Ballistic performance of titanium-based layered composites made using blended elemental powder metallurgy and hot isostatic pressing

Abstract

Metal matrix composites tiles based on Tie6Ale4V (Ti64) alloy, reinforced with 10, 20, and 40 (vol%) of either TiC or TiB particles were made using press-and-sinter blended elemental powder metallurgy (BEPM) and then bonded together into 3-layer laminated plates using hot isostatic pressing (HIP). The laminates were ballistically tested and demonstrated superior performance. The microstructure and properties of the laminates were analyzed to determine the effect of the BEPM and HIP processing on the ballistic properties of the layered plates. The effect of porosity in sintered composites on further diffusion bonding of the plates during HIP is analyzed to understand the bonding features at the interfaces between different adjacent layers in the laminate. Exceptional ballistic performance of fabricated structures was explained by a significant reduction in the residual porosity of the BEPM products by their additional processing using HIP, which provides an unprecedented increase in the hardness of the layered composites. It is argued that the combination of the used two technologies, BEPM and HIP is principally complimentary for the materials in question with the abilities to solve the essential problems of each used individually.

Headnote

ABSTRACT

Metal matrix composites tiles based on Tie6Ale4V (Ti64) alloy, reinforced with 10, 20, and 40 (vol%) of either TiC or TiB particles were made using press-and-sinter blended elemental powder metallurgy (BEPM) and then bonded together into 3-layer laminated plates using hot isostatic pressing (HIP). The laminates were ballistically tested and demonstrated superior performance. The microstructure and properties of the laminates were analyzed to determine the effect of the BEPM and HIP processing on the ballistic properties of the layered plates. The effect of porosity in sintered composites on further diffusion bonding of the plates during HIP is analyzed to understand the bonding features at the interfaces between different adjacent layers in the laminate. Exceptional ballistic performance of fabricated structures was explained by a significant reduction in the residual porosity of the BEPM products by their additional processing using HIP, which provides an unprecedented increase in the hardness of the layered composites. It is argued that the combination of the used two technologies, BEPM and HIP is principally complimentary for the materials in question with the abilities to solve the essential problems of each used individually.

Keywords:

Metal matrix composites

Powder metallurgy

Titanium hydride powder

Master alloy

Titanium carbide

Titanium boride

Hot isostatic pressing

Ballistic tests

1. Introduction

Titanium alloys are important structural materials for numerous applications due to a unique complex of physical and mechanical properties, as stated by Lutjering and Williams [1]. The engineering application of these alloys according to Peters et al. [2] is primarily based on their high specific strength well-balanced with other mechanical characteristics. That is why by Niinomi [3] titanium alloys are used in the biomedical, and based on Jones et al. [4] in aerospace, automotive, and military products. In the latter case, titanium alloys are very desired materials for the armor elements fabrication, as pointed out by Montgomery and Wells [5] and later confirmed by Fanning [6]. According to Prikhodko et al. [7], layered structures for this application are shown as one of the best approaches to improving the protective performance of armor. Respectfully, it is very important to have high hardness value at the front surface of the armor element to stop the piercing action of the projectile and high ductility in its core to prevent the fragmentation of the armor, as noted by Prikhodko et al. [7]. Powder metallurgy (PM) is one of the best methods of making layered structures, which is not an easy to achieve using more traditional titanium processing such as cast and wrought technology. Besides, according to Ivasishin et al. [8], it is very attractive due to its cost efficiency. It has been recently shown by another study of Ivasishin et al. [9] that the laminate structures of the alloy Tie6Ale4V (wt%) (Ti64) and metal matrix composites (MMC) based on this alloy can be made by relatively simple pressing and sintering Blended Elemental Powder Metallurgy (BEPM) using titanium hydride TiH2 as a base powder with additions of AleV master alloy. For the higher hardness composite layers, reinforcement particles of TiC or TiB can be added to the blend. Owing to the presence of such particles in top layer of the laminate, the surface hardness can be increased from about 340 HV common for conventional Ti64 alloy, to about 400e430 HV when 20% of TiC particles is added to the composite as shown by Markovsky et al. [10]. However, according to Benedetto et al. [11] the core of special armor piercing bullet is made of reinforced steel with a hardness above 700 HV, therefore, effective protection against the piercing action of such projectile requires a corresponding hardness of the armor. As Markovsky et al. [12] shown recently, a significant hardening effect can be achieved in titaniumbased composites via high temperature solution treatment after sintering of the structure, which was explained by increasing the hard fraction content during used high temperature aging and quenching. Though, it was discussed that the lack of complete structure densification prevents the material reaching its ultimate performance. The most common ways of porosity reduction in PMmade metal parts usually involve a post processing of the structure using hot plastic deformation. Unfortunately, it is not effective on multilayer structures as shown by Prikhodko et al. [13] due to the mismatch in the plastic flow of different layers during hot plastic deformation of the laminate. However, as Kim and Yang [14] note, when plastic deformation is applied isostatically, as it done in hot isostatic pressing (HIP) this difficulty can be overcome. This was experimentally confirmed by Dekhtyar et al. [15] for near-b titanium powder alloys. The combined effect of high temperature and pressure, which used in HIP could potentially eliminate the porosity as well as increase the content of the hard phase within the composite. Although the added treatment undoubtedly increases the cost of the entire process, its use can be justified by a significant improvement in the mechanical characteristics of the final products. In view of the above, the purpose of this study was to increase the hardness of composites on the base of titanium alloy Ti64 to improve their antiballistic protection by (i) adding a higher content of hard phases than was used before and (ii) reducing residual porosity. It was expected that for parts manufactured using BEPM, the goal could be achieved by further processing them using HIP.

