Modeling underwater explosions: Extending Charles to FIVER

FFI-Report 2021

About the publication

Report number

21/00963

ISBN

978-82-464-3344-8

Format

PDF-document

Size

868.3 KB

Language

English

Download publication
Andreas N. Osnes
Modeling of explosions and implosions under water is important for design of vehicles and installations both under water and on the sea surface. There exists simple models, often in the form of empirical calibrations, for the loads experienced at a given distance from an explosion. These models, however, are often insufficient, for example if one is interested in the response of a ship to an explosion at short or moderate distances from the ship. It is therefore necessary to be able to simulate such problems with high accuracy, and account for the interactions between explosion gases, water, air and the ship simultaneously. This is a very difficult problem, but modern methods and computational resources are able to perform such simulations, which provide valuable insights into realistic underwater explosion scenarios. This report is concerned with the implementation of a method for simulation of explosions and implosions under water in the flow solver "Charles", which is used at the Norwegian Defence Research Establishment (FFI). The extended simulation capabilities this implementation provides will strengthen the ability to simulate these problems at FFI, and increase FFI’s knowledge about underwater explosion problems. The "finite volume method with exact two-phase Riemann solver" (FIVER) has been implemented in Charles. This method is especially suited to simulate interactions between materials with very different properties, under the extreme thermodynamic conditions following immediately after detonation of a high explosive. The method is of a type that is readily implemented in the flow solver. In this work, the method has been implemented and tested for one-dimensional problems. In future work, the flow solver will be developed further, and the method will be implemented for two and three dimensions, so that realistic problems can be simulated.

Newly published