Flash X-ray cinematography analysis of dwell and penetration of small caliber projectiles with three types of SiC ceramics

Elmr STRASSBURGER*，Steffen BAUERSteffen WEBER，Heiner GEDON

aFraunhofer Institute for High-Speed Dynamics，Ernst-Mach-Institute （EMI），Am Christianswuhr 2，79400 Kandern，Germany

bDLR-German Aerospace Center，Institute of Structures and Design，Pfaffenwaldring 38-40，70569 Stuttgart，Germany

cWIWeB-Bundeswehr Research Institute for Materials，F(xiàn)uels and Lubricants，Institutsweg 1，85435 Erding，Germany

Flash X-ray cinematography analysis of dwell and penetration of small caliber projectiles with three types of SiC ceramics

Elmar STRASSBURGERa，*，Steffen BAUERa，Steffen WEBERb，Heiner GEDONc

aFraunhofer Institute for High-Speed Dynamics，Ernst-Mach-Institute （EMI），Am Christianswuhr 2，79400 Kandern，Germany

bDLR-German Aerospace Center，Institute of Structures and Design，Pfaffenwaldring 38-40，70569 Stuttgart，Germany

cWIWeB-Bundeswehr Research Institute for Materials，F(xiàn)uels and Lubricants，Institutsweg 1，85435 Erding，Germany

In order to improve the performance of ceramic composite armor it is essential to know the mechanisms during each phase of the projectile-target interaction and their inf l uence on the penetration resistance.Since the view on the crater zone and the tip of a projectile penetrating a ceramic is rapidly getting obscured by damaged material，a f l ash X-ray technique has to be applied in order to visualize projectile penetration.For this purpose，usually several f l ash X-ray tubes are arranged around the target and the radiographs are recorded on f i lm.At EMI a f l ash X-ray imaging method has been developed，which provides up to eight f l ash radiographs in one experiment.A multi-anode 450 kV f l ash X-ray tube is utilized with this method.The radiation transmitted through the target is then detected on a f l uorescent screen.The f l uorescent screen converts the radiograph into an image in the visible wavelength range，which is photographed by means of a high-speed camera.This technique has been applied to visualize and analyze the penetration of 7.62 mm AP projectiles into three different types of SiC ceramics.Two commercial SiC grades and MICASIC （Metal Inf i ltrated Carbon derived SiC），a C-SiSiC ceramic developed by DLR，have been studied.The inf l uences，not only of the ceramic but also the backing material，on dwell time and projectile erosion have been studied.Penetration curves have been determined and their relevance to the ballistic resistance is discussed.

Ballistic resistance；Dwell；Flash X-ray cinematography；SiC ceramics

1.Introduction

The penetration of a high-speed projectile into a target material can only be visualized by means of f l ash-radiography.For this purpose，usually several f l ash X-ray tubes are arranged around the target and the radiographs are recorded on X-ray f i lm.A simple method is the multi-exposure of one f i lm or alternative detector.This method can only be applied when the number of objects is limited and the objects can easily be distinguished.The upper limit of the number of projections is set by the saturation of the detector due to multi-exposures.

A different approach can be realized by a set of geometrically separated channels and an array of slit apertures，in order to prevent multi-exposure.Due to geometrical boundary conditions with respect to the target set-up and safe distances，thenumber of channels is also limited.Therefore，both methods allow only pseudocinematography of the process to be observed，since the radiographs of several experiments have to be combined in order to get a time-resolved image of the process.However，this requires a high reproducibility of the experiments，which can be diff i cult to achieve in a series of tests.The lower the reproducibility，the higher is the number of tests needed.For this reason it is desirable to have a f l ash X-ray system that provides a high-number of radiographs in just one experiment.A system which provides eight f l ash radiographs at a frame rate of up to 200 kHz has been developed at EMI ［1］.

Fig.1.Schematic of test conf i guration.

