从k-\epsilon到k-\omega

写在前头

写这一块的初衷很简单，就是在给FLUENT k-\omega模型内嵌湍动能生成项的时候不知道 /omega 方程内浮升力生成项对\omega的影响。因为即便对于\epsilon，浮升力的作用机理也是不完全清楚的，对\omega方程而言，只能暂时参考\epsilon方程浮升力生项的作用机理。
因为柯神、Wilcox、Menter这些神仙对\omega的定义不太一致，加上又有各种修正，导致k-w方程的变种太多。目前主流的CODE （e.g. openfoam，star-ccm+，comsol，fluent）基本都涵盖std和sst两类（Wilcox…Menter…），这里分两部分推导sd和sst中浮升力项。因为各两方程的k方程基本是一样的，所以关注的重点在\omega方程中各生成项和耗散项。
如果考虑浮升力作用，说明物理问题已经不属于不可压缩范畴，在高温压情况下，粘度、比热、导热系数这些物性也不可当做常务性处理。如果考虑密度脉动，Favre averaging的雷诺应力关系又不明确。所以这里考虑低温压、弱变物性的情况。因为给的是推导过程，所以过程难免冗余，见谅。

推导过程

1. Wilcox中对比耗散率的定义
   $$\epsilon ={\beta }^{\ast }\omega k$$

2. 物质导数定义
   $$\frac{D(\rho \epsilon )}{Dt}={\beta }^{\ast }\omega \frac{D(\rho k)}{Dt}+{\beta }^{\ast }k\frac{D(\rho \omega )}{Dt}$$
   这里${\beta }^{\ast }$暂时认为是常数，这符合STD模型特征，在SST模型中，${\beta }^{\ast }$被定义为当地雷诺数的函数，不能轻易分离出来。

3. 由式1和2可以得到\omega的全导数公式
   $$\frac{D(\rho \omega )}{Dt}=\frac{1}{{\beta }^{\ast }k}\frac{D(\rho \epsilon )}{Dt}-\frac{\omega }{k}\frac{D(\rho k)}{Dt}$$

4. k, \epsilon方程
   $$\frac{D(\rho k)}{Dt}=D_k^{\mu }+D_k^T+{\Pi }_k+G_k+P_k-{\Phi }_k$$ $$\frac{D(\rho \epsilon )}{Dt}=D_{\epsilon }^{\mu }+D_{\epsilon }^T+{\Pi }_{\epsilon }+G_{\epsilon }+P_{\epsilon }-{\Phi }_{\epsilon }$$
   前式和后式等号右边各项分别表示湍动能（耗散率）粘性扩散项、湍流扩散项、压力扩散项、浮升力生成项、湍动能（耗散率）生成项、耗散项。

5. STD k-\epsilon方程
   $$\frac{\partial }{\partial t}(\rho k)+\frac{\partial }{\partial x_i}(\rho k{u}_i)=\frac{\partial }{\partial x_j}(\mu \frac{\partial k}{\partial x_j})+\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_k}\frac{\partial k}{\partial x_j})+P_k+G_k-\rho \epsilon$$$$\frac{\partial }{\partial t}(\rho \epsilon )+\frac{\partial }{\partial x_i}(\rho \epsilon u_i)=\frac{\partial }{\partial x_j}(\mu \frac{\partial \epsilon }{\partial x_j})+\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{\partial \epsilon }{\partial x_j})+\underset{P_{\epsilon }}{\underbrace{C_{1\epsilon }{P}_k}}+\underset{G_{\epsilon }}{\underbrace{C_{1\epsilon }{C}_{3\epsilon }{G}_k}}-C_{2\epsilon }\rho \frac{{\epsilon }^2}{k}$$

