[GRASS-user] Movement Cost: Euclidean results?

[David]

Dear GRASS-users,

I want to obtain a movement cost layer (r.walk) which only regards pure 
walking cost (without a friction layer). I've generated a 32 bit geotiff 
DEM (not too big) and a constant value = 1 grid as friction surface (for 
being neutral). Unfortunately GRASS only gives me a symmetrical result 
as there would be no relief.
Is is not possible to perform a simple movement-cost analysis without 
any optional friction maps? Using a constant value grid (or lamda = 0) 
doesn't work unfortunately. I'm always getting symmetrical/euclidean 
results.

Thank you!

Regards
David