This so called f l ash X-ray cinematography technique was applied in order to study the dwell-penetration transition with small caliber AP projectiles impacting different SiC ceramics on three types of backing.The phenomenon of dwell with small caliber AP projectiles at impact velocities below 1000 m/s was already discovered in the pioneering studies ofWilkins ［2］，who examined the interaction of 7.62 mm AP projectiles and surrogate steel penetrators with thin ceramic/aluminum targets. Using the classic f l ash X-ray technique Wilkins observed that the steel projectiles did not penetrate the ceramics during a time interval of about 20 μs after impact.During this phase the projectiles were eroded to about half of their initial length.The phenomenon that a projectile does not （or only very little）penetrate a target over a period of time is designated as dwell. Several studies with small caliber projectiles on the dwell phenomenon have demonstrated that erosion or “wear”of the steel core is one key factor in the energy dissipation of the projectile and，thus，for the ballistic resistance.P.C.den Reijer ［3］studied the interaction of steel cylinders with thin Al2O3-ceramic/ aluminum targets using a pseudo-cinematography set-up and determined penetration curves and dwell times.Penetration velocities were determined by Gooch et al. ［4］for 7.62 mm APM2 projectiles with B4C ceramic，and the dwell and penetration behavior with B4C/aluminum targets was studied by Anderson et al. ［5］using two 1 MeV X-ray pulse generators.

2.Flash X-ray cinematography set-up

A schematic of the measurement set-up for f l ash X-ray cinematography is shown in Fig.1.Instead of several separate X-ray tubesonemulti-anode450 kVtubeisutilized.Inthemulti-anode tube eight anodes are arranged on a circle of ≈12 cm diameter. This conf i guration causes only a relatively small parallax for the projections from the different anodes.The process under observation can be X-rayed at eight different times.The radiation transmitted through the target is then detected on a f l uorescent screen.The position of the target is between the multi-anode tube and the f l uorescent screen，relatively close to the f l uorescent screen.The f l uorescent screen converts the radiograph into an image in the visible wavelength range，which is photographed by meansofanintensif i eddigitalhigh-speedcamera.Themaximum frame rate that can be achieved with such a system depends on the decay time of the f l uorescent screen，the time characteristics oftheintensif i erandthecamera.Frameratesof200，000 fpshave been achieved with a fast decaying f l uorescent screen and have been used in this study.

3.Materials and experimental conf i gurations

The objective of the study was to determine the inf l uence of the SiC-type and the backing on penetration，deformation and erosion of the projectile.For this purpose two commercial SiCgrades，EKasic F and EKasic T of 3M Technical Ceramics（formerly ESK Ceramic）and MICASIC （Metal Inf i ltrated Carbon derived SiC），a C-SiSiC ceramic developed by DLR，have been studied.A compilation of the physical and mechanical properties of the materials is presented in Table 1.

Table 1Mechanical and physical properties of tested SiC ceramics.

The SiC ceramics were tested on three different backing materials:High hardness steel of type ARMOX 500T （Tensile strength 1450-1750 MPa），aluminum 2017 （AlCuMg1，Tensile strength 400 MPa）and Corlight sandwich panels consisting of a NOMEX? T722 meta-aramid honeycomb core of 5.3 mm thickness，covered with one layer of glass f i ber in an epoxy matrix on eachside.Thetargetswiththesandwichpanelsconsistedofthree layers:ceramic，sandwich panel and an aluminum backing plate. Due to the limited penetration capability of the radiation，small targets had to be assembled for the f l ash X-ray cinematography. On the one hand，in the cinematography set-up a good contrast could be achieved with irradiated SiC material thicknesses in the range from 30 to 40 mm.Therefore，the ceramic layer of the targets consisted of single hexagonal tiles with a wrench size（inscribed circle diameter）of 30 mm with EKasic F and T，or 32 mm with C-SiSiC.On the other hand，the size is representativefortilesusedinrealceramicarmorconf i gurations.Theupper and lower lateral surfaces of the tile were surrounded by aluminum caps，which not only provided a conf i nement of the ceramic but also ensured a homogeneous intensity distribution of the transmittedradiationoverthecompletewidthoftheceramic.The photograph in Fig.2 shows a target consisting of a ceramic on asandwich panel with aluminum backing on its mounting.The dimensions of the backing plates were 40 mm × 80 mm.The thickness of the ARMOX 500T plates was 8.5 mm，the aluminum backing plates were 8 mm thick.The targets were held by a u-steel with a distance of 80 mm between the support points.

Fig.2.Target design for f l ash X-ray cinematography.

Three tests were conducted with each ceramic-backing combination in most cases.The radiographs were recorded at frame rates of 100 and 200 kHz，i.e.the time interval between two frames was 10 or 5 microseconds.Different recording times were chosen in the three tests for a given target conf i guration，so that ideally a penetration curve with 24 data points could be determined.

Armor piercing projectiles of caliber 7.62 mm with steel core were used in the tests.The total mass of the projectiles was 9.5 g.The steel cores had a length of 28.8 mm and a mass of 4.5 g.The mean impact velocity was 841 m/s with a standard deviation of 8 m/s.

