Hypersonic impact flash characteristics of a long-rod projectile collision with a thin plate target

Yi-jiang Xue,Qing-ming Zhang,Dan-yang Liu,Ren-rong Long,Yang-yu Lu,Tian-fei Ren,Liang-fei Gong

State Key Laboratory of Explosion Science and Technology,Beijing Institute of Technology,Beijing 100081,China

Keywords:Hypersonic impact flash Radiation intensity Impact flash mechanism Environmental luminescence

ABSTRACT Impact flash occurs when objects collide at supersonic speeds and can be used for real-time damage assessment when weapons rely on kinetic energy to destroy targets.However,the mechanism of impact flash remains unclear.A series of impact flash experiments of flat-head long-rod projectiles impacting thin target plates were performed with a two-stage light gas gun.The impact flash spectra for 6061 aluminum at 1.3-3.2 km/s collision speeds were recorded with a high-speed camera,a photoelectric sensor,and a time-resolved spectrometer.The intensity of the impact flash exhibited a pulse characteristic with time.The intensity(I)increased with impact velocity(V0)according to I∝Vn0,where n=4.41 for V0＞2 km/s.However,for V0<2 km/s,n=2.21,and the intense flash duration is an order of magnitude less than that of higher V0.When V0＞2 km/s,a continuous spectrum(thermal radiation background)was observed and increased in intensity with V0.However,for V0<2 km/s,only atomic line spectra were detected.There was no aluminum spectral lines for V0<2 km/s,which indicated that it had not been vaporized.The initial intense flash was emission from excited and ionized ambient gases near the impact surface,and had little relationship with shock temperature rise,indicating a new mechanism of impact flash.?2020 The Authors.Production and hosting by Elsevier B.V.on behalf of China Ordnance Society.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1.Introduction

The study of impact flashes originated from the Apollo program[1,2],and is still an important topic in astrophysics[3-10].The characteristics and spectrum of the impact flash can be used to determine the material composition and the surface structure of celestial bodies[11-13].It also has important applications in realtime damage assessment for kinetic energy weapons,such as armor-piercing projectiles[14]and missile defense systems[15],where velocities exceed 1 km/s.Impact flash is the most effective means of performing real-time damage assessment of the target and its authenticity,which can then provide a basis for decision making concerning second strikes.

Many hypervelocity impact flash experiments using different materials have been performed[16-26].The impact flash radiation intensity has a pulse characteristic with time,and appears almost simultaneous with the projectile/target contact.Several or tens of microseconds later(determined by the impact velocity and target material),the radiation intensity reaches a peak value;then it decays to 20%(or less)of the peak intensity,after which it decays for milliseconds at a low intensity.The impact flash radiation intensity I and the impact velocity V0exhibit a power function relationship given by I∝Vn0.For V0in the range 3-7 km/s,n=8 for a copper ball and copper target collision[17],n=4.0-4.5 for an aluminum ball and aluminum target collision[20],and n=3.8-5 for iron particle collisions with a gold target[18].These results are very important references for the damage assessment of kinetic-energy-based weapons.However,the speed of the weapons is usually lower(V0<3 km/s)than that of celestial bodies.There are no systematic studies of impact flashes in this speed range,and almost no experimental data is available for reference.

Plasma is the main source of the impact flash when V0＞10 km/s[27,28].For V0=5-7 km/s,atomic aluminum lines at 396.15 nm and 394.40 nm were observed[17,26],and well as the copper atomic line at 510.55 nm[23],indicating gaseous aluminum or copper.Because the impact velocities of these tests were very high,it was inferred that the shock temperature vaporizes the material.However,the mechanism of the impact flash is unknown when the collision speed is reduced and the shock temperature rise is insufficient to vaporize the material.

Here,a two-stage light gas gun was used to accelerate projectiles into targets.A high-speed camera,a photoelectric sensor,and a time-resolved spectrometer equipped with an intensified charge-coupled device(ICCD)were used to record impact flashes and their spectra for V0in the range 1.3-3.2 km/s.Time and frequency domain diagnostic equipment was used to multidimensionally analyze the characteristics of the hypersonic impact flash.The spectral analysis was used to identify the emitting material and its state,which could be used to infer the mechanism of the impact flash.

