- Email: [email protected]
Filling the gap between hypervelocity and low velocity impacts
Journal Pre-proof
Filling the Gap between Hypervelocity and Low Velocity Impacts W. Arnold , T. Hartmann , E. Rottenkolber PII: DOI: Reference:
S0734-743X(19)30733-X https://doi.org/10.1016/j.ijimpeng.2020.103531 IE 103531
To appear in:
International Journal of Impact Engineering
Please cite this article as: W. Arnold , T. Hartmann , E. Rottenkolber , Filling the Gap between Hypervelocity and Low Velocity Impacts, International Journal of Impact Engineering (2020), doi: https://doi.org/10.1016/j.ijimpeng.2020.103531
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier Ltd.
Highlight
Experiments with standard military shaped charges on the hypervelocity range and with elongated STANAG-projectiles on the low velocity range respectively were known and published elsewhere
New Experiments with a newly designed Simplified Shaped Charges (SSC) in between in the intermediate velocity range (to close the gap) were conducted and a new initiation model for the whole range from hypervelocity to the low velocity is proposed
A Simplified Shaped Charge (SSC) was introduced generating slower and thicker jets closer to STANAG projectiles
New ERL-curves were obtained with the new SSC-charges using the same setup as in previous studies
The behavior at the root points of the ERL-curves (ERL = VI) with the STANAG-projectiles and the SCCs with thick jets were conclusive and showed good agreement
The ERL-curves for the SCC-jets exhibit extensive and ramp-like ERL-slopes, whereas the STANAG-curves have a rather steep slope
The “S = constant-rule” is not valid and the stimulus S = v²∙d is inappropriate for predicting the explosive reaction of a charge on jet / projectile impacts thus requiring a new initiation model
A new linear unified initiation model for the whole velocity regime (hyper- to low velocities) is proposed: vcrit = A - B∙d
1
Filling the Gap between Hypervelocity and Low Velocity Impacts W. Arnolda, T. Hartmannb, E. Rottenkolberb a
MBDA-TDW Gesellschaft für verteidigungstechnische Wirksysteme mbH, Hagenauer Forst, D86529 Schrobenhausen, Germany, b
NUMERICS GmbH, Mozartring 6, D-85238 Petershausen, Germany
Abstract During more than one decade of studying initiation phenomenology numerous papers were published by the authors. A multitude of experimental data on hypervelocity impact initiation of plastic bonded high explosive charges by shaped charge jets (SCJ) and some data in the ordnance velocity impact regime obtained with STANAG projectiles and explosively formed projectiles (EFP) were generated. Amongst other findings the results showed that the established assumption that the critical stimulus is constant was wrong and that a new initiation model is needed taking the new results into account. Towards such a new model further investigation was necessary trying to make a link between the initiation phenomenology of shaped charge jet impacts in the hypervelocity regime and of projectile impacts in the lower velocity regime. To bridge the existing data gap a new approach was taken applying newly designed “simplified shaped charges” (SSC). This paper describes the way to this new approach and summarizes the related test results as well as their implications.
2
Keywords: Initiation; Plastic Bonded Explosive; Simplified Shaped Charge Jet (SSCJ); STANAG Projectile; Explosive Reaction Level (ERL); Penetration Mode and Semi-Impact Mode initiation; Unified Initiation Model;
1.
Introduction During more than one decade of studying initiation phenomenology numerous papers at the
previous HVIS and other symposia ([1] - [12]) were published. Most of them dealt with the hypervelocity impact initiation of plastic bonded high explosive charges by shaped charge jets (SCJ) and a few ones in the ordnance velocity impact regime with STANAG projectiles and explosively formed projectiles (EFP) ([2] & [11]). A recent finding of the investigations of shaped charge jet (SCJ) attacks suggests that the critical stimulus S = v²∙d (v = SCJ / projectile velocity; d = SCJ / projectile diameter) for the initiation of a munition can no longer be seen as a constant (S ≠ const.) ([11] & [10]). Also, known equations like Jacobs-Roslund [13], are not capable to describe low velocity and hypervelocity impacts with the same parameter set. Therefore, a new initiation model is needed taking these findings into account [10]. On the way to such a new model further investigation was necessary trying to make a link between the initiation phenomenology of shaped charge jet impacts in the hypervelocity regime and of projectile impacts in the lower velocity regime. Instead of using elongated STANAG projectiles [16] to further bridge this gap a Simplified Shaped Charge (SSC) was introduced in [12], which generates a slower and thicker jet thus allowing more flexibility in the jet / projectile parameters. The new SSC charge used a caliber of 145 mm with a copper liner of 90° angle. In this study an SSC with a wider liner angle of 110° will be applied to generate even thicker and slower jets.
3
2.
Hypervelocity and Low Velocity Regimes
2.1. Hypervelocity Regime (SCJ) The hypervelocity regime with jet tip velocities causing increasing reaction inside the explosive from 2200 m/s up to 5800 m/s and generated by high performance military shaped charges was already thoroughly investigated and published in [10]. Nevertheless, for the reason of completeness these tests and their results are briefly summarized in the following. Thereby, the maximum caliber of the tested (COTS = components off-the-shelf) shaped charges was 200 mm.
Shaped Charge Characterization In order to compare different ERL-curves (Explosive Reaction Level) obtained with different shaped charges (different calibers) it is very important that the SC jets are very carefully characterized and that the calibration curves (S = v2d vs. barrier thickness P) are really soundly generated. The procedure of this characterization and the determination of the calibration curves have already been presented in detail in [10]. At the end of an extensive evaluation process the necessary calibration curves were obtained. They are is shown in Figure 1. A calibration curve is an essential basis for the experimental initiation trials and the subsequent assessment of the results: it can be used as a prerequisite for the test planning to adapt the stimulus S = v 2d of the attacking SCJ by varying the barrier thickness P, and is indispensable for relating the resulting ERL to the SCJ stimulus in the ERL-curve as shown later.
4
Fig. 1. Calibration curves of the investigated shaped charges [10].
Experimental Setup In all conducted trials the so-called standard charge filled with the TDW high explosive KS32 (HMX/PB 85/15, ρ = 1.64 g/cm³) was used for the investigations. The charge consists of a high explosive cylinder with 100 mm in diameter and 200 mm in length and a mild steel casing with 10 mm thickness and screwed lids on both sides (standard threads) as shown in Figure 2. A mild steel barrier with varying thickness P was applied to tune the SCJ-stimulus S = v2·d. Thereby, the stand-off to the steel barrier P was kept constant at two calibers (e.g. 400 mm for the shown SC-200, with Cal. 200 mm), which ensures a continuous jet even at low stimuli. To assess high order reactions of the charge, all test setups additionally comprised a 4 mm thick mild steel witness plate (1 m x 2 m, not shown in Figure 2) in a distance of 2 m from the test charge.
5
Fig. 2. Sketch of the test set-up with SC, steel barrier and standard charge – here with the SC200 (Cal. 200 mm).
Experimental Results A complete ERL-curve (from ERL = VI to ERL = I) was determined by incrementally varying the barrier thickness P and applying the corresponding calibration curve from Figure 1. As an example, the results (ERL vs. Stimulus S) for the SC-200 are presented in Figure 3. The photo insets (with the respective test numbers) give a vivid illustration of the increasing reaction level of the standard charge when the stimuli are continuously increased.
6
Fig. 3. ERL-curve for the SC-200 (photo insets demonstrating assessment of the ERLs). The experiments were assessed according to STANAG 4439 [14] introducing six Explosive Reaction Levels (ERL) from “Propulsion” (ERL = VI) to “Detonation” (ERL = I). Lower level reactions were estimated from the high explosive remnants of the charge and from broken pieces / fragments of the casing. The high-level reaction modes (ERL = I & II = (partial) detonation) were determined from the fragment pattern caused in the event, i.e. by evaluation of the witness plate. Figure 4 summarizes the results obtained with five (in service) military SC charges (taken from [10]) with increasing SC caliber and SCJ diameter. The ERL-curves for the varying stimulus S = v²∙d clearly illustrate the above-mentioned new finding: the so far accepted assumption that the critical stimulus S = const. - as e.g. described in [15] - is clearly not valid. Further, the “strange” behavior with the SC-200 crossing three other ERL-curves indicates that the stimulus S is not an appropriate parameter to compare the ERL caused by different shaped charges. When plotted over the velocity v (as will be shown later) the data becomes much more consistent.
7
Fig. 4. ERL-curves obtained with five military SCs.
Experimental Assessment High ERL = I / Detonation: The stimulus found to be required to reach ERL = I is in complete agreement with the results already achieved and discussed in [10]. The experiments show no agreement with the v2d-rule (S = v2d = const.) and thus do not agree with the STANAG 4526. In the higher ERL region all five curves do not overlap, but appear as five distinct individual curves. Also, the order of the curves is striking: the higher the SC caliber (i.e. the larger the SCJdiameter) the more insensitive is the reaction of the standard charge. The disagreement with the v2d-rule (and accordingly STANAG 4526) is highlighted by plotting the determined threshold stimuli S for ERL = I (and ERL = VI respectively) against the SC caliber in Figure 5. The threshold stimulus S = v2d is obviously not constant but significantly increasing with increasing SC caliber (except for ERL = VI with SC-200, which is discussed below).
