Method of characterising the structure of a void sensitized explosive composition

Abstract

A method of characterizing the structure of a void sensitized liquid energetic material, which method comprises defining the material in terms distribution function, the distribution function representing the fraction of liquid energetic material that occurs at a given point within the void sensitized liquid energetic material.

Claims

1. A method of achieving a designed bulk detonation energy output in terms of a ratio of shock energy to heave energy for an explosives composition that comprises sensitizing voids distributed within liquid energetic material and that is formulated by blending together a void sensitized liquid energetic material and a void-free liquid energetic material, the method comprising: identifying a distribution function template that is representative of a desired ratio of shock energy to heave energy, the distribution function template being characterized by a blending ratio of void sensitized liquid energetic material to void-free liquid energetic material and by a characteristic dimension of void-free liquid energetic material relative to a mean diameter of voids in the void sensitized liquid energetic material; selecting a density for the explosive composition; providing a void sensitized liquid energetic material having a mean void diameter and a void-free liquid energetic material that, based on their respective densities and on the blending ratio, would achieve the density selected for the explosive composition; and formulating the explosive composition by blending together at the blending ratio the void sensitized liquid energetic material and the void-free liquid energetic material, wherein blending is carried out to achieve the characteristic dimension of void-free liquid energetic material relative to the mean diameter of voids in the void sensitized liquid energetic material consistent with the distribution function.

2. The method of claim 1, wherein the distribution function is non-Gaussian.

3. The method of claim 1, wherein the sensitizing voids are gas bubbles.

4. The method of claim 1, wherein the sensitizing voids are at least one of microballoons, or polystyrene.

5. The method of claim 1, wherein the liquid energetic material is an emulsion.

6. The method of claim 1, wherein the sensitizing voids are gas bubbles having a mean diameter between 20 and 2000 microns.

7. The method of claim 1, wherein the sensitizing voids have a density below 0.25 g/cc and wherein the voids have a mean diameter in the range of 20 to 2000 microns.

8. The method of claim 1, wherein the sensitizing voids are at least one of gas bubbles, glass microballoons, plastic microballoons, or expanded polystyrene beads having with a density below 0.25 g/cc and wherein the voids have a mean diameter in the range of 20 to 2000 microns.

9. The method of claim 8, wherein the mean diameter of voids is in the range 40 to 500 microns.

10. The method of claim 1, wherein the void-sensitized liquid energetic material and void-free liquid energetic material are blended by laminar mixing using a static mixer.

Description

BRIEF DISCUSSION OF DRAWINGS

(1) FIG. 1 shows Distribution Function templates for conventional void-sensitized explosive formulations;

(2) FIG. 2 shows Distribution Functions templates for conventional and non-conventional void-sensitized explosive formulations;

(3) FIG. 3 shows the differential of Distribution Functions for conventional and non-conventional void-sensitized explosive formulations;

(4) FIG. 4 is an X-ray image of a conventional void-sensitized explosive formulation;

(5) FIG. 5 shows the differential of Distribution Functions for conventional and non-conventional void-sensitized explosive formulations;

(6) FIG. 6 is a plot comparing VOD against inverse/diameter for two conventional void-sensitized explosive formulations and for one non-conventional void-sensitized explosive formulation;

(7) FIG. 7 is a schematic illustrating an apparatus referred to in the examples;

(8) FIG. 8 is a schematic illustrating a mixing element referred to in the examples;

(9) FIGS. 9-11 are graphs illustrating results obtained in the examples;

(10) FIG. 12 is a schematic illustrating a container used for obtaining emulsion samples for determining distribution function;

(11) FIG. 13 is a processed image of an explosive material as referred to in the examples;

(12) FIGS. 14-16 are plots of bubble position against distance as referred to in the examples;

(13) FIG. 17 is a plot of cumulative fraction versus separation distance for formulations referred to in the examples; and

(14) FIG. 18 is a plot of normalized distribution function rate versus cumulative fraction for formulations referred to in the examples.

