Skip to main content

Stability of thermoelastic Timoshenko beam with suspenders and time-varying feedback


This paper considers a one-dimensional thermoelastic Timoshenko beam system with suspenders, general weak internal damping with time varying coefficient, and time-varying delay terms. Under suitable conditions on the nonlinear terms, we prove a general stability result for the beam model, where exponential and polynomial decay are special cases. We also gave some examples to illustrate our theoretical finding.


In this paper, we consider a thermoelastic Timoshenko beam with suspension cables, weak internal damping, and a time-varying delay damping of the form

$$ \textstyle\begin{cases} \rho u_{tt}(x,t)-\alpha u_{xx}(x,t)-\lambda ( \varphi -u )(x,t) \\ \quad {} + \gamma _{1}a(t)g_{1}(u_{t}(x,t))+ \gamma _{2}a(t)g_{2}(u_{t}(x,t- \tau (t)))=0, \\ \rho _{1}\varphi _{tt}(x,t)-k (\varphi _{x} +\psi )_{x}(x,t) +\lambda (\varphi -u )(x,t)+\gamma _{3}\varphi _{t}(x,t)=0, \\ \rho _{2} \psi _{tt}(x,t)-b \psi _{xx}(x,t) + k (\varphi _{x} + \psi )(x,t)- m\theta _{x}(x,t)=0, \\ \rho _{3} \theta _{t}(x,t)-\beta \theta _{xx}(x,t) -m \psi _{xt}(x,t)=0, \end{cases} $$

where \((x,t)\in (0,1)\times (0,\infty )\), φ represents the transverse displacement (in vertical direction) of the beam cross section, ψ is the angle of rotation of a cross-section. The vertical displacement of the vibrating spring (the cable) is represented by the function u, θ depicts the thermal moment of the beam, \(\lambda >0\) is the common stiffness of the string, and \(\alpha >0\) is the elastic modulus of the string (holding the cable to the deck). The positive constants ρ, \(\rho _{1}\), \(\rho _{2}\) are the density of the mass material of the cable, the mass density, and the moment of mass inertia of the beam, respectively. Also, b, k, β, m represent the rigidity coefficient of the cross-section, the shear modulus of elasticity, the thermal diffusivity, and the coupling coefficient which depends on the material properties, respectively. The function \(\tau (t)>0\) is the time-varying delay, \(\gamma _{1}\) and \(\gamma _{2} \) are real positive damping constants, \(g_{1}\) and \(g_{2}\) are the damping functions, and \(a(t)\) is a nonlinear weight function. We supplement (1.1) with the boundary conditions

$$ \textstyle\begin{cases} u(0,t)=\varphi _{x}(0,t)=\psi (0,t)=\theta _{x}(0,t)=0, \quad t>0, \\ u(1,t)=\varphi (1,t)=\psi _{x}(1,t)=\theta (1,t)=0, \quad t>0, \end{cases} $$

and the initial data

$$ \textstyle\begin{cases} u(x,0)=u_{0}(x),\qquad \varphi (x,0)= \varphi _{0}(x), \\ \psi (x,0)= \psi _{0}(x), \qquad \theta (x,0)=\theta _{0}(x), & \text{in } (0,1), \\ u_{t}(x,0)=u_{1}(x),\qquad \varphi _{t}(x,0)= \varphi _{1}(x), \\ \psi _{t}(x,0)= \psi _{1}(x), & \text{in } (0,1), \\ u_{t}(x,t-\tau (0))=f_{0}(x,t-\tau (0)),\quad \text{in } (0,1)\times (0, \tau (0)). \end{cases} $$

The stability of the above thermoelastic Timoshenko system with suspension cables would be our penultimate focus in this work. The Timoshenko beam model is arguably very popular and most used when the vibration of a beam exhibits significant transverse shear strain. A model to describe this phenomenon was introduced by Timoshenko [35] in 1921, see also [15, 18]. The nonlinear vibration of suspension bridges have captured the attention of different researchers and a number of research articles were written on the topic. The somewhat unpredictable large oscillations of suspension bridges have been modeled in diverse ways, one may see [1, 14, 25]. In any attempt to adequately describe the complicated dynamics of a suspension bridge, a robust model would be one with a considerable amount of degrees of freedom. Without prejudice, some simplified models have been considered, but do not account for a number of realistic behavior of suspension bridges, e.g., torsional oscillations. Of an advantage is the fact that rigorous mathematical analysis is easily carried out with such simpler models. A typical simplified model is the one-dimensional vibrating beam model, where torsional motion is neglected by ignoring sectional dimensions of the beam when they are negligible compared to length of the beam. The emergence of string-beam systems which model a nonlinear coupling of a beam and main cable (the string) were born out of the pioneering works of Lazer, McKenna, and Walter [23, 25, 26] (see also [7] and its references). Though initially modeled through the classic Euler–Bernoulli beam theory, the Timoshenko beam theory is also proven to perform better in predicting a beam response to vibrations than a model based on the classical Euler–Bernoulli beam theory. Indeed, the Timoshenko beam theory takes into account both rotary inertia and shear deformation effects, these are often neglected when applying Euler–Bernoulli beam theory.

In the Timoshenko beam with suspenders which is modeled by (1.1), the suspenders are cables which are elastic in nature and are attached to the beam with elastic springs. The temperature dissipation here is assumed to be governed by the Fourier law of heat conduction. For \(a(t)\equiv 1\), \(g_{1}(s)\equiv s\) and \(\gamma _{2}\equiv 0\), \(g_{2}\equiv 0\) in system (1.1), Bochichio et al. [6] proved a well-posedness and an exponential stability result. A number of works have been done on different thermoelastic Timoshenko models without suspenders (see [10, 12, 16, 17, 28] and references in them). Time delays occur in systems modeling many phenomena in areas such as biosciences, medicine, physics, robotics, economics, chemical, thermal, and structural engineering. These phenomena depend on both present and some past history of occurrences, see [8, 9, 13, 21, 34] and the examples therein. In the case of constant delay and constant weight, the delay term usually accounts for the past history of strain, only up to some finite time \(\tau (t)\equiv \tau \).

A step further involves results in the literature about constant weights (\(\gamma _{1}a(t)\equiv \gamma _{1}\), \(\gamma _{2}a(t)\equiv \gamma _{2}\) constants) and time-varying delay \(\tau (t)\). Works presenting the exponential stability result for wave equation with boundary or internal time-varying delay appeared in Nicaise et al. [32, 33]. Enyi and Mukiawa in [11] presented a general decay result for a viscoelastic plate equation under the condition \(|\gamma _{2}|<|\gamma _{1}|\sqrt{(1-d)}\). Furthermore, in [4, 24], the authors presented some existence and stability results for wave equation with internal time-varying delay and time-varying weights; and for suspension bridge models, see Mukiawa [3, 27, 29, 30].

Motivated by the works in [3, 6, 29], in the current paper, we are concerned with the stability result for the thermoelastic Timoshenko system with suspension cables, time-varying internal feedback, and time-varying weight given in (1.1)–(1.3). The result in [6] is a particular case of our result in this paper.

We arrange this paper in the following manner. In Sect. 2, we state the needed assumptions. In Sect. 3, we present the proof of some technical and needed lemmas for our main result. In the last Sect. 4, we present and prove our main stability result. Throughout this paper, c and \(c_{i}\), \(i=1,2,\dots \), are generic positive constants, which are not necessarily the same from line to line.

