Typically, the cost of a space-borne imaging system is driven by the size and mass of the primary aperture. The solution that we propose uses a method to construct an imaging system in space in which the nonlinear optical properties of a cloud of micron-sized particles, shaped into a specific surface by electromagnetic means, and allows one to form a very large and lightweight aperture of an optical system, hence reducing overall mass and cost. Recent work at JPL has investigated the feasibility of a granular imaging system, concluding that such a system could be built and controlled in orbit. We conducted experiments and simulation of the optical response of a granular lens. In all cases, the optical response, measured by the Modulation Transfer Function, of hexagonal reflectors was closely comparable to that of a conventional spherical mirror. We conducted some further analyses by evaluating the sensitivity to fill factor and grain shape, and found a marked sensitivity to fill factor but no sensitivity to grain shape. We have also found that at fill factors as low as 30%, the reflection from a granular lens is still excellent. Furthermore, we replaced the monolithic primary mirror in an existing integrated model of an optical system (WFIRST Coronagraph) with a granular lens, and found that the granular lens that can be useful for exoplanet detection provides excellent contrast levels. We will present our testbed and simulation results in this paper.