(15) FIG. 19 is a plot of distribution function rate versus cumulative fraction for simulated formulations referred to in the examples

DETAILED DISCUSSION OF THE INVENTION

(16) As noted above, in the context of the present specification, the distribution function (DF) for a void-sensitized liquid energetic material is a statistical representation of the fraction of liquid energetic material that is within a given distance from any void surface. This can be illustrated with reference to FIG. 1 below. FIG. 1 shows DF templates that are representative of conventional emulsion explosives in which a liquid energetic material is sensitized by the inclusion of voids. The voids have a random distribution in the liquid energetic material.

(17) In FIG. 1 the y-axis is the fraction of liquid energetic material within a distance r from any void surface and the x-axis represents the radial distance from the nearest void surface. The solid line, DF0 template, represents a theoretical emulsion in which the voids, are at the centers of an array of 50 micron cubes, and r is the distance from the nearest void surface. The dotted line, DF1 template, represents a conventional emulsion of the same density as the cubic array, but with a random distribution of the voids, 95% having separations between 35 to 60 microns (a random generator picks positions in a 50 micron cubic grid so that voids can be placed randomly in the grid until the target voidage (density) is reached). This random distribution of voids is consistent with what one would observe in conventional emulsion explosives that are formulated by distributing sensitizing voids within a liquid energetic material.

(18) In practice, the randomness of the distribution of the voids will depend on the mixing procedure used, and the corresponding DF may vary from the DF1 template slightly. Nevertheless, it is believed that such changes would not be dramatic: the curve would still be sigmoid in nature and there would be no abrupt changes in the slope of the curve. In relation to such conventional void-sensitized liquid energetic materials the present invention resides in the application of DF to describe/represent the internal structure of the material. The application of statistical modeling involving DF to explosives is unique in this regard.

(19) The present invention is also concerned however with characterizing the internal structure of explosives materials that are new with respect to how voids are distributed within a liquid energetic material, and to the corresponding DF templates associated with such new explosives materials. Noting the random manner in which voids are present in conventional void-sensitized explosive materials, in general terms this new internal structure may be described as involving a non-random (or designed) distribution of voids. In view of this fundamental difference in void distribution, these new explosive materials will have different DF templates when compared with the DF templates associated with conventional materials.

(20) This embodiment of the present invention may be illustrated with reference to unique forms of explosive formulation that have a non-random distribution of voids in a liquid energetic material. Specifically, this explosive is manufactured by blending a void-free energetic liquid with conventional void sensitized energetic liquid. These formulations are referred to as mixtures of emulsion, designated MoE. Careful blending is undertaken to ensure that the finished formulation includes discrete regions of the individual component liquid energetic materials. The explosive can be conveniently prepared by laminar mixing of streams of the individual components using a static mixer (see for example FIG. 7 and the accompanying discussion). By this mixing methodology the streams of the individual components are split into sheets that have a mean thickness typically in the range 0.2 to 50 mm. It is to be understood however that sheets of larger thicknesses could be employed without deviating from the spirit of the invention. The characteristics of the sheets can be adjusted by adjusting the mixing methodology, for example by varying the number of mixing elements in the static mixer. DF templates for a number of formulations with varying dimensions of the void-free regions of liquid energetic material were modeled using the DF procedure described above. FIG. 2 is a plot as per FIG. 1 showing how the DF varies for each formulation.

(21) In relation to FIG. 2: Template (DF0) and Template (DF1) are the same as in FIG. 1, and correspond to the theoretical and conventional void-sensitized emulsions. Template (DF2) relates to a 50:50 blend of the conventional void sensitized emulsion and void-free emulsion in which the regions of void-free emulsion have dimensions ranging from 2 to 4 times the diameter of the voids in the sensitized emulsion. Template (DF3) relates to a 50:50 blend of the conventional void sensitized emulsion and an void-free emulsion in which the regions of void-free emulsion have dimensions ranging from 3 to 6 times the diameter of the voids in the sensitized emulsion. Template (DF4) relates to another equal blend of the conventional void-sensitized emulsion and an void-free emulsion, but in this case the regions of void-free emulsion have dimensions ranging from 4 to 8 times the diameter of the voids in the sensitized emulsion. Template (DF6) exhibits simply a coarser blend of sensitized and void-free emulsions in which the regions of void-free emulsion have dimensions ranging from 6 to 10 times the diameter of the voids in the sensitized emulsion.

