial stress at this transformation point may be considered as the theoretical tensile stress σts . The axial strains corresponding to the Bain’s deformation path are marked in Figure 2a by black vertical dashed line and the corresponding tensile strength σOUD ts for the OUD model is determined in the case of Fe 3 Al to be 24.7GPa. There is also a question whether the D0 3 is elastically stable for such large strain values. We will focus on this question in Section 3.5. Regarding the uniaxial loading (OUL), it revealed a completely different response (see Figure 2a). As a result of the optimization of the lateral stresses to zero values during the entire deformation path, e.g., allowing the Poisson’s contraction, OUL predicts remarkably lower values of the axial stress σax and the related tensile strength ( σOUL ts = 4.1GPa), i.e., only 17% of the value found in the case of the OUD model. As far as the strains corresponding to the bcc-to-fcc transition according to the Bain’s deformation path are concerned, they are marked in Figure 2a by black vertical dashed line but the value for the OUL model does not have any impact on the tensile strength due to its location after the maximum of the stress. Importantly, from the comparison of the OUD and OUL approaches above it is evident, that the strength σts of the D0 3 phase is very sensitive to the lateral stresses σlt and even small increases/decreases of these stresses increase/decrease the tensile strength σts . For this reason, we investigated the influence of σlt on the tensile strength σts in more detail and these results are described and discussed in Section 3.3. The stress-strain dependence for the OUL model also shows a plateau for the strain values in the range 0.05–0.1. This plateau means zero value of the corresponding elastic constant in this region and possibility of some structure instability. Hence, it is a question which maximum at the stress-strain should be considered as the theoretical tensile strength σts. To answer this question we computed elastic constants for several points of the stress-strain curves and also employed the molecular dynamics simulations to include temperature effect (lattice vibration) into the simulations. These results are discussed in Sections 3.4 and 3.5. Because the stress-strain curve in Figure 2a for the OUL model contains the plateau and has a different shape compared to the other structures there is a question whether the precision of the ab initio simulations in this case is sufficient. For this reason, we also performed the tensile test using very precise simulation settings (the cut-off energy was increased to 500eV, the tolerance of the DFT cycle to 10 −8 eV and the k-point grid to 11 ×11 ×11); the obtained stress-strain curve is practically identical to that one obtained earlier. In the next stage, we performed the tensile tests for the disordered Fe-Al phase with 18.75 at.% Al and the obtained stress-strain behavior is illustrated in Figure 2b. For the disordered phase and the OUD model the tensile strength σOUD ts was determined to be equal to 22.9GPa which is slightly lower than for the D0 3 (24.7GPa). On the contrary, the tensile strength for the OUL reach almost doubled value ( σOUL ts = 7.8GPa) when compared with the Fe 3 Al phase ( σOUL ts = 4.1GPa). Thus, the disordered phase is less sensitive to the transversal stresses than the ordered Fe 3 Al phase and it has higher tensile strength σOUL ts under the uniaxial loading. We note that the lateral stresses σx and σy reach very high values close to the axial one ( σax ). This indicates that the uniaxial deformation also leads to a stress state which is very close to the hydrostatic one (as mentioned above for the Fe 3 Al phase). Also, the stress-strain curve for the OUL is smooth, does not contain any irregularities and its maximum point is followed by sudden drop which indicates a fracture in the structure. Nanomaterials 2018,8, 873 8 of 20 The obtained tensile strengths σts are summarized in Table 1for all Fe-Al alloys and the deformation models applied in this work. The OUD value of the tensile strength of the Fe 3 Al compound in