Loading [MathJax]/jax/output/SVG/jax.js
Research article

Harmonic and subharmonic solutions of quadratic Liénard type systems with sublinearity

  • Received: 12 July 2021 Accepted: 06 September 2021 Published: 10 September 2021
  • MSC : 34A12, 34C25

  • In this paper, we prove the existence of harmonic solutions and infinitely many subharmonic solutions of dissipative second order sublinear differential equations named quadratic Liénard type systems. The method of the proof is based on the Poincaré-Birkhoff twist theorem.

    Citation: Chunmei Song, Qihuai Liu, Guirong Jiang. Harmonic and subharmonic solutions of quadratic Liénard type systems with sublinearity[J]. AIMS Mathematics, 2021, 6(11): 12913-12928. doi: 10.3934/math.2021747

    Related Papers:

    [1] Yinyin Wu, Fanfan Chen, Qingchi Ma, Dingbian Qian . Subharmonic solutions for degenerate periodic systems of Lotka-Volterra type with impulsive effects. AIMS Mathematics, 2023, 8(9): 20080-20096. doi: 10.3934/math.20231023
    [2] Yinyin Wu, Dingbian Qian, Shuang Wang . Periodic bouncing solutions for sublinear impact oscillator. AIMS Mathematics, 2021, 6(7): 7170-7186. doi: 10.3934/math.2021420
    [3] Chunlian Liu, Shuang Wang . Multiple periodic solutions of second order parameter-dependent equations via rotation numbers. AIMS Mathematics, 2023, 8(10): 25195-25219. doi: 10.3934/math.20231285
    [4] Yan Ning, Daowei Lu, Anmin Mao . Existence and subharmonicity of solutions for nonsmooth p-Laplacian systems. AIMS Mathematics, 2021, 6(10): 10947-10963. doi: 10.3934/math.2021636
    [5] Sabbavarapu Nageswara Rao, Abdullah Ali H. Ahmadini . Multiple positive solutions for system of mixed Hadamard fractional boundary value problems with (p1,p2)-Laplacian operator. AIMS Mathematics, 2023, 8(6): 14767-14791. doi: 10.3934/math.2023755
    [6] Qingtian Shi . Equivalent characterizations of harmonic Teichmüller mappings. AIMS Mathematics, 2022, 7(6): 11015-11023. doi: 10.3934/math.2022615
    [7] Chunlian Liu, Shuang Wang, Fanfan Chen . Resonance with Landesman-Lazer conditions for parameter-dependent equations: a multiplicity result via the Poincaré-Birkhoff theorem. AIMS Mathematics, 2024, 9(10): 28320-28340. doi: 10.3934/math.20241374
    [8] Ye Xue, Zhiqing Han . Existence and multiplicity of solutions for Schrödinger equations with sublinear nonlinearities. AIMS Mathematics, 2021, 6(6): 5479-5492. doi: 10.3934/math.2021324
    [9] Yuyan Wei, Xili Tan, Peiyu Sun, Shuang Guo . Weak and strong law of large numbers for weakly negatively dependent random variables under sublinear expectations. AIMS Mathematics, 2025, 10(3): 7540-7558. doi: 10.3934/math.2025347
    [10] Shaoyuan Xu, Yan Han, Qiongyue Zheng . Fixed point equations for superlinear operators with strong upper or strong lower solutions and applications. AIMS Mathematics, 2023, 8(4): 9820-9831. doi: 10.3934/math.2023495
  • In this paper, we prove the existence of harmonic solutions and infinitely many subharmonic solutions of dissipative second order sublinear differential equations named quadratic Liénard type systems. The method of the proof is based on the Poincaré-Birkhoff twist theorem.



    Generally speaking, the equations of dissipative systems usually contain the first derivative term, which reflects the energy loss of the system. Second order dissipative differential equations arise widely in various research fields such as celestial mechanics, fluid mechanics [1,2], relativistic mechanics [3,4], engineering [5,6].

    The purpose of this paper is to study a class of special dissipative second order differential equations, named a quadratic Liénard type systems in [7],

    ¨x+f(x)(˙x)2+g(x)=p(t), (1.1)

    where p(t) is a continuous periodic function (let T>0 be its minimum period), and f(x),g(x) are local Lipschitz continuous functions. Equation (1.1) models one-dimensional oscillator studied at the classical and also at the quantum level [8].

    Equation (1.1) is a special case of the form

    ¨u+f(u)(˙u)2=˜g(t,u),

    which arises from nonlinear elastic mechanics [9]. Take F(u) such that F(u)=f(u), then multiplying by eF(u), the latter equation is written as

    ddt[eF(u)˙u]=eF(u)˜g(t,u),

    which in turn is just

    d2dt2[Ψ(u)]=eF(u)˜g(t,u),

    where Ψ(u)=eF(u). Now, the change x=Ψ(u) leads to the desired newtonian formulation ¨x=˜f(t,x). For the nonlinear vibrations of a radially forced thick-walled hollow sphere made of an elastic, homogeneous, isotropic, and incompressible material, Del Pino and Manásevich [10] have shown that ˜f has the strong singularity at the origin and the superlinearity at infinite by an asymptotic analysis. Then by Poincaré-Birkhoff twist theorem, they have prove the existence of infinitely many periodic solutions. Another concrete example of this equation is the Rayleigh-Plesset equation

    ρx¨x+32(˙x)2=pB(t)p(t).