2. Materials and experimental procedure

Six composite tiles on the base of the alloy Ti64 reinforced with different amount of either TiC or TiB particles, were made using BEPM. Each individual tile was reinforced with 10, 20, or 40 (vol%) of TiC or TiB particles. The maximum used content of 40% was chosen based on the results of prior experiment described below and the recent study of Markovsky et al. [12], which, reported the highest hardness value of MMC reinforced with TiC at this content. Following Prikhodko et al. [7] and Markovsky [10] two other tiles' compositions, 20% and 10% of TiC, were selected based on the optimal ratios for reinforcement between adjacent layers to optimize the mechanical properties of laminates. For structures with TiB, similar tiles' compositions, 40%, 20%, and 10% were taken to compare the effect of different hardening phase. All six powder blends were prepared using hydrogenated titanium (TiH2) sponge crush (3.5H (wt%), particle size <100 mm) as the base powder instead of conventional titanium powder. The advantages of using TiH2 have been reported by Ivasishin et al. [16] and confirmed later by Ivasishin and Moxson [17]. For making individual MMC tiles on the base of alloy Ti64, TiH2 powder was blended with corresponding amounts of 60Ale40V (wt%) master alloy powder (particle size <63 mm) and either TiC powder (1e30 mm) or TiB2 powder (1e20 mm). In the latter case, based on Sahay et al. [18], TiB2 was expected to be converted during vacuum sintering to monoboride following the reaction TiB2þTi/TiB. The powder blends were then die-pressed at 150 MPa at room temperature to obtain 909010 mm flat preforms. The preforms were sintered in the vacuum furnace at 1250 C for 4 h and cooled with the furnace. All treatment provided titanium dehydrogenation, sintering and homogenization of powder compacts, formation of nearly dense and uniform matrix alloy Ti64 with specified amounts of reinforcement phase. Detailed structure examination of tiles was not possible in this study since the goal was ballistic examination of the bonded tiles and only posttest structure examination was conducted at full scale. However, relatively small samples for structure characterisation were taken from the tiles before their HIP bonding. Besides, the initial structure of composites was carefully studied in an additional prior experiment using small cylinder samples described below. The sintered MMC tiles were joined together using HIP to produce three-layered hybrid plates. Two different threelayer plates were made, one with TiC and the other with TiB. The reinforcement in adjacent layers of each triplet varied by 10%, 20% and 40%. HIP was done at 900 C, 100 MPa for 3 h. Tiles were placed in one prismatic can with the stainless steel spacer to separate different plates in stack. Plate laminates after the HIP had some specific bulges of technological origin in the center of the sample, which were removed by spark erosion and polished for their ballistic test.

This study carried out an additional prior experiment, which was done using small-sized cylinder samples. Ten MMC samples, five with TiC and five with TiB particles with a volume fraction varied from 20% to 60% in 10% increments, were used to test samples densification in HIP treatment and to verify HIP processing parameters that could be used to make armor plates suitable for the ballistic test. Composite cylinders with an initial dimension 1015 mm (dia.l) were manufactured through the sintering using standard in this study BEPM protocol. This test made it possible to finalize the HIP parameters for the manufacture of the composite laminates plates suitable for ballistic examination. For the HIP processing cylinder samples were placed in one tubular can with the stainless steel spacers to separate each sample. After the HIP cylinders were removed from the tubular can. For the SEM structure study about 1 mm thick layer was cut using diamond saw from one end of each cylinder and polished; this guaranteed a structure representing the bulk.

Structure of the materials was characterized using light optical microscopy (LOM), done with Olympus IX70 microscope (Olympus, Japan), and digital microscope (DM) VHX-6000 (KEYENCE, USA). Polished samples for LOM were etched with Kroll reagent (2% HF, 3% HNO3, 95% H2O). Samples after ballistic tests, including fractured surfaces, were studied using scanning electron microscope (SEM) VEGA 3 (Tescan, Czech Republic). The volume fraction of porosity of the samples was estimated as the pores area fraction measured using ImageJ software. Given the size of the tiles and considering the plates multilayered structure, their porosity measurement using a more accurate Archimedes method was not possible, and their representative sampling for this purpose was not acceptable due to need of their ballistic examination. However, our preliminary measurements made on another set of samples with similar reinforcement (below 40%) content show good agreement between these two methods.

