Localized oscillations of an Euler--Bernoulli beam with time-varying parameters on a visco-elastic foundation: asymptotics, adiabatic invariant, and equivalent Hamiltonian system

arXiv:2606.22444 · math-ph, math.MP, physics.class-ph · Submitted 2026-06-21 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Mathematical Physics. The day in mathematical physics: rigorous models of physical systems.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Localized oscillations of an Euler--Bernoulli beam with time-varying parameters on a visco-elastic foundation".

Mira: Here's the gist of the paper. Localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator is studied using various analytic approaches that yield the same formula for the amplitude of oscillation in both conservative and dissipative cases.

Kai: Mathematical Formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by a specific equation involving time-varying parameters such as flexural stiffness and mass density.

Mira: The problem is formulated equivalently with zero right-hand side for spatial coordinates other than zero and four boundary conditions at the point x=0.

Kai: The total energy of the beam is defined as a sum of kinetic and potential energy terms involving the mass density, flexural stiffness, and foundation stiffness.

Paper summary: Kai: Looking at this whole piece again, the title itself tells you a lot about what they are tackling: localized oscillations of an Euler--Bernoulli beam with time-varying parameters on a visco-elastic foundation. It really summarizes the physical setup they built for their study.

Mira: And it's interesting how they bring in the concepts of asymptotics, adiabatic invariance, and the equivalent Hamiltonian system to analyze this problem. They show that even though these are different mathematical tools, for this specific setup, all three end up giving you that same formula for the amplitude of oscillation.

Lev: So what does that convergence mean for us on a practical level? It means we don't have to pick one method and hope it works; if we can find a physical system where these conditions hold—like slow variation of parameters—we get the same result regardless of which mathematical tool you choose to use.

Kai: Exactly. And they also confirm that for the non-dissipative case, they get a specific formula for the amplitude, and they also confirm that their Hamiltonian system method works even for complicated systems where standard asymptotic approaches are really difficult to apply.

Mira: And the important part is what it implies about how we look at these complex dynamics. It shows that finding an equivalent Hamiltonian system can be useful when applying a straightforward asymptotic approach is proving too hard. That's a big statement about the power of that framework for analysis.

Lev: From my side, it’s still about translation to hardware feasibility. If this mathematical framework gives us a simple way to find these modes, we need to ensure that the actual simulation or measurement overhead doesn't make it too slow for real-time feedback on those oscillations.

Kai: So the main thing is that they've provided these formulas, like the one for C zero at zero time, which they state is C0 equals sqrt(two) times p eight over k minus rho omega squared sub zero times three rho root four B plus root two M p four k minus rho omega squared sub zero times T equals zero.

Mira: Yeah, that specific formula for the amplitude, C0, that they derive, is what they use to describe this localized oscillation of the Euler--Bernoulli beam on a visco-elastic foundation and its coupling with a discrete oscillator. It’s their main takeaway result.

Lev: So it's essentially showing that for this specific problem, these different mathematical frameworks all lead to the same description of the system's response, which is what we need when building systems that have complex dynamics running on real hardware.

Conclusion: Kai: So we're wrapping up this paper on localized oscillations of an Euler--Bernoulli beam on a visco-elastic foundation with time-varying parameters and it’s all about how different math tools—asymptotics, adiabatic invariance, and the Hamiltonian system—all point to the same amplitude formula for those wiggles.

Mira: Yeah, the authors are showing that when you have properties changing slowly over time and you’re on a foundation with damping, these different approaches converge on one single description of how big those localized oscillations get.

Lev: What this means practically is that if you can find a physical system where those slow changes happen, like if you tune the beam stiffness over time in a controlled way, you don't have to worry about which complex math method to use; the result for that amplitude stays the same.

Kai: They really prove that finding an equivalent Hamiltonian system is useful when using those standard asymptotic methods gets too messy for complicated systems. It’s like having a backup plan when your primary calculation hits a wall.

Mira: And they also confirm that for the non-dissipative case, they get that specific amplitude formula involving the terms p eight over k minus rho omega squared sub zero times three rho root four B plus root two M p four k minus rho omega squared sub zero times T equals zero. That’s a concrete number describing those oscillations at large times.

