Strain-Extended Continuum Models

Updated 4 February 2026

Strain-extended continuum models are advanced frameworks that add higher-order strain gradients and microstructural variables to classical mechanics.
They predict phenomena like strain localization, size effects, and dispersion by incorporating additional kinematic descriptors and energetic terms.
Applications span metals, semiconductors, graphene, polymers, and granular media, effectively bridging atomistic insights with continuum formulations.

A strain-extended continuum model is any theoretical or computational framework for continuum mechanics in which the constitutive response or kinematic description is systematically enriched to account for higher-order strain gradients, additional microstructural variables, or nonlocal effects. These models are essential in describing material behavior at small scales, in the presence of microstructure, or under conditions where classical (Cauchy) elasticity and plasticity fail to capture observed phenomena such as strain localization, dispersion, size effects, or the emergence of effective internal length scales.

1. Fundamental Kinematics and Governing Field Variables

Strain-extended continuum models generalize the classical kinematics by introducing additional field variables and higher spatial derivatives of primary kinematic or internal variables. The principal descriptors include:

Infinitesimal displacement field: $u(x, t)$ .
Macro-strain tensor: $\varepsilon = \nabla^s u$ , the symmetric part of the displacement gradient.
Generic internal variable: $\alpha(x, t)$ , which can represent quantities such as plastic strain, micro-distortion, or crystallinity parameters.
Gradient of internal variable: $\nabla \alpha$ , explicit in strain-gradient or micromorphic theories.
Eringen-type micromorphic continua: Introduce a two-field kinematics with independent “micro-strain” $\varepsilon_m$ , its gradient $\gamma = \nabla \varepsilon_m$ , and the relative strain $\delta(u, \varepsilon_m) = \varepsilon(u) - \varepsilon_m$ . The full set of kinematic measures is $\mathcal{E} = \{\delta, \varepsilon_m, \gamma\}$ . Length scales are introduced to preserve dimensional consistency and capture material size effects (McBride et al., 2018).

2. Energetic, Dissipative, and Constitutive Structures

The thermodynamic framework of strain-extended models encompasses a generalized Helmholtz free energy and a dissipation potential:

Total free energy: $\psi = \psi(\mathcal{E}, \mathcal{I})$ , where $\mathcal{E}$ gathers all kinematic measures and $\mathcal{I}$ stands for internal variables.
Typical gradient plasticity energy: $\psi = \psi(\varepsilon, \alpha, \nabla \alpha)$ .
Dissipation potential: $\phi = \phi(\dot{\alpha}, \nabla \dot{\alpha})$ , convex and positively homogeneous for rate-independent cases, regularized for rate-dependence (McBride et al., 2018).

The attachment of the energetic and dissipative components to the stresses and generalized forces is realized via first and higher spatial derivatives, promoting the emergence of length scales and nonlocal effects.

3. Microstresses, Stresses, and Conjugate Forces

Different classes of strain-extended continuum models introduce a spectrum of stress-like quantities:

Macro-stress: $\sigma = \partial \psi / \partial \varepsilon$ .
Microforces: $\xi = \partial \psi / \partial \alpha$ , $\tau = \partial \psi / \partial (\nabla \alpha)$ , often called double-stresses.
Energetic vs. Dissipative stresses: Defined via the Coleman–Noll procedure, leading to splits such as $S = S' + S''$ , where $S'$ is energetic and $S''$ dissipative; subdifferentials of the dissipation potential define dissipative microstresses (McBride et al., 2018).

In models such as strain-gradient elasticity, double stresses play a central role in higher-order boundary conditions and microscale traction balances (Giuseppe et al., 2017, Rosi et al., 2016).

4. Serial and Parallel Model Architectures, Flow Relations, and Yield Conditions

McBride, Reddy & Steinmann classify strain-extended plasticity and related models as either “serial” or “parallel”:

Serial (rheological series, classical plasticity):
- Additive strain split: $\varepsilon = \varepsilon^e + \alpha$ , energy depends on the elastic strain $\varepsilon - \alpha$ .
- Unique stress–strain relation at all times.
- Flow rules and yield conditions remain local: for rate-independent potentials, the normality rule applies pointwise in stress space.
Parallel (rheological parallel, rigid-plastic-like, or fully-dissipative):
- Energy may depend on total strain only; internal variables enter only via dissipation.
- Stress decomposed as $\sigma = \sigma'(\varepsilon) + \sigma''$ , with possible indeterminacy for non-smooth potentials (rigid-plastic plateaux).
- Local determination of the admissible stress region may be impossible for non-smooth φ; global (nonlocal) yield conditions are necessary (McBride et al., 2018).

Yield conditions and flow relations are constructed in generalized stress space including both energetic and dissipative stress contributions and often encompass normal cone inclusions for higher-order microstresses.

