A simplified approach to modelling blasts in computational fluid dynamics (CFD)

Abstract

This paper presents a time-efficient numerical approach to modelling high explosive (HE) blastwave propagation using Computational Fluid Dynamics (CFD). One of the main issues of using conventional CFD modelling in high explosive simulations is the ability to accurately define the initial blastwave properties that arise from the ignition and consequent explosion. Specialised codes often employ Jones-Wilkins-Lee (JWL) or similar equation of state (EOS) to simulate blasts. However, most available CFD codes are limited in terms of EOS modelling. They are restrictive to the Ideal Gas Law (IGL) for compressible flows, which is generally unsuitable for blast simulations. To this end, this paper presents a numerical approach to simulate blastwave propagation for any generic CFD code using the IGL EOS. A new method known as the Input Cavity Method (ICM) is defined where input conditions of the high explosives are given in the form of pressure, velocity and temperature time-history curves. These time history curves are input at a certain distance from the centre of the charge. It is shown that the ICM numerical method can accurately predict over-pressure and impulse time history at measured locations for the incident, reflective and complex multiple reflection scenarios with high numerical accuracy compared to experimental measurements. The ICM is compared to the Pressure Bubble Method (PBM), a common approach to replicating initial conditions for a high explosive in Finite Volume modelling. It is shown that the ICM outperforms the PBM on multiple fronts, such as peak values and overall overpressure curve shape. Finally, the paper also presents the importance of choosing an appropriate solver between the Pressure Based Solver (PBS) and Density-Based Solver (DBS) and provides the advantages and disadvantages of either choice. In general, it is shown that the PBS can resolve and capture the interactions of blastwaves to a higher degree of resolution than the DBS. This is achieved at a much higher computational cost, showing that the DBS is much preferred for quick turnarounds.