Lev: So yeah, the convergence is solid for that specific setup, but we still have to figure out how much computational overhead it takes to actually run that equivalent Hamiltonian system on real hardware compared to just using the invariant formula they derived earlier.

Kai: That’s right, we need those numbers on how much processing power this requires before we can really talk about building something functional for a lab testbed.

Mira: It shows that understanding the underlying physical assumptions—like the slow variation and the adiabatic invariance—is just as important as finding the final amplitude formula itself.

Lev: Next up, we need to see what those real-world constraints are for simulating this kind of time-varying system in a real quantum or classical environment.

Institute for Problems in Mechanical Engineering RAS

math-ph, math.MP, physics.class-ph

Submitted: 2026-06-21

Updated: 2026-11-05

Comments: 15 pages, 1 figure

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 81/100

The gist: Localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator is studied using various analytic approaches that yield the same formula for…

Key concepts

Euler–Bernoulli beam on a visco-elastic foundation
This describes a beam structure whose stiffness and mass properties change with time, resting on a foundation that exhibits both elastic (spring-like) and viscous (damping) characteristics. The oscillation of the beam is influenced by these time-varying properties.
Adiabatic Invariant
This concept suggests that for certain systems, the action of a trapped wave remains constant even as the system parameters change slowly over time. This invariance allows researchers to find simple formulas for the oscillation amplitude by relating it to this conserved quantity.
Equivalent Hamiltonian System
This is a mathematical framework used to describe complex physical systems. Instead of solving the original difficult equations directly, this method reformulates the problem into a simpler system defined by generalized coordinates and impulses, making analysis easier.

Terminology

Summary

Localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator is studied using various analytic approaches that yield the same formula for the amplitude of oscillation in both conservative and dissipative cases. p. 0

Where the paper says it

3 4 Adiabatic invariant 5 5 Equivalent Hamiltonian system 6 6 Numerics 7 7 Conclusion 8 A The dispersion relation 8 B The Green function 9 C The spectral problem 9 D Inhomogeneous non-stationary problem 10 1 Introduction In this paper we apply recently developed in Gavrilov et al. (2024); Shishkina and Gavrilov (2026) approaches to the problem concerning localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator.

Mathematical Formulation

The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by a specific equation involving time-varying parameters such as flexural stiffness and mass density. The problem is formulated equivalently with zero right-hand side for spatial coordinates other than zero and four boundary conditions at the point x=0. p. 1

Where the paper says it

1) with zero right-hand side for x ̸= ±0: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = 0 (2. 2) and four boundary conditions at x = 0: [w] = 0, (2.

The total energy of the beam is defined as a sum of kinetic and potential energy terms involving the mass density, flexural stiffness, and foundation stiffness. p. 1
Where the paper says it

2 Mathematical formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by the following equation: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = p(t) − Mw˙ · − Kw − 2ϵΓw˙ δ(x). 1) Here w(x, t) is the displacement, B = B(ϵt) is the flexural stiffness, ρ = ρ(ϵt) is the mass density of the beam, ϵγ(ϵt) is the damping in the foundation, M = M(ϵt) is the value of mass in the discrete oscillator, K = K(ϵt) is the stiffness for the discrete oscillator, k = k(ϵt) is the stiffness for the Winkler foundation, ϵΓ(ϵt) is the damping in the oscillator, p(t) is the given external force on the beam, x is the spatial coordinate, t is time, ϵ is a formal small parameter.

Asymptotic Analysis

The authors employ an asymptotic ansatz for the solution under assumptions that include slow variation in time and smallness of a formal parameter epsilon. The first-order approximation equation derived by substituting the ansatz into the governing equation at x → 0 is given by a complex expression involving various system parameters. p. 3

Where the paper says it

8) into the governing equation (2. 2) at x → 0 and equating the coefficients of like powers ϵ, one obtains that to the first-order approximation: 4Bω3W0 ′ X + 6Bω2ω ′ XW0 + ρ 2Ω0W0 ′ T + Ω0 ′ T W0 + Ω0ρ ′ T W0 + 2γΩ0W0 = 0.

