数学物理学报, 2022, 42(4): 1150-1172 doi:

论文

一维具有阻尼和摩擦项的可压缩流体欧拉方程组当压力消失时黎曼解的极限

邵志强,

福州大学数学与统计学院 福州 350108

Concentration and Cavitation in the Pressureless Limit of Euler Equations of Compressible Fluid Flow with Damping and Friction

Shao Zhiqiang,

School of Mathematics and Statistics, Fuzhou University, Fuzhou 350108

收稿日期: 2021-09-24  

基金资助: 福建省自然科学基金.  2019J01642

Received: 2021-09-24  

Fund supported: the NSF of Fujian Province.  2019J01642

作者简介 About authors

邵志强,E-mail:zqshao@fzu.edu.cn , E-mail:zqshao@fzu.edu.cn

Abstract

In this paper, we study the Riemann problem for the Euler equations of compressible fluid flow with a composite source term. The source can cover a Coulomb-like friction or a damping or both. Different from the homogeneous system, Riemann solutions of the inhomogeneous system are non self-similar. Concentration and cavitation in the pressureless limit of solutions to the Riemann problem for the Euler equations of compressible fluid flow with a composite source term are investigated in detail as the adiabatic exponent tends to one. We rigorously proved that, as the adiabatic exponent tends to one, any two-shock Riemann solution tends to a delta shock solution of the pressureless Euler system with a composite source term, and the intermediate density between the two shocks tends to a weighted $ \delta$-mesaure which forms the delta shock; while any two-rarefaction-wave Riemann solution tends to a two-contact-discontinuity solution of the pressureless Euler system with a composite source term, whose intermediate state between the two contact discontinuities is a vacuum state.

Keywords: Pressureless limit ; Euler equations of compressible fluid flow ; Composite source term ; Delta shock wave ; Vacuum state ; Riemann problem

PDF (483KB) 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

邵志强. 一维具有阻尼和摩擦项的可压缩流体欧拉方程组当压力消失时黎曼解的极限. 数学物理学报[J], 2022, 42(4): 1150-1172 doi:

Shao Zhiqiang. Concentration and Cavitation in the Pressureless Limit of Euler Equations of Compressible Fluid Flow with Damping and Friction. Acta Mathematica Scientia[J], 2022, 42(4): 1150-1172 doi:

1 引言

考虑带有复合源项的一维可压缩流体欧拉方程组

$ \begin{align} \left\{\begin{array}{ll} \rho_{t}+(\rho u)_{x}=0, \\ { } u_{t}+(\frac{u^{2}}{2}+p (\rho))_{x}=-\alpha u+\beta, \end{array}\right . \end{align} $

其中, $ \beta $是常数, 表示Coulomb-like摩擦项, $ \alpha> 0 $表示恒耗散系数, $ \rho $表示密度, $ u $表示速度, 以及$ p(\rho) $表示流体压力, 且非线性方程$ p(\rho)=\frac{\theta}{2}\rho^{\gamma-1}, $ 其中$ \theta=\frac{\gamma-1}{2} $, $ \gamma\in(1, 2) $是常数. 绝热指数$ \gamma $在不同的情况下起着重要作用, 例如当$ \gamma=-1 $, $ p(\rho)=-\frac{1}{2\rho^{2}} $称为Chaplygin气体压力, 这是Chaplygin[1], Tsien[2]和von Karman[3]为了计算飞机机翼上升力的数值近似而引入的.

Shen[4]研究具有Coulomb-like摩擦项的零压欧拉方程组, 并通过变量代换

$ \begin{align} v(t, x)=u(t, x)-\beta t \end{align} $

获得其非自相似解. Faccanoni和Mangeney[5]首次通过引入变量代换(1.2)研究具有Coulomb-like摩擦项的浅水方程组的Riemann问题.

如果$ \alpha=0 $$ \beta=0 $, 则系统(1.1)为齐次可压缩流体欧拉方程组[69]

$ \begin{align} \left\{\begin{array}{ll} \rho_{t}+(\rho u)_{x}=0, \\ { } u_{t}+(\frac{u^{2}}{2}+p (\rho))_{x}=0.\end{array}\right . \end{align} $

系统(1.3)是Earnshaw[6]于1858年首次用于研究等熵流体, 具有不同的物理背景. 例如, 它是具有远距离相互作用的牛顿动力学标度极限系统, 用于质量在$ {\Bbb R} $中的连续分布[1011], 同时也是Vlasov方程的流体动力的极限[12].

许多学者已深入研究系统(1.3)的解[79, 1320]. 特别的, DiPerna[14]首次使用Glimm方法对$ 1 <\gamma<3 $情形的Cauchy问题研究全局弱解的存在性定理. Li[15]在DiPerna[14]研究的基础上获得$ -1 < \gamma < 1 $情形的Cauchy问题全局弱解. Lu[7]和Cheng[8]基于补偿紧致性理论和动力学方程的基本思想, 分别研究$ \gamma > 3 $, $ 1 <\gamma < 3 $情形的Cauchy问题的全局熵解的存在性定理. 对压力$ p $为在实轴上取实值的一整函数情形, Sarrico[17]构造性的得到Riemann问题在含间断函数和Dirac测度的广义函数的适当的空间中解. 对压力$ p(\rho)=\int^{\rho}_{0}s e^{s}{\rm d}s $情形, Song and Xiao[18]研究带有源项的系统(1.3)的Cauchy问题全局弱解的存在性定理. 因此, 研究具有源项的可压缩流体流动欧拉方程组的解, 如摩擦、阻尼和松弛效应是很自然的. 这里, 我们集中讨论(1.1)式中提出的源项. 如果$ \alpha=0 $, 那么这种源项, 有时被称为Coulomb-like摩擦项, 是Savage和Hutter[21]早期提出来描述颗粒流动行为的. 目前, 对一些具有Coulomb-like摩擦项的双曲型方程组解的数学研究非常活跃, 有兴趣的读者可以参考文献[45, 2233].

$ \gamma\rightarrow 1 $, 系统(1.1)趋于带有复合源项的一维零压欧拉方程

$ \begin{align} \left\{\begin{array}{ll} \rho_{t}+(\rho u)_{x}=0, \\ { } u_{t}+(\frac{u^{2}}{2})_{x}=-\alpha u+\beta.\end{array}\right . \end{align} $

如果取$ u_{a}=-1 $, 则系统(1.4)涵盖了重要的欧拉液滴模型[3439]. 此外, 如果$ \alpha=0 $$ \beta=-1 $, 那么系统(1.4) 在文献[25]中模拟了具有大弗劳德数浅流中的剧烈不连续性.

对一维等熵气体欧拉方程组

$ \begin{align} \left\{\begin{array}{ll} \rho_{t}+(\rho u)_{x}=0, \\(\rho u)_{t}+ (\rho u^{2}+p(\rho) )_{x}=0, \end{array}\right . \end{align} $

当压力趋于零或常数时, 欧拉方程组(1.5)收敛到零压气体动力学系统. 在早期的开创性论文中, Chen和Liu[40]通过在模型$ p(\rho)=\varepsilon\rho^{\gamma}/\gamma $$ ( \gamma >1) $中取极限$ \varepsilon \rightarrow 0+ $, 首次证明了多方气体欧拉方程组(1.5) 的Riemann解的delta激波和真空状态的形成, 它严格地描述了数学中的集中和空化现象. 此外, 他们还得到了文献[41]中非等熵流体欧拉方程相同的结果。Li[42]研究了欧拉方程(1.5) 对等温情形$ (\gamma = 1) $的相同问题. 最近, 当$ \gamma \rightarrow 0 $, 即压力趋于常数时, Ibrahim等人在文献[43]中证明了集中现象也存在于模型$ p(\rho) =\rho^{\gamma} $$ (0 < \gamma < 1) $. 也就是说, 他们用欧拉方程(1.5)的黎曼解的极限行为严格地证明了delta激波的形成. 对于其他一些物理模型, 也有许多结果[4458].

在文献[4043]的激励下, 本文研究了可压缩流体流动的非齐次欧拉方程组(1.1)的黎曼解当绝热指数趋于1时的极限. 与齐次方程不同, 黎曼解是非自相似的, 当$ \gamma \rightarrow1 $时, 我们发现在$ 1 < \gamma < 2 $的情况下, 也存在相同的质量集中现象和真空现象.

本文的结构如下. 在第2和第3节中, 我们分别介绍了方程(1.4), (1.1)的黎曼解的一些结果. 在第4节中, 我们严格证明了当$ \gamma \rightarrow1 $时, 系统(1.1) 的Riemann解形成delta激波和真空状态. 最后, 结论在第5节进行了论述.

2 预备知识

在本节, 我们给出系统(1.4)的Riemann问题的解. 对于系统(1.4)相应的齐次型系统, Riemann问题的解可参考文献[59, 60].

通过变量代换

$ \begin{align} v(t, x)=\frac{\beta}{\alpha}+(u(t, x)-\frac{\beta}{\alpha})e^{\alpha t}, \end{align} $

系统(1.4) 可改写为以下齐次守恒形式

$ \begin{align} \left\{\begin{array}{ll} { } \rho_{t}+(\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}))_{x}=0, \\ { } v_{t}+(\frac{1}{2}(\frac{\beta}{\alpha}e^{\alpha t}+v-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) )_{x}=0.\end{array}\right . \end{align} $

接下来, 我们将研究具有下列初值的系统(2.2) 的Riemann问题

$ \begin{align} (\rho, v)(0, x) =\left\{\begin{array}{ll} (\rho_{-}, u_{-}), \, \, \, \, x< 0, \\(\rho_{+}, u_{+}), \, \, \, \, x> 0, \end{array} \right. \end{align} $

其中$ \rho_{\pm}>0 $$ u_{\pm} $都是常数.

可以看出, 通过变量代换$ (\rho, u)(t, x) =(\rho(t, x), \frac{\beta}{\alpha}+(v(t, x)-\frac{\beta}{\alpha})e^{-\alpha t})) $, 系统(1.4)的Riemann解可通过直接求解系统(2.2)–(2.3)的Riemann问题而得.

