1 绪论
本文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 在拉格朗日坐标系下, 它具有形式
(1.1) $\begin{equation} \left\{\begin{array}{ll} v_{t}-u_{x}=0, & t>0, x \in \mathbb{R}, \\ u_{t}+p(v)_{x}=\left(\mu(v) \frac{u_{x}}{v}\right)_{x}, & t>0, x \in \mathbb{R}. \end{array}\right. \end{equation} $
(1.2) $\begin{equation} \left.(v, u)\right|_{t=0}=\left(v_{0}, u_{0}\right), uad \lim _{x \rightarrow \pm \infty}\left(v_{0}, u_{0}\right)(x)=\left(v_{\pm}, u_{\pm}\right). \end{equation}$
这里 $t>0$ 是时间变量, $x \in \mathbb{R} $ 是拉格朗日空间变量, $v_{\pm}>0, u_{\pm}$ 是给定的常数. 主要的因变量是流体的比容 $v$ 和速度 $u$ . 在本文中, 压强 $p$ 和粘性系数 $\mu$ 满足下式
(1.3) $\begin{equation} p(v)=a v^{-\gamma}, uad \mu(v)=b v^{-\kappa}. \end{equation}$
其中$\gamma>1$ 代表绝热指数, $a>0, b>0$ 和 $\kappa\geq 0$ 是气体常数. 不失一般性,我们可以假设在下文中$a=b=1$ .
对于理想多方气体, 基本热力学量密度 $\rho=v^{-1}$ , 温度 $\theta$ , 内能 $e$ , 熵 $s$ , 和压力$p$ , 满足如下状态方程
(1.4) $\begin{equation}p=R \rho \theta=\tilde{b} v^{-l} e^{s / c_{v}}, uad e=c_{v} \theta.\end{equation}$
其中正常数 $l>1, R>0, \tilde{b}>0, c_{v}>0$ .
根据稀薄气体的动理学理论,当Boltzmann方程通过Chapman-Enskog展开推导出可压缩Navier-Stokes方程组时, 粘性系数$\mu$ 和热传导系数$\kappa$ 是依赖于温度的函数[2 ,3 ,21 ,23 ] , 且满足
(1.5) $\begin{equation} \mu(\theta) \propto \theta^{\frac{1}{2}+\frac{2}{s-1}}, uad \kappa(\theta) \propto \theta^{\frac{1}{2}+\frac{2}{s-1}}. \end{equation}$
其中$s(>5)$ 是常数. 由于 $\frac{1}{2}+\frac{2}{s-1}>\frac{1}{2}$ , 上式可改写为
(1.6) $\begin{equation} \mu(\theta)=\bar{\mu} \theta^{\iota}, uad \kappa(\theta)=\bar{\theta} \theta^{\iota}, uad \iota \geq \frac{1}{2}. \end{equation} $
其中 $\bar{\mu}$ 和 $\bar{\theta}$ 是两个正常数.
本文我们只考虑等熵且气体为多方气体的情形, 在这种情形下压强 $p$ 满足 $\gamma$ -律, 即
(1.7) $\begin{equation}p(\rho)=a \rho^{\gamma}\end{equation}$
的情形, 这里 $a>0$ 表示正的常数, $\gamma \geq 1$ 表示气体的绝热指数. 此时, 由(1.4)和(1.7)式可得
(1.8) $\begin{equation}p=R \rho \theta=a \rho^{\gamma},\end{equation}$
从而我们可以得到温度 $\theta$ 和密度 $\rho$ 满足如下关系式
(1.9) $\begin{equation}\theta=\frac{a}{R} \rho^{\gamma-1}.\end{equation}$
从(1.6)和(1.9)式可以推导出, 粘性系数 $\mu$ 是关于密度 $\rho$ 的函数并且它们满足下面的关系式
(1.10) $\begin{equation}\mu(\rho)=\mu_{0} \rho^{(\gamma-1) \iota}, uad \iota \geq \frac{1}{2}, uad \mu_{0}=\bar{\mu}\left(\frac{a}{R}\right)^{\iota}.\end{equation}$
在阐述本文中研究的主要问题之前, 我们首先对关系式 (1.3) 作一些补充说明. 对于等熵多方流, 由(1.5)和(1.9)式可知, 粘性系数 $\mu$ 对于 $\theta$ 的依赖性可以转化为 $\mu$ 对密度的依赖性
(1.11) $\begin{equation} \mu(\rho) \propto \rho^{\frac{(s+3)(\gamma-1)}{2(s-1)}}, uad \gamma>1, uad s>5, \end{equation}$
这事实上和(1.4)式是等价的, 其中 $\kappa=\frac{(s+3)(\gamma-1)}{2(s-1)}$ . 需要指出的是,尽管从物理角度来说, 当 $s>5$ 时意味着$\frac{\gamma-1}{2}<\kappa \leq \gamma-1$ , 为了讨论方便, 我们在下文中将假设 $\kappa$ 与 $\gamma$ 无关.
本文主要研究柯西问题(1.1)-(1.2)粘性激波的非线性稳定性. 我们知道,系统(1.1)连接 $\left(v_{-}, u_{-}\right)$ 和 $\left(v_{+}, u_{+}\right)$ 这两个状态的的粘性激波是一个行波解, 它满足
(1.12) $\begin{equation} (V, U)(-\infty)=\left(v_{l}, u_{l}\right), uad(V, U)(+\infty)=\left(v_{r}, u_{r}\right), \end{equation}$
其中 $s$ 是激波速度, $\left(v_{l}, u_{l}\right)$ 和 $\left(v_{r}, u_{r}\right)$ 是给定的远场状态, 且满足 $\left(v_{+}, u_{+}\right) \in S_{1} S_{2}\left(v_{-}, u_{-}\right)$ , 其中
(1.13) $\begin{equation} S_{1} S_{2}\left(v_{-}, u_{-}\right):=\left\{(v, u): u<u_{-}-\left(v-v_{-}\right) s_{i}\left(v, v_{-}\right), i=1,2\right\}, \end{equation}$
$s_{i}\left(v, v_{-}\right)=(-1)^{i} \sqrt{\left(p(v)-p\left(v_{l}\right)\right) /\left(v_{l}-v\right)}$ . 在上述假设下, 根据文献[38 ], 我们可以找到一个唯一的 $(\bar{v}, \bar{u}) \in S_{1}\left(v_{-}, u_{-}\right)$ 使得 $\left(v_{+}, u_{+}\right) \in$ $S_{2}(\bar{v}, \bar{u})$ , 其中
(1.14) $\begin{equation} S_{i}\left(v_{l}, u_{l}\right):=\left\{(v, u) ; u=u_{l}-\left(v-v_{-}\right) s_{i}\left(v, v_{l}\right), u<u_{l}\right\} \end{equation} $
是通过$\left(v_{l}, u_{l}\right)$ 的第 $i$ 族激波曲线.
容易知道系统(1.1)有一个连接$\left(v_{-}, u_{-}\right)$ 和$(\bar{v},\bar{u})$ 的第一族粘性激波$\left(V_{1}, U_{1}\right)\left(x-s_{1} t\right)$ , 与一个连接$(\bar{v}, \bar{u})$ 和 $\left(v_{+}, u_{+}\right)$ 的第二族粘性激波$\left(V_{2}, U_{2}\right)\left(x-s_{2} t\right)$ . 这两族粘性激波在相差一个适当平移下是唯一确定的. 其中 $ s_{1}=s_{1}\left(v_{-}, \bar{v}\right)<0,$ $s_{2}=s_{2}\left(v_{+}, \bar{v}\right)>0. $
在这种情况下, Navier-Stokrs方程组柯西问题 (1.1)-(1.2)解的大时间渐近行为可以通过第一族粘性激波$\left(V_{1}, U_{1}\right)\left(x-s_{1} t+\alpha_{1}\right)$ 和第二族粘性激波 $\left(V_{2}, U_{2}\right)\left(x-s_{2} t+\alpha_{2}\right)$ 的叠加来描述
(1.15) $\begin{equation} (V, U)\left(t, x ; \alpha_{1}, \alpha_{2}\right):=\left(V_{1}, U_{1}\right)\left(x-s_{1} t+\alpha_{1}\right)+\left(V_{2}, U_{2}\right)\left(x-s_{2} t+\alpha_{2}\right)-(\bar{v}, \bar{u}), \end{equation}$
平移 $\alpha_{1}$ 和 $\alpha_{2}$ 是由如下式子给定的
(1.16) $\begin{equation}\alpha_{1}=\frac{s_{2} A+B}{\delta_{1}\left(s_{1}-s_{2}\right)}, uad \alpha_{2}=\frac{s_{1} A+B}{\delta_{2}\left(s_{1}-s_{2}\right)}. \end{equation}$
(1.17) $\begin{equation} \begin{array}{ll} A =\int_{\mathbb{R} }\left[v_{0}(x)-V_{1}(x)-V_{2}(x)+\bar{v}\right] {\rm d} x<+\infty, \\[3mm] B =\int_{\mathbb{R} }\left[u_{0}(x)-U_{1}(x)-U_{2}(x)+\bar{u}\right] {\rm d} x<+\infty. \end{array} \end{equation}$
我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性.
当粘性系数依赖于密度(即满足(1.3)式)时, 与之对应也有许多相关结果. 文献[17 ,18 ]证明了当初始值具有大扰动且含真空情形时, 一维可压缩 Navier-Stokes 方程组稀疏波的非线性稳定性. 文献[34 ] 研究了在小初始扰动下, 一维可压缩 Navier-Stokes 方程组粘性激波的非线性稳定性(文中要求 $\kappa \geq\frac{\gamma-1}{2}$ ). 最近, 文献[7 ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ .
需要指出的是, 当 $\gamma=2, \kappa=1$ 时, (1.1)-(1.3)式对应于 Saint-Venant 浅水波方程组, 该方程组在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . (1.3)式中的结果很显然排除了这一重要的情形. 本学位论文的主要研究目的就是扩大 $\kappa$ 的取值范围, 改进(1.3)式中的结果至 $\kappa\geq 0$ .
现在我们来陈述本文的主要结果. 为此, 首先我们需要先引入以下符号:1 -粘性激波和2 -粘性激波的强度各自表示为 $\delta_{1}:=\left|v_{-}-\bar{v}\right|$ , $\delta_{2}:=\left|\bar{v}-v_{+}\right|$ . 再定义 $\delta:=\left|u_{+}-u_{-}\right|$ ,
(1.18) $\begin{equation} \left(\phi_{0}, \psi_{0}\right)(x):=\int_{-\infty}^{x}\left(v_{0}(y)-V\left(0, y ; \alpha_{1}, \alpha_{2}\right), u_{0}(y)-U\left(0, y ; \alpha_{1}, \alpha_{2}\right)\right) {\rm d} y. \end{equation}$
其次, 我们假设初始值 $\left(v_{0}, u_{0}\right)$ ,粘性激波的强度$\delta_{1}, \delta_{2}$ 和平移 $\alpha_{1}, \alpha_{2}$ 满足如下条件.
(H$_0)$ 存在与 $\delta$ 无关的常数 $\ell \geq 0$ 和 $C>0$ 使得对任意 $x \in \mathbb{R} $ ,有
(1.19) $\begin{equation} C^{-1} \delta^{\ell} \leq v_{0}(x) \leq C\left(1+\delta^{-\ell}\right); \end{equation}$
(H$_1)$ $\left(v_{+}, u_{+}\right) \in S_{1} S_{2}\left(v_{-}, u_{-}\right)$ 并且$(\bar{v}, \bar{u}) \in S_{1}\left(v_{-}, u_{-}\right)$ 使得 $\left(v_{+}, u_{+}\right) \in S_{2}(\bar{v}, \bar{u});$
(H$_2$ ) 假设粘性激波的强度$\delta_{1}, \delta_{2}$ , 平移 $\alpha_{1}, \alpha_{2}$ ( 由(1.16)定义)和初始值 $\left(v_{0}, u_{0}\right)$ 满足
(1.20) $\begin{equation} \begin{array}{ll} \left(v_{0}-V\left(0, \cdot ; \alpha_{1}, \alpha_{2}\right), u_{0}-U\left(0, \cdot ; \alpha_{1}, \alpha_{2}\right)\right) \in H^{1}(\mathbb{R} ) \cap L^{1}(\mathbb{R} ), \\ \left(\phi_{0}, \psi_{0}\right) \in L^{2}(\mathbb{R} ). \end{array} \end{equation}$
对于与$\delta$ 无关的正常数$C$ , 我们假设
(1.21) $\begin{equation}\label{t1.18} C^{-1} \delta_{2} \leq \delta_{1} \leq C \delta_{2}, uad \alpha_{2}-\alpha_{1} \leq C \delta^{-1}, uad \mbox{ 当 } \delta \rightarrow 0_{+} ; \end{equation}$
(H$_3$ ) $v_{-}$ 和 $v_{+}$ 是与 $\delta$ 无关的正常数.
