数学物理学报, 2021, 41(5): 1372-1381 doi:

论文

具有非线性记忆项的半线性双波动方程解的全局非存在性

欧阳柏平,1, 肖胜中,2

1 广州华商学院 广州 511300

2 广东农工商职业技术学院 广州 510507

Nonexistence of Global Solutions for a Semilinear Double-Wave Equation with a Nonlinear Memory Term

Ouyang Baiping,1, Xiao Shengzhong,2

1 Guangzhou Huashang College, Guangzhou 511300

2 Guangdong AIB Polytechnic College, Guangzhou 510507

通讯作者: 欧阳柏平, E-mail: oytengfei79@tom.com

收稿日期: 2021-01-11  

基金资助: 国家自然科学基金.  11371175
广东省普通高校创新团队.  2020WCXTD008
广东财经大学华商学院校内项目.  2020HSDS01
广州华商学院科研团队.  2021HSKT01

Received: 2021-01-11  

Fund supported: the NSFC.  11371175
the Innovation Team Project in Colleges and Universities of Guangdong Province.  2020WCXTD008
the Science Foundation of Huashang College Guangdong University of Finance & Economics.  2020HSDS01
the Science Research Team Project in Guangzhou Huashang College.  2021HSKT01

作者简介 About authors

肖胜中,E-mail:1246683963@qq.com , E-mail:1246683963@qq.com

Abstract

In this paper, we investigate the blow-up of solutions to a semilinear double-wave equation with a nonlinear memory term. By establishing some auxiliary functions and using iteration methods associated with a nonlinear integral inequality, the estimate of upper bound for the lifespan is obtained.

Keywords: Semilinear double-wave equation ; Blow-up ; Nonlinear memory term ; Lifespan

PDF (325KB) 元数据 多维度评价 相关文章 导出 EndNote| Ris| Bibtex  收藏本文

本文引用格式

欧阳柏平, 肖胜中. 具有非线性记忆项的半线性双波动方程解的全局非存在性. 数学物理学报[J], 2021, 41(5): 1372-1381 doi:

Ouyang Baiping, Xiao Shengzhong. Nonexistence of Global Solutions for a Semilinear Double-Wave Equation with a Nonlinear Memory Term. Acta Mathematica Scientia[J], 2021, 41(5): 1372-1381 doi:

1 引言

文献[1]研究了如下具有非线性记忆项的半线性波动方程解的爆破问题

$ \begin{eqnarray} \left\{\begin{array}{ll} u_{tt}-\Delta u = {\cal N}_{\gamma, p}(u), & (x, t)\in {{\Bbb R}} ^n \times (0, T), \\ (u, u_t)(0, x) = \varepsilon (u_0, u_1)(x), & x\in {{\Bbb R}} ^n, \end{array}\right. \end{eqnarray} $

其中

$ \begin{eqnarray} \begin{array}{ll} { } {\cal N}_{\gamma, p}(u) = C_{\gamma}\int_0^t (t-s)^{-\gamma} |u(s, x)|^p {\rm d}s, C_{\gamma} = \frac{1}{\Gamma(1-\gamma)}, \\ p>1, \gamma\in (0, 1), \varepsilon>0, \ \Gamma\ \mbox{为欧拉积分.} \end{array} \end{eqnarray} $

作者运用切片化程序迭代技巧得到了次临界($ 1<p<p_{\rm crit}(n, \gamma) $)和临界($ p = p_{\rm crit}(n, \gamma) $)两种情况下解的爆破结果. 进一步, 作者推出了临界指数, 即一般的Strauss指数$ p_{\rm crit} = p_{\rm crit}(n, \gamma) $. $ p_{\rm crit} $指数由以下二次方程的正实根表示($ n\geq 2 $)

也就是说

对于$ n = 1 $, 有$ p_{\rm crit}(1, \gamma) = \infty $. 确实, 极限情况$ p_{\rm crit}(n, 1) $就是所谓的Strauss指数, 是半线性波动方程在$ {{\Bbb R}} ^n $上的临界指数. 目前已有很多文献对其进行了研究[2-8], 主要集中在柯西问题解的爆破, 其研究方法主要是基于微分不等式技巧和Kato引理. 然而, Kato引理不适用于高阶半线性发展方程, 特别是具有非线性记忆项的高阶发展方程. 因为Kato引理只能解决二阶具有多项式型非线性项的常微分方程. 为了解决带有时间变量和非线性积分项的高阶偏微分方程解的爆破问题, 需要寻找其他的办法. 幸运的是, 近来已有学者运用迭代方法研究了某些双曲方程[9-16]. 还有一些学者研究了其他抛物问题解的爆破[17-26].