the [001] direction σts , 22.0 GPa, is rather high and of the same order of magnitude as the OUL value of 20 GPa reported for this material for the loading along the [111] direction [ 70 ]. A similar difference (12.7 GPa for the [001] direction and 27.3 GPa for the [111] one) was also found in Fe [ 71 ]. As we are not aware of any other calculations of strength for this material, the values shown in Table 1are the first ab initio calculated values of strength for loading along the [001] direction also for the Fe 3 Al in the D03structure. Table 1. The tensile strengths σts in the [001] direction for the Fe 3 Al compound and the disordered Fe-Al phase with 18.75 at.% Al together with their nanocomposite from ab initio calculations, from quasi-static simulations and from the molecular dynamics simulations at temperature of 1K. The table contains the tensile strengths obtained from maximum at stress-strain dependence (ab initio; except of the value of 24.7 GPa for Fe 3 Al OUD, which corresponds rather to a structural transformation), the ab initio tensile strength obtained from elastic instability (ab initio + ei), molecular static (MD (qs)) and molecular dynamics at the temperature of 1K. ab initio ab initio + ei MD (qs) MD (1K) Fe3Al OUD 24.7 22.0 18.6 16.7 Fe3Al OUL 4.1 4.1 6.0 6.0 Fe-Al disordered OUD 22.9 - 17.7 17.3 Fe-Al disordered OUL 7.8 - 8.5 8.4 nanocomposite OUD 23.0 21.4 16.6 16.0 nanocomposite OUL 5.6 5.5 7.0 7.2 3.2. Mechanical Response of the Fe3Al/Fe-Al Nanocomposite The next part of our ab initio simulations was focused to the determination of properties of the nanocomposite consisting of the ordered Fe 3 Al and the disordered Fe-Al phase. We applied the uniaxial deformation (OUD) and uniaxial loading (OUL) to the entire simulation cell shown in Figure 1 and compared the results with those summarized in the previous section for the individual phases. The stress-strain curves for the nanocomposite are depicted in Figure 2c and it can be seen that the shape of these curves is smooth for both OUD and OUL deformation models. Here, the uniaxial tensile strength σOUD ts = 23GPa is almost the same as for the perfect disordered phases. This fact indicates that the strength found for uniaxial straining (here represented by triaxial loading state) of Fe 3 Al will be determined by the strength of the weaker disordered phase. On the other hand, the tensile strength for the uniaxial loading is σOUL ts = 5.6GPa is located between the values of the strength calculated for Fe 3 Al and disordered Fe-Al phase. The maximum achieved strain values also indicate that the presence of the disordered structure increases the brittleness of the nanocomposite because these maximum strains are equal to those computed for this phase. In summary, if we neglect small increases of the strength in case of uniaxial loading we can conclude that the presence of the disordered phase has a negative effect on the mechanical characteristics, in particular a small reduction of the strength for uniaxial deformation and significant reduction of strains corresponding to the theoretical tensile strength. 3.3. Influence of Lateral Stresses on the Strength The previous results revealed a very high influence of the lateral stresses σlt on the tensile strength σts for all studied materials. Hence, in this section we analyze this effect in detail in order to clarify its impact on the mechanical characteristics of the studied Fe-Al-based systems. Here, using the quasi-static ( ab initio and molecular dynamics) simulations we performed several sets of the tensile tests where each individual test was realized under predefined constant value of the lateral stress σlt . For example, to obtain the tensile strength for Fe 3 Al phase as a function of the lateral stress σlt we performed five tensile tests where each test was realized under a particular constant value of σlt . Nanomaterials 2018,8, 873 9 of 20 As mentioned in Section 2these values were chosen to be σlt = (0, 5, 10, 15, 20GPa). We note that the tensile test obtain for σlt = 0GPa is identical to the OUL model (the uniaxial loading). The obtained tensile strength σts