Some mathematical properties of the uniformly sampled quadratic phase function and associated issues in digital Fresnel diffraction simulations