2
$\begingroup$

I need to build a system in Simulink that solves a PDE, but I can't find any literature or books where it is described how to do it (especially any stuff according to modeling PDE in Simulink).

Please give some advice where I can find such literature.

Thanks beforehand!

$\endgroup$
7
  • $\begingroup$ Hi and welcome to SciComp. Could you give some more details on the problem itself? What PDE? What application? $\endgroup$
    – GertVdE
    Commented Jun 30, 2013 at 19:13
  • $\begingroup$ I need to make a model in Simulink, this model must solve an equation $$\frac{\partial u(x,t)}{\partial t}-\operatorname{div}\left(A(x)\nabla u(x,t)\right)=f(x,t),\text{where } A(x) \text{ is a matrix 2x2} $$ - general equation of diffusion in $(x_{1},x_{2},t)$. I have an experience in building models of ODEs in Simulink, but I have any idea what to do in this case. So that I'm trying to find someone who can help me. $\endgroup$
    – cool
    Commented Jun 30, 2013 at 20:48
  • $\begingroup$ AFAIK, Simulink is inherently focused on ODEs, and I know of now possibility to directly simulate PDEs. You may manually discretize the model spatially, although that's cumbersome in Simulink. $\endgroup$ Commented Jul 2, 2013 at 10:35
  • $\begingroup$ @FlorianBrucker, although it's cumbersome, but I would like to try it. Could you give me the source, where you saw the method how to do it? $\endgroup$
    – cool
    Commented Jul 2, 2013 at 12:25
  • $\begingroup$ I don't have a source or an example. Discretize your space dimensions using method of lines or finite differences. This yields a system of ODEs which you can implement in Simulink. An application that I know is simulating heat flow in a rod, where the continuous rod is discretized into a low number (~10) of 0d nodes. Note that such things are much easier in Modelica, because it allows you to vary the number of nodes via a parameter. $\endgroup$ Commented Jul 2, 2013 at 15:43

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.