as a function of the lateral stress σlt is illustrated in Figure 3 for Fe 3 Al, disordered Fe-Al phase as well as their nanocomposite. Here, the red curves represent the data obtained from the quantum-mechanical simulations whereas the blue ones are the data from the molecular quasi-static simulations. From these dependencies it is obvious that the tensile strength σts for all structures linearly increases with increasing of the lateral stress σlt and therefore it can be approximated by linear functions σts(σlt)using the formula σts =γσlt +σOUL ts , (3) where the σlt is the selected value of the lateral stress, the γ represents the slope of this dependence and the σOUL ts is the tensile strength in corresponding direction for the OUL model. Let us note that the same equation was used by ˇ Cerný and Pokluda for perfect bcc, fcc and hcp crystals [ 72 , 73 ]. Those papers also contain the values of the coefficients γ (denoted as kmax or s in the above mentioned papers) and the theoretical strengths for uniaxial loading σOUL ts (marked as σmax,0 or σr ). Hence, the present results obtained for Fe-Al systems can be compared with perfect crystals, in particular with Fe. However, perfect Al crystal was not considered in Refs. [ 72 , 73 ]. For this reason, we supplement the present results with the data calculated for perfect fcc Al crystal, i.e., from the stress-strain curves for all deformation models considered in the present work. 0 5 10 15 20 applied lateral stress (GPa) 5 10 15 20 tensile strength (GPa) D03 EAM-FS (MS) D03 DFT-GGA composite EAM-FS (MS) composite DFT-GGA disordered EAM-FS (MS) disordered DFT-GGA Figure 3. The effect of the transverse stresses on the tensile strength as studied by quantum-mechanical calculations (marked as DFT-GGA) and atomistic Embeded Atom Method (EAM) potentials (marked as EAM-FS). The results obtained in Refs. [ 72 , 73 ] revealed that the most fcc and bcc crystals with linear dependencies of σts on σlt have the slope γ mostly positive with higher values for bcc crystals than fcc ones. Interestingly, our results computed for perfect fcc Al crystal showed behavior similar to Ni or Cu where the tensile strength σts is insensitive to the applied lateral stress σlt . This means that the tensile strength σts of Al always reaches the same value for all present deformation models. Of course, there are some small differences between the computed values, however, these differences are smaller than the stress convergence criteria introduced in the computational details, e.g., the σts is always located in the range of h 11.31;11.43 i GPa. For this reason we consider the slope to be γ = 0. We must point out that the range of the lateral stress σlt used for Al is only within the values 0, 5, 10 GPa. On the other hand, the slope γ for perfect Fe is 0.63 [ 72 ] and more interestingly 0.79–0.89 for the Fe-Al-based Nanomaterials 2018,8, 873 16 of 20 (EAM) simulations show that temperature significantly affects the mechanical properties compared to those obtained from quasi-static simulations. For example, at room temperature of 300K the strength decreases to as low as 75% of the zero-Kelvin static lattice value. Author Contributions: Conceptualization, P.Š. and M.F.; Methodology, P.Š., D.H. and M.V. ; Resources, M.F. and M.Š.; Writing—Original Draft Preparation, P.Š. and M.F.; Writing—Review & Editing, M.V., D.H., and M.Š.; Visualization, P.Š. and M.F.; Project Administration, M.F. and M.Š.; Funding Acquisition, M.F. and M.Š. Funding: The authors acknowledge the Czech Science Foundation for the financial support received under the Projects No. 17-22139S (P.Š. and M.F.) and 16-24711S (M.V. and M.Š.). Additional resources were provided by the Academy of Sciences of the Czech Republic through the Fellowship of J. E. Purkynˇe (M.F.) and by the Ministry of Education, Youth and Sports of the Czech Republic under the Project CEITEC 2020, LQ1601 (M.Š and M.V.). Acknowledgments: M.F., P.Š. and M.Š. also acknowledge supports from the Academy of Sciences of the Czech Republic (Institutional Project No. RVO:68081723) and from the Ministry of Education, Youth and Sports of the Czech Republic via the research infrastructure IPMINFRA, LM2015069 (P.