Embedded Atom Method Research Articles

Development of interatomic potential suitable for molecular dynamics simulation of Ni oxidation and Ni-NiO interface.

Large-scale molecular dynamics (MD) simulations enabled by computationally efficient semiempirical potentials are an invaluable tool for materials modeling. In the case of metallic alloys, embedded atom method (EAM) and Finnis-Sinclair (FS) potentials are a reasonable choice based on their good balance of quality and computational cost. However, these semiempirical potentials are not suitable for simulating ionic systems, which prevents their use in studying many technologically relevant metal-oxide systems. The charge transfer ionic potential (CTIP), which can utilize EAM/FS potentials available in the literature together with a variable charge representation of electrostatic interactions, should be a reasonable choice for performing reliable and computationally efficient MD simulations of such systems. However, only a few such potentials are available in the literature, and their computational cost is much higher compared to EAM/FS potentials. In the present work, we have attempted to remedy these deficiencies by combining several modifications to the CTIP model proposed in the literature and efficiently implementing them into the widely used Large-scale Atomic/Molecular Massively Parallel Simulator MD code. Using these modifications, we have developed a new Ni-O CTIP parameterization, which has been tested in several different scenarios of interest. First, the early stages of Ni surface oxidation were simulated, demonstrating the nucleation and growth of a crystalline NiO film across the surface. Second, solidification and vitrification in the Ni-O system were investigated, demonstrating that the new CTIP parameterization provides reasonable agreement with the experimentally determined equilibrium phase diagram. Finally, we studied the interaction of dislocations in a Ni matrix with a NiO inclusion using a simulation cell with an unprecedented number of atoms for a variable charge MD simulation. Thus, the approach utilized in the present study is an efficient method to simulate large scale atomic mechanisms in metal-oxide systems.

Chemical short-range order and its influence on selected properties of non-dilute random alloys

In recent years, non-dilute random alloys (NDRAs) have been gaining interest as promising materials for high-temperature structural applications due to their unique ability to form solid solution phases that portray exceptional mechanical properties and resilience under extreme conditions. This study deals with the effects of chemical short-range order (CSRO) on selected material properties including lattice parameters, lattice distortion (LD), unstable stacking fault energy (USFE), and melting points (MPs). The main body of this study is on 21 ternary alloys, where the interatomic interactions are described by an embedded-atom method potential. Our results demonstrate that random structures generally exhibit larger lattice parameters and greater LD compared to CSRO structures, while CSRO structures possess higher USFEs and MPs. Furthermore, there is no discernible pattern in the segregation of the same and different atom pairs in the NDRAs, as revealed by the notably varying local ordering or segregation for most of them. Last, the effects of interatomic potential and the number of constituent elements are studied using other interatomic potentials and/or non-ternary systems, suggesting that our finding is consistent across different potentials/alloys.

Effect of composition, temperature, and grain size on mechanical behavior and deformation mechanism of lightweight magnesium alloy.

To address the severe fuel crisis and environmental pollution, the use of lightweight metal materials, such as AZ alloy, represents an optimal solution. This study investigates the mechanical behavior and deformation mechanism of AZ alloys under uniaxial compressive using molecular dynamics (MD) simulations. The influence of various compositions, grain sizes (GSs), and temperatures on the compressive stress, the ultimate compressive strength (UCS), compressive yield stress (CYS), Young's modulus (E), shear strain, phase transformation, dislocation distribution, and total deformation length is thoroughly examined. The results show that although AZ91 has the highest Al content, it exhibits the lowest UCS, CYS, and fraction atoms with shear strain larger than 0.2 (FSS0.2) compared to AZ31 and AZ61. At the same time, the total dislocation length of AZ31 is the largest. The effect of GS and temperature on the mechanical response and deformation mechanism of AZ31 alloy indicates that a GS of 7.6nm is the critical value to determine the mechanical properties and deformation intensity of AZ31 alloy. Moreover, the E, UCS, and CYS values decrease gradually as temperature increases. The compressive stress, E, UCS, CYS, and FSS0.2 of single crystalline AZ31 are higher than those of polycrystalline AZ31 with all GSs. The MD simulation results show that the sample experiences the formation of stacking faults at both single crystalline and polycrystalline AZ31 while forming the shear band at the single crystalline, leading to strong oscillations in compressive stress. In contrast, polycrystalline AZ31, across all GSs, exhibits high shear strain zones, causing oscillations in compressive stress. The ATOMSK program is used to create the polycrystalline AZ structures. The MD method is employed to investigate the influence of various compositions, GSs, and temperatures on the mechanical properties and deformation mechanism of AZ alloys. All the simulations are performed by LAMMPS software. The visualization tool (OVITO) is used to inspect, analyze, and illustrate the simulation results. The EAM potential is applied to the interactions between Al-Zn, Al-Mg, and Zn-Mg.

