Repository hosted by TU Delft Library

Home · Contact · About · Disclaimer ·

BeamBlast: Blast Path-finding algorithms

Publication files not online:

Author: Halswijk, W.H.C.
Source:16th ISIEMS International Symposium for the Interaction of the Effects of Munitions with Structures, Destin, Florida, USA, 9-13 November 2015
Identifier: 529329
Keywords: Physics · Blast · Path · Algorithms · Reflection · Diffraction · V/L Assessment · Observation, Weapon & Protection Systems · WS - Weapon Systems · TS - Technical Sciences


TNO is developing a flexible blast effects module suited to be used in simulations of MOUT scenarios. The aim of the module is to fill the gap between semi-empirical free-air blast relations (with a lower fidel-ity and complexity) and CFD (with larger computational efforts). The module is suited for vulnerabil-ity/lethality (V/L) assessments and is implemented in the V/L TARVAC simulation suite. The TNO blast effects module consists of the BeamBlast model and response models. BeamBlast in-cludes an engineering level path-based blast propagation model, that has proven to be robust (capable of complex geometries) and fast (results within seconds or minutes). Once a set of paths is found, a blast loading model calculates from the set of blast paths a combined pressure time graph. The blast-path model looks for paths around a free triangle-based geometry that agree to the gas-dynamics law concerning diffractions and different kinds of reflections. The user has flexibility with specifying the recursion depth of the algorithm, which strongly influences the completeness of the result-ing set of paths. If the scenario involves an indoor explosion, the recursion number is usually required to be high; an outdoor scenario does not, but may involve more complex geometry. The algorithms find all paths, within the given recursion depth, that lead from the source to the observer. In many cases, there will be multiple diffraction paths around the same object. This is in contrast with Dijkstra-based path-finding algorithms, which find only the approximate shortest route. Furthermore, the paths are not dependent on nodes, but follow the shortest path around corners, which increases the accu-racy of the path length and angles involved. This paper explains the algorithms involved, shows a few early performance indicators, and describes the current developments for improvement.