|
One of the main activities of the Explosion Prevention and Protection Research Group of TNO is the quantification of explosion effects and the prediction of possible consequences. The research presented is aimed at setting up new guidelines for safety distances for urban areas around hazardous sites. To investigate the distribution of the blast load on buildings experimentally, the buildings are scaled down and exposed to a plane shockwave in a 40x40 cm2 shock tube. For perfect gasses the Hopkinson scaling law applies. This law states that if all characteristic times are scaled by the same factor as the length scale then all pressures, temperatures, densities and velocities will remain unchanged. In the shocktube the shockwave will propagate and interact with objects. The density distribution around the objects can be visualized using an interferometer and from the interferograms the pressure distribution can be computed. These data can be used to validate numerical results obtained with the Computational Fluid Dynamics code BLAST. This code, developed at the Prins Maurits Laboratory of TNO, simulates the interaction of three-dimensional blast waves with structures. The data obtained can be used for the accurate prediction of the blast load on individual buildings in an urban area. In the tests, rectangular blocks were used to obtain the two dimensional test situation shown in figure 1 .At some typical locations on the structures, the pressure signal was measured by means of piezo-resistive pressure-transducers. Interferograms were recorded using a phase-stepping double-reference-beam holographic interferometer. The principle of the interferometer setup is that the initial state and the disturbed state are both recorded on one single holographic plate, using two slightly tilted reference beams. When analyzing, the two reference beams simultaneously reconstruct the initial state and the disturbed state. The two states will interfere and the resulting interference pattern is recorded using a CCD camera. This interferometer setup is very attractive because, first of all, it is possible to record an interference pattern hardly influenced by the mechanical vibrations caused by the shock tube, secondly, this interferometer is easy to setup, and thirdly, most lens errors cancel out in the resulting image because of the double reference beam. A typical example of a reconstruction of an interferogram recorded with the holographic interferometer is shown in figure I. where the shockwave travels from the left to the right, interacting with two models of buildings. To be able to calculate tile pressure distribution, more than one reconstruction of the interferogram is needed. If the difference in optical pathlength between the reference beams is varied over an unknown, but constant, distance, at least four reconstructions are needed. From four different recordings the pressure distribution can be calculated. First, all pixels that do not have enough contrast are being removed, then the phase is calculated, for example using Carré's algorithm, the result of this can be seen in flgure 2. Subsequently the discrete 271 steps have to be removed, which is called phase unwrapping, and finally the grey values in this image have to be converted to pressure values. The resulting pattern of isobars can be seen in figure 3.This image can be compared to tile numerical simulation in figure 4 that was obtained using the BLAST-code. ibis paper presents an optical study of blast wave propagation and interaction with multiple structures and a method for obtaining quantitative information on the pressure distribution from a number of phase-stepped images.
|