Ballistic tests were carried out using stand shown in Fig. 1(a) in a ballistic test laboratory at the Jarosław Da˛browski Military University of Technology in Warsaw. The used 7.6251 mm NATO ammunition was equipped with flat-nose core AP projectile weighting 126.6 grains1 - i.e. 9.45 g (Fig. 1(b)), containing a hard steel core with a blunt tip (Fig. 1(c)). Rounds applied during investigations were produced by MESKO S.A., Poland. Some other important characteristics of used ammunition are listed in Table 1. In order to record the moment of projectile impact and its deformation and destruction, high-speed camera was used with the resolution reduced to 256256 dpi, the frame rate was around 130,000 frames/s. It is important to note that the core hardness of such a bullet is 870 HV, while according to Poondla et al. [19], the highest hardness value of titanium alloy, expected after the special hardening heat treatment, usually does not exceed 350e400 HV. This is the main reason for the relatively low ballistic resistance of titanium-based armor when its compared to rolled homogeneous armor (RHA) steels, which is still considered as one of the most commonly used armor material.

3. Results

3.1. Microstructure of materials after BEPM fabrication and HIP processing

Microstructures taken from cylindrical samples after BEPM and HIP are shown in Fig. 2 for TiC and Fig. 3 for TiB composites respectfully. The fraction of porosity measured on these samples before and after the HIP is listed in Table 2. As can be seen, HIP was very effective for porosity reduction in composites, especially when reinforcement fraction was not higher than 40%. This is true for both types of composites with TiC and TiB. When the particle content was 40% and below, 10e25 times porosity reduction of TiB composites was achieved compared to only 2e4 times reduction in TiC composites under the similar conditions. When the particle content exceeds 40%, the effect of HIP treatment was not as obvious, and the results listed for 60% composites may seem really strange at first glance, suggesting no change in porosity for the TiB composite and its increase for the TiC composite. Needless to say, the latter result must be treated with a high degree of skepticism and require some clarification to understand it correctly. Closer up observation of the sample in Figs. 2(i) and 2(j) indicate that some parts of the structure were lost during sample preparation, so the measured "absence of material" at the surface does not represent the porosity per se which exist in bulk; it rather reflects the structure become very fragile. More detailed images supporting this argument are presented in Fig. S1. However, there is some very important implication of this result. As recently shown by Markovsky et al. [12], the Ti64 alloy composites with 40% of TiC after BEPM fabrication can form continuously reinforced structures that do not have enough ductile phase regions for composite to be successfully plastically deformed. In other words, effective plastic deformation of the matrix is prevented by a rigid and continuous framework made of very hard and brittle reinforcement phase. It is difficult to expect that continuously reinforced composite can be well deformed during HIP to change porosity. On the other hand, with the lacking of a plastic phase, the pressure applied during HIP can damage the fragile structural framework. Summarizing these results, composites made using BEPM with TiC and TiB particles below 40% can be effectively densified by applying further HIP treatment. It is worth to note that when porosity of titanium alloys is below 2%, their performance is not affected strongly. In present study the porosity after HIP was exceeding the threshold value when the particles content was higher than 30%. It is also accepted that when the porosity slightly exceeds the 2% threshold, the reduction in mechanical properties is still not abrupt. However, recently published data on high temperature aging of TiC composites at 1000 C [12] show that the hardness rate increase in the range of 30%e40% of particles content is almost twice as fast as when the content is below 30%. Thus, we expected that even though the porosity was slightly above the threshold, the benefit of increasing hardness would overcome the disadvantage of slightly increased porosity, and for this reason the use of 40% composites were accepted in present study. Based on this evaluation the composites with reinforcement content 40% and below were selected for further HIP experiment to fabricate the plates suitable for the ballistic test.

The microstructure of the tiles after the sintering is shown in Fig. 4. The structure is a uniform lamellar aþb, typical for sintered Ti64 alloy with some inclusions of reinforcing phase particles. As shown by Ivasishin and Moxson [17], sintering results in complete dehydrogenation of the starting TiH2 powder and formation of a titanium-base matrix where aluminum and vanadium uniformly distributed. Current results show that depending on type and content of reinforcing particles, the fraction of residual pores can be relatively high.

The TiC particles are nearly equiaxed but unevenly distributed in MMC (Figs. 4(a)e4(c)). This is due to the tendency for TiC particles form conglomerates at blending stage, and it is commonly observed according to Ivasishin et al. [9]. As the amount of TiC particles increases, their non-uniform distribution becomes more apparent. In addition, the number and size of pores also increases

The structure of MMC reinforced with TiB particles is somewhat different (Figs. 4(d)e4(f)). Despite the similar appearance of the matrix alloy showing a fine-grained aþb microstructure, the reinforcing particles are more evenly distributed. The needle-like TiB inclusions were formed during sintering as a result of reaction TiB2þTi¼2TiB. When TiB2 content was 10% the needles are uniformly distributed over the matrix (Fig. 4(d)). The fraction of residual pores is noticeably greater than in composite with similar amount of TiC. Most likely this is a result of additional porosity formation via Kirkendall effect during the transformation of TiB2 particles into TiB, which was reported before by Sahay et al. [18] and later confirmed by Ivasishin et al. [9]. When the TiB2 content increases to 20%, it become apparent that not all particles are converted into TiB and exist not transformed together with the newly formed TiB needles (Fig. 4(e)). Besides, the fraction of residual pores increases with increasing the borides content (Fig. 2(e)). This trend intensifies when borides content reaches 40% (Fig. 2(f)). The ratio between non-transformed TiB2 and transformed TiB phases increases in favor of the latter one. In addition, the number and size of the aþb matrix alloy regions continue to decrease. Although the pore fraction increases significantly, the pores are fairly evenly distributed (Fig. 4(f)).