The resulting first-order approximations of the boundary conditions lead to a differential equation for W0. p. 3
Where the paper says it

2) at x → 0 and equating the coefficients of like powers ϵ, one obtains that to the first-order approximation: 4Bω3W0 ′ X + 6Bω2ω ′ XW0 + ρ 2Ω0W0 ′ T + Ω0 ′ T W0 + Ω0ρ ′ T W0 + 2γΩ0W0 = 0. 19) The corresponding first-order approximations of boundary conditions (2.

Adiabatic Invariant and Hamiltonian System

For the non-dissipative case, the paper demonstrates that the action of the trapped wave is an adiabatic invariant by postulating its constancy. This invariance leads to a formula for the amplitude proportional to a quantity involving system parameters and frequency terms. p. 0

Where the paper says it

Namely, these are asymptotics, the method based on the adiabatic invariance of the action of a trapped wave, and the consideration of the equivalent Hamiltonian system. All approaches result in the same formula for the amplitude of oscillation.

The equivalent Hamiltonian system is defined with a Hamiltonian H that depends on generalized coordinates Q and generalized impulse P. p. 5
Where the paper says it

2) is the corresponding Hamiltonian, Q = U, P = M dU dt (5. 3) are the generalized co-ordinate and the generalized impulse, respectively.

The evolution of the amplitude U can be found using the WKB approach within this framework. p. 5
Where the paper says it

5) are the effective mass and the effective stiffness, respectively; |C0| is defined by Eq. The evolution of the amplitude U can be found by the WKB approach.

Spectral Problem and Non-Stationary Response

The spectral problem for the unperturbed system yields a dispersion relation where the wave number omega is given by two roots, omega = ωδ = a(i + δ). The solution at x=0 for the amplitude W(0) is derived from this spectral analysis. p. 8

Where the paper says it

Put p = 0 and search the solution in the form of Eq. For the amplitude W(x) one obtains Eq.

For the inhomogeneous non-stationary problem with a vanishing external force as t → ∞, the asymptotic evaluation results in undamped oscillations with the trapped mode frequency Omega0. p. 2
Where the paper says it

We also require the frequency equation for the trapped mode and the dispersion relation at x = ±0 hold for all t. We use the following asymptotic ansatz: w(x, t) = H(T) X δ=±1 Wδ (X, T) exp ϕ δ + c.

The large time response is given by C(0) cos Ω0(0)t − arg C(0) + o(1), t → ∞. p. 4
Where the paper says it

12), one gets w(0, t) = 4|A|H(T)A(T) exp (−D(T)) cos π 4 cos(ψ − arg A) + O(ϵ) (3. 35) where ψ ≡ Z t 0 Ω0(ϵtˆ) dt.

Conclusion

The most important results are the formulas describing localized oscillation of an Euler-Bernoulli beam lying on the visco-elastic foundation and coupled with a discrete oscillator. The method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task. p. 7

Where the paper says it

(2024); Shishkina and Gavrilov (2026) for the analogous problem concerning localized oscillation of a string. This confirms that the method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task.

The paper confirms that the method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task. p. 7
Where the paper says it

(2024); Shishkina and Gavrilov (2026) for the analogous problem concerning localized oscillation of a string. This confirms that the method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task.

The non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Omega0. p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

The large times, the non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Omega0. p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

The final amplitude is given by C0(0) = sqrt(2)p8 / (k − ρomega20) (3ρ√4 B + √2M p4 k − ρomega20) T = 0. p. 4
Where the paper says it

where A(T) = p8 k − ρΩ2 q 0 Ω0 3ρ √4 B + √ 2M p4 k − ρΩ2 0 , (3. 32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT .

p. 4
Where the paper says it

where A(T) = p8 k − ρΩ2 q 0 Ω0 3ρ √4 B + √ 2M p4 k − ρΩ2 0 , (3. 32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT .

The dispersion relation

Here we consider the corresponding stationary unperturbed problem with constant parameters and zero dissipation. The dispersion relation for the operator in the left-hand side of the above equation is Bρω4 + Ω2∗ − Ω2 = 0. p. 7

Where the paper says it

1) one gets −Ω 2 ρW + BW′′′′ + kW = p(t) + MΩ 2W(0) − KW(0) δ(x). 3) The dispersion relation for the operator in the left-hand side of the above equation is B ρ ω 4 + Ω2 ∗ − Ω 2 = 0, (A.