最近, 文献[27]考虑了一个四阶双曲方程的双波动方程. 需要强调的是, 它与以往的半线性波动方程(比如二阶双曲方程(1.1))具有非常大的不同.

本文研究如下具有非线性记忆项的半线性双波动方程柯西问题解的爆破

$ \begin{eqnarray} \left\{\begin{array}{ll} (\partial^2_t-\Delta)^2u = {\cal N}_{\gamma, p}(u), & (x, t)\in {{\Bbb R}} ^n \times (0, T), \\ (u, u_t, u_{tt}, u_{ttt})(0, x) = \varepsilon (u_0, u_1, u_2, u_3)(x), & x\in {{\Bbb R}} ^n, \end{array}\right. \end{eqnarray} $

其中非线性项$ {\cal N}_{\gamma, p}(u) $由(1.2)式给出, $ p>1, \gamma\in (0, 1) $.

本文的目标是得到柯西问题(1.3)式解的爆破条件以及生命跨度估计. 事实上, 如果考虑同样的Lamé常数, 方程(1.3)式将与四阶弹性波方程[28]高度相关. 不幸的是, 有关半线性弹性波方程柯西问题临界指数的研究目前尚未开展. 文献[29]考虑了弹性波方程柯西问题解的爆破. 到目前为止, 尚未发现有文献关于具有非线性记忆项的半线性四阶双曲方程柯西问题的研究. 本文运用泛函分析方法和迭代技巧来研究(1.3)式中柯西问题解的爆破, 得到了(1.3)式柯西问题解的全局非存在性条件以及生命跨度上界估计, 为高阶波动方程研究提供了一个思路, 从一定程度上说, 是有意义的.

本文结构安排如下: 第$ 2 $节介绍本文的主要结果. 第$ 3 $节给出主要结果的证明. 第$ 4 $节对本文进行总结以及后续的研究设想.

2 主要结果

首先, 引入(0.3)式柯西问题能量解的定义.

定义 2.1  设$ (u_0, u_1, u_2, u_3)\in H^3({{\Bbb R}} ^n)\times H^2({{\Bbb R}} ^n)\times H^1({{\Bbb R}} ^n)\times L^2({{\Bbb R}} ^n) $. $ u $是问题(1.3)式在$ [0, T) $上的能量解, 如果

使得$ u\in L^p_{loc}([0, T)\times {{\Bbb R}} ^n) $满足$ (u_0, u_1, u_2, u_3) \in H^3({{\Bbb R}} ^n)\times H^2({{\Bbb R}} ^n)\times H^1({{\Bbb R}} ^n)\times L^2({{\Bbb R}} ^n) $和下面的积分关系

$ \begin{array}{l} \int_{{{\Bbb R}} ^n} u_{ttt}(t, x)\phi(t, x) {\rm d}x-\int_{{{\Bbb R}} ^n}u_{ttt}(0, x)\phi(0, x){\rm d}x-\int^t_0\int_{{{\Bbb R}} ^n}u_{ttt}(s, x)\phi_t(s, x){\rm d}x{\rm d}s\\+2\int^t_0\int_{{{\Bbb R}} ^n}\nabla u_{tt}(s, x)\cdot\nabla\phi(s, x) {\rm d}x{\rm d}s-\int^t_0\int_{{{\Bbb R}} ^n}\nabla \Delta u(s, x)\cdot \nabla \phi(s, x){\rm d}x{\rm d}s\\ = C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\phi(s, x)\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s, \end{array} $

其中$ \phi(t, x)\in {\cal C}^\infty_0([0, T]\times {{\Bbb R}} ^n), t\in (0, T). $

(2.1)式中运用分部积分, 有

$ \begin{array}{l}\int_{{{\Bbb R}} ^n} u_{ttt}(t, x)\phi(t, x) {\rm d}x-\varepsilon\int_{{{\Bbb R}} ^n}u_3(x)\phi(0, x){\rm d}x-\int_{{{\Bbb R}} ^n} u_{tt}(t, x)\phi_t(t, x) {\rm d}x\\+\varepsilon\int_{{{\Bbb R}} ^n}u_2(x)\phi_t(0, x){\rm d}x+\int_{{{\Bbb R}} ^n} u_t(t, x)\phi_{tt}(t, x){\rm d}x-\varepsilon\int_{{{\Bbb R}} ^n}u_1(x)\phi_{tt}(0, x){\rm d}x\\-\int_{{{\Bbb R}} ^n} u(t, x)\phi_{ttt}(t, x) {\rm d}x+\varepsilon\int_{{{\Bbb R}} ^n}u_0(x)\phi_{ttt}(0, x){\rm d}x\\+\int^t_0\int_{{{\Bbb R}} ^n}u(s, x)\phi_{tttt}(s, x){\rm d}x{\rm d}s-2\int_{{{\Bbb R}} ^n} u_t(t, x)\Delta\phi(t, x){\rm d}x+2\varepsilon\int_{{{\Bbb R}} ^n} u_1(x)\Delta\phi(0, x){\rm d}x \\+2\int_{{{\Bbb R}} ^n} u(t, x)\Delta\phi_t(t, x){\rm d}x-2\varepsilon\int_{{{\Bbb R}} ^n} u_0(x)\Delta\phi_t(0, x){\rm d}x-2\int^t_0\int_{{{\Bbb R}} ^n} u(s, x)\Delta\phi_{tt}(s, x) {\rm d}x{\rm d}s\\-\int^t_0\int_{{{\Bbb R}} ^n}\nabla \Delta u(s, x)\cdot \nabla \phi(s, x){\rm d}x{\rm d}s\\ = C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\phi(s, x)\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s. \end{array} $

