Links between detonation wave propagation and reactive flow models.
- Damian C.
- Stephen J.
An accurate reactive flow model is necessary to be able to predict the initiation properties of explosives by complicated shock structures, but a very fine the spatial resolution is needed in reactive flow to reproduce the detailed dynamics of a detonation wave. However, it is not often necessary to use a reactive flow model to simulate the motion of a fully-developed detonation wave. In many situations the same results can be obtained with a coarse computational mesh using programmed burn techniques. In the WBL model [Lambourn89,Swift93], an eikonal detonation wave propagates through a body of explosive at a speed which depends on the curvature of the wave. The model describes the motion of the leading shock of the detonation wave. Here we use the level set method for integrating the WBL equations in time [Collyer98,Bdzil93,Osher88,Aslam98]. This method is attractive because complicated detonation wave shapes can be represented simply. It was found possible to initialize the level set field by a set of source points derived from a reactive flow simulation, by taking 'trigger states' from the reactive flow. The level set scheme was generalized further to take account of motion of the material behind the detonation wave, allowing it to be used for simulations coupled with reactive flow, where detonation may propagate through preshocked and moving material. The modified level set scheme was implemented in 1D and 2D Lagrangian hydrocodes. Trial calculations were performed of initiation and detonation in the TATB-based explosive LX-17, using the Lee - Tarver model. A CJ detonation was simulated in order to verify that the modified level set algorithm operated correctly. The detonation speed was in very good agreement with the expected value. Single-shock initiation was simulated. The position - time history of the leading shock from the coupled model was in excellent agreement with full reactive flow; the pressure profiles were similar but not identical, because of the difference in material properties behind the WBL wave and the omission of the von Neumann spike from the WBL profiles. As a more interesting test, we simulated the shock-to-detonation transition on reflection of a weak shock from a rigid boundary. The position - time history of the leading shock was in good agreement. The pressure profiles varied much more than in the single-shock case, because the WBL calculation used the same propagation parameters and for simplicity imposed the same state at the end of the detonation zone as was used in the single-shock simulation. We have previously used quasisteady flow analysis to derive a reaction rate from experimental measurements of the relation between detonation speed and wave curvature, or vice versa [Swift93]. Reactive flow models have been developed for HMX-based explosives based on mesoscale representations of the components of the explosive [Mulford01], and using a temperature-dependent reaction rate which should be valid over a wide range of loading conditions. The quasisteady analysis scheme was extended to allow arbitrary reaction models to be investigated.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE
- OSTI ID:
- 976256
- Report Number(s):
- LA-UR-02-4998; TRN: US201009%%404
- Resource Relation:
- Conference: Submitted to: 12th International Detonation Symposium, 11-16 August 2002, San Diego, CA.
- Country of Publication:
- United States
- Language:
- English
Similar Records
Modeling Three-Dimensional Shock Initiation of PBX 9501 in ALE3D
Modeling The Shock Initiation of PBX-9501 in ALE3D