The two roots of the dispersion given by expression are omega = ωδ = a(i + δ). p. 7
Where the paper says it

4) where ω is the wave-number. We treat the right part of the beam x > 0, therefore, we consider the two roots of the dispersion given by expression ω = ω δ = a(i + δ).

p. 7
Where the paper says it

4) where ω is the wave-number. We treat the right part of the beam x > 0, therefore, we consider the two roots of the dispersion given by expression ω = ω δ = a(i + δ).

p. 7
Where the paper says it

4) where ω is the wave-number. We treat the right part of the beam x > 0, therefore, we consider the two roots of the dispersion given by expression ω = ω δ = a(i + δ).

The Green function

Consider equation ρG¨ + BG′′′′ + kG = exp(−iomegat)δ(x) for x > 0. The expression for the Green function (B. p. 8

Where the paper says it

6), one has W = (1 − i)√ 2 8B1/4ρ 3/4(Ω2 ∗ − Ω2) 3/4 . 9) Expression for the Green function (B.

2) can be transformed to the following equivalent form G(x, Ω) = √28Ba3 exp(-ax) cos ax − π/4. p. 8
Where the paper says it

9) Expression for the Green function (B. 2) can be transformed to the following equivalent form: G(x, Ω) = √ 2 8Ba3 exp(−ax) cos ax − π 4 , (B.

The spectral problem

For the amplitude W(x), one obtains Eq. (A.3) where p = 0. The form of free oscillations at Ω = Ω0 for x > 0 can be calculated as follows w(x, t) = CG˜ (x, Ω) exp(−iΩ0t) + c. p. 9

Where the paper says it

7) K M ≤ Ω 2 0 ≤ k ρ , K ≥ 0. 8) The form of free oscillations at Ω = Ω0 for x > 0 can be calculated as follows: w(x, t) = CG˜ (x, Ω) exp(−iΩ0t) + c.

c., where C˜ is an arbitrary complex constant. p. 4
Where the paper says it

32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT . 33) Here A is an arbitrary complex constant.

p. 4
Where the paper says it

32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT . 33) Here A is an arbitrary complex constant.

p. 4
Where the paper says it

32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT . 33) Here A is an arbitrary complex constant.

Inhomogeneous non-stationary problem

The solution of Eq. (D.6) yields w(0, τ) = 1/2π Z +∞−∞ pF (omega) exp−iΩτ domega / K − Momega2 + 2√2B1/4ρ3/4(Ω2∗ − Ω2)3/4. p. 9

Where the paper says it

4) Expressing wF (0, Ω) from the above equation we get wF (0, Ω) = pF (Ω) K − MΩ2 + 2√ 2B1/4ρ 3/4(Ω2 ∗ − Ω2) 3/4 . 5) The inverse transform of wF (0, Ω) yields w(0, τ ) = 1 2π Z +∞ −∞ pF (Ω) exp−iΩτ dΩ K − MΩ2 + 2√ 2B1/4ρ 3/4(Ω2 ∗ − Ω2) 3/4 .

p. 9
Where the paper says it

4) Expressing wF (0, Ω) from the above equation we get wF (0, Ω) = pF (Ω) K − MΩ2 + 2√ 2B1/4ρ 3/4(Ω2 ∗ − Ω2) 3/4 . 5) The inverse transform of wF (0, Ω) yields w(0, τ ) = 1 2π Z +∞ −∞ pF (Ω) exp−iΩτ dΩ K − MΩ2 + 2√ 2B1/4ρ 3/4(Ω2 ∗ − Ω2) 3/4 .

For the large times, the non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Ω0. p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

p. 8
Where the paper says it

Here we consider the modes corresponding to the frequencies from the discrete part of the spectrum (the trapped modes), which lies below the cut-off frequency Ω∗: 0 < |Ω| < Ω∗. 1) In what follows we demonstrate that in the problem under consideration the only one trapped mode can exist.

