Computational materials researchers Jonas Ritter and Michael Zaiser have formulated a hybrid modeling framework that couples bond-based peridynamics with Mindlin-type gradient elasticity to resolve dislocation stress fields accurately from atomistic cores to macroscopic sample surfaces. Dislocation modeling presents an enduring multiscale challenge in mechanical physics because crystal defects produce intense, localized stress fields near their atomistic cores while simultaneously dictating long-range boundary value problems across finite specimens. In classical linear elasticity, these fields produce non-physical singularities directly at the dislocation line and at infinity unless regularized by non-local interactions and surface boundary constraints. By introducing distributed Burgers vectors into an integro-differential continuum representation, the investigators regularize core singularities and bridge the resulting non-local displacement fields with analytical gradient elasticity corrections, establishing a self-consistent continuum mechanics workflow.
Peridynamic theory, as initially formulated by Silling, eliminates spatial derivative requirements by expressing momentum balance through integral formulations of pair force densities acting across a finite neighborhood called the peridynamic horizon. Within bond-based formulations, continuous bodies are discretized into interacting material points whose pair forces depend directly on the relative stretch between interconnected bonds within an interaction sphere of radius delta. In an isotropic bulk material, pair forces scale linearly with bond stretch according to a constitutive micromodulus function that depends exclusively on initial bond length. When applied in the small-deformation limit, the strain energy density reduces to a quadratic functional of the displacement field, enabling analytical comparisons with classical Lamé constants where the effective Poisson ratio evaluates strictly to one-fourth under central-force bond assumptions.
To formalize the mathematical equivalence between non-local integral continuum mechanics and higher-order elasticity, Ritter and Zaiser expanded relative displacement fields into multidimensional Taylor series up to third-order displacement gradients. Separating radial integration over the interaction horizon from angular integrations over solid angles reveals that the zeroth-order energy term directly reproduces classical linear elasticity. Simultaneously, the second-order gradient expansion yields an elastic energy formulation conforming to Mindlin’s isotropic first strain gradient elasticity. The five characteristic strain gradient coefficients of Mindlin’s theory map directly to the peridynamic horizon radius and Lamé constants, demonstrating that the intrinsic length scale governing bond-based peridynamics inherently incorporates the higher-order constitutive features of strain gradient theories.
Incorporating crystal defects into this nonlocal integral framework required translating Volterra dislocation concepts into stress-free bond distortions governed by continuum dislocation theory. Because plastic distortion represents an incompatible, force-free geometric transformation, the authors evaluated modified bond vectors by integrating plastic distortion tensors along the straight line segments connecting interacting material points. When plastic slip occurs, bond vectors shift without generating internal pair forces, modifying both their rest length and their directional orientation. To preserve physical material isotropy and accurately track atomic neighborhood relations during localized slip, the spherical interaction horizon must be evaluated in the plastically deformed intermediate configuration rather than the material reference configuration, ensuring that material points convecting across slip boundaries are correctly updated.
The researchers investigated straight edge dislocations deforming in two-dimensional plane strain, first analyzing singular slip distributions where plastic distortion localizes exclusively onto a singular slip plane. When discretized, this singular Volterra-type jump causes bond vectors crossing the slip plane to add or subtract the Burgers vector depending on crossing orientation. Numerical relaxation demonstrated that although the nonlocal peridynamic integral successfully prevents the infinite stress singularities of classical linear elasticity, singular dislocation distributions retain a non-analytic discontinuity. Specifically, as numerical collocation grids are refined within the horizon, stress components exhibit an asymptotic step jump across the glide plane, which can produce severe numerical artifacts if smoothed with standard spatial interpolation techniques.
To eliminate this non-analytic behavior, the authors extended their methodology to dislocations characterized by distributed Burgers vectors, smearing plastic distortion smoothly across the core neighborhood. Two distinct continuous spatial profiles were analyzed: a uniform distribution spread evenly over a circular cross-section of horizon radius, and an analytical distribution derived from the modified Bessel function Green’s function of the two-dimensional Helmholtz operator. Dispersing the Burgers vector eliminates the step jump in core stress fields, producing completely smooth stress and strain distributions. Furthermore, numerical solutions demonstrated exact scaling properties shared across classical elasticity, gradient elasticity, and peridynamic simulations, in which normalized stress fields collapse onto invariant master curves across varying horizon lengths.