$ t\rightarrow T $, 易知$ u $满足(1.3)式弱解的定义.

本文有以下定理.

定理 2.1  设$ p>1 $, 有

其中$ p_{0} $是如下方程的最大根

$ (u_0, u_1, u_2, u_3)\in H^3({{\Bbb R}} ^n)\times H^2({{\Bbb R}} ^n)\times H^1({{\Bbb R}} ^n)\times L^2({{\Bbb R}} ^n) $是非负的紧支集函数, 其支集包含于$ B_R(R>0) $, 使得$ u_0 $$ u_1 $$ u_2 $$ u_3 $不恒为$ 0 $. 其中, $ B_R $是一个原点为球心半径为$ R $的球. 特别地, $ u_3(x)+u_2(x)>u_1(x)+u_0(x) $$ 2u_1(x)>u_0(x)+u_2(x) $. 如果$ u $是问题(0.3)式的能量解, 其生命跨度$ T(\varepsilon) $满足supp $ u(t, \cdot)\subset B_{t+R}, t\in (0, T) $, 则存在一个正常数$ \varepsilon_0 = \varepsilon_0(u_0, u_1, u_2, u_3, n, p, \gamma, R) $, 使得当$ \varepsilon\in (0, \varepsilon_0] $$ u $在有限时间内爆破. 进一步, 可得解的生命跨度上界估计, 即

其中$ \widetilde{\widetilde{C}} $是独立于$ \varepsilon $的正常数并且

$ \begin{equation} \Upsilon(p, n, \gamma) = (n+5-2\gamma)p+2-(n-3)p^2. \end{equation} $

3 主要结果的证明

$ \begin{equation} F(t) = \int_{{{\Bbb R}} ^n} u(t, x){\rm d}x. \end{equation} $

$ \phi $满足$ \phi\equiv1, \{(s, x)\in [0, t]\times {{\Bbb R}} ^n:|x|\leq R+s\} $. 于是, 由(2.2)式, 得

$ \begin{equation} \int_{{{\Bbb R}} ^n} u_{ttt}(t, x){\rm d}x-\varepsilon\int_{{{\Bbb R}} ^n}u_3(x){\rm d}x = C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s. \end{equation} $

从而有

$ \begin{equation} F'''(t) = F'''(0)+ C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s. \end{equation} $

对(3.3)式关于时间$ t $积分三次, 得到

$ \begin{eqnarray} F(t)& = &F(0)+F'(0)t+\frac{1}{2}F''(0)t^2+\frac{1}{6}F'''(0)t^3\\+C_{\gamma}\int_0^t \int_0^s\int_0^\tau\int_0^\sigma\int_0^\eta\int_{{{\Bbb R}} ^n}(\eta-\xi)^{-\gamma} |u(\xi, x)|^p {\rm d}x{\rm d}\xi{\rm d}\eta{\rm d}\sigma{\rm d}\tau{\rm d}s. \end{eqnarray} $

由supp $ u(t, \cdot)\subset B_R+t, t\in (0, T) $和Hölder不等式, 可得

$ \begin{eqnarray} \int_{{{\Bbb R}} ^n} |u(\xi, x)|^p{\rm d}x\geq C(R+\xi)^{-n(p-1)}(F(\xi))^p. \end{eqnarray} $

根据(3.4)式及初始值的非负假设, 易知$ F(\xi)\geq0, \xi \in [0, t]. $

联立(3.4)–(3.5)式, 有

$ \begin{eqnarray} F(t)\geq CC_{\gamma}(R+t)^{-n(p-1)-\gamma}\int_0^t \int_0^s\int_0^\tau\int_0^\sigma\int_0^\eta(F(\xi))^p {\rm d}\xi{\rm d}\eta{\rm d}\sigma{\rm d}\tau{\rm d}s. \end{eqnarray} $

