🔢 mathematics

Nonlinear vibrations of Euler–Bernoulli beams on fractional viscoelastic foundations: energy stability, modal reduction and finite-element approximation

This paper establishes the global well-posedness and energy stability of a nonlinear Euler–Bernoulli beam on a fractional viscoelastic foundation, while developing an unconditionally stable finite-element scheme with optimal error estimates and a reduced-order model that quantifies the interplay between geometric nonlinearity and hereditary damping.

Original authors: Alvaro H. Salas

Published 2026-09-24

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Alvaro H. Salas

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Structures like railway tracks, pipelines, and the tiny mechanical arms inside modern sensors often rest on layers of material that are neither perfectly solid nor purely fluid. These supporting layers, which might be soil, polymer composites, or biological tissues, have a peculiar quality: they do not react to a push or a pull the way a simple spring does. Instead, their resistance depends on how long they have been deformed and how fast they were deformed in the past. This "memory" of the material creates a complex damping effect that slows down vibrations in ways that standard physics models struggle to capture accurately. When engineers try to predict how a beam will vibrate under these conditions, especially if the vibrations are large enough to change the stiffness of the foundation itself, the math becomes incredibly difficult. The challenge lies in balancing the beam's own bending forces, the nonlinear resistance of the ground, and this long-term memory of the material, all while ensuring the computer simulations do not produce impossible results like energy appearing out of nowhere.

In a recent study, researchers developed a new way to model and simulate these complex vibrations, focusing on a classic type of beam known as an Euler–Bernoulli beam. This model is widely used because it accurately describes how long, thin structures bend when the movements are relatively small compared to their length. The team specifically looked at beams resting on a foundation that combines a standard spring-like force with a cubic force that gets stronger as the beam bends further, along with a memory term that accounts for the material's history. To handle the memory aspect, they used a mathematical tool called a fractional derivative, which acts like a weighted average of all past movements, giving more importance to recent events while still remembering the distant past. The researchers proved that their model is mathematically sound, meaning it always produces a unique solution and that the total energy of the system behaves correctly, always decreasing or staying constant due to the damping, never increasing spontaneously.

To test this theory, the team created a computer method that breaks the beam into small segments and calculates the forces at each point over time. They used a specific type of mathematical shape, known as a cubic polynomial, to represent the curve of the beam between these points, ensuring the simulation was smooth and continuous. By running these calculations step-by-step, they showed that their method remains stable even when the time steps are large, a crucial feature for long simulations. They verified the accuracy of their approach by comparing the computer results against a known mathematical solution, finding that the error decreased predictably as they made the segments smaller and the time steps shorter. The simulations confirmed that the memory of the foundation material significantly changes how vibrations die out; different settings for the memory strength altered both how quickly the beam stopped shaking and the timing of its movements.

The study also explored what happens when the foundation becomes "hardening," meaning it gets stiffer as the beam bends more. In these scenarios, the researchers observed that the beam's vibration patterns changed shape, a phenomenon known as a hardening phase portrait, where the frequency of the swing shifts as the amplitude grows. Furthermore, they simplified the complex beam model into a single, representative oscillator to see how it would respond to a rhythmic external force. This simplified model revealed that the memory of the foundation does more than just slow the beam down; it also shifts the frequency at which the beam resonates, or vibrates most strongly. Depending on the memory settings, the peak of this resonance could move, and the range of frequencies that cause large vibrations could widen or narrow. The researchers found that for certain memory settings, the system could have multiple possible vibration amplitudes for the same driving force, a behavior typical of nonlinear systems.

Ultimately, this work provides a reliable bridge between the complex physics of materials with memory and the practical tools engineers use to design safer structures. The team demonstrated that by treating the memory force as a hereditary effect rather than a simple friction, they could derive precise rules for how energy dissipates. Their computer code, which is available for others to use, successfully reproduces these physical behaviors, showing that the beam's energy remains stable and that the vibrations attenuate in a way that depends on the specific memory properties of the foundation. The findings suggest that ignoring the fractional memory of the supporting material could lead to inaccurate predictions of how structures behave under stress, particularly when those structures are subjected to large, nonlinear movements. By confirming that their mathematical framework holds up under rigorous testing, the researchers offer a solid foundation for future designs involving complex, memory-dependent materials.

