Table of Contents
Fetching ...

Seismic wave propagation in viscoelastic media under Atangana-Baleanu fractional dynamics: Model formulation and numerical simulations

Taylan Demir, Atakan Koçyiğit

TL;DR

The paper develops a 1D fractional viscoelastic wave model for seismic propagation using the Atangana–Baleanu–Caputo derivative with a non-singular Mittag–Leffler kernel to capture memory effects. It couples a constitutive law with an ABC derivative to yield a fractional wave equation, discretizes in space with centered finite differences, and advances in time via an Adams–Bashforth–Moulton predictor–corrector scheme. The main findings show that memory parameters $\alpha$ and $\kappa$ modulate attenuation and dispersion, producing non-exponential (Mittag–Leffler) energy decay and smoother wave-packet dispersion compared to the classical case. The work demonstrates a robust numerical framework implemented in Python suitable for extension to heterogeneous media and data-driven parameter estimation, offering a tractable path toward more realistic seismic attenuation models.

Abstract

We propose a one-dimensional viscoelastic seismic-wave model driven by the Atangana-BaleanuCaputo fractional derivative with a non-singular Mittag-Leffler kernel. A finite-difference discretization in space and an Adams-Bashforth-Moulton predictor-corrector scheme in time are used to compute solutions for several fractional orders. Simulations indicate that fractional memory alters both attenuation and dispersion, leading to non-exponential energy decay compared with the classical integer-order case.

Seismic wave propagation in viscoelastic media under Atangana-Baleanu fractional dynamics: Model formulation and numerical simulations

TL;DR

The paper develops a 1D fractional viscoelastic wave model for seismic propagation using the Atangana–Baleanu–Caputo derivative with a non-singular Mittag–Leffler kernel to capture memory effects. It couples a constitutive law with an ABC derivative to yield a fractional wave equation, discretizes in space with centered finite differences, and advances in time via an Adams–Bashforth–Moulton predictor–corrector scheme. The main findings show that memory parameters and modulate attenuation and dispersion, producing non-exponential (Mittag–Leffler) energy decay and smoother wave-packet dispersion compared to the classical case. The work demonstrates a robust numerical framework implemented in Python suitable for extension to heterogeneous media and data-driven parameter estimation, offering a tractable path toward more realistic seismic attenuation models.

Abstract

We propose a one-dimensional viscoelastic seismic-wave model driven by the Atangana-BaleanuCaputo fractional derivative with a non-singular Mittag-Leffler kernel. A finite-difference discretization in space and an Adams-Bashforth-Moulton predictor-corrector scheme in time are used to compute solutions for several fractional orders. Simulations indicate that fractional memory alters both attenuation and dispersion, leading to non-exponential energy decay compared with the classical integer-order case.

Paper Structure

This paper contains 7 sections, 17 equations, 4 figures.

Figures (4)

  • Figure 1: The modal amplitude is time dependent, where $A(t)=E_{\alpha}(-\lambda t^{\alpha})$ shows different behaviours with respect to the order of fractional decay $\alpha$. The case of $\alpha=1$ shows only an exponential decay. Both cases for $\alpha=0.8$ and $\alpha=0.6$ exhibit Mittag-Leffler type decay (or relaxation) but at a much slower rate (i.e., more heavily tailed), indicating that a fractional memory effect is contributing significantly to the behaviour of the decay.
  • Figure 2: Semi-logarithmic plots of modal energy $E(t)=\tfrac{1}{2}(A'(t)^2+\pi^2 A(t)^2)$ The energy behaves like a linear function when plotted with respect to the logarithm of time because it decays almost exactly in an exponential manner when $\alpha=1$. On the other hand, both curves for $\alpha=0.8$ and for $\alpha=0.6$ decay in a slower, non-exponential manner (Mittag-Leffler) due to the enhanced memory effects of the fractional order kernel.
  • Figure 3: The Space-time contour plot shows the way the displacement $w(x,t)=A(t)\sin(\pi x)$ behaves over the course of time for a representative fractional order (here $\alpha=0.8$). The fact that the endpoints are held fixed (i.e. tested using the condition $w(0,t)=w(1,t)=0$) is evident. The main feature of the interior is the formation and decay of a standing wave as the amplitude decays over time through the Mittag-Leffler type of the decay function generated by the Atangana-Baleanu fractional kernel.
  • Figure 4: Displacement, $w(x,t)$, shown in a 3D graph with values defined by the equation of a standing wave based on the fixed support conditions; Also shown is the amplitude of the wave progressively decreasing over time based on the fractional order differential equation created through the Mittag-Leffler type relaxation introduced by the Atangana-Baleanu fractional derivative.