8
Fig. 5. Upper and lower threshold stimuli vs. SC caliber. Low ERL = VI / Beginning of Reactions: The beginning of a reaction (ERL = VI) is more difficult to estimate. But even when considering this fact, a new quality of behavior can be observed with the SC-200: a crossover of the curves at lower ERL. Numerical simulations showed that while increasing the SC caliber (and SCJ diameter) and reducing the SCJ velocity, a growing bulging effect of the casing at the exit side can be noticed. Therefore, an influence of the steel casing could be assumed. Considering that the lower level initiation of the high explosive occurs in the “penetration mode” at the exit side, an influence of this bulging on the virtual initiation volume in front of the jet may be expected. This will be further discussed later below.
Another potential influence is the reflection of the shock wave from the inner side of the steel casing at the exit side. This potential influence was studied by trials under varying impact angles of the SCJ thus varying the reflection angles as well. At the same time the shot line length 9
through the high explosive (HE = KS32) varies. As indicated in Figure 6, three different impact angles θ were applied, leading to three different shot line lengths s HE through the HE:
Impact angle θ = 0°, shot line length sHE = 100 mm
Impact angle θ = 30°, shot line length sHE = 115 mm
Impact angle θ = 60°, shot line length sHE = 200 mm
A 44 mm caliber Shaped Charge (SC 44) was used for all test series. Based on a calibration curve “barrier thickness P vs. jet stimulus v2d” available from previous studies, the barrier thickness was varied to keep the line of sights (LOS) and thus the stimulus S constant. The standoff s/o was 90 mm (~2 calibers) in all tests.
Fig. 6. Standard test charge with steel barrier P and three different SCJ impact angles θ.
Several test charges were assembled for each angle variation series. The results are shown in Figure 7. The stimulus necessary to achieve a certain ERL decreases with increasing impact
10
angle. Or in other words: applying the same stimulus, the reactivity of KS32 is higher with a larger impact angle. If we assume that the shock wave reflection from the steel casing at the exit side is the most significant parameter, an opposed trend would be expected.
The superposition between
incoming and reflected shock wave is highest at 0° and lowest at 60°. That means the highest reactivity of the HE at the exit side would be expected at 0° - which is not the case. Obviously, the length of the perforated HE is much more significant and thus the determining factor for the reaction levels. The longer the SCJ path through the HE, the more time is available especially for building up higher reactivity levels.
Fig. 7. Summary of impact angle variation results for KS32.
11
2.2. Ordnance Velocity Regime (STANAG Projectile) There were no (COTS) high performance military shaped charges available generating jets with a diameter of more than about 8 mm (SC-200). Besides, threats from fast fragments are typically represented by STANAG projectiles (projectile diameter: 14.3 mm) [16]. Therefore, these projectiles were used for the lower velocity regime.
From Shaped Charges to STANAG Projectiles While making the transition from SCJ towards STANAG projectile impacts, several changes of initiation phenomena are expected. This transition process starts from a continuous copper jet and ends up with the standard steel STANAG projectile with L/D = 1. The individual steps include:
•
Continuous Cu liner SCJ with velocity gradient,
•
Particulated Cu liner SCJ with different jet particle velocities,
•
Modified STANAG projectile: elongated (L/D > 1) and material changed to Cu,
•
Modified STANAG projectile made of steel but elongated (L/D > 1),
•
Standard steel STANAG projectile (L/D = 1).
In the first step from a continuous to a particulated (Cu liner) SCJ, a first change in the initiation phenomenology could be observed [10]: the second SCJ particle (and all the following ones) now hits moving, but practically bare high explosive (KS32) leading to a higher sensitivity (larger ERL). Taking the next step, only one elongated (L/D > 1) Cu projectile will be hitting the test charge instead of multiple SCJ particles. Then the next step towards the L/D = 1 STANAG steel projectile makes the transition from Cu to steel, and the final step is done when using an original STANAG steel projectile [16] with L/D = 1. Such a short projectile can erode very quickly while
12
penetrating the charge. This might lead to a higher required velocity to initiate the charge, as the initiation now - due to the quick erosion process - must take place at the entry side of the charge (instead of the exit side). In any case an at least partial transition from a penetration mode to an impact mode initiation (see section 4) must occur, which might be accompanied by a further change in required projectile velocities.
This short summary of the initiation phenomena induced by the transition from SCJs towards STANAG projectiles already shows that it was not fully clear or foreseeable what would happen when this transition is actually executed. Therefore, further experimental trials with original STANAG projectiles (L/D = 1) and elongated ones made out of steel and copper were conducted [11]. But before the trials were started numerical simulations were conducted to support the setup and interpretation of the experimental results.
Numerical Simulations Numerical simulations with a commercial hydrocode SPEED [18] were applied to determine the minimum projectile length required to perforate the complete charge and to investigate the erosion process, the velocity reduction during the penetration of the charge, acceptable maximum yaw angle (< 6°), impact velocity vs. projectile velocity behind the steel casing, casing materials etc. The original STANAG projectile (L/D = 1) and elongated ones (L/D = 2, 2.5 & 3) were studied (Figure 8). Besides steel projectiles also Cu projectiles (making the link to copper SC jets) were regarded. The high explosive was modelled with an inert PBX-simili.
13
Fig. 8. Simulation models of the STANAG projectile (L/D = 1) and elongated ones (L/D = 2, 2.5 & 3). A sequence of the penetration of an elongated Cu-projectile (L/D = 3) is shown in Figure 9. The impact velocity (on the steel casing) was 2000 m/s. After perforation of the 10 mm thick steel casing of the charge the projectile is already strongly eroded (ca. 40%). When arriving at the middle of the charge the penetration velocity is close to 1000 m/s. In the simulation the length of the projectile is sufficient to reach the rear side of the charge but it is fully eroded and cannot completely perforate the casing.
Fig. 9. Sequence from 0 – 200 µs of the penetration of an elongated Cu-projectile (L/D = 3).
14
Experimental Setup Firing tests with STANAG-projectiles and their derivatives: L/D = 1 & 3, made from steel and copper materials were planned and carried out. The trials were conducted at the Fraunhofer Ernst-Mach-Institute (EMI) in Germany. A powder gun was used by the EMI to accelerate the projectiles. The projectiles were mounted into a standard sabot which was stripped-off before hitting the target. The upper velocity limit was about 2600 m/s. The larger two-stage light-gas gun (LGG) allowing much higher velocities was out of operation at that time. The test setup was designed to be as close as possible to the setup used for the SCJ trials (see e.g. [10]). The pictures in Figure 10 (taken at the EMI impact chamber) illustrate the setup. The projectile enters the chamber through the opening on the left. A high-speed video camera records the projectile´s flight path and permits to determine the impact velocity, the projectile pitch and the impact point on the charge. A mirror is applied to observe the shot from an orthogonal direction and to control projectile yaw. The Aluminum witness plate (2 mm thick) in the background is used to detect higher ERL levels (ERL = I & II). For the lower level reactions, the casing fragments and the KS32 residues respectively were collected and used for the ERL assessment as in the SCJ trials. The close-up on the right of Figure 10 shows the standard charge mounted in the projectile´s shot line.
15
Fig. 10. Test setup at the EMI impact chamber (left) and a close-up of the TDW standard charge (right).
Experimental Results The diagram in Figure 11 combines the data of the SCJ trials (see above) with those from the STANAG projectile trials. Different from Figure 4, the jet / projectile velocities (instead of S = v2·d) were used as x-axis. The dashed lines from ERL = III to ERL = I indicate an straight forward extrapolation to higher velocities, which was necessary since the velocity of the powder gun was limited to about 2600 m/s. The STANAG projectile data fit well into the overall ERL trends and the data generally look consistent. It should be noted that this is not at all the case when the ERL is plotted over the stimulus S = v²·d. Hence, also this result confirms that the stimulus S = v²·d is not an appropriate parameter for describing the initiation behavior.
16
Fig. 11. Comparison of the SCJ results with the STANAG projectile results.
Despite the fact that the STANAG data fit in very well, there are still a couple of questions that must remain unanswered for the moment, e.g.: •
Why are ERL-slopes different between SCJ and STANAG projectile results?
•
What is the reason for the ramp-like ERL-slope especially for the SC-200?
•
What is the reason for the gap between SCJ and STANAG projectile results?
•
What is the influence on the transition between impact mode and penetration mode initiation?
In order to try to answer these questions and to take the next step towards a new unified initiation model further trials were necessary. Instead of using larger caliber shaped charges ( > 200 mm, COTS were not available) for getting thicker jet diameters (d > 8 mm) and lower jet velocities a new ”Simplified Shaped Charge” (SSC) with a thicker jet and jet tip velocities in the intermediate regime (between high performance / hypervelocity SCs and STANAG projectiles in
17
the ordnance velocity regime) was designed.
3.
Intermediate Velocity Regime
3.1. SSC-Jet Characterization The first SSC charge was based on an SC-145 with a liner angle of 90°, a design intentionally being very close to the SC-200 to make the results comparable afterwards. In this work additionally a second SSC charge with the same caliber of 145 mm but a wider liner angle of 110° was selected (Figure 12) to further increase the SSC-jet diameter while reducing the jet velocity and jet length.
Fig. 12. Simplified Shaped Charge (SSC) design and test setup: caliber 145 mm without wave shaper and liner angle of 110°.
The design and the jet characterization of both SSC were realized by numerical simulations with SPEED and flash X-ray picture evaluation, respectively. The idea was to start with a jet diameter comparable to the largest of the already tested performance shaped charges (SC-200) in order to study potential differences in the initiation behavior with the new simplified shaped charge design. Based on existing COTS shaped charges, a number of different designs with varying liner thickness and angle were numerically simulated until a promising design was found. Figure 13 exemplarily shows the numerical simulation model (top), the simulated SC-jet (middle) and a 18
sketch of the final design (bottom) for the SSC-145 with 90°. The jet diameter in the simulation was slightly above 8 mm and thus very close to the SC-200 [10].
