🚀 go-pugleaf

I'm trying to do a calculation of the full compressible Navier Stokes
equations for flow past an axisymmetric blunt body.
So, I put the equations into axisymmetric form and multiply
through by 1/Jacobian = r = distance from centerline (of axisymmetry).
I need to do this to put the equations back into conservative form for 
my method.
The obvious problem is that the equations are no longer valid along
the axis of symmetry (where r = 0). The Jacobian here is infinite.
I'm not quite sure what the implications of an infinite Jacobian are,
but from looking at my equations it becomes obvious that they cannot 
be used to solve the flowfield at r=0.

I was wondering what sorts of methods ae in common practice to solve
for the flowfield near r=0.
The solution is important here since this is where the flow passes
through the strongest portion of the shockwave. (It's a normal shock
at r=0 and the heating is high, also this is the where the stagnation
streamline lies.)
So, I need a good method, this might be the most important part of the flow!

Any help is greatly appreciated.

(I've thought about bounding the domain of the grid away from r=0,
 by 1/2 of a grid cell, but what boundry condition would I use?)

Ron Fedkiw

--
Ron Fedkiw (rfedkiw@redwood.math.ucla.edu)

A plan is made by someone who is sitting and thinking ...
while others are doing.