    We can refer to [11,12] for the related development.

    When Eq (1.1) has no dissipative term, it is just a Duffing equation

    ¨x+g(x)=p(t).

    Under various conditions such as superlinearity, sublinearity and semilinearity, the existence and multiplicity of periodic solutions has been established, see [13,14,15] for instance. We can refer to [16,17,18,19] for more research on the problem of periodic solutions. When the system has nonconvex potentials or periodic nonlinearities, the existence of subharmonic solutions has been established in [20,21], where the authors exploited different techniques by using critical point theory. Moreover, with a similar approach based on the Poincaré-Birkhoff fixed point theorem, the existence of subharmonic solutions for different second-order differential systems has been widely studied by Zanolin and his collaborators, see [22,23,24,25].

    When the forced term p vanishes, Eq (1.1) is an autonomous equation

    ¨x+f(x)(˙x)2+g(x)=0. (1.2)

    Owing to the second term f(x)(˙x)2 of (1.1), equation is not conservative in usual phase space (x,˙x). However, in the generalized coordinates (x,p), Eq (1.1) can be transformed into a Hamiltonian system whose energy is conserved; see [26] for details.

    In case of

    f(x)=λx1λx2,g(x)=αx21λx2,

    Eq (1.2) is a one-dimensional Mathews-Lakshmanan (ML) oscillator with real constant parameters λ,α (see [27,28]). The ML oscillator can be regarded as the zero-dimensional version of a scalar nonpolynomial field equation or as a velocity dependent potential oscillator. The kinetic term in its Hamiltonian function features a position-dependent mass term that produces a variable "spring constant" of the oscillator. As a consequence, the classical Euler-Lagrange equation associated with the model admits simple, sinusoidal solutions [29]. The ML oscillator exhibits simple harmonic periodic solutions but with amplitude dependent frequency,

    x(t)=Acos(Ωt+δ),Ω=α1λA2,

    where A is the amplitude and δ is the initial phase. We can refer to [28] for details.

    When the functions f(x) and g(x) are of class C1, a sufficient condition for the monotonicity of the period T or for the isochronicity of the origin O has been established for Eq (1.2) in [30] by Sabatini. In the analytic case of f(x) and g(x), Chouikha has given a necessary and sufficient condition for the isochronicity of the origin O; see [31]. We can refer to [32,33] for the development. Moreover, a complete classification of the Lie point symmetry groups is given in [7] for Eq (1.2).

    Recently, Atslega [34] provides some conditions on the functions f(x) and g(x) which ensure the existence of solutions of Eq (1.2) with the Neumann boundary conditions

    ˙x(0)=0,˙x(1)=0.

    A natural problem is, if the periodic external force p(t) is added, how about the existence and multiplicity of solutions for Eq (1.1) under the periodic boundary value condition

    x(0)=x(mT),˙x(0)=˙x(mT).

    Comparing with the case considered by Del Pino and Manásevich [10], we do not give any assumption of singularity.

    In this paper, we consider the sublinear case, and need the following conditions:

    (H1) g(x) satisfies the sublinear condition

    lim|x|g(x)x=0.

    (H2) F(x) and xf(x) are bounded on (,+), where F(x) is a prime function of f(x);

    (H3)for all x0, there is xf(x)<0.

    (H4) limx+g(x)=+, limxg(x)=.

    We remark that condition (H4) implies the sign condition of g, that is, there exist d>0 such that xg(x)>0, for all |x|>d. Without loss of generality, in view of assumption (H2) we assume that 0F(x)a with a positive constant a>0. Moreover, in condition (H2), if f(x) is a monotone function, then the boundedness of xf(x) holds naturally. In fact, the boundedness of F implies that infinite integral

    +0f(x)dx,0f(x)dx

    are all convergent, and the monotonicity of f yields the limit limxxf(x)=0.

    The condition (H3) is only a sign condition of the functions f(x) and g(x). A simple example which satisfies all conditions above is given in the following

    ¨xx1+x4(˙x)2+x1/3=sint.

    In order to state our main results, we recall some definitions form [35]. Assume x(t) is a periodic solution of Eq (1.1) with its minimum positive period T0. If T0=T, we call x(t) is a harmonic solution; if T0=mT with some positive integer m2, we call x(t) is a m-order subharmonic solution.

    Now we give our main results as follows.