系统(2.2)可以重新表示为一个拟线性形式

$ \begin{align} \left(\begin{array}{cccc}\rho\\v \end{array}\right)_{t}+\left( \begin{array}{ccc} { } \frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}{\quad} & \rho e^{-\alpha t}\\ 0 {\quad} &{ } \frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t} \end{array}\right)\left(\begin{array}{cccc}\rho\\v \end{array}\right)_{x}=\left(\begin{array}{cccc}0\\0 \end{array}\right). \end{align} $

通过(2.4)式, 很容易看出系统(2.2)有两个相同的特征值$ \lambda =\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t} $, 其对应的右特征向量为$ \overrightarrow{r} = (1, 0)^{T}. $ 由于$ \nabla\lambda\cdot \overrightarrow{r}\equiv 0, $ 所以系统(2.2)是完全线性退化的, 相应的基本波都是接触间断.

假设接触间断的波速为$ \sigma(t)=x'(t), $ 则满足下列Rankine-Hugoniot条件

$ \begin{align} \left\{ \begin{array}{ll} { } -\sigma(t)[\rho]+[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})]=0, \\ { } -\sigma(t)[v]+[\frac{1}{2}(\frac{\beta}{\alpha}e^{\alpha t}+v-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) ]=0, \end{array} \right. \end{align} $

其中$ [\rho]=\rho_{r}-\rho_{l} $$ \rho $穿过间断线的跳跃, $ [\rho(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})]=\rho_{r}(\frac{\beta}{\alpha}+(v_{r}-\frac{\beta}{\alpha})e^{-\alpha t}) -\rho_{l}(\frac{\beta}{\alpha}+(v_{l}-\frac{\beta}{\alpha})e^{-\alpha t}) $$ \rho(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) $穿过间断线的跳跃, 其中$ \rho_{l}=\rho(t, x(t)-0) $$ \rho_{r}=\rho(t, x(t)+0) $.

如果$ \sigma(t)\neq 0, $ 则由(2.5)式可得

$ \begin{eqnarray} &&\frac{1}{2}(\rho_{r}-\rho_{l})\bigg((\frac{\beta}{\alpha}e^{\alpha t}+v_{r}-\frac{\beta}{\alpha}) (\frac{\beta}{\alpha}+(v_{r}-\frac{\beta}{\alpha})e^{-\alpha t})-(\frac{\beta}{\alpha}e^{\alpha t} +v_{l}-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v_{l}-\frac{\beta}{\alpha})e^{-\alpha t})\bigg){}\\ &=&(v_{r}-v_{l})(\rho_{r}(\frac{\beta}{\alpha}+(v_{r}-\frac{\beta}{\alpha})e^{-\alpha t}) -\rho_{l} (\frac{\beta}{\alpha}+(v_{l}-\frac{\beta}{\alpha})e^{-\alpha t})). \end{eqnarray} $

简化(2.6)式可得

因此, 两个非真空常状态$ (\rho_{l}, v_{l}) $$ (\rho_{r}, v_{r}) $可由间断线$ J $连接当且仅当

$ v_{l}=v_{r}. $

下面, 我们构造系统(2.2)–(2.3)的Riemann解, 即分别通过接触间断、真空状态和$ \delta $ -激波连接两个常数状态$ (\rho_{\pm}, u_{\pm}) $.

对于$ u_{-}< u_{+} $情形, Riemann解包含两条间断线和真空状态, 即

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} (\rho_{-}, u_{-}), { } &-\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } \Big(0, \frac{\beta}{\alpha}+\frac{\alpha(x-\frac{\beta}{\alpha}t)}{1 -e^{-\alpha t}} \Big), {\quad} & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) \leq x\\ &{\qquad} { } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ (\rho_{+}, u_{+}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})<x<+\infty. \end{array}\right . \end{align} $

Riemann解可表示为

$ \begin{align} (\rho_{-}, u_{-})+J_{1}+Vac+J_{2}+(\rho_{+}, u_{+}), \end{align} $

其中“+”表示“紧接着”.

对于$ u_{-}= u_{+} $情形, Riemann解包含一条间断线

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} (\rho_{-}, u_{-}), &{ } -\infty<x<\frac{\beta}{\alpha}t +\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ (\rho_{+}, u_{+}), {\quad} &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})<x<+\infty. \end{array}\right . \end{align} $

Riemann解可表示为

$ \begin{align} (\rho_{-}, u_{-})+ J + (\rho_{+}, u_{+}). \end{align} $

对于$ u_{-}> u_{+} $情形, Riemann解不再是经典波, 出现了$ \delta $ -激波. 则Riemann解为

$ \begin{align} (\rho_{-}, u_{-})+ \delta S + (\rho_{+}, u_{+}), \end{align} $

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} (\rho_{-}, u_{-}), &-\infty<x<x(t), \\(w(t)\delta(x-x(t)), v_{\delta}), {\quad} &x=x(t), \\(\rho_{+}, u_{+}), & x(t)<x<+\infty, \end{array}\right . \end{align} $

其中$ x(t) $, $ w(t) $$ u_{\delta}(t) =\frac{\beta}{\alpha}+(v_{\delta}-\frac{\beta}{\alpha})e^{-\alpha t} $分别表示$ \delta $ -激波的位置, 权重和速度, $ v_{\delta} $表示$ v $在这个$ \delta $ -激波曲线上对应的函数值. 这个$ \delta $ -激波$ \delta S $满足如下的广义R-H条件

$ \begin{align} \left\{ \begin{array}{ll} { } \frac{{\rm d}x(t)}{{\rm d}t}=u_{\delta}(t), \\ { } \frac{{\rm d}w(t)}{{\rm d}t}=u_{\delta}(t) [\rho]-[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})], \\ { } u_{\delta}(t) [v]=[\frac{1}{2}(\frac{\beta}{\alpha}e^{\alpha t}+v-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) ], \end{array} \right. \end{align} $

其中$ [\rho] = \rho_{+}- \rho_{-} $, 且$ (x, w)(0) =(0, 0). $

计算(2.13)式可得

$ \begin{align} \begin{array}{l}{ } v_{\delta}=\frac{1}{2}(u_{-}+u_{+}), \; \; x(t)=\frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\delta}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } w(t)=\frac{1}{2\alpha}(\rho_{-}+\rho_{+})(u_{-}-u_{+})(1 -e^{-\alpha t}). \end{array} \end{align} $

由此可知, 这个$ \delta $ -激波满足广义熵条件

$ \begin{align} \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}< u_{\delta}(t) < \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}. \end{align} $

综上所述, 通过变量代换(2.1), 我们可获得系统(1.4)的Riemann解, 即

(1) 当$ u_{-} >u_{+} $, 有

$ \begin{align} (\rho, u)(t, x)=\left\{ \begin{array}{ll} { } (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}), & { x<x(t) , }\\ (w(t)\delta(x-x(t)), u_{\delta} (t)), & { x=x(t) , }\\ { } (\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), & { x>x(t) , } \end{array} \right. \end{align} $

其中

$ \begin{align} \begin{array}{l} { } x(t)=\frac{\beta}{\alpha}t+\frac{1}{\alpha}(\frac{1}{2}(u_{-}+u_{+})-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } w(t)=\frac{1}{2\alpha}(\rho_{-}+\rho_{+})(u_{-}-u_{+})(1 -e^{-\alpha t}), \\ { } u_{\delta}(t)=\frac{\beta}{\alpha}+(\frac{1}{2}(u_{-}+u_{+})-\frac{\beta}{\alpha})e^{-\alpha t}.\end{array} \end{align} $

(2) 当$ u_{-}<u_{+} $, 有

$ \begin{align} (\rho, u)(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } \Big(0, \frac{\beta}{\alpha}+\frac{\alpha(x-\frac{\beta}{\alpha} t)}{e^{\alpha t}-1}\Big), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) \leq x\\ &{\qquad} { } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } (\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), {\quad} & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})<x<+\infty, \end{array}\right . \end{align} $

其中两条间断线$ J_{1} $$ J_{2} $的位置和波速与系统(2.2)–(2.3)相同.

(3) 当$ u_{-}=u_{+} $, 有

$ \begin{align} (\rho, u)(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } (\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})<x<+\infty, \end{array}\right . \end{align} $

其中间断线$ J_{} $的位置和波速与系统(2.2)–(2.3)相同.

3 系统(1.1)的Riemann问题

本节, 我们构造带有源项的一维可压缩流体欧拉方程组(1.1)的Riemann解.

通过变量代换(2.1), 系统(1.1)可写为下列形式

$ \begin{align} \left\{\begin{array}{ll} { } \rho_{t}+(\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}))_{x}=0, \\ { } v_{t}+(\frac{1}{2}(\frac{\beta}{\alpha}e^{\alpha t}+v-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) +\frac{\gamma-1}{4}\rho^{\gamma-1}e^{\alpha t} )_{x}=0.\end{array}\right . \end{align} $

接下来, 我们将研究具有下列初值的系统(3.1)的Riemann问题

$ \begin{align} (\rho, v)(0, x) =\left\{\begin{array}{ll} (\rho_{-}, u_{-}), \, \, \, \, x< 0, \\(\rho_{+}, u_{+}), \, \, \, \, x> 0, \end{array} \right. \end{align} $

其中$ \rho_{\pm}>0 $$ u_{\pm} $是常数.

系统(3.1)可写为下列拟线性形式

$ \begin{align} \left(\begin{array}{cc}\rho\\v \end{array}\right)_t+\left( \begin{array}{ccc} { } \frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}{\quad} & \rho e^{-\alpha t} \\ { } \frac{(\gamma-1)^{2}}{4}\rho^{\gamma-2}e^{\alpha t}{\quad} &{ } \frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t} \end{array}\right)\left(\begin{array}{cc}\rho\\v \end{array}\right)_x=\left(\begin{array}{cc}0\\0 \end{array}\right). \end{align} $

由(3.3)式可知, 系统(3.1)的特征值为

对应的右特征向量为

满足

因此, 系统(3.1)关于$ \rho>0 $是严格双曲型的, 且两个特征值均为真正非线性的, 相应的基本波为激波或稀疏波.

系统(3.1)的Riemann不变量可选为

分别满足$ \bigtriangledown w \cdot \overrightarrow{r_1}=0 $$ \bigtriangledown z \cdot \overrightarrow{r_2}=0 $.