Suitability of Available Interatomic Potentials for Sn to Model Its 2D Allotropes.

The suitability of a range of interatomic potentials for elemental tin was evaluated in order to identify an appropriate potential for modeling the stanene (2D tin) allotropes. Structural and mechanical features of the flat (F), low-buckled (LB), high-buckled (HB), full dumbbell (FD), trigonal dumbbell (TD), honeycomb dumbbell (HD), and large honeycomb dumbbell (LHD) monolayer tin (stanene) phases, were gained by means of the density functional theory (DFT) and molecular statics (MS) calculations with ten different Tersoff, modified embedded atom method (MEAM), and machine-learning-based (ML-IAP) interatomic potentials. A systematic quantitative comparison and discussion of the results is reported.

Mechanical Properties of Nickel, Palladium, and Platinum Nanowires: A Molecular Dynamics Study

Transition metal group 10 nanowires exhibit exceptional mechanical properties, particularly those with FCC structures, making them promising candidates for electromechanical devices. This study used classical MD simulations with EAM potentials to explore the mechanical behavior of nickel, palladium, and platinum nanowires with varying diameters. Our findings reveal a strong correlation between nanowire diameter and mechanical properties. Increasing diameter reduces surface effects, leading to higher tensile strength. Deformation mechanisms are complex, involving phase transformations such as FCC to HCP and BCC. These results will contribute to the fundamental understanding of nanoscale mechanics and pave the way for the design of advanced nanodevices

Molecular dynamics study on the nanofriction and wear mechanism of transverse grain boundaries in nickel-based alloys.

This study employs molecular dynamics simulations to investigate the nanoscale tribological behavior of a single transverse grain boundary in a nickel-based polycrystalline alloy. A series of simulations were conducted using a repetitive rotational friction method to explore the mechanisms by which different grain boundary positions influence variations in wear depth, friction force, friction coefficient, dislocation, stress, and internal damage during repeated friction processes. The results reveal that the grain boundary structure enhances the strength of the nanoscale nickel-based polycrystalline alloy. When the friction surface is far from the transverse grain boundary, the grain boundary's obstructive effect is weaker, leading to larger ranges of atomic displacement and migration of internal defects. This results in smaller fluctuations in friction force and coefficient, along with the formation of numerous densely packed downward defect bundles. At the grain boundary, two grains undergo relative slip along the grain boundary interface, while atoms below the grain boundary remain largely unaffected. When the grain boundary is closer to the friction surface, more wear debris atoms accumulate in front of and on the sides of the friction grinding ball, increasing the friction force during the process. If the friction grinding ball breaches the grain boundary layer, its supporting and strengthening effects are diminished, leading to a significantly greater wear depth compared to when the grain boundary remains intact. In this paper, nanoscale modeling was performed in the large-scale atomic/molecular parallel simulator simulation environment (LAMMPS). Three potential functions, namely EAM potential, Morse potential, and Tersoff potential, are used to simulate the interaction between atoms during the processing. The model was visualized and analyzed in three dimensions by Open Visualization Tool (OVITO).