The structure of all composites changed significantly after the tiles were HIP bonded into three-layer plates. The interfaces between layers are well consolidated with no visible pores. Most likely the load applied during the HIP resulted in plastic deformation promoting additional dislocations formation, which facilitate the diffusion and better bonding of layers.

The structure of TiC composites after the HIP is shown in Fig. 5. The changes take place for the shape, size, and also the chemical composition of carbide particles with possibility of formation of diffusion zones enriched with carbon (Figs. 5(a)e5(c)). As a result of 3-h HIP at 900 C and applied load, the TiC particles and their aggregates, grow into larger ones, and their outlines become more rounded (Figs. 5(a)e5(c)). The observed changes are result of carbon diffusion from TiC into the alloy, as has been recently shown and discussed in details elsewhere [12]. As a result of such diffusion, TiC can be partially transformed into Ti2C which is an equilibrium phase at elevated temperatures. This causes an increase of reinforcement phase volume fraction. As can be seen, some larger carbide particles systematically exhibit darker tone in their core, whereas the outer regions are much brighter. This indicates heterogeneity in particle composition and it is confirmed with the EDS chemical analysis (Figs. 5(d) and 5(e)). The composition of carbide is close to the stoichiometric TiC in the particle core (points 1 and 2 in Fig. 5(e) and Table 3), but it is carbon-depleted in the outer regions of the particle (points 4 and 3 ibid). Brighter layers adjacent to carbon-depleted regions of particles also contain a marked amount of aluminum, which is slightly higher than in surrounding matrix alloy (point 3 vs. 5). This is a likely result of the formation of the ternary phase Ti3AlC, the possibility of which was recently shown by Markovsky et al. [12] under similar conditions. Chemical composition of ternary phase is quite different from the two-phase aþb Ti64 alloy, primarily due to high amount of carbon, and aluminum.

The used HIP greatly modified the structure and chemistry of composites. It increases the volume fraction of the strengthening phase by partial converting it from TiC into Ti2C by using titanium from the matrix; in other words, it increases the fraction of the harder phase in the expense of the softer one. Another important result is formation of a new interface between the carbide particles and surrounding matrix alloy, which results of chemical reaction between the reinforcement phase and the matrix alloy. All this possible due to diffusion mobility of the elements involved, additionally activated by used high temperature and pressure. Moreover, diffusion of carbon continues further into the already formed alloy matrix from the initial powder matrix, which can be identified by the presence of vanadium, which has the lowest diffusion mobility in titanium according to Zwicker [20]. Finally, another very important consequence of HIP is significant reduction of porosity in all composite layers and their tight bonding (Fig. 5(f)). Without a doubt, all these structural changes should improve the mechanical properties of the laminates.

Composites reinforced with TiB still show some residual porosity after the HIP (Figs. 6(a)e6(c)), but its fraction was drastically reduced after the treatment. The porosity measured in 10% TiB layer was below 1%, and the particles did not change their dispersion character (compare Figs. 6(a) and 4(d)). In the 20% TiB layer, the total pore fraction also decreased significantly after HIP, while the transformation of TiB2 into TiB was essentially complete (compare Figs. 6(b) and 4(e)). In addition, in the 20% TiB layer, majority of particles after HIP exhibit rather plate morphology than needle. The 40% TiB layer differed from 10% to 20% layers by the presence of some of TiB2 particles still not transformed into TiB (Fig. 6(c)). It should be noted that the composite alloy matrix cannot be identified as a two-phase aþb structure. Most likely, the matrix alloy is titanium depleted, since a significant amount of it was reacted to convert TiB2 into TiB. This inevitably makes the alloy enriched with alloying elements to a level that provides formation of relatively small inclusions of a high-alloyed b-phase. Interfaces between layers of TiB laminates (Figs. 6(d) and 6(e)) appear more defective, less consolidated and not as smooth as in TiC structures (compare with Fig. 5(f)). The features of phase and structural transformations in MMC hardened with TiB occurring under HIP are quite complex and diverse, so that a separate work will be performed for their detailed description.

Three-layer plates joined using HIP were machined and polished from both sides (Fig. 7). The HV hardness measured in each triplet layer separately is shown in Table 4. Due to the relatively large size of the carbides, it was possible to measure hardness distinctly in the reinforcing particles and matrix alloy. This was not possible in TiB composites due to the higher particle dispersion, so only the average hardness of the layers was measured.