2.In situ experiments

2.1.Projectile and target

Previous studies generally used spherical projectiles and thick targets[26].Here,flat-head,long-rod projectiles and thin targets were examined.According to Ref.[29],the original design of the long-rod projectile and the sabot structure are depicted in Fig.1(a),using sub-caliber launch technology.After multiple launches,the projectile cannot meet the experimental requirements,as shown in Fig.1(b).If the end of the long-rod is the surface that is mainly subjected to the force,it is easy to amplify the error caused by machining during the driving process.Thus,the flight posture is difficult to control after ejection of the long-rod projectile,and a very significant attack angle occurs.A long-rod projectile and sabot structure used after many improvements are shown in Fig.1(c).The force-bearing surface of the long-rod projectile and the sabot is parallel to the launch tube,and the shape centroid of the sabot is also coincident with the mass centroid of the long-rod projectile.The optimized experimental result is shown in Fig.1(d),where the long-rod projectile maintains the flight posture after leaving the tube muzzle.

The long-rod projectile used here is shown in Fig.1(e).It has a diameter of 6 mm and a 50-mm length.The sabot has a diameter of 14.5 mm and a 20-mm length.The thickness of the target plate is 4 mm.Both the projectile and target are 6061 aluminum.The target plate is perpendicular to the projectile trajectory and is placed in a vacuum chamber where the ambient pressure is 100 Pa.The range of V0is 1.3-3.2 km/s.The long-rod projectile recovered by steelplate retardation is shown in Fig.1(f),the right image is the collision surface and the left is the tail of the projectile.

Sabot detachment could easily affect the flight posture of the long-rod projectile,which in turn affects the repeatability of the impact flash experiments.The intense flash at the moment of collision is the main concern here,and there is a delay time for the sabot/target collision.Therefore,we did not take off the sabot,and successfully closed the electronic shutter of the ICCD spectrometer before the sabot hit the target plate,which avoided its affection on the spectra.The specific experimental results in Section 3.1 proved the feasibility of this scheme.

Fig.1.(a)Original design of a long-rod projectile and the sabot structure.(b)Flight trajectory of the original design photographed with a high-speed camera.(c)Optimized design of the long-rod projectile and sabot.(d)Flight trajectory of the optimized design photographed with a high-speed camera.(e)6061 aluminum long-rod projectile and sabot used here.(f)The collision surface is at right,and the tail of the projectile is at left.

2.2.Experimental system

As shown in Fig.2,the experimental arrangement consists of three subsystems:launching,diagnostics,and triggering.The launching system consists of four parts:a two-stage light gas gun,a flying chamber,a target chamber,and a three-channel electromagnetic velocity measurement device(EVMD)at the exit of the launching tube to measure V0.The EVMD generates current when the projectile passes through.V0can be calculated from the time intervals of the currents in the three channels recorded by a digital storage oscilloscope,and the 50-mm distance between each channel.Because the ambient pressure of the flying chamber and the target chamber is only 100 Pa,speed attenuation of the projectile can be ignored and the EVMD data is directly used as the velocity V0at which the projectile collides with the plate target.

The diagnostic system includes a high-speed camera(Photron Corp.,Japan),a silicon-based photoelectric sensor(Thorlabs Corp.,USA),and an ICCD spectrometer(Andor Corp.,UK).The photoelectric sensor has a 350-1050 nm range and a response time less than 10 ns.The impact flash irradiance is converted by the sensor into a voltage signal that is recorded over time with a digital storage oscilloscope.The ICCD spectrometer has a 200-1100 nm range,with a resolution less than 0.02 nm.Wavelength and intensity calibrations were performed separately before each experiment.The fiber optic probe of the ICCD spectrometer and the photoelectric sensor are symmetrically arranged on both sides of the impact point,and respectively record the spectral signal of impact flash and its intensity over time.The high-speed camera is outside the target chamber,and the impact flash is recorded through a window of the chamber for a side/front view of the collision,as depicted in Fig.2.The diagnostic equipment is synchronized by the triggering device,and the high-speed camera,photoelectric sensor,and ICCD spectrometer can be simultaneously trigged to record signals when the projectile is hitting the target.