4.Experimental results

Three tests were conducted with C-SiSiC on steel backing. Fig.3 shows the eight f l ash radiographs from one of these tests. The radiographs were recorded at a frame rate of 200 kHz，starting with the f i rst frame nominally at 8 μs after impact.The actual times of the X-ray f l ashes are indicated in the single frames in Fig.3.The time of impact was determined by means of a thin shortcut trigger foil on the impact side of the ceramic. Due to the high absorption of the X-rays in the steel，the backing plate appears black in the radiographs.It can be recognized from the radiographs that the projectile hardly penetrated the ceramic during the f i rst 20 μs.In particular the radiograph at 18.8 μs illustrates that not only the jacket material but also the core material were moving radially outward along the surface of the ceramic.During this phase of projectiletarget interaction the projectile lost mass and its length were signif i cantly reduced.The so-called dwell-phase，where no projectile penetration occurred，was followed by the penetration phase.This is illustrated by the path-time plot in Fig.4，which shows the position of the projectile tail，tip and the back surface of the steel plate （bulge）for all three tests with the same conf i guration.Since the tip of the projectile was eroded quickly the term tip position is only used for the sake of simplicity in the following and means the position of the interface between the projectile and the ceramic.The path-time history of the projectile tip （projectile-ceramic interface）is represented in more detail in Fig.5.The penetration curve could be divided into two sections:During the f i rst 15 μs no signif i cant penetration was observed.A linear regression of the data during this phase yielded an average penetration velocity of 30 m/s.The dwell phase ended after 16 μs and penetration at an average velocity of 192 m/s was then observed until 60 μs after impact.The projectile was stopped since no signif i cant penetration of the steel backing could be noticed after the test.

Fig.4.Position-time plot of the projectile tail and tip as well as the bulge of the steel backing （12 mm C-SiSiC+8.5 mm ARMOX 500T）.

The uncertainty in the measurement of the projectile tip position is mainly caused by the resolution of the radiographical and optical system and the fact that，at material edges and interfaces，a gradual transition of gray shades occurs.These inf l uences led to an uncertainty in the position measurement of ±1 mm.The uncertainty in the time measurement of the X-ray f l ashes can be neglected.The application of error propagation calculation yielded the uncertainty in the point of time of the intersection of the straight lines describing the penetrationduring and after dwell.The uncertainty of the dwell times was determined as ±2 μs，which allows for discrimination between different ceramic/backing combinations.The uncertainty of the penetration velocity after the dwell phase was ±10 m/s in the case described above （Fig.5）.

Fig.3.8 Flash radiographs illustrating projectile penetration into 12 mm C-SiSiC on 8.5 mm ARMOX 500T steel backing，Test no.18262.

Fig.5.Detailed view of the penetration curve （position of projectile-ceramic interface）.

The projectile-target interaction in the case of the ceramic EKasic F on aluminum backing is illustrated in Fig.6，which shows eight f l ash radiographs recorded between 5 μs and 41 μs after impact.Due to the similar X-ray absorption in the SiC and the aluminum，the contrast between the two materials in the radiographs was only weak.The aluminum backing appeared only slightly darker than the SiC.In some of the radiographs recorded later than 15 μs，it can be recognized that a small gap opened between the ceramic and the backing，which appeared as a bright line in the radiographs.The projectile did not signif i cantly penetrate the ceramic for more than 20 μs.The position-time curve in Fig.7 illustrates the course of the projectile-target interaction during the three tests with this material combination.Three sections could be distinguished in the penetration curve.The f i rst phase without penetration （up to 9 μs）was followed by a phase with a low average penetration velocity of 41 m/s，during which the projectile penetrated only about 1 mm of the ceramic.The dwell phase lasted up to 24 μs and then the projectile penetrated the ceramic at an average velocity of 327 m/s.

Fig.7.Position-time plot of the projectile penetration for EKasic F on aluminum.