Fig. 13. Numerical simulations with the Simplified shaped charges with liner angle 90° and thickness 4.35 mm: simulation model (top), simulated SC-jet (middle) and sketch of the final SSC-145 design (bottom).
The newly manufactured SSC-145 charges were fired in front of an X-ray facility capturing the SSC-jet at two different exposure times. The FRX picture of the SSC-145 with 90° with a closeup of the jet is shown in Figure 14. A copper wire was used as reference for the evaluation procedure. For the evaluation of the FXR picture the TDW in-house software EDI [18] was used. Applying a Sobel Filter edge detection algorithm, EDI automatically detects the single particles and determines the jet characteristics (particle velocities, mass / diameter profiles, drift velocities,
19
etc.). Thereby, the evaluation results were in very good agreement with the numerical simulations.
Fig. 14. X-ray picture of the SSC-145 jet with 90° at two different exposure times together with a reference copper wire.
The measured average SSC-145 (110°) jet diameter was 11.3 mm and thus (as intended) larger than that of the SC-145 (90°) with only 8.3 mm and already very close to the STANAG projectile diameter of 14.3 mm. Hence, the design again fully met the objective of generating a slower and thicker jet. Figure 15 shows the relation of the SC jet diameter and the SC caliber for the five military COTS shaped charges. The diagram also indicates the (extrapolated) necessary caliber of military SCs to yield the jet diameters of the new SSC-145 charges (210 mm & 280 mm).
20
Fig. 15. Averaged SC jet diameters vs. SC caliber for all applied shaped charges. 3.2. SSC-Jet Initiation Trials and ERL-Results Like in the earlier tests (see above & [1] – [11]) the standard charge filled with KS32 was used. Again, a mild steel barrier with varying thickness P was applied to tune the SSCJ-stimulus S = v²·d. Thereby, the stand-off to the steel barrier P was kept constant at two calibers (290 mm). The initiation test results were again assessed according to STANAG 4439 [14] determining the respective ERLs. A complete ERL-curve (from ERL = VI/V to ERL = I) was determined by incrementally varying the barrier thickness. The results (ERL vs. Stimulus S = v²·d) exemplarily for the investigated SSC-145 (110°) are presented in Figure 16. The photo insets (with the respective test numbers) again illustrate the increasing reaction level of the standard charge when the stimulus is continuously increased.
21
Fig. 16. ERL-curve for the SSC-145 (110°) (photo insets demonstrating assessment of the ERLs).
As already discussed above the stimulus S = v²·d is not the appropriate parameter for comparing results. It led to the crossing of the ERL-curves in Figure 4 and the results are inconsistent when the STANAG projectile data is added. Therefore, the initiation ERL-curves with all SSC-145 (90° & 110°) results were plotted over the velocity in Figure 17.
22
Fig. 17. Comparison of all ERL-curves achieved so far with the SCs, SSCs and the STANAG projectiles.
Comparing the SSC-145 (90°) with the SC-200 curves it can be seen that both curves are nearly identical. The initiation behavior obtained with the high-performance SC-200 could thus be reproduced with the simplified design (despite the shorter and slower jet) requiring significantly thinner barriers P to tune the jet.
The results of the ERL-curve achieved with the SSC-145 (110°) shaped charge compared with the ones of the SSC-145 (90°) and the SC-200 in Figure 17 show several interesting features:
•
The SSC-145 (110°) ERL-curve shows almost the same flat slope and plateau as the curves for SC-145 (90°) and SC-200,
•
It is shifted to the left towards lower velocities getting closer to the STANAG results,
•
The distinctive plateau of the SC-145 (110°) seems to be one ERL-level higher than with the SC-145 (90°): ERL = V -> IV,
•
At the high-level reactions (ERL = I/II) it seems that a partial detonation took place (less holes in the witness plate),
•
At the low-level reactions (ERL = VI/V) it seems that the lowest level reaction (VI) could not be achieved (instead VI/V).
The “virtual” caliber of a “representative military SC” would amount to about 280 mm (see Figure 15). Comparing the velocity shifts of the other shaped charges: SC-115 -> SC-150 -> SC-
23
200 results in about the same order of magnitude - obviously caused by the growing SCJ diameters. The reason for the occurrence and shift of the distinctive plateau to higher reaction levels from ERL = V to ERL = IV for the SC-145 (110°) might be the strongly increased jet diameter leading to higher mechanical (and chemical?) lateral effects. Considering this effect, the simply straight forward extrapolation for curves of the elongated STANAG projectiles to higher order reactions unthinkingly conducted in Figure 11 above (dashed lines) is presumably not correct. There might be another plateau also for the STANAG projectile (see the indicated arrow in Figure 17) instead of the steep-sloped curves leading to ERL = I. The reason why no clear lowest reaction level ERL = VI was achieved could be also those high lateral effects allowing no “low damage mode” (see the broken casings for the STANAG projectiles in [11]).
4.
Initiation: from Semi-Impact Mode to Penetration Mode In order to understand the phenomenology of SCJ initiation but also to make the link to
projectile initiation, it is important to understand the potential transition from an Impact Mode to a Penetration Mode initiation when the SCJ stimulus (SCJ-velocity and/or SCJ-diameter) decreases. In [4] this transition was measured and discussed in detail and it is only briefly summarized here. The setup and the charge geometry for the trials in [4] were rather different from the setup generally used in this work (Figure 2). In [4] an air-gap or a thin steel-plate was used to realize the mentioned mode transition. For further details of this work it is recommended to read the paper in Ref. [4].
24
The impact mode initiation is typical for SCJs attacking a bare high explosive (HE). Due to the high pressure loads at the impact side an instantaneous initiation is enforced. The transition to the penetration mode takes place while either a gap between casing and HE is closed successively or the casing thickness is increased gradually. By doing this the very high pressure loads directly introduced into the uncased HE changes into an attenuated ramp-like pressure load. Thus, the initiation mode coming from an Impact Mode also changes into a Penetration Mode as shown in Figure 18 taken from [4] for the PBX KS32 and for the TNT-bonded Comp B (RDX/TNT 65/35, ρ = 1.71 g/cc). The test setup is sketched in the figure as well. The run distance to detonation Δs was measured by a rotating mirror camera in streak mode [4].
Fig. 18. Comparison of run distances to detonation ∆s for Comp B and KS32 while varying the air gap between barrier and HE charge (P = 100 mm).
In the trials the thickness of the steel barrier was not varied but held constant at P = 100 mm (that means the stimulus S = v²∙d is constant as well) but the air gap x between the barrier and the HE charge was varied. The variation of the air gap between 15 mm and about 500 mm
25
shows two differences between a more insensitive PBX KS32 and the sensitive TNT-bonded Comp B:
Comp B reacts more sensitive: shorter ∆s at small air gaps.
The HE intrinsic so called “corner turning radius” also correlates with the sensitivity: for the sensitive Comp B it is very small (2 – 4 mm) whereas for the more insensitive KS32 it is much larger. Thus, the break-through of detonation wave can be measured only after about ∆s ~ 20 mm.
In both cases the transition between the two modes can be clearly observed. The air gap was decreased starting with the standard 15 mm and then reduced to 4 mm, 2 mm and 1 mm. The bulging effect of the steel barrier influences the initiation behavior clearly at 2 – 4 mm air gap with a strongly increasing run distance, i.e. a typical transition to Penetration Mode initiation. At 1 mm finally, no detonation at all can be achieved within the 100 mm sample length. With an air gap of 4 mm on the other hand the run distance is already short and shows a clear transition to Impact Mode initiation. The results demonstrate impressively the increasing importance of the bulging effect and the decreasing importance of the impact shock while reducing the air gap. Thus, an enforced transition from Impact Mode to Penetration Mode initiation takes place. With the “artificial” air gap the transition was enforced: starting with a smooth ramp-like shock loading (cased charge) then changing to a very strong and steep shock wave loading (bare charge). But also, with the cased charge such a transition can be enforced when the SCJ stimulus S = v²∙d is increased continuously. In [7] a “trajectory measuring method” was introduced allowing to measure (indirectly) the run distance to detonation ∆s with the help of a rotating mirror camera system. The setup is comparable to the setup-sketch in Figure 18 (without
26
air gap). The camera measures continuously the detonation break-throughs of the detonation front on the surface of the test sample (diameter 48 mm, length 100 mm). In Figure 19 examples taken from [7] are shown. On the left side the calculated trajectories (blue straight lines) in the test sample are shown which are perpendicular to the propagating detonation front (red curves) which can be constructed with the help of the calculated trajectories (the SCJ was shot into the sample from the bottom). By the rotating mirror camera, the first breakthrough of detonation wave is measured. With the trajectory method the “inner” run to detonation distance ∆s (point-source where the detonation front arises) can be calculated. Now the stimulus S = v²∙d was increased step by step from beginning with 93 mm3/µs2 up to 154 mm3/µs2 (right side). It can be clearly seen that the inner run distance ∆s gets shorter and shorter with increasing stimulus (the SCJ is shot into the sample from the left side). The behavior is like an enforced transition from Penetration Mode initiation to an “Semi-Impact” Mode initiation which happens at the entrance side of the charge.