Construction and Experimental Validation of Embedded Potential Functions for Ta-Re Alloys.

Ta/Re layered composite material is a high-temperature material composed of the refractory metal tantalum (Ta) as the matrix and high-melting-point, high-strength rhenium (Re) as the reinforcement layer. It holds significant potential for application in aerospace engine nozzles. Developing the Ta/Re potential function is crucial for understanding the diffusion behavior at the Ta/Re interface and elucidating the high-temperature strengthening and toughening mechanism of Ta/Re layered composites. In this paper, the embedded atom method (EAM) potential function for tantalum/rhenium binary alloys (Ta-Re alloys) is derived using the force-matching method and validated through first-principles calculations and experimental characterization. The results show that for the lattice constant of a bcc structure containing 54 atoms, surface formation energies per unit area of Ta-Re alloys obtained based on the potential function are 12.196 Å, E100 = 0.16 × 10-2 eV, E110 = 0.10 × 10-2 eV, and E111 = 0.08 × 10-2 eV, with error values of 0.015 Å, 0.04 × 10-2 eV, 0.02 × 10-2 eV, and 0.01 × 10-2 eV, respectively, compared with the calculations from first principles calculations. It is noteworthy that the errors in the average binding energies of Ta-rich (Ta39Re20, where the number of Ta atoms is 39 and Re atoms is 20) and Re-rich (Ta20Re39, where the number of Ta atoms is 20 and Re atoms is 39) cluster atoms, calculated by the potential function and first-principles methods, are only 1.64% to 1.98%. These results demonstrate the accuracy of the constructed EAM potential function. Based on this, three compositions of Ta-Re alloys (Ta48Re6, Ta30Re24, and Ta6Re48; the numerical subscripts represent the number of atoms of each corresponding element) were randomly synthesized, and a comparative analysis of their bulk moduli was conducted. The results revealed that the experimental values of the bulk modulus showed a decreasing and then an increasing tendency with the calculated values, which indicated that the potential function has a very good generalization ability. This study can provide theoretical guidance for the modulation of Ta/Re laminate composite properties.

Micromechanical behavior of BCC phases in TiTaNbZrMo and Ti5Ta35Nb20Zr20Mo20 refractory high entropy alloys for total joint replacement

TiTaNbMoZr, a prominent refractory high entropy (RHEA) alloy, shows potential for articulating surfaces in total joint replacement, owing to its remarkable mechanical properties and biocompatibility. This study investigates the micromechanical deformation behavior and strengthening mechanisms of TiTaNbZrMo (Ta1) and Ti5Ta35Nb20Zr20Mo20 (Ta1.75), using experiments and molecular dynamics (MD) simulations. Firstly, a fitted embedded atom method interatomic potential was validated for the constituent BCC phases in RHEAs. Experimental and MD simulations of nanoindentation indicated that Ta1.75 RHEA possesses higher hardness and elastic modulus than Ta1 RHEA. Nanoindentation simulations showed that higher lattice distortion in BCC phases of Ta1.75 RHEA reduced migration of atoms and increased the kink densities hindering further dislocation nucleation and mobility. Moreover, during retraction stage, Ta1.75 RHEA exhibited little relaxation of the dislocation network, smaller plastic zone, and higher density of sessile dislocations. These findings enhance our understanding of strengthening mechanisms in RHEAs and may aid future alloy design.

