The formation mechanism of the martian moons Phobos and Deimos is still enigmatic. A debris disk origin hypothesis, as opposed to a capture hypothesis, has the advantage of explaining the current low eccentricity and low inclination of the two moons. Hesselbrock and Minton (2017) proposed that Phobos may be a product of an ongoing ring-moon cycle that was initiated by an early massive debris disk generated from a giant impact. However, Canup and Salmon (2018) suggested that Deimos would have been accreted onto the large satellites produced by the early cycles in the Hesselbrock and Minton scenario. They concluded that the early martian debris disk would have only been massive enough to produce Phobos, and not the larger inner satellites required by the ring-moon cycle model.
The N-body integrator used by Canup and Salmon, called HydroSyMBA, treated all collisions between bodies as pure mergers. We revisit the evolution of the martian moons within a debris disk with our own version of the SyMBA integrator called Swiftest/SyMBA, which includes collisional fragmentation. Swiftest/SyMBA has been under development by the Purdue Swiftest team. Swiftest is able to resolve close encounters and differentiate collisional outcomes between hit and runs, disruptions, and mergers. We compare the evolutionary outcomes of three main debris disk (high, medium, and low mass) combined with various state of excitation (low or high initial eccentricity) and compare it to the previous model of Canup and Salmon. The results illustrate the consequence of fragmentation on the final state of the debris disk in terms of embryos mass, semi-major axis, and density. The rate of accretion of the martian debris disk is evaluated with collisional dynamics, and could explain how Deimos could have survived within a massive debris disk.