Fig. 19. Calculated trajectories (straight lines) and detonation wave fronts (curved lines) for a typical SCJI test (sample diameter 48 mm, length 100 mm; left). Trajectory and detonation front reconstruction from tests with varying stimulus S. 27
In [11] a “Semi-Impact Mode” could be ascertained with the steel STANAG projectile (L/D = 1) and the 10 mm thick steel casing of the standard charge as shown in Figure 20. The numerical simulations confirmed that the original steel STANAG projectile with diameter D = 14.3 mm and L/D = 1 was not able to perforate the charge completely - due to the erosion of the projectile. Therefore, only reactions at the impact side are possible (or no reactions at all). This can be seen in Figure 20 on the left showing the simulation and the ripped open casing with the rupture at the entry side, obviously caused by reaction pressures. When the L/D of the projectile is increased from L/D = 1 to 3, the projectile is able to perforate the charge and a low-level reaction (ERL = VI) can be observed – now at the exit side (Figure 20, right).
Fig. 20. Numerical simulation and ripped open steel casing for a Semi-Impact Mode initiation with a steel L/D = 1 projectile (left) and a Penetration Mode initiation with a steel L/D = 3 projectile (right).
Up to now, the charges attacked by an SCJ reacted at lowest level (ERL = VI) with a Penetration Mode initiation at the exit side of the charge after 100 mm run to reaction distance (nucleation and growth of reactions). Figure 21 demonstrates typical examples (taken from [10]) of increasing exit holes. 28
Fig. 21. Exit hole diameters in the steel casings achieved with the SC-115 and SC-200 shaped charges.
As discussed above the SC-145 (110°) shaped charge produces a rather short and thick jet. For the lowest stimuli (ERL = V - VI) the diameter and length of the initiating jet particle is already very close to the STANAG projectile of 14.3 mm. It therefore could be expected that this “last” particle behaves rather similar to a STANAG projectile. And indeed, the reaction in the standard charge starts at the entry side as shown in Figure 22. The reaction was strong enough to cause a distinctive rupture of the casing at the entry side. It thus seems that also with the shaped charge SC-145 (110°) a transition from a Penetration Mode initiation to a Semi-Impact Mode initiation took place. The small keyhole at the exit side was probably caused by remaining parts of the jet.
29
Fig. 22. Semi-Impact Mode initiation with the SC-145 (110°) leading to the ripping open of the casing at the entry side. Exit keyhole from residual jet particles.
Despite the comparable behavior of “STANAG projectiles” and “short and thick SC jets”, the deeper reason for the extensive (about 2000 m/s) ramp-like ERL-slopes paired with the wide plateau (about 1000 m/s) in Figure 17 for shaped charges delivering thicker jets (SC-200 and SC-145 (90° & 110°)) cannot conclusively be answered.
5.
Unified Initiation Model It was already realized in [11] that the stimulus S = v²∙d is not an appropriate parameter for
the description of the observed initiation phenomena. Neither the so far assumed “v²d-rule” (S = v²∙d = const.) nor another conclusive relation to the parameter S can be observed for the different SC and projectile initiation tests. Moreover, none of these relations can coherently link the initiation from SC jets to those from projectiles. Figure 23 therefore exhibits the correlation of jet / projectile diameter and the velocities leading to higher and lower order reactions including the new results. The results strongly suggest that a linear relationship:
30
v = A - B∙d
(1)
is more appropriate to describe the initiation phenomena correctly. A linear relation also covers both the SCJ results and the results obtained in tests with STANAG projectiles, which is not possible with relations based on the stimulus S. For the standard charge investigated in this paper, the experimental results yield “averaged” values (average between ERL = I and ERL = VI) of A = 5500 m/s and B = 2.2∙105 1/s.
Fig. 23. Approximately linear initiation behavior of the data for ERL = I and ERL = VI.
6.
Conclusions Shaped charge initiation trials were carried out with different shaped charges (SCs) with
different calibers form 44 mm to theoretical 280 mm. The standard charge known from earlier tests and filled with the PBX KS32 (HMX/PB 85/15, ρ = 1.64 g/cm³) was attacked by the shaped charges jets. In the course of this extensive trial campaign the stimuli S = v2d (v = SCJ-velocity, d = SCJ-diameter) of the different shaped charges were varied by varying the barrier plate 31
thickness P and determining the reaction behavior (ERL = explosive reaction level) of the standard charge. The experiments revealed that the different shaped charges with varying SC-calibers result in distinct ERL-curves. This is in clear disagreement with the “v2d-rule” (S = v2d = const.) and thus in disagreement with the STANAG 4526. Further investigating this finding and particularly trying to make a link between the initiation phenomenology in shaped charge and projectile attacks, shaped charge variations of the newly developed simplified SSC-145 with liner angles of 90° and 110° were successfully designed and tested. The jets were well characterized by using flash X-ray pictures of the jets as in earlier studies. The slower and thicker part of the SSC jet was already comparable to the STANAG projectile with a diameter of 14.3 mm. The behavior at the root points of the ERL-curves (ERL = VI) with the STANAG projectiles and the SCCs with thick jets was found to be very comparable. Also, a transition of the Penetration Mode initiation to a Semi-Impact Mode initiation could be observed. However, the question why the ERL-curves for the SCC jets exhibit extensive and ramp-like ERL-slopes paired with a plateau, whereas the STANAG-curves seem to have an assumed rather steep slope still remains unanswered. But maybe the elongated STANAG projectile curves exhibit such a plateau as well when the impact velocity is further increased? This will have to be further investigated in the future. Nevertheless, also the new test results confirmed that the “S = v²∙d = constant-rule” is not valid and that the stimulus S = v²∙d is inappropriate to predict the explosive reaction of a charge subjected to jet / projectile impact. Instead, the entirety of results exhibited a nearly linear relationship between velocity and diameter of the jet / projectile. Hence, a new linear initiation model is proposed: vcrit = A - B∙d. This relation might also finally link the initiation phenomena
32
observed in shaped charge with those in projectile attacks. At the moment it can at least be stated that the achieved empirical data are comprehensive enough to draw the conclusion that the linear relationship between the SC-velocity v and the SC-diameter d describes the measured data much better than the recent constant v²∙d-relationship. Nevertheless, this new relationship has to be further confirmed by additional trials in the future. Also, an explanation of the physical and / or chemical meaning of the two parameters A and B (if there is one) has to be postponed into the future when more data and new ideas become available. Declaration of interests ☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Author Statement The paper is a comprehensive summary of our work on initiation of a PBX with shape charge jets for the last decade. It shows a numerous new finding, including a new linear initiation model covering shaped charge jets and (STANAG or EFP) projectiles.
Acknowledgements The authors would like to thank the BAAINBw Team K 1.2 at Koblenz for the funding of this study.
33
References [1]
W. Arnold, E. Rottenkolber, “Sensitivity of High Explosives against Shaped Charge Jets”, Insensitive Munitions & Energetic Materials (IMEMTS) 2007, Miami, USA
[2]
W. Arnold, “High Explosive Initiation by High Velocity Projectile Impact”, HVIS 2010, Freiburg, GE
[3]
W. Arnold, M. Graswald, “Shaped Charge Jet Initiation of High Explosives equipped with an Explosive Train”, IMEMTS 2010, Munich, GE
[4]
W. Arnold, E. Rottenkolber, “Shaped Charge Jet Initiation Phenomena of Plastic Bonded High Ex-plosives”, IMEMTS 2012, Las Vegas, USA
[5]
W. Arnold, E. Rottenkolber, “High Explosive Initiation Behavior by Shaped Charge Jet Impacts”, Hypervelocity Impact Symposium HVIS 2012, Baltimore, USA
[6]
W. Arnold, E. Rottenkolber, “Significant Charge Parameters influencing the Shaped Charge Jet Initiation”, IMEMTS 2013, San Diego, USA
[7]
W. Arnold, E. Rottenkolber, T. Hartmann, “Analysis of Shock and Jet Initiation Tests of High Explosives”, 19th Int. Symposium on Detonation ISD 2014, San Francisco, USA
[8]
W. Arnold, E. Rottenkolber, T. Hartmann, “Challenging v²d”, IMEMTS 2015, Rome, IT
[9]
W. Arnold, E. Rottenkolber, T. Hartmann, “Testing and Modeling the Initiation of Insensitive Explosives by Projectile Impact”, 11th EUROPYRO International Seminar 2015, Toulouse, France
[10]
W. Arnold, T. Hartmann, E. Rottenkolber, “Towards a Unified Initiation Model”, IMEMTS 2016, Nashville, TN, USA
[11]
W. Arnold, T. Hartmann, E. Rottenkolber, “Filling the Gap between the Initiation Behavior of Shaped Charge Jets and Fragments”, IMEMTS 2018, Portland, OR, USA
[12]
W. Arnold, T. Hartmann, E. Rottenkolber, “Initiation Phenomenology from Hypervelocity to 34
Low Velocity Impacts”, 16th Detonation Symposium 2018, Cambridge, MD, USA [13]
T.P. Liddiard, L.A. Roslund, “Projectile/Fragment Impact Sensitivity of Explosives” NSWC TR 89-184, Naval Surface Warfare Center, Dahlgren, VA, June 1993
[14]
STANAG 4439 (Edition 3), “Policy for Introduction and Assessment of Insensitive Munitions (MURAT)”, 2006
[15]
STANAG 4526 (Edition 2), “Shaped Charge Jet Munitions Test Procedure”, 2004
[16]
STANAG 4496 (Edition 1), “Fragment Impact, Munitions Test Procedure”, 2006
[17]
SPEED 2.2 (Shock Physics Explicit Eulerian Dynamics), “Theory & User Manual”, NUMERICS, 2015
[18]
EDI 4.2 (Edge Detection in Images) “Theory & User Manual”, TDW/NUMERICS, 2014
35
Filling the Gap between Hypervelocity and Low Velocity Impacts W. Arnold , T. Hartmann , E. Rottenkolber PII: DOI: Reference:
S0734-743X(19)30733-X https://doi.org/10.1016/j.ijimpeng.2020.103531 IE 103531
To appear in:
International Journal of Impact Engineering
Please cite this article as: W. Arnold , T. Hartmann , E. Rottenkolber , Filling the Gap between Hypervelocity and Low Velocity Impacts, International Journal of Impact Engineering (2020), doi: https://doi.org/10.1016/j.ijimpeng.2020.103531
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier Ltd.