Thermophysical properties of undercooled Zr–Fe–Nb alloys investigated by electrostatic levitation and molecular dynamics calculation

The density, surface tension, and viscosity of liquid Zr76.0−xFe24.0Nbx (x = 6.6, 10.0, 15.0) alloys were measured by using the electrostatic levitation technique. The maximum undercooling achieved for these alloys was 151, 91, and 119 K, respectively. To evaluate the thermophysical properties in a wider temperature range, molecular dynamics simulations were performed by using the embedded atom method potential. Both measured and simulated results indicate that the liquid density increases linearly with decreasing temperature and also gradually rises with increasing Nb content. Additionally, the simulated and experimental results for surface tension and viscosity were analyzed. In all three alloys, surface tension increases linearly with decreasing temperature. The relationship between viscosity and temperature follows an Arrhenius-type equation, with both surface tension and viscosity increasing as the Nb content increases. The calculated results of density, surface tension, and viscosity are in good agreement with the experimental results. Furthermore, the specific heat, emissivity, and diffusion coefficient of liquid alloys were calculated. The specific heat for liquid Zr76.0−xFe24.0Nbx (x = 6.6, 10.0, 15.0) alloys is (36.47 ± 1.68), (35.20 ± 2.28), and (41.04 ± 3.73) J mol−1 K−1, respectively. Emissivity decreases linearly with temperature. The diffusion coefficient decreases, while the diffusion activation energy increases with a higher Nb content.

Molecular dynamic simulations of adiabatic self-reaction in Al@Ni core-shell nanoparticles

Abstract Al@Ni core-shell nanoparticles are attracting significant attentions for using as a promising material. In this paper, the self-reaction thermochemical behavior of Al@Ni nanoparticles was studied by molecular-dynamics (MD) simulations to characterize the final reaction temperature and self-heating rate. The analysis was carried out in canonical-constant temperature (NVT) and microcanonical-constant energy (NVE) ensembles using an embedded atom method. The ignition temperature was in the range of 700 K ∼ 1300 K, the core-shell thickness ratio of 1:1 ∼ 7:1 and the nanoparticle radius of 2 nm ∼ 8 nm were considered. When the ignition temperature was 1100 K, the nanoparticle radius was 8 nm, the core-shell thickness ratio was 3:1, the final reaction temperature was highest of 1955 K. As the ignition temperature was 1100 K, the nanoparticle radius was 4 nm, the core-shell thickness ratio was 3:1, the self-heating rate was highest of 3.04×1012K/s. These results obtained from the MD simulations of Al@Ni nanoparticles are helpful to understand the energy release mechanism.

Uniaxial Tension Behavior of FeNiCrCoAl and FeNiCrCoTi Complex Concentrated Alloys: A Molecular Dynamics Approach

Complex concentrated alloys have been thoroughly examined for their excellent mechanical properties. In the present study, phase transitions and mechanisms of dislocation for FeNiCrCoAl and FeNiCrCoTi complex concentrated alloys were examined using classical molecular dynamics simulations under uniaxial tension. The (LAMMPS) code was used to simulate CCAs systems by using EAM potentials. The effect of strain rate change has been taken into account. The results show that at the early plastic stage, the main deformation behavior is the transition from FCC to HCP phase. Moreover, tensile characteristics are negatively affected by strain rate rise. At strain rate value of , for FeNiCrCoAl, elastic modulus and density are computed to be and . Whereas, for FeNiCrCoTi, elastic modulus and density are calculated to be and . It can be noticed that elasticity modulus of FeNiCrCoAl is two times greater than that of FeNiCrCoTi. In contrast, young modulus of complex concentrated alloys decreases succinctly when the strain rate increases to a value of . Under tensile deformation, the movement direction and impact on mechanical properties of the prevalent 1/6 <112> Shockley partial dislocations are analyzed.

