In nature, communal hunting is often performed by predators by charging through an aggregation of prey. However, it has been noticed that variations exist in the geometric shape of the charging front; in addition, distinct differences arise between the shapes depending on the particulars of the feeding strategy. For example, each member of a dolphin foraging group must contribute to the hunt and will only be able to eat what it catches. On the other hand, some lions earn a "free lunch" by feigning help and later feasting on the prey caught by the more skilled hunters in the foraging group. We model the charging front of the predators as a curve moving through a prey density modeled as a reaction-diffusion process and we optimize the shape of the charging front in both the free lunch and no-free-lunch cases. These different situations are simulated under a number of varied types of predator-prey interaction models, and connections are made to multi-agent robot systems.