Highlight
Experiments with standard military shaped charges on the hypervelocity range and with elongated STANAG-projectiles on the low velocity range respectively were known and published elsewhere
New Experiments with a newly designed Simplified Shaped Charges (SSC) in between in the intermediate velocity range (to close the gap) were conducted and a new initiation model for the whole range from hypervelocity to the low velocity is proposed
A Simplified Shaped Charge (SSC) was introduced generating slower and thicker jets closer to STANAG projectiles
New ERL-curves were obtained with the new SSC-charges using the same setup as in previous studies
The behavior at the root points of the ERL-curves (ERL = VI) with the STANAG-projectiles and the SCCs with thick jets were conclusive and showed good agreement
The ERL-curves for the SCC-jets exhibit extensive and ramp-like ERL-slopes, whereas the STANAG-curves have a rather steep slope
The “S = constant-rule” is not valid and the stimulus S = v²∙d is inappropriate for predicting the explosive reaction of a charge on jet / projectile impacts thus requiring a new initiation model
A new linear unified initiation model for the whole velocity regime (hyper- to low velocities) is proposed: vcrit = A - B∙d
1
Filling the Gap between Hypervelocity and Low Velocity Impacts W. Arnolda, T. Hartmannb, E. Rottenkolberb a
MBDA-TDW Gesellschaft für verteidigungstechnische Wirksysteme mbH, Hagenauer Forst, D86529 Schrobenhausen, Germany, b
NUMERICS GmbH, Mozartring 6, D-85238 Petershausen, Germany
Abstract During more than one decade of studying initiation phenomenology numerous papers were published by the authors. A multitude of experimental data on hypervelocity impact initiation of plastic bonded high explosive charges by shaped charge jets (SCJ) and some data in the ordnance velocity impact regime obtained with STANAG projectiles and explosively formed projectiles (EFP) were generated. Amongst other findings the results showed that the established assumption that the critical stimulus is constant was wrong and that a new initiation model is needed taking the new results into account. Towards such a new model further investigation was necessary trying to make a link between the initiation phenomenology of shaped charge jet impacts in the hypervelocity regime and of projectile impacts in the lower velocity regime. To bridge the existing data gap a new approach was taken applying newly designed “simplified shaped charges” (SSC). This paper describes the way to this new approach and summarizes the related test results as well as their implications.
2
Keywords: Initiation; Plastic Bonded Explosive; Simplified Shaped Charge Jet (SSCJ); STANAG Projectile; Explosive Reaction Level (ERL); Penetration Mode and Semi-Impact Mode initiation; Unified Initiation Model;
1.
Introduction During more than one decade of studying initiation phenomenology numerous papers at the
previous HVIS and other symposia ([1] - [12]) were published. Most of them dealt with the hypervelocity impact initiation of plastic bonded high explosive charges by shaped charge jets (SCJ) and a few ones in the ordnance velocity impact regime with STANAG projectiles and explosively formed projectiles (EFP) ([2] & [11]). A recent finding of the investigations of shaped charge jet (SCJ) attacks suggests that the critical stimulus S = v²∙d (v = SCJ / projectile velocity; d = SCJ / projectile diameter) for the initiation of a munition can no longer be seen as a constant (S ≠ const.) ([11] & [10]). Also, known equations like Jacobs-Roslund [13], are not capable to describe low velocity and hypervelocity impacts with the same parameter set. Therefore, a new initiation model is needed taking these findings into account [10]. On the way to such a new model further investigation was necessary trying to make a link between the initiation phenomenology of shaped charge jet impacts in the hypervelocity regime and of projectile impacts in the lower velocity regime. Instead of using elongated STANAG projectiles [16] to further bridge this gap a Simplified Shaped Charge (SSC) was introduced in [12], which generates a slower and thicker jet thus allowing more flexibility in the jet / projectile parameters. The new SSC charge used a caliber of 145 mm with a copper liner of 90° angle. In this study an SSC with a wider liner angle of 110° will be applied to generate even thicker and slower jets.
3
2.
Hypervelocity and Low Velocity Regimes
2.1. Hypervelocity Regime (SCJ) The hypervelocity regime with jet tip velocities causing increasing reaction inside the explosive from 2200 m/s up to 5800 m/s and generated by high performance military shaped charges was already thoroughly investigated and published in [10]. Nevertheless, for the reason of completeness these tests and their results are briefly summarized in the following. Thereby, the maximum caliber of the tested (COTS = components off-the-shelf) shaped charges was 200 mm.
Shaped Charge Characterization In order to compare different ERL-curves (Explosive Reaction Level) obtained with different shaped charges (different calibers) it is very important that the SC jets are very carefully characterized and that the calibration curves (S = v2d vs. barrier thickness P) are really soundly generated. The procedure of this characterization and the determination of the calibration curves have already been presented in detail in [10]. At the end of an extensive evaluation process the necessary calibration curves were obtained. They are is shown in Figure 1. A calibration curve is an essential basis for the experimental initiation trials and the subsequent assessment of the results: it can be used as a prerequisite for the test planning to adapt the stimulus S = v 2d of the attacking SCJ by varying the barrier thickness P, and is indispensable for relating the resulting ERL to the SCJ stimulus in the ERL-curve as shown later.
4
Fig. 1. Calibration curves of the investigated shaped charges [10].
Experimental Setup In all conducted trials the so-called standard charge filled with the TDW high explosive KS32 (HMX/PB 85/15, ρ = 1.64 g/cm³) was used for the investigations. The charge consists of a high explosive cylinder with 100 mm in diameter and 200 mm in length and a mild steel casing with 10 mm thickness and screwed lids on both sides (standard threads) as shown in Figure 2. A mild steel barrier with varying thickness P was applied to tune the SCJ-stimulus S = v2·d. Thereby, the stand-off to the steel barrier P was kept constant at two calibers (e.g. 400 mm for the shown SC-200, with Cal. 200 mm), which ensures a continuous jet even at low stimuli. To assess high order reactions of the charge, all test setups additionally comprised a 4 mm thick mild steel witness plate (1 m x 2 m, not shown in Figure 2) in a distance of 2 m from the test charge.
5
Fig. 2. Sketch of the test set-up with SC, steel barrier and standard charge – here with the SC200 (Cal. 200 mm).
Experimental Results A complete ERL-curve (from ERL = VI to ERL = I) was determined by incrementally varying the barrier thickness P and applying the corresponding calibration curve from Figure 1. As an example, the results (ERL vs. Stimulus S) for the SC-200 are presented in Figure 3. The photo insets (with the respective test numbers) give a vivid illustration of the increasing reaction level of the standard charge when the stimuli are continuously increased.
6
Fig. 3. ERL-curve for the SC-200 (photo insets demonstrating assessment of the ERLs). The experiments were assessed according to STANAG 4439 [14] introducing six Explosive Reaction Levels (ERL) from “Propulsion” (ERL = VI) to “Detonation” (ERL = I). Lower level reactions were estimated from the high explosive remnants of the charge and from broken pieces / fragments of the casing. The high-level reaction modes (ERL = I & II = (partial) detonation) were determined from the fragment pattern caused in the event, i.e. by evaluation of the witness plate. Figure 4 summarizes the results obtained with five (in service) military SC charges (taken from [10]) with increasing SC caliber and SCJ diameter. The ERL-curves for the varying stimulus S = v²∙d clearly illustrate the above-mentioned new finding: the so far accepted assumption that the critical stimulus S = const. - as e.g. described in [15] - is clearly not valid. Further, the “strange” behavior with the SC-200 crossing three other ERL-curves indicates that the stimulus S is not an appropriate parameter to compare the ERL caused by different shaped charges. When plotted over the velocity v (as will be shown later) the data becomes much more consistent.
7
Fig. 4. ERL-curves obtained with five military SCs.
Experimental Assessment High ERL = I / Detonation: The stimulus found to be required to reach ERL = I is in complete agreement with the results already achieved and discussed in [10]. The experiments show no agreement with the v2d-rule (S = v2d = const.) and thus do not agree with the STANAG 4526. In the higher ERL region all five curves do not overlap, but appear as five distinct individual curves. Also, the order of the curves is striking: the higher the SC caliber (i.e. the larger the SCJdiameter) the more insensitive is the reaction of the standard charge. The disagreement with the v2d-rule (and accordingly STANAG 4526) is highlighted by plotting the determined threshold stimuli S for ERL = I (and ERL = VI respectively) against the SC caliber in Figure 5. The threshold stimulus S = v2d is obviously not constant but significantly increasing with increasing SC caliber (except for ERL = VI with SC-200, which is discussed below).