    Theorem 1.1. Assume that conditions (H1)(H4) hold, then there exists a sufficiently large positive integer m0 such that, for each positive integer mm0, Eq (1.1) possesses at least two distinct m-order subharmonic solution.

    Theorem 1.2. Assume that conditions (H1)(H4) hold, then Eq (1.1) has at least one harmonic solution.

    Theorem 1.1 implies that Eq (1.1) possesses infinitely many subharmonic solutions. The proof of Theorem 1.1 is based on the Poincaré-Birkhoff twist theorem, and the main difficulty is how to verify the twist property of the Poincaré map corresponding to Eq (1.1). Compared with the superlinear case [36], when g(x) satisfies the sublinear condition, the rotation speed of the solutions of (1.2) around the origin is very slow outside a sufficiently large disk.

    The paper is organized as follows. Preliminaries and some lemmas which are useful for proving our theorems are stated in Section 2. We will prove Theorem 1.1 and 1.2 in Section 3.

    Firstly, Eq (1.1) is a special case of Benoulli type equation with n=1 and p(t)0 (see [26]), whose characteristic equation is

    d˙xdx=f(x)˙xg(x)˙x,

    where we regard t as an independent parameter variable. Then we have the first integral of (1.2)

    I(x,˙x)=(˙x)2e2F(x)+2x0[g(s)]e2F(s)ds.

    In the following, we introduce the generalized coordinates and the generalized momentum. The generalized coordinates x is used to describe a conservative mechanical system whose configuration is completely specified by the value of a certain single variable x. The variable need not represent a displacement; it might, for example, be an angle, or even the reading on a dial forming part of the system. The generalized momentum y is a function of both the generalized velocity and generalized coordinates; see [37, p.31 and p.374] or [26].

    Let

    y=2˙xe2F(x)+φ(x),

    then the Hamiltonian function corresponding to (1.1) is given by

    H(x,y)=14e2F(x)y2+2x0g(s)e2F(s)ds2x0e2F(s)p(t)ds.

    Here, we take φ(x)0. For simplification, we rewrite the Hamiltonian function in the following form

    H(t,x,y)=14e2F(x)y2+V(x)+P(t,x), (2.1)

    where

    V(x)=x02e2F(s)g(s)ds,P(t,x)=x02e2F(s)p(t)ds.

    The corresponding Hamiltonian system is defined by

    {˙x=12e2F(x)y,˙y=12e2F(x)f(x)y22e2F(x)(g(x)p(t)). (2.2)

    Lemma 2.1. The existence of periodic solutions of Eq (1.1) is equivalent to the existence of periodic solutions of Hamiltonian system (2.2).

    Proof. Equation (1.1) is equivalent to the following plane differential system

    {˙x=v,˙v=f(x)v2g(x)+p(t). (2.3)

    By introducing a global differential homeomorphism

    x=x,y=2e2F(x)v,

    system (1.1) is transformed into (2.2). Thus, we complete the proof.

    Lemma 2.2. Assume that conditions (H2)(H4) hold, then the solutions of Eq (2.2) are well defined on (,+).

    Proof. Assume that (x(t;x0,y0),y(t;x0,y0)) is a solution of Eq (2.2) with the initial value condition (x(t0),y(t0))=(x0,y0), which is defined on the maximum existence interval (α,β). Let

    V(t)=14e2F(x(t;x0,y0))y2(t;x0,y0)+V(x(t;x0,y0)),

    then we have

    dV(t)dt=P(t,x)x˙x=2e2F(x)p(t)(12e2F(x)y)=yp(t).

    From the conditions (H2) and (H3), we have

    |dV(t)dt|=|yp(t)|12y2(t)+|p|212e2F(a)2F(x(t))y2(t)+|p|2C1[14e2F(x(t))y2(t)+V(x(t))]+C2C1V(t)+C2, (2.4)

    where C1,C2 are positive constants, since V(x)+ as x.

    If β<+, by the continuation of solutions, we have

    limtβ(|x(t)|+|y(t)|)=+.

    In view of condition (H4), it follows that limtβV(t)=+. By using the Grownwall inequality for (2.4), for all t[t0,β), we know that

    V(t)C2C1+eC1(tt0)V(t0)+C2C1eC1(tt0),

    which implies V(t) is bounded on [t0,β) since β is finite. Therefore, we obtain a contradiction and β=+.

    If α>, with the same argument we obtain

    V(t)C2C1+eC1(t0t)V(t0)+C2C1eC1(t0t),

    which also is a contradiction.

    Thus the proof of the lemma is completed.

    Let (x(t;x0,y0),y(t;x0,y0)) be the solution of Eq (2.2) with the initial value (x(t0;x0,y0),y(t0;x0,y0))=(x0,y0). Let

    x1=x(T;x0,y0),y1=y(T;x0,y0).

    By Lemma 2.2, the Poincaré mapping of (2.2)

    P:(x0,y0)(x1,y1)

    is well defined. Moreover, it is a area-preserving smooth homeomorphism in the phase plane since (2.2) is a Hamiltonian system.