定理 1.1 在假设 (H$_0)$ -( H$_3)$ 下, 我们进一步假设 $\kappa\geq0, \gamma>1$ 并且存在与 $\delta$ 无关的正常数$C, \alpha$ 和 $\beta$ 使得
(1.22) $\begin{equation}\label{t1.19} \left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{H^{1}(\mathbb{R} )} \leq C \delta^{\alpha}, uad\left\|\phi_{0 x x}\right\|_{L^{2}(\mathbb{R} )} \leq C\left(1+\delta^{-\beta}\right) \end{equation}$
成立. 如果参数$\ell, \alpha$ 和 $\beta$ 满足
(1.23) $\begin{equation}\left\{\begin{array}{l} (3 \gamma+5 \kappa+6) \ell<\min \{\frac{1}{2}, \alpha\}, \\ [3mm] \lbrack l(\kappa+1)+\beta\rbrack(2\gamma+2\kappa+3)<\min\{\frac{1}{2},2\alpha-(\gamma+1)l\}, \end{array}\right. \end{equation}$
那么存在一个足够小的 $\delta_{0}>0$ 使得如果 $0<\delta \leq \delta_{0}$ , 那么柯西问题(1.1)-(1.2) 有一个唯一的整体解 $(v(t,x), u(t,x))$ 满足
(1.24) $\begin{equation} \begin{array}{ll} \left(v(t, x)-V\left(t, x ; \alpha_{1}, \alpha_{2}\right), u(t, x)-U\left(t, x ; \alpha_{1}, \alpha_{2}\right)\right) \in C\left([0, \infty) ; H^{1}(\mathbb{R} )\right), \\ v(t, x)-V\left(t, x ; \alpha_{1}, \alpha_{2}\right) \in L^{2}\left(0, \infty ; H^{1}(\mathbb{R} )\right), \\ u(t, x)-U\left(t, x ; \alpha_{1}, \alpha_{2}\right) \in L^{2}\left(0, \infty ; H^{2}(\mathbb{R} )\right). \end{array} \end{equation} $
并且存在一个与$\delta$ 无关的正常数$C$ 使得
(1.25) $\begin{equation} C^{-1} \delta^{-2\tilde{\eta_2}} \le v(t,x) \le C \delta^{-\frac{l(\kappa+1)+\beta}{2}} \end{equation}$
$\tilde{\eta_2}=\min\left\{0,\frac{\eta_1-l(\kappa+1)-\beta}{2\kappa+\gamma-1}\right\}, uad\eta_1=\min\left\{\alpha-\frac{(\kappa+1)l}{2},\frac{1}{4}\right\}. $
(1.26) $\begin{equation}\lim _{t \rightarrow \infty} \sup _{x \in \mathbb{R} }\left|(v, u)(t, x)-(V, U)\left(t, x ; \alpha_{1}, \alpha_{2}\right)\right|=0.\end{equation}$
注 1.1 在(1.23)式中当 $\alpha$ 充分大, $\ell$ 和 $\beta$ 充分小时, $\gamma$ 和 $\kappa$ 的取值范围可包含 $\gamma=2, \kappa=1$ , 即包含对应于 Saint-Venant 浅水波方程组的情形.
注 1.2 注意到文献[7 ] 的主要结果中 $\alpha\leq\beta$ ,这使得比容的初始值具有大振幅. 而在我们的结果中, 由(1.23)式 知 $\alpha>\beta$ , 这不能保证比容的初始值具有大振幅. 事实上, 由(1.22)式知
$\begin{eqnarray*} \left\|\phi_{0x}\right\|_{L^{\infty}(\mathbb{R} )}\leq C\left\|\phi_{0x}\right\|_{L^{2}(\mathbb{R} )}^{\frac{1}{2}} \left\|\phi_{0xx}\right\|_{L^{2}(\mathbb{R} )}^{\frac{1}{2}}\leq C\delta^{\frac{\alpha-\beta}{2}}. \end{eqnarray*} $
$\alpha>\beta, \delta\rightarrow 0$ 时, $\left\|\phi_{0x}\right\|_{L^{\infty}(\mathbb{R} )}\rightarrow 0$ .
下面将介绍本文的一些主要想法. 研究方程组(1.1)-(1.2)粘性激波的非线性稳定性, 关键在于推导比容与时间 $t$ 无关的一致 上下界估计. 为此, 文献[7 ] 通过 $L^{2}$ 能量方法得出 $\left\|\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(t)\right\|_{L^{2}(\mathbb{R} )}$ ($\widetilde{v}=\frac{v}{V}$ ) 的与时间 $t$ 无关的先验估计.再通过 Kanel' 的方法[19 ] 得出比容 v 的一致上下界估计. 注意到该方法在推导比容上下界时, 我们要求 $0\leq\kappa<\frac{1}{2}$ . 为了扩大 $\kappa$ 的范围, 我们的主要想法如下: 首先可利用 Kanel'的方法[19 ] 推导比容的一致下界估计. 在推导比容一致上界估计时, 我们先作关于 $\|\phi_{x}\|_{L^{2}(\mathbb{R} )}$ 和 $\|\phi_{xx}\|_{L^{2}(\mathbb{R} )}$ 的精细估计. 再利用 Sobolev 不等式
$\begin{eqnarray*} \|v\|_{L^{\infty}(\mathbb{R} )}\leq C+C\|\phi_{x}\|_{L^{\infty}(\mathbb{R} )} \leq C+C\|\phi_{x}\|_{L^{2}(\mathbb{R} )}^{\frac{1}{2}} \|\phi_{xx}\|_{L^{2}(\mathbb{R} )}^{\frac{1}{2}}, \end{eqnarray*}$
便可得出比容的一致上界估计. 在得出比容的一致上下界估计后, 通过精细设计的连续性技巧, 我们可将方程组(1.1)-(1.2)的局部解延拓为整体解, 并得出整体解的大时间渐近行为.
本文所用数学符号的相关说明: 在本文中, $c$ 和$C$ 被用来表示各种与$\delta$ 无关的一般的正的常数, $\delta$ 是粘性冲击波的强度. 对于正常数 $C$ , 我们将使用 $A \lesssim B(B \gtrsim A)$ 来表示 $A \leq C B(B \geq CA)$ . 符号$A \sim B$ 意味着$A \lesssim B$ 以及 $B \lesssim A$ . 对于函数空间, $L^q\left(\mathbb{R} \right)\left(1\leq q\leq \infty\right)$ 表示在 $\mathbb{R} $ 上通常意义下的 Lebesgue 空间, 其范数记为 $\|\cdot\|_{L^q\left(\mathbb{R} \right)}$ . $W^{k,p}\left(\mathbb{R} \right)\left(1\leq p\leq \infty\right)$ 表示通常意义下的 Sobolev 空间, 且 $H^{k}\left(\mathbb{R} \right):= W^{k,2}\left(\mathbb{R} \right)$ . 为简单起见, 我们记 $\|\cdot\|_{q}:=\|\cdot\|_{H^{q}\left(\mathbb{R} \right)}$ , $\|\cdot\|:=\|\cdot\|_{L^{2}\left(\mathbb{R} \right)}$ . 在后文中我们将使用记号$(V, U)(t, x)$ 来表示 $(V, U)\left(t, x ; \alpha_{1}, \alpha_{2}\right)$ .
2 粘性激波的一些性质
本节介绍性激波$(V_i,U_i)(t,x)\ (i=1,2)$ 及其叠加的一些基本性质.
我们首先陈述粘性激波$(V_i,U_i)(t,x)\ (i=1,2)$ 的存在性以及当$x-s_it \to \pm \infty$ 时的衰减估计. 事实上, 由文献[12 ] 及假设 (H$_3$ ) 可得
引理 2.1 在假设条件(H$_0)$ -( H$_3)$ 成立的前提下, 一维可压缩 Navier-Stokes 方程组(1.1)-(1.2)存在连接$(v_-,u_-)$ 和$(\overline{v},\overline{u})$ 的速度为$s_1$ 的第一族粘性激波$(V_1(x-s_1t),U_1(x-s_1t))$ , 以及连接$(\overline{v},\overline{u})$ 和$(v_+,u_+)$ 的速度为$s_2$ 的第二族粘性激波$(V_1(x-s_1t),U_1(x-s_1t))$ . 这两族粘性激波在相差一个平移的意义下是唯一的, 并且存在只依赖与$v_-$ 和$v_+$ 的正常数$c$ , 使得对$i=1,2$ 有
(2.1) $\begin{equation} \begin{array}{lll} \left|\left(V_{1}(\xi)-\bar{v}, U_{1}(\xi)-\bar{u}\right)\right| \lesssim \delta_{1} e^{-c \delta_{1}|\xi|}, & \forall \xi>0, \\ \left|\left(V_{2}(\xi)-\bar{v}, U_{2}(\xi)-\bar{u}\right)\right| \lesssim \delta_{2} e^{-c \delta_{2}|\xi|}, & \forall \xi<0, \\ \left|V_{1}^{\prime}(\xi)\right| \lesssim\left|V_{1}(\xi)-v_{-}\right|\left|V_{1}(\xi)-\bar{v}\right|, & \forall \xi \in \mathbb{R}, \\ \left|V_{2}^{\prime}(\xi)\right| \lesssim\left|V_{2}(\xi)-\bar{v}\right|\left|V_{2}(\xi)-v_{+}\right|, & \forall \xi \in \mathbb{R}, \\ U_{i}^{\prime}(\xi) <0, & \forall \xi \in \mathbb{R}, \\ \left|\left(U_{i}^{\prime}(\xi), V_{i}^{\prime \prime}(\xi), U_{i}^{\prime \prime}(\xi)\right)\right| \lesssim\left|V_{i}^{\prime}(\xi)\right| \lesssim \delta_{i}^{2} e^{-c \delta_{i}|\xi|}, & \forall \xi \in \mathbb{R}. \end{array} \end{equation}$
注意到$(V_i,U_i)(t,x)\ (i=1,2)$ 是可压缩Navier-Stokes方程(1.1)的精确解, 它们的叠加$(V,U)(t,x)$ 满足
$\left\{\begin{array}{l}V_{t}-U_{x}=0, \\ U_{t}+p(V)_{x}=\left(\mu(V) \frac{U_{x}}{V}\right)_{x}-g_{x},\end{array} uad t>0, x \in \mathbb{R} \right.$
(2.2) $\begin{equation}g=\mu(V) \frac{U_{x}}{V}-\mu\left(V_{1}\right) \frac{U_{1 x}}{V_{1}}-\mu\left(V_{2}\right) \frac{U_{2 x}}{V_{2}}-p(V)+p\left(V_{1}\right)+p\left(V_{2}\right)-p(\bar{v}).\end{equation}$
下面的引理是对$g(t,x)$ 的一些估计, 这在后续的能量估计发挥重要的作用.
(2.3) $\begin{equation}\int_{0}^{\infty}\|g(t)\| {\rm d} t \lesssim \delta^{\frac{1}{2}}, uad \int_{0}^{\infty}\left(\left\|g_{x}(t)\right\|+\left\|g_{x x}(t)\right\|\right) {\rm d} t \lesssim \delta^{\frac{3}{2}}.\end{equation}$
证 我们仅考虑$\alpha_1<\alpha_2$ 的情形, 对于其他的情形可以类似处理. 在这种情况下, 因为$U=U_1+U_2-\bar{u}$ , 所以有
(2.4) $\begin{matrix}g&=& \mu_{0} \frac{U_{x}}{V^{\kappa+1}}-\mu_{0} \frac{U_{1 x}}{V_{1}^{\kappa+1}}-\mu_{0} \frac{U_{2 x}}{V_{2}^{\kappa+1}}-p(V)+p\left(V_{1}\right)+p\left(V_{2}\right)-p(\bar{v}) \\&=& \mu_{0} U_{1 x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{1}^{\kappa+1}}\right)+\mu_{0} U_{2 x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{2}^{\kappa+1}}\right) \\&&-\left(p(V)-p\left(V_{1}\right)\right)+\left(p\left(V_{2}\right)-p(\bar{v})\right) \\&=&-\mu_{0} U_{1 x}\left(\int_{0}^{1}(\kappa+1)\left(\theta V+(1-\theta) V_{1}\right)^{-(\kappa+1)} {\rm d}\theta\right)\left(V-V_{1}\right) \\&&+\mu_{0} U_{2 x}\left(\int_{0}^{1}(\kappa+1)\left(\theta V+(1-\theta) V_{1}\right)^{-(\kappa+1)} {\rm d}\theta\right)\left(V-V_{2}\right) \\&&-\left(\int_{0}^{1} p^{\prime}\left(\theta V+(1-\theta) V_{1}\right) {\rm d}\theta\right)\left(V-V_{1}\right) +\left(\int_{0}^{1} p^{\prime}\left(\theta V_{2}+(1-\theta) \bar{v}\right) {\rm d}\theta\right)\left(V_{2}-\bar{v}\right) \\&=&-\mu_{0} U_{1 x}\left(\int_{0}^{1}(\kappa+1)\left(\theta V+(1-\theta) V_{1}\right)^{-(\kappa+1)} {\rm d}\theta\right)\left(V-V_{1}\right) \\&&+\mu_{0} U_{2 x}\left(\int_{0}^{1}(\kappa+1)\left(\theta V+(1-\theta) V_{1}\right)^{-(\kappa+1)} {\rm d}\theta\right)\left(V-V_{2}\right)\\&&-\left(\int_{0}^{1} \int_{0}^{1} p^{\prime \prime}\left(\theta_{1}\left(V_{1}-\bar{v}+\theta V_{2}+(1-\theta) \bar{v}\right)\right) {\rm d}\theta {\rm d}\theta_{1}\right)\left(V_{1}-\bar{v}\right)\left(V_{2}-\bar{v}\right).\end{matrix}$
$\left|U_{1 x}\right| \lesssim\left|V_{1}-\bar{v}\right|, uad\left|U_{2 x}\right| \lesssim\left|V_{2}-\bar{v}\right|, uad V-V_{1}=V_{2}-\bar{v}, uad V-V_{2}=V_{1}-\bar{v},$
(2.5) $\begin{matrix}|g(t, x)| & \lesssim&\left(\left|V_{1}-\bar{v}\right|\left|V-V_{1}\right|+\left|V_{2}-\bar{v}\right|\left|V-V_{2}\right|+\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|\right) \\& \lesssim&\left|V_{1}-\bar{v}\right|\left|V_{2}-\bar{v}\right|.\end{matrix}$
同理, 对于 $g_{x}(t, x)$ , 我们有
(2.6) $\begin{matrix}g_{x}&=& \underbrace{\mu_{0} U_{1 x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{1}^{\kappa+1}}\right)_{x}}_{J_{1}}+\underbrace{\mu_{0} U_{1 x x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{1}^{\kappa+1}}\right)}_{J_{2}} \\&&+\underbrace{\mu_{0} U_{2 x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{2}^{\kappa+1}}\right)_{x}}_{J_{3}}+\underbrace{\mu_{0} U_{2 x x}\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{2}^{\kappa+1}}\right)}_{J_{4}} \\&&+\underbrace{\left(-p^{\prime}(V) V_{x}+p^{\prime}\left(V_{1}\right) V_{1 x}+p^{\prime}\left(V_{2}\right) V_{2 x}-p^{\prime}(\bar{v}) \bar{v}_{x}\right)}_{J_{5}}.\end{matrix}$
(2.7) $\begin{matrix}\left|\left(\frac{1}{V^{\kappa+1}}-\frac{1}{V_{1}^{\kappa+1}}\right)_{x}\right| &=&|\kappa+1|\left|\frac{V_{x}}{V^{\kappa+2}}-\frac{V_{1 x}}{V_{1}^{\kappa+2}}\right| \lesssim\left|\left(V-V_{1}\right)_{x}\right| \\&\lesssim&\left|V_{2 x}\right| \lesssim\left|V_{2}-\bar{v} \|\right| V_{2}-V_{+}\left|\lesssim \delta_{2}\right| V_{2}-\bar{v} \mid,\end{matrix}$
$\left|\left(U_{1 x}, U_{1 x x}\right)\right| \leq\left|V_{1}-\bar{v}\right|,$
(2.8) $\begin{equation}\begin{array}{ll}\left|J_{1}\right| \lesssim \delta_{2}\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|, \\\left|J_{2}\right| \lesssim\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|.\end{array}\end{equation}$
(2.9) $\begin{equation}\begin{array}{ll}\left|J_{3}\right| \lesssim \delta_{1}\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|,\\\left|J_{4}\right| \lesssim\left|V_{1}-\bar{v}\right|\left|V_{2}-\bar{v}\right|.\end{array}\end{equation}$
(2.10) $\begin{matrix}J_{5}&=&-\left(\int_{0}^{1} p^{\prime \prime}\left(\theta V+(1-\theta) V_{1}\right) {\rm d}\theta\right) V_{1 x}\left(V-V_{1}\right) \\&&-\left(\int_{0}^{1} p^{\prime \prime}\left(\theta V+(1-\theta) V_{2}\right) {\rm d}\theta\right) V_{2 x}\left(V-V_{2}\right) \\& \lesssim&\left(\left|V-V_{1}\right|\left|V_{1 x}\right|+\left|V-V_{2}\right|\left|V_{2 x}\right|\right) \\& \lesssim&\left(\delta_{1}+\delta_{2}\right)\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|,\end{matrix}$
(2.11) $\begin{equation}\left|g_{x}(t, x)\right| \lesssim\left(\delta_{1}+\delta_{2}+1\right)\left|V_{1}-\bar{v}\left\|V_{2}-\bar{v}|\lesssim(\delta+1)| V_{1}-\bar{v}\right\| V_{2}-\bar{v}\right|.\end{equation}$
(2.12) $\begin{equation}\left|g_{x x}(t, x)\right| \lesssim(\delta+1)\left|V_{1}-\bar{v} \| V_{2}-\bar{v}\right|.\end{equation}$
然后, 我们将上平面 $\mathbb{R} _{+} \times \mathbb{R} $ 分为六个区域
$\begin{eqnarray*}&&\Omega_{1}=\{(t, x): \left.0 \leq t \leq t_{0}, x \leq s_{2} t-\alpha_{2}\right\}, \\&&\Omega_{2}=\{(t, x): \left.0 \leq t \leq t_{0}, s_{2} t-\alpha_{2}<x \leq s_{1} t-\alpha_{1}\right\}, \\&&\Omega_{3}=\{(t, x): \left.0 \leq t \leq t_{0}, x \geq s_{1} t-\alpha_{1}\right\}, \\&&\Omega_{4}=\{(t, x): \left.t>t_{0}, x \leq s_{1} t-\alpha_{1}\right\}, \\&&\Omega_{5}=\{(t, x): \left.t>t_{0}, s_{1} t-\alpha_{1}<x \leq s_{2} t-\alpha_{2}\right\}, \\&&\Omega_{6}=\{(t, x): \left.t>t_{0}, x \geq s_{2} t-\alpha_{2}\right\},\end{eqnarray*}$
其中 $t_{0}=\left(\alpha_{2}-\alpha_{1}\right) /\left(s_{2}-s_{1}\right)$ .