(22) It will be noted that the formulations in which the voids are provided with a non-random (designed) distribution give rise to DFs that have increasingly different shapes from those for conventional emulsions, i.e. DF0 and DF1. For formulations having a non-random void distribution, the plot of DF against radial distance (r) departs from that of conventional formulations with this departure becoming more exaggerated as the dimensions of the void-free emulsion increases.

(23) For DF2, DF3, DF4 and DF6 the exact shape of the curve will vary depending on such factors as the voidage level of the sensitized emulsion and the void distribution of that emulsion.

(24) An alternative method of displaying the differences between DFs for the conventional and non-random void sensitized formulations is to plot the differential of the DF with respect to the distance from the nearest void surface r, against the DF. This produces a graph that is similar in form to the conventional way of displaying reaction kinetics in the modelling of detonation. In this the reaction rate is plotted against the fraction of material reacted.

(25) Such a DF rate plot is shown in FIG. 3 where the y-axis is the rate of change of the distribution function from the nearest void surface (r) (DF rate) and the x-axis is the unity normalized distribution function.

(26) In relation to FIG. 3: Template (DF0) and Template (DF1) correspond to the theoretical and conventional emulsion blends as shown in FIG. 1. DFa3, DFa5, DFa8 and DFal14 are 50:50 blends of a conventional emulsion and an unsensitized emulsion in which the conventional emulsion is distributed as droplets/globules in a continuum of the unsensitized emulsion, the diameters of the droplets/globules being approximately 3, 5, 8 and 14 times the average diameter of the voids.

(27) Various aspects are worthy of comment: The first point to notice with this method of displaying information is the dome shape of the distribution function curves. For the conventional emulsions the dome is more or less symmetrical, remaining convex over DF values (x-axis) from 0 to 1. However, this is not the case for the non-conventional formulations, where the domed portion of the curve extends approximately only from DF values (x-axis) 0 to 0.5, after which the curve has a point of inflexion and transitions to a concave shape. It will be shown later that emulsions that exhibit this characteristic point of inflexion and concave shape in their DF curve exhibit reduced VODs relative to conventional emulsions with symmetrical, convex DF curves. For the non-conventional formulations the maximum value of DF rate over the DF range from 0 to 1 is significantly less than for the conventional formulations. The non-conventional formulations exhibit increasingly lower values of DF rate (y-axis) and reduced slope gradient at values of DF above 0.5. This is the consequence of distance between (r) the sensitizing voids becoming greater. The emulsions prepared by conventional methods exhibit comparable DF rate of non-conventional materials only at DF values between 0.85 and 1.0. The DF rate templates for the non-conventional formulations correspond to emulsion blend ratios of sensitized to dense emulsions from 10% to 90%, which roughly correspond to the transition from the dome region to the lower DF rate region occurring at DF values between 10% and 85%.

(28) Experimental measurements of the distribution functions (DFs) of conventional emulsions (random distribution of voids) were carried out using an X-ray tomography method to record the positions and sizes of voids in a 10 mm10 mm1 mm sample of a gassed emulsion. The two dimensional digital record of this was analyzed using commercial image analysis software that identified the outer edges of all the voids, and provided a digital output of the coordinates of the centre and length of the circumference of each void. This data was then used to generate templates for the DF rate plots. An X-ray tomography image and analysis of a conventional gas-void emulsion is shown in FIG. 4. The circumference of lighter of the voids is analysed, noting also that certain features were identified as ammonium nitrate crystals where the emulsion has broken down.

(29) The data from this two dimensional analysis was also used to generate DF rate graphs. This was done by calculating the distance of each pixel of the digital image that corresponds to emulsion, from the nearest void surface, a computationally intensive operation. The resultant graph of the experimental DF is shown in FIG. 5. FIG. 5 is a representation of distribution function rate (DF rate) for the experimental X-ray image analysis of the experimental data. DFex is the experimental data for a conventional emulsion in which voids cover about 20% of the area, the traces therefore stopping below this value on the x-axis. DFsim is a simulated conventional emulsion in which the void size distribution and average void concentration is set approximately equal to that of the experimental data.

