Date of this Version
Understanding the behavior of soil under blast loading is very important to engineers in mining, tunneling, and military construction. Due to the very complex structure of a soil mass it is very difficult to describe its constitutive relation, especially when it has different water contents and it is under blast loading conditions. New protective system designs subjected to blast loading need to be proved its validation prior to predict effect of explosive before implementation. Full-scale, buried explosive tests are costly. Finite element simulations play a significant role in the design of protective systems, for example a bottom platform of lightweight vehicles, against underground explosion.
The Perzyna viscoplastic cap model has been shown to be a valid model for use in the simulations of dry soil behavior under both static and dynamic loading. This model is a dramatic improvement over the inviscid cap model for soil behavior under high strain rate loading, such as from an explosion. However, soil should be modeled as a three-phase porous media to accommodate various degrees of water saturation. This is especially true for the soil mass surrounding the source of energy release, as each of the three phases responds differently to shock loading. To improve the model accuracy, a revised model comprising a Gruneisen equation of state (EOS) for each of the three phases has been developed. These equations of state for solid, water and air have been integrated with a viscoplastic cap model to simulate behaviors of soil with different degrees of water saturation.
These EOS models as well as the viscoplastic cap model are implemented into LS-DYNA as user-supplied subroutines for numerical simulation of six explosive tests in dry soil as well as in saturated soil. The shock front time of arrival, the air pressure directly above the buried explosive, and the ejecta heights predicted by the revised cap model agree fairly well with the experimental data. Four elements from finite element mash are selected to observe three phases volume fractions change. There is noticeable improvement in the prediction of saturated soil behavior than dry soil behavior under blast loading. It is concluded that the revised model is adequate for blast loading behavior simulations for soil with different degrees of water saturation.