将上半平面 $\mathbb{R} _{+} \times \mathbb{R} $ 做以上划分后, 利用引理2.1和(2.5),(2.11), (2.12)式可以很容 易的得到(2.3)式. 引理证毕.
接下来我们给出下面的定义, 主要的目的是重新改写方程(1.1).
$(\phi, \psi)(t, x):=\int_{-\infty}^{x}\left(v(t, y)-V\left(t, y ; \alpha_{1}, \alpha_{2}\right), u(t, y)-U\left(t, y ; \alpha_{1}, \alpha_{2}\right)\right){\rm d}y,$
利用方程(1.1)和(2.2)式重新构建原问题得到
(2.13) $\begin{equation}\left\{\begin{array}{l}\phi_{t}-\psi_{x}=0, \\ \psi_{t}+p(v)-p(V)=\left(\mu(v) \frac{u_{x}}{v}-\mu(V) \frac{U_{x}}{V}\right)+g, \\\left.(\phi, \psi)\right|_{t=0}=\left(\phi_{0}, \psi_{0}\right).\end{array}\right.\end{equation}$
$X_{m, M}(0, T)=\left\{\begin{array}{c|c}(\phi(t, x), \psi(t, x)) & \begin{array}{c}(\phi(t, x), \psi(t, x)) \in C\left([T] ; H^{2}(\mathbb{R} )\right), \\\phi_{x}(t, x) \in L^{2}\left(0, T ; H^{1}(\mathbb{R} )\right), \\\psi_{x}(t, x) \in L^{2}\left(0, T ; H^{2}(\mathbb{R} )\right), \\m \leq V(t, x)+\phi_{x}(t, x) \leq M \end{array} \end{array}\right\},$
我们之所以定义这个函数空间是因为我们将要在这个空间中寻找方程组(2.13)的解.
3 定理1.1的证明
在函数集$X_{m, M}(0, T)$ 中, Cauchy问题(2.13)解的局部存在性定理如下.
引理 3.1 若 $\left(\phi_{0}(x), \psi_{0}(x)\right)\in H^{2}(\mathbb{R} )$ 且满足$\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}\leq M_0$ , 对任意 $x \in \mathbb{R} $ 有
$m \leq V(0, x)+\phi_{0 x}(x) \leq M,$
则存在一个仅依赖于 $m, M$ 和 $\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}$ 的 $t_{0}>0$ 使得柯西问题(2.13)存在唯一的解 $(\phi(t, x)$ , $\psi(t, x)) \in X_{m / 2,2 M}\left(0, t_{0}\right)$ , 并且对任意的$0\leq t \leq t_{0}$ 有
(3.1) $\begin{equation}\|\psi(t)\| \leq 2\left\|\psi_{0}\right\|, uad\left\|\psi_{x}(t)\right\| \leq 2\left\|\psi_{0 x}\right\|, uad\|(\phi, \psi)(t)\|_{2} \leq 2\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}.\end{equation}$
在本文中我们的目的是为了得到整体解, 假设性质(3.1)中得到的局部解已经延拓到了时间 $t=T \geq t_{0}$ . 为了能够证明 $T=\infty$ , 我们需要在一定的先验假设下推导出 $(\phi(t, x)$ , $\psi(t, x))$ 的某些与时间 $t$ 无关的先验估计, 从而在此基础上利用连续性技巧将局部解延拓 到整体解. 本文所给出的先验假设是 $(\phi(t, x), \psi(t, x)) \in X_{1 / m, M}(0, T)$ , 这样就有
(3.2) $\begin{equation}\frac{1}{m} \leq V(t, x)+\phi_{x}(t, x)=v(t, x) \leq M, uad \forall(t, x) \in[T] \times \mathbb{R}.\end{equation}$
其中 $m$ 和 $M$ 是正常数.我们注意到 $c$ 和 $C$ 是与 $T, m, M$ 及 $\delta$ 无关的正常数, 同时在末知函数为 $\phi$ 和 $\psi$ 的情 况下, 文中经常会用到记号 $(v, u)=\left(V+\phi_{x}, U+\psi_{x}\right)$ . 此外, 为了方便起见将 $ N_{\psi}(T):=\sup _{[T]}\|\psi(t)\|_{L^{\infty}(\mathbb{R} )}$ , 简记为 $N_{\psi}$ . 并且不失一般性, 我们可以令 $m \geq 1$ 和 $M \geq 1$ .
引理 3.2 在定理1.1的假设下, 存在与 $\delta$ 无关的足够小的正常数 $\delta_{1}$ , 使得如果 $\delta$ 满足$0<\delta \leq \delta_{1}$ , 则对任意的 $0 \leq t \leq T$ 有
(3.3) $\begin{matrix}&&\|(\phi, \psi)(t)\|^{2}+\int_{0}^{t} \int_{\mathbb{R} }\left(\left|V_{t}\right| \psi^{2}+\psi_{x}^{2}\right) {\rm d} x {\rm d} \tau \\&\lesssim &\left\|\left(\phi_{0}, \psi_{0}\right)\right\|^{2}+\delta^{\frac{1}{2}}+\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau+C_{1}\left(m, M, \delta, N_{\psi}\right) \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau.\end{matrix}$
(3.4) $\begin{equation}C_{1}\left(m, M, \delta, N_{\psi}\right)=N_{\psi} m^{\gamma+2}+N_{\psi}^{2} M^{2 \kappa} m^{\kappa+1}+M^{2 \kappa} m^{2(\kappa+1)} \delta^{2}.\end{equation}$
(3.5) $\begin{matrix}&& \psi_{t}+p^{\prime}(V) \phi_{x}-\mu(V) \frac{\psi_{x x}}{V}+\left(p(v)-p(V)-p^{\prime}(V) \phi_{x}\right) \\&=&\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) \psi_{x x}+\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x}+g.\end{matrix}$
将$(2.13)_1$ 和(3.5)式分别乘以 $\phi$ 和 $-p^{\prime}(V)^{-1} \psi$ 再相加, 且
$\begin{eqnarray*}&&\phi \phi_{t}-\phi \psi_{x}-\psi \phi_{x} =\left(\frac{1}{2} \phi^{2}\right)_{t}-(\phi \psi)_{x} \\&&-\frac{\psi \psi_{t}}{p^{\prime}(V)} =\left(-\frac{\psi^{2}}{2 p^{\prime}(V)}\right)_{t}-\frac{p^{\prime \prime}(V) V_{t} \psi^{2}}{2 p^{\prime}(V)^{2}} \\&&\frac{\mu(V) \psi_{x x} \psi}{V p^{\prime}(V)} =\left(\frac{\mu(V) \psi_{x} \psi}{V p^{\prime}(V)}\right)_{x}-\frac{\mu(V) \psi}{V p^{\prime}(V)} \psi_{x}^{2}-\left(\frac{\mu(V) \psi}{V p^{\prime}(V)}\right)^{\prime} V_{x} \psi \psi_{x},\end{eqnarray*}$
(3.6) $\begin{matrix}&&\left(\frac{1}{2} \phi^{2}-\frac{\psi^{2}}{2 p^{\prime}(V)}\right)_{t}-\frac{p^{\prime \prime}(V) V_{t} \psi^{2}}{2 p^{\prime}(V)^{2}}-\frac{\mu(V) \psi_{x}^{2}}{V p^{\prime}(V)}-\left\{\phi \psi-\frac{\mu(V) \psi_{x} \psi}{V p^{\prime}(V)}\right\}_{x} \\&=&\left(\frac{\mu(V)}{V p^{\prime}(V)}\right)^{\prime} V_{x} \psi \psi_{x}+\left(p(v)-p(V)-p^{\prime}(V) \phi_{x}\right) \frac{\psi}{p^{\prime}(V)} \\&&-\frac{\psi}{p^{\prime}(V)}\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) \psi_{x x}-\frac{\psi}{p^{\prime}(V)}\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x}-\frac{g \psi}{p^{\prime}(V)}.\end{matrix}$
$\mu(V)=\mu_{0} V^{-\kappa}, p(V)=a V^{-\gamma}, p^{\prime}(V)=-\gamma a V^{-\gamma-1}, p^{\prime \prime}(V)=\gamma(\gamma+1) a V^{-\gamma-2},$
$\frac{p^{\prime \prime}(V)}{p^{\prime}(V)^{2}} \sim V^{\gamma}, uad \frac{\mu(V)}{V p^{\prime}(V)} \sim V^{\gamma-\kappa}.$
由于 $\delta$ 与 $v_{-}$ 和 $v_{+}$ 的取值无关, 我们可以假设 $\delta$ 取足够小的数, 这样就可以 推导出 $V(t, x)$ 存在与 $\delta$ 无关的正的上下界. 于是将等式(3.6)的两边关于 $t, x$ 在 $[t] \times \mathbb{R} $ 上 积分有
(3.7) $\begin{matrix}&&\|(\phi, \psi)(t)\|^{2}+\int_{0}^{t} \int_{\mathbb{R} }\left[\left|V_{t}\right| \psi^{2}+\psi_{x}^{2}\right] \\&\lesssim &\left\|\left(\phi_{0}, \psi_{0}\right)\right\|^{2}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|V_{x} \psi \psi_{x}\right|}_{I_{1}}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|\left(p(v)-p(V)-p^{\prime}(V) \phi_{x}\right) \psi\right|}_{I_{2}} \\&&+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|\left[\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right] \psi \psi_{x x}\right|}_{I_{3}}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} } \|\left[\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right] \psi U_{x} \mid}_{I_{4}}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }|\psi g|}_{I_{5}}.\end{matrix}$
(3.8) $\begin{equation}\left|p(v)-p(V)-p^{\prime}(V) \phi_{x}\right| \lesssim m^{\gamma+2} \phi_{x}^{2}, uad\left|\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right| \lesssim \frac{M^{\kappa}\left|\phi_{x}\right|}{v^{\kappa+1} V^{\kappa+1}}.\end{equation}$
注意到 $C \delta^{2} \geq-V_{t}=-U_{x}$ , 于是利用 Cauchy 不等式和Holder不等式可以得到, 对任意 的 $\epsilon>0$ 有
$\begin{eqnarray*}I_{1} & \leq& \epsilon \int_{0}^{t} \int_{\mathbb{R} } \psi_{x}^{2} {\rm d} x {\rm d} \tau+C(\epsilon) \int_{0}^{t} \int_{\mathbb{R} } V_{x}^{2} \psi^{2} {\rm d} x {\rm d} \tau, \\I_{2} & \lesssim& N_{\psi} m^{\gamma+2} \int_{0}^{t} \int_{\mathbb{R} } \phi_{x}^{2} {\rm d} x {\rm d} \tau, \\I_{3} & \lesssim& M^{\kappa} \int_{0}^{t} \int_{\mathbb{R} } \frac{V^{\gamma-\kappa}\left|\phi_{x} \| \psi \psi_{x x}\right|}{v^{\kappa+1}} {\rm d} x {\rm d} \tau \\& \lesssim &\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau+N_{\psi}^{2} M^{2 \kappa} m^{\kappa+1} \int_{0}^{t} \int_{\mathbb{R} } \phi_{x}^{2} {\rm d} x {\rm d} \tau, \\I_{4} & \lesssim& M^{\kappa} \int_{0}^{t} \int_{\mathbb{R} } \frac{V^{\gamma-\kappa}\left|U_{x} \| \psi\right|}{v^{\kappa+1}} {\rm d} x {\rm d} \tau \\& \lesssim& \epsilon \int_{0}^{t} \int_{\mathbb{R} }\left|U_{x}\right| \psi^{2} {\rm d} x {\rm d} \tau+M^{2 \kappa} \delta^{2} m^{2 \kappa+2} \int_{0}^{t} \int_{\mathbb{R} } \psi_{x}^{2} {\rm d} x {\rm d} \tau, \\I_{5} & \lesssim &\int_{0}^{t}\|V\|_{L^{\infty}\left(\mathbb{R} _{+} \times \mathbb{R} \right)}^{\frac{\gamma+1}{2}}\left\|V^{\frac{\gamma+1}{2}} \psi(\tau)\right\|\|g(\tau)\| {\rm d} \tau \\& \lesssim &\delta^{1 / 2}+\int_{0}^{t}\|\psi(\tau)\|^{2}\|g(\tau)\| {\rm d} \tau.\end{eqnarray*}$
取 $\epsilon$ 足够的小, 将上面 $I_{j}(j=1,2,3,4,5)$ 的估计代入(3.7)式, 结合(2.3)式并利用Gronwall不等 式完成了引理的证明.
