---
title: 'Nonlinear Disordered Chains: Dynamics & Transport'
url: https://www.emergentmind.com/topics/nonlinear-disordered-chains
type: topic
---

# Nonlinear Disordered Chains: Dynamics & Transport

Nonlinear disordered chains are one-dimensional lattice systems in which both spatial inhomogeneity (disorder) and nonlinear dynamical interactions play central roles in determining excitation dynamics, chaos, transport, and thermalization. These chains, broadly typified by the disordered Klein–Gordon (KG) chain, discrete nonlinear Schrödinger (DNLS) chain, disordered Fermi–Pasta–Ulam–Tsingou (FPUT) models, and their granular or SSH (Su–Schrieffer–Heeger) variants, serve as paradigmatic platforms for the study of the interplay between Anderson localization and deterministic chaos induced by nonlinearity. This interplay gives rise to highly nontrivial regimes of energy diffusion, ultra-slow delocalization, and anomalous thermal and transport properties.

## 1. Fundamental Models and Disorder Structures

The core mathematical formulations underlying nonlinear disordered chains include:

- **Disordered Klein–Gordon chain**:
  $$
  H(q,p)=\sum_{x}\left[\frac{1}{2}p_x^2+\frac{1}{2}\omega_x^2q_x^2+U(q_x-q_{x+1})\right], \quad U(r)=\frac{g}{4}r^4,
  $$
  with $(q_x,p_x)$ and i.i.d. $\omega_x^2$ on-site disorder.

- **Discrete nonlinear Schrödinger (DNLS) chain**:
  $$
  H(\psi)=\sum_x\left[\omega_x|\psi_x|^2+\frac{g}{4}|\psi_x-\psi_{x+1}|^4\right],
  $$
  where $\psi_x$ are complex amplitudes, $\omega_x$ is the onsite disorder, and $g$ sets quartic nonlinearity [2502.02344].

Variations include the SSH chain with Kerr nonlinearity [2212.06235], Hertzian and FPUT lattices (continuous/discontinuous nonlinearities) [1811.10288], chains with structured (fractal/correlated) disorder [1701.09029, 1107.1114], and strong-disorder Fröhlich–Spencer–Wayne (FSW) models [1103.4034].

Typical disorder is implemented via random onsite frequencies or spring constants uniformly distributed in finite intervals. In some models, disorder is introduced geometrically or via correlated/fractional algorithms to probe different resonance scenarios.

## 2. Chaos Nucleation and Local Resonant Dynamics

In weakly nonlinear regimes, deterministic chaos is not spatially uniform but emerges from rare, local resonant structures—specifically, *resonant triples* or higher-order resonant clusters where Anderson-localized mode frequencies nearly coincide [1206.1149]. The density of chaotic sites (“chaotic fraction per site” $w(\varepsilon)$ for mean energy density $\varepsilon$) in such chains obeys universal scaling laws:
$$
w(\varepsilon)\simeq A^2\varepsilon^2,\qquad A\approx1.37\times10^3,
$$
leading to exponentially rare chaotic “seeds” at weak nonlinearity. These segments act as stochastic pumps, generating Arnold diffusion in surrounding oscillators, driving extremely slow relaxation and eventual delocalization [1005.5033, 1206.1149].

The typical distance between chaotic sites is
$$
\ell_r\sim\frac{1}{w(\varepsilon)}\sim\frac{1}{A^2\varepsilon^2},
$$
so Anderson localization persists robustly in finite chains at low energy, but can be statistically overcome for sufficiently large system sizes or higher nonlinearity [1206.1149, 1005.5033].

## 3. Wave Packet Spreading and Transport Laws

The time-evolution of initially localized energy or norm excitations exhibits radically modified spreading compared to linear Anderson systems:

- **Subdiffusive power-law spreading**:
  In typical KG/DNLS chains, the second moment $m_2(t)$, participation number $P(t)$, and higher-order moments display universal subdiffusion exponents [1312.5102, 2212.06235, 1103.4034]:

  $$
  m_2(t)\sim t^{1/3},\qquad P(t)\sim t^{1/6},
  $$
  corresponding to a compact, thermalized spreading front (compactness index $\zeta=P^2/m_2\approx 3$ indicates interior thermalization) [1103.4034].

- **Non-algebraic ultra-slow decay**:
  Recent rigorous results demonstrate that the local energy peak in infinite chains decays more slowly than any power law (sub-power–law) [2502.02344]. Explicitly, beyond a finite random $T^*$,

  $$
  M(t)\ge\exp\left[-2(\ln t)^{3/4}\right],
  $$
  so $M(t)>t^{-\alpha}$ for every $\alpha>0$ asymptotically, indicating ultra-slow Arnold-type diffusion. This rigorous result refutes any true asymptotic power-law decay of energy maxima in infinite chains [2502.02344].

- **Modified transport in strongly nonlinear or non-smooth chains**:
  In discontinuous nonlinear (e.g., Hertzian) models, gap propagation events can rapidly destroy localization at lower energies than in continuous-nonlinearity systems (FPUT), directly triggering energy equipartition. Smooth FPUT chains require higher energies for delocalization and can show re-entrant localization depending on excitation energy [1811.10288].