The asymptotic order of the error term in formula (D. p. 4
Where the paper says it

35) where ψ ≡ Z t 0 Ω0(ϵtˆ) dt. 36) For X → 0 and T → 0, the leading-order term of Eq.

p. 4
Where the paper says it

35) where ψ ≡ Z t 0 Ω0(ϵtˆ) dt. 36) For X → 0 and T → 0, the leading-order term of Eq.

7) depends on the properties of pF.

References

M V Fedoryuk Metod perevala [The Saddle-Point Method] Nauka [Science], Moscow, 1977. S N Gavrilov Non-stationary problems in dynamics of a string on an elastic foundation subjected to a moving load Journal of Sound and Vibration, 222(3):345–361. p. 0

Where the paper says it

3 4 Adiabatic invariant 5 5 Equivalent Hamiltonian system 6 6 Numerics 7 7 Conclusion 8 A The dispersion relation 8 B The Green function 9 C The spectral problem 9 D Inhomogeneous non-stationary problem 10 1 Introduction In this paper we apply recently developed in Gavrilov et al. (2024); Shishkina and Gavrilov (2026) approaches to the problem concerning localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator.

S N Gavrilov I O Poroshin E V Shishkina and Yu A Mochalova Formal asymptotics for oscillation of a discrete mass-spring-damper system of time-varying properties, embedded into a one-dimensional medium described by the telegraph equation with variable coefficients Nonlinear Dynamics, 112(23):20967–21002, 2024. p. 1
Where the paper says it

2 Mathematical formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by the following equation: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = p(t) − Mw˙ · − Kw − 2ϵΓw˙ δ(x). 1) Here w(x, t) is the displacement, B = B(ϵt) is the flexural stiffness, ρ = ρ(ϵt) is the mass density of the beam, ϵγ(ϵt) is the damping in the foundation, M = M(ϵt) is the value of mass in the discrete oscillator, K = K(ϵt) is the stiffness for the discrete oscillator, k = k(ϵt) is the stiffness for the Winkler foundation, ϵΓ(ϵt) is the damping in the oscillator, p(t) is the given external force on the beam, x is the spatial coordinate, t is time, ϵ is a formal small parameter.

E V Shishkina S N Gavrilov On the adiabatic invariance of the action of a trapped wave arXiv preprint 2602. p. 1
Where the paper says it

And finally, only the asymptotic approach was used in Shishkina et al. (2019), whereas two additional analytic techniques, namely, the method based on the adiabatic invariance of the action of a trapped wave, and the consideration of the equivalent Hamiltonian system, are utilized in the current paper in the non-dissipative case.

18815. E V Shishkina S N Gavrilov and Yu A Mochalova Non-stationary localized oscillations of an infinite BernoulliEuler beam lying on the Winkler foundation with a point elastic inhomogeneity of time-varying stiffness Journal of Sound and Vibration, 440C:174–185, 2019. p. 0
Where the paper says it

3 4 Adiabatic invariant 5 5 Equivalent Hamiltonian system 6 6 Numerics 7 7 Conclusion 8 A The dispersion relation 8 B The Green function 9 C The spectral problem 9 D Inhomogeneous non-stationary problem 10 1 Introduction In this paper we apply recently developed in Gavrilov et al. (2024); Shishkina and Gavrilov (2026) approaches to the problem concerning localized oscillation of an Euler–Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator.

J C Strikwerda Finite difference schemes and partial differential equations SIAM, Philadelphia, 2004. p. 6
Where the paper says it

To discretize PDE (2. 2) in the particular case under investigation, we use the following scheme Strikwerda (2004): w i+1 j − 2w i j + w i−1 j (∆t) 2 + w i j+2 − 4w i j+1 + 6w i j − 4w i j−1 + w i j−2 (∆x) 4 + u i+1 j + u i−1 j 2 = 0, (6.

V S Vladimirov Equations of Mathematical Physics Marcel Dekker, New York, 1971. p. 1
Where the paper says it

6) Here, and in what follows, [µ] ≡ µ(ξ + 0) − µ(ξ − 0) for any arbitrary quantity µ. The initial conditions can be formulated in the following form Vladimirov (1971): w t<0 ≡ 0.