下面让我们把注意力放在系统(3.1)的基本波上. 我们首先用稀疏波作为基本解. 我们知道如果$ \alpha=0 $$ \beta=0 $, 也就是说, 在齐次情况下, 存在自相似解$ (\rho, v)(t, x)=(\rho, v)(\xi) $$ (\xi=x/t) $. 但是现在情况则不存在像齐次情形那样的经典自相似解$ (\rho, v)(x/t) $. 在这种情况下, 解决方法必须另辟蹊径. 由于稀疏波是一种满足系统(3.1)的连续解, 可以通过确定每个特征场的积分曲线得到. 注意到$ k $ - 黎曼不变量$ (k = 1, 2) $$ k $ -稀疏波中是守恒的。因此, 就像文献[23]一样, 我们仍然可以使用稀疏波作为基本解.

在相平面上, 给定一个状态$ (\rho_{-}, u_{-}) $, 稀疏波曲线是通过1 -稀疏波或2 -稀疏波在右侧连接的状态集合, 即

$ \begin{align} R_{1}(\rho_{-}, u_{-}):\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{1}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}-\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v+\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \rho<\rho_{-}, v>u_{-}, \\ { } \lambda_{1}(\rho_{-}, u_{-})<\lambda_{1}(\rho, v), \end{array} \right. \end{align} $

$ \begin{align} R_{2}(\rho_{-}, u_{-}):\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{2}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}+\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v-\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}-\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \rho>\rho_{-}, v>u_{-}, \\ \lambda_{2}(\rho_{-}, u_{-})<\lambda_{2}(\rho, v).\end{array} \right. \end{align} $

对方程(3.4)的第二个方程$ v $关于$ \rho $求导, 可得

则1 -稀疏波曲线$ R_{1}(\rho_{-}, u_{-}) $$ (\rho, v) $相平面上是单调递减的凸函数. 同样的, 对(3.5)式的第二个方程$ v $关于$ \rho $求导得$ \frac{{\rm d}v}{{\rm d}\rho}> 0 $$ \frac{{\rm d}^{2}v}{{\rm d}\rho^{2}} < 0 $, 则2 -稀疏波曲线$ R_{2}(\rho_{-}, u_{-}) $$ (\rho, v) $相平面上是单调递增的凹函数. 而且通过(3.4)式, 对于1 -稀疏波曲线

$ R_{1}(\rho_{-}, u_{-}) $$ v $ -轴上交于点$ (0, \widetilde{v}_{\ast}) $, 其中$ \widetilde{v}_{\ast} $ is determined by $ \widetilde{v}_{\ast}= u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t} $. 对于$ R_{2}(\rho_{-}, u_{-}) $, 从(3.5)式可知$ \lim\limits_{\rho\rightarrow +\infty}v=+\infty. $

我们首先计算在1 -稀疏波内部点$ (t, x) $上的Riemann解$ (\rho, v) $. 由(3.4) 式的第一个方程可得

$ \begin{align} \frac{x}{t}-\frac{\beta}{\alpha} =(v-\frac{\beta}{\alpha})\frac{1 -e^{-\alpha t}}{\alpha t}-\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}. \end{align} $

结合(3.6)式和(3.4)式的第二个方程可得

$ \begin{align} \left\{\begin{array}{ll} { } \rho(t, x)=\Big(\frac{\rho^{\frac{\gamma-1}{2}}_{-}\frac{e^{\alpha t}-1}{\alpha t} +u_{-}\frac{1 -e^{-\alpha t}}{\alpha t}-\frac{x}{t}+\frac{\beta}{\alpha} \frac{\alpha t-1 +e^{-\alpha t}}{\alpha t}}{\frac{\gamma-1}{2}+\frac{e^{\alpha t}-1}{\alpha t}}\Big)^{\frac{2}{\gamma-1}}, \\ { } v(t, x)=\frac{\frac{x}{t}+\frac{\gamma-1}{2}(u_{-}e^{-\alpha t}+\rho^{\frac{\gamma-1}{2}}_{-})+ \frac{\beta}{\alpha}\frac{1 -e^{-\alpha t}-\alpha t}{\alpha t}}{\frac{\gamma-1}{2}e^{-\alpha t} +\frac{1 -e^{-\alpha t}}{\alpha t}}.\end{array}\right . \end{align} $

同样地, 我们计算在2 -稀疏波内部点$ (t, x) $上的Riemann解$ (\rho, v) $, 由(3.5) 式的第一个方程得到

$ \begin{align} \frac{x}{t}-\frac{\beta}{\alpha} =(v-\frac{\beta}{\alpha})\frac{1 -e^{-\alpha t}} {\alpha t}+\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}. \end{align} $

结合(3.8)式和(3.5) 式的第二个方程可知

$ \begin{align} \left\{\begin{array}{ll} { } \rho(t, x)=\Big(\frac{\rho^{\frac{\gamma-1}{2}}_{-}\frac{e^{\alpha t}-1}{\alpha t}-u_{-}\frac{1 -e^{-\alpha t}}{\alpha t}+\frac{x}{t}+\frac{\beta}{\alpha} \frac{1 -e^{-\alpha t}-\alpha t}{\alpha t}}{\frac{\gamma-1}{2}+\frac{e^{\alpha t}-1}{\alpha t}}\Big)^{\frac{2}{\gamma-1}}, \\ { } v(t, x)=\frac{\frac{x}{t}+\frac{\gamma-1}{2}(u_{-}e^{-\alpha t}-\rho^{\frac{\gamma-1}{2}}_{-}) +\frac{\beta}{\alpha}\frac{1 -e^{-\alpha t}-\alpha t}{\alpha t}}{\frac{\gamma-1}{2}e^{-\alpha t} +\frac{1 -e^{-\alpha t}}{\alpha t}}.\end{array}\right . \end{align} $

$ \sigma(t)=\frac{{\rm d}x (t)}{{\rm d}t} $是有界间断$ x=x(t) $的波速, 则系统(3.1) 满足以下Rankine-Hugoniot条件

$ \begin{align} \left\{ \begin{array}{ll} { } -\sigma(t)[\rho]+[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})]=0, \\ { } -\sigma(t)[ v]+[\frac{1}{2}(\frac{\beta}{\alpha}e^{\alpha t}+v-\frac{\beta}{\alpha})(\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) +\frac{\gamma-1}{4}\rho^{\gamma-1}e^{\alpha t} ]=0, \end{array} \right. \end{align} $

其中$ [\rho]=\rho-\rho_{-} $等. 由(3.10)式获得

$ \begin{array}{c} \sigma (t)=\frac{[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})]}{[\rho]},\\ \frac{v-u_{-}}{\rho-\rho_{-}} =\pm e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho+\rho_{-})(\rho-\rho_{-})}}, \end{array} $

其中$ (\rho_{-}, u_{-}) $$ (\rho_{}, v_{}) $分别是左状态和右状态.

1 -激波曲线$ S_{1} (\rho_{-}, u_{-}) $

Lax熵条件说明1 -激波的波速$ \sigma_{1} (t) $满足

$ \begin{align} \sigma_{1} (t)<\lambda_{1} (\rho_{-}, u_{-}), \, \, \, \, \lambda_{1} (\rho, v)<\sigma_{1} (t)<\lambda_{2} (\rho, v). \end{align} $

通过(3.10)式的第一个方程可知

$ \begin{align} \sigma_{1} (t)&{ } =\frac{\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}) -\rho_{-} (\frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t})}{\rho-\rho_{-}} \\ &{ } = \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}+\frac{\rho e^{-\alpha t}}{\rho-\rho_{-}}(v-u_{-}). \end{align} $

将(3.13)式代入(3.12)式的第一个不等式, 可得

这意味着$ v - u_{-} $$ \rho- \rho_{-} $有不同的符号. 因此, 通过(3.11) 式可得

$ v>u_{-} $, 则$ \rho<\rho_{-} $

对于$ \bar{\rho}\in(\rho, \rho_{-}). $ 通过直接计算可得

这与$ \sigma_{1} (t)<\lambda_{1} (\rho_{-}, u_{-}) $矛盾. 因此, 给定一个状态$ (\rho_{-}, u_{-}) $, 1 -激波曲线$ S_{1} (\rho_{-}, u_{-}) $是相平面上所有通过1 -激波右连接$ (\rho_{-}, u_{-}) $的状态集合, 即

$ \begin{align} S_{1} (\rho_{-}, u_{-}):\, \, \left\{\begin{array}{ll} { } \sigma_{1} (t)=\frac{\beta}{\alpha} +(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t} -\rho\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho+\rho_{-})(\rho-\rho_{-})}}, \\ { } v=u_{-}- e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})} {(\rho+\rho_{-})(\rho-\rho_{-})}}(\rho-\rho_{-}), \; \; \; \\ \rho>\rho_{-}, v<u_{-}.\end{array} \right. \end{align} $

2 -激波曲线$ S_{2} (\rho_{-}, u_{-}) $

同样地, 2 -激波$ S_{2} $的波速$ \sigma_{2} (t) $满足

则给定一个状态$ (\rho_{-}, u_{-}) $, 2 -激波曲线$ S_{2} (\rho_{-}, u_{-}) $是相平面上所有通过2 -激波右连接$ (\rho_{-}, u_{-}) $的状态集合, 即

$ \begin{align} S_{2} (\rho_{-}, u_{-}):\, \, \left\{\begin{array}{ll} { } \sigma_{2} (t)=\frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t} +\rho\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho+\rho_{-})(\rho-\rho_{-})}}, \\ { } v=u_{-} +e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})} {(\rho+\rho_{-})(\rho-\rho_{-})}}(\rho-\rho_{-}), \; \; \; \\ \rho<\rho_{-}, v<u_{-}.\end{array} \right. \end{align} $

对(3.14)式的第二个方程$ v $的两边关于$ \rho $求导, 则当$ \rho>\rho_{-} $时, 有

即在相平面$ (\rho, v) $上, 1 -激波曲线$ S_{1} (\rho_{-}, u_{-}) $单调递减. 同样地, 对于(3.15)式, 当$ \rho<\rho_{-} $时, $ \frac{{\rm d}v}{{\rm d}\rho} > 0, $ 即2 -激波曲线$ S_{2} (\rho_{-}, u_{-}) $在相平面$ (\rho, v) $上是单调递增. 由(3.15)式可得

$ S_{2} (\rho_{-}, u_{-}) $$ v $ -轴交于$ (0, \widetilde{v}_{\ast\ast} ) $, 其中$ \widetilde{v}_{\ast\ast} =u_{-}-e^{\alpha t}\sqrt{\frac{\gamma-1}{2}}\rho_{-}^{\frac{\gamma-1}{2}}. $$ S_{1} (\rho_{-}, u_{-}) $, 由(3.14)式可得$ \lim\limits_{\rho\rightarrow +\infty}v= -\infty $.

在相平面$ (\rho, v) $上, 通过给定状态$ (\rho_{-}, u_{-}) $, 我们画出基本波$ R_{j} (\rho_{-}, u_{-}) $$ S_{j} (\rho_{-}, u_{-}) $$ (j=1, 2). $ 基本波将相平面$ (\rho, v) $划分为五个区域(见图 1). 根据在不同区域的右状态$ (\rho_{+}, u_{+}) $, 我们构造(3.1)和(3.2)式的唯一全局Riemann解即

图 1

图 1   基本波曲线


(1) $ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}): $$ (\rho_{-}, u_{-})+R_{1} +(\rho_{\ast}, v_{\ast})+R_{2} +(\rho_{+}, u_{+}); $

(2) $ (\rho_{+}, u_{+})\in II(\rho_{-}, u_{-}): $$ (\rho_{-}, u_{-})+S_{1} +(\rho_{\ast}, v_{\ast})+R_{2} +(\rho_{+}, u_{+}); $

(3) $ (\rho_{+}, u_{+})\in III(\rho_{-}, u_{-}): $$ (\rho_{-}, u_{-})+R_{1} +(\rho_{\ast}, v_{\ast})+S_{2} +(\rho_{+}, u_{+}); $

(4) $ (\rho_{+}, u_{+})\in IV(\rho_{-}, u_{-}): $$ (\rho_{-}, u_{-})+S_{1} +(\rho_{\ast}, v_{\ast})+S_{2} +(\rho_{+}, u_{+}); $

(5) $ (\rho_{+}, u_{+})\in V(\rho_{-}, u_{-}): $$ (\rho_{-}, u_{-})+R_{1} +{\rm Vac}+R_{2} +(\rho_{+}, u_{+}). $

准确地说, 若$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $, 则Riemann解包含两条稀疏波$ R_{1} $$ R_{2} $以及中间状态$ (\rho_{\ast}, v_{\ast}) $, 其中$ (\rho_{\ast}, v_{\ast}) $只取决于

通过简单计算, 不难发现

$ \begin{align} v_{\ast}=\frac{1}{2}(u_{-}+u_{+}+(\rho_{-}^{\frac{\gamma-1}{2}}-\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}), \, \, \, \, \, \rho_{\ast}=\Bigg(\frac{(u_{-}-u_{+})e^{-\alpha t}+\rho_{-}^{\frac{\gamma-1}{2}} +\rho_{+}^{\frac{\gamma-1}{2}}}{2}\Bigg)^{\frac{2}{\gamma-1}}. \end{align} $

因此, 系统(3.1)和(3.2)的Riemann解可写成

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} (\rho_{-}, u_{-}), &{ } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t, \\ R_{1}, &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad}{ } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ (\rho_{\ast}, v_{\ast}), &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t<x\\ &{\qquad} { } < \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) +\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ R_{2}, &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha} (v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad}{ } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) +\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t, \\ (\rho_{+}, u_{+}), & { } \frac{\beta}{\alpha}t+\frac{1} {\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t<x<+\infty, \end{array}\right . \end{align} $

其中$ R_{1} $上状态$ (\rho_{1}, v_{1}) $可以由(3.7)式计算, 而$ R_{2} $上状态$ (\rho_{2}, v_{2}) $可以由(3.9)式分别用$ \rho_{\ast} $$ v_{\ast} $代替$ \rho_{-} $$ u_{-} $计算.

$ (\rho_{+}, u_{+})\in II(\rho_{-}, u_{-}) $, 则Riemann解包含1 -激波$ S_{1} $和2 -稀疏波$ R_{2} $以及中间状态$ (\rho_{\ast}, v_{\ast}) $, 其中$ (\rho_{\ast}, v_{\ast}) $只取决于

$ \begin{align} \left\{\begin{array}{ll} { } v_{\ast}=u_{-}-e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}- \rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}}(\rho_{\ast}-\rho_{-}), \\ { } v_{\ast}-\rho_{\ast}^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{+}-\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}. \end{array} \right. \end{align} $

系统(3.1)和(3.2)的Riemann解可写成

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, u_{-}), &{ } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})-\rho_{\ast} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})} {(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}} \, t, \\ (\rho_{\ast}, v_{\ast}), &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\rho_{\ast} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}} \, t<x\\ &{\qquad} { } <\frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) +\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ R_{2}, &{ } \frac{\beta}{\alpha}t+\frac{1} {\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad}{ } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) +\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t, \\ (\rho_{+}, u_{+}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) +\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t<x<+\infty, \end{array}\right . \end{align} $

其中, $ R_{2} $上状态$ (\rho_{2}, v_{2}) $也可以用前面的方法计算, $ S_{1} $的位置由下式确定

$ \begin{align} x=\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\rho_{\ast} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}} \, t. \end{align} $

$ (\rho_{+}, u_{+})\in III(\rho_{-}, u_{-}) $, 则Riemann解包含1 -稀疏波$ R_{1} $和2 -激波$ S_{2} $以及中间状态$ (\rho_{\ast}, v_{\ast}) $, 其中$ (\rho_{\ast}, v_{\ast}) $只取决于

$ \begin{align} \left\{\begin{array}{ll} { } v_{\ast}+\rho_{\ast}^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \\ { } u_{+}=v_{\ast} +e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}}(\rho_{+} -\rho_{\ast}). \end{array} \right. \end{align} $

系统(3.1)和(3.2)的Riemann解可写成

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} (\rho_{-}, u_{-}), &{ } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t, \\ R_{1}, &{ } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad} { } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})- \frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ (\rho_{\ast}, v_{\ast}), &{ } \frac{\beta}{\alpha}t +\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t<x\\ &{\qquad}{ } <\frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\rho_{+} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} \, t, \\ (\rho_{+}, u_{+}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})+\rho_{+} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}- \rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} \, t<x<+\infty, \end{array}\right . \end{align} $

其中$ R_{1} $上状态$ (\rho_{1}, v_{1}) $可以由(3.7)式计算, $ S_{2} $的位置由下式确定

$ \begin{align} x=\frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})+\rho_{+} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1} -\rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} \, t. \end{align} $

$ (\rho_{+}, u_{+})\in IV(\rho_{-}, u_{-}) $, 则Riemann解由1 -激波$ S_{1} $和2 -激波$ S_{2} $以及中间状态$ (\rho_{\ast}, v_{\ast}) $组成, 其中

$ \begin{align} \left\{\begin{array}{ll} { } v_{\ast}=u_{-}-e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1} -\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast} -\rho_{-})}}(\rho_{\ast}-\rho_{-}), \\ { } u_{+}=v_{\ast} +e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} (\rho_{+}-\rho_{\ast}).\end{array} \right. \end{align} $

系统(3.1)和(3.2)的Riemann解可写成

$ \begin{align} (\rho, v)(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, u_{-}), &{ } -\infty<x<\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})-\rho_{\ast} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})} {(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}} \, t, \\ (\rho_{\ast}, v_{\ast}), &{ } \frac{\beta}{\alpha}t +\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})-\rho_{\ast} \sqrt{\frac{\frac{\gamma-1}{2} (\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}} t<x\\ &{\qquad}{ } <\frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\rho_{+} \sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast}) (\rho_{+}-\rho_{\ast})}} \, t, \\ (\rho_{+}, u_{+}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha} (v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\rho_{+} \sqrt{\frac{\frac{\gamma-1}{2} (\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} \, t<x<+\infty, \end{array}\right . \end{align} $

其中$ S_{1} $$ S_{2} $的位置分别由(3.20)和(3.23)式确定.

通过变量代换(1.2), 可得到系统(1.1)的Riemann解为

(1) $ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}): $$ (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t})+R_{1} +(\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t})+R_{2} +(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}); $

(2) $ (\rho_{+}, u_{+})\in II(\rho_{-}, u_{-}): $$ (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t})+S_{1} +(\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t}) +R_{2} +(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}); $

(3) $ (\rho_{+}, u_{+})\in III(\rho_{-}, u_{-}): $$ (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta} {\alpha})e^{-\alpha t})+R_{1} +(\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha}) e^{-\alpha t})+S_{2} +(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}); $

(4) $ (\rho_{+}, u_{+})\in IV(\rho_{-}, u_{-}): $$ (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha}) e^{-\alpha t})+S_{1} +(\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t})+S_{2} +(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}); $

(5) $ (\rho_{+}, u_{+})\in V(\rho_{-}, u_{-}): $$ (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha}) e^{-\alpha t})+R_{1} +{\rm Vac}+R_{2} +(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}). $

我们以$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $为例来说明. 如果$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $, 则系统(1.1)的Riemann解可表示为

$ \begin{align} (\rho, u)(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}), &-\infty<x\\ &{\qquad} { } <\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t, \\ { } (\rho_{1}, \frac{\beta}{\alpha}+(v_{1}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) -\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad} { } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})-\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ { } (\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})- \frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t<x\\ &{\qquad}{ } < \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t, \\ { } (\rho_{2}, \frac{\beta}{\alpha}+(v_{2}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(v_{\ast}-\frac{\beta}{\alpha})( 1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{\ast}^{\frac{\gamma-1}{2}}t\leq x\\ &{\qquad} { } \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha}) (1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t, \\ { } (\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), & { } \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t})+\frac{\gamma-1}{2}\rho_{+}^{\frac{\gamma-1}{2}}t\\ &{\qquad} <x<+\infty, \end{array}\right . \end{align} $

其中状态$ (\rho_{1}, v_{1}) $, $ (\rho_{\ast}, v_{\ast}) $$ (\rho_{2}, v_{2}) $也可以像以前一样计算.

4 系统(1.1)的Riemann解的极限

在本节, 我们分别讨论$ u_{+}<u_{-} $$ u_{+}>u_{-} $两种情形. 令$ \gamma \rightarrow 1 $, 研究系统(1.1) 的Riemann解的变化, 即分别研究系统(1.1)的delta激波和真空状形成.

引理4.1  若$ u_{+}<u_{-} $, 则存在一个足够小的$ \gamma_{0} > 0 $, 使得当$ 1<\gamma <1 +\gamma_{0} $时, $ (\rho_{+}, u_{+})\in IV( \rho_{-}, u_{-}) $.

  若$ \rho_{+} = \rho_{-} $, 则对$ \forall\gamma\in (1, 2) $, $ (\rho_{+}, u_{+})\in IV( \rho_{-}, u_{-}) $成立. 因此下面我们只需要考虑$ \rho_{+} \neq\rho_{-} $情况.

通过(3.14)和(3.15)式, 易知所有与左状态$ (\rho_{-}, u_{-}) $通过1 -激波$ S_{1} $或2 -激波$ S_{2} $连接的$ (\rho, v) $满足

$ \begin{equation} S_{1}: \, \, \, \, v=u_{-}-e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho+\rho_{-})(\rho-\rho_{-})}}(\rho-\rho_{-}), \; \; \; \rho>\rho_{-}, \end{equation} $

$ \begin{equation} S_{2}: \, \, \, \, v=u_{-}+e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho+\rho_{-})(\rho-\rho_{-})}}(\rho-\rho_{-}), \; \; \rho<\rho_{-}. \end{equation} $

$ \rho_{+} \neq\rho_{-} $$ (\rho_{+}, u_{+})\in IV( \rho_{-}, u_{-}) $, 则由图 1, (4.1) 和(4.2) 式, 可得

$ \begin{equation} u_{+}<u_{-}-e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{+}+\rho_{-})(\rho_{+}-\rho_{-})}}(\rho_{+}-\rho_{-}), \; \; \; \mathrm {} \, \, \rho_{+}>\rho_{-}, \end{equation} $

$ \begin{equation} u_{+}<u_{-}+e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{+}+\rho_{-})(\rho_{+}-\rho_{-})}}(\rho_{+}-\rho_{-}), \; \; \; \mathrm {} \, \, \rho_{+}<\rho_{-}. \end{equation} $

由(4.3)和(4.4)式, 有

$ \begin{equation} e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{-}^{\gamma-1})}{\rho_{+}^{2}-\rho_{-}^{2}}}< \frac{u_{-}-u_{+}}{|\rho_{+}-\rho_{-}|}. \end{equation} $

由于

$ \begin{equation} \lim\limits_{{\gamma\rightarrow1}}e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{-}^{\gamma-1})}{\rho_{+}^{2}-\rho_{-}^{2}}}= e^{\alpha t}\lim\limits_{{\gamma\rightarrow1}}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{-}^{\gamma-1})}{\rho_{+}^{2}-\rho_{-}^{2}}}=0, \end{equation} $

因此存在$ \gamma_{0}>0 $足够小, 当$ 1<\gamma <1 +\gamma_{0} $时, 我们有不等式(4.5)成立. 所以, 当$ 1<\gamma <1 +\gamma_{0} $时, $ (\rho_{+}, u_{+})\in IV( \rho_{-}, u_{-}) $成立. 证毕.

$ 1<\gamma <1+\gamma_{0} $, 即$ (\rho_{+}, u_{+})\in IV( \rho_{-}, u_{-}) $, 假设中间状态$ (\rho_{\ast}, v_{\ast}) $通过波速为$ \sigma_{1}(t) $的1 -激波$ S_{1} $连接$ (\rho_{-}, u_{-}) $, 且通过波速为$ \sigma_{2}(t) $的2 - 激波$ S_{2} $连接$ (\rho_{+}, u_{+}) $, 则

$ \begin{equation} S_{1} :\, \, \left\{\begin{array}{ll} { } \sigma_{1} (t)=\frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t} -\rho_{\ast}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})} {(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}}, \\ { } v_{\ast}=u_{-}-e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1} -\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast} -\rho_{-})}}(\rho_{\ast}-\rho_{-}), \; \; \; \rho_{\ast}>\rho_{-}, \end{array} \right. \end{equation} $

$ \begin{equation} S_{2} :\, \, \left\{\begin{array}{ll} { } \sigma_{2} (t)=\frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t} +\rho_{+}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}}, \\ { } u_{+}=v_{\ast} +e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} (\rho_{+}-\rho_{\ast}), \; \; \; \rho_{\ast}>\rho_{+}.\end{array} \right. \end{equation} $

由(4.7)和(4.8)式, 有

$ \begin{eqnarray} u_{-}-u_{+}&=&e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-} ^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}}(\rho_{\ast}-\rho_{-}){}\\ &&+ e^{\alpha t}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{\ast}+\rho_{+})(\rho_{+}-\rho_{\ast})}}(\rho_{\ast}-\rho_{+}), \, \, \, \, \rho_{\ast}>\rho_{\pm}. \end{eqnarray} $

则有下列引理.

引理4.2  $ \lim\limits_{\gamma\rightarrow1}\rho_{\ast}=+\infty, $$ \lim\limits_{\gamma\rightarrow1}\frac{\gamma-1}{2}\rho_{\ast}^{\gamma-1}=:a=\frac{((u_{-}-u_{+})e^{-\alpha t})^{2}}{4}. $

  令$ \lim\limits_{\gamma\rightarrow1}\inf\rho_{\ast}=\alpha $, $ \lim\limits_{\gamma\rightarrow1}\sup\rho_{\ast}=\beta $.

$ \alpha<\beta $, 则由$ \rho_{\ast}(\gamma) $的连续性可知, 存在数列$ \{\gamma_{n}\}_{n=1}^{\infty}\subseteq(1, 2) $使得

其中$ c\in(\alpha, \beta). $ 将数列代入(4.9)式的右边, 并令$ n\rightarrow +\infty $, 则有

$ \begin{equation} \lim\limits_{n\rightarrow +\infty}\sqrt{\frac{\frac{\gamma_{n}-1}{2}(\rho_{\ast}(\gamma_{n})^{\gamma_{n}-1}-\rho_{\pm}^{\gamma_{n}-1})} {(\rho_{\ast}(\gamma_{n})+\rho_{\pm})(\rho_{\ast}(\gamma_{n})-\rho_{\pm})}}(\rho_{\ast}(\gamma_{n})-\rho_{\pm})=0. \end{equation} $

因此, 由(4.9)式可知

这与假设$ u_{-}>u_{+} $矛盾. 从而可知$ \alpha=\beta $, 即$ \lim\limits_{\gamma\rightarrow1}\rho_{\ast}(\gamma)=\alpha. $

$ \alpha\in(0, +\infty), $ 那么在(4.9)式取极限时我们也可以得到一个矛盾. 因此$ \alpha=0 $$ \alpha=+\infty $. 由条件$ \rho_{\ast}>\max\{\rho_{-}, \rho_{+}\} $, 易知

对(4.9)式的右边取极限$ \gamma\rightarrow 1 $, 则有

从而可知$ a=\frac{((u_{-}-u_{+})e^{-\alpha t})^{2}}{4}. $ 证毕.

引理4.3  若$ u_{-}>u_{+}, $

$ \begin{equation} \lim\limits_{\gamma\rightarrow1}u_{\ast}= \lim\limits_{\gamma\rightarrow1}(\frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t}) =\lim\limits_{\gamma\rightarrow1}\sigma_{1}(t)=\lim\limits_{\gamma\rightarrow1}\sigma_{2}(t)=u_{\delta}(t), \end{equation} $

$ \begin{equation} \lim\limits_{\gamma\rightarrow1}\int^{\sigma_{2}(t)}_{\sigma_{1}(t)}\rho_{\ast} {\rm d}\xi=u_{\delta}(t)[\rho]-[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})] =\frac{1}{2}(\rho_{-}+\rho_{+})(u_{-}-u_{+})e^{-\alpha t}, \end{equation} $

其中$ [\rho] = \rho_{+}- \rho_{-} $$ u_{\delta}(t)=\frac{\beta}{\alpha}+(\frac{1}{2} (u_{-}+u_{+})-\frac{\beta}{\alpha})e^{-\alpha t}. $

  由(2.1), (4.7)和(4.8)式以及引理4.2, 可得

因此$ \lim\limits_{\gamma\rightarrow1}u_{\ast}= \lim\limits_{\gamma\rightarrow1} \sigma_{1}(t)=\lim\limits_{\gamma\rightarrow1}\sigma_{2}(t)=u_{\delta}(t). $

类似地, 由(4.7)式, (4.8)式和引理4.2, 可得

$ \begin{eqnarray} \lim\limits_{\gamma\rightarrow1}\int^{\sigma_{2}(t)}_{\sigma_{1}(t)}\rho_{\ast}{\rm d}\xi &=&\lim\limits_{\gamma\rightarrow1}\rho_{\ast}(\sigma_{2}(t)-\sigma_{1}(t)){}\\ &=&\lim\limits_{\gamma\rightarrow1}(\rho_{\ast}(v_{\ast}-u_{-})e^{-\alpha t}+\rho_{\ast}\rho_{+}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} {} \\ & &+\rho_{\ast}^{2}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}}) {}\\ & =&\lim\limits_{\gamma\rightarrow1}(-\rho_{\ast}(\rho_{\ast}-\rho_{-})\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast} +\rho_{-})(\rho_{\ast}-\rho_{-})}} {}\\ && +\rho_{\ast}\rho_{+}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})} {(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}} +\rho_{\ast}^{2}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast}+\rho_{-})(\rho_{\ast}-\rho_{-})}}) {}\\ &=&\lim\limits_{\gamma\rightarrow1}(\rho_{\ast}\rho_{-}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{\ast}^{\gamma-1}-\rho_{-}^{\gamma-1})}{(\rho_{\ast} +\rho_{-})(\rho_{\ast}-\rho_{-})}} +\rho_{\ast}\rho_{+}\sqrt{\frac{\frac{\gamma-1}{2}(\rho_{+}^{\gamma-1}-\rho_{\ast}^{\gamma-1})}{(\rho_{+}+\rho_{\ast})(\rho_{+}-\rho_{\ast})}}) {} \\ &=&\rho_{-}\sqrt{a}+\rho_{+}\sqrt{a}=(\rho_{-}+\rho_{+})\frac{1}{2} (u_{-}-u_{+})e^{-\alpha t}. \end{eqnarray} $

直接计算, 有

$ \begin{eqnarray} & &u_{\delta}(t)[\rho]-[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})]{}\\ &=&\frac{\beta}{\alpha}[\rho]+\frac{1}{2}(u_{-}+u_{+})e^{-\alpha t}[\rho]-\frac{\beta}{\alpha} e^{-\alpha t}[\rho]-[\rho (\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t})] {}\\ & =&\frac{1}{2}(u_{-}+u_{+})[\rho]e^{-\alpha t}-[\rho v]e^{-\alpha t}{}\\ & =&\big(\frac{1}{2}(u_{-}+u_{+})(\rho_{+}-\rho_{-})-\rho_{+}u_{+}+\rho_{-}u_{-}\big)e^{-\alpha t}{}\\ &=&\frac{1}{2}(\rho_{-}+\rho_{+})(u_{-}-u_{+})e^{-\alpha t}. \end{eqnarray} $

则由(4.13)和(4.14)式可立即得到(4.12)式. 证毕.

注4.1  从引理4.2 –4.3可以得出这样的结论: 当$ \gamma\rightarrow1 $时, 两条激波曲线$ S_{1} $$ S_{2} $重合, 中间状态的密度$ \rho_{\ast} $变为奇异, $ \rho_{\ast} $的极限具有一个奇点, 该奇点是一个带有速度$ u_{\delta}(t) $的加权狄拉克$ \delta $函数.

注4.2  从引理4.3可以得出这样的结论: 当$ \gamma\rightarrow1 $时, 两条激波$ S_{1} $$ S_{2} $的速度和系统(1.1)的中间$ u_{\ast} $收敛于$ u_{\delta}(t) $, 它决定了带阻尼和摩擦的零压气体欧拉方程组的$ \delta $ -激波解, 且两条激波之间的中间密度$ \rho_{\ast} $趋于一个加权的$ \delta $测度, 从而形成$ \delta $ -激波.

根据上面的分析, 我们得到了以下结果.

定理4.1  若$ u_{+}<u_{-} $, 当$ \gamma\rightarrow1 $时, 带有初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.1), 在第3节构造的包含两条激波的黎曼解收敛于具有同样初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.4) 的$ \delta $ -激波解.

引理4.4  如果$ u_{-}<u_{+}<u_{-}+2, $ 则存在$ \gamma_{1}>0 $, 使得$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $, 当$ 1<\gamma<1+\gamma_{1}. $

  由(3.4)和(3.5)式可推导出通过1 -稀疏波$ R_{1} $或2 -稀疏波$ R_{2} $与左状态$ (\rho_{-}, u_{-}) $右连接的所有可能的状态$ (\rho, v) $满足

$ \begin{equation} R_{1}(\rho_{-}, u_{-}): v+\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, v>u_{-}, \, \, \, \rho<\rho_{-}, \end{equation} $

$ \begin{equation} R_{2}(\rho_{-}, u_{-}): v-\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}-\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \, \, v>u_{-}, \, \, \, \rho>\rho_{-}. \end{equation} $

同样地, 由(3.5)式可知, 所有通过2 -稀疏波$ R_{2} $与左状态$ (0, \widetilde{v}_{\ast}) $右连接的状态$ (\rho, v) $满足

$ \begin{equation} R_{2}(0, \widetilde{v}_{\ast}): v-\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \, \, v>u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \rho>0. \end{equation} $

$ u_{-}<u_{+}<u_{-}+2 $, $ \rho_{+}\neq\rho_{-} $$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}), $ 则结合(4.15)–(4.17)式, 由图 1可得

$ \begin{equation} u_{+}>u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}-\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \rho_{+}<\rho_{-}, \end{equation} $

$ \begin{equation} u_{+}>u_{-}-\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}+\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \rho_{+}>\rho_{-}, \end{equation} $

$ \begin{equation} u_{+}<u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}+\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \rho_{+}>0. \end{equation} $

由(4.18)–(4.20)式, 可知

$ \begin{equation} |(\rho_{-}^{\frac{\gamma-1}{2}}-\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}|<u_{+}-u_{-}<(\rho_{-}^{\frac{\gamma-1}{2}}+\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}, \, \, \, \, \, \, \, \, \, \, \rho_{+}>0, \, \rho_{-}>0. \end{equation} $

由于$ \lim\limits_{{\gamma\rightarrow1}}(\rho_{-}^{\frac{\gamma-1}{2}}-\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}=0<u_{+}-u_{-} $$ \lim\limits_{{\gamma\rightarrow1}}(\rho_{-}^{\frac{\gamma-1}{2}}+\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}=2e^{\alpha t}>2>u_{+}-u_{-}, $ 因此, 存在足够小$ \gamma_{1}>0 $, 使得当$ 1<\gamma <1 +\gamma_{1} $时, 我们有不等式(4.21). 于是, 当$ 1<\gamma <1 +\gamma_{1} $时, 很明显$ (\rho_{+}, u_{+})\in I( \rho_{-}, u_{-}) $. 证毕.

$ u_{-}<u_{+}<u_{-}+2, $ 由引理4.5, 对任意给定$ \gamma\in (1, 1+\gamma_{1}), $ 带有Riemann初值的系统(1.1)的Riemann解为

$ \begin{equation} (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t})+R_{1}+(\rho_{\ast}, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t})+R_{2}+(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), \end{equation} $

其中

$ \begin{equation} R_{1}:\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{1}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}-\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v+\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \rho_{\ast}\leq\rho\leq\rho_{-}, \end{array} \right. \end{equation} $

$ \begin{equation} R_{2}:\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{2}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}+\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v-\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{+}-\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \rho_{\ast}\leq\rho\leq\rho_{+}.\end{array} \right. \end{equation} $

因此, 从(4.23)和(4.24)式可得

$ \begin{equation} u_{+}-u_{-}=\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}-2\rho_{\ast}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \, \, \, \, \, \, \, \, \, \rho_{\ast}\leq\rho_{\pm}. \end{equation} $

$ \gamma\rightarrow1 $时, 真空状态出现.

定理4.2  如果$ u_{-}<u_{+}< u_{-}+2 $. 对任意$ \gamma \in(1, 2) $, 假设$ (\rho, u)(t, x) $是带有初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.1)在第三节构造的包含两条疏散波的黎曼解. 则当$ \gamma\rightarrow1 $时, 真空状态出现, 且两条疏散波变成连接$ (\rho_{\pm}, \frac{\beta}{\alpha}+(u_{\pm}-\frac{\beta}{\alpha})e^{-\alpha t}) $和真空状态$ (\rho =0) $的两条接触间断, 其中真空状态是具有相同初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.4)的真空解.

  若$ \lim\limits_{\gamma\rightarrow 1}\rho_{\ast}=K\in(0, \min(\rho_{-}, \rho_{+})), $ 则在(4.25)中令$ \gamma\rightarrow 1 $取极限, 可得$ u_{+}=u_{-} $, 这与$ u_{-}<u_{+} $矛盾. 因此$ \lim\limits_{\gamma\rightarrow 1}\rho_{\ast}=0 $, 即当$ \gamma\rightarrow 1 $时, 出现真空状态. 进一步地, 由(4.23)和(4.24)式可得

$ \begin{equation} \lim\limits_{\gamma\rightarrow 1}v=u_{-} \, \, {\rm on}\, \, R_{1}, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \, \lim\limits_{\gamma\rightarrow 1}v=u_{+} \, \, {\rm on}\, \, R_{2}, \end{equation} $

$ \begin{equation} \left\{ \begin{array}{ll} { } \lambda_{1}=\frac{\beta}{\alpha}+(\frac{\gamma+1}{2}v-\frac{\gamma-1}{2} u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}-\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}, \\ { } \lambda_{2}=\frac{\beta}{\alpha}+(\frac{\gamma+1}{2}v-\frac{\gamma-1}{2} u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}+\frac{\gamma-1}{2}\rho_{-}^{\frac{\gamma-1}{2}}. \end{array} \right. \end{equation} $

从(4.26)和(4.27)式可得

$ \begin{equation} \lim\limits_{\gamma\rightarrow 1}\lambda_{1}= \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}, \, \, \, \, \, \, \, \, \, \, \, \lim\limits_{\gamma\rightarrow 1}\lambda_{2} =\frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}. \end{equation} $

证毕.

引理4.5  如果$ u_{+}>u_{-}+2. $

$ (1) $$ t<\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}) $时, 存在$ \gamma_{2}>0 $, 使得$ (\rho_{+}, u_{+})\in V(\rho_{-}, u_{-}) $$ 1<\gamma<1+\gamma_{2}; $

$ (2) $$ t>\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}) $时, 存在$ \gamma_{3}>0 $, 使得$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $$ 1<\gamma<1+\gamma_{3}; $

$ (3) $$ t=\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}) $, $ \rho_{-}<1 $$ \rho_{+}<1 $时, 对任意给定$ \gamma\in (1, 2), $$ (\rho_{+}, u_{+})\in V(\rho_{-}, u_{-}) $;

$ (4) $$ t=\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}), $$ \rho_{-}>1 $$ \rho_{+}>1 $时, 存在$ \gamma_{4}>0 $, 使得$ (\rho_{+}, u_{+})\in I(\rho_{-}, u_{-}) $$ 1<\gamma<1+\gamma_{4}. $

  若$ u_{+}>u_{-}+2 $$ (\rho_{+}, u_{+})\in V(\rho_{-}, u_{-}), $ 则结合(4.17)式及图 1

$ \begin{equation} u_{+}>u_{-}+(\rho_{-}^{\frac{\gamma-1}{2}}+\rho_{+}^{\frac{\gamma-1}{2}})e^{\alpha t}, \, \, \rho_{+}>0. \end{equation} $

(1) 当$ t<\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}) $时, 由于

因此, 存在足够小$ \gamma_{2}>0 $, 使得当$ 1<\gamma <1 +\gamma_{2} $时, 我们有不等式(4.29). 于是, 当$ 1<\gamma <1 +\gamma_{2} $时, 显然$ (\rho_{+}, u_{+})\in V( \rho_{-}, u_{-}) $.

(2) 当$ t>\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}) $时, 由于

因此, 存在足够小$ \gamma_{3}>0 $, 使得当$ 1<\gamma <1 +\gamma_{3} $时, 我们有不等式(4.21). 于是, 当$ 1<\gamma <1 +\gamma_{3} $时, 显然$ (\rho_{+}, u_{+})\in I( \rho_{-}, u_{-}) $.

(3) 当$ t=\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}), $$ \rho_{-}<1 $$ \rho_{+}<1 $时, 由于

因此, 当$ 1<\gamma <2 $时, 显然$ (\rho_{+}, u_{+})\in V( \rho_{-}, u_{-}) $.

(4) 当$ t=\frac{1}{\alpha}\ln(\frac{u_{+}-u_{-}}{2}), $$ \rho_{-}>1 $$ \rho_{+}>1 $时, 由于

因此, 存在足够小$ \gamma_{4}>0 $, 使得当$ 1<\gamma <1 +\gamma_{4} $时, 我们有不等式(4.21). 于是, 当$ 1<\gamma <1 +\gamma_{4} $时, 很明显$ (\rho_{+}, u_{+})\in I( \rho_{-}, u_{-}) $. 证毕.

$ u_{+}>u_{-}+2 $时, 为了简单和不失一般性, 我们可以假设对于任意给定的$ \gamma\in (1, 1+\gamma_{2}), $$ (\rho_{+}, u_{+})\in V( \rho_{-}, u_{-}) $. 那么, 具有黎曼初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.1) 的黎曼解是两个稀疏波$ R_{1} $, $ R_{2} $和它们之间的中间真空状态(用vac表示), 可以表示为

$ \begin{equation} (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t})+R_{1}+\mathrm {Vac}+R_{2}+(\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), \end{equation} $

其中$ R_{1} $$ R_{2} $

$ \begin{equation} R_{1}:\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{1}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}-\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v+\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, 0\leq\rho\leq\rho_{-} \end{array} \right. \end{equation} $

$ \begin{equation} R_{2}:\, \, \left\{\begin{array}{ll} { } \frac{{\rm d}x}{{\rm d}t}=\lambda_{2}^{\gamma}=\frac{\beta}{\alpha}+(v-\frac{\beta}{\alpha})e^{-\alpha t}+\frac{\gamma-1}{2}\rho^{\frac{\gamma-1}{2}}, \\ { } v-\rho^{\frac{\gamma-1}{2}}e^{\alpha t}=u_{+}-\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, 0\leq\rho\leq\rho_{+};\end{array} \right. \end{equation} $

两个稀疏波之间的真空状态为$ (0, \frac{\beta}{\alpha}+(v_{\ast}-\frac{\beta}{\alpha})e^{-\alpha t}) $$ v_{1}\leq v_{\ast}\leq v_{2} $, 其中

$ \begin{equation} v_{1}=u_{-}+\rho_{-}^{\frac{\gamma-1}{2}}e^{\alpha t}, \, \, \, \, \, \, \, \, \, \, \, \, v_{2}=u_{+}-\rho_{+}^{\frac{\gamma-1}{2}}e^{\alpha t}. \end{equation} $

我们可以看到真空状态出现在黎曼解中, 位于两个稀疏波之间.

从(4.31)式中, 很容易看到

这意味着当$ \gamma\rightarrow 1 $时, 稀疏波$ R_{1} $将退化为接触不连续$ J_{1} $, 其中$ J_{1} $的位置由

给出, 且$ J_{1} $的传播速度为

类似地, 从(4.32)式, 很容易看到

这意味着当$ \gamma\rightarrow 1 $时, 稀疏波$ R_{2} $将退化为接触不连续$ J_{2} $, 其中$ J_{2} $的位置由

给出, 而$ J_{2} $的传播速度为

因此, 我们证明了以下结果.

定理4.3  如果$ u_{+}>u_{-}+2 $$ (\rho_{+}, u_{+})\in V( \rho_{-}, u_{-}) $. 对任意给定的$ \gamma\in (1, 1+\gamma_{2}), $ 假设$ (\rho_{\gamma}, u_{\gamma})(t, x) $是带有黎曼初值$ (\rho_{\pm}, u_{\pm}) $的系统(1.1)在第3节构造的包含两条稀疏波的黎曼解. 则当$ \gamma\rightarrow1 $时, 黎曼解趋于两条连接状态$ (\rho_{\pm}, \frac{\beta}{\alpha}+(u_{\pm}-\frac{\beta}{\alpha})e^{-\alpha t}) $的接触间断且两条接触间断之间的真空状态, 具体为

$ \begin{align} \lim\limits_{ \gamma\rightarrow 1}(\rho_{\gamma}, u_{\gamma})(t, x)=\left\{\begin{array}{ll} { } (\rho_{-}, \frac{\beta}{\alpha}+(u_{-}-\frac{\beta}{\alpha})e^{-\alpha t}), \, \, \, \, x< \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } {\rm Vac}, \, \, \, \, \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{-}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}) \leq x \leq \frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \\ { } (\rho_{+}, \frac{\beta}{\alpha}+(u_{+}-\frac{\beta}{\alpha})e^{-\alpha t}), \, \, \, \, x>\frac{\beta}{\alpha}t+\frac{1}{\alpha}(u_{+}-\frac{\beta}{\alpha})(1 -e^{-\alpha t}), \end{array} \right. \end{align} $

这正是具有相同黎曼初值$ (\rho_{\pm}, u_{\pm}) $的无压气体动力学系统(1.4)的黎曼解.

5 结论

本文给出并研究了具有复合源项且绝热指数趋于1的可压缩流体流动欧拉方程的黎曼解的极限. 源可以表示类库仑摩擦或阻尼, 或两者兼有. 我们发现系统(1.1)中的复合源项具有使特征曲线为抛物曲线的能力, 使系统(1.1)的黎曼解中的所有稀疏波和激波都是弯曲的. 因此, 系统(1.1)的黎曼解不再是自相似的. 同样, 在复合源项的影响下, 系统(1.4) 的黎曼解中的$ \delta $ - 激波也弯曲成$ (t, x) $平面上的抛物线. 具体来说, 作者遵循一条很好的研究路径: 通过变量代换去除源(以引入非自相似解为代价), 并在包括压力的情况下将解与无压系统的解进行比较. 当前的结果从物理角度揭示了出现的$ \delta $ -激波可以用来解释质量集中现象.

参考文献

Chaplygin S .

On gas jets

Sci Mem Moscow Univ Math Phys, 1904, 21: 1- 121

[本文引用: 1]

Tsien H S .

Two dimensional subsonic flow of compressible fluids

J Aeron Sci, 1939, 6: 399- 407

DOI:10.2514/8.916      [本文引用: 1]

Von Karman T .

Compressibility effects in aerodynamics

J Aeron Sci, 1941, 8: 337- 356

DOI:10.2514/8.10737      [本文引用: 1]

Shen C .

The Riemann problem for the pressureless Euler system with the Coulomb-like friction term

IMA J Appl Math, 2016, 81: 76- 99

[本文引用: 2]

Faccanoni G , Mangeney A .

Exact solution for granular flows

Int J Numer Anal Meth Geomech, 2012, 37: 1408- 1433

[本文引用: 2]

Earnshaw S .

On the mathematical theory of sound

Philos Trans, 1858, 150: 1150- 1154

[本文引用: 2]

Lu Y G .

Existence of global entropy solutions to a nonstrictly hyperbolic system

Arch Ration Mech Anal, 2005, 178: 287- 299

DOI:10.1007/s00205-005-0379-0      [本文引用: 2]

Cheng Z .

Global entropy solutions to a variant of the compressible Euler equations

Appl Math Lett, 2008, 21: 410- 415

DOI:10.1016/j.aml.2007.03.022      [本文引用: 1]

Klainerman S , Majda A .

Singular limits of quasilinear hyperbolic systems with large parameters and the incompressible limit of compressible fluids

Comm Pure Appl Math, 1981, 34: 481- 524

DOI:10.1002/cpa.3160340405      [本文引用: 2]

Oelschläger K .

On the connection between Hamiltonian many-particle systems and the hydrodynamical equation

Arch Ration Mech Anal, 1991, 115: 297- 310

DOI:10.1007/BF00375277      [本文引用: 1]

Oelschläger K .

An integro-differential equation modelling a Newtonian dynamics and its scaling limit

Arch Ration Mech Anal, 1997, 137: 99- 134

DOI:10.1007/s002050050024      [本文引用: 1]

Caprino S , Esposito R , Marra R , Pulvirenti M .

Hydrodynamic limits of the Vlasov equation

Comm Partial Differential Equations, 1993, 18: 805- 820

DOI:10.1080/03605309308820951      [本文引用: 1]

Cheng H , Yang H , Zhang Y .

Riemann problem for the Chaplygin Euler equations of compressible fluid flow

Int J Nonlinear Sci Numer Simul, 2010, 11: 985- 992

[本文引用: 1]

DiPerna R J .

Global solutions to a class of nonlinear hyperbolic systems of equations

Comm Pure Appl Math, 1973, 26: 1- 28

DOI:10.1002/cpa.3160260102      [本文引用: 2]

Li C Z .

Existence in the large for certain systems of quasi linear hyperbolic conservation laws

J Differential Equations, 1982, 45: 85- 102

DOI:10.1016/0022-0396(82)90056-0      [本文引用: 1]

Pang Y , Zhang Y , Wen Y .

Delta shock wave to the compressible fluid flow with the generalized Chaplygin gas

Int J Nonlinear Mech, 2018, 99: 311- 318

DOI:10.1016/j.ijnonlinmec.2017.12.014     

Sarrico C O R , Paiva A .

The multiplication of distributions in the study of a Riemann problem in fluid dynamics

J Nonlinear Math Phys, 2017, 24: 328- 345

[本文引用: 1]

Song G , Xiao J .

Existence of global weak solutions to a special system of Euler equation with a source (Ⅱ): General case

J Math Anal Appl, 2009, 352: 943- 953

DOI:10.1016/j.jmaa.2008.11.051      [本文引用: 1]

Sheng S , Shao Z .

Concentration of mass in the pressureless limit of the Euler equations of one-dimensional compressible fluid flow

Nonlinear Analysis: Real World Applications, 2020, 52: 103039

DOI:10.1016/j.nonrwa.2019.103039     

Sheng S , Shao Z .

The limits of Riemann solutions to Euler equations of compressible fluid flow with a source term

Journal of Engineering Mathematics, 2020, 125: 1- 22

DOI:10.1007/s10665-020-10066-3      [本文引用: 1]

Savage S B , Hutter K .

The motion of a finite mass of granular material down a rough incline

J Fluid Mech, 1989, 199: 177- 215

DOI:10.1017/S0022112089000340      [本文引用: 1]

Shen C .

The Riemann problem for the Chaplygin gas equations with a source term

Z Angew Math Mech, 2016, 96: 681- 695

DOI:10.1002/zamm.201500015      [本文引用: 1]

Sun M .

The exact Riemann solutions to the generalized Chaplygin gas equations with friction

Commun Nonlinear Sci Numer Simul, 2016, 36: 342- 353

DOI:10.1016/j.cnsns.2015.12.013      [本文引用: 1]

Daw D A E , Nedeljkov M .

Shadow waves for pressureless gas balance laws

Appl Math Lett, 2016, 57: 54- 59

DOI:10.1016/j.aml.2016.01.004     

Edwards C M , Howison S D , Ockendon H , Ockendon J R .

Non-classical shallow water flows

IMA J Appl Math, 2008, 73: 137- 157

[本文引用: 1]

Guo L , Li T , Yin G .

The limit behavior of the Riemann solutions to the generalized Chaplygin gas equations with a source term

J Math Anal Appl, 2017, 455: 127- 140

DOI:10.1016/j.jmaa.2017.05.048     

Guo L , Li T , Yin G .

The vanishing pressure limits of Riemann solutions to the Chaplygin gas equations with a source term

Commun Pure Appl Anal, 2017, 16: 295- 309

DOI:10.3934/cpaa.2017014     

Zhang Y , Zhang Y .

Riemann problems for a class of coupled hyperbolic systems of conservation laws with a source term

Commun Pure Appl Anal, 2019, 18: 1523- 1545

DOI:10.3934/cpaa.2019073     

Zhang Y , Zhang Y .

The Riemann problem for the Suliciu relaxation system with the double-coefficient Coulomb-like friction terms

Int J Non-Linear Mech, 2019, 116: 200- 210

DOI:10.1016/j.ijnonlinmec.2019.07.004     

Zhang Y , Zhang Y , Wang J .

Zero-exponent limit to the extended Chaplygin gas equations with friction

Bull Malays Math Sci Soc, 2021, 44: 3571- 3599

DOI:10.1007/s40840-021-01133-8     

Zhang Q .

Concentration in the flux approximation limit of Riemann solutions to the extended Chaplygin gas equations with friction

J Math Phys, 2019, 60: 101508

DOI:10.1063/1.5085233     

Yin G , Chen J .

Existence and stability of Riemann solution to the Aw-Rascle model with friction

Indian J Pure Appl Math, 2018, 49: 671- 688

DOI:10.1007/s13226-018-0294-3     

Sheng S , Shao Z .

The vanishing adiabatic exponent limits of Riemann solutions to the isentropic Euler equations for power law with a Coulomb-like friction term

J Math Phys, 2019, 60: 101504

DOI:10.1063/1.5108863      [本文引用: 1]

Bourgault Y , Habashi W G , Dompierre J , Baruzzi G S .

A finite element method study of Eulerian droplets impingement models

Internat. J Numer Methods Fluids, 1999, 29: 429- 449

DOI:10.1002/(SICI)1097-0363(19990228)29:4<429::AID-FLD795>3.0.CO;2-F      [本文引用: 1]

Boutanios Z, Bourgault Y, Cober S, et al. 3D droplets impingement analysis around an aircraft's nose and cockpit using FENSAP-ICE[C]//36th AIAA Aerospace Science Meeting and Exhibit, Reno, NV, USA, 1998. DOI: 10.2514/6.1998-200

Bourgault Y, Thiriet M. Critical aspects of flow and aerosol simulations in the airway tract[C]//European Conference on Computational Fluid Dynamics-ECCOMAS CFD 2008, Venice, Italy, 2008

Keita S , Bourgault Y .

Eulerian droplet model: Delta-shocks and solution of the Riemann problem

J Math Anal Appl, 2019, 472: 1001- 1027

DOI:10.1016/j.jmaa.2018.11.061     

Bourgault Y. Computing gas-particle flows in airways with an Eulerian model[C]//European Conference on Computational Fluid Dynamics - ECCOMAS CFD 2006, Egmond ann Zee, The Netherlands, 2006

Keita S , Bourgault Y .

Eulerian models with particle pressure for air-particle flows

European Journal of Mechanics-/B Fluids, 2019, 78: 263- 275

DOI:10.1016/j.euromechflu.2019.08.001      [本文引用: 1]

Chen G Q , Liu H .

Formation of δ -shocks and vacuum states in the vanishing pressure limit of solutions to the Euler equations for isentropic fluids

SIAM J Math Anal, 2003, 34: 925- 938

DOI:10.1137/S0036141001399350      [本文引用: 2]

Chen G Q , Liu H .

Concentration and cavitation in the vanishing pressure limit of solutions to the Euler equations for nonisentropic fluids

Phys D, 2004, 189: 141- 165

DOI:10.1016/j.physd.2003.09.039      [本文引用: 1]

Li J .

Note on the compressible Euler equations with zero temperature

Appl Math Lett, 2001, 14: 519- 523

DOI:10.1016/S0893-9659(00)00187-7      [本文引用: 1]

Ibrahim M , Liu F , Liu S .

Concentration of mass in the pressureless limit of Euler equations for power law

Nonlinear Anal Real World Appl, 2019, 47: 224- 235

DOI:10.1016/j.nonrwa.2018.10.015      [本文引用: 2]

Li H , Shao Z .

Delta shocks and vacuum states in vanishing pressure limits of solutions to the relativistic Euler equations for generalized Chaplygin gas

Commun Pure Appl Anal, 2016, 15: 2373- 2400

DOI:10.3934/cpaa.2016041      [本文引用: 1]

Shen C .

The limits of Riemann solutions to the isentropic magnetogasdynamics

Appl Math Lett, 2011, 24: 1124- 1129

DOI:10.1016/j.aml.2011.01.038     

Shen C , Sun M .

Formation of delta shocks and vacuum states in the vanishing pressure limit of Riemann solutions to the perturbed Aw-Rascle model

J Differential Equations, 2010, 249: 3024- 3051

DOI:10.1016/j.jde.2010.09.004     

Shen C , Sun M , Wang Z .

Limit relations for three simple hyperbolic systems of conservation laws

Math Methods Appl Sci, 2010, 33: 1317- 1330

DOI:10.1002/mma.1248     

Sun M .

Concentration and cavitation phenomena of Riemann solutions for the isentropic Euler system with the logarithmic equation of state

Nonlinear Anal Real World Appl, 2020, 53: 103068

DOI:10.1016/j.nonrwa.2019.103068     

Tong M , Shen C .

The limits of Riemann solutions for the isentropic Euler system with extended Chaplygin gas

Applicable Analysis, 2019, 98: 2668- 2687

DOI:10.1080/00036811.2018.1469009     

Sheng W , Wang G , Yin G .

Delta wave and vacuum state for generalized Chaplygin gas dynamics system as pressure vanishes

Nonlinear Anal Real World Appl, 2015, 22: 115- 128

DOI:10.1016/j.nonrwa.2014.08.007     

Sun M .

The limits of Riemann solutions to the simplified pressureless Euler system with flux approximation

Math Methods Appl Sci, 2018, 41: 4528- 4548

DOI:10.1002/mma.4912     

Yang H , Wang J .

Delta shocks and vacuum states in vanishing pressure limits of solutions to the isentropic Euler equations for modified Chaplygin gas

J Math Anal Appl, 2014, 413: 800- 820

DOI:10.1016/j.jmaa.2013.12.025     

Yang H , Zhang Y .

Flux approximation to the isentropic relativistic Euler equations

Nonlinear Anal, 2016, 133: 200- 227

DOI:10.1016/j.na.2015.12.002     

Yin G , Sheng W .

Delta shocks and vacuum states in vanishing pressure limits of solutions to the relativistic Euler equations for polytropic gases

J Math Anal Appl, 2009, 355: 594- 605

DOI:10.1016/j.jmaa.2009.01.075     

Yin G , Sheng W .

Delta shocks and vacuum states in vanishing pressure limits of solutions to the relativistic Euler equations

Chin Ann Math, 2008, 29B: 611- 622

Yin G , Sheng W .

Delta wave formation and vacuum state in vanishing pressure limit for system of conservation laws to relativistic fluid dynamics

Z Angew Math Mech, 2015, 95: 49- 65

DOI:10.1002/zamm.201200148     

Zhang Y , Yang H .

Flux-approximation limits of solutions to the relativistic Euler equations for polytropic gas

J Math Anal Appl, 2016, 435: 1160- 1182

DOI:10.1016/j.jmaa.2015.11.012     

Zhang Y , Zhang Y .

Delta-shocks and vacuums in the relativistic Euler equations for isothermal fluids with the flux approximation

J Math Phys, 2019, 60: 011508

[本文引用: 1]

Sheng W, Zhang T. The Riemann Problem for the Transportation Equations in Gas Dynamics. Providence, RI: Amer Math Soc, 1999

[本文引用: 1]

Joseph K T .

A Riemann problem whose viscosity solutions contain delta measures

Asymptot Anal, 1993, 7: 105- 120

[本文引用: 1]

/