Technical Summary: Nonlinear Vibrations of Euler–Bernoulli Beams on Fractional Viscoelastic Foundations

Problem Statement
The paper addresses the modeling and analysis of nonlinear Euler–Bernoulli beams resting on a viscoelastic foundation with fractional memory. The foundation reaction combines linear and cubic restoring forces with a Caputo fractional derivative of order $0 < \alpha < 1$ , representing a hereditary damping mechanism. This model interpolates between long-memory damping and classical viscous limits. The study focuses on the simply supported boundary value problem, aiming to establish a rigorous analytical framework for existence and stability, develop structure-preserving numerical approximations, and connect the distributed partial differential equation (PDE) model to reduced-order nonlinear oscillator dynamics.

Methodology
The research employs a unified approach spanning continuous analysis, finite element discretization, and modal reduction:

Continuous Analysis: A weak formulation is established in the space $H^2(0, L) \cap H^1_0(0, L)$ . The core of the stability analysis relies on identifying the positivity of the weakly singular convolution kernel associated with the Caputo derivative. This allows for the derivation of an exact integrated energy identity and global a priori bounds.
Well-Posedness: Global weak existence and uniqueness are proven using a Galerkin argument combined with compactness techniques (Aubin–Lions–Simon lemma), tailored to handle the cubic nonlinearity of the foundation.
Numerical Discretization:
- Spatial: Conforming cubic Hermite finite elements are used to approximate the fourth-order beam operator, ensuring the necessary $C^1$ continuity.
- Temporal: A backward-Euler scheme is coupled with the L1 convolution formula to discretize the Caputo derivative.
- Stability: The fully discrete scheme is designed to inherit a discrete memory-positivity property, ensuring unconditional energy stability without Courant–Friedrichs–Lewy (CFL) restrictions.
Modal Reduction: A one-mode projection is performed to derive a fractional Duffing oscillator equation. Harmonic balance analysis is applied to this reduced model to quantify the influence of geometric/foundation nonlinearity and hereditary damping on resonance.

Key Contributions
The paper makes the following specific contributions:

Energy Stability: It identifies a positive memory quadratic form for the Caputo term, yielding an exact integrated energy identity and global stability bounds for hardening foundations ( $k_3 \ge 0$ ).
Well-Posedness: It establishes global weak existence and uniqueness for the coupled beam-foundation system under standard smoothness assumptions.
Structure-Preserving Numerics: It proposes a fully discrete method (cubic Hermite elements + backward-Euler–L1) that is unconditionally energy stable. The method admits an energy-norm error estimate of order $O(h^2 + \tau)$ under smoothness assumptions.
Reduced Dynamics: It derives a fractional Duffing equation from the first simply supported mode, providing an explicit harmonic-balance amplitude equation that separates the storage and loss components of the fractional impedance.
Reproducibility: The study includes a manufactured solution test verifying the predicted second-order spatial convergence rate and provides Python code to regenerate all numerical results.

Results
Numerical experiments and theoretical analysis yield the following results:

Convergence: Manufactured-solution tests confirm that the $H^2$ seminorm error converges at the predicted second-order rate ( $O(h^2)$ ) when the time step is reduced proportionally to $h^2$ .
Memory Effects: Simulations demonstrate that the fractional order $\alpha$ significantly influences transient attenuation and phase. Increasing $\alpha$ toward the viscous limit ( $\alpha \to 1$ ) generally strengthens attenuation over finite intervals for the tested parameters.
Energy Stability: The discrete energy remains bounded and non-increasing for the unforced problem, with the gap between total energy and initial energy accounting for algorithmic dissipation inherent in the first-order time stepping.
Nonlinear Behavior: Phase portraits illustrate the hardening effect of the cubic foundation coefficient ( $k_3$ ), showing distorted trajectories and frequency shifts.
Resonance: The harmonic-balance analysis of the reduced model reveals that the fractional order shifts the location and width of resonance branches, with the fractional impedance exhibiting both real (storage) and imaginary (loss) components.

Significance and Scope
The paper claims to provide a "reproducible bridge" between hereditary beam mechanics, nonlinear vibration theory, and structure-preserving computation. Its primary significance lies not in introducing a new fractional derivative, but in deriving a coherent framework where the constitutive memory law, stability estimates, finite-element discretization, and reduced nonlinear dynamics remain mutually consistent.

The work emphasizes that the fractional term should be viewed as a hereditary foundation force encoded in a positive space-time quadratic form, rather than a simple pointwise viscous damper. It clarifies that the distributed PDE model and the fractional Duffing oscillator are complementary descriptions: the PDE provides boundary conditions and spatial modes, while the reduced oscillator offers rapid frequency-response diagnostics. The study is modest in its scope, focusing on hardening foundations to ensure global coercivity and noting that softening laws require additional stabilization or amplitude restrictions. The proposed numerical scheme prioritizes robustness and explicit energy inequalities over high-order temporal accuracy, serving as a stable baseline for further development.

More like this