Large Eddy Simulations (LES) of a novel type of wing/body junction called the anti-fairing are performed in the current thesis to study the complex turbulent flow physics involved in the junction area and also to obtain a clear understanding of the drag reduction capabilities of the anti-fairing. In regards to that, two separate LES are performed: one for the baseline case with a Rood wing/flat plate combination and another with the Rood wing/anti-fairing combination. A detailed comparative study is performed between the two cases to observe important differences in junction flow characteristics. Both the simulations are performed on a 25 million immersed boundary Cartesian mesh by solving the incompressible Navier-Stokes equations using the in-house finite volume LES solver called INCA. Results from the LES study confirms the existence of the propulsive pressure mechanism of drag reduction for the anti-fairing case, previously proposed by Belligoli et al. However, the results also show that there exists a secondary drag reduction mechanism caused by a combination of increase in approach boundary layer momentum thickness and dampening of the turbulence associated with the horseshoe vortex (HSV) upstream of the wing. This secondary mechanism has been found to be caused by the convex dent present at the start of the anti-fairing geometry. The total drag reduction for the anti-fairing case comes out to be 1.8%. A new parameter called junction drag is defined which accounts for the drag only due to the presence of a junction. The reduction in junction drag obtained for the anti-fairing case is about 6.8%. Apart from the LES analysis, a RANS analysis has also been performed to further investigate the drag reduction capabilities of anti-fairing for different approach boundary layer thicknesses and anti-fairing depths. All the RANS analysis have been performed on a 5 million body-fitted mesh by solving the incompressible Navier-Stokes using the open source finite-volume solver OpenFOAM. Results from the RANS analysis indicate that there exists an optimum depth for the anti-fairing which corresponds to the least drag. Furthermore, it is found that the effect of approach boundary layer thickness is mostly on changing the base drag of the case where no anti-fairing is present, rather than actually affecting the performance of the anti-fairing at different depths.