利用(3.3)式, 我们希望得到对 $\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau$ 以及 $\int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau$ 的估计. 首先对时空 估计$\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau$ .
引理 3.3 在定理1.1的假设下, 当 $\delta$ 取足够小的数时, 则对任意的 $0 \leq t \leq T$ 有
(3.9) $\begin{matrix}&&\left\|\left(\phi, \psi, \psi_{x}, \sqrt{\Phi}, M^{-\frac{\gamma+1}{2}} \phi_{x}\right)(t)\right\|^{2}+\int_{0}^{t} \int_{\mathbb{R} }\left[\psi_{x}^{2}+\frac{\psi_{x x}^{2}}{v^{\kappa+1}}\right] {\rm d} x {\rm d} \tau \\& \lesssim&\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\delta^{\frac{1}{2}}+C_{2}\left(m, M, \delta, N_{\psi}\right) \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau.\end{matrix}$
其中 $\Phi_{0}=\left.\Phi\right|_{t=0}$ ,
(3.10) $\begin{equation}\begin{array}{ll} \Phi=\Phi(v, V)=p(V)(v-V)-\int_{V}^{v} p(\eta) {\rm d} \eta, \\ C_{2}\left(m, M, \delta, N_{\psi}\right)=N_{\psi} m^{\gamma+2}+N_{\psi}^{2} M^{2 \kappa} m^{\kappa+1}+M^{2 \kappa} m^{2(\kappa+1)} \delta^{2}+m^{\gamma+2} \delta^{2}+M^{2 \kappa} m^{\kappa+1} \delta^{4}.\end{array}\end{equation}$
证 将方程(2.13)关于 $x$ 求一阶偏导数, 并分别乘以 $p(V)-p(v)$ 和 $\psi_{x}$ , 将所得的 结果相加得到
(3.11) $\begin{matrix}&&\left\{\Phi+\frac{1}{2} \psi_{x}^{2}\right\}_{t}+\mu(v) \frac{\psi_{x x}^{2}}{v}+\left\{\psi_{x}\left(p(v)-p(V)-\mu(v) \frac{\psi_{x x}}{v}-\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x}\right)\right\}_{x} \\&=&-\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \psi_{x x}+\psi_{x} g_{x}-V_{t}\left(p(v)-p(V)-p^{\prime}(V) \phi_{x}\right).\end{matrix}$
$\begin{eqnarray*}(p(V)-p(v))(v-V)_{t} &=& p(V)(v-V)_{t}-p(v)(v-V)_{t} \\&=&(p(V)(v-V))_{t}-p^{\prime}(V) V_{t}(v-V)-p(v) v_{t}+p(v) V_{t} \\&=&(p(V)(v-V))_{t}-\left(p(v) v_{t}-p(V) V_{t}\right)-p(V) V_{t}\\&&+p(v) V_{t}-p^{\prime}(V) V_{t}(v-V) \\&=&\left(p(V)(v-V)-\int_{V}^{v} p(\eta) {\rm d} \eta\right)_{t}+V_{t}\left(p(v)-p(V)-p^{\prime}(V)(v-V)\right) \\&=& \Phi(v, V)_{t}+V_{t}\left(p(v)-p(V)-p^{\prime}(V)(v-V)\right)\end{eqnarray*}$
$\begin{eqnarray*}\left(\frac{\mu(v) u_{x}}{v}-\frac{\mu(V) U_{x}}{V}\right)_{x} \psi_{x} &=&\left(\frac{\mu(v) \psi_{x x}}{v}+\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x}\right)_{x} \psi_{x} \\&=&\left(\frac{\mu(v) \psi_{x x} \psi_{x}}{v}\right)_{x}-\frac{\mu(v) \psi_{x x}}{v}+\left(\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \psi_{x}\right)_{x}\\&&-\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \psi_{x x}.\end{eqnarray*}$
将等式(3.11)两边关于 $t, x$ 在 $[t] \times \mathbb{R} $ 上积分有
(3.12) $\begin{matrix}&&\left\|\left(\sqrt{\Phi}, \psi_{x}\right)(t)\right\|^{2}+\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau \\&\lesssim &\left\|\left(\sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \psi_{x x}\right| {\rm d} x {\rm d} \tau}_{I_{6}} \\&&+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|\psi_{x} \| g_{x}\right| {\rm d} x {\rm d} \tau}_{I_{7}}+\underbrace{\int_{0}^{t} \int_{\mathbb{R} }\left|V_{t}\left(p(v)-p(V)-p^{\prime}(V) \phi_{x}\right)\right| {\rm d} x {\rm d} \tau}_{I_{8}}.\end{matrix}$
这里 $\Phi_{0}=\left.\Phi\right|_{t=0}$ . 下面我们估计(3.12)式右端诸项, 由于
$v^{\kappa+1}-V^{\kappa+1}=(\kappa+1)\left(\int_{0}^{1}(\theta v+(1-\theta) V)^{\kappa} {\rm d} \theta\right) \phi_{x} \lesssim M^{\kappa}\left|\phi_{x}\right|.$
(3.13) $\begin{matrix}I_{6} &=&\int_{0}^{t} \int_{\mathbb{R} }\left|\frac{\mu_{0}\left(V^{\kappa+1}-v^{\kappa+1}\right)}{v^{\kappa+1} V^{\kappa+1}} U_{x} \psi_{x x}\right| {\rm d} x {\rm d} \tau \\& \leq &\epsilon \int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau+C(\epsilon) M^{2 \kappa} m^{\kappa+1} \delta^{4} \int_{0}^{t} \int_{\mathbb{R} } \phi_{x}^{2} {\rm d} x {\rm d} \tau.\end{matrix}$
对 $I_{7}$ 利用Holder不等式并结合(2.3)式可以得到
(3.14) $\begin{equation}I_{7} \leq \int_{0}^{t}\left\|g_{x}(\tau)\right\|\left\|\psi_{x}(\tau)\right\|^{2} {\rm d}\tau+\int_{0}^{t}\left\|g_{x}(\tau)\right\| {\rm d}\tau \lesssim \int_{0}^{t}\left\|g_{x}(\tau)\right\|\left\|\psi_{x}(\tau)\right\|^{2} {\rm d}\tau+\delta^{\frac{3}{2}}.\end{equation}$
(3.15) $\begin{equation}I_{8} \lesssim m^{\gamma+2} \delta^{2} \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d}\tau.\end{equation}$
将(3.13),(3.14),(3.15)式代入(3.12)式, 并注意到
$\begin{equation}\Phi(v, V)=-\phi_{x}^{2} \int_{0}^{1} \int_{0}^{1} \theta_{1} p^{\prime}\left(\left(1-\theta_{2} \theta_{1}\right) V+\theta_{2} \theta_{1} v\right) {\rm d} \theta_{1} {\rm d} \theta_{2} \gtrsim M^{-\gamma-1} \phi_{x}^{2},\end{equation}$
于是利用(3.2)式及Gronwall不等式可以推得
$\begin{eqnarray*}&&\left\|\left(\sqrt{\Phi}, \psi_{x}, M^{-\frac{\gamma+1}{2}} \phi_{x}\right)(t)\right\|^{2}+\int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d} x {\rm d} \tau \\&\lesssim &\left\|\left(\sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\left[m^{\gamma+2} \delta^{2}+M^{2 \kappa} m^{\kappa+1} \delta^{4}\right] \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d}\tau+\delta^{\frac{3}{2}}.\end{eqnarray*}$
因为 $m \geq 1$ , 将上面的不等式和(3.3)式经过适当的线性组合, 从而完成引理 $3.3$ 的证明.
我们接下来控制 $\int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau$ . 为了达到这个目的, 我们将(2.13)式两边同时乘以 $\phi_{x}$ 整理得到
(3.17) $\begin{equation}\left(\phi_{x} \psi\right)_{t}+\psi_{x}^{2}-\left(\psi \phi_{t}\right)_{x}-\frac{\mu(v) \psi_{x x} \phi_{x}}{v}-g \phi_{x} =-(p(v)-p(V)) \phi_{x}+\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \phi_{x}.\end{equation}$
注意到先验假设(3.2)以及 $U_{x}(t, x)<0$ , 我们有
$-(p(v)-p(V)) \phi_{x} \gtrsim M^{-(\gamma+1)} \phi_{x}^{2},$
$\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \phi_{x}=-\frac{\mu_{0}(\kappa+1) \int_{0}^{1}(\theta v+(1-\theta) V)^{\kappa} {\rm d} \theta \phi_{x}^{2}}{v^{\kappa+1} V^{\kappa+1}} U_{x}>0,$
从而将(3.17)式关于 $t$ 和 $x$ 在 $[t] \times \mathbb{R} $ 上积分, 并结合上面所得的不等式及Cauchy 和Holder 不等式可以得到
(3.18) $\begin{matrix}M^{-\gamma-1} \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d}\tau &\lesssim&\left\|\phi_{x}(t)\right\|\|\psi(t)\|+\left\|\phi_{0 x}\right\|\left\|\psi_{0}\right\|+\int_{0}^{t}\left\|\psi_{x}(\tau)\right\|^{2} {\rm d}\tau\\&&+M^{\gamma+1} m^{\kappa+1} \int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d}x {\rm d}\tau+\sup _{0 \leq \tau \leq t}\left\|\phi_{x}(\tau)\right\| \int_{0}^{t}\|g(\tau)\| {\rm d}\tau.number\end{matrix}$
$\left|\int_{0}^{t} \int_{\mathbb{R} } \frac{\mu(v) \psi_{x x} \phi_{x}}{v} {\rm d}x {\rm d}\tau\right| \leq \epsilon M^{-\gamma-1} \int_{0}^{t} \int_{\mathbb{R} } \phi_{x}^{2} {\rm d}x {\rm d}\tau+C(\epsilon) M^{\gamma+1} m^{\kappa+1} \int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v^{\kappa+1}} {\rm d}x {\rm d}\tau,$
$\int_{0}^{t} \int_{\mathbb{R} }\left|g \phi_{x}\right| {\rm d}x {\rm d}\tau \leq \sup _{0 \leq \tau \leq t}\left\|\phi_{x}(\tau)\right\| \int_{0}^{t}\|g\| {\rm d}\tau.$
由(2.3),(3.9),(3.18)式可以得到以下的估计
(3.19) $\begin{matrix}\int_{0}^{t}\left\|\phi_{x}^{2}(\tau)\right\| {\rm d}\tau &\lesssim & M^{2(\gamma+1)} m^{\kappa+1}\left\{\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\delta^{\frac{1}{2}}\right\} \\&&+M^{2(\gamma+1)} m^{\kappa+1} C_{2}\left(m, M, \delta, N_{\psi}\right) \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau.\end{matrix}$
$M^{2(\gamma+1)} m^{\kappa+1} C_{2}\left(m, M, \delta, N_{\psi}\right) \leq M^{2(\gamma+\kappa+1)} m^{\gamma+3 \kappa+3}\left(N_{\psi}+\delta^{2}\right),$
引理 3.4 在定理1.1的假设下, 存在一个与 $\delta$ 无关的充分小的 $\delta_{2}>0$ 使得如果
(3.20) $\begin{equation}m^{\gamma+3 \kappa+3} M^{2\gamma+2\kappa+3}\left(N_{\psi}+\delta^{2\eta_{1}}\right) \leq \delta_{2},\end{equation}$
则对任意的 $0 \leq t \leq T$ 有
(3.21) $\begin{equation}\int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau \lesssim m^{\kappa+1} M^{2 \gamma+2}\left[\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\delta^{\frac{1}{2}}\right]\end{equation}$
(3.22) $\begin{equation}\left\|\left(\phi, \psi, \psi_{x}, \sqrt{\Phi}, M^{-\frac{\gamma+1}{2}} \phi_{x}\right)(t)\right\|^{2}+\int_{0}^{t} \int_{\mathbb{R} }\left[\psi_{x}^{2}+\frac{\psi_{x x}^{2}}{v^{\kappa+1}}\right] {\rm d} x {\rm d} \tau \lesssim \left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\delta^{\frac{1}{2}}.\end{equation}$
证 将$(2.13)_2$ 式与 $\phi_{x}$ 相乘有
(3.23) $\begin{matrix}\left(\phi_{x} \psi\right)_{t}+\psi_{x}^{2}-\left(\psi \phi_{t}\right)_{x}-\frac{\mu(v) \psi_{x x} \phi_{x}}{v}-g \phi_{x} &=&-(p(v)-p(V)) \phi_{x}+\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \phi_{x} \\&\geq &-\int_{0}^{1} p^{\prime}\left(V+\theta \phi_{x}\right) {\rm d} \theta \phi_{x}^{2} \gtrsim M^{-\gamma-1} \phi_{x}^{2}.\end{matrix}$
$\left(\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right) U_{x} \phi_{x}=-\frac{(\kappa+1) \int_{0}^{1}(\theta v+(1-\theta) V)^{\kappa} {\rm d} \theta \phi_{x}^{2}}{v^{\kappa+1} V^{\kappa+1}} U_{x} \geq 0.$
在$[t] \times \mathbb{R} $ 上对(3.23)式进行积分, 再利用Holder's 不等式便有
$\begin{eqnarray*}M^{-\gamma-1} \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau &\lesssim&\left\|\phi_{x}(t)\right\|\|\psi(t)\|+\left\|\phi_{0 x}\right\|\left\|\psi_{0}\right\|+\int_{0}^{t}\left\|\psi_{x}(\tau)\right\|^{2} {\rm d} \tau\\&&+m^{\kappa+1} M^{\gamma+1} \int_{0}^{t} \int_{\mathbb{R} } \frac{\psi_{x x}^{2}}{v} {\rm d} x {\rm d} \tau+\sup _{0 \leq \tau \leq t}\left\|\phi_{x}(\tau)\right\| \int_{0}^{t}\|g(\tau)\| {\rm d} \tau.\end{eqnarray*}$
(3.24) $\begin{matrix}\int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau &\lesssim & m^{\kappa+1} M^{2 \gamma+2}\left[\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|^{2}+\delta^{\frac{1}{2}}\right] \\&&+m^{\gamma+3 \kappa+3} M^{2(\gamma+\kappa+1)}\left(N_{\psi}+\delta^{2}\right) \int_{0}^{t}\left\|\phi_{x}(\tau)\right\|^{2} {\rm d} \tau.\end{matrix}$
$C_{2}\left(m, M, \delta, N_{\psi}\right) m^{\kappa+1} M^{2 \gamma+2} \lesssim m^{\gamma+3 \kappa+3} M^{2(\gamma+\kappa+1)}\left(N_{\psi}+\delta^{2}\right),$
为了推导出 $v$ 有正的上下界, 我们令 $\tilde{v}:=v / V$ , 见文献[31 ],在下面的引理中我们得到了$\tilde{v}$ 相关的估计.