(30) It will be noted that DFex and DFsim in FIG. 5 exhibit a convex shape consistent with the convex shape of plots for DF0 and DF1 in FIG. 3.

(31) From the foregoing it should be apparent how to generate DF profile templates for void sensitized formulations. The approach may be especially useful for generating DF templates for non-conventional formulations that are typically prepared by blending a conventional void sensisitized emulsion with a void-free (or differently sensitized) continuum of liquid energetic material.

(32) FIG. 6 shows a plot of velocity of detonation (VOD) divided by ideal VOD versus inverse diameter, where the ideal VOD is calculated by application of hydrodynamic theory, for example the Orica Ltd program IDEX. The figure plots results for two conventional explosive formulations and one non-conventional explosive formulation for charge diameters in range between 40-300 mm.

(33) The conventional charges were samples of AN-based emulsion explosives prepared by a conventional methodology at densities equal to 1.22 and 1.02 g/cm.sup.3 for EM 100 both exhibiting a random distribution of sensitizing voids. The total sensitizing voids volume was equal to about 5.3% for EM 100 at 1.22 g/cm.sup.3 and 23% for EM 100 of the AN-based liquid energetic material continuum. The latter was the same for both formulations. With regard to VOD data the solid lines in FIG. 6 are fits to a theoretical model of non-ideal detonation.

(34) The main point to note from, this experiment is that the emulsion prepared by a conventional method as per DFsim/DFex templates exhibits an approximately straight line relationship of VOD/idealVOD against inverse diameter. The DF rate profiles for these conventional formulations are reasonably matched to be in line with the DFsim/DFex template in FIG. 5 above.

(35) A non-conventional emulsion explosive formulation (denoted MOE25) was prepared according to a selected DF rate design template produced in accordance with the present invention. The non-conventional formulation was a blend of 25% mass void sensitized liquid energetic material (density 1.02 g/cc) and 75% mass void-free liquid energetic material continuum (density 1.32 g/cc). The liquid energetic material used was the same as used in formulating the conventional EM 100 control samples. The resulting explosive charges of MOE25 had a density of 1.23 g/cc.

(36) Experimental samples were prepared in a specially designed emulsion experimental rig shown in FIG. 7 and described in Example 1.

(37) Notably, the relationship between VOD against inverse diameter for this non-conventional formulation was very different from that of the conventional control sample. Indeed, considering that the liquid energetic material continuum used is identical, it is remarkable to see the vast difference between the VOD characteristics for these formulations.

(38) More importantly, the non-conventional formulation shows a characteristic highly concave variation of unconfined normalised detonation velocity (VOD/idealVOD) versus inverse diameter. In contrast, the formulations prepared by conventional methodology exhibit an approximately straight or slightly concave shape from the critical diameter to the ideal VOD.

(39) It is well known to those skilled in the art that at a given explosive density, the shock energy increases with increasing VOD, and that a reduction in VOD corresponds to an increase in heave energy.

(40) For a given liquid energetic material, it is important to note that lower VODs can be obtained in conventional formulations by reducing density, i.e. by increasing the level of voidage include in the liquid energetic material. However, an undesirable effect of this is reduced energy density output and thus lower heave and shock energy.

(41) In distinct contrast, the formulation provided in the present invention enables reduced VOD to be achieved without reducing overall energy density. Thus, such non-conventional formulations may provide a remarkable enhancement in energy density as well as enhanced and unique partitioning of heave energy to shock energy.