3.2. Ballistic tests

Each layered plate was tested with two shots. The first shot was always fired to the side with 40% composite layer, and the second shot was always fired into the opposite, 10% composite layer. Images in Fig. 8 shows a sequence of frames taken at the same interval from the test recorded with a high-speed camera. The supplement video (S2) was taken during the first shot, fired at TiB laminate. As one can see, the bullet literally completely shredded on impact. Significant fraction of the bullet energy is converted into flash and fire upon its contact with the armor plate surface. It can also be seen that the destruction of the high strength steel core (Figs. 8(g)e8(i)) occurs after the destruction of the copper jacket of the projectile (Figs. 8(d)e8(f)), which is typical effect.

The first shot practically did not cause any destruction to the plate. The vast majority of debris and crushed pieces observed during the shot are result of a disintegrated bullet. This was confirmed assessing the location of the bullet impact after the shot. A small crater slightly larger than 3 mm in diameter and up to 1 mm in depth was the only damage observed (Fig. 9(a)). The second shot fired at the opposite side, made a significant distraction to the plate (Fig. 9(b)). It created five radial cracks, starting at the bullet struck point and propagating to the edges of the plate. The cracks were throughout the entire plate thickness. The size of the crater after the second shot, was much larger, exceeding 20 mm in diameter and 6 mm in depth (Figs. 9(c)e9(d)). As can be seen in Fig. 9(d), the crack initiated from small crater formed after the first shot on the 40% side, which was apparently a ready-made stress concentrator. After the crack initiated, it spread over all 3-layers of the plate precisely between two opposite craters, with some deviations at the boundaries between the layers (Fig. 9(d)).

The first shot at the TiC laminate plate made a crater 6e8 mm in diameter and about 3 mm deep. A significantly larger crater with a diameter of about 20 mm and a depth of 6 mm was made after the second shot to the opposite side of the plate (Figs. 10(a) and 10(c)). In this test, the plate was also broken only after the second shot and crack was likewise initiated in the crater made after the first shot (Fig. 10(b)). However, in the TiC plate, the shape of the crack was significantly different from that formed in the TiB laminate. It deviates slightly more on individual larger TiC particles and at the interface between the 40% and 20% layers. Besides, the crack changes direction by 90 and spreads along the interface between 20% and 10% layers (Fig. 10(d)).

The differences in damage on opposite sides of the laminated plate with TiB after the first and the second shots are difficult to explain only by comparing the hardness values (N 4, and N 6 in Table 4), especially when compared to similar results obtained on TiC laminate (Fig. 10). This three-layer plate with carbides, having fairly close values of matrix hardness and even higher hardness of TiC particles in comparison with the average hardness of layers with TiB (compare N 1, N 2, and N 3 against N 4, N 5, and N 6, respectively, in Table 4), but is characterized by essentially bigger crater after the first shot (compare Figs. 10(a) and 9(a)).

Thus, comparing the results for both plates after the second firing, it can be concluded that the craters' parameters are quite close, as well as the hardness of the matrix of the layer reinforced with TiC and average hardness of the layer reinforced by TiB are rather close showing (422±7 HV) and (417±2 HV) correspondingly. This suggests that with given combination of aþb matrix of Ti64 alloy and presence of 10% of strengthening particles, both materials behave similarly when fired with given bullets type, regardless of the nature/type of the particles.

The fracture surface near the dent from the shot into 10% TiB layer and details of microstructure are shown in Fig. 11. A general view of the dent in Fig. 11(a) shows a smooth bell-shaped crater.

The initial minor cracks do not destroy the entire plate; in contrast they extend at some depth of the layer creating the stress concentration from the first shot. The direction of its propagation is indicated by red arrows, and the depth of advance is marked by a dotted line. Various locations on the crater rim are marked by letters A-F in Fig. 11(a) and their details are shown in Fig. 11(beg). The images show that the nature of the fracture is virtually the same at all sites. The fracture shows ductile dimples and small rounded inclusions, which are visible in the inset in Fig. 11(d). The melt of lead (Pb labeled) from the bullet can be seen in larger (Figs. 11(b), 11(e), 11(f)), and smaller cracks as small drops.

A thin section cut through the depth of the sample at a fairly significant distance from the crater, at least 8 mm from the fracture (Zone G in Fig. 11(a)), shows many small, randomly oriented cracks (Fig. 11(h)). There is brittle micro-cleavages area observed directly below the center of the crater (Zone H in Fig. 11(a)), made by bullet impact near larger borides particles and their aggregates (Figs. 11(i) and 11(j)). The evaluation of structure supported by the EDS chemical analysis, indicate that both the residual TiB2 and the newly formed TiB particles are extensively cracked. At the same time, microvolumes of the Ti64 alloy saturated with boron2 crack mainly along their boundaries, apparently due to their lower plasticity

