So far, however, the hole-cutting process has played no active part in the actual folding process - that is, we have so far assumed that rotation curves should be folded using all of the available data. This latter assumption is justified if the transition regions are as asymmetric about the dynamical centres as the exterior parts of rotation curves are assumed to be.
But, by the very nature of "dynamical transition regions'' in the generality of physical systems, the possibility exists that the dynamics in such regions is, to some extent, chaotic. In this latter case, the shapes of the interior parts of rotation curves could be then significantly disturbed from the asymmetry assumed for the exterior parts of rotation curves - and this would lead to systematic inaccuracies in the final solutions due entirely to the inclusion of the transition regions in the folding process. It would then follow that using only the exterior parts of rotation curves for the folding process will lead to more accurate folding.
Copyright The European Southern Observatory (ESO)