O-211 Probabilistic Modeling of Initiation due to Fragment Impact
October 2019Ernest L. Baker, Christiaan Leibbrandt, Martijn van der Voort
A probabilistic analytic model of initiation for the STANAG 4496 Insensitive Munitions (IM) Fragment Impact (FI) test was developed. The deterministic Jacobs-Roslund (JR) detonation initiation model was augmented to include impact point offset and fragment tilt against a curved munition case. This augmentation was based on geometrical aspects of the impact and the shock produced by impact. The model was calibrated using the 1-D hydrocode, GODLAG. The velocities and thicknesses of the 1-D GODLAG impactors were calibrated based on matching peak pressures and pressure histories from previous fully 3-D fragment impact hydrocode modeling using a 155mm diameter munition configuration. A state of the art probability tool was subsequently used for variational impact conditions using the newly developed analytic IM FI explosive initiation model. The probability computational program AgenaRisk was used. AgenaRisk is a Bayesian Belief Network (BBN) implementation and computes probabilistic relationships. Each variable, or node, is given a certain probability distribution. By expressing the mutual dependencies of the different variables, probabilistic computations are made. By implementing the augmented JR model in AgenaRisk, a probabilistic initiation response model was obtained. A test case was run using the same 155mm diameter munition configuration that was used for the model calibration. The varied parameters included the fragment velocity, tilt and impact position offset. The results showed a large reduction in initiation probability can occur as a result of impact position offset and fragment tilt.