8
Fig. 5. Upper and lower threshold stimuli vs. SC caliber. Low ERL = VI / Beginning of Reactions: The beginning of a reaction (ERL = VI) is more difficult to estimate. But even when considering this fact, a new quality of behavior can be observed with the SC-200: a crossover of the curves at lower ERL. Numerical simulations showed that while increasing the SC caliber (and SCJ diameter) and reducing the SCJ velocity, a growing bulging effect of the casing at the exit side can be noticed. Therefore, an influence of the steel casing could be assumed. Considering that the lower level initiation of the high explosive occurs in the “penetration mode” at the exit side, an influence of this bulging on the virtual initiation volume in front of the jet may be expected. This will be further discussed later below.
Another potential influence is the reflection of the shock wave from the inner side of the steel casing at the exit side. This potential influence was studied by trials under varying impact angles of the SCJ thus varying the reflection angles as well. At the same time the shot line length 9
through the high explosive (HE = KS32) varies. As indicated in Figure 6, three different impact angles θ were applied, leading to three different shot line lengths s HE through the HE:
Impact angle θ = 0°, shot line length sHE = 100 mm
Impact angle θ = 30°, shot line length sHE = 115 mm
Impact angle θ = 60°, shot line length sHE = 200 mm
A 44 mm caliber Shaped Charge (SC 44) was used for all test series. Based on a calibration curve “barrier thickness P vs. jet stimulus v2d” available from previous studies, the barrier thickness was varied to keep the line of sights (LOS) and thus the stimulus S constant. The standoff s/o was 90 mm (~2 calibers) in all tests.
Fig. 6. Standard test charge with steel barrier P and three different SCJ impact angles θ.
Several test charges were assembled for each angle variation series. The results are shown in Figure 7. The stimulus necessary to achieve a certain ERL decreases with increasing impact
10
angle. Or in other words: applying the same stimulus, the reactivity of KS32 is higher with a larger impact angle. If we assume that the shock wave reflection from the steel casing at the exit side is the most significant parameter, an opposed trend would be expected.
The superposition between
incoming and reflected shock wave is highest at 0° and lowest at 60°. That means the highest reactivity of the HE at the exit side would be expected at 0° - which is not the case. Obviously, the length of the perforated HE is much more significant and thus the determining factor for the reaction levels. The longer the SCJ path through the HE, the more time is available especially for building up higher reactivity levels.
Fig. 7. Summary of impact angle variation results for KS32.
11
2.2. Ordnance Velocity Regime (STANAG Projectile) There were no (COTS) high performance military shaped charges available generating jets with a diameter of more than about 8 mm (SC-200). Besides, threats from fast fragments are typically represented by STANAG projectiles (projectile diameter: 14.3 mm) [16]. Therefore, these projectiles were used for the lower velocity regime.
From Shaped Charges to STANAG Projectiles While making the transition from SCJ towards STANAG projectile impacts, several changes of initiation phenomena are expected. This transition process starts from a continuous copper jet and ends up with the standard steel STANAG projectile with L/D = 1. The individual steps include:
•
Continuous Cu liner SCJ with velocity gradient,
•
Particulated Cu liner SCJ with different jet particle velocities,
•
Modified STANAG projectile: elongated (L/D > 1) and material changed to Cu,
•
Modified STANAG projectile made of steel but elongated (L/D > 1),
•
Standard steel STANAG projectile (L/D = 1).
In the first step from a continuous to a particulated (Cu liner) SCJ, a first change in the initiation phenomenology could be observed [10]: the second SCJ particle (and all the following ones) now hits moving, but practically bare high explosive (KS32) leading to a higher sensitivity (larger ERL). Taking the next step, only one elongated (L/D > 1) Cu projectile will be hitting the test charge instead of multiple SCJ particles. Then the next step towards the L/D = 1 STANAG steel projectile makes the transition from Cu to steel, and the final step is done when using an original STANAG steel projectile [16] with L/D = 1. Such a short projectile can erode very quickly while
12
penetrating the charge. This might lead to a higher required velocity to initiate the charge, as the initiation now - due to the quick erosion process - must take place at the entry side of the charge (instead of the exit side). In any case an at least partial transition from a penetration mode to an impact mode initiation (see section 4) must occur, which might be accompanied by a further change in required projectile velocities.
This short summary of the initiation phenomena induced by the transition from SCJs towards STANAG projectiles already shows that it was not fully clear or foreseeable what would happen when this transition is actually executed. Therefore, further experimental trials with original STANAG projectiles (L/D = 1) and elongated ones made out of steel and copper were conducted [11]. But before the trials were started numerical simulations were conducted to support the setup and interpretation of the experimental results.
Numerical Simulations Numerical simulations with a commercial hydrocode SPEED [18] were applied to determine the minimum projectile length required to perforate the complete charge and to investigate the erosion process, the velocity reduction during the penetration of the charge, acceptable maximum yaw angle (< 6°), impact velocity vs. projectile velocity behind the steel casing, casing materials etc. The original STANAG projectile (L/D = 1) and elongated ones (L/D = 2, 2.5 & 3) were studied (Figure 8). Besides steel projectiles also Cu projectiles (making the link to copper SC jets) were regarded. The high explosive was modelled with an inert PBX-simili.
13
Fig. 8. Simulation models of the STANAG projectile (L/D = 1) and elongated ones (L/D = 2, 2.5 & 3). A sequence of the penetration of an elongated Cu-projectile (L/D = 3) is shown in Figure 9. The impact velocity (on the steel casing) was 2000 m/s. After perforation of the 10 mm thick steel casing of the charge the projectile is already strongly eroded (ca. 40%). When arriving at the middle of the charge the penetration velocity is close to 1000 m/s. In the simulation the length of the projectile is sufficient to reach the rear side of the charge but it is fully eroded and cannot completely perforate the casing.
Fig. 9. Sequence from 0 – 200 µs of the penetration of an elongated Cu-projectile (L/D = 3).
14
Experimental Setup Firing tests with STANAG-projectiles and their derivatives: L/D = 1 & 3, made from steel and copper materials were planned and carried out. The trials were conducted at the Fraunhofer Ernst-Mach-Institute (EMI) in Germany. A powder gun was used by the EMI to accelerate the projectiles. The projectiles were mounted into a standard sabot which was stripped-off before hitting the target. The upper velocity limit was about 2600 m/s. The larger two-stage light-gas gun (LGG) allowing much higher velocities was out of operation at that time. The test setup was designed to be as close as possible to the setup used for the SCJ trials (see e.g. [10]). The pictures in Figure 10 (taken at the EMI impact chamber) illustrate the setup. The projectile enters the chamber through the opening on the left. A high-speed video camera records the projectile´s flight path and permits to determine the impact velocity, the projectile pitch and the impact point on the charge. A mirror is applied to observe the shot from an orthogonal direction and to control projectile yaw. The Aluminum witness plate (2 mm thick) in the background is used to detect higher ERL levels (ERL = I & II). For the lower level reactions, the casing fragments and the KS32 residues respectively were collected and used for the ERL assessment as in the SCJ trials. The close-up on the right of Figure 10 shows the standard charge mounted in the projectile´s shot line.
15
Fig. 10. Test setup at the EMI impact chamber (left) and a close-up of the TDW standard charge (right).
Experimental Results The diagram in Figure 11 combines the data of the SCJ trials (see above) with those from the STANAG projectile trials. Different from Figure 4, the jet / projectile velocities (instead of S = v2·d) were used as x-axis. The dashed lines from ERL = III to ERL = I indicate an straight forward extrapolation to higher velocities, which was necessary since the velocity of the powder gun was limited to about 2600 m/s. The STANAG projectile data fit well into the overall ERL trends and the data generally look consistent. It should be noted that this is not at all the case when the ERL is plotted over the stimulus S = v²·d. Hence, also this result confirms that the stimulus S = v²·d is not an appropriate parameter for describing the initiation behavior.
16
Fig. 11. Comparison of the SCJ results with the STANAG projectile results.
Despite the fact that the STANAG data fit in very well, there are still a couple of questions that must remain unanswered for the moment, e.g.: •
Why are ERL-slopes different between SCJ and STANAG projectile results?
•
What is the reason for the ramp-like ERL-slope especially for the SC-200?
•
What is the reason for the gap between SCJ and STANAG projectile results?
•
What is the influence on the transition between impact mode and penetration mode initiation?
In order to try to answer these questions and to take the next step towards a new unified initiation model further trials were necessary. Instead of using larger caliber shaped charges ( > 200 mm, COTS were not available) for getting thicker jet diameters (d > 8 mm) and lower jet velocities a new ”Simplified Shaped Charge” (SSC) with a thicker jet and jet tip velocities in the intermediate regime (between high performance / hypervelocity SCs and STANAG projectiles in
17
the ordnance velocity regime) was designed.
3.
Intermediate Velocity Regime
3.1. SSC-Jet Characterization The first SSC charge was based on an SC-145 with a liner angle of 90°, a design intentionally being very close to the SC-200 to make the results comparable afterwards. In this work additionally a second SSC charge with the same caliber of 145 mm but a wider liner angle of 110° was selected (Figure 12) to further increase the SSC-jet diameter while reducing the jet velocity and jet length.
Fig. 12. Simplified Shaped Charge (SSC) design and test setup: caliber 145 mm without wave shaper and liner angle of 110°.
The design and the jet characterization of both SSC were realized by numerical simulations with SPEED and flash X-ray picture evaluation, respectively. The idea was to start with a jet diameter comparable to the largest of the already tested performance shaped charges (SC-200) in order to study potential differences in the initiation behavior with the new simplified shaped charge design. Based on existing COTS shaped charges, a number of different designs with varying liner thickness and angle were numerically simulated until a promising design was found. Figure 13 exemplarily shows the numerical simulation model (top), the simulated SC-jet (middle) and a 18
sketch of the final design (bottom) for the SSC-145 with 90°. The jet diameter in the simulation was slightly above 8 mm and thus very close to the SC-200 [10].