    With the same argument as [38,39], we have the elastic property for system (2.2).

    Lemma 2.3. For any constants d>0 and L>0, there is a sufficiently large constant c>d, such that the solution (x(t;x0,y0),y(t;x0,y0)) of system (2.2) satisfies the inequality

    x2(t;x0,y0)+y2(t;x0,y0)d2,t[t0,t0+L], (2.5)

    if the initial value (x0,y0) satisfies the condition

    x20+y20c2.

    If the solution does not pass through the origin, we can use the polar coordinates to represent the solution. Now we transform system (2.2) into the polar coordinates, that is,

    {˙r=12rsinθcosθe2F(rcosθ)+12r2sin3θe2F(rcosθ)f(rcosθ)2sinθe2F(rcosθ)(g(rcosθ)p(t)),˙θ=12rsin2θcosθe2F(rcosθ)f(rcosθ)2cosθre2F(rcosθ)(g(rcosθ)p(t))12sin2θe2F(rcosθ). (2.6)

    From Lemma 2.3, if r0>c, the solution can be written in the form of polar coordinates

    r(t)=r(t;r0,θ0)>d,θ(t)=θ(t;r0,θ0),t[0,T], (2.7)

    which satisfies the initial value r(0)=r0,θ(0)=θ0. Furthermore, we have

    r(t;r0,θ0+2π)=r(t;r0,θ0),θ(t;r0,θ0+2π)=θ(t;r0,θ0)+2π.

    Then the function r(t;r0,θ0) is the period of 2π for θ0, and the function θ(t;r0,θ0) is 2π-appreciation with respect to θ0.

    When r0>c, we can get the Poincaré mapping P in the polar coordinate form

    r1=r0+R(r0,θ0),θ1=θ0+Θ(r0,θ0),

    where

    R(r0,θ0)=r(T;r0,θ0)r0,Θ(r0,θ0)=θ(T;r0,θ0)θ0.

    Lemma 2.4. Assume that conditions (H3) and (H4) hold, then there exists a sufficiently large disc

    Dd={(x,y)|x2+y2d2}

    such that the trajectory of (2.6) outside the disc Dd rotates clockwise around the origin O.

    Proof. By the second equality of (2.6), we have

    dθdt=12sin2θe2F(x)(xf(x)1)2cos2θe2F(x)(g(x)p(t))x.

    From conditions (H3) and (H4), there exists a positive constant N such that

    g(x)p(t)x>0,|x|N.

    Moreover, from (H3) we have xf(x)<0, for all |x|N. Therefore, when |x|N, we have

    dθ(t)dt<0.

    When |x|N, we have

    dθdt=12sin2θe2F(x)(xf(x)1)2e2F(x)(g(x)p(t))x2+y2x.

    Notice that if d is large enough, for |x|N, we have θ[π/4,3π/4][5π/4,7π/4] so that sin2θ is uniformly positive. On the other hand, for any ε>0, we can take a sufficiently large d, we obtain

    |(g(x)p(t))x2+y2x|ε.

    Consequently, we have

    dθ(t)dt<0,t[0,T]. (2.8)

    By combining Lemma 2.3 and Lemma 2.4, we know that if the initial value (θ0,r0) such that r0>c, then for t[0,L], the solution of (2.6) rotates clockwise around the origin O and is always outside the disc Dd.

    Denote by Δτ(r0,θ0) the time for which the solution (θ(t;r0,θ0),r(t;r0,θ0)) of system (2.2) rotates one around the origin.

    Lemma 3.1. Assume that conditions (H1)(H4) hold, we have

    limr0+Δτ(r0,θ0)=+.

    Proof. Taking two large enough constants A>0 and B>0 such that

    AB<ε1,

    we consider the following regions respectively:

    D1={(x,y)R2:|x|A,y>B},D2={(x,y)R2:xA,|y|<},D3={(x,y)R2:|x|A,y<B},D4={(x,y)R2:xA,|y|<},

    and D=D1D2D3D4Dd, where Dd is defined in Lemma 2.4. According to Lemma 2.3 and Lemma 2.4, when r0 is sufficiently large, the solution of (2.6) rotates clockwise around the origin O and is always outside the region D. Let [t1,t2], [t2,t3] be the time intervals for the solution staying at D1, D2, respectively, and so on.

    From the first equality of Eq (2.2), we obtain

    t2t1=AA2e2F(x)ydx<4Ae2aB.

    Similarly, we have

    t4t3=AA2e2F(x)ydx<4Ae2aB.

    From the second equality of (2.6), we obtain

    dθdt=12rsin2θcosθe2F(rcosθ)f(rcosθ)2cosθe2F(rcosθ)(g(rcosθ)p(t))r12sin2θe2F(rcosθ).

    Let

    F=dθdt.

    Then we have

    Δτ(r0,θ0)=2π01Fdθ.