Š., M.F.). Computational resources were made available by the Ministry of Education, Youth and Sports of the Czech Republic under the Projects CESNET (Project No. LM2015042), CERIT-Scientific Cloud (Project No. LM2015085) and IT4Innovations National Supercomputer Center (Project No. LM2015070) within the program Projects of Large Research, Development and Innovations Infrastructures. Conflicts of Interest: The authors declare no conflict of interest. References 1. Hotar, A.; Kejzlar, P.; Palm, M.; Minarik, J. The effect of Zr on high-temperature oxidation behaviour of Fe3Al-based alloys. Corros. Sci. 2015,100, 147–157. [CrossRef] 2. Brito, P.; Schuller, E.; Silva, J.; Campos, T.; de Araujo, C.R.; Carneiro, J.R. Electrochemical corrosion behaviour of (100), (110) and (111) Fe3A single crystals in sulphuric acid. Corros. Sci. 2017,126, 366–373. [CrossRef] 3. Sauthoff, G. Intermetallics; VCH Verlagsgesellschaft: Weinheim, Germany, 1995. 4. Liu, C.T.; Stringer, J.; Mundy, J.N.; Horton, L.L.; Angelini, P. Ordered intermetallic alloys: An assessment. Intermetallics 1997,5, 579–596. [CrossRef] 5. Stoloff, N.S. Iron aluminides: Present status and future prospects. Mater. Sci. Eng. A 1998 ,258, 1–14. [CrossRef] 6. Liu, C.T.; Lee, E.H.; McKamey, C.G. An environmental-effect as the major cause for room-temperature embrittlement in FeAl. Scr. Metall. 1989,23, 875–880. [CrossRef] 7. Lynch, R.J.; Heldt, L.A.; Milligan, W.W. Effects of alloy composition on environmental embrittlement of B2 ordered iron aluminides. Scr. Metall. 1991,25, 2147–2151. [CrossRef] 8. Liu, C.T.; McKamey, C.G.; Lee, E.H. Environmental-effects on room-temperature ductility and fracture in Fe3Al. Scr. Metall. 1990,24, 385–389. [CrossRef] 9. Lynch, R.J.; Gee, K.A.; Heldt, L.A. Environmental embrittlement of single-crystal and thermomechanically processed B2-ordered iron aluminides. Scr. Metall. 1994,30, 945–950. [CrossRef] 10. Li, X.; Prokopcakova, P.; Palm, M. Microstructure and mechanical properties of Fe-Al-Ti-B alloys with additions of Mo and W. Mater. Sci. Eng. A Struct. Mater. Prop. Microstruct. Process. 2014 ,611, 234–241. [CrossRef] 11. Azmi, S.A.; Michalcova, A.; Sencekova, L.; Palm, M. Microstructure and mechanical properties of Fe-Al-Nb-B alloys. MRS Adv. 2017,2, 1353–1359. [CrossRef] 12. Lazinska, M.; Durejko, T.; Czujko, T.; Bojar, Z. The Effect of the Traverse Feed Rate on the Microstructure and Mechanical Properties of Laser Deposited Fe 3 Al (Zr,B) Intermetallic Alloy. Materials 2018 ,11, 792, doi:10.3390/ma11050792. [CrossRef] [PubMed] 13. Kratochvil, P.; Danis, S.; Minarik, P.; Pesicka, J.; Kral, R. Strengthening of Fe 3 Al Aluminides by One or Two Solute Elements. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 2017,48A, 4135–4139. [CrossRef] 14. Zamanzade, M.; Barnoush, A.; Motz, C. A Review on the Properties of Iron Aluminide Intermetallics. Crystals 2016,6, 10, doi:10.3390/cryst6010010. [CrossRef] 15. Dobeš, F.; Dymáˇcek, P.; Friák, M. Small punch creep of Fe-Al-Cr alloy with Ce addition and its relation to uniaxial creep tests. Kov. Mater. Metal. Mater. 2018,56, 205–212. [CrossRef] Nanomaterials 2018,8, 873 17 of 20 16. Friák, M.; Oweisová, S.; Pavl˚u, J.; Holec, D.; Šob, M. An ab initio study of thermodynamic and mechanical stability of Heusler-based Fe2AlCo polymorphs. Materials 2018,11, 1543. [CrossRef] [PubMed] 17. Slávik, A.; Miháliková, I.; Friák, M.; Všianská, M.; Šob, M. Quantum-mechanical study of magnetic properties of superalloy nanocomposite phase Fe 2 AlTi. In Proceedings of the NANOCON 2017 Conference Proceedings (9th International Conference on Nanomaterials—Research & Application, Brno, Czech Republic, 18–20 October 2017; Tanger Ltd.: Ostrava, Czech Republic, 2018; pp. 63–68. 18. Miháliková, I.; Friák, M.; Slávik, A.; Všianská, M.; Koutná, N.; Holec, D.; Šob, M. First-principles study of interface energies in Fe-Al-based superalloy nanocomposites. In Proceedings of the NANOCON 2017 Conference Proceedings (9th International Conference on Nanomaterials—Research & Application, Brno, Czech Republic, 18–20 October 2017; Tanger Ltd.: Ostrava, Czech Republic, 2018; pp. 69–74. 