On Tue, Feb 14, 2012 at 11:29 AM, Leland M Barnard <[email protected]>wrote:

> Is it possible to specify how gradients in the solution variable are
> determined, e. g. by forward difference, backward difference or center
> difference?
>

I believe that is what the central difference and upwind schemes convection
terms currently do. The direction depends on the coefficient values.


> For example, is is possible to write a convection term of the form
> U*gradC that calculates gradC using a backwards difference, so that it
> depends only on the value of C in cell n and n-1, and not in cell n+1?
>

The upwind convection term does that.

Hope that helps.

-- 
Daniel Wheeler
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
  [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]

Reply via email to