    By condition (H1), for any positive constant δ1, we can take B1 such that

    0<g(x)p(t)x<δ,|x|B.

    We take AB to ensure that θ(t)[π/4,π/4] for all t[t2,t3]. Moreover, xf(x) is bounded on x(,+). Then, for K>0 large enough such that xf(x)K, we have

    t3t2=θ(t2)θ(t3)dθHπ4π4dθ12sin2θ+12Ksin2θ+2e2aδcos2θ=π(1+K)e2aδ.

    Similarly, we have

    t5t4π(1+K)e2aδ.

    Consequently, if 0<δ1, then

    Δτ(r0,θ0)=(t2t1)+(t3t2)+(t4t3)+(t5t4)2π(1+K)e2aδ1.

    Thus we complete the proof.

    In the proof of Lemma 3.1, it seems that the estimates of time intervals [t1,t2] and [t3,t4] are not needed. However, it has shown that the main part of Δτ(r0,θ0) is the solution staying at D2 and D4.

    Proof of Theorem 1.2. By using Lemma 3.1, we conclude that the time required for the solution rotating clockwisely one around the origin is sufficiently large.

    Therefore, when |z0|=a is sufficiently large, θ(t,z0) satisfies the condition

    2π<θ(T,z0)θ(0,z0)<0. (3.1)

    Then the Poincaré mapping P of Eq (2.2) satisfies the conditions of the Poincaré-Bohl fixed point theorem on the disk Da. Therefore, P has at least one fixed point ζ0, then z=z(t,ζ0) is the harmonic solution of Eq (2.2).

    Thus we complete the proof.

    According to Theorem 1.2, Eq (2.2) has at least one harmonic solution z=z0(t)=(x0(t),y0(t)). Let

    x=u+x0(t),y=v+y0(t).

    Substituting into Eq (2.2), we have

    {˙u=12ve2F(u+x0(t))+12y0(t)[e2F(u+x0(t))e2F(x0(t))],˙v=12e2F(u+x0(t))f(u+x0(t))(y0(t)+v)22e2F(u+x0(t))g(u+x0(t))+2p(t)[e2F(u+x0(t))e2F(x0(t))]+2e2F(x0(t))g(x0(t))12e2F(x0(t))f(x0(t))y20(t). (3.2)

    Obviously, Eq (3.2) has a trivial solution (u,v)=(0,0). From the uniqueness of the solution, if (u0,v0)0, we have (u(t;u0,v0),v(t;u0,v0))(0,0) for all tR. Therefore, it can be written in the form of polar coordinates

    Λ:u(t)=ρ(t)cosφ(t),v(t)=ρ(t)sinφ(t),

    where ρ(t)>0 and φ(t) are continuous functions of t.

    Lemma 3.2. Let t1>0 and assume that the angle φ of Λ satisfies the condition

    φ(t1)φ(0)<2Nπ. (3.3)

    Then for any t2>t1, we have

    φ(t2)φ(0)<2Nπ+π. (3.4)

    Proof. From the first equality of Eq (3.2), when Λ intersects with the positive half axis of the v-axis, that is, u=0 and v>0, then

    ˙u(t)=12ve2F(x0(t))>0.

    When Λ intersects with the negative half axis of the v-axis, that is, u=0 and v<0, then ˙u(t)<0. Therefore, when the trajectory Λ intersects with the v-axis, it crosses in a clockwise direction. That is, when Λ goes from the positive (or negative) v-axis to the negative (or positive) v-axis, the angle φ of Λ gets an increment of π. And when Λ is in the right (or left) half plane, no matter how active, the increment of the angle will not exceed +π. Therefore, when t1tt2, we have

    φ(t2)φ(t1)<π.

    By using inequality (3.3), we get

    φ(t2)φ(0)=(φ(t1)φ(0))+(φ(t2)φ(t1))<2Nπ+π.

    Therefore, the proof of Lemma 3.2 is now completed.

    Let

    x=x(t)=u(t)+x0(t),y=y(t)=v(t)+y0(t)

    be the solution of Eq (2.2) on the (x,y) plane. The trajectory on the plane of (x,y) is also denoted as Λ.

    Consider the moving point

    P(t)=(x(t),y(t))Λ,Q(t)=(x0(t),y0(t))Λ0(tR),

    and the triangle ΔOPQ, where O is the origin on the plane (x,y) and Q is the origin on the plane (u,v). Let the constant c0 satisfy

    c0>sup0tTx20(t)+y20(t). (3.5)

    That is, the closed orbit Λ0 is within the disk Dc0(O). Then when the poin P(t) is outside the disk Dc0(O) (i.e. r(t,z0)=|P(t)|>c0), the angle ΔOPQ at the vertex P of the triangle σ(t) is an acute angle. Therefore, using |θ(t,ζ0)φ(t)|=σ(t), we can derive

    |θ(τ2,ζ0)θ(τ1,ζ0)|<|φ(τ2)φ(τ1)|+π2(τ2>τ1>0). (3.6)

    Lemma 3.3. Let 0<t1<mT. Suppose the argument θ(t,z0) of the trajectory Λ satisfies

    θ(t1,z0)θ(0,z0)<2Nπ. (3.7)

    Then

    θ(mT,z0)θ(0,z0)<(2N32)π. (3.8)

    Proof. From Lemma 3.2, for any t2>t1, we have

    φ(t2)φ(t1)<π.

