Characterisation of Blast Loading from Ideal and Non-Ideal Explosives

Explosive detonation in its simplest form can be characterised by an instantaneous release of energy at an infinitely small point in space as a solid explosive material. This is a result of chemical decomposition of an explosive which reforms as high pressure and temperature gases which expand radially. This supersonic expansion of detonation products compresses the surrounding medium resulting in a shock wave discontinuity which propagates away from an explosive epicentre at high speeds. This has the potential of significant damage to anything the shock wave interacts with.

Shock wave quantification work conducted in 1940`s through to the 1980`s was done so to understand the effects of large scale explosive detonation which was an immediate threat due to the discovery of the nuclear bomb. Highly skilled experimental and theoretical scientists were assigned the task of capturing the effects of large scale detonations through innovative solutions and development of pressure gauges. The in-depth fundamental understanding of physics, combustion and fluid dynamics the researchers utilised resulted in the well-favoured semi-empirical blast predictions for simplistic free-field spherical/hemispherical blasts.\\

A broad amount of literature has been published on free-air characterisation of spherical/hemispherical explosives, with the detonation process and subsequent shock wave formation mechanics being well understood. However, there is yet to be a definitive and robust understanding of how deterministic a shock waves spatial and temporal parameters are for simplistic scenarios. This goes as far as some studies suggesting that semi-empirical tools are not as effective as previously assumed. Often the use of numerical simulations provide reasonable insights to blast loading conditions imparted on structures and scenarios with higher complexities. However, when the validation data used is assumed to exhibit erroneousness, the schemes are no longer characteristically high in fidelity. The lack of quantified variability and confidence in the data which is published, are significant issues for engineers when designing infrastructure that is both robust enough to withstand extreme loading, and not overly conservative that there are cost and material waste implications. This issue is investigated thoroughly within this thesis, highlighting the sensitivity of blast parameters across the scaled distance ranges, and determining their predictability with both numerical simulation and semi-empirical tools.

The vast majority of free-field characterisation has been conducted using military grade explosive which exhibit ideal detonation behaviours; meaning the detonation reaction is effectively instantaneous. Ideal explosives, by the theoretical definition, can be categorised by a simplistic instantaneous energy release. In far-field regimes, any explosive with ideal-like compositions and behaviours should be scalable with mass. This assumption is not valid for homemade explosives (HME), such as ANFO (Ammonium Nitrate + Fuel Oil), whose compositions are usually homogenous, resulting in a finite reaction zone length. These can be long enough to cause failures in detonations and exhibit a variety of different energy releases depending on the mass of the charge resulting in HME's having different TNT equivalence values depending on their scale.

Early works of ANFO characterisation was done so in the desire to replace TNT, to assess its capability of producing similar yields for a fraction of the manufacturing costs. This meant the hemispherical detonations of ANFO which have led to its overall classification, were done using charges of over 100kg and therefore non-ideal reaction zone effects become negligible in comparison to the overall charge size. Yields presented in this region were consistently measured at around 80\% of a similar TNT detonation and has therefore been incorrectly assumed a rule for ANFO across all mass ranges within published literature.\\

There is a distinct lack of characterisation of non-ideal explosives throughout the mass scales, posing a significant implication for designing structures to withstand the threat of HMEs. With the knowledge that energy is released at a much slower rate when detonating these compositions, the assumption that large scale trials accurately capturing the behaviour of a small charge masses, when scaled down, is not verified. Most HMEs will be hand held devices or, at the very least, backpack size, meaning the threat currently is not predictive with confidence through validated data conducted under well-controlled conditions. Small scale ANFO trials have demonstrated this to be the case within this thesis, with theoretical mechanisms proposed which offering a prediction method of the behaviour of non-ideal detonation across all mass scales.

Findings in this PhD thesis will offer a conclusion on whether shock waves in free-field scenarios are deterministic for both ideal and no-ideal explosives, with a particular emphasis on the far-field range. The results presented are developments in the accurate quantification of shock wave loading conditions a structure is subjected to through explosive detonation and should be used by engineers to establish robust, probabilistic but accurate designs.