2.3.Triggering system

A 10-μm-thick insulated copper foil is attached to the surface of the target plate.The foil and the metal target plate are connected to a resistance-capacitance(RC)pulse circuit that turns on instantly when the projectile hits the plate.A current pulse is thus formed and triggers the digital delay generator(DDG)to output multiple TTL-level signals,which simultaneously start the high-speed camera,the photoelectric sensor,and the ICCD spectrometer.The response time of the whole process is less than 100 ns,and the delay time between the DDG and the ICCD spectrometer is zero during experiments.For V0=3.0 km/s,signals with and without the copper foil output by the photoelectric sensor are plotted in Fig.3.The copper foil has no significant effect on the impact flash.

Fig.3.Effect of copper foil on the impact flash.

3.Results

3.1.Impact flash process

The impact and associated flash were recorded with the highspeed camera and the photoelectric sensor.Photographs of the impact flashes for V0=1.5 km/s and V0=2.5 km/s are shown in Fig.4.As shown in Fig.2,the camera is has a side/front view of the plate target.There is no background light source;therefore,the camera exposure relies entirely on the flash intensity.At time t=0,the projectile is starting to impact the target.For V0=1.5 km/s,a circular emission area appeared at the moment of collision(Fig.4a t=0).Because the interval between each camera frame was too long,changes in the emitting region were not observed.In the second photograph at 11.5μs,the impact flash has decayed,while in the third photograph at 23μs,material is ejected from the collision point,with almost no light emission.

For V0=2.5 km/s(Fig.4(b)),the circular illuminating area was about twice the size of that observed for V0=1.5 km/s.The entire field of view in the second photograph at 16μs was overexposed because of the intense flash.The“overexposure”lasted less than 32μs,because the third photograph started to reveal some detail.In the fourth photograph at 96μs,the flash intensity was much weaker,and“dust plume”-like weak luminescence appeared.More high-speed camera results at different impact velocities are listed in Table 1.The results indicated that higher V0produced longer“overexposures”and flash durations.

Impact flash signals vs.time recorded from the photoelectric sensor for various V0are shown in Fig.5.The sensor output voltage increases with the intensity of the impact flash.Because the photoelectric sensor was not calibrated,only relative values were obtained.The signal changes with time for different V0,and all have pulsed characteristics.The entire impact flash process can be separated into a pulse stage and an attenuation stage,as shown in Fig.5(b).More sensor results for different V0are shown in Table 1.They indicate that as V0increases,the peak radiation intensity increases along with the duration of the intense pulse flash.However,the time at which the peak intensity occurs(peak time)lags.When compared with the spherical aluminum projectile colliding with the plate target[1],the long-rod aluminum projectile colliding with the thin plate has a slightly longer pulse width,but still has the same order of magnitude.

To analyze the time when the sabot hits the plate target at different V0,a model using the experimental dimensions was established for smoothed particle hydrodynamics(SPH)numerical simulations of the impact process.The calculated results for different V0are shown in Fig.6.The material model of 6061 aluminum was reported previously[30,31].During the collision,the plate target is greatly deformed by the projectile and many plate fragments are ejected.Some of the fragments move in a direction opposite to the impact velocity and eventually collide with the sabot.The time when the sabot and the debris began to interact was 7.4μs,4.8μs,and 3.8μs after impact at V0=1.5 km/s,2.4 km/s,and 3.0 km/s,respectively.The results in Fig.5 indicate that the flash signals were hardly affected by the sabot during the first half of the peak time.Therefore,in order to avoid or minimize sabot effects on the spectra,the electronic shutter width of the ICCD spectrometer was set to half of the peak time of the sensor signal for the corresponding impact velocity,as shown in Fig.5(c).

3.2.Spectral characteristics

Based on the relative intensity calibration,the spectrum of the V0=1.43 km/s impact flash captured by the ICCD spectrometer is plotted in Fig.7.The entire spectrum over the range 300-800 nm consisted of atomic lines;no bands or continuum spectra were observed.In addition to the two strong spectral lines,the average amplitude of the other spectral lines was 1600(a.u.).