6. 公式3和4合并
   $$\frac{D(\rho \omega )}{Dt}=(\frac{1}{{\beta }^{\ast }k}{D}_{\epsilon }^{\mu }-\frac{\omega }{k}{D}_k^{\mu })+(\frac{1}{{\beta }^{\ast }k}{D}_{\epsilon }^T-\frac{\omega }{k}{D}_k^T)+(\frac{1}{{\beta }^{\ast }k}{\Pi }_{\epsilon }-\frac{\omega }{k}{\Pi }_k)+(\frac{1}{{\beta }^{\ast }k}{P}_{\epsilon }-\frac{\omega }{k}{P}_k)+(\frac{1}{{\beta }^{\ast }k}{G}_{\epsilon }-\frac{\omega }{k}{G}_k)-(\frac{1}{{\beta }^{\ast }k}{\Phi }_{\epsilon }-\frac{\omega }{k}{\Phi }_k)$$

7. 比耗散率生成项
   $$P_{\omega }=P_{\omega }=\frac{1}{{\beta }^{\ast }k}{P}_{\epsilon }-\frac{\omega }{k}{P}_k=C_{1\epsilon }\frac{\omega }{k}{P}_k-\frac{\omega }{k}{P}_k=(C_{1\epsilon }-1)\frac{\omega }{k}{P}_k$$

8. 粘性扩散项
   $$D_{\omega }^{\mu }=\frac{1}{{\beta }^{\ast }k}{D}_{\epsilon }^{\mu }-\frac{\omega }{k}{D}_k^{\mu }=\frac{1}{{\beta }^{\ast }k}\frac{\partial }{\partial x_j}[\mu \frac{\partial {\beta }^{\ast }\omega k}{\partial x_j}]-\frac{\omega }{k}\frac{\partial }{\partial x_j}[\mu \frac{\partial k}{\partial x_j}]$$$$=\frac{1}{k}\frac{\partial }{\partial x_j}[\mu k\frac{\partial \omega }{\partial x_j}+\mu \omega \frac{\partial k}{\partial x_j}]-\frac{\omega }{k}\frac{\partial }{\partial x_j}[\mu \frac{\partial k}{\partial x_j}]$$$$=\frac{1}{k}\frac{\partial }{\partial x_j}[\mu k\frac{\partial \omega }{\partial x_j}]+\frac{1}{k}\frac{\partial }{\partial x_j}[\mu \omega \frac{\partial k}{\partial x_j}]-\frac{\omega }{k}\frac{\partial }{\partial x_j}[\mu \frac{\partial k}{\partial x_j}]$$$$=\frac{\partial }{\partial x_j}(\mu \frac{\partial \omega }{\partial x_j})+\frac{1}{k}\mu \frac{\partial \omega }{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{\omega }{k}\frac{\partial }{\partial x_j}[\mu \frac{\partial k}{\partial x_j}]+\frac{1}{k}\mu \frac{\partial k}{\partial x_j}\frac{\partial \omega }{\partial x_j}-\frac{\omega }{k}\frac{\partial }{\partial x_j}[\mu \frac{\partial k}{\partial x_j}]$$$$=\frac{\partial }{\partial x_j}(\mu \frac{\partial \omega }{\partial x_j})+\frac{2}{k}\mu \frac{\partial \omega }{\partial x_j}\frac{\partial k}{\partial x_j}$$