When comparing the cross sections of the two different plates, one with TiB and another with TiC shown in Figs. 9 and 10, it can be seen that the TiB composite structure exhibits a more brittle character and looks more like ceramic. This is likely due to the more uniform distribution of TiB particles in the matrix alloy compared to TiC composite. Less uniformly distributed TiC particles leave larger regions of the matrix alloy that locally exhibit higher plasticity. TiB composite structures do not have such regions and behave more like uniform ceramics that are unable to stop or heal a crack when it starts. Whereas the existence of large alloy regions in TiC composites leaves room for plastic mechanisms to be involved, resulting in effective retardation or possibly complete arrest of the crack. Fig. 12 shows 40% and 20% layers of TiC laminate after two shots. The crater in the 40% layer formed after the first shot into this layer, but the crack appeared only after the second shot into the opposite side, in the 10% layer (not shown in this image). The crack as it grows deviates from straight-line propagation on both at the boundary between layers (Fig. 12(a)) and on individual large TiC particles (Fig. 12(b)). Separately, it should be noted that the image clearly shows the presence of two parallel cracks, which are most likely the result of a fracture of a tubular shape, if presented in a 3D perspective. The distance between these two cracks or the diameter of the tube in 3D consideration, is close to 1.2 mm, which correlate well with the diameter of the hard steel core blunt tip of the AP bullet shown in Fig. 3(c). This suggests that the observed specific shape of the damage is a result of the impact of steel core blunt bullet part. It confirms that studied laminate resists well the impact of a projectile designed specifically to break through a plug-in armor material.

This type of structure damage is well consistent with the test results reported on relatively softer materials that used the same armor-piercing bullets. For example, according to Markovsky et al. [21] during ballistic testing of rolled plate made of a conventional alloy Ti64 10e14 mm thick, the same type AP bullets made a similar plug. Unlike conventional Ti64 alloy, MMC based laminate plates may have different damage patterns when the blunt core bullet completely disintegrates (as with the TiB laminate shown in Fig. 8) or can cause the above-mentioned plug-shaped crack to form, which is predetermined by the interplay of hardness and plasticity of the material.

4. Discussion

A general scheme describing the behavior of the studied threelayers MMC plates produced using BEPM combined with HIP can be proposed by summing the above results. The first shot fired into the layer, reinforced with 40% of hard particles, forms a relatively small crater and, naturally, an adjacent zone of accumulated stresses3 (Fig. 13(a)). The second shot fired into the opposite side of the plate, reinforced with 10% of hard particles, forms a larger crater (Fig. 13(b)) due to the lower hardness of this layer since it has a smaller number of hard particles. However, as a result of many factors, strong bending forces arise during the second shot, which in some way interact with the stress fields accumulated from the first shot. Some structural consequences of this interaction are different for composites reinforced with TiC and TiB, but in both, the craters formed after the first shot and adjacent region act as stress concentrator to initiate the cracks. In a 3-layer laminate with TiC, the crack nucleates near the crater in a 40% layer and propagates towards the boundary with a 20% layer adjacent to it, while deviating on individual TiC particles (Fig. 10). Having reached the boundary between the layers, it also deviates somewhat, passes through the second layer in a similar way to the first, and having reached the boundary with the 3rd layer (10% TiC), changes its direction by 90. In this case, the crack is formed in the vertical plane (perpendicular to the layers) for only 40% and 20% of TiC layers, and at the 3-rd layer containing 10% TIC, after the crack direction is rotated 90, it propagates along the boundary between 20% and 10% of TiC layers (Fig. 13(c)).

In the case of a similar double-shot test of the layered composite plate strengthened by TiB particles, both nucleation and crack propagation are noticeably different from TiC part. Even a very small crater that appeared after the first shot A (Fig. 13 (a)) also effectively acts as a stress concentrator and initiates cracks under the action of bending forces D arising from the second impact (Fig. 13(b)). However, unlike the case of TiC laminate, the result of such double-shot test is not a single crack formation, but 5 cracks radiating from the center of the crater formed after the first shot, sown here on schematics as C1 and observed in real test in Fig. 9(b). In addition, all these cracks spread across all 3-layers, only slightly deviating from the rectilinear direction, mainly at the boundaries between the layers.