(42) In practice implementation of the design aspect of the present invention is likely to involve the following sequence of steps, given by way of illustration with reference to a particular example: 1. Select the density of the void-free liquid energetic material being used and the desired density of the high energy density/high heave charge to be formulated. For example, the density of the void-free liquid energetic material may be 1.32 g/cc and the required density of the explosive charge to be produced is 1.23 g/cc. 2. Calculate the total volume of the voidage that needs to be incorporated to achieve the required density. Calculated voidage volume is (100)(1.23/1.32100)=6.8%. Note: this is not necessary for gas sensitized emulsions. However, it is helpful in case of micro-balloons as sensitizing agent or other material voids when the particle density is known. The required mass of balloons to achieve voidage-density can be then calculated. 3. Select the mean size of the voids to be used for sensitization. For example, the mean size of the voids might be 150 m (Measure the size distribution if desired). 4. Select the DF template to obtain desirable VOD (shock/heave ratio), for example, the DF4 template. This template represents 50/50 volume fine blend of conventional void sensitized liquid energetic material and void-free liquid energetic material. 5. Calculate the required density of sensitized energetic material that gives the final density of 1.23 g/cc when mixed 50/50 with void-free liquid energetic material, i.e. 1.14 g/cc. 6. Blend 50% sensitized conventional liquid energetic material (density of 1.14 g/cc) and 50% void-free liquid energetic material (density of 1.32 g/cc) utilizing process consistent with achieving the DF4 template. 7. The DF4 template requires the high density regions to have dimensions equal to 4-8 times the diameter of the voids. Calculate the size of the dense emulsion regions as (150 m4)=600 m and (150 m8)=1200 m. 8. Select the static mixer blending head with laminar flow design such that individual streams of sensitized and void-free components are provided within the thickness specified by DF4 template. This is 600-1200 m.

(43) Embodiments of the present invention are illustrated with reference to the following non-limiting examples.

EXAMPLES

(44) Description of Equipment

(45) Experimental samples were prepared in a specially designed emulsion experimental rig. The corresponding process diagram is shown in FIG. 7. With reference to that figure the experimental rig comprises two emulsion holding hoppers ANE1 and ANE2. Two metering pumps PC Pump 1 and PC Pump 2 supply streams of the emulsions into an inter-changeable mixing head. The mass flow of the individual fluid streams is set up by calibration of the metering pumps and cross-checking against the total mass flow via into the inter-changeable mixing head. Blending is done in a continuous manner in the closed pipe of a interchangeable mixing head module.

(46) The inter-changeable mixing head is comprised of two parts. The first part has two separate inlet channels for the entry of each emulsion stream and a baffle just before the entrance to the first static mixer element to ensure separation of the individual streams in the mixing section. The inter-changeable mixing head is 50 mm diameter and length of 228 mm.

(47) A Kenics static mixer (having 3 elements; see FIG. 8) was used for layering the void sensitized emulsion into the void-free high density emulsion continuum through laminar flow of two continuous streams of the emulsions. Laminar mixing is achieved by repeated division, transposition and recombination of liquid layers around a static mixer. In this way, the components of emulsion to be mixed are spread into a large number of layers. A clearly defined and uniform shear field is generated through mixing. Addition of further static mixer elements (for example No 4, 5& 6) reduces the thickness of the layers produced.

(48) The density change of the gassing emulsion was determined in a plastic cup of known mass and volume. The emulsion was initially filled to the top of the cup and leveled off. As the gassing reaction progressed, the emulsion rose out of the top of the cup and was leveled off periodically and weighed. The density was determined by dividing the mass of emulsion in the cup by the cup volume. Charges larger than 70 mm in diameter were initiated with a single 400 g Pentex PPP booster, whist smaller charges were initiated with a 150 g Pentex H booster. Velocity of detonation (VOD) was determined using an MREL Handitrap VOD recorder.

(49) Procedure for Determining Distribution Function

(50) Product samples were delivered from the pump rig described above into a 100 mm diameter cylindrical plastic container consisting of a 150 mm tall base, a 10 mm sample slice and a 30 mm tall top slice, as shown in FIG. 12. The three slices were joined together with masking tape to produce a cylinder which was filled to the top with emulsion. After filling, the upper 30 mm slice was removed and the emulsion scraped level on the top of the 10 mm slice with a flat stiff blade. A clear perspex plate was placed over the top of the 10 mm slice, and the entire container inverted. The 150 mm section was then removed, leaving the 10 mm section filled with emulsion sitting on the flat perspex plate. The emulsion was allowed to gas to completion prior to photography. The slice was illuminated from underneath using an x-ray viewer and photographed from above with a digital camera.

(51) The photograph of the product structure was analysed using the Image) program. A rectangular section of the image was selected for distribution function analysis. FIG. 13 shows a typical image after processing and the rectangular section selected for DF analysis. The software enabled automatic detection of the bubbles in the photograph and produced a table showing the x and y position of the voids, the void perimeters and the void area. This data was exported to Mathcad for radial distribution function analysis.