Functional settings and assumptions

In this section, we state some needed assumptions on the damping coefficients, nonlinear functions, and the time-varying delay. As in [5, 32, 33], we assume the following conditions:


Function \(a:[0,+\infty )\rightarrow (0,+\infty )\) is a nonincreasing \(C^{1}\)-function such that there exists a positive constant C satisfying

$$ \bigl\vert a'(t) \bigr\vert \leq Ca(t),\qquad \int _{0}^{+\infty}a(t)\,dt=+\infty . $$

Fuction \(g_{1}:\mathbb{R}\rightarrow \mathbb{R}\) is a nondecreasing \(C^{0}\)-function such that there exist positive constants \(C_{1}\), \(C_{2}\), r and a convex increasing function \(\chi \in C^{1}([0,+\infty ))\cap C^{2}((0,+\infty ))\) satisfying \(\chi (0)=0\) or χ is a nonlinear strictly convex \(C^{2}\)-function on \((0, r]\) with \(\chi '(0), \chi ''>0\) such that

$$\begin{aligned}& s^{2}+g_{1}^{2}(s) \le \chi ^{-1}\bigl(s g_{1}(s)\bigr),\quad \text{for all } \vert s \vert \le r, \end{aligned}$$
$$\begin{aligned}& C_{1}s^{2} \le sg_{1}(s)\le C_{2}s^{2}, \quad \text{for all } \vert s \vert \ge r. \end{aligned}$$

Function \(g_{2}:\mathbb{R}\rightarrow \mathbb{R}\) is an increasing and odd \(C^{1}\)-function such that for some positive constants \(C_{3}\), \(\alpha _{1}\), \(\alpha _{2}\),

$$\begin{aligned}& \bigl\vert g_{2}'(s) \bigr\vert \leq C_{3}, \end{aligned}$$
$$\begin{aligned}& \alpha _{1}\bigl(sg_{2}(s)\bigr) \leq G(s) \leq \alpha _{2}\bigl(sg_{1}(s)\bigr), \end{aligned}$$


$$ G(s)= \int _{0}^{s}g_{2}(r)\,dr. $$

There exist \(\tau _{0},\tau _{1}>0\) such that

$$\begin{aligned}& 0 < \tau _{0}\leq \tau (t)\leq \tau _{1},\quad \forall t>0, \end{aligned}$$
$$\begin{aligned}& \tau \in W^{2,\infty}(0,T), \quad \forall T>0, \end{aligned}$$
$$\begin{aligned}& \tau '(t) \leq d< 1,\quad \forall t>0. \end{aligned}$$

The damping coefficients satisfy

$$ \gamma _{2} \alpha _{2}(1-d\alpha _{1})< \alpha _{1}(1-d)\gamma _{1}. $$

Remark 2.1

Using the monotonicity of \(g_{2}\) and the mean value theorem for integrals, we deduce that

$$ G(s)= \int _{0}^{s}g_{2}(r)\,dr< sg_{2}(s). $$

It follows from (2.5) that \(\alpha _{1}<1\).

Similarly, as in Nicaise and Pignotti [31], we introduce the following change of variable:

$$ z(x,\sigma , t)=u_{t}\bigl(x,t-\tau (t)\sigma \bigr), \quad \text{for } (x, \sigma ,t)\in (0,1)\times (0,1)\times (0,\infty ). $$

It follows that

$$ \tau (t)z_{t}(x,\sigma ,t)+\bigl(1-\tau '(t)\sigma \bigr)z_{\sigma}(x,\sigma ,t)=0. $$

Therefore, system (1.1) becomes

$$ \textstyle\begin{cases} \rho u_{tt}(x,t)-\alpha u_{xx}(x,t)-\lambda ( \varphi -u )(x,t)+ \gamma _{1}a(t)g_{1}(u_{t}(x,t)) \\ \quad {}+ \gamma _{2}a(t)g_{2}(z(x,1,t))=0, \\ \rho _{1}\varphi _{tt}(x,t)-k (\varphi _{x} +\psi )_{x}(x,t) +\lambda (\varphi -u )(x,t)+\gamma _{3}\varphi _{t}(x,t)=0, \\ \rho _{2} \psi _{tt}(x,t)-b \psi _{xx}(x,t) + k (\varphi _{x} + \psi )(x,t)- m\theta _{x}(x,t)=0, \\ \rho _{3} \theta _{t}(x,t)-\beta \theta _{xx}(x,t) -m \psi _{xt}(x,t)=0, \\ \tau (t)z_{t}(x,\sigma ,t)+(1-\tau '(t)\sigma )z_{\sigma}(x,\sigma ,t)=0, \end{cases} $$

subjected to the boundary conditions

$$ \textstyle\begin{cases} u(0,t)=\varphi _{x}(0,t)=\psi (0,t)=\theta _{x}(0,t)=0, \quad t>0, \\ u(1,t)=\varphi (1,t)=\psi _{x}(1,t)=\theta (1,t)=0, \quad t>0, \\ z(x,0,t)=u_{t}(x,t), \quad x\in (0,1), t>0, \end{cases} $$

and initial data

$$ \textstyle\begin{cases} u(x,0)=u_{0}(x),\qquad \varphi (x,0)= \varphi _{0}(x), \\ \psi (x,0)= \psi _{0}(x),\qquad \theta (x,0)=\theta _{0}(x), & \text{in } (0,1), \\ u_{t}(x,0)=u_{1}(x),\qquad \varphi _{t}(x,0)= \varphi _{1}(x), \\ \psi _{t}(x,0)= \psi _{1}(x), & \text{in } (0,1), \\ z(x,\sigma ,0)= u_{t}(x,-\tau (0)\sigma )=f_{0}(x,-\tau (0)\sigma ), & \text{in } (0,1)\times (0,1). \end{cases} $$

We introduce the following spaces:

$$\begin{aligned} & H_{a}^{1}(0,1)=\bigl\{ \phi \in H^{1}(0,1): \phi (0) = 0\bigr\} , \\ & H_{b}^{1}(0,1)= \bigl\{ \phi \in H^{1}(0,1): \phi (1) = 0\bigr\} , \\ & H^{2}_{a}(0,1)=\bigl\{ \phi \in H^{2}(0,1): \phi _{x}\in H_{a}^{1}(0,1)\bigr\} , \\ & H^{2}_{b}(0,1)=\bigl\{ \phi \in H^{2}(0,1): \phi _{x}\in H_{b}^{1}(0,1) \bigr\} . \end{aligned}$$

For completeness, we state without proof the existence and uniqueness result for problem (1.1)–(1.3). The result can be established using the Faedo–Galerkin approximation method, see [5] or standard nonlinear semigroup method, see [19, 20].

Theorem 2.1


$$ \begin{aligned} (u_{0},\varphi _{0},\psi _{0},\theta _{0})&\in H^{2}(0,1)\cap H^{1}_{0}(0,1)\times H^{2}_{a}(0,1)\cap H^{1}_{b}(0,1) \times H^{2}_{b}(0,1) \\ &\quad {}\cap H^{1}_{a}(0,1) \times H^{2}_{a}(0,1)\cap H^{1}_{b}(0,1) \end{aligned} $$


$$ (u_{1},\varphi _{1},\psi _{1})\in H^{1}_{0}(0,1)\times H^{1}_{a}(0,1) \times H^{1}_{b}(0,1),\qquad f_{0}\bigl(\cdot ,-\tau (0) \bigr)\in H^{1}_{0} \bigl( (0,1);H^{1}(0,1) \bigr) $$

be given such that

$$ f_{0}(\cdot ,0)=u_{1}. $$

Suppose conditions \((A_{1})\)\((A_{4})\) hold. Then, problem (1.1)(1.3) has a unique global weak solution in the class

$$\begin{aligned} &u\in L^{\infty} \bigl( [0,+\infty);H^{2}(0,1) \cap H^{1}_{0}(0,1) \bigr), \qquad u_{t}\in L^{\infty} \bigl( [0,+\infty); H^{1}_{0}(0,1) \bigr), \\ & u_{tt} \in L^{\infty} \bigl( (0,+\infty );L^{2}(0,1) \bigr), \\ &\varphi \in L^{\infty} \bigl( [0,+\infty);H^{2}_{a}(0,1) \cap H^{1}_{b}(0,1) \bigr),\qquad \varphi _{t} \in L^{\infty} \bigl( [0,+\infty);H^{1}_{b}(0,1) \bigr), \\ &\varphi _{tt} \in L^{\infty} \bigl( (0,+\infty );L^{2}(0,1) \bigr), \\ &\psi \in L^{\infty} \bigl( [0,+\infty);H^{2}_{b}(0,1) \cap H^{1}_{a}(0,1) \bigr), \qquad \psi _{t} \in L^{\infty} \bigl( [0,+\infty);H^{1}_{a}(0,1) \bigr), \\ &\psi _{tt} \in L^{\infty} \bigl( (0,+\infty );L^{2}(0,1) \bigr), \\ &\theta \in L^{\infty} \bigl( [0,+\infty);H^{2}_{a}(0,1) \cap H^{1}_{b}(0,1) \bigr),\qquad \theta _{t}\in L^{\infty} \bigl( (0,+\infty );L^{2}(0,1) \bigr). \end{aligned}$$

Technical lemmas

In this section, we prove some important lemmas which will be essential in establishing the main result. Let μ̄ be a positive constant satisfying

$$ \frac{\gamma _{2}(1-\alpha _{1})}{\alpha _{1}(1-d)}< \bar{\mu}< \frac{\gamma _{1}-\gamma _{2}\alpha _{2}}{\alpha _{2}} $$

and set

$$ \mu (t)=\bar{\mu}a(t). $$

The energy functional of system (2.14)–(2.16) is defined by

$$ \begin{aligned} E(t)={}& \frac{1}{2} \int _{0}^{1} \bigl[\rho u_{t}^{2} + \rho _{1} \varphi _{t}^{2} + \rho _{2} \psi _{t}^{2} + \alpha u_{x}^{2} + k( \varphi _{x}+\psi )^{2}+b\psi _{x}^{2}+ \lambda (\varphi -u)^{2} \bigr]\,dx \\ &{}+ \frac{1}{2} \int _{0}^{1}\rho _{3}\theta ^{2}\,dx + \mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1} G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned} $$

