Diameter Effect Curve and Detonation Front Curvature Measurements for ANFO

Abstract

Diameter effect and front curvature measurements are reported for rate stick experiments on commercially available prilled ANFO (ammonium-nitrate/fuel-oil) at ambient temperature. The shots were fired in paper tubes so as to provide minimal confinement. Diameters ranged from 77 mm (» failure diameter) to 205 mm, with the tube length being ten diameters in all cases. Each detonation wave shape was fit with an analytic form, from which the local normal velocity D n , and local total curvature k, were generated as a function of radius R, then plotted parametrically to generate a D n (k) function. The observed behavior deviates substantially from that of previous explosives, for which curves for different diameters overlay well for small k but diverge for large k, and for which k increases monotonically with R. For ANFO, we find that D n (k) curves for individual sticks 1) show little or no overlap¾with smaller sticks lying to the right of larger ones, 2) exhibit a large velocity deficit with little k variation, and 3) reach a peak k at an intermediate R.Near critical diameter, ANFO exhibits a severe velocity deficit; however, with increasing diameter a slight upturn in the diameter effect curve is observed. We developed an extension of the widely used Campbell/Engelke (3) curve fit that accurately captures both behaviors together. INTRODUCTIONRate stick tests on lightly confined prilled ANFO at ambient temperature were conducted to measure steady-state detonation velocity and detonation front curvature as a function of diameter. It is well known that ANFO's detonation velocity varies considerably with charge diameter. Yancik (1) , Petes (2) , and others have shown that several other physical properties can affect ANFO's detonation velocity. These include the type of confinement, explosive density, particle size and distribution, fuel oil content, moisture content and temperature. Furthermore, these factors can be interdependent. Differences in detonation velocity can also be attributed to slightly different physical properties of the ammonium nitrate (AN) produced by different manufacturers. Because there are no standard specifications by which AN is produced, batch-to-batch variation from a single production plant is not uncommon. Although performance data for ANFO have been studied for over four decades and become somewhat self-consistent since being introduced in the mid-1950's, variability in performance can still be expected from this highly non-ideal explosive.

Interactions of Inert Confiners with Explosives

“…For this explosive-inert confiner pair, figure 17 shows that the shock polars for the two material almost overlay over a wide region, and this behavior persists for a large range of detonation speeds. Interestingly, experiments by Catanach and Hill [12] and Bdzil et al [13] found very different matches in terms of the measured shock normal angles at the edge of the explosive for this explosive-inert pair and for two different lots of ANFO: Catanach & Hill [12] found edge shock angles of around 50 o (this lot had experienced some separation and had a reduced fuel oil content), while Bdzil et al's [13] results gave edge angles of around 25-29 o . The near overlay of the shock polars seen in figure 17 explains why a unique and consistent interaction between the two materials is not seen, and why the measured shock angle is so sensitive to lot.…”

Determination of the velocity-curvature relationship for unknown front shapes

“…where z is the front height, r is the local radius, and R is the charge radius [1,2]. Equation 1 is fit globally across an entire experimental rate-stick front shape by varying parameters A i and η where 0 < η < 1.…”

Reactive flow modeling of small scale detonation failure experiments for a baseline non-ideal explosive

“…3 In addition, ammonium nitrate prill size and porosity may result in ANFO detonation velocities ranging between 1.5 and 4.0 km/s, owing to different levels of fuel oil absorption within the prills 4,5 or other factors. A broad spectrum of nonideal behavior may be observed in shock front curvature, 6 diameter effects, 7 and interactions with confining material, 8 which cannot be determined without large scale (currently $1 kg or greater) rate stick tests. 6,8 The wide range of non-ideal explosives prohibits large scale characterization of every composition of interest.…”

“…A broad spectrum of nonideal behavior may be observed in shock front curvature, 6 diameter effects, 7 and interactions with confining material, 8 which cannot be determined without large scale (currently $1 kg or greater) rate stick tests. 6,8 The wide range of non-ideal explosives prohibits large scale characterization of every composition of interest. Previous attempts to develop small scale characterization tests (e.g., Floret, 9 mushroom, 10 and tiny plate 11 tests) relied on high explosives that can sustain a detonation wave with only a few grams of material.…”