Computational materials researchers have developed a three-dimensional discrete dislocation dynamics model that explicitly simulates the physical transmission of crystal dislocations through planar grain boundaries. The framework, introduced by researchers M. Stricker and D. Weygand and published in the Journal of Materials Science: Materials Theory, overcomes long-standing constraints in microstructural plasticity simulations by replacing purely effective stress superpositions with an explicit physical transmission mechanism. In metallic engineering components, macroscopic plastic deformation and flow stress are predominantly governed by grain size and the obstacle strength of internal interfaces, traditionally approximated through empirical relationships such as the Hall-Petch scaling law where flow stress scales inversely with the square root of grain size. As dislocations move through a crystal lattice under applied mechanical load, their travel is obstructed at grain boundaries, causing dislocations to pile up and multiply stress concentrations at the interface by the number of accumulated defect lines. The newly reported numerical model resolves how these localized stress concentrations drive lattice dislocations across crystal interfaces into neighboring grains, integrating geometric and energetic criteria to determine outgoing slip paths while maintaining stress transparency, displacement discontinuity, and interfacial topological continuity.
Dislocation-grain boundary interactions involve complex physical processes including obstruction, partial transmission, absorption, emission, and the generation of residual interfacial defects. In earlier computational approaches, polycrystal simulations often relied on effective transmission models that superimposed the stress and displacement fields of dislocation pile-ups on opposing sides of a boundary without permitting actual defect penetration. While effective models produce stress fields that appear indistinguishable from physical penetration under balanced conditions, they fail in configurations featuring high dislocation density contrasts, such as an interface bounded by one heavily deformed grain and one initially dislocation-free grain. In those asymmetric configurations, effective representations cannot generate the necessary balancing pile-up on the dislocation-free side, leading to unphysical mechanical predictions. Existing explicit formulations within discrete dislocation dynamics frameworks have suffered from other severe operational limits: they were either confined to two-dimensional formulations, restricted to special cases with zero or very low crystallographic misorientations, or dependent on hybrid coupling with atomistic molecular dynamics simulations that are computationally prohibitive at the mesoscale.
To establish a general three-dimensional transmission formulation for arbitrary planar interfaces, the researchers adopted three sequential event criteria originally proposed by Lee and coworkers to govern incoming defect selection and predict the active outgoing slip system. When an incoming dislocation approaches a flat grain boundary, its trajectory aligns along the interface plane, forming segments that become candidates for transmission once their aligned length exceeds a defined critical threshold length between one hundred and two hundred lattice constants, provided they lie within a critical distance of the order of the Burgers vector. For pile-up configurations, testing is restricted strictly to the leading dislocation. The outgoing slip plane and outgoing Burgers vector in the adjacent crystal are selected by minimizing the crystallographic angle between incoming and outgoing glide plane normals and minimizing the residual Burgers vector using the L2 norm. In cases where these geometrical criteria do not identify a unique crystallographic solution, the model evaluates the local stress tensor and selects the candidate outgoing slip system that maximizes the resolved shear stress induced by the approaching defect structure.
The central structural innovation of the model is the introduction of an additional grain boundary dislocation constructed in a butterfly-shaped geometry at the interface between the two grains. Rather than transferring dislocation segments directly across the boundary or sharing nodes across mismatched crystalline lattices, the model inserts an independent interface dislocation loop whose nodes and segments remain numerically independent of both incoming and outgoing lattice defects. End-nodes along the interface reactions are treated analogously to dislocation junctions and are constrained to move only along the junction line direction. The Burgers vector of this butterfly-shaped interface dislocation is defined as half the vector sum of the incoming and outgoing Burgers vectors. Superimposing this butterfly defect onto the incoming dislocation creates a residual Burgers vector equal to half of the vector difference between the incoming and outgoing Burgers vectors in the first grain, with an equal and opposite residual contribution produced on the outgoing plane. Closing segments of the butterfly loop possess the average Burgers vector, ensuring full topological closure and the correct displacement discontinuity across the interface.
Partitioning the residual Burgers vector across both sides of the interface provides substantial energetic advantages over earlier models. Because elastic self-energy scales quadratically with the magnitude of the Burgers vector, splitting the total Burgers vector mismatch into two balanced residual traces of half-magnitude reduces the excess elastic energy of the residual interfacial traces by half compared to retaining the full mismatch on a single boundary side. Furthermore, decomposing the butterfly Burgers vector into normal and tangential components relative to the boundary plane provides a direct physical interpretation: the tangential component acts as a glissile interfacial dislocation facilitating strain transmission via boundary misfit traces, whereas the normal sessile component acts as a combined vacancy-interstitial defect that alternates localized tension and compression. This alternating stress state facilitates interface relaxation and models the physical step produced along the intersection trace, effectively replicating key atomic-scale phenomena within a continuum mesoscale framework.
The mathematical formulation was embedded into an established discrete dislocation dynamics framework wherein elastic boundary conditions are solved over a global time step and the underlying dislocation network evolves dynamically using sub-time steps between one-tenth and one-half of a picosecond. Transmission evaluations occur once per global time step. Because general high-angle grain boundaries exhibit atomic restructuring capable of screening localized interface stress concentrations, the simulation treats interface junctions as possessing screened, effectively vanishing far-field stress fields, whereas the connecting grain boundary dislocation links retain full stress field interactions. This treatment lowers the required external stress for transmission at a fixed critical segment length or permits transmission once aligned segments reach sufficient span. Dynamic stability of the resulting transmission loop is verified through nodal force calculations, ensuring that a nucleated test loop expands outward only when local driving stresses remain mechanically favorable.
