On the use of a coordinate transformation for the solution of the Navier-Stokes equations