引理 3.5 在定理1.1的假设下, 当 $\delta$ 足够小使得(3.20)式成立, 则可以得到
(3.25) $\begin{equation}\left\|\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(t)\right\|^{2}+\int_{0}^{t} \int_{\mathbb{R} } \frac{\phi_{x x}^{2}}{v^{\gamma+\kappa+2}} {\rm d}x {\rm d}\tau \lesssim\left\|\left(\phi_{0}, \phi_{0}, \sqrt{\Phi_{0}}, \phi_{0 x}, \mu\left(v_{0}\right) \frac{\tilde{v}_{0 x}}{\tilde{v}_{0}}\right)\right\|^{2}+\delta^{\frac{1}{2}}.\end{equation}$
(3.26) $\begin{equation}\left[\mu(v) \frac{u_{x}}{v}-\mu(V) \frac{U_{x}}{V}\right]_{x}=\left(\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right)_{t}+\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t},\end{equation}$
将$(2.13)_2$ 式两边关于 $x$ 求偏导, 可以得到
(3.27) $\begin{equation}\left[\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}-\psi_{x}\right]_{t}-(p(v)-p(V))_{x}=-\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}-g_{x}.\end{equation}$
在(3.27)式两边同时乘以 $\mu(v) \tilde{v}_{x} / \tilde{v}$ 整理得到
$\begin{eqnarray*}&&\left[\frac{1}{2}\left(\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right)^{2}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}} \psi_{x}\right]_{t}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(p(v)-p(V))_{x} \\&=&-\left(\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right)_{t} \psi_{x}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}} g_{x},\end{eqnarray*}$
$-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(p(v)-p(V))_{x}=-\frac{p^{\prime}(v) \mu(v) V^{2}}{v} \tilde{v}_{x}^{2}-\mu(v) \frac{V_{x} \tilde{v}_{x}}{v}\left(p^{\prime}(v) v-p^{\prime}(V) V\right),$
(3.28) $\begin{matrix}&&\left[\frac{1}{2}\left(\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right)^{2}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}} \psi_{x}\right]_{t}-\frac{p^{\prime}(v) \mu(v) V^{2}}{v} \tilde{v}_{x}^{2}-\mu(v) \frac{V_{x} \tilde{v}_{x}}{v}\left(p^{\prime}(v) v-p^{\prime}(V) V\right) \\&=&-\left[\psi_{x}\left(\mu(v) \frac{u_{x}}{v}-\mu(V) \frac{U_{x}}{V}\right)\right]_{x}+\mu(v) \frac{\psi_{x x}^{2}}{v}+\psi_{x x}\left[\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right] U_{x} \\&&+\left[\psi_{x}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right]\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}-\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}} g_{x}.\end{matrix}$
(3.29) $\begin{matrix}\mu(v) \frac{V_{x} \tilde{v}_{x}}{v}\left(p^{\prime}(v) v-p^{\prime}(V) V\right) & \leq &\epsilon \frac{\left|p^{\prime}(v)\right| \mu(v)}{v} \tilde{v}_{x}^{2}+C(\epsilon) \frac{V_{x}^{2} \mu(v)}{v\left|p^{\prime}(v)\right|}\left|p^{\prime}(v) v-p^{\prime}(V) V\right|^{2} \\& \leq& \epsilon \frac{\left|p^{\prime}(v)\right| \mu(v)}{v} \tilde{v}_{x}^{2}+C(\epsilon) V_{x}^{2} M^{\gamma-\kappa} m^{2 \gamma+2} \phi_{x}^{2}.\end{matrix}$
(3.30) $\begin{equation}\psi_{x x}\left[\frac{\mu(v)}{v}-\frac{\mu(V)}{V}\right] U_{x} \lesssim\left|\psi_{x x} U_{x}\right| \frac{M^{\kappa}\left|\phi_{x}\right|}{v^{\kappa+1}} \lesssim \frac{\psi_{x x}^{2}}{v^{\kappa+1}}+\delta^{4} M^{2 \kappa} m^{\kappa+1} \phi_{x}^{2}.\end{equation}$
(3.31) $\begin{matrix}\left|\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}\right| &=&\left|(\mu(v)-\mu(V))\left(\frac{U_{x}}{V}\right)_{x}+\left(\mu^{\prime}(v) \psi_{x x}+\left(\mu^{\prime}(v)-\mu^{\prime}(V)\right) U_{x}\right) \frac{V_{x}}{V}\right| \\&\lesssim & \delta^{2} m^{\kappa+1}\left|\phi_{x}\right|+\delta^{2} \frac{\left|\psi_{x x}\right|}{v^{\kappa+1}}+\delta^{4} m^{\kappa+2}\left|\phi_{x}\right|.\end{matrix}$
(3.32) $\begin{equation}\left|\psi_{x}\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}\right| \lesssim \psi_{x}^{2}+\delta^{4}\left[m^{2 \kappa+4} \phi_{x}^{2}+m^{\kappa+1} \frac{\psi_{x x}^{2}}{v^{\kappa+1}}\right]\end{equation}$
(3.33) $\begin{equation}\left|\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\left[(\mu(v)-\mu(V)) \frac{V_{x}}{V}\right]_{t}\right| \lesssim \epsilon \frac{\left|p^{\prime}(v)\right| \mu(v)}{v} \tilde{v}_{x}^{2}+C(\epsilon) M^{\gamma-\kappa} \delta^{4}\left[m^{2 \kappa+4} \phi_{x}^{2}+m^{\kappa+1} \frac{\psi_{x x}^{2}}{v^{\kappa+1}}\right].\end{equation}$
将(3.28)式关于 $t, x$ 在 $[t] \times \mathbb{R} $ 上积分, 结合 (3.29)-(3.50),(3.21)-(3.22) 和 (3.20)式, 利用Gronwall不 等式以及(2.3)式, 我们可以得到(3.25)式, 引理得证.
有了上面的准备工作, 现在我们来推导 $v(t, x)$ 的一致上下界估计.
引理 3.6 在定理1.1 的假设下, 对任意的 $(t, x) \in[T] \times \mathbb{R} $ , 我们有
(3.34) $\begin{equation}C^{-1} \delta^{-2\tilde{\eta_2}} \le v(t,x) \le C \delta^{-\frac{l(\kappa+1)+\beta}{2}}.\end{equation}$
证 我们首先利用Kanel'的方法来推导比容的一致上下界估计, $\Phi(v, V)$ 可以重新写为
$\Phi(v, V)=V^{-\gamma+1} \tilde{\Phi}(\tilde{v}), uad \tilde{\Phi}(\tilde{v})=\tilde{v}-1+\frac{1}{\gamma-1}\left(\tilde{v}^{-\gamma+1}-1\right).$
$\tilde{\Phi}(z) \sim \left\{\begin{array}{ll}z, & z \rightarrow+\infty,\\z^{-\gamma+1}, & z \rightarrow 0^{+}.\end{array}\right.$
为了使用Kanel'的方法[19 ], 我们构造新的函数$\Psi(\tilde{v})$
$\Psi(\tilde{v}):=\int_{1}^{\tilde{v}} \sqrt{\tilde{\Phi}(z)} \frac{\mu(z)}{z} {\rm d} z.$
$\Psi(\tilde{v}) \sim \left\{\begin{array}{ll} \tilde{v}^{\frac{1}{2}-\kappa},& \tilde{v} \rightarrow+\infty,\\ \tilde{v}^{\frac{1-\gamma}{2}-\kappa},& \tilde{v} \rightarrow 0^{+}.\end{array}\right.$
(3.35) $\begin{equation}|\Psi(\tilde{v})| \gtrsim \tilde{v}^{\frac{1}{2}-\kappa}+\tilde{v}^{\frac{1-\gamma}{2}-\kappa}-C.\end{equation}$
(3.36) $\begin{matrix}|\Psi(\tilde{v}(t, x))|&=&\left|\int_{-\infty}^{x} \frac{\partial \Psi(\tilde{v})}{\partial y}(t, y) {\rm d} y\right| \leq \int_{\mathbb{R} } \sqrt{\tilde{\Phi}(\tilde{v})}\left|\mu(\tilde{v}) \frac{\tilde{v}_{x}}{\tilde{v}}\right|(t, x) {\rm d} x \\& \leq&\|\sqrt{\tilde{\Phi}(\tilde{v})(t)}\|\left\|\mu(\tilde{v}) \frac{\tilde{v}_{x}}{\tilde{v}}(t)\right\|.\end{matrix}$
注意到$\min \left\{v_{-}, \bar{v}, v_{+}\right\} \leq V(t, x) \leq \max \left\{v_{-}, \bar{v}, v_{+}\right\}$ , $v_{\pm}, u_{\pm}$ 是与$\delta$ 无关的, 我们可以从(3.22), (3.25), (3.35)和(3.36)式推导出
$B_{0}^{\frac{2}{1-\gamma-2 \kappa}} \lesssim v(t, x) \lesssim B_{0}^{\frac{2}{1-2 \kappa}}.$
$B_{0}=\left[\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\| +\delta^{\frac{1}{4}}\right]\left[\left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}, \mu\left(v_{0}\right) \frac{\tilde{v}_{0 x}}{\tilde{v}_{0}}\right)\right\|+\delta^{\frac{1}{4}}\right].$
下面我们对比容的上下界做更精细的估计. 先估计$v(t,x)$ 的下界, 由(3.9)和(3.21)式知
(3.37) $\begin{equation}\Vert\sqrt{\Phi}\Vert\lesssim\delta^{\min\left\{\alpha-\frac{(\kappa+1)l}{2},\frac{1}{4}\right\}}=\delta^{\eta_1}.\end{equation}$
(3.38) $\begin{equation}\left\|\left(\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}\right)(t)\right\|^{2}\lesssim\left\|\left(\phi_{0}, \phi_{0}, \sqrt{\Phi_{0}}, \phi_{0 x}, \mu\left(v_{0}\right) \frac{\tilde{v}_{0 x}}{\tilde{v}_{0}}\right)\right\|^{2}+\delta^{\frac{1}{2}}\lesssim \delta^{-2l(\kappa+1)-2\beta}.\end{equation}$
(3.39) $\begin{equation}\left(\frac{v}{V}\right)^{\frac{1-\gamma-2\kappa}{2}}\lesssim 1+\delta^{\eta_1-l(\kappa+1)-\beta},\end{equation}$
(3.40) $\begin{equation}\left(\frac{V}{v}\right)^{\frac{2\kappa+\gamma-1}{2}}\lesssim 1+\delta^{\eta_1-l(\kappa+1)-\beta}.\end{equation}$
(3.41) $\begin{equation}\frac{1}{v}\lesssim 1+\delta^\frac{\eta_1-l(\kappa+1)-\beta}{2\kappa+\gamma-1}=1+\delta^{2\eta_2}.\end{equation}$
若$\eta_2 \ge 0$ , 即$\min\left\{\alpha-\frac{(\kappa+1)l}{2},\frac{1}{4}\right\}\ge l(\kappa+1)+\beta$ 时, 有$\frac{1}{v}\lesssim 1$ . 若$\eta_2 < 0$ , 即$\min\left\{\alpha-\frac{(\kappa+1)l}{2},\frac{1}{4}\right\}< l(\kappa+1)+\beta$ 时, 有$\frac{1}{v}\lesssim 1+\delta^{2\eta_2}$ , 因此
(3.42) $\begin{equation}\delta^{-2\tilde{\eta_2}} \lesssim v(t,x).\end{equation}$
下面估计$v(t,x)$ 的上界. 因为$v=V+\phi_x$ , 所以$\|v\|_{L^{\infty}}\lesssim 1+\|\phi_x\|^{\frac{1}{2}}\|\phi_{xx}\|^{\frac{1}{2}}$ , 由(3.9)和(3.21)式知
$\left\|M^{-\frac{2}{\gamma+1}}\phi_x\right\|^2 \lesssim \left\|\left(\phi_0,\psi_0,\sqrt{\Phi_0},\phi_{0x}\right)\right\|^2+\delta^\frac{1}{2}.$
(3.43) $\begin{equation}\left\|M^{-\frac{2}{\gamma+1}}\phi_x\right\| \lesssim \left\|\left(\phi_0,\psi_0,\sqrt{\Phi_0},\phi_{0x}\right)\right\|+\delta^\frac{1}{4}\lesssim \delta^{\min\{\alpha-\frac{(\kappa+1)l}{2},\frac{1}{4}\}}=\delta^{\eta_1}.\end{equation}$
(3.44) $\begin{equation}\|\phi_x\| \lesssim \delta^{\eta_1} M^{\frac{\gamma+1}{2}}.\end{equation}$
(3.45) $\begin{matrix}\mu(v)\frac{\tilde{v}_{x}}{\tilde{v}}&=&\mu(v)\frac{(\frac{v}{V})_x}{\frac{v}{V}}=\mu(v)\frac{(\frac{v_xV-V_xv}{V^2})}{\frac{v}{V}}\\&=&\mu(v)\frac{v_xV-V_xv}{vV}=\mu(v)(\frac{V_x}{v}-\frac{v_x}{V})\\&=&\mu(v)(\frac{V_x+\phi_{xx}}{v}-\frac{V_x}{V})\\&=&\frac{\mu(v)\phi_{xx}}{v}+\mu(v)(\frac{V_x}{v}-\frac{V_x}{V})\\&=&\frac{\mu(v)\phi_{xx}}{v}+\mu(v)V_x(\frac{1}{v}-\frac{1}{V})\\&=&\frac{\mu(v)\phi_{xx}}{v}+\mu(v)V_x \frac{V-v}{vV}\\&=&\frac{\mu(v)\phi_{xx}}{v}- \frac{\mu(v)V_x \phi_x}{vV},\end{matrix}$
(3.46) $\begin{equation}\frac{\mu(v)\phi_{xx}}{v}=\mu(v)\frac{\tilde{v}_{x}}{\tilde{v}}+\frac{\mu(v)V_x \phi_x}{vV}.\end{equation}$
(3.47) $\begin{equation}\phi_{xx}=v^{1+\kappa}\frac{\tilde{v}_{x}}{\tilde{v}}+\frac{V_x \phi_x}{V}.\end{equation}$