19. Friák, M.; Slávik, A.; Miháliková, I.; Holec, D.; Všianská, M.; Šob, M.; Palm, M.; Neugebauer, J. Origin of the low magnetic moment in Fe2AlTi: An ab initio study. Materials 2018,11, 1732. [CrossRef] [PubMed] 20. Kratochvil, P.; Pesicka, J.; Kral, R.; Svec, M.; Palm, M. Evaluation of solid-solution hardening of Fe-27 at. pct Al by vanadium and comparison to precipitation strengthening by vanadium carbides. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 2015,46A, 5091–5094. [CrossRef] 21. Sencekova, L.; Palm, M.; Pesicka, J.; Vesely, J. Microstructures, mechanical properties and oxidation behaviour of single-phase Fe 3 Al (D0 3 ) and two-phase α -Fe-Al (A2) + Fe 3 Al (D0 3 ) Fe-Al-V alloys. Intermetallics 2016 , 73, 58–66. [CrossRef] 22. Shahid, R.N.; Scudino, S. Strengthening of Al-Fe 3 Al composites by the generation of harmonic structures. Sci. Rep. 2018,8, doi:10.1038/s41598-018-24824-y. [CrossRef] [PubMed] 23. Verona, M.N.; Setti, D.; Cortes Paredes, R.S. Microstructure and Properties of Fe 3 Al-Fe 3 AlC (x) Composite Prepared by Reactive Liquid Processing. Metall. Mater. Trans. B Process Metall. Mater. Process. Sci. 2018 , 49, 529–536. [CrossRef] 24. Prakash, U. Intermetallic matrix composites based on iron aluminides. In Intermetallic Matrix Composites: Properties and Applications; Mitra, R., Ed.; Woodhead Publishing Series in Composites Science and Engineering; Elsevier Science: Amsterdam, The Netherlands, 2018; pp. 21–35. 25. Sharifitabar, M.; Khaki, J.V.; Sabzevar, M.H. Formation mechanism of TiC-Al 2 O 3 -Fe 3 Al composites during self-propagating high-temperature synthesis of TiO 2 -Al-C-Fe system. Ceram. Int. 2016 ,42, 12361–12370. [CrossRef] 26. Duan, X.; Gao, S.; Dong, Q.; Zhou, Y.; Xi, M.; Xian, X.; Wang, B. Reinforcement mechanism and wear resistance of Al 2 O 3 /Fe-Cr-Mo steel composite coating produced by laser cladding. Surf. Coat. Technol. 2016 , 291, 230–238. [CrossRef] 27. Kong, J.; Wei, Y.; Li, J.; Huang, J.; Wang, T. Microwave-assisted combustion synthesis of Fe 3 Al bulk nanocrystalline intermetallic matrix composites. Adv. Powder Technol. 2015,26, 778–782. [CrossRef] 28. Imandoust, A.; Zarei-Hanzaki, A.; Ou, K.L.; Yu, C.H. D0 3 Ordered Phase Strengthening in Dual Phase Twinning-Induced Plasticity Steel. J. Mater. Eng. Perform. 2015,24, 2085–2090. [CrossRef] 29. Cheng, J.; Yin, B.; Qiao, Z.; Yang, J.; Liu, W. Mechanical and dry-sliding tribological properties of Fe 3 Al based composites reinforced by novel W0.5Al0.5C0.5 particulates. Mater. Des. 2015,66, 67–76. [CrossRef] 30. Molina, A.; Torres-Islas, A.; Serna, S.; Acosta-Flores, M.; Rodriguez-Diaz, R.A.; Colin, J. Corrosion, Electrical and Mechanical Performance of Copper Matrix Composites Produced by Mechanical Alloying and Consolidation. Int. J. Electrochem. Sci. 2015,10, 1728–1741. 31. Bai, Y.; Xing, J.; Guo, Y.; Li, J.; He, Y.; Ma, S. Effect of Cr on Microstructure, Mechanical Properties, and Wear Behavior of In Situ 20 wt.%Al 2 O 3 /Fe-25Al Composites. J. Mater. Eng. Perform. 2015 ,24, 936–945. [CrossRef] 32. Panda, D.; Kumar, L.; Alam, S.N. Development of Al-Fe 3 Al Nanocomposite by Powder Metallurgy Route. Mater. Today Proc. 2015,2, 3565–3574, doi:10.1016/j.matpr.2015.07.070. [CrossRef] 33. Dobes, F.; Kratochvil, P.; Kejzlar, P. Creep of three-phase alloy Fe-30%Al-5.2%Zr. Kovove Mater. Metal. Mater. 2015,53, 127–132. [CrossRef] 34. Jiraskova, Y.; Pizurova, N.; Titov, A.; Janickovic, D.; Friak, M. Phase separation in Fe-Ti-Al alloy - Structural, magnetic, and Mossbauer study. J. Magn. Magn. Mater. 2018,468, 91–99. [CrossRef] 35. Kattner, U.; Burton, B. Al-Fe (Aluminium-Iron). In Phase Diagrams of Binary Iron Alloys; Okamoto, H., Ed.; ASM International: Geauga County, OH, USA, 1993; pp. 12–28. 36. Sundman, B.; Ohnuma, I.; Dupin, N.; Kattner, U.R.; Fries, S.G. An assessment of the entire Al-Fe system including D0(3) ordering. Acta Mater. 2009,57, 2896–2908. [CrossRef] Nanomaterials 2018,8, 873 18 of 20 37. Allen, S.; Cahn, J. Mechanisms of phase-transformations within miscibility gap of Fe-rich Fe-Al alloys. Acta Metall. Mater. 