The equivalent Hamiltonian system

For the problem under consideration the equivalent Hamiltonian system is P˙ = −∂H/∂Q, Q˙ = ∂H/∂P. The coefficients M and K are defined as M(ϵt) = 1/Ω0(ϵt)C0(ϵt) and K(ϵt) = M(ϵt)Ω20 (5. p. 5

Where the paper says it

3) are the generalized co-ordinate and the generalized impulse, respectively. The coefficients M and M(ϵt) = 1 Ω0(ϵt)|C0(ϵt)| , (5.

5). The evolution of the amplitude U can be found by the WKB approach Shishkina and Gavrilov (2026). p. 5
Where the paper says it

The evolution of the amplitude U can be found by the WKB approach. One has Shishkina and Gavrilov (2026): U(t) = |C(ϵt)| cos ψ − arg C + o(1), (5.

Where the gains come from

The formula for J = const is obtained by assuming that the action of the trapped wave J = E¯w / Ω0 is an adiabatic invariant. This leads to C ∝ p8 / (k − ρomega20) (√2M p4 k − ρomega20 + 3ρ√4 B). p. 4

Where the paper says it

where A(T) = p8 k − ρΩ2 q 0 Ω0 3ρ √4 B + √ 2M p4 k − ρΩ2 0 , (3. 32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT .

This is in agreement with Eq. p. 4
Where the paper says it

40) we get the following formula: C ∝ p8 1 − Ω2 √ 0 Ω0 (3. 42) that is in agreement with the main result of the previous paper Shishkina et al.

(3.40). The second formula in Eq. p. 5
Where the paper says it

8) |C0| = |C0(ϵt)| def = 1 √ MΩ0 , (5. , we again obtain the second formula in Eq.

(5.7) is obtained by the WKB approach. p. 5
Where the paper says it

5) are the effective mass and the effective stiffness, respectively; |C0| is defined by Eq. The evolution of the amplitude U can be found by the WKB approach.

C0 = 1 / √MΩ0 (5. p. 5
Where the paper says it

7) arg C = arg C(0), arg C0 = arg C0(0), (5. 8) |C0| = |C0(ϵt)| def = 1 √ MΩ0 , (5.

9). C0 = p C0(ϵt) (5. p. 1
Where the paper says it

2 Mathematical formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by the following equation: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = p(t) − Mw˙ · − Kw − 2ϵΓw˙ δ(x). 1) Here w(x, t) is the displacement, B = B(ϵt) is the flexural stiffness, ρ = ρ(ϵt) is the mass density of the beam, ϵγ(ϵt) is the damping in the foundation, M = M(ϵt) is the value of mass in the discrete oscillator, K = K(ϵt) is the stiffness for the discrete oscillator, k = k(ϵt) is the stiffness for the Winkler foundation, ϵΓ(ϵt) is the damping in the oscillator, p(t) is the given external force on the beam, x is the spatial coordinate, t is time, ϵ is a formal small parameter.

10). The term T = 0 is obtained from the final expression for C(0). p. 4
Where the paper says it

35) where ψ ≡ Z t 0 Ω0(ϵtˆ) dt. 36) For X → 0 and T → 0, the leading-order term of Eq.

A The dispersion relation

The dispersion relation for the operator in the left-hand side of the above equation is Bρω4 + Ω2∗ − Ω2 = 0. The two roots of the dispersion given by expression are omega = ωδ = a(i + δ).

B The Green function

The expression for the Green function (B.10) can be transformed to the following equivalent form G(x, Ω) = √28Ba3 exp(-ax) cos ax − π/4. p. 8

Where the paper says it

9) Expression for the Green function (B. 2) can be transformed to the following equivalent form: G(x, Ω) = √ 2 8Ba3 exp(−ax) cos ax − π 4 , (B.

p. 8
Where the paper says it

9) Expression for the Green function (B. 2) can be transformed to the following equivalent form: G(x, Ω) = √ 2 8Ba3 exp(−ax) cos ax − π 4 , (B.

C The spectral problem

The form of free oscillations at Ω = Ω0 for x > 0 can be calculated as follows w(x, t) = CG˜ (x, Ω) exp(−iΩ0t) + c.c., where C˜ is an arbitrary complex constant.