接下来, 通过对$ F(t) $的下界序列进行迭代完成定理的证明. 实际上, (3.6) 式已给出了下界序列的迭代框架. 为了推导$ F(t) $的第一下界估计, 引入如下函数

其中$ \Phi $是正的光滑函数且有下面的性质

$ \Psi = \Psi(t, x) = {\rm e}^{-t}\Phi(x). $易知, $ \Psi $是方程$ (\partial^2_t-\Delta)^2\Psi = 0 $的解. 给定辅助函数

$ \begin{eqnarray} F_0(t) = \int_{{{\Bbb R}} ^n}u(t, x)\Psi(t, x){\rm d}x. \end{eqnarray} $

(3.3)式对$ t $求导数, 得

$ \begin{eqnarray} F''''(t) = C_{\gamma}\int_0^t\int_{{{\Bbb R}} ^n}(t-s)^{-\gamma} |u(s, x)|^p {\rm d}x{\rm d}s. \end{eqnarray} $

运用Hölder不等式于(3.7)式, 有

$ \begin{eqnarray} \int_{{{\Bbb R}} ^n} |u(s, x)|^p {\rm d}x\geq |F_0(s)|^p\bigg(\int_{B_{R+s}} |\Psi(s, x)|^{\frac{p}{p-1}} {\rm d}x\bigg)^{-(p-1)}. \end{eqnarray} $

将测试函数$ \Psi $应用于(2.1)式, 可推出

$ \begin{array}{l} \int_{{{\Bbb R}} ^n} u_{ttt}(t, x)\Psi(t, x) {\rm d}x-\int_{{{\Bbb R}} ^n}u_{ttt}(0, x)\Psi(0, x){\rm d}x-\int^t_0\int_{{{\Bbb R}} ^n}u_{ttt}(s, x)\Psi_t(s, x){\rm d}x{\rm d}s\\+2\int^t_0\int_{{{\Bbb R}} ^n}\nabla u_{tt}(s, x)\cdot\nabla\Psi(s, x) {\rm d}x{\rm d}s-\int^t_0\int_{{{\Bbb R}} ^n}\nabla \Delta u(s, x)\cdot \nabla \Psi(s, x){\rm d}x{\rm d}s\\ = C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\Psi(s, x)\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s. \end{array} $

由分部积分并注意到$ \Psi $的性质, (3.10)式化为

$ \begin{array}{l} \int_{{{\Bbb R}} ^n} u_{ttt}(t, x)\Psi(t, x) {\rm d}x+\int_{{{\Bbb R}} ^n} u_{tt}(t, x)\Psi(t, x) {\rm d}x\\-\int_{{{\Bbb R}} ^n} u_t(t, x)\Psi(t, x) {\rm d}x -\int_{{{\Bbb R}} ^n} u(t, x)\Psi(t, x) {\rm d}x \\ = \varepsilon I[u_0, u_1, u_2, u_3]+ C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\Psi(s, x)\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s, \end{array} $

其中

联立(3.7)–(3.11)式, 可得

$ \begin{array}{l} F'''_0(t)+4F''_0(t)+4F'_0(t)\\ =\varepsilon I[u_0, u_1, u_2, u_3]+ C_{\gamma}\int_0^t \int_{{{\Bbb R}} ^n}\Psi(s, x)\int_0^s(s-\tau)^{-\gamma} |u(\tau, x)|^p {\rm d}\tau {\rm d}x{\rm d}s\\ \geq \varepsilon I[u_0, u_1, u_2, u_3]. \end{array} $

$ G(t) = F''_0(t)+2F'_0(t) $, 重写(3.12)式为

$ \begin{eqnarray} G'(t)+2G(t)\geq \varepsilon I[u_0, u_1, u_2, u_3]. \end{eqnarray} $

积分(3.13)式, 得到

$ \begin{eqnarray} G(t)\geq \left(G(0)-\frac{1}{2}\varepsilon I[u_0, u_1, u_2, u_3]\right){\rm e}^{-2t}+\frac{1}{2}\varepsilon I[u_0, u_1, u_2, u_3]. \end{eqnarray} $

从而有

$ \begin{eqnarray} F''_0(t)+2F'_0(t)\geq \left(G(0)-\frac{1}{2}\varepsilon I[u_0, u_1, u_2, u_3]\right){\rm e}^{-2t}+\frac{1}{2}\varepsilon I[u_0, u_1, u_2, u_3]. \end{eqnarray} $

