2016A&A...585A..69G

The origin and evolution of the shapes of bipolar planetary nebulae are poorly understood. We postulate that their history can be traced through their internal velocity fields in a procedure similar to the one well established for spherical objects. Such an analysis requires 3D photoionization and kinematical modelling that is computationally very time consuming. We apply an axially symmetric pseudo-3D photoionization model, pyCloudy, to derive the structures of six bipolar nebulae and two suggested post-bipolars in an attempt to constrain the bipolar planetary nebulae evolution. HST images and VLT/UVES spectroscopy are used for the modelling. The targets are located in the direction of the Galactic bulge. A 3D model structure is used as input to the photoionization code in order to fit the HST images. Line profiles of different ions constrain the velocity field. The model and associated velocity fields allow us to derive masses, velocities, and ages. The 3D models find much lower ionized masses than required in 1D models: ionized masses are reduced by factors of 2-7. The selected bi-lobed planetary nebulae show a narrow range of ages: the averaged radii and velocities result in values between 1300 and 2000yr. The lobes are fitted well with velocities that increase linearly with radius. These Hubble-type flows have been found before, and suggest that the lobes form at a defined point in time. The lobes appear to be slightly younger, by ∼500yr, than the main (host) nebulae; they seem to form at an early phase of PN evolution and fade after 1-2kyr. We find that 30-35% of bulge PNe pass through a bipolar phase.