An explicitly magnetic modified embedded atom method formalism for coupled spin dynamics and molecular dynamics

Abstract In this paper, we augment the modified embedded atom method formalism to include magnetic spin-spin interactions for elements with a persistent magnetic moment. While previous spin coupling methods have been based on pair potentials, our Magnetic MEAM formalism, which we term MagMEAM, incorporates the many-body and angular effects of MEAM allowing for the strength of the magnetic interaction to vary with atomic environment. In particular, this allows potentials using this formalism to differentiate the magnetic interaction of different stable phases of magnetic elements such as the ferritic and austenitic phases of iron. This, in turn, allows for a more robust and realistic description of magnetism in polymorphic materials than was previously possible. The motivation for MagMEAM, including the insufficiency of magnetic pair potentials, is presented and the structure of the formalism is developed. A sample iron potential is developed using this formalism and shown to exceed the capabilities of existing magnetic pair potentials by simultaneously reproducing the magnetic energy of both martensite and austenite as well as the dynamic mechanical and magnetic properties of martensite. This newly designed formalism will allow for deeper explorations in the the complex interaction between different phases of polymorphic magnetic materials at the molecular dynamics scale.

Irradiation Damage Evolution Dependence on Misorientation Angle for Σ 5 Grain Boundary of Nb: An Atomistic Simulation-Based Study

Abstract Niobium as refractory element holds ability to arrest the primary radiation damage in reactor's fissionable conditions and can withstand high-temperature applications. We have inclined the investigation of irradiated Nb Σ 5 symmetric-tilt angled grain boundary (STGB) models at two high-angled grain boundary misorientation: 53.13 deg (Σ 5(2–10)/(120)) and 36.87 deg (Σ 5(3–10)/(130)), respectively. A hybrid of Ziegler–Biersack–Littmark (ZBL) and embedded atom method (EAM) potentials were superimposed to simulate radiation damage. Statistical averaging of the displacement cascades was conducted to study the dynamic evolution of the point defects and interstitial clusters at varying magnitudes of primary knock-on atom (PKA) energies, irradiation temperatures, and PKA directions. The irradiated grain bounary (GB) models were compared with an irradiated bulk Nb specimen, and the results of the study indicate that the irradiated Nb system with greater misorientation angle, i.e., Nb Σ 5 (ɵ = 53.13 deg) survived with lower number Frenkel pair defects as well as the population small-sized interstitial clusters. The point defect cluster analysis indicated the highest population of interstitial clusters survived in Nb STGB models were irradiated along <1 3 5> PKA direction and 100 keV recoil energies respectively.

General-purpose machine-learned potential for 16 elemental metals and their alloys

Machine-learned potentials (MLPs) have exhibited remarkable accuracy, yet the lack of general-purpose MLPs for a broad spectrum of elements and their alloys limits their applicability. Here, we present a promising approach for constructing a unified general-purpose MLP for numerous elements, demonstrated through a model (UNEP-v1) for 16 elemental metals and their alloys. To achieve a complete representation of the chemical space, we show, via principal component analysis and diverse test datasets, that employing one-component and two-component systems suffices. Our unified UNEP-v1 model exhibits superior performance across various physical properties compared to a widely used embedded-atom method potential, while maintaining remarkable efficiency. We demonstrate our approach’s effectiveness through reproducing experimentally observed chemical order and stable phases, and large-scale simulations of plasticity and primary radiation damage in MoTaVW alloys.

K-Means Clustering in Fingerprint-Based Configuration Selection for Fitting Interatomic Potentials.

In this study, we present a method for selecting an arbitrary number of distinct configurations from a larger data set by applying k-means clustering to atomistic configuration fingerprints based on the CrystalNN model and radial distribution function (RDF). This approach improves the accuracy of fitting classical molecular dynamics interatomic potentials to density functional theory (DFT) data for both energies and forces while requiring fewer configurations than random selection. We demonstrate this improvement by fitting an embedded-atom method (EAM) potential for titanium, using various configurational sizes from an initial set of 1800 configurations. The k-means clustering consistently achieves better precision and lower standard deviations for a smaller number of configurations than random selection. The results also suggest that only about 30 configurations are sufficient to obtain an EAM model that describes well the full set of 1800 configurations in terms of energies and forces. Additionally, t-distributed stochastic neighbor embedding (t-SNE) method was used to reduce the configuration fingerprints into 2D space, and it revealed an overlap between two configuration subsets with and without Ti vacancy, indicating similar atomic environments. This similarity is captured by k-means clustering but not by random selection. Furthermore, when the overlapping configurations with vacancies were excluded from the k-means algorithm and used only as a test set, their energy and force predictions showed similar precision to those when they were included. This indicates that the overlapping configurations in the 2D t-SNE space indeed imply potential information redundancy among the atomistic configurations.