对(3.15)式求积分, 可得

$ \begin{array}{l} F_0(t)\geq\varepsilon \int_{{{\Bbb R}} ^n}u_0(x)\Phi(x){\rm d}x+\frac{\varepsilon}{4}(1-{\rm e}^{-2t}) \int_{{{\Bbb R}} ^n}(3u_1(x)-2u_0(x)-u_3(x))\Phi(x){\rm d}x\\+\frac{\varepsilon}{4}t \int_{{{\Bbb R}} ^n}(u_3(x)+u_2(x)-u_1(x)-u_0(x))\Phi(x){\rm d}x\\+\frac{\varepsilon}{4} t{\rm e}^{-2t}\int_{{{\Bbb R}} ^n}(u_3(x)+u_0(x)-u_1(x)-u_2(x))\Phi(x){\rm d}x\\ \geq\varepsilon \int_{{{\Bbb R}} ^n}u_0(x)\Phi(x){\rm d}x+2\eta \int_{{{\Bbb R}} ^n}(2u_1(x)-u_0(x)-u_2(x))\Phi(x){\rm d}x\\+\frac{\varepsilon}{4}t \int_{{{\Bbb R}} ^n}(u_3(x)+u_2(x)-u_1(x)-u_0(x))\Phi(x){\rm d}x, \end{array} $

其中$ \eta = \min\left\{\frac{\varepsilon}{4}(1-{\rm e}^{-2t}), \frac{\varepsilon}{4} t{\rm e}^{-2t}\right\}>0. $

由(3.16) 式, 存在一个$ t_0 $使得

$ \begin{equation} F_0(t)\geq \widetilde{C}\varepsilon t , \end{equation} $

其中$ t\geq t_0, \widetilde{C} $为正常数.

又由假设有

另外, 由$ \Psi $的渐近性可得

$ \begin{equation} \int_{B_{R+s}} |\Psi(s, x)|^{\frac{p}{p-1}} {\rm d}x\leq \widetilde{K}(R+s)^{(n-1)(1-\frac{p'}{2})}, \end{equation} $

其中$ \widetilde{K}>0. $

联立(3.9)式和(3.17)–(3.18)式, 有

$ \begin{eqnarray} \int_{{{\Bbb R}} ^n} |u(s, x)|^p {\rm d}x\geq C_0\varepsilon^p(R+s)^{(n-1)-\frac{(n-1)p}{2}}s^p, \end{eqnarray} $

其中$ C_0 = \widetilde{C}^p\widetilde{K}^{-(p-1)}, s\geq t_0. $

由(3.8)–(3.19)式, 可得

$ \begin{eqnarray} F''''(t)\geq C_1\varepsilon^p(R+t)^{(n-1)-\frac{(n-1)p}{2}-\gamma}(t-t_0)^{p+1}, \end{eqnarray} $

其中$ t\geq t_0, C_1 = \frac{C_0C_{\gamma}}{p+1}. $

对(3.20)式求积分四次, 得到

$ \begin{array}{l} F(t)\geq F(t_0)+F'(t_0)t+F''(t_0)\frac{t^2}{2}+F'''(t_0)\frac{t^3}{6}\\ +C_1\varepsilon^p\int^t_{t_0}\int^s_{t_0}\int^\tau_{t_0}\int^\sigma_{t_0}(R+\eta)^{(n-1)-\frac{(n-1)p}{2}-\gamma}(\eta-t_0)^{p+1}{\rm d}\eta{\rm d}\sigma{\rm d}\tau{\rm d}s\\ \geq C_1\varepsilon^p\int^t_{t_0}\int^s_{t_0}\int^\tau_{t_0}\int^\sigma_{t_0}(R+\eta)^{(n-1)-\frac{(n-1)p}{2}-\gamma}(\eta-t_0)^{p+1}{\rm d}\eta{\rm d}\sigma{\rm d}\tau{\rm d}s\\ \geq C_2\varepsilon^p(R+t)^{-\frac{(n-1)p}{2}-\gamma}(t-t_0)^{n+p+4}, \end{array} $

其中$ C_2 = \frac{C_1}{(n+p+1)(n+p+2)(n+p+3)(n+p+4)}, t\geq t_0. $

重记(3.21)式为

$ \begin{eqnarray} F(t)\geq K_0(R+t)^{-\alpha_0}(t-t_0)^{\beta_0}, \quad t\geq t_0, \end{eqnarray} $

其中$ K_0 = C_2\varepsilon^p, \alpha_0 = \frac{(n-1)p}{2}+\gamma, \beta_0 = n+p+4 $.

接下来, 将运用迭代框架(3.6) 式推出$ F(t) $的下界序列, 即

$ \begin{eqnarray} F(t)\geq K_j(R+t)^{-\alpha_j}(t-t_0)^{\beta_j}, \end{eqnarray} $

其中$ \{K_j\}_{j\in N} $, $ \{\alpha_j\}_{j\in N} $, $ \{\beta_j\}_{j\in N} $是后面会定义的非负实数列.