9. 湍流粘性扩散项
   湍流扩散项和压力扩散项两项一般会通过类比Boussinesq EVMS的方法合并为一项，这里统称为湍流粘性扩散项
   $$D_{\omega }^{{\mu }_t}=D_{\omega }^T+{\Pi }_{\omega }=\frac{1}{{\beta }^{\ast }k}(D_{\epsilon }^T+{\Pi }_{\epsilon })-\frac{\omega }{k}(D_k^T+{\Pi }_k)$$$$=\frac{1}{{\beta }^{\ast }k}\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{\partial {\beta }^{\ast }\omega k}{\partial x_j})-\frac{\omega }{k}\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_k}\frac{\partial k}{\partial x_j})$$$$=\frac{1}{k}\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{\partial \omega k}{\partial x_j})-\frac{\omega }{k}\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_k}\frac{\partial k}{\partial x_j})$$$$=\frac{1}{k{\sigma }_{\epsilon }}\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial \omega k}{\partial x_j}+\frac{{\mu }_t}{k{\sigma }_{\epsilon }}\frac{{\partial }^2\omega k}{\partial x_j^2}-\frac{\omega }{k}\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_k}\frac{\partial k}{\partial x_j})$$$$=\frac{1}{{\sigma }_{\epsilon }}\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial \omega }{\partial x_j}+\frac{\omega }{k{\sigma }_{\epsilon }}\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{{\mu }_t}{k{\sigma }_{\epsilon }}\frac{\partial }{\partial x_j}(k\frac{\partial \omega }{\partial x_j}+\omega \frac{\partial k}{\partial x_j})-\frac{\omega }{{\sigma }_{k}k}({\mu }_t\frac{{\partial }^{2}k}{\partial x_j^2}+\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j})$$$$=\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial \omega }{\partial x_j}+\frac{\omega }{k{\sigma }_{\epsilon }}\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{{\partial }^2\omega }{\partial x_j^2}+\frac{{\mu }_t}{k{\sigma }_{\epsilon }}\frac{\partial k}{\partial x_j}\frac{\partial \omega }{\partial x_j}+\frac{{\mu }_t\omega }{k{\sigma }_{\epsilon }}\frac{{\partial }^{2}k}{\partial x_j^2}+\frac{{\mu }_t}{k{\sigma }_{\epsilon }}\frac{\partial \omega }{\partial x_j}\frac{\partial k}{\partial x_j}-\frac{\omega }{{\sigma }_{k}k}({\mu }_t\frac{{\partial }^{2}k}{\partial x_j^2}+\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j})$$$$=\frac{2{\mu }_t}{k{\sigma }_{\epsilon }}\frac{\partial \omega }{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{{\mu }_t\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{{\partial }^{2}k}{\partial x_j^2}+\frac{\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{1}{{\sigma }_{\epsilon }}\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial \omega }{\partial x_j}+\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{{\partial }^2\omega }{\partial x_j^2}$$$$=\frac{2{\mu }_t}{k{\sigma }_{\epsilon }}\frac{\partial \omega }{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{{\mu }_t\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{{\partial }^{2}k}{\partial x_j^2}+\frac{\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j}+\frac{\partial }{\partial x_j}(\frac{{\mu }_t}{{\sigma }_{\epsilon }}\frac{\partial \omega }{\partial x_j})$$输运方程内部出现湍流粘度的偏导数显然是不符合一般编程规范的，这里引入EVMS把湍流粘度去掉$$\frac{\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{\partial {\mu }_t}{\partial x_j}\frac{\partial k}{\partial x_j}=\frac{\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{\partial k}{\partial x_j}\frac{\partial }{\partial x_j}(\rho C_{\mu }\frac{k}{w})=C_{\mu }\frac{\omega }{k}(\frac{1}{{\sigma }_{\epsilon }}-\frac{1}{{\sigma }_k})\frac{\partial k}{\partial x_j}[\rho \frac{\partial }{\partial x_j}(\frac{k}{w})+\frac{k}{w}\frac{\partial \rho }{\partial x_j}]$$多出来了的密度项的偏导数！！有兴趣的可以继续把密度从偏导数内部分离出来。

10. 浮升力生成项
    $$G_{\omega }=\frac{1}{{\beta }^{\ast }k}{G}_{\epsilon }-\frac{\omega }{k}{G}_k$$$$=\frac{1}{{\beta }^{\ast }k}(C_{1\epsilon }{C}_{3\epsilon }\frac{{\beta }^{\ast }\omega k}{k}{G}_k)-\frac{\omega }{k}{G}_k$$$$=\frac{\omega }{k}(C_{1\epsilon }{C}_{3\epsilon }-1)G_k$$

11. 耗散项
    $${\Phi }_{\omega }=\frac{1}{{\beta }^{\ast }k}{\Phi }_{\epsilon }-\frac{\omega }{k}{\Phi }_k=\frac{1}{{\beta }^{\ast }k}{C}_{2\epsilon }\frac{{\epsilon }^2}{k}-\frac{\omega }{k}\epsilon =(C_{2\epsilon }-1){\beta }^{\ast }k{\omega }^2$$

12. 合并以上各项可得到