    According to inequalities (3.6) and (3.7), there is

    θ(t2,z0)θ(0,z0)|θ(t2,z0)θ(t1,z0)|+|θ(t1,z0)θ(0,z0)||φ(t2)φ(t1)|+π2+(2Nπ)<(2N32)π.

    Take t2=mT, then we complete the proof of Lemma 3.3.

    Now we began to prove Theorem 1.1 as follows.

    Proof of Theorem 1.1. The proof follows the method of Fonda, Manásevich and Zanolin (see [25]), and we also can refer to the book [35]. The proof is essentially the same as the one in [25], however, for the sake of the integrity of the article, we repeat this process in our setting.

    First of all, according to Lemma 2.4, we know that there exists a constant c0>0 such that, for any trajectory of Eq (2.2), namely,

    Γ:z=(x(t),y(t))=(r(t,z0)cosθ(t,z0),r(t,z0)sinθ(t,z0)),

    we have

    θ(t)<0,ifr(t,z0)c0. (3.9)

    Assume that inequality (3.5) holds, where (x0(t),y0(t)) is a harmonic solution of Eq (2.2).

    Secondly, for an arbitrarily large t1>0, there is a sufficiently large d0>c0 such that when 0tt1, we get

    r(t,z0)>c0,θ(t)<0,if|z0|d0.

    Thus

    θ(t1,z0)θ(0,z0)<2Nπ, (3.10)

    where the integer N0.

    Next we have to prove that the constant N of inequality (3.10) can be arbitrarily large, if t1 is sufficiently large.

    In fact, assume by contradiction that there is a constant K>0 such that

    θ(t1,z0)θ(0,z0)>K(t11). (3.11)

    There include two situations. If t1+, r(t1,z0)+, then we obtain

    limtθ(t,z0)=θ>,

    that is, the trajectory Γz0 takes the ray θ=θ as the asymptote. From the second equality of Eq (2.6), we have

    dθdt=12sin2θe2F(x)xf(x)2cos2θe2F(x)(g(x)p(t))x12sin2θe2F(x).

    Thus we have θ=kπ, where k is a certain integer. That is, the trajectory Γz0 takes the positive x-axis (or negative x-axis) as the asymptotes. Let the positive x axis be the asymptote. Therefore, the tangent slope of the trajectory Γz0 has a limit

    limxdydx=limx12e2F(x)f(x)y22e2F(x)(g(x)p(t))12e2F(x)y=0.

    which yields that, when x+, we have y0. It contradicts with the condition (H4), and the above situation cannot happen.

    We consider the other case. Suppose c0>0, which is given by (3.5). Then there is t11 such that r(t1,z0)=c0. Since |z0| can be sufficiently large, there exists a trajectory Γ starting from the circle |z|=c0 with a negative direction asymptote, which is similar to the first case. Then we can also deduce contradictions. This proves that inequality (3.11) does not hold.

    Therefore, the constant N>0 in the inequality (3.10) can be arbitrarily large, if t1 is sufficiently large.

    Then, take an appropriately large constant a0>c0, so that the initial value of the trajectory Γ which satisfies |z(0)|=a0 has the following properties:

    (P1) For a given prime number Q2, there is t1>0 such that

    |z(t,z0)|>c0(0t<t1),
    |z(t1,z0)|=c0,θ(t1,z0)θ(0,z0)<(2Q+2)π.

    or

    (P2) For any sufficiently large t1>0, such that

    |z(t,z0)|>c0(0t<t1),
    θ(t1,z0)θ(0,z0)<(2Q+2)π.

    Let

    E={t1>0:theproperty(P1)holds}.

    Then when E, we have an upper bound

    t1=supt1Et1.

    And when E= (that is, for any t1>0, the property (P2) holds). Let t1=0 and take the integer

    m0=max{2,t1}.

    Then when mm0, from Lemma 3.3, we obtain

    θ(mT,z0)θ(0,z0)<(2(Q+1)32)π<(2Q+12)π,|z(0)|=a0. (3.12)

    On the other hand, from Lemma 3.1, there is a sufficiently large constant bm>0(bm>a0) such that

    2π<θ(mT,z0)θ(0,z0)<0,|z(0)|=bm. (3.13)

    Considering the annular domain

    Am:a20x2+y2b2m,

    we denote the m iterations of the Poincaré-Birkhoff of Eq (2.2) as Pm. Obviously, the composition of the Poincaré maps for Hamiltonian Systems is an area-preserving homeomorphism. And by (3.12) and (3.13), it is twisted on the annular domain Am. Therefore, according to the Poincaré-Birkhoff twist theorem, Pm has at least two fixed points ζ(k)m(k=1,2) in Am, and satisfies the conditions

    θ(mT,ζ(k)m)=2Qπ,(k=1,2).

    where Q2 is a prime number. Obviously, z=zm(t,ζ(k)m) is the mT periodic solution of Eq (2.2).