Comparative study of irradiation resistance for multicomponent concentrated HfNbTiZr and dilute V-4Cr-4Ti alloys irradiated with He ions

To clarify the efficiency of irradiation resistance, investigation of body-centered cubic concentrated HfNbTiZr and dilute V-4Cr-4Ti alloys, irradiated by 40 keV He ions up to 5 × 1016, 1 × 1017 and 5 × 1017 cm–2 fluences at room temperature, was carried out. Similar to V-4Cr-4Ti, HfNbTiZr possesses high phase stability and surface erosion resistance to irradiation with He ions up to 5 × 1017 cm-2. Using transmission electron microscopy, a more than 2-fold increase in overall swelling, as well as its intensification with increasing fluence was observed for HfNbTiZr compared to V-4Cr-4Ti. Combining atomistic calculations and simulations based on the Modified Embedded Atom Method interatomic potential and Density Functional Theory, the energetics of defects and helium-vacancy complexes, as well as their dynamics, were studied for alloys. It was shown that in the HfNbTiZr and dilute vanadium alloys the number of radiation-induced vacancies (v) can be comparable. According to the binding energy curves, there is a tendency for higher He accumulation in helium-vacancy complexes due to the increased He/v ratio in HfNbTiZr compared to V-4Cr-4Ti (∼1.5 versus ∼1.1). It was found that the kick-out of lattice atoms is enhanced in HfNbTiZr and is suppressed in V-4Cr-4Ti. Therefore, the more intense He bubble growth in HfNbTiZr may be due to the kick-out mechanism, which leads to a decrease in the He/v ratio and stimulates helium-vacancy complexes to trap additional He atoms. Our results can be used to improve the bubble swelling resistance in the design of new multicomponent concentrated alloys.

From electrons to phase diagrams with machine learning potentials using pyiron based automated workflows

We present a comprehensive and user-friendly framework built upon the pyiron integrated development environment (IDE), enabling researchers to perform the entire Machine Learning Potential (MLP) development cycle consisting of (i) creating systematic DFT databases, (ii) fitting the Density Functional Theory (DFT) data to empirical potentials or MLPs, and (iii) validating the potentials in a largely automatic approach. The power and performance of this framework are demonstrated for three conceptually very different classes of interatomic potentials: an empirical potential (embedded atom method - EAM), neural networks (high-dimensional neural network potentials - HDNNP) and expansions in basis sets (atomic cluster expansion - ACE). As an advanced example for validation and application, we show the computation of a binary composition-temperature phase diagram for Al-Li, a technologically important lightweight alloy system with applications in the aerospace industry.

Mechanical properties of hcp Fe at high pressures and temperatures from large-scale molecular dynamics simulations

We study the elastic behavior of hexagonal close-packed (hcp) Fe at the high temperature and pressure conditions of the Earth Core, using an embedded-atom method interatomic potential adjusted to those conditions. We calculate diffusivity, elastic constants, density, bulk modulus, shear modulus, and sound velocities vs temperature. We obtain reasonable agreement with ab initio simulations and with other empirical potential simulations. Our densities and shear modulus are slightly higher than those in the preliminary reference earth model for the core. Phase stability is discussed in terms of the Born criteria and free energies, finding that hcp is mechanically stable and that the free energy difference between hcp and body-centered cubic (bcc) is very small compared to the thermal energy. We compare our simulated shear modulus G to several analytical models, obtaining excellent agreement with the Atom in Jelium model by Swift and co-workers. Assuming that the yield strength Y is equal to the shear modulus G, Y=G/30, we find reasonable agreement with a recent parametrization of the Steinberg–Guinan model. These results can lead to future large-scale, multi-million simulations of Fe under core conditions for samples with microstructure like grain boundaries and twins, which might be present under those conditions.