(3.48) $\begin{eqnarray}\|\phi_{xx}\|&\le& \|v^{1+\kappa}\frac{\tilde{v}_{x}}{\tilde{v}}\|+\|\frac{V_x \phi_x}{V}\| \\&\lesssim &\|v\|^{1+\kappa}_{L^\infty}\|\mu(v)\frac{\tilde{v}_{x}}{\tilde{v}}\|+\|V_x\|_{L^\infty}\|\frac{1}{V}\|_{L^\infty}\|\phi_x\| \\&\lesssim &M^{1+\kappa}\|\mu(v)\frac{\tilde{v}_{x}}{\tilde{v}}\|+\delta^2\|\phi_x\|\\&\lesssim& M^{1+\kappa}\delta^{-l(\kappa+1)-\beta}+\delta^{2+\eta_1}M^{\frac{1+\gamma}{2}}.\end{eqnarray}$
(3.49) $\begin{matrix}\|v\|_{L^\infty}&\lesssim &1+\delta^{\frac{\eta_1}{2}}M^{\frac{\gamma+1}{4}}(M^{\frac{\kappa+1}{2}}\delta^{-\frac{l(\kappa+1)+\beta}{2}}+\delta^{\frac{2+\eta_1}{2}}M^{\frac{\gamma+1}{4}})\&\lesssim &1+\delta^{\frac{\eta_1-l(\kappa+1)-\beta}{2}}M^{\frac{\gamma}{2}+\frac{\kappa}{2}+\frac{3}{4}}\\&=&1+\delta^{-\frac{l(\kappa+1)+\beta}{2}}\delta^{\frac{\eta_1}{2}}M^{\frac{\gamma}{2}+\frac{\kappa}{2}+\frac{3}{4}}\\&\lesssim& 1+\delta^{-\frac{l(\kappa+1)+\beta}{2}}.\end{matrix}$
对于$(\phi, \psi)$ 二阶导数的估计. 因为
$\frac{\tilde{v}_{x}}{\tilde{v}}=\frac{\phi_{x x}}{v}-\frac{V_{x} \phi_{x}}{V v},$
$\left\|\frac{\phi_{x x}}{v}\right\|^{2} \leq C\left(\left\|\frac{\tilde{v}_{x}}{\tilde{v}}\right\|^{2}+\left\|\frac{v_{x} \phi_{x}}{v V}\right\|^{2}\right) \leq C(\delta)\left(\left\|\frac{\tilde{v}_{x}}{\tilde{v}}\right\|^{2}+\left\|\phi_{x}\right\|^{2}\right). $
$\int_{0}^{t}\left\|\frac{\phi_{x x}}{v}\right\|^{2} {\rm d}\tau \leq C(\delta) \int_{0}^{t}\left\|\frac{\tilde{v}_{x}}{\tilde{v}}\right\|^{2}+\left\|\phi_{x}\right\|^{2} {\rm d}\tau.$
(3.50) $\begin{equation}\left\|\phi_{x x}(t)\right\|^{2}+\int_{0}^{t}\left\|\phi_{x x}(\tau)\right\|^{2} {\rm d} \tau \leq C\left(\delta,\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}\right).\end{equation}$
对于 $\left\|\psi_{x x}(t)\right\|$ 的估计, 我们用$\partial_{x}(2.13)_{2}$ 式 乘以$\psi_{x x x}$ , 将得到的恒等式在$[t] \times \mathbb{R} $ 上进行积分, 并使用Sobolev不等式, Young不等式以及 Gronwall不等式, 可以发现
(3.51) $\begin{equation}\left\|\psi_{x x}(t)\right\|^{2}+\int_{0}^{t}\left\|\psi_{x x x}(\tau)\right\|^{2} {\rm d} \tau \leq C\left(\delta,\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}\right).\end{equation}$
由(3.50), (3.51)式以及引理3.1-引理3.5可以得到下面的引理.
引理 3.7 如果 $\delta$ 很小使得(3.20)式成立, 那么对于任意 $0 \leq t \leq T$ 都有
(3.52) $\begin{equation}\|(\phi, \psi)(t)\|_{2}^{2}+\int_{0}^{t}\left(\left\|\phi_{x}(\tau)\right\|_{1}^{2}+\left\|\psi_{x}(\tau)\right\|_{2}^{2}\right) {\rm d} \tau \leq C\left(\delta,\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}\right).\end{equation}$
接下来我们证明定理1.1. 因为 $\Phi_{0}(x) \lesssim\left(|V(0, x)|^{-\gamma-1}+|v(0, x)|^{-\gamma-1}\right) \phi_{0 x}^{2}$ , 从(2.1)式和假设(H$_{0}$ ) , (H$_{3})$ , 我们可以得到
$\left\|\sqrt{\Phi_{0}}\right\| \lesssim\left(1+\delta^{-(\gamma+1) \ell / 2}\right)\left\|\phi_{0 x}\right\|, uad\left\|\mu\left(v_{0}\right) \frac{\tilde{v}_{0 x}}{\tilde{v}_{0}}\right\| \lesssim \delta^{-\ell(\kappa+1)}\left(\left\|\phi_{0 x x}\right\|+\delta^{2}\left\|\phi_{0 x}\right\|\right).$
因此, 如果(1.22)和(1.23)式成立, 那么当 $0<\delta<1$ 时,有
(3.53) $\begin{equation}\begin{array}{ll} \left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}\right)\right\|+\delta^{\frac{1}{4}} \lesssim \delta^{\min\left\{\alpha-\frac{\gamma+1}{2} \ell, \frac{1}{4}\right\} },\\[3mm] \left\|\left(\phi_{0}, \psi_{0}, \sqrt{\Phi_{0}}, \psi_{0 x}, \mu\left(v_{0}\right) \frac{\tilde{v}_{0 x}}{\tilde{v}_{0}}\right)\right\|+\delta^{\frac{1}{4}} \lesssim \delta^{-\ell(\kappa+1)-\beta}.\end{array}\end{equation}$
根据引理3.1, 存在一个只依赖于$\delta$ 和 $\left\|\left(\phi_{0}, \psi_{0}\right)\right\|_{2}$ 的正常数$t_{1}$ , 使得柯西问题(2.13)有唯一的解 $(\phi(t, x), \psi(t, x)) \in$ $X_{m_{0}, M_{0}}\left(0, t_{1}\right)$ 且对每一个$0 \leq t \leq t_{1}$ ,$(\phi(t, x), \psi(t, x))$ 都满足(3.1)式, 其中 $m_{0}=2^{-1}C_1^{-1}\delta^{\ell}$ , $M_{0}=2C_1(1+\delta^{-\ell})$ . 这样我们利用(1.22)式和Sobolev不等式可知
$N_{\psi}\left(t_{1}\right)=\sup _{\left[t_{1}\right]}\|\psi(t)\|_{L^{\infty}(\mathbb{R} )} \leq \sup _{\left[t_{1}\right]}\|\psi(t)\|^{\frac{1}{2}}\left\|\psi_{x}(t)\right\|^{\frac{1}{2}} \leq 2 C_{2} \delta^{\alpha}.$
$m_{0}^{-\gamma-3 \kappa-3} M_{0}^{2\gamma+2\kappa+3}\left(N_{\psi}\left(t_{1}\right)+\delta^{2\eta_1}\right) \lesssim 1.$
所以如果$(1.23)_1$ 成立, 我们可以选取一个足够小的常数 $\delta_{1}<1$ , 使得当$0<\delta \leq \delta_{1}$ 时,引理3.1-引理3.7在$T=t_{1}, m=m_{0}^{-1}$ , $M=M_{0}$ 时成立. 因此从引理3.6中我们可以推导出对于每一个 $0 \leq t \leq t_{1}$ ,有
(3.54) $\begin{equation}C_4^{-1} \delta^{-2\tilde{\eta_2}} \le v(t,x) \le C_4 \delta^{-\frac{l(\kappa+1)+\beta}{2}},\end{equation}$
从(3.22)式中我们知道对每一个$0 \leq t \leq t_{1}$ 有
(3.55) $\begin{equation}\|\psi(t)\|_{1} \leq C_{5} \delta^{\min \left\{\alpha-\frac{\gamma+1}{2} \ell, \frac{1}{4}\right\}}.\end{equation}$
接下来如果我们以 $\left(\phi\left(t_{1}, x\right), \psi\left(t_{1}, x\right)\right)$ 作为初始值, 我们可以再次利用引理3.1推导出, 上面构造的唯一局部解$(\phi(t, x), \psi(t, x))$ 可以延拓到时间 $\left[t_{1}+t_{2}\right]$ 内, 并满足
$\|\psi(t)\|_{L^{\infty}(\mathbb{R} )} \leq\|\psi(t)\|_{1} \leq 2\left\|\psi\left(t_{1}\right)\right\|_{1} \leq 2 C_{5} \delta^{\min \left\{\alpha-\frac{\gamma+1}{2} \ell, \frac{1}{4}\right\}}$
(3.56) $\begin{equation}2^{-1}C_4^{-1} \delta^{-2\tilde{\eta_2}} \le v(t,x) \le 2C_4 \delta^{-\frac{l(\kappa+1)+\beta}{2}}.\end{equation}$
因此, 对任意$t_{1} \leq t \leq t_{1}+t_{2}$ , 有
$N_{\psi}\left(t_{1}+t_{2}\right) \leq \max \left\{N_{\psi}\left(t_{1}\right), 2 C_{5} \delta^{\min \left\{\alpha-\frac{\gamma+1}{2} \ell, \frac{1}{4}\right\}}\right\} \leq C_{6} \delta^{\min \left\{\alpha-\frac{\gamma+1}{2} \ell, \frac{1}{4}\right\}}.$
$m_{1}=2^{-1}C_4^{-1} \delta^{-2\tilde{\eta_2}}, M_{1}=2C_4 \delta^{-\frac{l(\kappa+1)+\beta}{2}}.$
易知, 如果参数$\alpha>0, \beta$ 和 $\ell$ 满足 $(1.23)_2$ 式, 那么存在一个足够小的 $\delta_{2}>0$ 使得当 $0<\delta \leq \delta_{2}$ 时, 取 $T=t_{1}+t_{2}, m=m_{1}^{-1}$ , $M=M_{1}$ , 引理3.1-引理3.7仍成立. 如果我们取 $\left(\phi\left(t_{1}+t_{2}, x\right), \psi\left(t_{1}+t_{2}, x\right)\right)$ 作为初始值并且再次使用引理3.1, 那么我们可以将上述的解$(\phi(t, x), \psi(t, x))$ 延拓到$t=t_{1}+2 t_{2}$ . 重复上述步骤, 那么我们可以将 $(\phi(t, x), \psi(t, x))$ 逐步延拓为唯 一的整体解并且对于所有的$t \geq 0$ (3.52), (3.54)和(3.55)式成立. 这样便完成了定理1.1的证明.
参考文献
View Option
[1]
Bresch D , Desjardins B . Existence of global weak solutions for a 2D viscous shallow water equations and convergence to the quasi-geostrophic model
Commun Math Phys , 2013 , 238 (1/2 ): 211 -223
DOI:10.1007/s00220-003-0859-8
URL
[本文引用: 3]
[2]
Cercignani C , Illner R , Pulvirenti M . The Mathematical Theory of Dilute Gases . New York : Springer-Verlag , 1994
[本文引用: 1]
[3]
Chapman S , Cowling T . The Mathematical Theory of Non-uniform Gases (3rd ed) . London : Cambrige University Press , 1970
[本文引用: 1]
[4]
Chen G Q , Perepelistsa M . Shallow water equations: viscous solutions and inviscid limit
Z Angew Math Phys , 2012 , f63 : 1067 -1084
[本文引用: 3]
[5]
Duan R , Liu H X , Zhao H J . Nonlinear stability of rarefaction waves for the compressible Navier-Stokes equations with large initial perturbation
Trans Amer Math Soc , 2009 , 361 (1 ): 453 -493
DOI:10.1090/S0002-9947-08-04637-0
URL
[6]
Haspot B . Global existence of strong solution for shallow water system with large initial data on the irrotational part
J Differential Equations , 2017 , 262 (10 ): 4931 -4978
DOI:10.1016/j.jde.2017.01.010
URL
[本文引用: 3]
[7]
He L , Tang S J , Wang T . Stability of viscous shock waves for the one-dimesional compressible Navier-Stokes equations with density-dependent viscosity
Acta Math Sci , 2016 , 36B (1 ): 34 -48
[本文引用: 8]
[8]
Hong H . Global stability of viscous contact wave for 1-D compressible Navier-Stokes equations
J Differential Equations , 2012 , 252 (5 ): 3482 -3505
DOI:10.1016/j.jde.2011.11.015
URL
[本文引用: 1]
[9]
Huang B K , Wang L S , Xiao Q H . Global nonlinear stability of rarefaction waves for compressible Navier-Stokes equations with temperature and density dependent transport coefficients
Kinet Relat Models , 2016 , 3 : 469 -514
[本文引用: 1]
[10]
Huang B K , Liao Y K . Global stability of combination of viscous contact wave with rarefaction wave for compressible Navier-Stokes equations with temperature-dependent viscosity
Math Models Methods Appl Sci , 2017 , 27 (12 ): 2321 -2379
DOI:10.1142/S0218202517500464
URL
[本文引用: 1]
We study the nonlinear stability of a composite wave pattern, which is a combination of a viscous contact wave with a rarefaction wave, to the Cauchy problem of one-dimensional compressible Navier–Stokes equations for a viscous and heat conducting ideal polytropic gas with large initial perturbation when the transport coefficients depend on both temperature and density. Our main idea is to use the “smallness mechanism” induced by the structures of the equations under consideration and the smallness of the strengths of the two elementary waves to control the possible growth of the solutions caused by the nonlinearities of the equations, the interactions between the solutions themselves and the wave pattern, and the interactions of waves between different families. The main ingredient in the analysis is to derive the uniform positive lower and upper bounds on the specific volume and the temperature.
