This paper presents a new framework for determining the Stillinger-Weber (SW) potential parameters for modeling fracture in graphene and carbon nanotubes. In addition to fitting the equilibrium material properties, the approach allows fitting the potential to the forcing behavior as well as the mechanical strength of the solid, without requiring ad hoc modification of the nearest-neighbor interactions for avoiding artificial stiffening of the lattice at larger deformation. Consistent with the first-principles results, the potential shows the Young's modulus of graphene to be isotropic under symmetry-preserving and symmetry-breaking deformation conditions. It also shows the Young's modulus of carbon nanotubes to be diameter-dependent under symmetry-breaking loading conditions. The potential addresses the key deficiency of existing empirical potentials in reproducing experimentally observed glass-like brittle fracture in graphene and carbon nanotubes. In simulating the entire deformation process leading to fracture, the SW-potential costs several factors less computational time compared to the state-of-the-art interatomic potentials that enables exploration of the fracture processes in large atomistic systems which are inaccessible otherwise.