Formation of Co-orbital satellites

Coorbital Satellites of Saturn: Congenital Formation
http://arxiv.org/abs/1002.4617

Abstract wrote:

Saturn is the only known planet to have coorbital satellite systems. In the present work we studied the process of mass accretion as a possible mechanism for coorbital satellites formation. The system considered is composed of Saturn, a proto-satellite and a cloud of planetesimals distributed in the coorbital region around a triangular Lagrangian point. The adopted relative mass for the proto-satellite was 10^-6 of Saturn's mass and for each planetesimal of the cloud three cases of relative mass were considered, 10^-14, 10^-13 and 10^-12 masses of Saturn. In the simulations each cloud of planetesimal was composed of 10^3, 5 x 10^3 or 10^4 planetesimals. The results of the simulations show the formation of coorbital satellites with relative masses of the same order of those found in the saturnian system (10^-13 - 10^-9). Most of them present horseshoe type orbits, but a significant part is in tadpole orbit around L_4 or L_5. Therefore, the results indicate that this is a plausible mechanism for the formation of coorbital satellites.

Maybe we could scalate it to planetary bodies around stellar masses. Coorbital Earths

Probably not, this tends to be unstable due to the large masses of the planets. It is guessed that something like this happened in our solar system, with a body forming in one of Earth's Lagrange points. But as it's not stable in the long term, the two came too close together, and collided, resulting in the formation of the moon.

IIRC there are several simulations which suggest Trojan planets should form fairly easily in the orbits of giant planets, so presumably some systems will retain them if they do in fact form in reality.

I recall those simulations proposing that such trojan planets were terrestrial in nature (thus a high ratio between the masses of the planets). The scenario ciceron pointed out would have a low planet mass ratio. I suspect this makes a difference for dynamic stability.

Hello everyone,
I found this very interesting discussion!!
I'm a brazilian PhD student and one of the author of the work mentioned above. You pointed some points very interesting and some parts of my current work. We are performing simulations to answer this and other questions.

Awesome! If you don't mind, keep us updated!

Of course! Any new result I post here again.

It's very much appreciated , thank you a lot.