(52) The distribution function (DF) plots the fraction of emulsion that is within a given distance of a void surface. The DF procedure involved calculating the distance from each emulsion pixel to the nearest bubble surface. This program calculated the distance between a pixel and all of the bubble surfaces and returned the distance to the nearest bubble surface. The procedure was then repeated for all emulsion pixels. The frequency of emulsion points residing within a given distance to a bubble surface was then determined and plotted as a cumulative distribution. The differential of the cumulative fraction with respect to distance was also plotted against the cumulative fraction (also referred to as distribution function rate).

Example 1

Gassed Emulsion at 1.22 g/cm3

(53) This example demonstrates the performance of conventional gassed emulsion with random void distribution at a density of 1.22 g/cm.sup.3.

(54) The starting emulsion at a density of 1.32 g/cm.sup.3 was delivered by a progressive cavity pump at a rate of 3 kg/min. A 4% mass sodium nitrite solution was injected into the flowing emulsion stream at a rate of 16 g/min by means of a gasser (gear) pump and dispersed in a series of static mixers. 1 m long cardboard tubes with internal diameters ranging from 40 to 180 mm were loaded with emulsion and allowed to gas. Charges were fired once the sample cup reached the target density of 1.22 g/cm.sup.3.

(55) A sample of the emulsion was taken for DF analysis according to the procedure described above. FIG. 16 shows the void positions for conventional gassed emulsion. The cumulative distribution function is plotted in FIG. 17 and the differential plotted in FIG. 18. The cumulative distribution function shows a steep curve, with the cumulative fraction rising to unity within a distance of approximately 0.7 mm. This indicates that 100% of the emulsion in the sample lies within 0.7 mm of a void surface. The differential of the distribution function (FIG. 18) shows a characteristic convex shape.

(56) The VOD ranged from 2.9 km/s for the 70 mm diameter charge to 4.3 km/s at 180 mm. Charges smaller than 70 mm failed to sustain detonation. The VOD results are illustrated in FIG. 9.

Example 2

MOE25 at 1.22 g/cm3

(57) This example demonstrates the performance of MOE25, i.e. a mixture of emulsion with 25% mass sensitized and 75% unsensitized emulsion and was prepared using the apparatus described above.

(58) The base emulsion (density 1.32 g/cm.sup.3) was delivered by two progressive cavity pumps, PC1 and PC2. The base emulsion formulation was identical to Example 1 and was the same for both pumps. PC1 pumped ungassed emulsion at a flow rate of 4 kg/min. PC2 delivered emulsion at 1.3 kg/min with gasser (4% NaNO.sub.2 solution) injected by a gasser (gear) pump. The emulsion was blended by a static mixer consisting of three helical mixing elements and loaded into cardboard tubes with internal diameters ranging from 70 to 180 mm. The gassed emulsion target density was 0.99 g/cm.sup.3 providing an overall density of 1.22 g/cm.sup.3 for the mixture of gassed and ungassed emulsion.

(59) A sample of the emulsion was taken for DF analysis according to the procedure described above. The void positions in this sample are shown in FIG. 15. The cumulative distribution function is plotted in FIG. 17 and the differential plotted in FIG. 18. Compared to the gassed emulsion curve, the cumulative distribution for MOE25 exhibits a significantly shallower slope, with a long tail that extends out to a distance of approximately 6 mm. The plot of the distribution function differential can also be distinguished from the gassed emulsion sample by the presence of a point of inflexion in the curve and a concave tail section.

(60) These changes in the distribution function and differential distribution function are reflected in the VOD measurements, shown in FIG. 10. The VOD ranged from 2.5 km/s for the 90 mm charge to 3.7 km/s at 180 mm, a significant reduction relative to conventional gassed emulsion described in Example 1. Charges with diameters smaller than 90 mm failed to sustain detonation. The reduced VOD in this example demonstrates the effect of the distribution function and differential distribution function on the shock/heave energy ratio. The shallower slope of this distribution function, the point of inflexion and the concave portion of the differential distribution function result in increased heave energy relative to conventional gassed emulsion, which exhibits a steeply sloped distribution function and convex differential distribution function.

