Nonaffine Mean-Squared Displacement

• Nonaffine mean-squared displacement is a metric that measures local deviations from homogeneous, linear (affine) transformations in crystalline, amorphous, and granular systems.
• It reveals system-specific scaling laws and power-law behaviors that characterize defect nucleation, yielding events, and dynamical heterogeneity across thermal and athermal conditions.
• Robust measurement protocols, including local strain minimization and projection techniques, enable detailed mapping of elastic versus plastic deformations for material design.

Nonaffine mean-squared displacement (MSD) quantifies local deviations from affine (homogeneous, linear) transformations in crystalline and amorphous solids, granular media, and glasses. It is a central metric for probing microscopic rearrangements, plasticity, structural disorder, and critical dynamics in driven and thermalized many-body systems. Nonaffine MSD measures the residual particle or atomic motions after subtracting the best-fit local linear strain and rotation, providing insight into defect nucleation, yielding, jamming, and dynamical heterogeneity.

1. Mathematical Formulation and Definitions

Nonaffine displacement fields are constructed by decomposing particle or atomic motions into affine and nonaffine components. For a particle $i$ over a time interval $\Delta t$, its displacement is:

$$\Delta \mathbf{r}_i(t; \Delta t) = \mathbf{r}_i(t) - \mathbf{r}_i(t-\Delta t)$$

To isolate the nonaffine part in systems with flow or macroscopic deformation, the local affine motion, either as a spatially averaged field (granular flows) or a best-fit strain tensor (crystals/amorphous solids), is subtracted. Two principal definitions are used:

• Background-subtracted (flow-local) nonaffinity:

$$\Delta \mathbf{r}_{i,\mathrm{NA}}(t;\Delta t) = \Delta \mathbf{r}_i(t;\Delta t) - \Delta \mathbf{r}_{\mathrm{mean}}(\mathbf{x}_i; \Delta t)$$

where $\Delta \mathbf{r}_{\mathrm{mean}}$ is the local mean (affine) displacement field (Illing et al., 3 Feb 2025).

• Best-fit local (strain-minimized) nonaffinity:

Following the Falk–Langer D$^2_{\min}$ construction (Illing et al., 3 Feb 2025), for reference particle $0$ with $N_\mathrm{nbs}$ neighbors:

$$D^2_{\min}(t, \Delta t) = \sum_{n=1}^{N_\mathrm{nbs}} \sum_{i=1}^2 \left\{ r^i_n(t) - r^i_0(t) - \sum_{j=1}^2 [\delta_{ij} + \epsilon_{ij}] [r^j_n(t-\Delta t) - r^j_0(t-\Delta t)] \right\}^2$$

where $\epsilon_{ij}$ is the best-fit strain tensor. D$^2_{\min}$ is a local measurable with dimensions of length squared.

In coarse-grained crystal settings, nonaffinity is more generally cast as the minimum residual over a region $\Omega$:

$$\chi = \min_{D} \sum_{i\in\Omega} \left| [u_i-u_0] - D \cdot (R_i - R_0) \right|^2 = \Delta^T P \Delta$$

with $P$ the projection operator onto the nonaffine subspace and $\Delta$ the vector of relative displacements (Ganguly et al., 2012, Popli et al., 2019). The mean nonaffine MSD is then $\langle\chi\rangle = \textrm{Tr}[P C]$, with $C$ the displacement covariance.

2. Quantitative Scaling, Exponents, and Systematics

Extensive studies reveal system-specific, universal, and polydispersity-dependent scalings for the nonaffine mean-squared displacement:

• Crystalline Harmonic Solids:

For $d$-dimensional lattices with volume $|\Omega|$ and temperature $T$, the nonaffine mean $\langle\chi\rangle \propto k_B T |\Omega|$ with lattice-specific prefactors. In explicit cases (e.g. 2D triangular lattice, six nearest neighbors):