D Inhomogeneous non-stationary problem

The solution of Eq. (D.6) yields w(0, τ) = 1/2π Z +∞−∞ pF (omega) exp−iΩτ domega / K − Momega2 + 2√2B1/4ρ3/4(Ω2∗ − Ω2)3/4. For the large times, the non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Ω0. The asymptotic order of the error term in formula (D.7) depends on the properties of pF. The final amplitude is given by C0(0) = sqrt(2)p8 / (k − ρomega20) (3ρ√4 B + √2M p4 k − ρomega20) T = 0. The argument of C is arg C(0) = arg pF Ω0(0) + arg C0(0), (D. p. 4

Where the paper says it

12), one gets w(0, t) = 4|A|H(T)A(T) exp (−D(T)) cos π 4 cos(ψ − arg A) + O(ϵ) (3. 35) where ψ ≡ Z t 0 Ω0(ϵtˆ) dt.

13). The argument of C0 is arg C0 = π/2, (D. p. 4
Where the paper says it

34) Now according to Eqs. 12), one gets w(0, t) = 4|A|H(T)A(T) exp (−D(T)) cos π 4 cos(ψ − arg A) + O(ϵ) (3.

14). The final amplitude is given by C(0) = C(ϵt) D(ϵt) cos π/4 cos ψ − arg C + o(1), t → ∞. p. 5
Where the paper says it

The evolution of the amplitude U can be found by the WKB approach. One has Shishkina and Gavrilov (2026): U(t) = |C(ϵt)| cos ψ − arg C + o(1), (5.

The most important results are Eqs. (3.37)–(3.40) describing localized oscillation of an Euler-Bernoulli beam lying on the visco-elastic foundation and coupled with a discrete oscillator. p. 7

Where the paper says it

17) 7 Conclusion The most important results of the paper are Eqs. 40) describing localized oscillation of an Euler-Bernoulli beam lying on the visco-elastic foundation and coupled with a discrete oscillator.

Another important result is that, for the non-dissipative case, we again get formula (5. p. 7
Where the paper says it

40) describing localized oscillation of an Euler-Bernoulli beam lying on the visco-elastic foundation and coupled with a discrete oscillator. Another important result is that, for the non-dissipative case, we again get formula (5.

10) that was previously obtained Gavrilov et al. p. 7
Where the paper says it

Another important result is that, for the non-dissipative case, we again get formula (5. 10) that was previously obtained Gavrilov et al.

(2024); Shishkina and Gavrilov (2026) for the analogous problem concerning localized oscillation of a string. p. 7
Where the paper says it

10) that was previously obtained Gavrilov et al. (2024); Shishkina and Gavrilov (2026) for the analogous problem concerning localized oscillation of a string.

This confirms that the method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task. p. 7
Where the paper says it

(2024); Shishkina and Gavrilov (2026) for the analogous problem concerning localized oscillation of a string. This confirms that the method of the equivalent Hamiltonian system can be obtained for complicated systems where application is the asymptotic approach is a difficult task.

The non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Omega0. The large times, the non-stationary response of the system under consideration is undamped oscillations with the trapped mode frequency Omega0. The final amplitude is given by C(0) = C(ϵt) D(ϵt) cos ψ − arg C + o(1), t → ∞. p. 5
Where the paper says it

The evolution of the amplitude U can be found by the WKB approach. One has Shishkina and Gavrilov (2026): U(t) = |C(ϵt)| cos ψ − arg C + o(1), (5.

p. 5
Where the paper says it

The evolution of the amplitude U can be found by the WKB approach. One has Shishkina and Gavrilov (2026): U(t) = |C(ϵt)| cos ψ − arg C + o(1), (5.

The final amplitude is given by C0 = √4 2A (3. p. 1
Where the paper says it

2 Mathematical formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by the following equation: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = p(t) − Mw˙ · − Kw − 2ϵΓw˙ δ(x). 1) Here w(x, t) is the displacement, B = B(ϵt) is the flexural stiffness, ρ = ρ(ϵt) is the mass density of the beam, ϵγ(ϵt) is the damping in the foundation, M = M(ϵt) is the value of mass in the discrete oscillator, K = K(ϵt) is the stiffness for the discrete oscillator, k = k(ϵt) is the stiffness for the Winkler foundation, ϵΓ(ϵt) is the damping in the oscillator, p(t) is the given external force on the beam, x is the spatial coordinate, t is time, ϵ is a formal small parameter.