[11]
Huang F M , Li J , Matsumura A . Asymptotic stability of combination of viscous contact wave with rarefaction waves for one-dimensional compressible Navier-Stokes system
Arch Ration Mech Anal , 2010 , 197 (1 ): 89 -116
DOI:10.1007/s00205-009-0267-0
URL
[本文引用: 1]
[12]
Huang F M , Matsumura A . Stability of a composite wave of two viscous shock waves for the full compressible Navier-Stokes equation
Comm Math Phys , 2009 , 289 (3 ): 841 -861
DOI:10.1007/s00220-009-0843-z
URL
[本文引用: 1]
[13]
Huang F M , Matsumura A , Xin Z P . Stability of contact discontinuities for the 1-D compressible Navier-Stokes equations
Arch Rational Mech Anal , 2006 , 179 : 55 -77
DOI:10.1007/s00205-005-0380-7
URL
[本文引用: 1]
[14]
Huang F M , Wang T . Stability of superposition of viscous contact wave and rarefaction waves for compressible Navier-Stokes system
Indiana Univ Math J , 2016 , 65 : 1833 -1875
DOI:10.1512/iumj.2016.65.5914
URL
[本文引用: 1]
[15]
Huang F M , Xin Z P , Yang T . Contact discontinuity with general perturbations for gas motions
Adv Math , 2008 , 2019 : 1246 -1297
[本文引用: 1]
[16]
Huang F M , Zhao H J . On the global stability of contact discontinuity for compressible Navier-Stokes equations
Rend Sem Mat Univ Padova , 2003 , 109 : 283 -305
[本文引用: 1]
[17]
Jiu Q S , Wang Y , Xin Z P . Vacuum behaviors around rarefaction waves to 1D compressible Navier-Stokes equations with density-dependent viscosity
SIAM J Math Anal , 2013 , 45 (5 ): 3194 -3228
DOI:10.1137/120879919
URL
[本文引用: 1]
[18]
Jiu Q S , Wang Y , Xin Z P . Stability of rarefaction waves to the 1D compressible Navier-Stokes equations with density-dependent viscosity
Comm Partial Differential Equations , 2011 , 36 (4 ): 602 -634
DOI:10.1080/03605302.2010.516785
URL
[本文引用: 1]
[19]
Kanel' J . A model system of equations for the one-dimensional motion of a gas
Differential Equations , 1968 , 4 : 374 -380
[本文引用: 7]
[20]
Kawashima S , Matsumura A . Asymptotic stability of traveling wave solutions of systems for one-dimensional gas motion
Comm Math Phys , 1985 , 101 (1 ): 97 -127
DOI:10.1007/BF01212358
URL
[本文引用: 1]
[21]
Kawashima S , Matsumura A , Nishida T . On the fliud-dynamical approximation tothe Boltzmann equation at the level of the Navier-Stokes equation
Comm Math Phys , 1979 , 70 : 97 -124
DOI:10.1007/BF01982349
URL
[本文引用: 1]
[22]
Kawashima S , Nakamura T , Nishibata S , Zhu P C . Stationary waves to viscous heat-conductive gases in half-space: existence, stability and convergence rate
Math Models Methods Appl Sci , 2010 , 20 (12 ): 2201 -2235
DOI:10.1142/S0218202510004908
URL
[本文引用: 1]
The main concern of this paper is to study large-time behavior of solutions to an ideal polytropic model of compressible viscous gases in one-dimensional half-space. We consider an outflow problem and obtain a convergence rate of solutions toward a corresponding stationary solution. Here the existence of the stationary solution is proved under a smallness condition on the boundary data with the aid of center manifold theory. We also show the time asymptotic stability of the stationary solution under smallness assumptions on the boundary data and the initial perturbation in the Sobolev space, by employing an energy method. Moreover, the convergence rate of the solution toward the stationary solution is obtained, provided that the initial perturbation belongs to the weighted Sobolev space. The proof is based on deriving a priori estimates by using a time and space weighted energy method.
[23]
Liu S Q , Yang T , Zhao H J . Compressible Navier-Stokes approximation to theBoltzmann equation
J Differential Equations , 2014 , 256 (11 ): 3770 -3816
DOI:10.1016/j.jde.2014.02.020
URL
[本文引用: 1]
[24]
Liu T P , Xin Z P , Yang T . Vacuum states for compressible flow
Discrete Contin Dyn Syst , 1998 , 4 (1 ): 1 -32
DOI:10.3934/dcds.1998.4.1
URL
[26]
Liu T P , Xin Z P . Nonlinear stability of rarefaction waves for compressible Navier-Stokes equations
Comm Math Phys , 1988 , 118 : 451 -465
DOI:10.1007/BF01466726
URL
[本文引用: 1]
[28]
Liu T P , Zeng Y N . Shock waves in conservation laws with physical viscosity
American Mathematical Soc , 2015
[本文引用: 1]
[29]
Mascia C , Zumbrun K . Stability of small-amplitude shock profiles of symmetric hyperbolic-parabolic systems
Comm Pure Appl Math , 2004 , 57 (7 ): 841 -876
DOI:10.1002/(ISSN)1097-0312
URL
[本文引用: 1]
[30]
Matsumura A , Nishihara K . Global asymptotics toward the rarefaction wave for solutions of viscous $p$ - system with boundary effect
Quart Appl Math , 2000 , 58 (1 ): 69 -83
DOI:10.1090/qam/2000-58-01
URL
[31]
Matsumura A , Nishihara K . Global stability of the rarefaction wave of a one-dimensional model system for compressible viscous gas
Comm Math Phys , 1992 , 144 (2 ): 325 -335
DOI:10.1007/BF02101095
URL
[本文引用: 2]
[32]
Matsumura A , Nishihara K . Asymptotic toward the rarefaction waves of the solutions of a one-dimensional model system for compressible viscous gas
Japan J Appl Math , 1986 , 3 : 1 -13
DOI:10.1007/BF03167088
URL
[本文引用: 1]
[33]
Matsumura A , Nishihara K . On the stability of travelling wave solutions of a one-dimensional model system for compressible viscous gas
Japan J Appl Math , 1985 , 2 (1 ): 17 -25
DOI:10.1007/BF03167036
URL
[本文引用: 1]
[34]
Matsumura A , Wang Y . Asymptotic stability of viscous shock wave for a one-dimensional isentropic model of viscous gas with density dependent viscosity
Methods Appl Anal , 2010 , 17 (4 ): 279 -290
DOI:10.4310/MAA.2010.v17.n3.a3
URL
[本文引用: 1]
[35]
Nishida T , Smoller J . Solutions in the large for some nonlinear hyperbolic conservation laws
Comm Pure Appl Math , 1973 , 26 : 183 -200
DOI:10.1002/(ISSN)1097-0312
URL
[本文引用: 1]
[36]
Nishihara K , Yang T , Zhao H J . Nonlinear stability of strong rarefaction waves for compressible Navier-Stokes equations
SIAM J Math Anal , 2004 , 35 (6 ): 1561 -1593
DOI:10.1137/S003614100342735X
URL
[37]
Qin X H , Wang Y . Stability of wave patterns to the inflow problem of full compressible Navier-Stokes equations
SIAM J Math Anal , 2009 , 41 (5 ): 2057 -2087
DOI:10.1137/09075425X
URL
[本文引用: 1]
[38]
Smoller J . Shock Waves and Reaction-Diffusion Equations. Grundlehren der Mathematischen Wissenschaften 285
New York:Springer-Verlag , 1994
[本文引用: 2]
[39]
Wan L , Wang T , Zou Q Y . Stability of stationary solutions to the outflow problem for full compressible Navier-Stokes equations with large initial perturbation
Nonlinearity , 2016 , 29 (4 ): 1329 -1354
DOI:10.1088/0951-7715/29/4/1329
URL
[本文引用: 1]
[40]
Wan L , Wang T , Zhao H J . Asymptotic stability of wave patterns to compressible viscous and heat-conducting gases in the half space
J Differential Equations , 2016 , 261 (11 ): 5949 -5991
DOI:10.1016/j.jde.2016.08.032
URL
[本文引用: 1]
[41]
Wang T , Zhao H J , Zou Q Y . One-dimensional compressible Navier-Stokes equations with large density oscillation
Kinet Relat Models , 2013 , 6 (3 ): 649 -670
DOI:10.3934/krm.2013.6.649
URL
[本文引用: 1]
Existence of global weak solutions for a 2D viscous shallow water equations and convergence to the quasi-geostrophic model
3
2013
... 该文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 主要研究目的是改进文献[7 ] 的结果至 $\gamma>1, \kappa\geq 0$ . 注意到 $\gamma=2, \kappa=1$ 时, 一维等熵可压缩Navier-Stokes方程组对应于Saint-Venant浅水波方程组, 该方程组描述了地表浅水运动的规律, 在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... This paper mainly studies the large-time asymptotic behavior of the global solution of the density dependent one-dimensional isentropic compressible Navier-Stokes equations Cauchy problem. The main purpose of this paper is to improve the result of [7 ] to $\gamma>1, \kappa \geq 0 $ . It is noted that when $\gamma=2,\kappa=1 $ , the one-dimensional isentropic compressible Navier-Stokes equations correspond to the Saint-Venant shallow water wave equations, which describe the law of surface shallow water movement and have important applications in physics and oceanography [1 ,4 ,6 ] . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... 需要指出的是, 当 $\gamma=2, \kappa=1$ 时, (1.1)-(1.3)式对应于 Saint-Venant 浅水波方程组, 该方程组在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . (1.3)式中的结果很显然排除了这一重要的情形. 本学位论文的主要研究目的就是扩大 $\kappa$ 的取值范围, 改进(1.3)式中的结果至 $\kappa\geq 0$ . ...
1
1994
... 根据稀薄气体的动理学理论,当Boltzmann方程通过Chapman-Enskog展开推导出可压缩Navier-Stokes方程组时, 粘性系数$\mu$ 和热传导系数$\kappa$ 是依赖于温度的函数[2 ,3 ,21 ,23 ] , 且满足 ...
1
1970
... 根据稀薄气体的动理学理论,当Boltzmann方程通过Chapman-Enskog展开推导出可压缩Navier-Stokes方程组时, 粘性系数$\mu$ 和热传导系数$\kappa$ 是依赖于温度的函数[2 ,3 ,21 ,23 ] , 且满足 ...
Shallow water equations: viscous solutions and inviscid limit
3
2012
... 该文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 主要研究目的是改进文献[7 ] 的结果至 $\gamma>1, \kappa\geq 0$ . 注意到 $\gamma=2, \kappa=1$ 时, 一维等熵可压缩Navier-Stokes方程组对应于Saint-Venant浅水波方程组, 该方程组描述了地表浅水运动的规律, 在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... This paper mainly studies the large-time asymptotic behavior of the global solution of the density dependent one-dimensional isentropic compressible Navier-Stokes equations Cauchy problem. The main purpose of this paper is to improve the result of [7 ] to $\gamma>1, \kappa \geq 0 $ . It is noted that when $\gamma=2,\kappa=1 $ , the one-dimensional isentropic compressible Navier-Stokes equations correspond to the Saint-Venant shallow water wave equations, which describe the law of surface shallow water movement and have important applications in physics and oceanography [1 ,4 ,6 ] . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... 需要指出的是, 当 $\gamma=2, \kappa=1$ 时, (1.1)-(1.3)式对应于 Saint-Venant 浅水波方程组, 该方程组在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . (1.3)式中的结果很显然排除了这一重要的情形. 本学位论文的主要研究目的就是扩大 $\kappa$ 的取值范围, 改进(1.3)式中的结果至 $\kappa\geq 0$ . ...
Nonlinear stability of rarefaction waves for the compressible Navier-Stokes equations with large initial perturbation
0
2009
Global existence of strong solution for shallow water system with large initial data on the irrotational part
3
2017
... 该文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 主要研究目的是改进文献[7 ] 的结果至 $\gamma>1, \kappa\geq 0$ . 注意到 $\gamma=2, \kappa=1$ 时, 一维等熵可压缩Navier-Stokes方程组对应于Saint-Venant浅水波方程组, 该方程组描述了地表浅水运动的规律, 在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... This paper mainly studies the large-time asymptotic behavior of the global solution of the density dependent one-dimensional isentropic compressible Navier-Stokes equations Cauchy problem. The main purpose of this paper is to improve the result of [7 ] to $\gamma>1, \kappa \geq 0 $ . It is noted that when $\gamma=2,\kappa=1 $ , the one-dimensional isentropic compressible Navier-Stokes equations correspond to the Saint-Venant shallow water wave equations, which describe the law of surface shallow water movement and have important applications in physics and oceanography [1 ,4 ,6 ] . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... 需要指出的是, 当 $\gamma=2, \kappa=1$ 时, (1.1)-(1.3)式对应于 Saint-Venant 浅水波方程组, 该方程组在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . (1.3)式中的结果很显然排除了这一重要的情形. 本学位论文的主要研究目的就是扩大 $\kappa$ 的取值范围, 改进(1.3)式中的结果至 $\kappa\geq 0$ . ...