因而后面的任务变成$ t\rightarrow \infty $$ F(t) $的爆破研究.

联立(3.6)式和(3.23)式, 可得

$ \begin{array}{l} F(t)\geq CC_{\gamma} K^p_j\int^t_{t_0}\int^s_{t_0}\int^{\tau}_{t_0}\int^{\sigma}_{t_0}\int^\eta_{t_0}(R+\xi)^{-n(p-1)-\gamma-\alpha_jp}(\xi-t_0)^{p\beta_j}d \xi {\rm d}\eta{\rm d}\sigma{\rm d}\tau{\rm d}s\\ \geq\frac{ CC_{\gamma} K^p_j(R+t)^{-n(p-1)-\gamma-\alpha_jp}(t-t_0)^{p\beta_j+5}}{(p\beta_j+1)(p\beta_j+2)(p\beta_j+3)(p\beta_j+4)(p\beta_j+5)}. \end{array} $

$ \begin{equation} \begin{array}{ll} { } K_{j+1} = \frac{ C C_{\gamma} K^p_j}{(p\beta_j+1)(p\beta_j+2)(p\beta_j+3)(p\beta_j+4)(p\beta_j+5)}, \\ \alpha_{j+1} = n(p-1)+\gamma+\alpha_jp, \\ \beta_{j+1} = p\beta_j+5, \end{array} \end{equation} $

于是, (3.24) 式重记为

$ \begin{eqnarray} F(t)\geq K_{j+1}(R+t)^{\alpha_{j+1}}(t-t_0)^{\beta_{j+1}}. \end{eqnarray} $

上式表明(3.23)式对于$ j+1 $是成立的.

下面研究$ \alpha_j, \beta_j, K_j $的估计. 由(3.25)式, 有

$ \begin{array}{l} \alpha_j=[n(p-1)+\gamma](1+p+p^2+\cdots+p^{j-1})+\alpha_0p^j\\ =(n(p-1)+\gamma)\frac{p^j-1}{p-1}+\alpha_0p^j\\ =\left(\alpha_0+n+\frac{\gamma}{p-1}\right)p^j-\left(n+\frac{\gamma}{p-1}\right), \\ \beta_{j}=\beta_0p^j+5p^{j-1}+5p^{j-2}+\cdots+5p+5\\ =\beta_0p^j+\frac{5(p^j-p)}{p-1}+5\\ =\left(\frac{5}{p-1}+\beta_0\right)p^j-\frac{5}{p-1}. \end{array} $

易知

因此, 由(3.25)式, 得到

$ \begin{eqnarray} K_j\geq CC_{\gamma}\left(\beta_0+\frac{5}{p-1}\right)^{-5}p^{-5j}K^p_{j-1} = Dp^{-5j}K^p_{j-1}. \end{eqnarray} $

对(3.28)式两边取对数并运用递推方法, 可得

$ \begin{equation} \log K_j\geq p^j\left(\log K_0-\frac{5p\log p}{(p-1)^2}+\frac{\log D}{p-1}\right)+\frac{5j\log p}{p-1}+\frac{5p\log p}{(p-1)^2}-\frac{\log D}{p-1}, \ \ \forall j\in N. \end{equation} $

$ j_0 = j_0(n, p, \gamma)\in N $是最小非负整数, 使得

于是, 由(3.29)式, 有

$ \begin{eqnarray} \log K_j\geq p^j\log \left(K_0p^{-\frac{5p}{(p-1)^2}}D^{\frac{1}{p-1}}\right) = p^j\log(E_0\varepsilon^p), \end{eqnarray} $

其中$ E_0 = E_0(n, p, \gamma)>0, j \geq j_0. $

联立(3.23), (3.27) 和(3.30)式, 可推出

$ \begin{array}{l} F(t)\geq \exp\left(p^j\log(E_0\varepsilon^p)\right)(R+t)^{-(\alpha_0+n+\frac{\gamma}{p-1})p^j+(n+\frac{\gamma}{p-1})}(t-t_0)^{(\beta_0+\frac{5}{p-1})p^j-\frac{5}{p-1}}\\ =\exp\left(p^j\left(\log(E_0\varepsilon^p)-\left(\alpha_0+n+\frac{\gamma}{p-1}\right)\log(R+t)+\left(\beta_0+\frac{5}{p-1}\right)\log (t-t_0)\right)\right)\\\times(t+R)^{n+\frac{\gamma}{p-1}}(t-t_0)^{-\frac{5}{p-1}}, \end{array} $