Example 3

MOE50 at 1.22 g/cm3

(61) This example demonstrates the performance of MOE50, i.e. a mixture of emulsion with 50% mass gassed and 50% ungassed emulsion.

(62) MOE50 was prepared using the apparatus mentioned in Example 2. The base emulsion (density 1.32 g/cm.sup.3) was delivered by two progressive cavity pumps, PC1 and PC2 and was identical to the previous two examples. PC1 pumped ungassed emulsion at a flow rate of 3 kg/min. PC2 delivered emulsion at 3 kg/min with gasser (4% NaNO.sub.2 solution) injected by a gasser (gear) pump. The emulsion was blended by a static mixer consisting of three helical mixing elements and loaded into cardboard tubes with internal diameters ranging from 70 to 180 mm. The gassed emulsion target density was 1.13 g/cm.sup.3 providing an overall density of 1.22 g/cm.sup.3 for the mixture of gassed and ungassed emulsion.

(63) A sample of the emulsion was taken for DF analysis according to the procedure described above. The void positions in this sample are shown in FIG. 14. The cumulative distribution function is plotted in FIG. 17 and the differential plotted in FIG. 18. The MOE50 sample exhibits a distribution function curve with an intermediate slope between conventional gassed emulsion and the MOE25 described in Examples 1 and 2, respectively. Likewise, the differential distribution function lies between the conventional gassed emulsion and MOE25, exhibiting a point of inflexion and a slight concave section.

(64) The VOD ranged from 2.8 km/s for the 80 mm charge to 3.9 km/s at 180 mm and is illustrated in FIG. 11. Charges with diameters smaller than 80 mm failed to sustain detonation. VOD results for MOE50 were between those of gassed emulsion and MOE25. This demonstrates that this explosive, with intermediate distribution and differential distribution functions relative to Examples 1 and 2, exhibits an intermediate shock/heave energy ratio. Importantly, the example demonstrates that the present invention allows tailoring of explosive performance (i.e. shock/heave energy balance) to suit different blasting applications by suitable selection of a distribution function template at the same overall explosive density. That is, the invention allows manipulation of the shock/heave energy balance whilst maintaining the same total energy of the explosive.

(65) The DF of an emulsion with a perfectly random distribution of voids, and that of two idealized (simulated) MoEs with the sensitized and unsensitized regions arranged as alternating flat sheets in which no voids have strayed into the unsensitized region, is shown in FIG. 19. The simulated emulsion DF is almost identical to the experimental emulsion. The idealised MoEs however have sharper corner turning in the graphs than the experimental MoEs. The replacement of the sharper corners of the idealized MoE with the smoother concave graphs of the experimental emulsion results from a slightly more diffuse distribution of the voids into the unsensitized regions in the experimental emulsion compared to the simulated MoEs.

(66) Noting the results obtained in the examples, the present invention also provides explosive compositions comprising sensitizing voids distributed in a liquid energetic materials that are believed to be new per se and that exhibit a characteristic distribution function that is different from known void-sensitized explosive formulations, such as emulsions, watergels and slurry formulations. More specifically, for the explosive compositions of the inventions a plot of distribution function rate versus distribution function includes a point of inflexion, and possibly a concave portion. In contrast corresponding plots for conventional explosive formulations exhibit a characteristic domed profile. As explained above, in this context the distribution function (or distance froth void function) is defined as the fraction of the liquid that is within a given distance from any void surface, and the distribution function rate is defined as the differential of the distribution function with respect to the distance from any void surface.

(67) In an embodiment, for the explosive compositions a plot of distribution function rate versus distribution function comprises a region extending from a distribution function value of 0% to between 10% and 90%, and wherein after the dome region the distribution function rate is between 1% and 50% of the peak of the dome. Preferably, the dome region extends from a distribution function value of 0% to between 15% and 85%, and in the region after the dome the distribution function rate is between 1.5% and 35% of the peak of the dome, Even more preferably the dome region extends from a distribution function value of 0% to between 20% and 80%, and in the region after the dome the distribution function rate is between 2% and 20% of the peak of the dome.