O-013 Streamlining Safety Testing for Munitions

Safety testing is generally done on very small numbers of munitions for reasons of cost. Therefore, obtaining high levels of confidence as the result of testing requires to design highly informative and reliable tests and to organize them into rational and cost-effective test programs.

Any individual test must obviously be representative of the threat or hazard it simulates, but it must also meet other conditions: be reproducible/reliable, be meaningful and informative (i.e. allow a good understanding of the test outcome) and be transferable (i.e. allow the extension of the results to other configurations than strictly that of the test). These requirements are not fully compatible, hence the need for compromises which may not always coincide with the current approach to standardized tests, because this approach places emphasis on representativity by trying to stick as closely as possible to the real scenarios. In particular, transfering results from a given test to other configurations can only be done by coupling experimental testing with mathematical modeling. This is true for "single" events such as fuel fire or single bullet impact, and still more for "compound" events such as multiple impacts.

Test programs can be of various types: "single label" or "multiple label" when their purpose is the attribution of a general qualification (e.g. "Insensitive Munition"), or "tailored" to the specific needs of a particular munition or family of munitions. Tailoring test programs requires a precise analysis of all the occurrences (both environmental and combat-generated) likely to be faced over a whole life cycle, but it enables specific threats or hazards to be considered (or eliminated if irrelevant), thus providing higher levels of confidence for lower costs.