Knowledge of the constitutive behavior of explosives is necessary for the development of exploding devices that are safe to handle and transport and that are effective in use. The plastic deformation of most explosives is strongly affected by pressure and by strain rate. Pressure-hardening effects may be determined directly by a uniaxial-compression test performed in a special hydraulic apparatus that applies a lateral pressure to the specimen. However, the apparatus is expensive and testing explosives presents some risk to it.

Previously, to investigate both pressure- and rate-hardening effects, we developed a modified Taylor cylinder-impact test in which the explosive specimen is encased in a confining sheath (or sleeve) of a well-characterized material, typically a ductile metal such as copper. The sheath, which is stiffer than the explosive, serves to confine the specimen to provide increased pressure and also maintains its integrity. Computations using a large-deformation dynamic plasticity code (hydrocode) are then fitted to the test results (the shape of the deformed specimen) to provide an estimate of the constitutive parameters.

To refine our results, we are currently developing a similar quasi-static compression test that uses a confining sheath to develop the pressure in the specimen. In our cylinder-impact tests, the copper sheath was able to generate pressures up to about 0.5 GPa. In quasi-static compression tests, however, a copper sheath fails at much lower pressures. To provide the higher pressures required, we are using instead a sheath of graphite- or glass-fiber composite material. The fibers are aligned in the circumferential direction to provide a very large hoop strength, with little axial stiffness. The axial load is thus carried mainly by the explosive specimen.

To design the sleeve and for post-test fitting of results, we have modified a finite-element hydrocode to include an anisotropic (orthotropic) elastic-plastic constitutive formulation to model the composite. Compressibility behavior is modeled using an established fit to Murnaghan_s pressure-volume relation. To describe the yield surface, both the simple Drucker-Prager criterion and a non-linear relation will be tried.

This test will be used to characterize first mock explosives, then, once validated, actual explosives. Advantages of the new test include economy and low risk of damage to test equipment.

This content is only available via PDF.
You do not currently have access to this content.