其中$ j\geq j_0, t\geq t_0. $

$ t\geq R+2t_0 $时, 有$ \log(R+t)\leq\log(2(t-t_0)). $因此(3.31)式可化为

$ \begin{equation} F(t)\geq \exp\left(p^j\left(\log(E_0\varepsilon^p2^{-(\alpha_0+n+\frac{\gamma}{p-1})}(t-t_0)^{\beta_0+\frac{5}{p-1}-(\alpha_0+n+\frac{\gamma}{p-1})})\right)\right) (R+t)^n(t-t_0)^{-\frac{5}{p-1}}. \end{equation} $

(3.32)式中, $ t-t_0 $的指数为

$ \begin{eqnarray} \beta_0+\frac{5}{p-1}-\left(\alpha_0+n+\frac{\gamma}{p-1}\right) = \frac{(n+5-2\gamma)p+2-(n-3)p^2}{2(p-1)} = \frac{\Upsilon(n, p, \gamma)}{2(p-1)}. \end{eqnarray} $

因此, 对于$ p>1(n = 1, 2, 3) $$ 1<p<p_0(n)(n\geq 4) $, $ t $的指数是正的.

固定$ \varepsilon_0 = \varepsilon_0(u_0, u_1, u_2, u_3, n, p, R, \gamma)>0 $, 得到

其中$ E_1 = (2^{-(\alpha_0+n+\frac{\gamma}{p-1})}E_0)^{\frac{2(p-1)}{\Upsilon(p, n, \gamma)}}. $

故对于$ \varepsilon\in(0, \varepsilon_0] $ and $ t>E_1^{-1}\varepsilon^{-\frac{2p(p-1)}{\Upsilon(p, n, \gamma)}}\geq R $, 有

(3.32)式中取$ j\rightarrow \infty $, 可得当$ \varepsilon\in(0, \varepsilon_0] $$ t>E_1^{-1}\varepsilon^{-\frac{2p(p-1)}{\Upsilon(p, n, \gamma)}} $时, $ F(t) $的下界爆破. 这表明问题(1.3)式不存在全局解. 进一步, 可得解的生命跨度的上界估计

从而完成了定理的证明.

4 总结

本文研究了一类具有非线性记忆项的半线性双波动方程解的爆破问题. 运用非线性积分不等式和迭代方法, 通过构造辅助函数, 得到了在次临界情况下该问题解的全局非存在性, 进一步推出了解的生命跨度的上界估计. 后续工作将对该问题在临界情况下解的全局存在性进行研究. 由于是在临界情况下讨论, 所以对辅助函数的构造难度加大, 因此问题变得更加复杂.

参考文献

Chen W, Palmieri A. Blow-up Result for a Semilinear Wave Equation with a Nonlinear Memory Term//Cicognani M, Santo D, Parmeggiani A, Reissig M. Anomalies in Partial Differential Equations. Switzerland: Springer, 2021: 77-97

[本文引用: 1]

John F .

Blow-up of solutions of nonlinear wave equations in three space dimensions

Manuscripta Math, 1979, 28, 235- 268

DOI:10.1007/BF01647974      [本文引用: 1]

Kato T .

Blow-up of solutions of some nonlinear hyperbolic equations

Comm Pure Appl Math, 1980, 33, 501- 505

DOI:10.1002/cpa.3160330403     

Strauss W A .

Nonlinear scattering theory at low energy

J Functional Analysis, 1981, 41, 110- 133

DOI:10.1016/0022-1236(81)90063-X     

Glassey R T .

Existence in the large for $ \Box u = F(u) $ in two space dimensions

Math Z, 1981, 178, 233- 261

DOI:10.1007/BF01262042     

Glassey R T .

Finite-time blow-up for solutions of nonlinear wave equations

Math Z, 1981, 177, 323- 340

DOI:10.1007/BF01162066     

Sideris T C .

Nonexistence of global solutions to semilinear wave equations in high dimensions

J Differential Equations, 1984, 52, 378- 406

DOI:10.1016/0022-0396(84)90169-4     

Schaeffer J .

The equation $ u_{tt}-\Delta u = |u|.p $ for the critical value of $ p $

Proc Roy Soc Edinburgh Sect A, 1985, 101, 31- 44

DOI:10.1017/S0308210500026135      [本文引用: 1]

Chen W .

Interplay effects on blow-up of weakly coupled systems for semilinear wave equations with general nonlinear memory terms

Nonlinear Anal, 2021, 202, 112160

DOI:10.1016/j.na.2020.112160      [本文引用: 1]

Chen W , Palmieri A .

Nonexistence of global solutions for the semilinear Moore-Gibson-Thompson equation in the conservative case

Discrete Contin Dyn Syst, 2020, 40, 5513- 5540

DOI:10.3934/dcds.2020236     

Chen W , Reissig M .