$\langle\chi\rangle = 6.865\,k_B T$

The spectrum is dominated by $d_\mathrm{dom}$ soft modes, associated with defect precursors and separated by a gap from harder modes: | Lattice | $d_\textrm{dom}$ | $\sigma_\textrm{dom}$ ($T=1$) | gap $\Delta\sigma$ | |--------------|:---------------:|:----------------------------:|:------------------:| | Triangular2D | 2 | 0.36 | 0.18 | | Square2D | 2 | 0.28 | 0.16 | | fcc3D | 3 | 0.30 | 0.15 | (Popli et al., 2019, Ganguly et al., 2012)

• Polydisperse Granular Flows:

Average r.m.s. nonaffine motion, $$\langle|\Delta r_\mathrm{NA}|\rangle/\langle R \rangle$$, monotonically decreases with particle size, with smallest particles showing $\sim$15–20% more than the largest. The mean $$\langle|\Delta r_\mathrm{NA}|\rangle/\langle R\rangle$$ remains almost constant with polydispersity $\delta$ (0.37 ± 0.01), but $\langle D^2_{\min}\rangle/\langle R\rangle^2$ grows linearly with $\delta$:

$$\langle D^2_{\min}\rangle/\langle R\rangle^2 = (0.11\pm0.01)\,\delta + (0.09\pm0.01)$$

Both nonaffinity metrics show power-law growth with time interval $\Delta t$:

$$\langle|\Delta r_\mathrm{NA}|\rangle/\langle R\rangle \sim (\Delta t/\Delta t_0)^{\alpha_1} \quad (\alpha_1 \simeq 0.74)$$

$$\langle D^2_{\min}\rangle/\langle R\rangle^2 \sim (\Delta t/\Delta t_0)^{\alpha_2} \quad (\alpha_2 \simeq 1.47 \simeq 2\alpha_1)$$

(Illing et al., 3 Feb 2025)

• Jamming and Athermal Compression:

As the packing fraction approaches jamming ($\phi\to\phi_J^-$), the nonaffine MSD diverges:

$$\Delta \equiv \langle|\delta r^{{NA}}|^2\rangle/\delta\ell^2 \sim (\delta\phi)^{-\beta},\quad \beta\simeq 2.7$$

and exhibits finite-size scaling, power-law displacement statistics, and fractal spatial localization with $d_f\simeq 2.04$ in $d=3$ (Ikeda et al., 2020).

• Amorphous Sheared Solids:

Under athermal quasistatic shear, the mean-squared nonaffine displacement scales with strain window $\Delta\gamma$:

$$\langle \Delta r_{na}^2\rangle \sim \begin{cases} (\Delta\gamma)^2 & \textrm{elastic} \ (\Delta\gamma)^\alpha,\,1.5\lesssim\alpha\lesssim 1.8 & \textrm{near yield} \ (\Delta\gamma)^1 & \textrm{steady flow} \ \end{cases}$$

The distribution $P(\Delta r_{na})$ is heavy-tailed, $P(\Delta r_{na})\propto (\Delta r_{na})^{-\tau}$ with strain-dependent $\tau$ between 2.4 and 4.8 (L et al., 2022).

3. Physical Interpretation, Defects, and Rearrangements

Nonaffine MSD encodes the propensity for localized, defect-related, or collective particle rearrangements:

• In compact crystals, the dominant nonaffine modes correspond to precursor configurations for lattice defects (e.g., $5$–$7$ dislocation pairs in 2D triangular lattices, stacking faults in fcc) (Popli et al., 2019). The statistics of nonaffine MSD reflect the underlying defect nucleation landscape.
• In amorphous materials, large nonaffine displacements cluster into shear transformation zones (STZs), mediate plastic instabilities, and, in the post-yield regime, evolve into system-spanning shear bands (L et al., 2022).
• For granular flows, nonaffinity is enhanced for small particles due to their inability to follow averaged directions set by large neighbors, creating short-range spikes in $\langle|\Delta r_\mathrm{NA}(d)|\rangle$. Large particles exhibit more affine collective motion, effectively 'screening' local flow fluctuations (Illing et al., 3 Feb 2025).
• Near jamming, the divergence of nonaffine MSD and concurrent localization on fractal sets indicate the approach to a rigidity transition, with scaling equivalence to viscosity divergence.