To provide a meaningful viewpoint on the potential value of the proposed approach, it would be very helpful to compare the relative ballistic performance of presently studied product to standard monolithic armor products, such as RHA steel, or Ti, Mg, or Albased. Unfortunately, the manner in which the ballistic test was carried out in this study does not allow such a comparison to be made very rigid due to the limitation of the size of the plates being tested and their number. However, we can make some reasonable estimates of the possible outcome of such a comparison. This can be done based on recently published data on Ti64 alloy-based armor manufactured using BEPM [7]. Using such parameter as specific kinetic energy (SKE) of bullet (kinetic energy/area of bullet crosssection), it was shown, the Ti64 alloy-based laminate composites require almost twice higher bullet energy to be pierced compared to the currently used Ti64 alloy armor [6]. Another conventional test, V50, has recently been used in a different study [21] that allowed comparison of Ti64 alloy-based materials made using BEPM with open data on commercially available armor made from uniform Ti alloys [6,22,23]. Results of such test can be presented by plotting V50 vs. thickness of the plate, where V50 stays for "Velocity50%", a ballistic test where bullets are fired at changed velocities. The velocity of the bullets where 50% of the bullets don't penetrate, is the V50 rating for that ballistic protection. This test allowed a direct comparison of materials with similar specific density, which can be used to compare layered composite structures made using BEPM with commercially available armor made of Ti64 and other Ti alloys. However, when there is a need to compare the V50 results for some materials with distinctively different density the results can be recalculated accounting for the mass efficiency of tested structures. In this case V50 can be plotted vs. thickness multiplied by density of material. Details of such analysis for different materials are presented elsewhere [21]. Here, we adopt a summary chart from this study on V50 plotted vs. thickness density data shown in Fig. 14 and illustrate how the data from the present study correlate with other materials including Ti-alloys [6,22,23], RHA steel [24e29], aluminum alloy 5083AL [28,30] and magnesium alloy AZ31B [31]. As can be seen, given the mass efficiency, Ti alloys (1) exhibit higher performance than RHA steel (5) and (6) and the listed aluminum (7) and magnesium (8) alloys. The results of 3D printed Ti64 armor, data (2) published in Ref. [21], are generally well consistent with V50 of commercial Ti rolled alloy armor (1). The data set (3) shows our earlier results presented on two and threelayer alloy Ti64-based laminates, which included composite layers with up to 20% of TiC and TiB [7]. Those results look superior over commercial Ti armor (1). For example, for the same thickness density, say at 60 mm g/cm3 , V50 is about 650 m/s and 800e900 m/s for commercial Ti alloys and for BEPM laminates correspondingly, which made the improvement between 19% and 27%. However, the dataset (3) reports on composites with reinforcement content not exceeding 20%, and the structure have not been fully optimized in terms of its porosity. The highest hardness value, 373 HV, was reported for a 10% TiC composite with a measured porosity of 3.6%. Other composites listed in this set (3) demonstrated slightly lower hardness and slightly higher porosity values [7]. Now, if we add materials from the present study for comparison, we should consider structures with hardness between 683 HV and 789 HV and porosity much lower or comparable to the dataset (3). There is no doubt that the V50 results for present structures manufactured using BEPM þ HIP should be significantly higher compared to the data set (3). This prospective is arrowed with (4) in Fig. 14. A similar conclusion can be made about significant increase of SKE required for piercing armor when compare present materials to data reported before for Ti-based laminates [7,21] and monolithic Ti armor product [6].

To summarize, the laminate structures of composites on the base of alloy Ti64 reinforced with TiC and TiB can be successfully formed by making each composite layer individually using BEPM and bonding them into a laminate structure using HIP. The combination of the two processing technologies is highly beneficial for the properties and performance of the final product. The most obvious positive effect from the addition of HIP to the production cycle of composites with BEPM is the almost complete removal of residual porosity in final parts. The removal of porosity is particularly important when the amount of reinforcement exceeds 10% or so, when this structure deficiency may be relatively high after just BEPM manufacture. The need for such structure improvement becomes particularly acute on laminated composites when standard porosity removal passages, such as hot plastic deformation, cannot be used due to mismatch of plastic flows of different layers as was argued by Prikhodko et al. [13]. As can be seen, the hardness of PM made composites can be significantly increased after their subsequent HIP treatment. Thus, HIP allows forming dense multilayer sandwich materials in which individual layers differ significantly in chemistry, structure and properties and provide excellent bonding between heterogeneous layers that are difficult or possible to manufacture in other ways. Finally, the high temperature associated with HIP treatment is beneficial for the strengthening of composites, particularly with TiC and TiB. As recently was shown by Markovsky et al. [12], TiC can be transformed in Ti64 at high temperature into various hard phases such as Ti2C and Ti3AlC, which are both result of phase transformation that increases the fraction of the hard phases in the expense of relatively soft matrix alloy. With respect to TiB composites, the extended effect of high temperature during HIP may be beneficial in allowing Ti2B to more fully transform into TiB, forming homogeneously distributed needle/ plate morphology of reinforcement. It is also important to note that the borides and carbides in the composites in question result in the formation of new strong interfaces between the matrix alloy and the strengthening phase since they are the results of a chemical reaction. Obviously, this leads to an improvement in the ability of the materials to withstand the applied external load or impact. In fairness note, joining the layers could be also done via the diffusion bonding as shown by Prikhodko et al. [32] or friction welding according to another study of Prikhodko et al. [33]. However, HIP bonding appears to be the most proficient way to build laminates (hard/ductile) from these materials, enabling their superior hardness in specific regions in the range 650e780 HV without compromising low specific weight of the whole part. In addition to layer bonding and mechanical properties improvement due to structure aging, HIP is also effective in increasing hardness by reducing porosity, in which diffusion bonding and friction welding are not as effective. Deformation of the entire structure is very useful for effective bonding creating dislocations that accelerate diffusion, and during HIP bonding, deformation of the structure is effortlessly achieved by closing the residual porosity. For the sake of fairness, it should be noted that the diffusion bonding has the potential to improve properties due to aging of the bulk structure, which is completely prevented by friction welding.