Lemma 3.1

Let \((u,\varphi , \psi , \theta ,z)\) be the solution of system (2.14)(2.16). Then, the energy functional (3.2) satisfies

$$ \begin{aligned} \frac{dE(t)}{dt}\leq {}&-a(t) [ \gamma _{1}-\bar{\mu}\alpha _{2}- \gamma _{2}\alpha _{2} ] \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx \\ &{}-a(t) \bigl[\bar{\mu}\bigl(1-\tau '(t)\bigr)\alpha _{1}- \gamma _{2}(1-\alpha _{1}) \bigr] \int _{0}^{1}z(x,1,t)g_{2}\bigl(z(x,1,t) \bigr)\,dx \\ & {}-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1} \theta _{x}^{2}\,dx \\ \leq{} &0, \quad \forall t\geq 0. \end{aligned} $$


Multiplying (2.14)1 by \(u_{t}\), (2.14)2 by \(\varphi _{t}\), (2.14)3 by \(\psi _{t}\), and (2.14)4 by θ, integrating the outcome over \((0,1)\), and applying integration by parts and the boundary conditions, we get

$$\begin{aligned} \begin{aligned} &\frac {1}{2}\frac {d}{dt} \int _{0}^{1} \bigl[ \rho u^{2}_{t} + \alpha u_{x}^{2}+\lambda (\varphi -u)^{2} \bigr]\,dx \\ & \quad =\lambda \int _{0}^{1}\varphi _{t}(\varphi -u)\,dx- \gamma _{1}a(t) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx- \gamma _{2}a(t) \int _{0}^{1}u_{t} g_{2} \bigl(z(x,1,t)\bigr)\,dx, \end{aligned} \end{aligned}$$
$$\begin{aligned} \begin{aligned} &\frac {1}{2}\frac {d}{dt} \int _{0}^{1} \bigl[ \rho _{1}\varphi _{t}^{2} +k(\varphi _{x}+\psi )^{2} \bigr]\,dx \\ & \quad =-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \lambda \int _{0}^{1} \varphi _{t}(\varphi -u)\,dx +k \int _{0}^{1} \psi _{t}(\varphi _{x}+ \psi )\,dx, \end{aligned} \end{aligned}$$
$$\begin{aligned} &\frac {1}{2}\frac {d}{dt} \int _{0}^{1} \bigl[ \rho _{2}\psi ^{2}_{t} + b\psi _{x}^{2} \bigr]\,dx= m \int _{0}^{1} \psi _{t}\theta _{x}\,dx-k \int _{0}^{1} \psi _{t}(\varphi _{x}+\psi )\,dx, \end{aligned}$$
$$\begin{aligned} &\frac {1}{2} \int _{0}^{1} \rho _{3}\theta ^{2}\,dx=-\beta \int _{0}^{1} \theta _{x}^{2}\,dx-m \int _{0}^{1}\psi _{t}\theta _{x}\,dx. \end{aligned}$$

Adding (3.4)–(3.7), we arrive at

$$\begin{aligned}& \begin{aligned} &\frac{1}{2} \int _{0}^{1} \bigl( \rho u^{2}_{t} +\alpha u_{x}^{2}+ \lambda (\varphi -u)^{2}+\rho _{1}\varphi _{t}^{2} +k(\varphi _{x}+ \psi )^{2}+\rho _{2}\psi ^{2}_{t} + b\psi _{x}^{2}+\rho _{3} \theta ^{2} \bigr)\,dx \\ &\quad = -\gamma _{1}a(t) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx- \gamma _{2}a(t) \int _{0}^{1}u_{t} g_{2} \bigl(z(x,1,t)\bigr)\,dx \\ &\qquad {}-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1} \theta _{x}^{2}\,dx. \end{aligned} \end{aligned}$$

Now, multiplying equation (2.14)5 by \(\mu (t)g_{2}(z(x,\sigma ,t))\) and integrating over \((0,1)\times (0,1)\), we obtain

$$ \begin{aligned} &\mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1} z_{t}(x,\sigma ,t) g_{2}\bigl(z(x, \sigma ,t)\bigr)\,d\sigma \,dx \\ &\quad {} +\mu (t) \int _{0}^{1} \int _{0}^{1}\bigl(1-\tau '(t)\sigma \bigr)z_{\sigma}(x, \sigma , t) g_{2}\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx=0. \end{aligned} $$

On account of (2.6), we can write

$$ \frac{\partial }{\partial \sigma} \bigl[ G\bigl(z(x,\sigma ,t)\bigr) \bigr]=z_{ \sigma}(x, \sigma , t) g_{2}\bigl(z(x,\sigma ,t)\bigr). $$

Therefore, (3.9) becomes

$$ \begin{aligned} &\mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1}z_{t}(x,\sigma ,t) g_{2}\bigl(z(x, \sigma ,t)\bigr)\,d\sigma \,dx \\ &\quad =-\mu (t) \int _{0}^{1} \int _{0}^{1}\bigl(1-\tau '(t)\sigma \bigr) \frac{\partial }{\partial \sigma} \bigl[ G\bigl(z(x,\sigma ,t)\bigr) \bigr]\,d\sigma \,dx. \end{aligned} $$

It follows that

$$ \begin{aligned} &\frac{d}{dt} \biggl(\mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x, \sigma ,t)\bigr)\,d\sigma \,dx \biggr) \\ &\quad =-\mu (t) \int _{0}^{1} \int _{0}^{1}\frac{\partial}{\partial \sigma} \bigl[\bigl(1-\tau '(t)\sigma \bigr)G\bigl(z(x,\sigma ,t)\bigr) \bigr]\,d\sigma \,dx \\ &\qquad {} + \mu '(t)\tau (t) \int _{0}^{l} \int _{0}^{1}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\ &\quad =\mu (t) \int _{0}^{1} \bigl( G\bigl(z(x,0,t)\bigr)-G \bigl(z(x,1,t)\bigr) \bigr)\,dx +\mu (t) \tau '(t) \int _{0}^{1} G\bigl(z(x,1,t)\bigr)\,dx \\ &\qquad {}+\mu '(t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\ &\quad =\mu (t) \int _{0}^{1}G\bigl(u_{t}(x,t)\bigr)\,dx-\mu (t) \bigl(1-\tau '(t)\bigr) \int _{0}^{1}G(z(x,1,t)\,dx \\ &\qquad {}+\mu '(t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned} $$

Recalling the definition of the energy functional (3.2), and adding (3.8) and (3.12), we obtain

$$ \begin{aligned} \frac{dE(t)}{dt}={}& -\gamma _{1}a(t) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx- \gamma _{2}a(t) \int _{0}^{1}u_{t} g_{2} \bigl(z(x,1,t)\bigr)\,dx \\ &{} +\mu (t) \int _{0}^{1}G\bigl(u_{t}(x,t)\bigr)\,dx-\mu (t) \bigl(1-\tau '(t)\bigr) \int _{0}^{1}G(z(x,1,t)\,dx \\ & {}-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1} \theta _{x}^{2}\,dx +\mu '(t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x, \sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned} $$

On the account of \((A_{1})\) and (2.5), we get

$$ \begin{aligned} \frac{dE(t)}{dt}\leq{} &- \bigl( \gamma _{1}a(t)-\mu (t)\alpha _{2} \bigr) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx- \gamma _{2}a(t) \int _{0}^{1}u_{t} g_{2} \bigl(z(x,1,t)\bigr)\,dx \\ &{} -\mu (t) \bigl(1-\tau '(t)\bigr) \int _{0}^{1}G(z(x,1,t)\,dx -\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1}\theta _{x}^{2}\,dx . \end{aligned} $$

Now, we consider the convex conjugate of G defined by

$$ G^{*}(s)=s\bigl(G'\bigr)^{-1}(s)-G \bigl(\bigl(G'\bigr)^{-1}(s)\bigr), \quad \forall s\geq 0, $$

which satisfies the generalized Young inequality (see [2])

$$ AB\leq G^{*}(A)+G(B), \quad \forall A,B >0. $$

Using (2.5) and the definition of G, we get

$$ G^{*}(s)=sg_{2}^{-1}(s)-G \bigl(g_{2}^{-1}(s)\bigr), \quad \forall s\geq 0. $$

