EFFECT OF PRE-BENDING AND HYSTERESIS OF A METAL PLATE ON BLAST-LOADING DEFORMATION: DATA VS. NUMERICAL SIMULATION

V-shaped hulls for vehicles, to mitigate buried blast loads, are typically formed by bending plate. Such an approach was carried out in fabricating small test articles and testing them with buried-explosive blast load in Southwest Research Institute’s (SwRI) Landmine Test Fixture. During the experiments, detailed time dependent deflections were recorded over a wide area of the test article surface using the Dynamic Deformation Instrumentation System (DDIS). This information allowed detailed comparison with numerical simulations that were performed with LS-DYNA. Though in general there is good agreement on the deflection, in the specific location of the bends in the steel the agreement decreases in the lateral cross section. Computations performed with empirical blast loads developed by SwRI and by more computationally intensive ALE methods in LS-DYNA produced the same results. Computations performed in EPIC showed the same result. The metal plate was then bent numerically so that the initial plate had both hardening and residual stresses from the fabrication. When blast loaded, though the deflection reduced due to the hardening in the bends in the plate, the qualitative disagreement with the lateral cross section remains. The study then focused on the material strength model for the steel. It was observed that the difference in behavior between the experiments and the computations occurs in a region where the hull metal is unloading from its formative bend. It is argued that using a kinematic yield surface with hysteresis, rather than an isotropic one with no hysteresis as is commonly done with the Johnson-Cook model, better models the unloading and hence can better match the deformation seen in the experiments.