4. Measurement Protocols and Methodological Considerations

Robust quantification of nonaffine mean-squared displacement requires careful specification of reference frames, averaging procedures, and spatial or temporal coarse-graining:

• Reference frame selection: For steady-state flows or sheared systems, background subtraction (local mean flow) is necessary to exclude global drift.
• Best-fit locality: The number and arrangement of neighboring particles, $N_\text{nbs}$, or the size of the coarse-graining region $\Omega$, directly controls sensitivity to local structural disorder versus extended collective modes (Ganguly et al., 2012, Popli et al., 2019).
• Thermal versus athermal protocols: In thermal solids, harmonic fluctuations dominate nonaffine noise; in athermally driven (quasistatic) settings, plastic events and avalanches are primary contributors.
• Temporal and ensemble averaging: Nonaffine MSDs are typically averaged over steady-state cycles or over statistically independent samples; time-interval scaling reveals dynamic universality classes (ballistic, superdiffusive, diffusive).
• Polydispersity control: In granular media, size distributions characterized by polydispersity index $\delta$ are necessary for disentangling collective size effects from single-particle behavior (Illing et al., 3 Feb 2025).

5. Correlations, Criticality, and Distributions

Beyond the mean, the spatial and statistical correlations of nonaffine displacement provide deep insight into both collective dynamics and phase transitions:

• Spatial correlations:

In crystals and glasses, the two-point correlation function $$G_\chi(r)=\langle\chi(0)\chi(r)\rangle-\langle\chi\rangle^2$$ exhibits short-range, nearly isotropic exponential decay with a finite correlation length $\xi_\chi$, of a few lattice spacings (Ganguly et al., 2012). In amorphous materials, correlations in the elastic regime are exponential; post-yield, they transition abruptly to robust power laws $C_{D^2}(r)\sim r^{-\lambda}$ with $\lambda\sim1.3$, reflecting system-spanning avalanches (L et al., 2022).

• Critical fields:

Introduction of a conjugate field $h_\chi$ to the nonaffine measure induces a transition to a "maximally nonaffine" regime, with diverging mean and correlation length, mirroring elastic instabilities (e.g., Peierls) and amorphous melting (Ganguly et al., 2012).

• Distribution tails:

The distribution of single-particle nonaffine displacements $P(\Delta r_{na})$ is universally broad, with power-law tails whose exponents relate to underlying dynamical and geometric universality (fractal participation, power-law event sizes) (Ikeda et al., 2020, L et al., 2022).

6. Material, Geometric, and Flow Specificities

Distinct materials and protocols yield different nonaffine MSD responses, which can guide the identification of reorganizational mechanisms:

• Crystalline solids: Soft nonaffine eigenmodes are isolated and well-separated from the phonon background in close-packed lattices, justifying a defect-centric view of plasticity (Popli et al., 2019).
• Polydisperse granular media: Nonaffine activity is highly size-dependent, and the role of polydispersity is primarily in enhancing the frequency and localization of rearrangements, not the global amplitude.
• Amorphous solids under shear: Transitioning from exponential to power-law spatial correlations and from quadratic to linear strain scaling in nonaffine MSD signifies yielding and the breakdown of affine elasticity (L et al., 2022).
• Proximity to jamming: Nonaffine MSD is the critical order parameter, diverging with the same exponents as viscosity, and becomes spatially intermittent on fractal structures (Ikeda et al., 2020).

7. Implications and Applications

Nonaffine mean-squared displacement serves as a diagnostic and predictive tool across condensed matter, soft matter, and materials science:

• Modulation of polydispersity can be employed to tune mixing, enhance flowability, or reduce effective viscosity in granular and suspension flows (Illing et al., 3 Feb 2025).
• Nonaffinity-based measures such as $\langle D^2_{\min}\rangle$ are robust predictors of the frequency and localization of plastic rearrangements.
• The critical divergence of nonaffine fluctuations allows the identification of rigidity/jamming transitions and the characterization of the underlying glass or defect landscape (Ganguly et al., 2012, Ikeda et al., 2020).
• In simulations and experiments, projection and decomposition techniques furnish a detailed map of elastic versus plastic zones, supporting the design of materials with tailored mechanical response.

For all these reasons, nonaffine mean-squared displacement is now a foundational metric in the study of disordered, jammed, and driven non-equilibrium materials.