Therefore, on account of (2.5) and (3.17), we have

$$ \begin{aligned} G^{*}\bigl(g_{2} \bigl(z(x,1,t)\bigr)\bigr)&=z(x,1,t)g_{2}\bigl(z(x,1,t)\bigr)-G \bigl(z(x,1,t)\bigr) \\ &\leq (1-\alpha _{1})z(x,1,t)g_{2}\bigl(z(x,1,t)\bigr). \end{aligned} $$

A combination of (3.14), (3.16), and (3.18) leads to

$$\begin{aligned} \frac{dE(t)}{dt} \leq &- \bigl( \gamma _{1}a(t)-\mu (t)\alpha _{2} \bigr) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx \\ &{} +\gamma _{2}a(t) \int _{0}^{1}(G(u_{t})+G^{*} \bigl( g_{2}\bigl(z(x,1,t)\bigr)\bigr)\,dx \\ &{} -\mu (t) \bigl(1-\tau '(t)\bigr) \int _{0}^{1}G(z(x,1,t)\,dx -\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1}\theta _{x}^{2}\,dx \\ \leq& - \bigl( \gamma _{1}a(t)-\mu (t)\alpha _{2} \bigr) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx +\gamma _{2}a(t)\alpha _{2} \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx \\ &{} +\gamma _{2}a(t) (1-\alpha _{1}) \int _{0}^{1}z(x,1,t)g_{2}\bigl(z(x,1,t) \bigr)\,dx \\ &{} -\mu (t) \bigl(1-\tau '(t)\bigr) \int _{0}^{1}G(z(x,1,t)\,dx-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1}\theta _{x}^{2}\,dx \\ \leq& - \bigl( \gamma _{1}a(t)-\mu (t)\alpha _{2}-\gamma _{2}a(t) \alpha _{2} \bigr) \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx \\ &{} - \bigl(\mu (t) \bigl(1-\tau '(t)\bigr)\alpha _{1}- \gamma _{2}a(t) (1- \alpha _{1}) \bigr) \int _{0}^{1}z(x,1,t)g_{2}\bigl(z(x,1,t) \bigr)\,dx \\ &{} -\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1} \theta _{x}^{2}\,dx. \end{aligned}$$

Recalling that \(\mu (t)=\bar{\mu}a(t)\), it follows from (3.19) that

$$ \begin{aligned} \frac{dE(t)}{dt}\leq{}& -a(t) [ \gamma _{1}-\bar{\mu}\alpha _{2}- \gamma _{2}\alpha _{2} ] \int _{0}^{1}u_{t} g_{1}(u_{t})\,dx \\ & {}-a(t) \bigl[\bar{\mu}\bigl(1-\tau '(t)\bigr)\alpha _{1}-\gamma _{2}(1- \alpha _{1}) \bigr] \int _{0}^{1}z(x,1,t)g_{2}\bigl(z(x,1,t) \bigr)\,dx \\ & {}-\gamma _{3} \int _{0}^{1}\varphi _{t}^{2}\,dx- \beta \int _{0}^{1} \theta _{x}^{2}\,dx. \end{aligned} $$

Using (2.9) and (3.1), we obtain the desired result. This finishes the proof. □

Lemma 3.2

The functional \(F_{1}\), defined by

$$ F_{1}(t):=-\rho _{2}\rho _{3} \int _{0}^{1}\psi _{t} \int _{0}^{x} \theta (y,t)\,dy\,dx, $$

satisfies, along the solution of system (2.14)(2.16) and for any \(\epsilon _{1}, \epsilon _{2}>0\), the estimate

$$ \begin{aligned} F'_{1}(t)\le{}& - \frac {m\rho _{2}}{2} \int _{0}^{1}\psi _{t}^{2}\,dx + \epsilon _{1} \int _{0}^{1}\psi _{x}^{2}\,dx+ \epsilon _{2} \int _{0}^{1}( \varphi _{x}+\psi )^{2}\,dx \\ &{}+c \biggl(1+\frac {1}{\epsilon _{1}} + \frac {1}{\epsilon _{2}} \biggr) \int _{0}^{1}\theta _{x}^{2}\,dx. \end{aligned} $$


Differentiating \(F_{1}\), using (2.14)3 and (2.14)4, then integrating by parts and exploiting the boundary conditions lead to

$$ \begin{aligned} F_{1}'(t)={}& b \rho _{3} \int _{0}^{1}\psi _{x}\theta \,dx +k\rho _{3} \int _{0}^{1}(\varphi _{x}+\psi ) \int _{0}^{x}\theta (y,t)\,dy\,dx \\ & {}+ m\rho _{3} \int _{0}^{1}\theta ^{2}\,dx -\rho _{2}\beta \int _{0}^{1} \psi _{t}\theta _{x}\,dx-\rho _{2}m \int _{0}^{1}\psi _{t}^{2}\,dx. \end{aligned} $$

Making use of Cauchy–Schwarz, Young’s, and Poincaré’s inequalities, we get (3.21). □

Lemma 3.3

The functional \(F_{2}\), defined by

$$ F_{2}(t):= \int _{0}^{1} \biggl(\rho uu_{t} +\rho _{1}\varphi \varphi _{t} + \rho _{2}\psi \psi _{t}+\frac{\gamma _{3}}{2}\varphi ^{2} \biggr)\,dx, $$

satisfies, along the solution of system (2.14)(2.16), the estimate

$$ \begin{aligned} F'_{2}(t)\le{}& - \int _{0}^{1} \biggl(\frac{\alpha}{2} u_{x}^{2}+ \lambda (\varphi -u)^{2}+k(\varphi _{x} +\psi )^{2}+\frac {b}{2}\psi _{x}^{2} \biggr)\,dx \\ & {}+ \int _{0}^{1} \bigl(\rho u_{t}^{2}+ \rho _{1}\varphi ^{2}_{t} + \rho _{2} \psi _{t}^{2} \bigr)\,dx +c \int _{0}^{1}\theta _{x}^{2}\,dx \\ &{} +c \int _{0}^{1} \bigl\vert g_{1}(u_{t}) \bigr\vert ^{2}\,dx+ c \int _{0}^{1} \bigl\vert g_{2} \bigl(z(x,1,t)\bigr) \bigr\vert ^{2}\,dx, \quad \forall t\geq 0. \end{aligned} $$


Directly differentiating \(F_{2}\), using (2.14)1, (2.14)2, and (2.14)3, then applying integration by parts and boundary conditions, we obtain

$$ \begin{aligned} F'_{2}(t)=&{}- \int _{0}^{1} \bigl( \alpha u_{x}^{2}+ \lambda (\varphi -u)^{2}+k( \varphi _{x} +\psi )^{2}+b\psi _{x}^{2} \bigr)\,dx \\ &{}+ \int _{0}^{1} \bigl(\rho u_{t}^{2} + \rho _{1}\varphi ^{2}_{t} + \rho _{2} \psi _{t}^{2} \bigr)\,dx +m \int _{0}^{1} \psi \theta _{x}\,dx \\ &{}-\gamma _{1}a(t) \int _{0}^{1} ug_{1}(u_{t})\,dx- \gamma _{2}a(t) \int _{0}^{1}ug_{2}\bigl(z(x,1,t)\bigr)\,dx. \end{aligned} $$

Using \((A_{1})\), Young’s and Poincaré’s inequalities, we obtain (3.23). □

Lemma 3.4

The functional

$$ F_{3}(t):=\bar{\mu}\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t) \sigma}G\bigl(z(x,\sigma ,t) \bigr)\,d\sigma \,dx, $$

satisfies, along the solution of system (2.14)(2.16), the estimate

$$ \begin{aligned} F'_{3}(t) \le &-2F_{3}(t)+ \frac{\bar{\mu}\alpha _{2}}{2} \int _{0}^{1} \bigl( u_{t}^{2}+ \bigl\vert g_{1}(u_{t}) \bigr\vert ^{2} \bigr)\,dx, \quad \forall t \geq 0. \end{aligned} $$


Differentiating \(F_{3}\), we get

$$ \begin{aligned} F'_{3}(t) ={}& \bar{\mu}\tau '(t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t) \sigma}G\bigl(z(x,\sigma ,t) \bigr)\,d\sigma \,dx \\ &{} -2\bar{\mu}\tau (t)\tau '(t) \int _{0}^{1} \int _{0}^{1}\sigma e^{-2 \tau (t)\sigma}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\ & {}+\bar{\mu}\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t) \sigma}z_{t}(x, \sigma ,t)g_{2}\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned} $$

