2014MNRAS.440L..16B

Some combination of binary interactions and accretion plausibly conspire to produce the ubiquitous collimated outflows from planetary nebulae (PN) and their presumed pre-PN (PPN) precursors. But which accretion engines are viable ? The difficulty in observationally resolving the engines warrants the pursuit of indirect constraints. We show how kinematic outflow data for 19 PPN can be used to determine the minimum required accretion rates. We consider main-sequence (MS) and white dwarf (WD) accretors and five example accretion rates inferred from published models to compare with the minima derived from outflow momentum conservation. While our primary goal is to show the method in anticipation of more data and better theoretical constraints, taking the present results at face value already rules out modes of accretion: Bondi-Hoyle-Lyttleton (BHL) wind accretion and wind Roche lobe overflow (M-WRLOF, based on Mira parameters) are too feeble for all 19/19 objects for an MS accretor. For a WD accretor, BHL is ruled out for 18/19 objects and M-WRLOF for 15/19 objects. RLOF from the primary at the Red Rectangle level can accommodate 7/19 objects, though RLOF modes with higher accretion rates are not yet ruled out. Accretion modes operating from within common envelope evolution can accommodate all 19 objects, if jet collimation can be maintained. Overall, sub-Eddington rates for an MS accretor are acceptable but 8/19 would require super-Eddington rates for a WD.