1976,24, 425–437. [CrossRef] 38. Oguma, R.; Matsumura, S.; Eguchi, T. Kinetics of B2-and D0 3 type ordering and formation of domain structures in Fe-Al alloys. J. Phys. Condens. Matter 2008,20, 275225. [CrossRef] [PubMed] 39. Watson, R.E.; Weinert, M. Transition-metal aluminide formation: Ti, V, Fe, and Ni aluminides. Phys. Rev. B 1998,58, 5981–5988. [CrossRef] 40. Gonzales-Ormeno, P.; Petrilli, H.; Schon, C. Ab-initio calculations of the formation energies of BCC-based superlattices in the Fe-Al system. Calphad 2002,26, 573–582. [CrossRef] 41. Perdew, J.P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996,77, 3865–3868. [CrossRef] [PubMed] 42. Lechermann, F.; Welsch, F.; Elsässer, C.; Ederer, C.; Fähnle, M.; Sanchez, J.M.; Meyer, B. Density-functional study of Fe3Al: LSDA versus GGA. Phys. Rev. B 2002,65. [CrossRef] 43. Lechermann, F.; Fähnle, M.; Meyer, B.; Elsässer, C. Electronic correlations, magnetism, and structure of Fe-Al subsystems: An LDA+U study. Phys. Rev. B 2004,69. [CrossRef] 44. Connetable, D.; Maugis, P. First principle calculations of the kappa-Fe 3 AlC perovskite and iron-aluminium intermetallics. Intermetallics 2008,16, 345–352. [CrossRef] 45. Amara, H.; Fu, C.C.; Soisson, F.; Maugis, P. Aluminum and vacancies in α -iron: Dissolution, diffusion, and clustering. Phys. Rev. B 2010,81, 174101. [CrossRef] 46. Liu, S.; Duan, S.; Ma, B. First-principles calculation of vibrational entropy for Fe-Al compounds. Phys. Rev. B 1998,58, 9705–9709. 47. Kulikov, N.I.; Postnikov, A.V.; Borstel, G.; Braun, J. Onset of magnetism in B2 transition-metal aluminides. Phys. Rev. B 1999,59, 6824–6833. [CrossRef] 48. Friák, M.; Neugebauer, J. Ab initio study of the anomalous volume-composition dependence in Fe-Al alloys. Intermetallics 2010,18, 1316–1321. [CrossRef] 49. Fähnle, M.; Drautz, R.; Lechermann, F.; Singer, R.; Diaz-Ortiz, A.; Dosch, H. Thermodynamic properties from ab-initio calculations: New theoretical developments, and applications to various materials systems. Phys. Status Solidi B-Basic Solid State Phys. 2005,242, 1159–1173. [CrossRef] 50. Friák, M.; Deges, J.; Krein, R.; Frommeyer, G.; Neugebauer, J. Combined ab initio and experimental study of structural and elastic properties of Fe3Al-based ternaries. Intermetallics 2010,18, 1310. [CrossRef] 51. Kirklin, S.; Saal, J.E.; Hegde, V.I.; Wolverton, C. High-throughput computational search for strengthening precipitates in alloys. Acta Mater. 2016,102, 125–135. [CrossRef] 52. Airiskallio, E.; Nurmi, E.; Heinonen, M.H.; Vayrynen, I.J.; Kokko, K.; Ropo, M.; Punkkinen, M.P.J.; Pitkanen, H.; Alatalo, M.; Kollar, J.; et al. High temperature oxidation of Fe-Al and Fe-Cr-Al alloys: The role of Cr as a chemically active element. Corros. Sci. 2010,52, 3394–3404. [CrossRef] 53. Medvedeva, N.I.; Park, M.S.; Van Aken, D.C.; Medvedeva, J.E. First-principles study of Mn, Al and C distribution and their effect on stacking fault energies in fcc Fe. J. Alloy Compd. 2014 ,582, 475–482. [CrossRef] 54. Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999,59, 1758–1775. [CrossRef] 55. Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996,54, 11169–11186. [CrossRef] 56. ˇ Cížek, J.; Lukáˇc, F.; Procházka, I.; Kužel, R.; Jirásková, Y.; Janiˇckoviˇc, D.; Anwand, W.; Brauer, G. Characterization of quenched-in vacancies in Fe-Al alloys. Phys. B 2012,407, 2659–2664. [CrossRef] 57. Ipser, H.; Semenova, O.; Krachler, R. Intermetallic phases with D0(3)-structure: A statistical-thermodynamic model. J. Alloy. Compd. 2002,338, 20–25. [CrossRef] 58. Kellou, A.; Grosdidier, T.; Raulot, J.M.; Aourag, H. Atomistic study of magnetism effect on structural stability in Fe 3 Al and Fe 3 AlX (X = H, B, C, N, O) alloys. Phys. Status Solidi B Basic Solid State Phys. 2008 ,245, 750–755. [CrossRef] 59. Perdew, J.P.; Wang, Y. Accurate and simple analytic representation of the electron-gas correlation energy. Phys. Rev. B 1992,45, 13244–13249. [CrossRef] 60. Monkhorst, H.J.; Pack, J.D. Special points for Brillouin-zone integrations. Phys. Rev. B 1976 ,13, 5188–5192. [CrossRef] 61. Zunger, A.; Wei, S.; Ferreira, L.; Bernard, J. Special quasirandom structures. Phys. Rev. Lett. 1990 ,65, 353–356. [CrossRef] [PubMed] Nanomaterials 2018,8, 873 19 of 20 62. Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995 ,117, 1–19. [CrossRef] 63. Mendelev, M.; Srolovitz, D.; Ackland, G.; Han, S. Effect of Fe Segregation on the Migration of a Non-Symmetric Σ5 Tilt Grain Boundary in Al. J. Mater. Res. 2005,20, 208–218. [CrossRef] 64. Pokluda, J.; ˇ Cerný, M.; Šob, M.; Umeno, Y. Ab initio calculations of mechanical properties: Methods and applications. Prog. Mater. Sci. 2015,73, 127–158. [CrossRef] 65. ˇ Cerný, M.; Šesták, P.; ˇ Rehák, P.; Všianská, M.; Šob, M. Ab initio tensile tests of grain boundaries in the fcc crystals of Ni and Co with segregated sp-impurities. Mater. Sci. Eng. A 2016,669, 218–225. [CrossRef] 66. ˇ Cerný, M.; Šesták, P.; ˇ Rehák, P.; Všianská, M.; Šob, M. Atomistic approaches to cleavage of interfaces. 2018, submitted. 67. Šob, M.; Wang, L.G.; Vitek, V. Local stability of higher-energy phases in metallic materials and its relation to the structure of extended defects. Comput. Mater. Sci. 1997,8, 100–106. [CrossRef] 68. Wang, L.G.; Šob, M.; Zhang, Z. Instability of higher-energy phases in simple and transition metals. J. Phys. Chem. Solids 2003,64, 863–872. [CrossRef] 69. Šob, M.; Wang, L.G.; Vitek, V. Ab initio calculation of the ideal tensile strength in copper and nickel aluminide. Kov. Mater. Metal. Mater. 1998,36, 145–152. 70. Legut, D.; Šob, M. Ideal Tensile Strength of Ni 3 Al and Fe 3 Al with D0 3 Structure; Book Series; Materials Science Forum: Zurich, Switzerland, 2008; Volume 567–568, pp. 77–80. 71. Friák, M.; Šob, M.; Vitek, V. Ab initio calculation of tensile strength in iron. Philos. Mag. 2003 ,83, 3529–3537. [CrossRef] 72. ˇ Cerný, M.; Pokluda, J. Influence of superimposed biaxial stress on the tensile strength of perfect crystals from first principles. Phys. Rev. B 2007,76, 024115. [CrossRef] 73. ˇ Cerný, M.; Pokluda, J. Ideal tensile strength of cubic crystals under superimposed transverse biaxial stresses from first principles. Phys. Rev. B 2010,82, 174106. [CrossRef] 74. Friák, M.; Šob, M.; Vítek, V. Proceedings of the International Conference Juniormat-01; Brno University of Technology: Brno, Czech Republic, 2001; pp. 117–120. 75. Clatterbuck, D.M.; Chrzan, D.C.; Morris, J.W., Jr. The inherent tensile strength of iron. Philos. Mag. Lett. 2002,82, 141–147. [CrossRef] 76. Clatterbuck, D.M.; Chrzan, D.C.; Morris, J.W., Jr. The ideal strength of iron in tension and shear. Acta Mater. 2003,51, 2271–2283. [CrossRef] 77. Šob, M.; Friák, M.; Legut, D.; Vítek, V. Complex Inorganic Solids: Structural, Stability, and Magnetic Properties of Alloys; Turchi, P.E.A., Gonis, A., Rajan, K., Meike, A., Eds.; Springer: New York, NY, USA, 2005. 78. Yue-Lin, L.; Ying, Z.; Rong-Jie, H.; Guang-Hong, L. Study of the theoretical tensile strength of Fe by a first-principles computational tensile test. Chin. Phys. B 2009,18, 1923. [CrossRef] 79. Li, W.; Wang, T. Ab initio investigation of the elasticity and stability of aluminium. J. Phys. Condens. Matter 1998,10, 9889–9904. [CrossRef] 80. Clatterbuck, D.M.; Krenn, C.R.; Cohen, M.L.; Morris, J.W., Jr. Phonon Instabilities and the Ideal Strength of Aluminum. Phys. Rev. Lett. 2003,91, 135501. [CrossRef] [PubMed] 81. ˇ Rehák, P.; ˇ Cerný, M.; Pokluda, J. The [100] Compressive Strength of Perfect Cubic Crystals under Superimposed Biaxial Stresses. Key Eng. Mater. 2011,465, 183–186. [CrossRef] 82. ˇ Cerný, M.; Pokluda, J. The theoretical tensile strength of fcc crystals predicted from shear strength calculations. J. Phys. Condens. Matter 2009 ,21, 145406, doi:10.1088/0953-8984/21/14/145406. [CrossRef] [PubMed] 83. Yang, R.; Tang, B.; Gao, T. A comparison of mechanical properties between Al and Al 3 Mg. Int. J. Mod Phys. B 2016,30, 1550243, doi:10.1142/S0217979215502434. [CrossRef] 84. Simmons, G.; Wang, H. Single Crystal Elastic Constants and Calculated Aggregate Properties: A Handbook; MIT Press: Cambridge, MA, USA, 1971; ISBN 0262190923. 85. Moakher, M.; Norris, A.N. The Closest Elastic Tensor of Arbitrary Symmetry to an Elasticity Tensor of Lower Symmetry. J. Elast. 