Using the last equation in (2.14), we can express the last term on the right hand-side of (3.26) as

$$\begin{aligned}& \tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma}z_{t}(x, \sigma ,t)g_{2}\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\& \quad = \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma} \bigl(\tau '(t) \sigma -1 \bigr) z_{\sigma}(x,\sigma ,t)g_{2} \bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\& \quad = \int _{0}^{1} \int _{0}^{1}\frac{\partial}{\partial \sigma} \bigl[ e^{-2 \tau (t)\sigma} \bigl(\tau '(t)\sigma -1 \bigr)G\bigl(z(x,\sigma ,t)\bigr) \bigr]\,d\sigma \,dx \\& \begin{aligned}&\qquad {}+2\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma} \bigl(\tau '(t)\sigma -1 \bigr)G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\ &\qquad {}-\tau '(t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma}G\bigl(z(x, \sigma ,t) \bigr)\,d\sigma \,dx \end{aligned}\\& \quad =-\bigl(1-\tau '(t)\bigr)e^{-2\tau (t)} \int _{0}^{1}G\bigl(z(x,1,t)\bigr)\,dx + \int _{0}^{l}G(u_{t})\,dx \\& \qquad {}+2\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma} \bigl(\tau '(t)\sigma -1 \bigr)G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx \\& \qquad {}-\tau '(t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t)\sigma}G\bigl(z(x, \sigma ,t) \bigr)\,d\sigma \,dx. \end{aligned}$$

Substituting (3.27) into (3.26), we arrive at

$$ \begin{aligned} F'_{3}(t) ={}&-2 \bar{\mu}\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2\tau (t) \sigma}G\bigl(z(x,\sigma ,t) \bigr)\,d\sigma \,dx +\bar{\mu} \int _{0}^{1}G(u_{t})\,dx \\ &{} -\bigl(1-\tau '(t)\bigr)e^{-2\tau (t)} \int _{0}^{1}G\bigl(z(x,1,t)\bigr)\,dx. \end{aligned} $$

Using condition (2.5) and Young’s inequality, we obtain (3.26). □

Lemma 3.5

Let \((u,\varphi , \psi ,\theta , z)\) be the solution of system (2.14)(2.16). Then, for \(N, N_{1}, N_{2}>0\) sufficiently large, the Lyapunov functional L, defined by

$$\begin{aligned} L(t):=NE(t)+N_{1}F_{1}(t)+ N_{2}F_{2}(t)+F_{3}(t), \end{aligned}$$

satisfies, for some positive constants \(c_{1}\), \(c_{2}\), η,

$$ c_{1}E(t) \le L(t) \le c_{2}E(t), \quad \forall t \ge 0, $$


$$ \begin{aligned} L'(t)\le -\eta E(t)+c \int _{0}^{1}\bigl(u^{2}_{t}+ \bigl\vert g_{1}(u_{t}) \bigr\vert ^{2} \bigr)\,dx+ c \int _{0}^{1} \bigl\vert g_{2} \bigl(z(x,1,t)\bigr) \bigr\vert ^{2}\,dx, \quad \forall t \ge 0. \end{aligned} $$


Applying Cauchy–Schwarz, Young’s, and Poincaré’s inequalities, we have

$$\begin{aligned} \bigl\vert L(t)-NE(t) \bigr\vert \leq& N_{1} \biggl\vert -\rho _{2} \int _{0}^{1}\psi _{t} \int _{0}^{x} \theta (y,t)\,dy\,dx \biggr\vert \\ &{} + N_{2} \biggl\vert \int _{0}^{1} \biggl(\rho uu_{t} +\rho _{1} \varphi \varphi _{t} + \rho _{2}\psi \psi _{t}+\frac{\gamma _{3}}{2} \varphi ^{2} \biggr)\,dx \biggr\vert \\ &{} + \biggl\vert \bar{\mu}\tau (t) \int _{0}^{1} \int _{0}^{1} e^{-2 \tau (t)\sigma}G\bigl(z(x,\sigma ,t) \bigr)\,d\sigma \,dx \biggr\vert \\ \leq& \frac{N_{2}\rho}{2} \int _{0}^{1}u_{t}^{2}\,dx + \frac{N_{2}\rho _{1}}{2} \int _{0}^{1}\varphi _{t}^{2}\,dx+ \frac{(N_{1}+N_{2})\rho _{2}}{2} \int _{0}^{1}\psi _{t}^{2}\,dx \\ &{} +\frac{N_{2}\gamma _{3}}{2} \int _{0}^{1}\varphi ^{2}\,dx+ \frac{N_{2}\rho}{2} \int _{0}^{1}u_{x}^{2}\,dx + \frac{N_{2}\rho _{1}}{2} \int _{0}^{1}\varphi _{x}^{2}\,dx \\ &{} + \frac{N_{2}\rho _{2}}{2} \int _{0}^{1}\psi _{x}^{2}\,dx + \frac{ N_{1}\rho _{2}}{2} \int _{0}^{1} \biggl( \int _{0}^{x}\theta (y,t)\,dy \biggr)^{2}\,dx \\ &{} +\bar{\mu}\tau (t) \int _{0}^{1} \int _{0}^{1} G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned}$$

Using the relations

$$ \begin{aligned} & \int _{0}^{1}\varphi ^{2}\,dx\leq 2 \int _{0}^{l}(\varphi -u)^{2}\,dx + 2 \int _{0}^{1} u_{x}^{2}\,dx, \\ & \int _{0}^{1}\varphi _{x}^{2}\,dx \leq 2 \int _{0}^{1}(\varphi _{x}+ \psi )^{2}\,dx +2 \int _{0}^{1}\psi _{x}^{2}\,dx, \end{aligned} $$

we arrive at

$$ \begin{aligned} \bigl\vert L(t)-NE(t) \bigr\vert \leq{} &\frac{N_{2}\rho}{2} \int _{0}^{1}u_{t}^{2}\,dx + \frac{N_{2}\rho _{1}}{2} \int _{0}^{1}\varphi _{t}^{2}\,dx+ \frac{(N_{1}+N_{2})\rho _{2}}{2} \int _{0}^{1}\psi _{t}^{2}\,dx \\ &{} +N_{2}\gamma _{3} \int _{0}^{l}(\varphi -u)^{2}\,dx+ \biggl( N_{2} \gamma _{3}+\frac{N_{2}\rho}{2} \biggr) \int _{0}^{1}u_{x}^{2}\,dx \\ & {}+N_{2}\rho _{1} \int _{0}^{1}(\varphi _{x}+\psi )^{2}\,dx + \biggl( \frac{N_{2}(\rho _{1}+\rho _{2})}{2} \biggr) \int _{0}^{1}\psi _{x}^{2}\,dx \\ &{} +\frac{ N_{1}\rho _{2}}{2} \int _{0}^{1}\theta ^{2}\,dx+\bar{\mu} \tau (t) \int _{0}^{1} \int _{0}^{1} G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx. \end{aligned} $$

From (3.33), we obtain

$$ \bigl\vert L(t)-NE(t) \bigr\vert \leq \bar{c}E(t). $$

By choosing N large enough such that

$$ c_{1}=N- \bar{c}>0,\qquad c_{2}=N+ \bar{c}>0, $$

estimate (4.14) follows. Next, we establish (3.31). Using Lemmas 3.13.4, we get

$$ \begin{aligned} L'(t)\le{}& -\rho \int _{0}^{1}u_{t}^{2}\,dx- [N \gamma _{3}-N_{2} \rho _{1} ] \int _{0}^{1}\varphi _{t}^{2}\,dx- \biggl[N_{1} \frac{m\rho _{2}}{2}-N_{2}\rho _{2} \biggr] \int _{0}^{1}\psi _{t}^{2}\,dx \\ & {}-\frac{N_{2}\alpha}{2} \int _{0}^{1}u_{x}^{2}\,dx-N_{2} \lambda \int _{0}^{1}(\varphi -u)^{2}\,dx- [N_{2}k-N_{1}\epsilon _{2} ] \int _{0}^{1}(\varphi _{x}+\psi )^{2}\,dx \\ &{} - \biggl[N_{2}\frac{b}{2}-N_{1}\epsilon _{1} \biggr] \int _{0}^{1} \psi _{x}^{2}\,dx- \biggl[N\beta -N_{1}c \biggl(1+\frac{1}{\epsilon _{1}} + \frac{1}{\epsilon _{2}} \biggr)-N_{2}c \biggr] \int _{0}^{1}\theta _{x}^{2}\,dx \\ &{} -\frac{2e^{-2\tau _{1}}}{a(0)}\mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx + \biggl[\rho +N_{2}\rho + \frac{\bar{\mu }\alpha _{2}}{2} \biggr] \int _{0}^{1}u_{t}^{2}\,dx \\ &{} + \biggl[cN_{2}+ \frac{\bar{\mu}\alpha _{2}}{2} \biggr] \int _{0}^{1} \bigl\vert g_{1}(u_{t}) \bigr\vert ^{2}\,dx +cN_{2} \int _{0}^{1} \bigl\vert g_{2} \bigl(z(x,1,t)\bigr) \bigr\vert ^{2}\,dx. \end{aligned} $$


