[nooredin]

Hi there
I am writing a code for solving 2D euler equation in a channel with variable cross section and the method is explicit. My code goes well for no variation. but as soon as i give it some slope it goes mad.
i have used the transformation procees explaned in hoffman's book. i.e. using U*J and Fbar and Ebar for evolution term.
Is there anybody to help me or any guidlines somewhere?
The problem is that the explanation in the book does not go through details.
I think the fault section should be something with dx, dy, x, and y. but i could not figure it out myself.