Stability of viscous shock waves for the one-dimesional compressible Navier-Stokes equations with density-dependent viscosity
8
2016
... 该文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 主要研究目的是改进文献[7 ] 的结果至 $\gamma>1, \kappa\geq 0$ . 注意到 $\gamma=2, \kappa=1$ 时, 一维等熵可压缩Navier-Stokes方程组对应于Saint-Venant浅水波方程组, 该方程组描述了地表浅水运动的规律, 在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... This paper mainly studies the large-time asymptotic behavior of the global solution of the density dependent one-dimensional isentropic compressible Navier-Stokes equations Cauchy problem. The main purpose of this paper is to improve the result of [7 ] to $\gamma>1, \kappa \geq 0 $ . It is noted that when $\gamma=2,\kappa=1 $ , the one-dimensional isentropic compressible Navier-Stokes equations correspond to the Saint-Venant shallow water wave equations, which describe the law of surface shallow water movement and have important applications in physics and oceanography [1 ,4 ,6 ] . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... 当粘性系数依赖于密度(即满足(1.3)式)时, 与之对应也有许多相关结果. 文献[17 ,18 ]证明了当初始值具有大扰动且含真空情形时, 一维可压缩 Navier-Stokes 方程组稀疏波的非线性稳定性. 文献[34 ] 研究了在小初始扰动下, 一维可压缩 Navier-Stokes 方程组粘性激波的非线性稳定性(文中要求 $\kappa \geq\frac{\gamma-1}{2}$ ). 最近, 文献[7 ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ . ...
... ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ . ...
... 注 1.2 注意到文献[7 ] 的主要结果中 $\alpha\leq\beta$ ,这使得比容的初始值具有大振幅. 而在我们的结果中, 由(1.23)式 知 $\alpha>\beta$ , 这不能保证比容的初始值具有大振幅. 事实上, 由(1.22)式知 ...
... 下面将介绍本文的一些主要想法. 研究方程组(1.1)-(1.2)粘性激波的非线性稳定性, 关键在于推导比容与时间 $t$ 无关的一致 上下界估计. 为此, 文献[7 ] 通过 $L^{2}$ 能量方法得出 $\left\|\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(t)\right\|_{L^{2}(\mathbb{R} )}$ ($\widetilde{v}=\frac{v}{V}$ ) 的与时间 $t$ 无关的先验估计.再通过 Kanel' 的方法[19 ] 得出比容 v 的一致上下界估计. 注意到该方法在推导比容上下界时, 我们要求 $0\leq\kappa<\frac{1}{2}$ . 为了扩大 $\kappa$ 的范围, 我们的主要想法如下: 首先可利用 Kanel'的方法[19 ] 推导比容的一致下界估计. 在推导比容一致上界估计时, 我们先作关于 $\|\phi_{x}\|_{L^{2}(\mathbb{R} )}$ 和 $\|\phi_{xx}\|_{L^{2}(\mathbb{R} )}$ 的精细估计. 再利用 Sobolev 不等式 ...
Global stability of viscous contact wave for 1-D compressible Navier-Stokes equations
1
2012
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Global nonlinear stability of rarefaction waves for compressible Navier-Stokes equations with temperature and density dependent transport coefficients
1
2016
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Global stability of combination of viscous contact wave with rarefaction wave for compressible Navier-Stokes equations with temperature-dependent viscosity
1
2017
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Asymptotic stability of combination of viscous contact wave with rarefaction waves for one-dimensional compressible Navier-Stokes system
1
2010
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Stability of a composite wave of two viscous shock waves for the full compressible Navier-Stokes equation
1
2009
... 我们首先陈述粘性激波$(V_i,U_i)(t,x)\ (i=1,2)$ 的存在性以及当$x-s_it \to \pm \infty$ 时的衰减估计. 事实上, 由文献[12 ] 及假设 (H$_3$ ) 可得 ...
Stability of contact discontinuities for the 1-D compressible Navier-Stokes equations
1
2006
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Stability of superposition of viscous contact wave and rarefaction waves for compressible Navier-Stokes system
1
2016
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Contact discontinuity with general perturbations for gas motions
1
2008
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
On the global stability of contact discontinuity for compressible Navier-Stokes equations
1
2003
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Vacuum behaviors around rarefaction waves to 1D compressible Navier-Stokes equations with density-dependent viscosity
1
2013
... 当粘性系数依赖于密度(即满足(1.3)式)时, 与之对应也有许多相关结果. 文献[17 ,18 ]证明了当初始值具有大扰动且含真空情形时, 一维可压缩 Navier-Stokes 方程组稀疏波的非线性稳定性. 文献[34 ] 研究了在小初始扰动下, 一维可压缩 Navier-Stokes 方程组粘性激波的非线性稳定性(文中要求 $\kappa \geq\frac{\gamma-1}{2}$ ). 最近, 文献[7 ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ . ...
Stability of rarefaction waves to the 1D compressible Navier-Stokes equations with density-dependent viscosity
1
2011
... 当粘性系数依赖于密度(即满足(1.3)式)时, 与之对应也有许多相关结果. 文献[17 ,18 ]证明了当初始值具有大扰动且含真空情形时, 一维可压缩 Navier-Stokes 方程组稀疏波的非线性稳定性. 文献[34 ] 研究了在小初始扰动下, 一维可压缩 Navier-Stokes 方程组粘性激波的非线性稳定性(文中要求 $\kappa \geq\frac{\gamma-1}{2}$ ). 最近, 文献[7 ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ . ...
A model system of equations for the one-dimensional motion of a gas
7
1968
... 该文主要研究粘性系数依赖于密度的一维等熵可压缩 Navier-Stokes 方程组 Cauchy 问题整体解的大时间渐近行为, 主要研究目的是改进文献[7 ] 的结果至 $\gamma>1, \kappa\geq 0$ . 注意到 $\gamma=2, \kappa=1$ 时, 一维等熵可压缩Navier-Stokes方程组对应于Saint-Venant浅水波方程组, 该方程组描述了地表浅水运动的规律, 在物理学和海洋学中有重要的应用[1 ,4 ,6 ] . 注意到文献[7 ] 中通过利用 Kanel 的方法[19 ] 来推导比容的一致上下界估计, 在得出比容的上界时, 该方法要求 $\kappa<\frac{1}{2}$ . 对该文所研究的问题而言, 需要首先利用Kanel'的方法[19 ] 来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... [19 ]来推导比容的一致上下界估计. 为了扩大 $\kappa$ 的取值范围, 还需要对比容的上下界作更为精细的能量估计. 在得出比容的一致上下界估计之后, 可通过精心设计的连续性技巧, 将 Navier-Stokes 方程组的局部解一步步延拓为整体解, 并得到对应的大时间渐近行为. ...
... This paper mainly studies the large-time asymptotic behavior of the global solution of the density dependent one-dimensional isentropic compressible Navier-Stokes equations Cauchy problem. The main purpose of this paper is to improve the result of [7 ] to $\gamma>1, \kappa \geq 0 $ . It is noted that when $\gamma=2,\kappa=1 $ , the one-dimensional isentropic compressible Navier-Stokes equations correspond to the Saint-Venant shallow water wave equations, which describe the law of surface shallow water movement and have important applications in physics and oceanography [1 ,4 ,6 ] . Note that in [7 ], the method[19 ] of Kanel is used to derive the uniform upper and lower bound estimation of specific volume. When obtaining the upper bound of specific volume, this method requires $\kappa<\frac{1}{2}$ . For the problem studied in this paper, we need to use Kanel's method[19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... [19 ] to derive the uniform upper and lower bound estimation of specific volume. In order to expand the value range of $\kappa$ , it is also necessary to make a more precise energy estimation of the upper and lower bounds of the specific volume. After obtaining the uniform upper and lower bound estimation of specific volume, the local solution of Navier-Stokes equations can be extended into the global solution step by step through carefully designed continuity techniques, and the corresponding large-time asymptotic behavior can be obtained. ...
... 下面将介绍本文的一些主要想法. 研究方程组(1.1)-(1.2)粘性激波的非线性稳定性, 关键在于推导比容与时间 $t$ 无关的一致 上下界估计. 为此, 文献[7 ] 通过 $L^{2}$ 能量方法得出 $\left\|\mu(v) \frac{\tilde{v}_{x}}{\tilde{v}}(t)\right\|_{L^{2}(\mathbb{R} )}$ ($\widetilde{v}=\frac{v}{V}$ ) 的与时间 $t$ 无关的先验估计.再通过 Kanel' 的方法[19 ] 得出比容 v 的一致上下界估计. 注意到该方法在推导比容上下界时, 我们要求 $0\leq\kappa<\frac{1}{2}$ . 为了扩大 $\kappa$ 的范围, 我们的主要想法如下: 首先可利用 Kanel'的方法[19 ] 推导比容的一致下界估计. 在推导比容一致上界估计时, 我们先作关于 $\|\phi_{x}\|_{L^{2}(\mathbb{R} )}$ 和 $\|\phi_{xx}\|_{L^{2}(\mathbb{R} )}$ 的精细估计. 再利用 Sobolev 不等式 ...
... [19 ]推导比容的一致下界估计. 在推导比容一致上界估计时, 我们先作关于 $\|\phi_{x}\|_{L^{2}(\mathbb{R} )}$ 和 $\|\phi_{xx}\|_{L^{2}(\mathbb{R} )}$ 的精细估计. 再利用 Sobolev 不等式 ...
... 为了使用Kanel'的方法[19 ], 我们构造新的函数$\Psi(\tilde{v})$ ...
Asymptotic stability of traveling wave solutions of systems for one-dimensional gas motion
1
1985
... 我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性. ...
On the fliud-dynamical approximation tothe Boltzmann equation at the level of the Navier-Stokes equation
1
1979
... 根据稀薄气体的动理学理论,当Boltzmann方程通过Chapman-Enskog展开推导出可压缩Navier-Stokes方程组时, 粘性系数$\mu$ 和热传导系数$\kappa$ 是依赖于温度的函数[2 ,3 ,21 ,23 ] , 且满足 ...
Stationary waves to viscous heat-conductive gases in half-space: existence, stability and convergence rate
1
2010
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Compressible Navier-Stokes approximation to theBoltzmann equation
1
2014
... 根据稀薄气体的动理学理论,当Boltzmann方程通过Chapman-Enskog展开推导出可压缩Navier-Stokes方程组时, 粘性系数$\mu$ 和热传导系数$\kappa$ 是依赖于温度的函数[2 ,3 ,21 ,23 ] , 且满足 ...
Vacuum states for compressible flow
0
1998
Shock waves for compressible Navier-Stokes equations are stable
1
1986
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Nonlinear stability of rarefaction waves for compressible Navier-Stokes equations
1
1988
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Pointwise decay to contact discontinuities for systems of viscous conservation laws
1
1997
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Shock waves in conservation laws with physical viscosity
1
2015
... 我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性. ...
Stability of small-amplitude shock profiles of symmetric hyperbolic-parabolic systems
1
2004
... 我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性. ...
Global asymptotics toward the rarefaction wave for solutions of viscous $p$ -system with boundary effect
0
2000
Global stability of the rarefaction wave of a one-dimensional model system for compressible viscous gas
2
1992
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
... 为了推导出 $v$ 有正的上下界, 我们令 $\tilde{v}:=v / V$ , 见文献[31 ],在下面的引理中我们得到了$\tilde{v}$ 相关的估计. ...
Asymptotic toward the rarefaction waves of the solutions of a one-dimensional model system for compressible viscous gas
1
1986
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
On the stability of travelling wave solutions of a one-dimensional model system for compressible viscous gas
1
1985
... 我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性. ...
Asymptotic stability of viscous shock wave for a one-dimensional isentropic model of viscous gas with density dependent viscosity
1
2010
... 当粘性系数依赖于密度(即满足(1.3)式)时, 与之对应也有许多相关结果. 文献[17 ,18 ]证明了当初始值具有大扰动且含真空情形时, 一维可压缩 Navier-Stokes 方程组稀疏波的非线性稳定性. 文献[34 ] 研究了在小初始扰动下, 一维可压缩 Navier-Stokes 方程组粘性激波的非线性稳定性(文中要求 $\kappa \geq\frac{\gamma-1}{2}$ ). 最近, 文献[7 ] 研究了在一类大初始扰动下, (1.1)-(1.3)式粘性激波的非线性稳定性. (1.1)-(1.3)式中初始值需满足“零质量”条件, 且初始密度具有大振幅. 注意到 文献[7 ] 中 $\kappa$ 式的取值范围为$0\leq\kappa<\frac{1}{2}$ . ...
Solutions in the large for some nonlinear hyperbolic conservation laws
1
1973
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Nonlinear stability of strong rarefaction waves for compressible Navier-Stokes equations
0
2004
Stability of wave patterns to the inflow problem of full compressible Navier-Stokes equations
1
2009
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Shock Waves and Reaction-Diffusion Equations. Grundlehren der Mathematischen Wissenschaften 285
2
1994
... $s_{i}\left(v, v_{-}\right)=(-1)^{i} \sqrt{\left(p(v)-p\left(v_{l}\right)\right) /\left(v_{l}-v\right)}$ . 在上述假设下, 根据文献[38 ], 我们可以找到一个唯一的 $(\bar{v}, \bar{u}) \in S_{1}\left(v_{-}, u_{-}\right)$ 使得 $\left(v_{+}, u_{+}\right) \in$ $S_{2}(\bar{v}, \bar{u})$ , 其中 ...
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Stability of stationary solutions to the outflow problem for full compressible Navier-Stokes equations with large initial perturbation
1
2016
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
Asymptotic stability of wave patterns to compressible viscous and heat-conducting gases in the half space
1
2016
... 注 1.3 关于一维可压缩 Navier-Stokes 基本波稳定性的相关结果,感兴趣的读者可参考 文献[8 ⇓ ⇓ -11 ,13 ⇓ ⇓ -16 ,22 ,25 ⇓ -27 ,31 ,32 ,35 ,37 ⇓ ⇓ -40 ]. ...
One-dimensional compressible Navier-Stokes equations with large density oscillation
1
2013
... 我们首先回顾一些已有的相关结果. 当粘性系数 $\mu(v)$ 是一个常数时, 文献[20 ,33 ] 研究了在初始值有小扰动且具有“零质量”条件下, (1.1)-(1.3)式粘性激波的非线性稳定性.若假设初始值不满足零质量扰动,相关的结果可参考文献[28 ,29 ]. 值得注意的是, 上述结果都是研究初始值具有小扰动的情形. 若假设初始值具有大的扰动, 问题变得更加困难和复杂. 对于这种情形, 文献[41 ]研究了在一类大初始扰动下,(1.1)-(1.3)式粘性激波的非线性稳定性. ...