$$ N_{2}=1, \qquad \epsilon _{1}=\frac {N_{2}b}{4N_{1}}, \qquad \epsilon _{2}= \frac {N_{2}k}{2N_{1}}, $$

we arrive at

$$ \begin{aligned} L'(t)\le{}& -\rho \int _{0}^{1}u_{t}^{2}\,dx- [N \gamma _{3}-\rho _{1} ] \int _{0}^{1}\varphi _{t}^{2}\,dx- \biggl[N_{1}\frac{m\rho _{2}}{2}- \rho _{2} \biggr] \int _{0}^{1}\psi _{t}^{2}\,dx \\ & {}-\frac{\alpha}{2} \int _{0}^{1}u_{x}^{2}\,dx-\lambda \int _{0}^{1}( \varphi -u)^{2}\,dx- \frac {k}{2} \int _{0}^{1}(\varphi _{x}+\psi )^{2}\,dx \\ & {}-\frac{b}{4} \int _{0}^{1}\psi _{x}^{2}\,dx- \biggl[N\beta -N_{1}c \biggl(1+\frac{4N_{1}}{b} +\frac{2N_{1}}{k} \biggr)-c \biggr] \int _{0}^{1} \theta _{x}^{2}\,dx \\ & {}-\frac{2e^{-2\tau _{1}}}{a(0)}\mu (t)\tau (t) \int _{0}^{1} \int _{0}^{1}G\bigl(z(x,\sigma ,t)\bigr)\,d\sigma \,dx + \biggl[2\rho + \frac{\bar{\mu }\alpha _{2}}{2} \biggr] \int _{0}^{1}u_{t}^{2}\,dx \\ &{} + \biggl[c+ \frac{\bar{\mu }\alpha _{2}}{2} \biggr] \int _{0}^{1} \bigl\vert g_{1}(u_{t}) \bigr\vert ^{2}\,dx +c \int _{0}^{1} \bigl\vert g_{2} \bigl(z(x,1,t)\bigr) \bigr\vert ^{2}\,dx. \end{aligned} $$

Now, we choose \(N_{1}\) large such that

$$ N_{1}\frac{m\rho _{2}}{2}-\rho _{2}>0. $$

Next, we select N very large so that (4.14) remains true and

$$ N\gamma _{3}-\rho _{1}>0,\qquad N\beta -N_{1}c \biggl(1+ \frac{4N_{1}}{b} +\frac{2N_{1}}{k} \biggr)-c>0. $$

Therefore, using the energy functional defined by (3.2), we obtain (3.31). □

Stability result

In this section, we are concerned with the main stability result, and is stated as follows.

Theorem 4.1

Let \((u,\varphi , \psi , \theta , z)\) be the solution of system (2.14)(2.16) and assume (A1)(A4) hold. Then, for some positive constants \(\delta _{1}\), \(\delta _{2}\), \(\delta _{3}\), and \(r_{0}\), the energy functional (3.2) satisfies

$$ E(t)\le \delta _{1}\chi _{1}^{-1} \biggl(\delta _{2} \int _{0}^{t} a(s)\,ds +\delta _{3} \biggr),\quad t \ge 0, $$


$$ \chi _{1}(t)= \int _{t}^{1}\frac {1}{\chi _{0}(s)}\,ds\quad \textit{and}\quad \chi _{0}(t)=t \chi '(r_{0} t). $$


We divide the proof into two cases:

Case I: χ is linear. Using (A2), we get

$$ C_{1} \vert s \vert \le \bigl\vert g_{1}(s) \bigr\vert \le C_{2} \vert s \vert , \quad \forall s\in \mathbb{R}. $$


$$\begin{aligned} g_{1}^{2}(s)\le C_{2}sg_{1}(s), \quad \forall s\in \mathbb{R}. \end{aligned}$$

Therefore, multiplying (3.31) by \(a(t)\) and using (3.3) and (4.2), we conclude that

$$\begin{aligned} a(t)L'(t)\le &-\eta a(t)E(t)+ca(t) \int _{0}^{1}u_{t}g_{1}(u_{t})\,dx+ca(t) \int _{0}^{1}z(x,1,t)g_{2}\bigl(z(x,1,t) \bigr)\,dx \\ \le & -\eta a(t)E(t)-cE'(t), \quad \forall t\in \mathbb{R}^{+}. \end{aligned}$$

Exploiting (A2) and (3.30), it follows that

$$ L_{0}(t):= a(t)L(t)+cE(t)\sim E(t) $$

and, for some constant \(\eta _{1}>0\), the functional \(L_{0}\) satisfies

$$ L'_{0}(t)\le -\eta _{1}a(t)L_{0}(t), \quad \forall t\ge 0. $$

A simple integration of (4.4) over \((0,t)\), using (4.3), yields

$$ E(t)\le \delta _{1} \exp \biggl( -\delta _{2} \int _{0}^{t}a(s)\,ds \biggr)=\delta _{1} \chi _{1}^{-1} \biggl(\delta _{2} \int _{0}^{t} a(s)\,ds \biggr), \quad \forall t\geq 0. $$

Case II: χ is nonlinear on \([0,r]\). Here, as in [22], we select \(0< r_{1}\le r\) so that

$$ sg_{1}(s)\le \min \bigl\{ r, \chi (r) \bigr\} , \quad \forall \vert s \vert \le r_{1}. $$

On account of (A2) and the continuity of \(g_{1}\) with the fact that \(|g_{1}(s)|>0\), for \(s\ne 0\), we conclude that

$$ \textstyle\begin{cases} s^{2}+g_{1}^{2}(s) \le \chi ^{-1}(sg_{1}(s)), \quad \forall \vert s \vert \le r_{1}, \\ C'_{1} \vert s \vert \le \vert g_{1}(s) \vert \le C'_{2} \vert s \vert , \hfill \quad \forall \vert s \vert \ge r_{1}. \end{cases} $$

Now, we introduce the following partitions:

$$\begin{aligned}& I_{1}= \bigl\{ x\in (0, 1): \vert u_{t} \vert \le r_{1} \bigr\} , \qquad I_{2}= \bigl\{ x\in (0, 1): \vert u_{t} \vert >r_{1} \bigr\} ,\\& \tilde{I_{1}}= \bigl\{ x\in (0, 1): \bigl\vert z(x,1,t) \bigr\vert \le r_{1} \bigr\} ,\qquad \tilde{I_{2}}= \bigl\{ x\in (0, 1): \bigl\vert z(x,1,t) \bigr\vert >r_{1} \bigr\} \end{aligned}$$

and the functional h, defined by

$$ h(t)= \int _{I_{1}}u_{t}g_{1}(u_{t})\,dx. $$

Using the fact that \(\chi ^{-1}\) is concave and Jensen’s inequality, it follows that

$$ \chi ^{-1}\bigl(h(t)\bigr)\ge c \int _{I_{1}}\chi ^{-1} \bigl(u_{t}g_{1}(u_{t}) \bigr)\,dx. $$

Combining (4.7) and (4.8), we have

$$\begin{aligned} a(t) \int _{0}^{1} \bigl(u_{t}^{2}+g_{1}^{2}(u_{t}) \bigr)\,dx&=a(t) \int _{I_{1}} \bigl(u_{t}^{2}+g_{1}^{2}(u_{t}) \bigr)\,dx+a(t) \int _{I_{2}} \bigl(u_{t}^{2}+g_{1}^{2}(u_{t}) \bigr)\,dx \\ &\le a(t) \int _{I_{1}}\chi ^{-1} \bigl(u_{t}g_{1}(u_{t}) \bigr)\,dx+ca(t) \int _{I_{2}}u_{t}g_{1}(u_{t})\,dx \\ &\le ca(t)\chi ^{-1}\bigl(h(t)\bigr)-cE'(t) . \end{aligned}$$


$$ \begin{aligned} a(t) \int _{0}^{1} g_{2}^{2} \bigl(z(x,1,t)\bigr)\,dx={}& a(t) \int _{\bar{I_{1}}}g_{2}^{2}\bigl(z(x,1,t)\bigr)\,dx + a(t) \int _{\bar{I_{2}}}g_{2}^{2}\bigl(z(x,1,t)\bigr)\,dx \\ \leq{}& ca(t) \int _{\bar{I_{1}}}z(x,1,t)g_{2}\bigl(z(x,1,t)\bigr)\,dx \\ &{} +a(t) \int _{\bar{I_{2}}}z(x,1,t)g_{2}\bigl(z(x,1,t)\bigr)\,dx \\ \leq{}& -cE'(t). \end{aligned} $$

