Scaling failure in blast loaded structures is considered to be impossible with the known scaling laws, when using fracture mechanics-based considerations (Jones, 1989). We will show in this research that scaling failure becomes possible when 2 alternative competing criteria are used, namely: maximum normal stress to describe separation (cracking) and an energy-based criterion that describes adiabatic shear failure. The failure criteria implemented in the numerical failure model are a strain energy density criterion, and a maximum principal stress criterion. These criteria were modeled using a user subroutine and inserted into a commercial finite element code. Numerical simulations of two test-cases were carried out: Failure of circular clamped plates under close-range, air blast loading, and penetration experiments carried out by Borvik et al. (1999, 2001). The scaling laws used in this study are the geometrical scaling laws for the geometrical parameters, along with the Hopkinson cube root scaling law for the blast. The dominant failure criterion in blast loading was maximum normal stress, while for the penetration simulations the dominant criterion was the energy-based criterion. This study shows that both the prototype and small-scale model undergo scaling for those failure criteria. This study presents a new alternative to the scaling of structural failure under dynamic loading conditions, which is both simple and efficient.
