On the quasi-continuum approximation of some localized patterns in the FPUT lattice

Technical Summary: Quasi-Continuum Approximation of Localized Patterns in the FPUT Lattice

Problem Statement
The Fermi–Pasta–Ulam–Tsingou (FPUT) lattice is a fundamental discrete model for nonlinear wave phenomena. While various wave patterns, such as solitons and dispersive shock waves, are well-understood in this context, the emergence of extreme events like rogue waves (freak waves) and their behavior on non-homogeneous backgrounds within the FPUT lattice remains an area of active investigation. Specifically, there is a need to determine whether the rich variety of exact analytical solutions available in continuous quasi-continuum models (such as the modified Korteweg-de Vries or mKdV equation) can be effectively mapped to the discrete FPUT lattice to predict and approximate complex localized wave structures, including those superposed on periodic traveling-wave backgrounds.

Methodology
The authors employ a multi-scale asymptotic analysis to derive a modified KdV (mKdV) equation from the discrete FPUT lattice equations.

1. Model Derivation: Starting with the standard FPUT lattice with a potential $V'(x) = K_2x + K_3x^2 + K_4x^3$, the authors introduce a strain variable and assume a plane-wave ansatz. By focusing on the $\beta$-FPUT model ($K_3=0$) and applying a multi-scale change of variables ($X = \epsilon(n - c_s t)$, $T = \epsilon^3 t$), they derive the mKdV equation at order $O(\epsilon^5)$. The resulting equation is $Y_T + \frac{3K_4}{2c_s}Y^2 Y_X + \frac{K_2}{24c_s}Y_{XXX} = 0$, which can be focusing or defocusing depending on the signs of $K_2$ and $K_4$.
2. Initial Condition Construction: The core methodology involves utilizing exact analytical solutions of the derived mKdV equation to construct initial conditions for the discrete FPUT lattice. The strain $y_n$ and velocity $s_n$ are initialized based on the mKdV solution $Y(X,T)$ and its derivatives, scaled by the small parameter $\epsilon$.
3. Numerical Simulation: The FPUT lattice equations are reformulated as a first-order system and integrated in time using a standard fourth-order Runge-Kutta (RK4) scheme on a sufficiently wide computational domain to minimize boundary effects.
4. Validation: The numerical evolution of the FPUT lattice is compared against the analytical mKdV solutions at specific time snapshots. The accuracy is quantified using the relative $L_\infty$ error, measuring the deviation between the numerical lattice solution and the scaled analytical solution.

Key Contributions and Results
The paper demonstrates the feasibility of using mKdV reductions to approximate localized wave patterns in the FPUT lattice across two distinct background scenarios:

• Homogeneous Backgrounds:

  • Second- and Third-Order Rational Solutions: The authors investigate rational solutions of the mKdV equation, which represent rogue-wave profiles. They show that initializing the FPUT lattice with second-order ($Y_2$) and third-order ($Y_3$) rational solutions yields numerical dynamics that closely match the analytical predictions.
  • Quantitative Agreement: For small $\epsilon$ (e.g., 0.05), the relative $L_\infty$ error remains bounded (not exceeding ~20% for the third-order case). The error peaks when the wave amplitude is maximal, and it increases with the order of the rational solution, yet the qualitative and semi-quantitative agreement confirms the validity of the approximation.

• Periodic Traveling-Wave Backgrounds:

  • Dark Breathers on Snoidal Backgrounds: In the defocusing regime, the authors analyze dark breathers on a sn-wave background. Numerical results show excellent alignment with the analytical solution, attributed to the modulational stability of the sn-wave background. The error decreases as the elliptic modulus $k$ increases.
  • Solitons on Dnoidal Backgrounds: In the focusing regime, solitons placed on a dn-wave background are examined. The numerical dynamics fit the analytical solutions very well, consistent with the known modulational stability of dnoidal waves.
  • Rogue Waves on Cnoidal Backgrounds: The authors explore rogue waves superposed on a cnoidal-wave background. While a qualitative analogy is clear, the numerical maximum amplitudes are consistently lower than the analytical predictions. The authors attribute this deviation to the modulational instability inherent to cnoidal waves, noting that this is a nontrivial limitation for long-time dynamics due to boundary effects in discrete lattices.

Significance and Claims
The paper claims that its primary contribution is establishing a bridge between the exact solutions of the continuous mKdV equation and the discrete dynamics of the FPUT lattice.

• Quasi-Continuum Approximation: The work validates the "quasi-continuum" approach, demonstrating that analytical waveforms derived from the mKdV reduction can serve as consistent initial conditions to observe specific localized patterns (solitons, breathers, and rogue waves) in the FPUT lattice.
• Extension to Periodic Backgrounds: Unlike previous studies that focused primarily on homogeneous backgrounds, this work extends the analysis to localized waves on periodic traveling-wave backgrounds (sn, dn, and cn waves), a scenario previously unprobed in the context of FPUT lattice dynamics.
• Physical Relevance: The authors suggest that these findings are particularly timely for mechanical metamaterials described by FPUT-type systems, where such wave patterns could potentially be realized experimentally.
• Future Directions: The paper modestly outlines future work, including the investigation of higher-order rational solutions, the extension to dimer FPUT lattices, and the application of similar asymptotic reductions to two-dimensional FPUT lattices (leading to Kadomtsev–Petviashvili reductions). It also notes the potential for rigorous proofs of existence and stability by connecting FPUT solitons to KdV solitons.

In summary, the paper provides a robust numerical and theoretical framework for importing complex waveforms from continuous models into discrete lattices, highlighting both the successes of the approximation and the specific limitations arising from lattice discreteness and modulational instability.