Validation commenced with a baseline bicrystal test case containing an artificial planar boundary with zero crystallographic misorientation, loaded in shear along the coordinate axis. The left crystal grain was initialized with an operating edge-type Frank-Read dislocation source while the right crystal grain was initialized completely dislocation-free. Under applied shear, the dislocation bowed out, encountered the artificial boundary, and transmitted through without geometric impediment. Because both crystals shared identical orientations, outgoing slip system minimization yielded the identical incoming glide system. The resulting grain boundary dislocation topologically collapsed into a flat configuration with zero-length closing segments. Macroscopically, the simulated bicrystal exhibited saw-tooth stress-strain behavior and cyclic dislocation density variations that precisely replicated single-crystal deformation without work hardening or spurious dislocation accumulation, proving that coupling through independent interface nodes and stress fields preserves continuity across high density contrasts.
The authors next investigated a general bicrystal featuring a misorientation of approximately thirty-eight degrees around a rotation axis near minus one, nine, minus five, incorporating both tilt and twist components. The left grain contained an edge-type Frank-Read source while the right grain remained dislocation-free, with crystal orientations aligned so the transmitted defect experienced slightly lower resolved shear stress than the incoming defect. Upon reaching the interface, the incoming dislocation triggered the construction of the butterfly interface dislocation and nucleated an outgoing loop on the predicted slip system. Analysis of stress and displacement tensor fields evaluated ten lattice constants away from the interface in the receiving grain demonstrated significant mutual cancellation of stresses along the boundary plane, confirming stress transparency. The stress-strain response revealed pronounced work hardening after the first transmission event, caused by backstresses exerted by the residual grain boundary dislocation that resisted subsequent emissions until applied stresses rose sufficiently to drive further transmission.
To evaluate numerical scalability across complex polycrystalline microstructures, the framework was deployed on a three-dimensional non-periodic cuboidal sample of eight micrometers on each side containing forty randomly oriented face-centered cubic grains generated via a Voronoi tessellation. Each grain was seeded with eight randomly selected Frank-Read sources distributed evenly across all four available glide planes, producing an initial dislocation density of approximately one trillion defects per square meter. The polycrystal was subjected to uniaxial tensile deformation along the vertical axis at a constant strain rate of two thousand per second up to total strains of eighteen hundredths of a percent. The simulation contrasted an impenetrable grain boundary baseline against penetrable boundaries governed by the butterfly transmission model under parameter choices designed to maximize transmission and interfacial relaxation.
Under impenetrable boundary conditions, dislocations accumulated rapidly into dense pile-ups against grain interfaces, producing steep and sustained macroscopic work hardening. In contrast, incorporating the explicit butterfly transmission model prevented pile-up saturation by allowing defects to transmit across boundaries, forming dense networks of residual interface junctions and connecting links along grain facets. Facilitating transmission across interfaces caused the flow stress to level off and oscillate around forty megapascals, demonstrating substantial plastic softening. While this polycrystal test demonstrates the numerical stability and computational feasibility of tracking hundreds of coupled transmission events across multiple grains, the authors noted current structural limitations: boundaries must currently remain planar, interface motion operates through effective displacement superposition rather than true physical migration, and temperature-dependent processes such as diffusive climb and thermal activation remain omitted pending experimental calibration against microscale mechanical tests.
Subject of Research: Discrete dislocation dynamics modeling of grain boundary transmission
Article Title: A model for physical dislocation transmission through grain boundaries and its implementation in a discrete dislocation dynamics tool
Article References: Stricker, M., & Weygand, D. (2024). A model for physical dislocation transmission through grain boundaries and its implementation in a discrete dislocation dynamics tool. Journal of Materials Science: Materials Theory, 8(1), Article 12. https://doi.org/10.1186/s41313-024-00065-7
Image Credits: AI Generated
DOI: 10.1186/s41313-024-00065-7
Keywords: Discrete Dislocation Dynamics, Grain Boundaries, Dislocation Transmission, Polycrystal Plasticity, Burgers Vector, Slip Systems, Work Hardening, Mesoscale Modeling, Crystal Plasticity, Hall-Petch Relationship, Bicrystal Simulation, Resolved Shear Stress, Interface Dislocations, Computational Materials Science, Frank-Read Source, Dislocation Pile-Up
Cite Scienmag News
Denise Maddox. (November 21, 2026). Discrete Dislocation Dynamics Model Simulates Physical Dislocation Transmission Across Grain Boundaries. Scienmag. https://scienmag.com/discrete-dislocation-dynamics-model-simulates-physical-dislocation-transmission-across-grain-boundaries/
Denise Maddox. "Discrete Dislocation Dynamics Model Simulates Physical Dislocation Transmission Across Grain Boundaries." Scienmag, 21 November 2026, https://scienmag.com/discrete-dislocation-dynamics-model-simulates-physical-dislocation-transmission-across-grain-boundaries/. Accessed 21 November 2026.
Denise Maddox. "Discrete Dislocation Dynamics Model Simulates Physical Dislocation Transmission Across Grain Boundaries." Scienmag. November 21, 2026. https://scienmag.com/discrete-dislocation-dynamics-model-simulates-physical-dislocation-transmission-across-grain-boundaries/