Blow-up of solutions to Nakao's problem via an iteration argument

J Differential Equations, 2021, 275, 733- 756

DOI:10.1016/j.jde.2020.11.009     

Chen W , Palmieri A .

A blow-up result for the semilinear Moore-Gibson-Thompson equation with nonlinearity of derivative type in the conservative case

Evol Equa Control Theory, 2020,

DOI:10.3934/eect.2020085     

Lai N A , Takamura H .

Nonexistence of global solutions of nonlinear wave equations with weak time-dependent damping related to Glassey's conjecture

Differ Integral Equa, 2019, 32, 37- 48

URL    

Lai N A , Takamura H , Wakasa K .

Blow-up for semilinear wave equations with the scale invariant damping and super-Fujita exponent

J Differential Equations, 2017, 263, 5377- 5394

DOI:10.1016/j.jde.2017.06.017     

Palmieri A , Takamura A .

Blow-up for a weekly coupled system of semilinear damped wave equations in the scattering case with power nonlinearities

Nonlinear Anal, 2019, 187, 467- 492

DOI:10.1016/j.na.2019.06.016     

Palmieri A, Takamura H. Nonexistence of global solutions for a weakly coupled system of semilinear damped wave equations in the scattering case with mixed nonlinear terms. 2019, arXiv: 1901.04038

[本文引用: 1]

Liu Y .

Blow-up phenomena for the nonlinear nonlocal porous medium equation under Robin boundary condition

Comput Math Appl, 2013, 66, 2092- 2095

DOI:10.1016/j.camwa.2013.08.024      [本文引用: 1]

Li Y , Liu Y , Lin C .

Blow-up phenomena for some nonlinear parabolic problems under mixed boundary conditions

Nonlinear Anal-Real, 2010, 11, 3815- 3823

DOI:10.1016/j.nonrwa.2010.02.011     

Liu Y , Luo S , Ye Y .

Blow-up phenomena for a parabolic problem with a gradient nonlinearity under nonlinear boundary conditions

Comput Math Appl, 2013, 65, 1194- 1199

DOI:10.1016/j.camwa.2013.02.014     

Chen W , Liu Y .

Lower bound for the blow-up time for some nonlinear parabolic equations

Bound Value Probl, 2016, 2016, 161

DOI:10.1186/s13661-016-0669-5     

Fang Z , Wang Y .

Blow-up analysis for a semi-linear parabolic equation with time-dependent coefficients under nonlinear boundary flux

Z Angew Math Phys, 2015, 66, 1- 17

DOI:10.1007/s00033-013-0377-2     

Tao X , Fang Z .

Blow-up phenomena for a nonlinear reaction diffusion system with time dependent coefficients

Comput Math Appl, 2017, 74, 2520- 2528

DOI:10.1016/j.camwa.2017.07.037     

Ma L , Fang Z .

Blow-up phenomena of solutions for a reaction-diffusion equation with weighted exponential nonlinearity

Comput Math Appl, 2018, 75, 2735- 2745

DOI:10.1016/j.camwa.2018.01.005     

Liu Z , Fang Z .

Blow-up phenomena for a nonlocal quasilinear parabolic problem equation with time-dependent coefficients under nonlinear boundary flux

Discrete Contin Dyn Syst Ser, 2016, 21, 3619- 3635

DOI:10.3934/dcdsb.2016113     

Ma L , Fang Z .

Blow-up phenomena for a semilinear parabolic equation with weighted with inner absorption under nonlinear boundary flux

Math Meth Appl Sci, 2017, 40, 115- 128

DOI:10.1002/mma.3971     

Zheng Y , Fang Z .

Blow-up analysis for a weakly coupled reaction-diffusion system with gradient sources terms and time-dependent coefficients(in Chinese)

Acta Mathematica Scientia, 2020, 40A (3): 735- 755

[本文引用: 1]

Liu Y , Jiang D , Yamamoto M .

Inverse source problem for a double hyperbolic equation describing the three-dimensional time cone model

SIAM J Appl Math, 2015, 75, 2610- 2635

DOI:10.1137/15M1018836      [本文引用: 1]

D'Abbicco M , Jannelli E .

Dissipative higher order hyperbolic equations

Commum Part Diff Equations, 2017, 42, 1682- 1706

DOI:10.1080/03605302.2017.1390674      [本文引用: 1]

Chen W , Reissig M .

Weakly coupled systems of semilinear elastic waves with different damping mechanisms in 3D

Mathematical Methods in the Applied Sciences, 2019, 42, 667- 709

DOI:10.1002/mma.5370      [本文引用: 1]

/