## 4. Kinetic Descriptions and Nonequilibrium Transport

Macroscopic transport equations emerge in various kinetic or hydrodynamic regimes:

- **Nonlinear diffusion**:
  In the regime of “homogeneous chaos” (when nonlinearity is weak enough not to shift frequencies but strong enough for chaos to be spatially dense), coarse-grained density $\rho(x,t)$ satisfies [1312.1351]:

  $$
  \partial_t\rho = \partial_x\left[D_0\rho^2\partial_x\rho\right],
  $$
  with $D_0$ explicitly determined by four-mode matrix elements and scaling as $D_0\sim(\Omega/W)^{4.9}$ for disorder bandwidth $W$ and hopping $\Omega$.

- **Thermal conductivity**:
  Disordered KG and FSW chains exhibit strong temperature dependence of the thermal conductivity. In the weak-chaos regime ($T\ll d$, $d$ mean mode spacing):

  $$
  \kappa\sim T^4 \quad (\text{weak chaos}),\qquad \kappa\sim T^2 \quad (\text{strong chaos}),
  $$
  with crossover governed by the effective nonlinear frequency shift and the spectrum of available resonances [1101.4530, 1911.03753].

- **Super-activated transport**:
  At very low temperature, thermal conductivity vanishes faster than any power of $T$:
  $$
  \kappa_\infty\sim\exp\left[-B|\ln(C\Delta/T)|^3\right],
  $$
  as predicted for transfer dominated by rare, high-order resonant clusters [1911.03753].

## 5. Thermalization, Ergodicity, and Long-Time Statistics

Thermalization properties exhibit a strong dependence on the structure of disorder and nonlinearity:

- **Grand-canonical site-norm distributions**:
  DNLS chains in strong-disorder regimes exhibit local equilibration to a grand-canonical distribution,
  $$
  P_n(I_n)=\frac{1}{Z_n}\exp\left[\beta(\mu-\epsilon_n)I_n-\frac{1}{2}\beta\chi I_n^2\right],
  $$
  where parameters are fixed by conservation of norm and energy [1107.1114].

- **Disorder correlations**:
  Spatial correlations in the disorder potential can dramatically accelerate thermalization, even for identical linear localization properties. Correlated disorder enhances resonant triple probabilities and speeds up norm/energy exchange between otherwise almost isolated localized modes [1107.1114].

- **Glassy and many-body localized regimes**:
  Chains with bounded (e.g. cosine) nonlinear interactions and carefully constructed fractal disorder can exhibit stretched-exponential relaxation and even energy trapping reminiscent of classical many-body localization at low temperature, marked by a freezing transition below a finite $T_f$ [1701.09029].

- **Persistence of chaos and central limit statistics**:
  Even after vastly long times ($t_{\max}\sim10^9$), chaos persists and no collapse to a quasi-periodic KAM torus is observed. The statistics of spatially local observables cross from non-Gaussian weak chaos to Gaussian strong chaos at long times [1312.5102].

## 6. Nonlinear Response and Topological Effects

Nonlinear transport in disordered and topological chains reveals unique phenomena beyond the linear response paradigm:

- **Nonlinear Landauer formula**:
  Expansion of the current yields
  $$
  G^{(n)} = \frac{e^{n+1}}{n!h}\left.\frac{d^{n-1}T}{dE^{n-1}}\right|_{E=\mu}, 
  $$
  with $T(E)$ the transmission. In 1D, all nonlinear conductances $G^{(n\ge2)}$ vanish exponentially with system size, except in chiral (sublattice-symmetric) models, where Dyson-type singularities yield algebraic decay and large nonlinear response at zero energy [2110.08304].

- **Effect of nonlinearity on topological markers**:
  In nonlinear disordered SSH chains, even weak Kerr nonlinearity suppresses the anomalous log–log diffusion that marks the topological phase transition in the linear limit and replaces it with universal subdiffusion ($m_2\sim t^{1/3}$), effectively erasing linear topological features from the spreading waveform [2212.06235].

## 7. Open Problems and Broader Implications

While the essential mechanisms—resonance-induced chaos, Arnold diffusion, subdiffusive delocalization, and the crucial role of rare events—are now well-established, several open questions remain. The microscopic origin of intermediate exponents observed in certain parameter regimes [1206.1149], the possibility of genuine ergodicity breaking or MBL-like transitions under classical bounded nonlinearities [1701.09029], and the full impact of disorder correlations or spatially inhomogeneous nonlinearity on the asymptotic dynamics continue to attract attention.

Research in nonlinear disordered chains has significant implications for understanding transport phenomena in optical lattices, cold atoms, granular metamaterials, and low-dimensional disordered solids, where both disorder and nonlinearity are ubiquitous. The confluence of mathematical rigor [2502.02344], kinetic theory [1312.1351], and large-scale simulation [1911.03753, 1103.4034] has established the field as a mature testing ground for fundamental questions in nonequilibrium statistical mechanics, dynamical systems, and condensed matter theory.

Source: https://www.emergentmind.com/topics/nonlinear-disordered-chains