The projectile-target interaction in case of a sandwich panel between the ceramic and the aluminum backing is shown in Fig.8.Due to its low X-ray absorbing power the sandwich panel appeared as a bright zone between the ceramic and the aluminum backing.The length of the projectile was also reduced signif icantlyinthisconf i guration，butpenetrationstartedearlierandthe penetration velocity was higher compared to the other targets. After 20 μs a bulge could be recognized at the back side of the ceramic.Since the brittle ceramic cannot withstand such strong deformations without failure，it can be assumed that the ceramic is strongly fragmented in the region of the bulge.However，neither the spatial resolution nor the difference in density of the fragmented ceramic compared to the intact ceramic is suff i cient to allow for a distinction between failed and intact materials in the radiographs.The recordings between 30 and 40 μs after impact show that the fragmented ceramic material was pushed into the sandwich panel by the penetrating projectile.The corresponding position-time curve for the projectile-ceramicinterface is presented in Fig.9 along with the data of the second test with this target conf i guration.The penetration curve can be approximated by two linear sections with different slopes. Penetration started shortly after projectile impact at an average velocity of 165 m/s.After 15 μs the penetration velocity increased up to an average value of 333 m/s and remained constant during the time interval of observation （50 μs）.

Fig.6.Flash radiographs from penetration of 12 mm EKasic F on 8 mm aluminum backing.

Fig.8.Flash radiographs from the penetration of EKasic T on a sandwich panel with aluminum backing.

Fig.9.Position-time plot of the projectile penetration for EKasic T on a sandwich panel with aluminum backing.

5.Discussion

The objective of the conducted test series was to reveal the possible inf l uences of the type of ceramic and backing on the duration of the dwell phase，the penetration velocity and the projectile erosion.The analysis of the radiographs demonstrated that，during the f i rst phase of the projectile-target interaction，the penetration velocity was not zero，but very low. Depending on the target conf i guration the penetration velocity increased after 10-25 μs up to 200-500 m/s.In order to determine the duration of the dwell phase a criterion had to be def i ned.Due to the scatter in the position data a criterion based on penetration velocity was preferred.In order to determine the penetration velocity a curve had to be f i tted to the position-time data.The best f i ts were achieved either with a third order polynomial or a section wise linear regression.When using a section wise linear approximation it is assumed that there is an abrupt transition from dwell to penetration occurring within a short time interval of a few microseconds.Whereas an approximation with a higher order polynomial is more representative for a smooth transition from dwell to penetration，which takes about 5-10 μs.Experimental and numerical investigations of the dwell-penetration transition with B4C-aluminum targets by Anderson and Walker ［6］indicated a quick transition，i.e. a rapid increase in the penetration velocity.Therefore，a sectionwise linear f i t was chosen and the time when the penetration velocity exceeded 100 m/s was def i ned as the end of the dwell phase.The results for the dwell times according to this criterion are summarized in Fig.10a and b.

Fig.10a highlights the inf l uence of the backing material on dwell time.For each ceramic the three columns indicate the dwelltimeswith thedifferentbacking materials.The highest dwell times were observed with the high hardness steel backing for all types of SiC ceramic.The maximum dwell times were 25 μs for EKasic F on ARMOX steel and 23 μs with aluminum backing.The magnitude of the inf l uence of the backing was strongly dependent on the ceramic material.With the C-SiSiC the variation of dwell time with backing was relatively small，between 11 μs for the sandwich panel and 15 μs for the steel.EKasic T exhibited the strongest dependency of performance on the backing.The dwell times varied from 19 μs with steel backing to almost zero in case of the sandwich panel. In Fig.10b each color represents a SiC type and the groups of columns are shown for each backing material.Since EKasic F andT exhibit similar mechanical properties，the question arises whether signif i cant differences could be detected in the dwellpenetration behavior.The highest dwell times，independent of the backing material，were measured with EKasic F，whereas EKasic T exhibited a high performance only on steel backing. On each backing material the dwell times observed with EKasic F were signif i cantly longer compared to EKasicT.With the sandwich panel/aluminum backing no dwell phase was observed with EKasicT.However，in the diagram of Fig.10 the value for the dwell time was set to 1 μs in order to illustrate the differences between all materials.The results show the correlation between the strength of the backing and the dwell times:The higher the strength of the backing the longer was the dwell time.

Fig.10.Comparison of dwell times for all target conf i gurations.

After the dwell phase the projectiles penetrated at a constant velocity during the time interval of observation.The maximum observed penetration velocities are compiled in Table 2.

It can be recognized that the penetration velocities were higher for the lower strength backing materials.A clear dependency on the ceramic material was not discernible from the data.

Table 2Comparison of maximum penetration velocities.

Table 3Residual projectile length at t=40 μs.

The residual length or mass of the projectile core can also be used as an indicator for the capability of the ceramic to erode the projectile and，therefore，for the ballistic performance.The radiographs allow for a measurement of the residual length shortly after the end of the dwell phase.The measured residual lengths at t=40 μs after impact are summarized in Table 3. Although the differences are relatively small it can be recognized that erosion was the strongest in the case of the ceramic EKasic F and that the residual length decreased as the strength of the backing increased.

