hippke
commented
3 years ago Thanks for the explanation. Let me rephrase: I want to assure that the more massive body orbits the less massive body. Or, in the case of a binary ensemble where both are similar in mass: That the Kepler ellipses are physically correct around the joint barycenter.
In the current code, it is then required to define the more massive body as the "planet" and the less massive body is the "moon". That's because the barycentric correction is defined that way. And to keep is physically correct, I need to set M_planet>M_moon.
Of course, we can still explore situations where the "moon" is larger in radius than the "planet". Both radii are free parameters and can be explored by the user and the sampler.
For a moon to be a moon, the planet-moon barycenter must be inside the planet. This is the case for the Earth-Moon system. If the barycenter is outside the primary, then the system should better be referred to as a double planet (see e.g. Stern & Levison 2002, DOI:10.1017/S1539299600013289). This is the case in the Pluto-Charon system.
So the actual status of the secondary (moon or double planet) depends on the radius of the primary and on the mutual distance between the two bodies in question.
As a consequence, there is more to the definition of a moon than just M_planet > M_moon. What is more, even if M_planet < M_moon, then one could simply switch the roles of the planet and the moon and the physics (and simulations) should still be correct.
I would suggest to drop this mass constraint in Pandora if it is not strictly necessary from a geometrical or simulation point of view.