Fig.4.High-speed camera photographs of impact flashes taken from a side/front view of the plate target.The exposure time of each photograph was 1.5μs(a)Impact flash for V0=1.5 km/s at 100 Pa ambient pressure.(b)Impact flash for V0=2.4 km/s at 100 Pa ambient pressure.

Table 1High-speed camera and photoelectric sensor results.

The spectrum for V0=2.07 km/s is shown in Fig.8.It still had a large number of spectral lines,but there was also a continuous thermal radiation background over the range 400-800 nm.The average amplitude of the spectral lines was around 2000(a.u.),and that of thermal radiation was about 5%of that.For V0=3.12 km/s,the spectrum shown in Fig.9 had an average amplitude for the spectral lines of about 6000(au),and a thermal background amplitude that was 30%of that.

The average irradiance measured by the optical fiber probe of the ICCD spectrometer was obtained by integrating the spectrum over 300-800 nm and dividing the result by the exposure time.Because the distance L0between the probe and the collision point was fixed and the light entering the area S of the optical fiber was much smaller than L02,the solid angle of each experiment was approximately the same.Thus,the irradiance can be considered equivalent to the relative intensity of the impact flash,and was plotted in Fig.10 for various V0.The black dots are for V0<2 km/s,and the red are for V0＞2 km/s.In both regions,V0and the intensity I of the impact flash can be fit with a power function,but the relationship is not continuous.For V0<2 km/s,the fit is I=AV2.210,whereas,for V0＞2 km/s,the fit is I=BV4.410,where A and B are constants.At 2 km/s,there is a transition region(blue dot)which cannot be fit by the two relations.

In the fits,the larger exponent indicated a steeper increase in flash intensity with V0.The fit for V0＞2 km/s was very close to the results for an aluminum ball projectile collision with an aluminum plate for V0<3-7 km/s[20].Even though projectiles had different shapes,the relationship between I and V0had the same fit with a similar exponent.Hence,the exponent likely characterizes aluminum projectiles,independent of shape.This finding is helpful for the study of inverse problems.

3.3.Spectral analysis

When the Solis spectrometer software(Andor Corp)was used to analyze the spectra for V0=1.43 km/s,none of the spectral lines could be attributed to aluminum,but rather to environmental gases.

Fig.5.Photoelectric sensor signals vs.time for various impact velocities.

Fig.6.SPH simulation results.(a)At t=7.4μs for V0=1.5 km/s.(b)At t=4.8μs for V0=2.4 km/s.(c)At t=3.8μs for V0=3.0 km/s.The simulation platform is ANSYS 14.0.

Fig.11 shows the analysis results for nitrogen and oxygen.It can be seen that lots of spectral lines can be identified by neutral nitrogen or oxygen atoms and their ion states.It indicates the excitation and ionization of the ambient gas may be the main source of the impact flash.The other lines of the spectrum plot have the possibility to be identified by some of the inert gas elements.However,since the amount of inert gas is very little,the decision on the excitation of inert gas element should be cautious until further results are obtained.

Because 6061 aluminum is an alloy with possible impurities and other elements,impact flash tests of pure aluminum(＞99.9999%)were performed under the same conditions.The spectral analysis is shown in Fig.12,where again no aluminum lines were identified.Also tested was 45 steel at V0=1.30 km/s;the spectral analysis is shown in Fig.13 which still consists of line emissions by gaseous elements in the environment.These results indicate that ambient gas emission is likely to be common in hypersonic impact flashes of metallic materials.

Fig.7.Spectrum for impact flash at V0=1.43 km/s.The ICCD gate width was 2μs.

Fig.8.Spectrum for impact flash at V0=2.07 km/s.The ICCD gate width was 2μs.

Fig.10.Radiation intensity I vs.impact velocity V0.

Fig.11.Spectral analysis of impact flash for 6061 aluminum at V0=1.43 km/s.Nitrogen and oxygen spectral lines were identified.The ICCD gate width was 2μs.

Fig.9.Spectrum for impact flash at V0=3.12 km/s.The ICCD gate width was 4μs.

Fig.12.Spectral analysis of impact flash for pure aluminum(99.9999%)at V0=1.40 km/s.Nitrogen and oxygen spectral lines were identified.The ICCD gate width was 2μs.

