home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!wupost!gumby!yale!yale.edu!ira.uka.de!chx400!urz.unibas.ch!solca1
- From: solca1@urz.unibas.ch
- Newsgroups: sci.math.num-analysis
- Subject: Non-Linear PDE,finite diff meth.
- Message-ID: <1992Jul30.182305.34205@urz.unibas.ch>
- Date: 30 Jul 92 18:23:05 MET
- Organization: University of Basel, Switzerland
- Lines: 38
-
- Hello everyboby
-
- I'm working on Reaction-Transport equation of the type :
-
- a + b <-> c with k1 -> , and k2 in <- direction as rate
- constant.
-
- The PDE I get as usual :
-
- dt(Ca(x,t))= D dx**2(Ca(x,t)) -k1Ca(x,t)Cb(x,t) + k2Cc(x,t)
-
- where D=diffusion const, dt=Dt (now D as partial Differential operator)
- dx**2 = Dxx (second partial derivative in x) .
- Ca(x,t) space-time concentration of specie A ,Cb the same for b .....
-
-
- Naturally you get the same pdes for Cb(x,t) and Cc(x,t) with little changes.
- I have a system of three non-linears PDEs which must solved simultanously .
-
- I have made a simulation on a Cellular Automata of this reaction-transport
- process and I'm interested in a numerical solution too. I have planed to use
- finite difference method to solve this system of PDE's .The problem is,
- must I first do a linearization of the system or can I simply use the usual
- algorithm for linear Pde.
- If, for instance, you have a linear pde , you use the implicit or the
- explicit scheme . Now what happend if your pde is non-linear ,is a $
- linearization obligatory or not ? In some books it is said that you can
- ungo this linearization, and in other the contrary is not said but
- nothing is said about ? If some body have some answers to my questions
- it would be glad of you to mail me .Thanks for any suggestion !!
-
- bye Jan .
-
-
-
-
-
-
-