The tensile behaviors of Ag-Cu alloys with different Cu contents: Molecular dynamics simulations study

The Ag-Cu alloys have received attention due to their excellent electrical and mechanical properties. However, the analytical research on why the mechanical properties of the Ag-Cu alloys increase after a small amount of Cu atoms are added is still lacked. In this work, the classical molecular dynamics (MD) simulations method was used to establish the Ag-Cu alloy models with different Cu contents. The tensile deformation mechanism of Ag-Cu alloy at room temperature was revealed from the atomic scale through the analysis of the stress-strain behavior and the evolution of dislocations. During the simulations, the embedded atom method (EAM) potential was used to describe the interaction between Ag atom and Cu atom. The uniaxial tensile deformation and nanostructure evolution of Ag-3 at.% Cu, Ag-6 at.% Cu and Ag-10 at.% Cu alloy models at the temperature of 300 K and the strain rate of 0.02 Å/ps were studied in detail. Results show that the internal stress was generated due to the large lattice mismatch among Ag atom, Cu cluster and the grain boundary. It results in the formation of dislocation and the negative initial stress at the beginning stage of Ag-Cu alloy models under tension. The dislocation firstly nucleates at the grain boundary and the growth rate of dislocation is slow before the crack propagation, and the complex dislocation reactions also exist. With the increase of Cu content, the tensile strengths of Ag-Cu alloys also increase. Cracks mainly form at the intersection of multiple grains, and the fracture mode changes from the intergranular fracture to the intergranular fracture and transgranular fracture after the unstable propagation of cracks. Meanwhile, a plenty of Cu atoms are distributed in the form of clusters in Ag matrix. They hinder the slip of dislocation and slow the relative sliding between atomic layers, and thus finally improves the deformation resistance to the plastic deformation of Ag-Cu alloy. The study on the tensile deformation behavior of Ag-Cu alloy at room temperature has certain guiding significance for the practical production and application of the design of Ag-Cu alloy.

Exact average many-body interatomic interaction model for random alloys

Understanding the physical origin of mechanisms in random alloys that lead to the formation of microstructures requires an understanding of their average behavior and, equally important, the role of local fluctuations around the average. Material properties of random alloys can be computed using direct simulations on random configurations. However, some properties are very difficult to compute, for others it is not even fully understood how to compute them using random sampling, in particular, interaction energies between multiple defects. To that end, we develop an atomistic model that does the averaging on the level of interatomic potentials. With such an average interatomic interaction model the problem of averaging via random sampling is bypassed since the problem of computing material properties on random configurations reduces to the problem of computing material properties on single crystals, the average alloy, which can be done using standard techniques.To be predictive, we develop our average model on the class of linear machine-learning interatomic potentials (MLIPs). To that end, using tools from higher-order statistics, we derive an analytic expansion of average many-body per-atom energies of linear MLIPs in terms of average tensor products of the feature vectors of the underlying machine-learning model that scales linearly with the size of an atomic neighborhood. In order to avoid forming higher-order tensors composed of products of feature vectors, we develop an implementation using equivariant tensor network (ETN) potentials, a class of linear MLIPs, that contracts the feature vectors to small-sized tensors, and then takes the average. We validate the average ETN potential by demonstrating the convergence of direct Monte Carlo simulations to the exact value for properties of the NbMoTaW medium-entropy alloy. Simulating average properties of dislocations has thus far only been possible with average embedded atom method potentials that, however, predict artificial polarized cores. Here, we show that average ETN potentials predict the compact screw dislocation core structure, in agreement with density functional theory. Hence, we anticipate that our model will become useful for understanding mechanistic origins of material properties and for developing predictive models of mechanical properties of random alloys.