    With the same argument in [25,35], owing to the fact that Q is prime and Q2, we know that mT is the minimum period of z=zm(t,ζ(k)m). Thus we complete the proof of Theorem 1.1.

    Based on the Poincaré-Birkhoff twist theorem, we have proved the existence of harmonic solutions and infinitely many subharmonic solutions of dissipative second order sublinear differential equations named quadratic Liénard type systems.

    This work is partially supported by National Natural Science Foundation Grant No. 11771105, Guangxi Natural Science Foundation Nos. 2017GXNSFFA198012, 2018GXNSFAA138177, Guangxi Distinguished Expert Project and Innovation Project of Guangxi Graduate Education (No. YCSW2020157).

    The authors declare that they have no conflicts of interest.



    [1] K. Johannessen, The Duffing oscillator with damping, Eur. J. Phys., 36 (2015), 065020. doi: 10.1088/0143-0807/36/6/065020
    [2] B. Baumann, J. Schwieger, M. Wolff, F. Manders, J. Suijker, Nonlinear behavior in high-intensity discharge lamps, J. Phys. D App. Phys., 49 (2016), 255201. doi: 10.1088/0022-3727/49/25/255201
    [3] Q. Liu, L. Huang, G. Jiang, Periodic oscillations of the relativistic pendulum with friction, Electron. J. Differ. Eq., 2017 (2017), 1–10. doi: 10.1186/s13662-016-1057-2
    [4] J. A. Cid, On the existence of periodic oscillations for pendulum-type equations, Adv. Nonlinear Anal., 10 (2020), 121–130. doi: 10.1515/anona-2020-0222
    [5] L. Gao, J. He, Y. Xu, C. Zhang, Application of Duffing oscillator in fault diagnosis, Mech. Des. Manuf., 3 (2009), 77–79.
    [6] Q. Liu, Z. Wang, Periodic impact behavior of a class of Hamiltonian oscillators with obstacles, J. Math. Anal. App., 365 (2010), 67–74. doi: 10.1016/j.jmaa.2009.09.054
    [7] A. K. Tiwari, S. N. Pandey, M. Senthilvelan, M. Lakshmanan, Classification of Lie point symmetries for quadratic Lienard type equation ¨x+f(x)(˙x)2+g(x)=0, J. Math. Phy., 54 (2013), 053506. doi: 10.1063/1.4803455
    [8] V. Marinca, N. Herisanu, Nonlinear Dynamical Systems in Engineering: Some Approximate Approaches, Berlin: Springer-Verlag, 2011.
    [9] P. J. Torres, Mathematical Models with Singularities, Paris: Atlantis Press, 2015.
    [10] M. A. Delpino, R. F. Manasevich, Infinitely many T-periodic solutions for a problem arising in nonlinear elasticity, J. Differ. Equations, 2 (1993), 260–277.
    [11] K. N. Shukla, A generalization of the Rayleigh-Plesset equation of bubble dynamics, ZAMM-Z. Angew Math. Me., 67 (2010), 470–471.
    [12] F. Hegedűs, C. Hős, L. Kullmann, Stable period 1, 2 and 3 structures of the harmonically excited Rayleigh-Plesset equation applying low ambient pressure, IMA J. Appl. Math., 78 (2013), 1179–1195. doi: 10.1093/imamat/hxs016
    [13] W. Ding, The fixed point of torsion map and periodic solution of ordinary differential equation, Acta Math. Sinica., 25 (1982), 227–235.
    [14] T. Ding, Study of the sublinear Duffing equation, Acta Math. Appl. Sin., 12 (1989), 449–455.
    [15] T. Ding, Nonlinear oscillations at a point of resonance, Sci. China Ser. A, 1 (1982), 1–13.
    [16] A. Fonda, F. Zanolin, On the use of time-maps for the solvability of nonlinear boundary value problems, Arch. Math., 59 (1992), 245–259. doi: 10.1007/BF01197322
    [17] D. Qian, Infinity of subharmonics for asymmetric Duffing equations with the Lazer-Leach-Dancer condition, J. Differ. Equations, 171 (2001), 233–250. doi: 10.1006/jdeq.2000.3847
    [18] Z. Wang, Lazer–Leach type conditions on periodic solutions of semilinear resonant Duffing equations with singularities, Z. Angew Math. Phys., 65 (2014), 69–89. doi: 10.1007/s00033-013-0323-3
    [19] A. Fonda, A. Sfecci, Periodic solutions of weakly coupled superlinear systems, J. Differ. Equations, 260 (2016), 2150–2162. doi: 10.1016/j.jde.2015.09.056