40). The final amplitude is given by C(0) = sqrt(2)p8 / (k − ρomega20) (3ρ√4 B + √2M p4 k − ρomega20) T = 0. p. 4
Where the paper says it

where A(T) = p8 k − ρΩ2 q 0 Ω0 3ρ √4 B + √ 2M p4 k − ρΩ2 0 , (3. 32) D(T) = Z T 0 3 √4 Bγ + √ 2Γ p4 k − ρΩ2 0 3 √4 Bρ + √ 2M p4 k − ρΩ2 0 dT .

The final amplitude is given by C(0) = C(ϵt) D(ϵt) cos ψ − arg C + o(1), t → ∞. The final amplitude is given by C0 = √2M / Ω0 (5. p. 4
Where the paper says it

4 Adiabatic invariant In the rest of the paper, we deal with the non-dissipative case Γ = 0, γ = 0. 1) Let us show that formula describing the evolution of the amplitude for the trapped wave can be obtained by postulating that the action of the trapped wave J = E{w¯} Ω0 C=C (4.

9). The final amplitude is given by C0 = p C0(ϵt) (5. p. 1
Where the paper says it

2 Mathematical formulation The transverse oscillation of an Euler–Bernoulli beam on the Winkler foundation with an embedded discrete oscillator is governed by the following equation: (ρw˙) · + Bw′′′′ + kw + 2ϵγw˙ = p(t) − Mw˙ · − Kw − 2ϵΓw˙ δ(x). 1) Here w(x, t) is the displacement, B = B(ϵt) is the flexural stiffness, ρ = ρ(ϵt) is the mass density of the beam, ϵγ(ϵt) is the damping in the foundation, M = M(ϵt) is the value of mass in the discrete oscillator, K = K(ϵt) is the stiffness for the discrete oscillator, k = k(ϵt) is the stiffness for the Winkler foundation, ϵΓ(ϵt) is the damping in the oscillator, p(t) is the given external force on the beam, x is the spatial coordinate, t is time, ϵ is a formal small parameter.

10).

Here we consider the corresponding stationary unperturbed problem with constant parameters and zero dissipation. The two roots of the dispersion given by expression are omega = ωδ = a(i + δ).

The expression for the Green function (B.10) can be transformed to the following equivalent form G(x, Ω) = √28Ba3 exp(-ax) cos ax − π/4.

The form of free oscillations at Ω = Ω0 for x > 0 can be calculated as follows w(x, t) = CG˜ (x, Ω) exp(−iΩ0t) + c.c., where C˜ is an arbitrary complex constant.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can accurately predict localized oscillation amplitudes of an Euler-Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator by employing three analytic approaches: asymptotics, method based on the adiabatic invariance of the action of a trapped wave, and consideration of the equivalent Hamiltonian system.

  2. The improved system can obtain the amplitude of oscillation using only the asymptotic approach in dissipative cases, while for non-dissipative cases, it can utilize both methods to ensure accuracy beyond just asymptotics.

  3. The system can solve complex non-stationary problems by formulating an equivalent Hamiltonian system with a generalized coordinate Q and generalized impulse P, where the effective mass M and stiffness K are derived from the parameters of the system at time t.

Abstract

We consider localized oscillations of an Euler--Bernoulli beam on a visco-elastic foundation coupled to a damped discrete oscillator. All system parameters vary slowly and independently in time. For the conservative case, we employ three different analytical approaches: an asymptotic method, a technique based on the adiabatic invariance of the action of a trapped wave, and an analysis of the equivalent Hamiltonian system. All three approaches yield the identical formula for the oscillation amplitude. In the dissipative case, the oscillation amplitude is obtained solely via the asymptotic approach. Finally, we validate our analytical findings against numerical simulations, demonstrating excellent agreement.

Sources

Related papers