Fig. 13. Numerical simulations with the Simplified shaped charges with liner angle 90° and thickness 4.35 mm: simulation model (top), simulated SC-jet (middle) and sketch of the final SSC-145 design (bottom).
The newly manufactured SSC-145 charges were fired in front of an X-ray facility capturing the SSC-jet at two different exposure times. The FRX picture of the SSC-145 with 90° with a closeup of the jet is shown in Figure 14. A copper wire was used as reference for the evaluation procedure. For the evaluation of the FXR picture the TDW in-house software EDI [18] was used. Applying a Sobel Filter edge detection algorithm, EDI automatically detects the single particles and determines the jet characteristics (particle velocities, mass / diameter profiles, drift velocities,
19
etc.). Thereby, the evaluation results were in very good agreement with the numerical simulations.
Fig. 14. X-ray picture of the SSC-145 jet with 90° at two different exposure times together with a reference copper wire.
The measured average SSC-145 (110°) jet diameter was 11.3 mm and thus (as intended) larger than that of the SC-145 (90°) with only 8.3 mm and already very close to the STANAG projectile diameter of 14.3 mm. Hence, the design again fully met the objective of generating a slower and thicker jet. Figure 15 shows the relation of the SC jet diameter and the SC caliber for the five military COTS shaped charges. The diagram also indicates the (extrapolated) necessary caliber of military SCs to yield the jet diameters of the new SSC-145 charges (210 mm & 280 mm).
20
Fig. 15. Averaged SC jet diameters vs. SC caliber for all applied shaped charges. 3.2. SSC-Jet Initiation Trials and ERL-Results Like in the earlier tests (see above & [1] – [11]) the standard charge filled with KS32 was used. Again, a mild steel barrier with varying thickness P was applied to tune the SSCJ-stimulus S = v²·d. Thereby, the stand-off to the steel barrier P was kept constant at two calibers (290 mm). The initiation test results were again assessed according to STANAG 4439 [14] determining the respective ERLs. A complete ERL-curve (from ERL = VI/V to ERL = I) was determined by incrementally varying the barrier thickness. The results (ERL vs. Stimulus S = v²·d) exemplarily for the investigated SSC-145 (110°) are presented in Figure 16. The photo insets (with the respective test numbers) again illustrate the increasing reaction level of the standard charge when the stimulus is continuously increased.
21
Fig. 16. ERL-curve for the SSC-145 (110°) (photo insets demonstrating assessment of the ERLs).
As already discussed above the stimulus S = v²·d is not the appropriate parameter for comparing results. It led to the crossing of the ERL-curves in Figure 4 and the results are inconsistent when the STANAG projectile data is added. Therefore, the initiation ERL-curves with all SSC-145 (90° & 110°) results were plotted over the velocity in Figure 17.
22
Fig. 17. Comparison of all ERL-curves achieved so far with the SCs, SSCs and the STANAG projectiles.
Comparing the SSC-145 (90°) with the SC-200 curves it can be seen that both curves are nearly identical. The initiation behavior obtained with the high-performance SC-200 could thus be reproduced with the simplified design (despite the shorter and slower jet) requiring significantly thinner barriers P to tune the jet.
The results of the ERL-curve achieved with the SSC-145 (110°) shaped charge compared with the ones of the SSC-145 (90°) and the SC-200 in Figure 17 show several interesting features:
•
The SSC-145 (110°) ERL-curve shows almost the same flat slope and plateau as the curves for SC-145 (90°) and SC-200,
•
It is shifted to the left towards lower velocities getting closer to the STANAG results,
•
The distinctive plateau of the SC-145 (110°) seems to be one ERL-level higher than with the SC-145 (90°): ERL = V -> IV,
•
At the high-level reactions (ERL = I/II) it seems that a partial detonation took place (less holes in the witness plate),
•
At the low-level reactions (ERL = VI/V) it seems that the lowest level reaction (VI) could not be achieved (instead VI/V).
The “virtual” caliber of a “representative military SC” would amount to about 280 mm (see Figure 15). Comparing the velocity shifts of the other shaped charges: SC-115 -> SC-150 -> SC-
23
200 results in about the same order of magnitude - obviously caused by the growing SCJ diameters. The reason for the occurrence and shift of the distinctive plateau to higher reaction levels from ERL = V to ERL = IV for the SC-145 (110°) might be the strongly increased jet diameter leading to higher mechanical (and chemical?) lateral effects. Considering this effect, the simply straight forward extrapolation for curves of the elongated STANAG projectiles to higher order reactions unthinkingly conducted in Figure 11 above (dashed lines) is presumably not correct. There might be another plateau also for the STANAG projectile (see the indicated arrow in Figure 17) instead of the steep-sloped curves leading to ERL = I. The reason why no clear lowest reaction level ERL = VI was achieved could be also those high lateral effects allowing no “low damage mode” (see the broken casings for the STANAG projectiles in [11]).
4.
Initiation: from Semi-Impact Mode to Penetration Mode In order to understand the phenomenology of SCJ initiation but also to make the link to
projectile initiation, it is important to understand the potential transition from an Impact Mode to a Penetration Mode initiation when the SCJ stimulus (SCJ-velocity and/or SCJ-diameter) decreases. In [4] this transition was measured and discussed in detail and it is only briefly summarized here. The setup and the charge geometry for the trials in [4] were rather different from the setup generally used in this work (Figure 2). In [4] an air-gap or a thin steel-plate was used to realize the mentioned mode transition. For further details of this work it is recommended to read the paper in Ref. [4].
24
The impact mode initiation is typical for SCJs attacking a bare high explosive (HE). Due to the high pressure loads at the impact side an instantaneous initiation is enforced. The transition to the penetration mode takes place while either a gap between casing and HE is closed successively or the casing thickness is increased gradually. By doing this the very high pressure loads directly introduced into the uncased HE changes into an attenuated ramp-like pressure load. Thus, the initiation mode coming from an Impact Mode also changes into a Penetration Mode as shown in Figure 18 taken from [4] for the PBX KS32 and for the TNT-bonded Comp B (RDX/TNT 65/35, ρ = 1.71 g/cc). The test setup is sketched in the figure as well. The run distance to detonation Δs was measured by a rotating mirror camera in streak mode [4].
Fig. 18. Comparison of run distances to detonation ∆s for Comp B and KS32 while varying the air gap between barrier and HE charge (P = 100 mm).
In the trials the thickness of the steel barrier was not varied but held constant at P = 100 mm (that means the stimulus S = v²∙d is constant as well) but the air gap x between the barrier and the HE charge was varied. The variation of the air gap between 15 mm and about 500 mm
25
shows two differences between a more insensitive PBX KS32 and the sensitive TNT-bonded Comp B:
Comp B reacts more sensitive: shorter ∆s at small air gaps.
The HE intrinsic so called “corner turning radius” also correlates with the sensitivity: for the sensitive Comp B it is very small (2 – 4 mm) whereas for the more insensitive KS32 it is much larger. Thus, the break-through of detonation wave can be measured only after about ∆s ~ 20 mm.
In both cases the transition between the two modes can be clearly observed. The air gap was decreased starting with the standard 15 mm and then reduced to 4 mm, 2 mm and 1 mm. The bulging effect of the steel barrier influences the initiation behavior clearly at 2 – 4 mm air gap with a strongly increasing run distance, i.e. a typical transition to Penetration Mode initiation. At 1 mm finally, no detonation at all can be achieved within the 100 mm sample length. With an air gap of 4 mm on the other hand the run distance is already short and shows a clear transition to Impact Mode initiation. The results demonstrate impressively the increasing importance of the bulging effect and the decreasing importance of the impact shock while reducing the air gap. Thus, an enforced transition from Impact Mode to Penetration Mode initiation takes place. With the “artificial” air gap the transition was enforced: starting with a smooth ramp-like shock loading (cased charge) then changing to a very strong and steep shock wave loading (bare charge). But also, with the cased charge such a transition can be enforced when the SCJ stimulus S = v²∙d is increased continuously. In [7] a “trajectory measuring method” was introduced allowing to measure (indirectly) the run distance to detonation ∆s with the help of a rotating mirror camera system. The setup is comparable to the setup-sketch in Figure 18 (without
26
air gap). The camera measures continuously the detonation break-throughs of the detonation front on the surface of the test sample (diameter 48 mm, length 100 mm). In Figure 19 examples taken from [7] are shown. On the left side the calculated trajectories (blue straight lines) in the test sample are shown which are perpendicular to the propagating detonation front (red curves) which can be constructed with the help of the calculated trajectories (the SCJ was shot into the sample from the bottom). By the rotating mirror camera, the first breakthrough of detonation wave is measured. With the trajectory method the “inner” run to detonation distance ∆s (point-source where the detonation front arises) can be calculated. Now the stimulus S = v²∙d was increased step by step from beginning with 93 mm3/µs2 up to 154 mm3/µs2 (right side). It can be clearly seen that the inner run distance ∆s gets shorter and shorter with increasing stimulus (the SCJ is shot into the sample from the left side). The behavior is like an enforced transition from Penetration Mode initiation to an “Semi-Impact” Mode initiation which happens at the entrance side of the charge.