Multiplying (3.31) by \(a(t)\) and using (4.9) and (4.10), we obtain

$$ a(t)L'(t)+cE'(t)\le -\eta a(t)E(t)+ca(t)\chi ^{-1}\bigl(h(t)\bigr). $$

It follows from (A1) that

$$ L_{1}'(t)\le -\eta a(t)E(t)+ca(t)\chi ^{-1}\bigl(h(t)\bigr), $$


$$ L_{1}(t)=a(t)L(t)+cE(t)\sim E (t)\quad \text{by virtue of (3.30)}. $$

Let \(r_{0}< r\) and \(\eta _{0}>0\) to be specified later. Then, combining (4.12) and the fact that

$$ E'\le 0,\qquad \chi '>0,\qquad \chi ''>0 \quad \text{on } (0, r], $$

the functional \(L_{2}\), defined by

$$ L_{2}(t):=\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)L_{1}(t)+\eta _{0}E(t), $$


$$ \kappa _{1}L_{2}(t) \le E(t) \le \kappa _{2}L_{2}(t) $$

for some positive constants \(\kappa _{1}\), \(\kappa _{2}\), and

$$\begin{aligned} L_{2}'(t)&=r_{0} \frac {E'(t)}{E(0)}\chi '' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr)L_{1}(t)+\chi ' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr)L_{1}'(t)+ \eta _{0}E'(t) \\ &\le -\eta a(t) E(t)\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)+ \underbrace{ca(t)\chi ' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr)\chi ^{-1} \bigl(h(t) \bigr)}_{A}+ \eta _{0}E'(t). \end{aligned}$$

To estimate the term A in (4.15), we consider the convex conjugate of χ denoted by \(\chi ^{*}\), defined by

$$ \chi ^{*}(y)=y\bigl(\chi ' \bigr)^{-1}(y)-\chi \bigl[\bigl(\chi '\bigr)^{-1}(y) \bigr]\le y\bigl(\chi ^{ \prime}\bigr)^{-1}(y),\quad \text{if }y\in (0, \chi '(r)], $$

and which satisfies the generalized Young’s inequality

$$ XY\le \chi ^{*}(X)+\chi (Y), \quad \text{if }X\in \big(0, \chi '(r)\big], Y \in (0, r]. $$

Taking \(X=\chi ' (r_{0}\frac {E(t)}{E(0)} )\) and \(Y=\chi ^{-1} (h(t) )\) and recalling Lemma 3.1 and (4.6), then (4.15)–(4.17) lead to

$$ \begin{aligned} L_{2}'(t)\le {}& -\eta a(t) E(t) \chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)+ca(t) \biggl[ \chi * \biggl( \chi ' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr) \biggr) +\chi \bigl( \chi ^{-1} \bigl(h(t) \bigr) \bigr) \biggr] \\ &{}+\eta _{0}E'(t) \\ ={}&-\eta a(t) E(t)\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)+ca(t) \chi * \biggl( \chi ' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr) \biggr) \\ &{}+ca(t)h(t)+ \eta _{0}E'(t) \\ \le {}&-\eta a(t) E(t)\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)+ cr_{0}a(t) \biggl(\frac{E(t)}{E(0)} \biggr)\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr) \\ &{}-cE'(t)+ \eta _{0}E'(t) \\ \le {}&-\bigl(\eta E(0)-cr_{0}\bigr)a(t) \biggl(\frac{E(t)}{E(0)} \biggr)\chi ' \biggl(r_{0}\frac {E(t)}{E(0)} \biggr)+(\eta _{0}-c)E'(t). \end{aligned} $$

By choosing \(r_{0}=\frac {\eta E(0)}{2c}\), \(\eta _{0}=2c\), and recalling that \(E'(t)\le 0\), we arrive at

$$\begin{aligned} L_{2}'(t)\le {}&-\eta _{1}a(t) \frac {E(t)}{E(0)}\chi ' \biggl(r_{0} \frac {E(t)}{E(0)} \biggr)=-\eta _{1}a(t)\chi _{0} \biggl( \frac {E(t)}{E(0)} \biggr), \end{aligned}$$

where \(\eta _{1}>0\) and \(\chi _{0}(t)=t\chi '(r_{0}t)\). Now, since χ is strictly convex on \((0,r]\), we conclude that \(\chi _{0}(t)>0\), \(\chi '_{0}(t) >0\) on \((0, 1]\). Using (4.14) and (4.19), it follows that the functional

$$ L_{3}(t)=\frac{\kappa _{1}L_{2}(t)}{E(0)} $$


$$ L_{3}(t)\sim E(t) $$

and, for some \(\delta _{2}>0\),

$$ L_{3}'(t)\le -\delta _{2}a(t) \chi _{0} \bigl(L_{3}(t) \bigr), $$

which yields

$$ \bigl[\chi _{1} \bigl(L_{3}(t) \bigr) \bigr]'\ge \delta _{2}a(t), $$


$$ \chi _{1}(t)= \int _{t}^{1}\frac {1}{\chi _{0}(s)}\,ds, \quad t\in (0,1]. $$

Integrating (4.22) over \([0, t]\), keeping in mind the properties of \(\chi _{0}\), and the fact that \(\chi _{1}\) is strictly decreasing on \((0, 1]\), we obtain

$$ L_{3}(t)\le \chi _{1}^{-1} \biggl( \delta _{2} \int _{0}^{t}a(s)\,ds+ \delta _{3} \biggr), \quad \forall t\in \mathbb{R}^{+}, $$

for some \(\delta _{3}>0\). Using (4.20) and (4.23), the proof of Theorem 4.1 is completed. □


We end this section by giving some examples to illustrate the obtained result.


$$ g_{0}\in C^{2} \bigl( [0,+\infty) \bigr) $$

be a strictly increasing function such that \(g_{0}(0)=0\) and, for some positive constants \(c_{1}\), \(c_{2}\) and r, the function \(g_{1}\) satisfies

$$ \begin{aligned} &g_{0}\bigl( \vert s \vert \bigr)\leq \bigl\vert g_{1}(s) \bigr\vert \leq g_{0}^{-1}\bigl( \vert s \vert \bigr), \quad \forall \vert s \vert \leq r, \\ &c_{1}s^{2}\leq sg_{1}(s)\leq c_{2}s^{2}, \quad \forall \vert s \vert \geq r. \end{aligned} $$

We consider the function

$$ \chi (s)= \biggl( \sqrt{\frac{s}{2}} \biggr)g_{0} \biggl(\sqrt{ \frac{s}{2}} \biggr). $$

It follows that χ is a \(C^{2}\)-strictly convex function on \((0,r]\) when \({\mathrm{g_{0}}}\) is nonlinear and therefore satisfies condition \((A_{2})\). Now, we give some examples of \(g_{0}\) such that \(g_{1}\) satisfies (5.1) near 0.

  1. 1.

    Let \(g_{0}(s)= \lambda s\), where \(\lambda >0\) a constant, then \(\chi (s)= \bar{\lambda} s\), where \(\bar{\lambda}=\frac {\lambda}{2}\) satisfies \((A_{2})\) near 0 and from (4.1), we get

    $$ E(t)\leq \bar{\delta}\exp \biggl( -\delta _{2} \int _{0}^{t}a(s)\,ds \biggr), \quad \forall t\geq 0. $$
  2. 2.

    Let \(g_{0}(s) = \frac {1}{s}e^{-\frac{1}{s^{2}}}\), then \(\chi (s)= e^{-\frac{2}{s}}\) satisfies \((A_{2})\) in the neighborhood of 0 and from (4.1), we obtain

    $$ E(t)\leq \delta _{1} \biggl( \ln \biggl(\delta _{2} \int _{0}^{t} a(s)\,ds+ \delta _{3} \biggr) \biggr)^{-1}, \quad \forall t\geq 0. $$
  3. 3.

    Let \(g_{0}(s) = e^{-\frac{1}{s}}\), then \(\chi (s)= \sqrt{\frac{s}{2}} e^{-\sqrt{\frac{2}{s}}}\) satisfies \((A_{2})\) near 0 and using (4.1), we obtain

    $$ E(t)\leq \delta _{1} \biggl( \ln \biggl(\delta _{2} \int _{0}^{t} a(s)\,ds + \delta _{3} \biggr) \biggr)^{-2}, \quad \forall t\geq 0. $$


In this work, we obtained some general decay results for a thermoelastic Timoshenko beam system with suspenders, general weak internal damping, time-varying coefficient, and time-varying delay terms. The damping structure in system (2.14)–(2.16) is sufficient enough to stabilize the system without any additional conditions on the coefficient parameters as it is the case with many Timoshenko beam systems in the literature. The result of the present paper generalizes the one established in Bochichio et al. [6] and allows a large class of functions that satisfy condition \((A_{2})\). We also gave some examples to illustrate our theoretical finding.