Quasi-static mechanical equilibria were computed numerically on a two-dimensional grid of 200 by 200 collocation points using explicit dynamic relaxation with adaptive damping based on Gershgorin’s circle theorem. To compare non-local discrete forces directly with classical Cauchy stress tensors, internal stresses were evaluated using the virial theorem across deformed bond neighborhoods. Calculations confirmed that when the Burgers vector distribution is matched to the Helmholtz Green’s function, the peridynamic stress profiles agree closely with analytical Mindlin and Helmholtz gradient elasticity solutions across the interior of the domain, confirming the analytical equivalence established in their strain gradient derivation.
Despite this interior agreement, bond-based peridynamics and analytical gradient elasticity display divergent physical behaviors near external specimen boundaries. In peridynamic simulations, traction-free conditions arise naturally at external surfaces without artificial constraints: normal stress components perpendicular to free boundaries vanish systematically, whereas tangential stress components remain non-zero, exactly fulfilling classical continuum boundary conditions. In contrast, closed-form gradient elasticity expressions derived for infinite bulk materials retain unphysical residual stresses along finite sample boundaries. This distinction highlights peridynamics as an ideal tool for resolving macroscopic boundary value problems in finite samples, while gradient elasticity provides efficient analytical descriptions of localized defect fields.
However, running peridynamic simulations with horizon radii scaled down to true atomistic lattice dimensions creates prohibitive computational costs that scale quadratically with local collocation density. Conversely, selecting a computationally convenient horizon much larger than the lattice parameter seriously underestimates peak stresses near the dislocation core by a scaling factor proportional to the Burgers vector divided by the horizon radius. To circumvent this limitation, the authors developed a three-part superposition strategy combining peridynamics with dual gradient elasticity corrections.
In this hybrid multiscale model, the total stress field is expressed as the numerical peridynamic stress field evaluated with an enlarged, computationally efficient horizon, plus an analytical gradient correction. The gradient correction represents the difference between an inner gradient elasticity field parameterized by an atomistic core length scale and an outer gradient elasticity field matching the expanded peridynamic horizon. Consequently, the correction vanishes completely beyond several horizon lengths, ensuring that long-range stress fields and surface boundary constraints are governed exclusively by the peridynamic solver, while atomistic core stresses are preserved.
The resulting hybrid multiscale formulation enables consistent mechanical modeling across disparate physical scales, allowing realistic dislocation evaluations in finite geometries containing free surfaces, microcracks, or external geometric boundaries. Ritter and Zaiser emphasize that because peridynamics handles spontaneous crack initiation, propagation, and dynamic branching naturally, coupling it with gradient elasticity establishes an efficient computational platform for investigating complex dynamic interactions between dislocations and evolving fracture processes.
Subject of Research: Multiscale continuum modeling of dislocation stress fields using peridynamics and gradient elasticity
Article Title: Multiscale modeling of dislocations: combining peridynamics with gradient elasticity
Article References: Ritter, J., & Zaiser, M. (2024). Multiscale modeling of dislocations: combining peridynamics with gradient elasticity. Journal of Materials Science: Materials Theory, 8(1), Article 2. https://doi.org/10.1186/s41313-024-00052-y
Image Credits: AI Generated
DOI: 10.1186/s41313-024-00052-y
Keywords: peridynamics, gradient elasticity, dislocations, multiscale modeling, edge dislocations, Burgers vector, plastic distortion, continuum mechanics, virial stress, Mindlin elasticity, Helmholtz equation, strain gradient, dynamic relaxation, dislocation cores, material science, computational mechanics
Cite Scienmag News
Denise Maddox. (November 21, 2026). Hybrid Peridynamic Model Resolves Dislocation Stresses from Atomistic to Sample Scales. Scienmag. https://scienmag.com/hybrid-peridynamic-model-resolves-dislocation-stresses-from-atomistic-to-sample-scales/
Denise Maddox. "Hybrid Peridynamic Model Resolves Dislocation Stresses from Atomistic to Sample Scales." Scienmag, 21 November 2026, https://scienmag.com/hybrid-peridynamic-model-resolves-dislocation-stresses-from-atomistic-to-sample-scales/. Accessed 21 November 2026.
Denise Maddox. "Hybrid Peridynamic Model Resolves Dislocation Stresses from Atomistic to Sample Scales." Scienmag. November 21, 2026. https://scienmag.com/hybrid-peridynamic-model-resolves-dislocation-stresses-from-atomistic-to-sample-scales/








