Estimation of upper bound probabilities for rare events resulting from nearby explosions
It is sometimes necessary to deploy, transport and store weapons containing high explosives (HE) in proximity. Accident analyses of these activities may include nearby explosion scenarios in which fragments from an exploding (donor) weapon impact a second (acceptor) weapon. Weapon arrays are designed to miti- gate consequences to potential acceptor weapons, but unless initiation of an accep- tor's HE is impossible, outcomes such as detonation must be considered. This paper describes an approach for estimating upper bound probabilities for fragment- dominated scenarios in which outcomes are expected to be rare events. Other aspectsl,z of nearby explosion problems were addressed previously. An example scenario is as follows. A donor weapon is postulated to detonate, and fragments of the donor weapon casing are accelerated outward. Some of the fragments may strike a nearby acceptor weapon whose HE is protected by casing materials. Most impacts are not capable of initiating the acceptor's HE. However, a sufficiently large and fast fragment could produce a shock-to-detonation transi- tion (SDT), which will result in detonation of the acceptor. Our approach will work for other outcomes of fragment impact, but this discussion focuses on detonation. Experiments show that detonating weapons typically produce a distribution of casing fragment sizes in which unusually large figments sometimes occur. Such fragments can occur because fragmentation physics includes predictable aspects as well as those best treated as random phenomena, such as the sizes of individual fragments. Likewise, some of the descriptors of fragment impact can be described as random phenomen% such as fragment orientation at impact (fragments typically are tumbling). Consideration of possibilities resulting from the various manifesta- tions of randomness can lead to worst-case examples tha~ in turn, lead to the out- comes of concern. For example, an unusually large fragment strikes an acceptor weapon with the worst possible orientation and in the acceptor's most vulnerable location. Intuitively, such an event clearly is very unlikely. Our approach is based on the quantification of such "unlikelihood." The randomness inherent in the physics is modeled explicitly. Worst-case events are pre&cted in simulations but appear with appropriately small frequencies and lead, in turn, to corresponding probability estimates. To make the complex physics involved in nearby explosions tractable, we keep some of worst-case approach. For example, we might assume that fragments strike with the worst-case orientation. Because some of the physics is modeled as worst case, our results become upper bound probability estimates. The remaining physics, which includes the random aspects, is modeled in a way that makes Monte Carlo esti- mation of probabilities possible. In the following, the different physical processes involved in nearby explosion scenarios are considered separately: fragmentation of the donor case, transport of the fragments to the acceptor, and impact of the fragment(s) and consequent effects on the acceptor warhead. The discussion focuses on the modeling choices, which inevitably limit the probability range over which outcome probabilities can be bounded by limiting the number of Monte Carlo simulations that are practical.
- Research Organization:
- Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE Office of Defense Programs (DP)
- DOE Contract Number:
- W-7405-ENG-36
- OSTI ID:
- 1392
- Report Number(s):
- LA-UR-98-1633; ON: DE00001392
- Resource Relation:
- Conference: PSAM 4, New York, NY, 09/13 - 09/18/1998
- Country of Publication:
- United States
- Language:
- English
Similar Records
Probabilistic Shock Iinitiation Thresholds and QMU Applications
Detonator cable initiation system safety investigation: Consequences of energizing the detonator and actuator cables