Fig. 19. Calculated trajectories (straight lines) and detonation wave fronts (curved lines) for a typical SCJI test (sample diameter 48 mm, length 100 mm; left). Trajectory and detonation front reconstruction from tests with varying stimulus S. 27
In [11] a “Semi-Impact Mode” could be ascertained with the steel STANAG projectile (L/D = 1) and the 10 mm thick steel casing of the standard charge as shown in Figure 20. The numerical simulations confirmed that the original steel STANAG projectile with diameter D = 14.3 mm and L/D = 1 was not able to perforate the charge completely - due to the erosion of the projectile. Therefore, only reactions at the impact side are possible (or no reactions at all). This can be seen in Figure 20 on the left showing the simulation and the ripped open casing with the rupture at the entry side, obviously caused by reaction pressures. When the L/D of the projectile is increased from L/D = 1 to 3, the projectile is able to perforate the charge and a low-level reaction (ERL = VI) can be observed – now at the exit side (Figure 20, right).
Fig. 20. Numerical simulation and ripped open steel casing for a Semi-Impact Mode initiation with a steel L/D = 1 projectile (left) and a Penetration Mode initiation with a steel L/D = 3 projectile (right).
Up to now, the charges attacked by an SCJ reacted at lowest level (ERL = VI) with a Penetration Mode initiation at the exit side of the charge after 100 mm run to reaction distance (nucleation and growth of reactions). Figure 21 demonstrates typical examples (taken from [10]) of increasing exit holes. 28
Fig. 21. Exit hole diameters in the steel casings achieved with the SC-115 and SC-200 shaped charges.
As discussed above the SC-145 (110°) shaped charge produces a rather short and thick jet. For the lowest stimuli (ERL = V - VI) the diameter and length of the initiating jet particle is already very close to the STANAG projectile of 14.3 mm. It therefore could be expected that this “last” particle behaves rather similar to a STANAG projectile. And indeed, the reaction in the standard charge starts at the entry side as shown in Figure 22. The reaction was strong enough to cause a distinctive rupture of the casing at the entry side. It thus seems that also with the shaped charge SC-145 (110°) a transition from a Penetration Mode initiation to a Semi-Impact Mode initiation took place. The small keyhole at the exit side was probably caused by remaining parts of the jet.
29
Fig. 22. Semi-Impact Mode initiation with the SC-145 (110°) leading to the ripping open of the casing at the entry side. Exit keyhole from residual jet particles.
Despite the comparable behavior of “STANAG projectiles” and “short and thick SC jets”, the deeper reason for the extensive (about 2000 m/s) ramp-like ERL-slopes paired with the wide plateau (about 1000 m/s) in Figure 17 for shaped charges delivering thicker jets (SC-200 and SC-145 (90° & 110°)) cannot conclusively be answered.
5.
Unified Initiation Model It was already realized in [11] that the stimulus S = v²∙d is not an appropriate parameter for
the description of the observed initiation phenomena. Neither the so far assumed “v²d-rule” (S = v²∙d = const.) nor another conclusive relation to the parameter S can be observed for the different SC and projectile initiation tests. Moreover, none of these relations can coherently link the initiation from SC jets to those from projectiles. Figure 23 therefore exhibits the correlation of jet / projectile diameter and the velocities leading to higher and lower order reactions including the new results. The results strongly suggest that a linear relationship:
30
v = A - B∙d
(1)
is more appropriate to describe the initiation phenomena correctly. A linear relation also covers both the SCJ results and the results obtained in tests with STANAG projectiles, which is not possible with relations based on the stimulus S. For the standard charge investigated in this paper, the experimental results yield “averaged” values (average between ERL = I and ERL = VI) of A = 5500 m/s and B = 2.2∙105 1/s.
Fig. 23. Approximately linear initiation behavior of the data for ERL = I and ERL = VI.
6.
Conclusions Shaped charge initiation trials were carried out with different shaped charges (SCs) with
different calibers form 44 mm to theoretical 280 mm. The standard charge known from earlier tests and filled with the PBX KS32 (HMX/PB 85/15, ρ = 1.64 g/cm³) was attacked by the shaped charges jets. In the course of this extensive trial campaign the stimuli S = v2d (v = SCJ-velocity, d = SCJ-diameter) of the different shaped charges were varied by varying the barrier plate 31
thickness P and determining the reaction behavior (ERL = explosive reaction level) of the standard charge. The experiments revealed that the different shaped charges with varying SC-calibers result in distinct ERL-curves. This is in clear disagreement with the “v2d-rule” (S = v2d = const.) and thus in disagreement with the STANAG 4526. Further investigating this finding and particularly trying to make a link between the initiation phenomenology in shaped charge and projectile attacks, shaped charge variations of the newly developed simplified SSC-145 with liner angles of 90° and 110° were successfully designed and tested. The jets were well characterized by using flash X-ray pictures of the jets as in earlier studies. The slower and thicker part of the SSC jet was already comparable to the STANAG projectile with a diameter of 14.3 mm. The behavior at the root points of the ERL-curves (ERL = VI) with the STANAG projectiles and the SCCs with thick jets was found to be very comparable. Also, a transition of the Penetration Mode initiation to a Semi-Impact Mode initiation could be observed. However, the question why the ERL-curves for the SCC jets exhibit extensive and ramp-like ERL-slopes paired with a plateau, whereas the STANAG-curves seem to have an assumed rather steep slope still remains unanswered. But maybe the elongated STANAG projectile curves exhibit such a plateau as well when the impact velocity is further increased? This will have to be further investigated in the future. Nevertheless, also the new test results confirmed that the “S = v²∙d = constant-rule” is not valid and that the stimulus S = v²∙d is inappropriate to predict the explosive reaction of a charge subjected to jet / projectile impact. Instead, the entirety of results exhibited a nearly linear relationship between velocity and diameter of the jet / projectile. Hence, a new linear initiation model is proposed: vcrit = A - B∙d. This relation might also finally link the initiation phenomena
32
observed in shaped charge with those in projectile attacks. At the moment it can at least be stated that the achieved empirical data are comprehensive enough to draw the conclusion that the linear relationship between the SC-velocity v and the SC-diameter d describes the measured data much better than the recent constant v²∙d-relationship. Nevertheless, this new relationship has to be further confirmed by additional trials in the future. Also, an explanation of the physical and / or chemical meaning of the two parameters A and B (if there is one) has to be postponed into the future when more data and new ideas become available. Declaration of interests ☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Author Statement The paper is a comprehensive summary of our work on initiation of a PBX with shape charge jets for the last decade. It shows a numerous new finding, including a new linear initiation model covering shaped charge jets and (STANAG or EFP) projectiles.
Acknowledgements The authors would like to thank the BAAINBw Team K 1.2 at Koblenz for the funding of this study.
33
References [1]
W. Arnold, E. Rottenkolber, “Sensitivity of High Explosives against Shaped Charge Jets”, Insensitive Munitions & Energetic Materials (IMEMTS) 2007, Miami, USA
[2]
W. Arnold, “High Explosive Initiation by High Velocity Projectile Impact”, HVIS 2010, Freiburg, GE
[3]
W. Arnold, M. Graswald, “Shaped Charge Jet Initiation of High Explosives equipped with an Explosive Train”, IMEMTS 2010, Munich, GE
[4]
W. Arnold, E. Rottenkolber, “Shaped Charge Jet Initiation Phenomena of Plastic Bonded High Ex-plosives”, IMEMTS 2012, Las Vegas, USA
[5]
W. Arnold, E. Rottenkolber, “High Explosive Initiation Behavior by Shaped Charge Jet Impacts”, Hypervelocity Impact Symposium HVIS 2012, Baltimore, USA
[6]
W. Arnold, E. Rottenkolber, “Significant Charge Parameters influencing the Shaped Charge Jet Initiation”, IMEMTS 2013, San Diego, USA
[7]
W. Arnold, E. Rottenkolber, T. Hartmann, “Analysis of Shock and Jet Initiation Tests of High Explosives”, 19th Int. Symposium on Detonation ISD 2014, San Francisco, USA
[8]
W. Arnold, E. Rottenkolber, T. Hartmann, “Challenging v²d”, IMEMTS 2015, Rome, IT
[9]
W. Arnold, E. Rottenkolber, T. Hartmann, “Testing and Modeling the Initiation of Insensitive Explosives by Projectile Impact”, 11th EUROPYRO International Seminar 2015, Toulouse, France
[10]
W. Arnold, T. Hartmann, E. Rottenkolber, “Towards a Unified Initiation Model”, IMEMTS 2016, Nashville, TN, USA
[11]
W. Arnold, T. Hartmann, E. Rottenkolber, “Filling the Gap between the Initiation Behavior of Shaped Charge Jets and Fragments”, IMEMTS 2018, Portland, OR, USA
[12]
W. Arnold, T. Hartmann, E. Rottenkolber, “Initiation Phenomenology from Hypervelocity to 34
Low Velocity Impacts”, 16th Detonation Symposium 2018, Cambridge, MD, USA [13]
T.P. Liddiard, L.A. Roslund, “Projectile/Fragment Impact Sensitivity of Explosives” NSWC TR 89-184, Naval Surface Warfare Center, Dahlgren, VA, June 1993
[14]
STANAG 4439 (Edition 3), “Policy for Introduction and Assessment of Insensitive Munitions (MURAT)”, 2006
[15]
STANAG 4526 (Edition 2), “Shaped Charge Jet Munitions Test Procedure”, 2004
[16]
STANAG 4496 (Edition 1), “Fragment Impact, Munitions Test Procedure”, 2006
[17]
SPEED 2.2 (Shock Physics Explicit Eulerian Dynamics), “Theory & User Manual”, NUMERICS, 2015
[18]
EDI 4.2 (Edge Detection in Images) “Theory & User Manual”, TDW/NUMERICS, 2014
35