3
$\begingroup$

I need to get a stability criterion for the numerical scheme for equation

$$\frac{\partial u}{\partial t}-\frac{\partial^2 u}{\partial x^2}-\frac{\partial^2 u}{\partial y^2}-\frac{\partial^2 u}{\partial z^2} -\alpha\frac{\partial u}{\partial x}-\beta\frac{\partial u}{\partial y}-\gamma\frac{\partial u}{\partial z}-f=0 $$

The scheme is explicit: $$ \frac{u^{n+1}_{i,j,k}-u^{n}_{i,j,k}}{\Delta t}=\frac{u^{n}_{i+1,j,k}-2u^{n}_{i,j,k}+u^{n}_{i-1,j,k}}{\Delta x^2}+\frac{u^{n}_{i,j+1,k}-2u^{n}_{i,j,k}+u^{n}_{i,j-1,k}}{\Delta y^2}+$$ $$+\frac{u^{n}_{i,j,k+1}-2u^{n}_{i,j,k}+u^{n}_{i,j,k-1}}{\Delta z^2}+\alpha\frac{u^{n}_{i+1,j,k}-u^{n}_{i-1,j,k}}{2\Delta x }+$$ $$+\beta\frac{u^{n}_{i,j+1,k}-u^{n}_{i,j-1,k}}{2\Delta y }+\gamma\frac{u^{n}_{i,j,k+1}-u^{n}_{i,j,k-1}}{2\Delta z }+f_{i,j,k}.$$ After that I found an expression with a help of von Neumann analysis for amplification coefficient:

$$\rho=1-4\Delta t\left(\frac{\sin^2{\theta /2}}{\Delta x^2}+\frac{\sin^2{\eta /2}}{\Delta y^2}+\frac{\sin^2{\zeta /2}}{\Delta z^2}\right)+i\Delta t\left(\alpha\frac{\sin{\theta}}{\Delta x}+\beta\frac{\sin{\eta}}{\Delta y}+\gamma\frac{\sin{\zeta }}{\Delta z}\right) $$

To find a condition for stability criterion I must find a condition for $\Delta t,\Delta x,\Delta y,\Delta z$ so that for all $\theta,\eta,\zeta$ an inequality $|\rho|<1$ is true. Here I stuck.Could you help me with this stuff?

$\endgroup$
  • $\begingroup$ What expression do you obtain when you substitute an arbitrary fourier mode? $\endgroup$ – Paul Mar 10 '14 at 17:21
  • $\begingroup$ @Paul (who asked the question) you should put your comment into the body of the question. Then look for conditions under which $|\rho|\le 1$. $\endgroup$ – David Ketcheson Mar 11 '14 at 4:18
3
$\begingroup$

Though your problem is in 3 dimensions, you can gain a lot of insight by considering the 1D case. Let's omit the y and z terms and concentrate on x and t. Then, your problem would be equivalent to the classic advection-diffusion equation with an explicit Forward-in-Time, Centered-in-Space (FTCS) finite difference method.

Knowing the name of the problem and the method, it's fairly easy to find derivations online. I personally found this stability analysis for the advection diffusion equation particularly insightful. Following this approach, you should be able to extrapolate a method for the 3D case.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.