Table 4 shows a compilation of the residual length and mass data for all residual steel cores that could be retrieved from the target chamber after the single tests.When comparing the data it has to be taken into account that in some cases all residual steel cores from tests with one target conf i guration could be recovered，whereas in other cases only one was found.Nevertheless，the data reveals that projectile erosion increased with the strength of the backing material.Fig.11 illustrates the residual projectiles from the tests with EKasic F on ARMOX steel.

Due to the limited size of the ceramic tiles and the absence of a lateral conf i nement，a reduction of the dwell times or increased penetration velocities compared to bigger targets could be possible.However，the conducted experiments demonstrate that all phenomena which can be observed with larger targets also occurred with the cinematography targets.Since allmaterials were tested under the same conditions，the materials can be compared on the basis of the presented test results.Due to the small dimensions of the metal backing plates higher f i nal deformations were observed compared to large targets. However，the strong deformation of the backing occurred only after the time interval of observation with f l ash X-ray cinematography.The areal density of the targets was high enough to stop the projectiles in all cases.

Table 4Residual length and mass of recovered steel cores.

Fig.11.Comparison between initial and residual steel cores from tests with EKasic F on ARMOX 500T.

6.Conclusions

The penetration of 7.62 mm AP projectiles with steel core into the three types of SiC was analyzed by means of f l ash X-ray cinematography.The dependency of the duration of the dwell phase and projectile erosion on the type of ceramic and backing was determined.A strong inf l uence of the backing type was observed.The higher the strength of the backing，the longer was the duration of the dwell phase.The magnitude of the inf l uence of the backing was different for the three types of ceramic.The weakest dependency was observed with the C-SiSiC.The longest dwell times were observed with EKasic F on all three backing types.The maximum penetration velocities were higher for the lower strength backing materials.A clear dependency of the penetration velocity on the ceramic material was not discernible from the data.The residual length of the projectiles shortly after the end of the dwell phase could be determined from the radiographs as a measure of projectile erosion.The strongest erosion was observed in the case of the ceramic EKasic F and the data indicated that residual length decreased as the strength of the backing increased.This result was also conf i rmed by the analysis of the residual steel cores that were recovered after the tests.The comprehensive set of data can be used for the calibration of material models.

Acknowledgement

The f i nancial support of the study by the Bundeswehr Research Institute forMaterials， Fuels and Lubricants（WIWeB）（grant number E/E210/AB015/9F120）is gratefully acknowledged.The EKasic ceramic samples were provided by 3M Technical Ceramics （formerly ESK Ceramics GmbH）.

［1］Thoma K，Helberg P，Strassburger E.Real time-resolved f l ash X-ray cinematographic investigation of interface defeat and numerical simulation validation.Proc.23rd Int.Symp.Ballistics 2007；2:1065-72.

［2］WilkinsML.Third ProgressReporton LightArmorProgram，UCRL-50460，Lawrence Livermore Laboratory，Livermore，CA，USA 1968.

［3］den Reijer PC.Impact on ceramic faced armor ［Ph.D.thesis］，Delft University of Technology 1991.

［4］Gooch WA，Burkins MS，Kingman P，Hauver G，Netherwood P，Benck R. Dynamic X-ray imaging of 7.62-mm APM2 projectiles penetrating boron carbide.Proc.18th Int.Symp.Ballistics 1999；2:901-8.

［5］Anderson CE，Burkins MS，Walker JD，Gooch WA.Time-resolved penetration of B4C tiles by the APM2 bullet.Comput Model Eng Sci 2005；8（2）:91-104.

［6］Anderson CE，Walker JD.An analytical model for dwell and interface defeat.Int J Impact Eng 2005；31:1119-32.

Received 7 October 2015；revised 18 January 2016；accepted 26 January 2016 Available online 3 March 2016

Peer review under responsibility of China Ordnance Society.

Paper for the 29th ISB.

*Corresponding author.Tel.:+49 7626 9157235.

E-mail address:strassburger@emi.fraunhofer.de （E.STRASSBURGER）.

http://dx.doi.org/10.1016/j.dt.2016.01.011

2214-9147/? 2016 China Ordnance Society.Production and hosting by Elsevier B.V.All rights reserved.

? 2016 China Ordnance Society.Production and hosting by Elsevier B.V.All rights reserved.