2006,85, 215–263. [CrossRef] 86. Zhou, L.; Holec, D.; Mayrhofer, P.H. First-principles study of elastic properties of cubic Cr 1−x Al x N alloys. J. Appl. Phys. 2013,113. [CrossRef] 87. Mouhat, F.; Coudert, F.X. Necessary and sufficient elastic stability conditions in various crystal systems. Phys. Rev. B 2014,90, 224104. [CrossRef] Nanomaterials 2018,8, 873 20 of 20 88. Titrian, H.; Aydin, U.; Friák, M.; Ma, D.; Raabe, D.; Neugebauer, J. Self-consistent Scale-bridging Approach to Compute the Elasticity of Multi-phase Polycrystalline Materials. MRS Proc. 2013,1524. [CrossRef] 89. Friák, M.; Counts, W.A.; Ma, D.; Sander, B.; Holec, D.; Raabe, D.; Neugebauer, J. Theory-Guided Materials Design of Multi-Phase Ti-Nb Alloys with Bone-Matching Elastic Properties. Materials 2012 ,5, 1853–1872. [CrossRef] 90. Zhu, L.-F.; Friák, M.; Lymperakis, L.; Titrian, H.; Aydin, U.; Janus, A.M.; Fabritius, H.-O.; Ziegler, A.; Nikolov, S.; Hemzalová, P.; et al. Ab initio study of single-crystalline and polycrystalline elastic properties of Mg-substituted calcite crystals. J. Mech. Behav. Biomed. Mater. 2013,20, 296–304. [CrossRef] [PubMed] 91. Tasnádi, F.; Abrikosov, I.A.; Rogström, L.; Almer, J.; Johansson, M.P.; Oden, M. Significant elastic anisotropy in Ti1−xAlxN alloys. Appl. Phys. Lett. 2010,97. [CrossRef] 92. Tasnádi, F.; Odén, M.; Abrikosov, I.A. Ab initio elastic tensor of cubic Ti0.5Al0.5N alloys: Dependence of elastic constants on size and shape of the supercell model and their convergence. Phys. Rev. B 2012 ,85. [CrossRef] 93. von Pezold, J.; Dick, A.; Friak, M.; Neugebauer, J. Generation and performance of special quasirandom structures for studying the elastic properties of random alloys: Application to Al-Ti. Phys. Rev. B 2010 ,81. [CrossRef] 94. Holec, D.; Tasnádi, F.; Wagner, P.; Friák, M.; Neugebauer, J.; Mayrhofer, P.H.; Keckes, J. Macroscopic elastic properties of textured ZrN-AlN polycrystalline aggregates: From ab initio calculations to grainscale interactions. Phys. Rev. B 2014,90. [CrossRef] 95. Koutná, N.; Holec, D.; Friák, M.; Mayrhofer, P.H.; Šob, M. Stability and elasticity of metastable solid solutions and superlattices in the MoN–TaN system: First-principles calculations. Mater. Des. 2018 ,144, 310–322. [CrossRef] 96. Park, N.; Lee, S.C.; Cha, P.R. Effects of alloying elements on the stability and mechanical properties of Fe 3 Al from first-principles calculations. Comput. Mater. Sci. 2018,146, 303–309. [CrossRef] 97. Niu, X.L.; Wang, L.J. Effect of transition-metal substitution on electronic and mechanical properties of Fe 3 Al: First-principles calculations. Comput. Mater. Sci. 2012,53, 128–132. [CrossRef] 98. Liu, Y.; Chong, X.; Jiang, Y.; Zhou, R.; Feng, J. Mechanical properties and electronic structures of Fe-Al intermetallic. Phys. B Condens. Matter 2017,506, 1–11. [CrossRef] 99. Leamy, H.; Gibson, E.; Kayser, F. Elastic stiffness coefficients of iron-aluminum alloys. I. experimental results and thermodynamic analysis. Acta Metall. 1967,15, 1827. [CrossRef] 100. Shaojun, L.; Suqing, D.; Benkun, M. First-principles calculation of vibrational entropy for Fe-Al compounds. Phys. Rev. B 1998,58, 9705–9709. [CrossRef] 101. Voigt, W. Lehrbuch der Kristallphysik; Teubner: Stuttgart, Germany, 1928. 102. Reuss, A. Account of the liquid limit of mixed crystals on the basis of the plasticity condition for single crystal. Z. Angew. Math. Mech. 1929,9, 49–58. [CrossRef] 103. Hershey, A.V. The elasticity of an isotropic aggregate of anisotropic cubic crystals. J. Appl. Mech. 1954 , 21, 236–240. 104. Friák, M.; Hickel, T.; Grabowski, B.; Lymperakis, L.; Udyansky, A.; Dick, A.; Ma, D.; Roters, F.; Zhu, L.F.; Schlieter, A.; et al. Methodological challenges in combining quantum-mechanical and continuum approaches for materials science applications. Eur. Phys. J. Plus 2011,126, 101. [CrossRef] 105. Pugh, S. Relations between the elastic moduli and the plastic properties of polycrystalline pure metals. Philos. Mag. 1954,45, 823–843. [CrossRef] 106. Pettifor, D. Theoretical predictions of structure and related properties of intermetallics. Mater. Sci. Technol. 1992,8, 345–349. [CrossRef] c 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).