    [20] A. Fonda, A. C. Lazer, Subharmonic solutions of conservative systems with nonconvex potentials, Proc. Amer. Math. Soc., 115 (1992), 183–190. doi: 10.1090/S0002-9939-1992-1087462-X
    [21] E. Serra, M. Tarallo, S. Terracini, Subharmonic solutions to second-order differential equations with periodic nonlinearities, Nonlinear Anal. Theor., 41 (2000), 649–667. doi: 10.1016/S0362-546X(98)00302-2
    [22] A. Boscaggin, F. Zanolin, Subharmonic solutions for nonlinear second order equations in presence of lower and upper solutions, Discrete Cont. Dyn. Sys., 33 (2013), 89–110. doi: 10.3934/dcds.2013.33.89
    [23] A. Boscaggin, G. Feltrin, Positive subharmonic solutions to nonlinear ODEs with indefinite weight, Commun. Contemp. Math., 20 (2018), 1750021. doi: 10.1142/S0219199717500213
    [24] A. Fonda, R. Toader, Subharmonic solutions of Hamiltonian systems displaying some kind of sublinear growth, Adv. Nonlinear Anal., 8 (2019), 583–602.
    [25] A. Fonda, R. Manásevich, F. Zanolin, Subharmonic solutions for some second-order differential equations with singularities, SIAM J. Math. Anal., 24 (1993), 1294–1311. doi: 10.1137/0524074
    [26] Y. Chen, Q. Liu, H. Su, Generalized Hamiltonian forms of dissipative mechanical systems via a unified approach, J. Geom. Phys., 160 (2021), 103976. doi: 10.1016/j.geomphys.2020.103976
    [27] P. Mathews, M. Lakshmanan, On a unique nonlinear oscillator, Quart. Appl. Math., 32 (1974), 215–218. doi: 10.1090/qam/430422
    [28] M. Lakshmanan, V. K. Chandrasekar, Generating finite dimensional integrable nonlinear dynamical systems, Eur. Phys. J. Spec. Top., 222 (2013), 665–688. doi: 10.1140/epjst/e2013-01871-6
    [29] A. Schulze-Halberg, Closed-form solutions and supersymmetric partners of the inverted Mathews-Lakshmanan oscillator, Eur. Phys. J. Plus., 130 (2015), 1–10. doi: 10.1140/epjp/i2015-15001-1
    [30] M. Sabatini, On the period function of x+f(x)x2+g(x)=0, J. Differ. Equations, 196 (2004), 151–168. doi: 10.1016/S0022-0396(03)00067-6
    [31] A. R. Chouikha, Isochronous centers of Lienard type equations and applications, J. Math. Anal. Appl., 331 (2004), 358–376.
    [32] I. Boussaada, A. R. Chouikha, J. M. Strelcyn, Isochronicity conditions for some planar polynomial systems, Bull. Sci. Math., 135 (2011), 89–112. doi: 10.1016/j.bulsci.2010.01.004
    [33] M. Bardet, I. Boussaada, A. R. Chouikha, J. M. Strelcyn, Isochronicity conditions for some planar polynomial systems Ⅱ, Bull. Sci. Math., 135 (2011), 230–249. doi: 10.1016/j.bulsci.2010.12.003
    [34] S. Atslega, On solutions of Neumann boundary value problem for the Liénard type equation, Math. Model. Anal., 13 (2008), 161–169. doi: 10.3846/1392-6292.2008.13.161-169
    [35] T. Ding, Application of Qualitative Methods of Ordinary Differential Equations, Higher Education Press, 2004.
    [36] Z. Zhang, K. Zhang, Q. Liu, W. Zhang, The existence of harmonic and subharmonic solutions for superlinear Liénard type equation with forced term, J. Nonlinear Funct. Anal., 2021.
    [37] D. W. Jordan, P. Smith, Nonlinear Odinary Dfferential Euations: An Introduction for Scientists and Engineers, Oxford University Press on Demand, 2007.
    [38] F. Nakajima, Even and periodic solutions of the equation ¨u+g(u)=p(t), J. Differ. Equations, 83 (1990), 277–299. doi: 10.1016/0022-0396(90)90059-X
    [39] T. Ding, F. Zanolin, Subharmonic solutions of second order nonlinear equations: A time-map approach, Nonlinear Anal., 20 (1993), 509–532. doi: 10.1016/0362-546X(93)90036-R
  • This article has been cited by:

    1. Yufei Chen, Bo Tang, Qihuai Liu, Envelope, singular solution, and finite-time stability, 2024, 0, 2163-2480, 0, 10.3934/eect.2024070
  • Reader Comments
  • © 2021 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(2406) PDF downloads(84) Cited by(1)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog