迎来QQ群吹逼,群号:538254037
3.1总览
使用雷诺分解,速度可以写为系综平均和脉动量相加的形式:
\[\underbrace{\mathbf{U}(\mathbf{x}, t)}_{\text {Instantaneous Velocity }}=\underbrace{\langle\mathbf{U}(\mathbf{x}, t)\rangle}_{\text {Mean Velocity }}+\underbrace{\mathbf{u}(\mathbf{x}, t)}_{\text {Fluctuating Velocity }} \]首先是连续方程:
\[\nabla \cdot\mathbf{U}(\mathbf{x}, t)=0 \]对上式整体取平均,然后根据上一篇博客中的随机场理论,梯度的平均等于平均值的梯度可得:
\[\langle \nabla \cdot\mathbf{U}(\mathbf{x}, t)\rangle = \nabla \cdot\langle\mathbf{U}(\mathbf{x}, t)\rangle=0 \]然后可得:
\[\begin{align} \nabla \cdot\langle\mathbf{U}(\mathbf{x}, t)\rangle&=0 \\ \nabla \cdot \mathbf{u}(\mathbf{x}, t)&=0 \end{align} \]根据物质导数或随流导数的定义可得:
\[\begin{align} \underbrace{\frac{D U_{j}}{D t}}_{\text {Material Derivative }}&=\underbrace{\frac{\partial U_{j}}{\partial t}}_{\text {Storage }}+\underbrace{\frac{\partial}{\partial x_{i}}\left(U_{i} U_{j}\right)}_{\text {Advection }} \\ \left\langle\frac{D U_{j}}{D t}\right\rangle&=\frac{\partial\left\langle U_{j}\right\rangle}{\partial t}+\frac{\partial}{\partial x_{i}}\left\langle U_{i} U_{j}\right\rangle . \end{align} \]但是 \(\left\langle U_{i} U_{j}\right\rangle\left[\mathrm{m}^{2} \mathrm{~s}^{-2}\right]\) 这一项未知,需要为这一项寻找表达式,展开上式后可得:
\[\begin{aligned} \left\langle U_{i} U_{j}\right\rangle & =\left\langle\left(\left\langle U_{i}\right\rangle+u_{i}\right)\left(\left\langle U_{j}\right\rangle+u_{j}\right)\right\rangle \\ & =\left\langle\left\langle U_{i}\right\rangle\left\langle U_{j}\right\rangle+u_{i}\left\langle U_{j}\right\rangle+u_{j}\left\langle U_{i}\right\rangle+u_{i} u_{j}\right\rangle \\ & =\left\langle U_{i}\right\rangle\left\langle U_{j}\right\rangle+\left\langle u_{i} u_{j}\right\rangle \end{aligned} \]在上式中,由于 \(u_{i}\) 和 \(u_{j}\) 的 PDF 均值为0,所以 \(u_{i}\left\langle U_{j}\right\rangle\) 和 \(u_{j}\left\langle U_{i}\right\rangle\) 也为 \(0\)。即:脉动量为正值或者负值的几率相当。
将展开后的带入物质导数,可得:
\[\begin{align} \left\langle\frac{D U_{j}}{D t}\right\rangle & =\frac{\partial\left\langle U_{j}\right\rangle}{\partial t}+\frac{\partial}{\partial x_{i}}\left(\left\langle U_{i}\right\rangle\left\langle U_{j}\right\rangle+\left\langle u_{i} u_{j}\right\rangle\right) \\ & =\frac{\partial\left\langle U_{j}\right\rangle}{\partial t}+\left\langle U_{i}\right\rangle \frac{\partial}{\partial x_{i}}\left\langle U_{j}\right\rangle+\frac{\partial}{\partial x_{i}}\left\langle u_{i} u_{j}\right\rangle \end{align} \]上式的第二个等号后面使用了 \(\partial\left\langle U_{i}\right\rangle / \partial x_{i}=0\) 条件。
使用
\[\frac{\bar{D}}{\bar{D} t} \equiv \frac{\partial}{\partial t}+\langle\mathbf{U}\rangle \cdot \nabla \]可以进一步简化上式。简化后的速度场的物质导数为:
\[\underbrace{\left\langle\frac{D U_{j}}{D t}\right\rangle}_{\text {Mean of Material Derivative }}=\underbrace{\frac{\bar{D}}{\bar{D} t}\left\langle U_{j}\right\rangle}_{\text {Mean Substantial Derivative of Mean }}+\underbrace{\frac{\partial}{\partial x_{i}}\left\langle u_{i} u_{j}\right\rangle}_{\text {Reynolds Stresses }} \]由上式可以看出,平均速度的平均物质导数与物质导数的平均是不一样的。由此可得平均动量方程:
\[\underbrace{\frac{\bar{D}\left\langle U_{j}\right\rangle}{\bar{D} t}}_{\text {Mean Substantial Derivative of Mean }}=\underbrace{\nu \nabla^{2}\left\langle U_{j}\right\rangle}_{\text {Surface Forces }}-\underbrace{\frac{\partial\left\langle u_{i} u_{j}\right\rangle}{\partial x_{i}}}_{\text {Reynolds Stresses }}-\underbrace{\frac{1}{\rho} \frac{\partial\langle p\rangle}{\partial x_{j}}}_{\text {Normal and Body Forces }} \]很多湍流模型是为了解决上式中的雷诺应力项,这一项也是湍流的封闭问题(closure problem)。
3.2 张量的性质
雷诺应力项 \(\left\langle u_{i} u_{j}\right\rangle\) 是一个二阶张量,具有对称性,即:\(\left\langle u_{i} u_{j}\right\rangle = \left\langle u_{j} u_{i}\right\rangle\)。这一张量的对角线(diagonal)元素 \(\left\langle u_{i} u_{i}\right\rangle\) 被称为正应力,而非对角线的元素被称为切应力(shear stress)。雷诺应力可被写为张量形式:
\[\left[\begin{array}{ccc} \left\langle u_{1}^{2}\right\rangle & \left\langle u_{1} u_{2}\right\rangle & \left\langle u_{1} u_{3}\right\rangle \\ \left\langle u_{2} u_{1}\right\rangle & \left\langle u_{2}^{2}\right\rangle & \left\langle u_{2} u_{3}\right\rangle \\ \left\langle u_{3} u_{1}\right\rangle & \left\langle u_{3} u_{2}\right\rangle & \left\langle u_{3}^{2}\right\rangle \end{array}\right] \]湍动能(turbulence kinetic energy):上述矩阵的迹(trace)
\[k \equiv \frac{1}{2}\langle\mathbf{u} \cdot \mathbf{u}\rangle=\frac{1}{2}\left\langle u_{i} u_{i}\right\rangle \]3.3 各向异性 (Anisotropy)
切应力和正应力之间的不同取决于坐标系的选择。比如,当坐标系旋转后,雷诺应力张量中的成分可能会发生变化。因此,有必要将雷诺应力写成各向同性 (Isotropic) 项和各项异性 (Anisotropy) 项。
\[\left\langle u_{i} u_{j}\right\rangle=\underbrace{\left\langle u_{i} u_{j}\right\rangle-\frac{2}{3} k \delta_{i j}}_{\text {Anisotropic Part }}+\underbrace{\frac{2}{3} k \delta_{i j}}_{\text {Isotropic Part }} \]各项异性项可被写为:\(a_{i j} \equiv\left\langle u_{i} u_{j}\right\rangle-\frac{2}{3} k \delta_{i j}\left[\mathrm{~m}^{2} \mathrm{~s}^{-2}\right]\)。一个重要的概念是:只有各项异性的项在湍流的输运中有效。因此可以将雷诺应力中的各项异性项和压力项写在一起:
\[\rho \frac{\partial\left\langle u_{i} u_{j}\right\rangle}{\partial x_{i}}+\frac{\partial\langle p\rangle}{\partial x_{j}}=\rho \frac{\partial a_{i j}}{\partial x_{i}}+\frac{\partial}{\partial x_{j}}\left(\langle p\rangle+\frac{2}{3} \rho k\right) \]各项同性项 \(\frac{2}{3} \rho k\left[\mathrm{kgm}^{-1} \mathrm{~s}^{-2}\right]\) 可以被压力项吸收。
3.4 平均剪切方程 (Mean Scalar Equation)
对被动标量 \(\phi(\mathbf{x},t)\) 使用雷诺分解:
\[\underbrace{\phi(\mathbf{x}, t)}_{\text {Instantaneous Scalar }}=\underbrace{\langle\phi(\mathbf{x}, t)\rangle}_{\text {Mean Scalar }}+\underbrace{\phi^{\prime}(\mathbf{x}, t)}_{\text {Fluctuating Scalar }} \]一个瞬时被动标量场 (instantaneous passive scalar field) 的控制方程为:
\[\underbrace{\frac{\partial \phi}{\partial t}}_{\text {Storage }}+\underbrace{\nabla \cdot(\mathbf{U} \phi)}_{\text {Advection }}=\underbrace{\Gamma \nabla^{2} \phi}_{\text {Diffusion }} \]唯一的非线性项 \(\mathbf{U}\phi\) 可写为:
\[\begin{aligned} \langle\mathbf{U} \phi\rangle & =\left\langle(\langle\mathbf{U}\rangle+\mathbf{u})\left(\langle\phi\rangle+\phi^{\prime}\right)\right\rangle \\ & =\langle\mathbf{U}\rangle\langle\phi\rangle+\left\langle\mathbf{u} \phi^{\prime}\right\rangle . \end{aligned} \]速度标量协方差 (velocity-scalar covariance) \(\langle\mathbf{u} \phi^{\prime}\rangle\) 被称为标量通量 (scalar flux)。它表示速度场的脉动引起的标量的通量:
\[\begin{align} \frac{\partial\langle\phi\rangle}{\partial t}+\nabla \cdot\left(\langle\mathbf{U}\rangle\langle\phi\rangle+\left\langle\mathbf{u} \phi^{\prime}\right\rangle\right)=\Gamma \nabla^{2}\langle\phi\rangle, \\ \frac{\bar{D}\langle\phi\rangle}{\bar{D} t}=\nabla \cdot\left(\Gamma \nabla\langle\phi\rangle-\left\langle\mathbf{u} \phi^{\prime}\right\rangle\right) . \end{align} \]这个公式也会带来新的封闭性问题,\(\langle\mathbf{u} \phi^{\prime}\rangle\) 这一项需要额外的建模或者参数化。
3.5 梯度扩散和湍流粘性假设
gradient-diffusion hypothesis or turbulent-viscosity hypothesis:
平均标量通量与平均标量梯度的负值成正比。
比例常数又被称为湍流扩散率 (turbulent diffusivity),其本身也是空间和时间的函数:\(\Gamma_{T}(\mathbf{x}, t)\left[\mathrm{m}^{2} \mathrm{~s}^{-1}\right]\):
\[\left\langle\mathbf{u} \phi^{\prime}\right\rangle=-\Gamma_{T} \nabla\langle\phi\rangle \]下标 \(T\) 代表湍流,\(\Gamma_{T}\) 为湍流扩散率,不要和分子扩散率 \(\Gamma\) 混淆。
此时可以将分子扩散率和湍流扩散率结合为有效扩散率 (effective diffusivity):
\[\underbrace{\Gamma_{\text {eff }}(\mathbf{x}, t)}_{\text {Effective Diffusivity }}=\underbrace{\Gamma}_{\text {Molecular Diffusivity }}+\underbrace{\Gamma_{T}(\mathbf{x}, t)}_{\text {Turbulent Diffusivity }} \]在这种模型下(注:这里是基于上述假设的建模,并不一定是真正的情况),公式可以用有效扩散率简化为:
\[\underbrace{\frac{\bar{D}\langle\phi\rangle}{\bar{D} t}}_{\text {Mean Substantial Derivative of Mean }}=\underbrace{\nabla \cdot\left(\Gamma_{\mathrm{eff}} \nabla\langle\phi\rangle\right)}_{\text {Diffusion of Mean }} \]对于平均动量方程,梯度扩散假设 (gradient-diffusion hypothesis) 相对更为困难。此时更应该基于各向异性 (anisotropic) 部分建模,建模时可以根据平均拉伸率:
\[\begin{align} \left\langle u_{i} u_{j}\right\rangle-\frac{2}{3} k \delta_{i j} & =-v_{T}\left(\frac{\partial\left\langle U_{i}\right\rangle}{\partial x_{j}}+\frac{\partial\left\langle U_{j}\right\rangle}{\partial x_{i}}\right) \\ & =-2 v_{T} \bar{S}_{i j} \end{align} \]\(\nu_{T}\): turbulent viscosity or eddy viscosity
此时方程可写为:
其中:
\[\underbrace{\nu_{\text {eff }}(\mathbf{x}, t)}_{\text {Effective Viscosity }}=\underbrace{\nu}_{\text {Molecular Viscosity }}+\underbrace{\nu_{T}(\mathbf{x}, t)}_{\text {Turbulent Viscosity }} \]为有效粘性。
\(\langle p\rangle+\frac{2}{3} \rho k\): modified pressure.
同分子扩散率一样,湍流扩散率也可以被无量纲化。
Turbulent Prandtl number:
Turbulent Schmidt number:
\[S c_{T}=\frac{\nu_{T}}{\Gamma_{T}} \] 标签:方程,langle,partial,frac,right,rangle,平均,left From: https://www.cnblogs.com/xubonan/p/18125815