5. Prototypical Examples: Models and Physical Interpretations

(a) Strain-Gradient Elasticity and Wave Dispersion

Strain-gradient theories (Mindlin type II, SGE) generalize the classical energy by including terms quadratic in strain gradients:

$W(\varepsilon, \eta) = \tfrac{1}{2} C_{ijkl} \varepsilon_{ij} \varepsilon_{kl} + \tfrac{1}{2} A_{ijklmn} \eta_{ijk} \eta_{lmn}$

where $\varepsilon_{ij}$ is strain and $\eta_{ijk} = \varepsilon_{ij,k}$ is the strain gradient. The governing equations acquire higher spatial derivatives and boundary conditions involve both standard tractions and “double-forces” related to $\tau_{ijk}$ . This framework quantitatively predicts dispersion and size effects in architectured materials for wavelengths down to 5–7 times the unit cell (Giuseppe et al., 2017, Rosi et al., 2016).

(b) Strain-Gradient Plasticity and Micromorphic Models

Micromorphic regularizations recast strain-gradient plasticity so that the plastic strain (or equivalent slip) is treated as an independent field, energetically penalized if it departs from the macro-strain. Classification into “hybrid,” “combined energetic-dissipative,” and “fully dissipative” micromorphic models clarifies the structure of microstresses and flow rules. The connection to physical dislocation microstructures is established via rigorous upscaling as a Γ-limit from discrete Volterra dislocations, yielding functional forms where the dislocation density (curl of plastic distortion) enters as an explicit penalization (Kupferman et al., 2023).

(c) Nonlocal Rational and High-Frequency Models

Methods such as nonlocal rational continuum enrichment derive continuum equations directly from atomistic dispersion, leading to exact operator forms (e.g., for a 1D lattice, wave operator is $\Delta_2 \partial_x^2$ with $\Delta_2$ a rational function of spatial derivatives). This ensures exact match of continuum and atomistic dispersion up to Brillouin zone edges, crucial in nanoscale beams, metamaterials, and high-frequency elasticity (Patra et al., 2017).

6. Applications Across Material Systems

Strain-extended continuum models are employed across a wide range of materials:

Metals and Crystalline Solids: To account for size-dependent plasticity, grain-boundary strengthening, and gradient effects near interfaces or in micron/submicron grains (Bayerschen et al., 2015, McBride et al., 2018).
Semiconductors: For strain engineering in nanowires and core-shell structures, continuum elasticity with spatially varying initial strains matches atomistic models except at interfaces (Grönqvist et al., 2010).
Graphene and 2D Materials: Bond-level strain variables inform gauge field theories beyond the continuum elasticity regime, essential for understanding local electronic structure in rippled membranes (Sloan et al., 2013).
Granular and Fibrous Media: Models incorporating strain (or strain-rate) gradients predict segregation and nonlocal flow in dense mixtures, as well as the transmission of forces in biopolymer gels exhibiting strong strain-stiffening or buckling softening (Liu et al., 2023, Wang et al., 2020).
Polymers and Soft Materials: Extended strain models encode strain-induced crystallization or strain-rate-sensitive viscoelasticity, with internal variables for network regularity or physics-informed data-driven stress decomposition (Aygün et al., 2020, Upadhyay et al., 2023).
Brittle Materials and Damage Mechanics: Fiber-bundle models with yield thresholds and statistical “renewal” of failing elements connect to power-law rheologies and aftershock decay in seismology (Nanjo, 2016).

7. Implications for Multiscale and Nonstandard Continuum Formulations

Modern computational and theoretical trends integrate strain-extended models as multiscale bridges:

From atomistics to continuum: The extended Irving–Kirkwood–Noll approach injects thermodynamic source and flux terms arising from thermostats or barostats directly into continuum balances, critical for consistent MD–continuum coupling (Giusteri et al., 2017).
Coupling and coarse-graining: Auxiliary degrees of freedom and nonstandard fluxes require explicit treatment in the mass, momentum, and energy balances of the macroscale model to faithfully represent fine-scale dynamics.

These extensions demand careful mathematical and algorithmic developments, such as treatment of nonlocal variational inequalities for yield surfaces, regularization of sharp interfaces, and robust imposition of higher-order boundary conditions.

References:

Dissipation-consistent modeling and classification: (McBride et al., 2018)
Strain-gradient elasticity and wave propagation: (Giuseppe et al., 2017, Rosi et al., 2016, Patra et al., 2017)
Plasticity and dislocation-based upscaling: (Kupferman et al., 2023, Bayerschen et al., 2015)
Heterogeneous and nanoscale elasticity: (Grönqvist et al., 2010, Sloan et al., 2013)
Multiscale coupling and auxiliary continuum balances: (Giusteri et al., 2017)
Soft matter and data-driven visco-hyperelasticity: (Upadhyay et al., 2023)
Granular segregation and biopolymer gels: (Liu et al., 2023, Wang et al., 2020)
Damage mechanics and brittle flow: (Nanjo, 2016)
Strain-induced crystallization: (Aygün et al., 2020)