Availability of data and materials

Data sharing is not applicable to this article as no new data were created or analyzed in this study.


  1. Ahmed, N.U., Harbi, H.: Mathematical analysis of dynamic models of suspension bridges. SIAM J. Appl. Math. 109, 853–874 (1998)

    MATH  Google Scholar 

  2. Arnold, V.I.: Mathematical Methods of Classical Mechanics. Springer, New York (1989)

    Book  Google Scholar 

  3. Audu, J., Mukiawa, S.E., Almeida Júnior, D.S.: General decay estimate for a two-dimensional plate equation with time-varying feedback and time-varying coefficient. Results Appl. Math. 12, 100219 (2021)

    Article  MATH  Google Scholar 

  4. Benaissa, A., Benguessoum, A., Messaoudi, S.A.: Energy decay of solutions for a wave equation with a constant weak delay and a weak internal feedback. Electron. J. Qual. Theory Differ. Equ. 11, 13 (2014)

    MATH  Google Scholar 

  5. Benaissa, A., Messaoudi, S.A.: Global existence and energy decay of solutions for the wave equation with a time varying delay term in the weakly nonlinear internal feedbacks. J. Math. Phys. 53, 123514 (2012)

    Article  MATH  Google Scholar 

  6. Bochicchio, I., Campo, M., Fernández, J.R., Naso, M.G.: Analysis of a thermoelastic Timoshenko beam model. Acta Mech. 231(10), 4111–4127 (2020)

    Article  MATH  Google Scholar 

  7. Bochicchio, I., Giorgi, C., Vuk, E.: Long-term damped dynamics of the extensible suspension bridge. Int. J. Differ. Equ. 2010, Article ID 383420 (2010).

    Article  MATH  Google Scholar 

  8. Caraballo, T., Marin-Rubio, P., Valero, J.: Autonomous and non-autonomous attractors for differential equations with delays. J. Differ. Equ. 208, 9–41 (2005)

    Article  MATH  Google Scholar 

  9. Chen, Y.H., Fang, S.C.: Neurocomputing with time delay analysis for solving convex quadratic programming problems. IEEE Trans. Neural Netw. 11, 230–240 (2000)

    Article  Google Scholar 

  10. Enyi, C.D., Feng, B.: Stability result for a new viscoelastic-thermoelastic Timoshenko system. Bull. Malays. Math. Soc. 44, 1837–1866 (2021)

    Article  MATH  Google Scholar 

  11. Enyi, C.D., Mukiawa, S.E.: Decay estimate for a viscoelastic plate equation with strong time-varying delay. Ann. Univ. Ferrara 66, 339–357 (2020)

    Article  MATH  Google Scholar 

  12. Enyi, C.D., Mukiawa, S.E., Apalara, T.A.: Stabilization of a new memory-type thermoelastic Timoshenko system. Appl. Anal. (2022).

    Article  Google Scholar 

  13. Feng, B.: Long-time dynamics of a plate equation with memory and time delay. Bull. Braz. Math. Soc. 49, 395–418 (2018)

    Article  MATH  Google Scholar 

  14. Giorgi, C., Pata, V., Vuk, E.: On the extensible viscoelastic beam. Nonlinearity 21, 713–733 (2008)

    Article  MATH  Google Scholar 

  15. Graff, K.F.: Wave Motion in Elastic Solids. Dover, New York (1975)

    MATH  Google Scholar 

  16. Guesmia, A., Messaoudi, S.A.: General energy decay estimates of Timoshenko systems with frictional versus viscoelastic damping. Math. Methods Appl. Sci. 32(16), 2102–2122 (2009)

    Article  MATH  Google Scholar 

  17. Guesmia, A., Messaoudi, S.A., Soufyane, A.: Stabilization of a linear Timoshenko system with infinite history and applications to the Timoshenko-heat systems. Electron. J. Differ. Equ. 2012(193), 1 (2012)

    MATH  Google Scholar 

  18. Han, S.M., Benaroya, H., Wei, T.: Dynamics of transversely vibrating beams using four engineering theories. J. Sound Vib. 225(5), 935–988 (1999)

    Article  MATH  Google Scholar 

  19. Kato, T.: Linear and quasilinear equations of evolution of hyperbolic type. In: C.I.M.E., II Ciclo, pp. 125–191 (1976)

    Google Scholar 

  20. Kato, T.: Abstract Differential Equations and Nonlinear Mixed Problems Lezioni Fermiane [Fermi Lectures]. Scuola Normale Superiore, Pisa (1985)

    Google Scholar 

  21. Kolmanoviskii, V., Mishkis, A.: Introduction of the Theory and Applications of Functional and Differential Equations, vol. 463. Kluwer Academic, Dordrecht (1999)

    Book  Google Scholar 

  22. Lasiecka, I., Tataru, D.: Uniform boundary stabilization of semilinear wave equations with nonlinear boundary damping. Differ. Integral Equ. 6, 507–533 (1993)

    MATH  Google Scholar 

  23. Lazer, A.C., McKenna, P.J.: Large-amplitude periodic oscillations in suspension bridges: some new connections with nonlinear analysis. SIAM Rev. 32(4), 537–578 (1990)

    Article  MATH  Google Scholar 

  24. Liu, W.J.: General decay rate estimate for the energy of a weak viscoelastic equation with an internal time-varying delay term. Taiwan. J. Math. 17, 2101–2115 (2013)

    Article  MATH  Google Scholar 

  25. McKenna, P.J., Walter, W.: Nonlinear oscillations in a suspension bridge. Arch. Ration. Mech. Anal. 98, 167–177 (1987)

    Article  MATH  Google Scholar 

  26. McKenna, P.J., Walter, W.: Travelling waves in a suspension bridge. SIAM J. Appl. Math. 50(3), 703–715 (1990)

    Article  MATH  Google Scholar 

  27. Mukiawa, S.E.: Decay result for a delay viscoelastic plate equation. Bull. Braz. Math. Soc. 51, 333–356 (2020).

    Article  MATH  Google Scholar 

  28. Mukiawa, S.E.: The effect of time-varying delay damping on the stability of porous elastic system. Open J. Math. Sci. 5(1), 147–161 (2021)

    Article  Google Scholar 

  29. Mukiawa, S.E.: Stability result of a suspension bridge problem with time-varying delay and time-varying weight. Arab. J. Math. (2021).

  30. Mukiawa, S.E.: A new optimal and general stability result for a thermoelastic Bresse system with Maxwell–Cattaneo heat conduction. Results Appl. Math. 10, 100152 (2021).

    Article  MATH  Google Scholar 

  31. Nicaise, S., Pignotti, C.: Stability and instability results of the wave equation with a delay term in the boundary or internal feedbacks. SIAM J. Control Optim. 45, 1561–1585 (2006)

    Article  MATH  Google Scholar 

  32. Nicaise, S., Pignotti, C.: Interior feedback stabilization of wave equations with time dependence delay. Electron. J. Differ. Equ. 2011, 41 (2011)

    MATH  Google Scholar 

  33. Nicaise, S., Pignotti, C., Valein, J.: Exponential stability of the wave equation with boundary time-varying delay. Discrete Contin. Dyn. Syst. 4, 693–722 (2011)

    MATH  Google Scholar 

  34. Richard, J.P.: Time-delay systems: an overview of some recent advances and open problems. Automatica 39, 1667–1694 (2003)

    Article  MATH  Google Scholar 

  35. Timoshenko, S.: On the correction for shear of the differential equation for transverse vibrations of prismatic bars. Philos. Mag. 41(245), 744–746 (1921)

    Article  Google Scholar 

Download references


The authors gratefully acknowledge the technical and financial support from the Ministry of Education and University of Hafr Al Batin, Saudi Arabia.


This research work was funded by institutional fund projects under No. IFP-A-2022-2-1-04.

Author information

Authors and Affiliations



All the authors have equal contributions in this paper. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Soh Edwin Mukiawa.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Mukiawa, S.E., Enyi, C.D. & Messaoudi, S.A. Stability of thermoelastic Timoshenko beam with suspenders and time-varying feedback. Adv Cont Discr Mod 2023, 7 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • 35D30
  • 74K10
  • 74H20
  • 74H40
  • 65M60
  • 65M15


  • General decay
  • Timoshenko beam
  • Thermoelasticity
  • Suspenders
  • Time delay