The spectral analysis of the impact flash for V0<2 km/s was similar.When V0＞2 km/s,the continuous spectrum representing high-temperature condensed matter increases in intensity with V0.Thus,it is likely to be the result of a shock temperature increase of the collision material.However,visible light emission of multiphase materials is very complex and is the result of photon,electron,and ion interactions.

Fig.13.Spectral analysis of impact flash for 45 steel V0=1.30 km/s.Nitrogen and oxygen spectral lines were identified.The ICCD gate width was 2μs.

4.Discussion

4.1.Impact flash emission source and its characteristics

Impact flash is an interdisciplinary problem of collision dynamics and light emission,and accurate measurements of impact flashes are critical.The experimental results here revealed impact flashes that are more complex than previously thought.We first define the light source,and then summarize its characteristics and suggest further research.

We define the illuminating area formed by the collision as a light source,as shown in Fig.4(a and b)at t=0.In the high-speed photographs,the emitting area is continuously expanding;thus,it is a dynamic light source,as shown in Fig.14,the ds is the area element of photoelectric sensor(observation point),and the L0is the original distance between light source and observation point.The luminous zone expands faster at higher V0.At t=16μs in Fig.4(b),the light source exceeded the field of view of the camera(about 80×80 mm)and overexposed the image.

Fig.14.Dynamic light source at different time.(a)Original time;(b)t=tn.

The dynamic feature of the light source has two main effects on the impact flash test.First,from a fixed observation point,the solid angle changes with time,which greatly increases the difficulty of performing absolute measurements.Second,from different observation points,the observed impact flash is constant.In Table 1,the“overexposure”state recorded by the camera was significantly longer than the pulse width recorded by the photoelectric sensor.This was most likely because the observation point of the highspeed camera was much farther from the collision point than was the sensor.

The spectral characteristics of the various V0suggest that the impact flash was visible light emission from multiphase materials,which is the second feature of the impact flash light source.Gaseous atoms emit sharp line spectra,gaseous molecules emit band spectra,and condensed matter(liquid sand solids)emit continuous spectra[32].For V0<2 km/s,the spectra mainly consisted of atomic lines.For V0＞2 km/s,in addition to a large number of line spectra,a continuous thermal radiation background appeared,indicating a mixture of gaseous and condensed states.This makes it very difficult to determine the temperature of the condensed phase from the impact flash.Previous results suggested that the temperature of the impact flash determined with a multichannel pyrometer was significantly higher than the shock temperature rise of the collision materials[22].

4.2.The mechanism of impact flash

When V0<2 km/s,the atomic spectral lines were not aluminum,which is significantly in contrast to previous results.In the spherical aluminum projectile impact flash at V0=5-6 km/s,there were two characteristic atomic aluminum lines at 396.15 nm and 394.40 nm[26],derived from two excited states.It was therefore generally believed that the collision material exhibits gasification and luminescence[17,21].The energy transport process is depicted in Fig.15(a).From shock wave compression to the impact flash,the energy carriers are all based on aluminum.

Fig.15.Energy transport processes and carrier.(a)Previous understanding;(b)New mechanism.

However,our spectral analysis results show that excited and ionized ambient gases are the main sources of the intense flash at the moment of collision at V0<2 km/s.The energy carriers of this new impact flash mechanism are shown in Fig.15(b),where the energy transport process is unknown.It is not clear whether the new mechanism persists or dominates when V0＞2 km/s.The first excited states of nitrogen and oxygen atoms are at 10.88 eV and 9.48 eV,respectively;whereas,that of aluminum atoms is at 3.13 eV.Thus,nitrogen and oxygen atoms are more difficult to excite than aluminum atoms.Hence,the energy that causes the ambient gas to emit light is not from the shock temperature rise.The new impact flash mechanism occurs when the shock residual temperature is not sufficient to vaporize aluminum.

4.3.Spectra from V0<2 km/s

When V0<2 km/s,no continuous spectra were observed.According to black-body thermal physics,condensed matter at a temperature greater than 0 K will radiate a spectrum of electromagnetic energy that correlates with the temperature[33].At low temperatures,the light may not be visible to the human eye,nor detected with optical devices such as ICCDs.Here,we roughly estimate the temperature threshold of condensed matter with the ICCD spectrometer.

The sensitivity of the photosensitive ICCD element is about 1×10-6Lx(photometry unit),that corresponds to Ee=4.6×10-9w/m2/nm(Irradiance of the photosensitive element)monochromatic emission at 555 nm[34].The original distance L0(Fig.14(a))between the optical fiber probe and the collision point is 100 mm.The relationship between expanding velocity(Ve)of emitting area and V0is Ve=αV0,and theαis about 4-4.7 with different V0according to Fig.4.When V0=2 km/s,theαis 4.4.Then Rn=Vetn+R0=20.6 mm at t=2μs is calculated,where the R0is the radius of the projectile.

As luminous zone expanding(Fig.14(b)),the relationship between the radiance of light source(Me)and Eeof ds is:

whereλis the emission wavelength,ηis the radiation efficiency,c1=3.742×10-16w?m2is the first radiation constant,and c2=1.439×10-2m K is the second radiation constant.

The calculation results for the radiance and various temperatures of the condensed matter source are plotted in Fig.16,where the vertical axis is Me,the horizontal axis is the wavelength,theis 1.12×10-7w/m2/nm and theηis 0.8.Emission at 555 nm and radiance equal to or greater thancan only be detected by the ICCD spectrometer for condensed matter temperatures above 822 K.The calculation results have two meanings.First,when V0<2 km/s,the thermal radiation could not be detected by ICCD spectrometer so the relationship between radiation intensity and impact velocity is not continuous.Second,if this assumption is true,then the calculated impact flash intensity for V0<2 km/s may need to be corrected,depending on the application requirements.But this does not affect the conclusion regarding the flash mechanism.Also,it can be inferred that the impact flash for V0<2 km/s has little relationship with shock temperature rise,because the aluminium material could not emit line spectrum below 822 K.

Fig.16.Results of radiance and temperature calculations of condensed matter emission.

5.Conclusions

Here,a series of impact flash experiments for a flat-head longrod projectile impacting a thin target plate were performed.A highspeed camera,a photoelectric sensor,and an ICCD spectrometer were used to characterize the flash and mechanism for V0=1.3-3.2 km/s.There were two characteristics of the impact flash light source.First,it was dynamic,where its geometric size continuously increased with time.Second,the light source was a mixture of gaseous and condensed materials.These two features make it challenging to accurately identify and evaluate the flash source with photometry.Conclusions were drawn as follows.

The impact flash intensity I of the long-rod projectile exhibited a pulse characteristic with time.The duration increased with V0.I was a power function of V0,i.e.I∝.For V0＞2 km/s,n=4.41,which is close to previous results for spherical aluminum projectiles at V0=3-7 km/s.However,for V0<2 km/s,n=2.21,indicating less of an increase in flash intensity with V0.The intense flash durations were also different in that the value for V0<2 km/s was an order of magnitude less than that for V0＞2 km/s.

When V0＞2 km/s,continuous spectra were observed and increased in intensity with V0.However,when V0<2 km/s,the intense flash spectra mainly consisted of atomic lines.Spectral analysis indicated that the lines derived from the excitation and ionization of ambient gases near the impact surface.In contrast to shock-induced vaporization luminescence of the collision materials,the finding here indicates a new mechanism for impact flashes.Although the energy transport process is uncertain,it has provided a new direction for future studies.Because the new mechanism was also observed in impact flash experiments with pure aluminum and 45 steel,ambient gas emission may be a common feature of hypersonic impact flashes of colliding metallic materials.

Declaration of competing interest

The authors declare no conflict of interests.

Acknowledgments

The authors would like to thank Mrs Li Chen,Mr Siyuan Ren,Mr Cheng Shang,Mr.Weijiang Hao,Mr.Yan Liu,Mr.Yaoxin Wang and others for their help in conducting the two-stage light-gas gun tests.This work was supported by the National Key R&D Program of China(Grant No.2016YFC0801204)and the National Program on Key Basic Research Project of China(973 Program,Grant No.613312).