5. Conclusions

(1) The laminate composite structures on the base of alloy Ti64 reinforced with TiC and TiB can be successfully formed by making each composite layer individually using BEPM and bonding them into a laminate using HIP. Combination of two processing methods allows (i) to minimalize the residual porosity often observed after BEPM manufacturing, (ii) formation of the compact multilayer sandwich structure with the individual layers differ by their chemistry, microstructure and properties and (iii) provide excellent bonding between dissimilar composite layers that is difficult or possible to joint in other ways.

(2) Ballistic examination of layered composite structures manufactured by BEPM þ HIP processing shows that Ti-based structures can successfully withstand the impact of a standard 7.62 caliber steel hardened bullet with the hardness of 870 HV. This result was due to the outstanding hardness values demonstrated by the fabricated composites, which were measured near 790 HV and 680 HV for 40% composites with TiC and TiB correspondingly

Declaration of competing interest

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.

Acknowledgments

The following authors, PM, OS, DS, VE, SP, acknowledge funding from the NATO Agency Science for Peace and Security (#G5787). Investigations of compression tests influence on phase composition and microstructure of materials were performed within the frames of the Agreement of Cooperation between G.V. Kurdyumov Institute for Metal Physics of N.A.S. of Ukraine and Jarosław Da˛browski Military University of Technology, Poland. Ballistic investigations were co-financed by Military University of Technology in Warsaw under research project UGB 829/2023/WAT. Separate works made in G.V. Kurdyumov Institute for Metal Physics of N.A.S. of Ukraine were partially financially supported by N.A.S. of Ukraine within the frames of project #III09e18.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.dt.2024.04.002.

Sidebar

ARTICLE INFO

Article history:

Received 28 November 2023

Received in revised form

28 January 2024

Accepted 2 April 2024

Available online 15 April 2024

Footnote

* Corresponding author.

E-mail address: [email protected] (S.V. Prikhodko).

Peer review under responsibility of China Ordnance Society

1 A grain is a unit of measurement of mass, and in the troy weight, avoirdupois, and apothecaries' systems, equal to exactly 64.79891 mg.

2 According to phase diagram TieB reported by Murray et al. [34] such concentrations are more than eutectic composition and should cause precipitation of TiB phase. However, since we have here not pure titanium, but an alloy containing also Al, V, and Fe, which was also subjected HIP, it is possible that it still a supersaturated solid solution.

3 Most likely, the area of accumulated internal stresses and probable microcracks spreads in a complex pattern over some considerable distance.

References

References

[1] Lutjering G, Williams JC. Titanium. second ed. Berlin/Heidelberg: Springer; 2007.

[2] Peters M, Kumpfert J, Ward CH, Leyens C. Titanium alloys for aerospace applications. Adv Eng Mater 2003;5(6):419e27.

[3] Niinomi M. Recent metallic materials for biomedical applications. Metall Mater Trans 2002;33A:477.

[4] Jones TL, Kondoh K, Mimoto T, Nakanishi N, Umeda J. The development of Ti6Al-4V alloy via oxygen solid solution strengthening for aerospace and defense applications. Key Eng Mater 2013;551:118e26.

[5] Montgomery JS, Wells MGH. Titanium armor applications in combat vehicles. JOM 2001;4:29e32.

[6] Fanning J. Military application for b titanium alloys. J Mater Eng Perform 2005;14:686e90.

[7] Prikhodko SV, Ivasishin OM, Markovsky PE, Savvakin DG, Stasiuk OO. Titanium armor with gradient structure: advanced technology for fabrication, advanced technologies for security applications. In: Palestini C, editor. NATO science for Peace and security cluster workshop on advanced technologies. Leuven, Belgium: Springer; September 2019, (2020. p. 17e8. 127-143.

[8] Ivasishin OM, Savvakin DG, Stasiuk OO, Golub VА, Mirnenko VІ, Sedov SG, Kurban VА, Antonyuk SL. Microstructure and properties of titanium-based materials promising for antiballistic protection. Progr Phys Met. 2019;20(2): 288e309.

[9] Ivasishin OM, Markovsky PE, Savvakin DG, Stasyuk OO, Sitzman SD, Norouzi Rad M, Prikhodko V. Multi-layered structures of Ti-6Al-4V alloy and TiC and TiB composites on its base fabricated using blended elemental powder metallurgy. J Mater Process Technol 2019;269:172e81.

[10] Markovsky PE, Ivasishin OM, Savvakin DG, Bondarchuk VI, Prikhodko S. Mechanical behavior of titanium-based layered structures fabricated using blended elemental powder metallurgy. J Mater Engineering and Perform 2019;28(9):5772e92.

[11] Di Benedetto G, Matteis P, Scavino G. Impact behavior and ballistic efficiency of armor-piercing projectiles with tool steel cores. Int J Impact Eng 2018;115: 10e8.

[12] Markovsky PE, Savvakin DG, Stasyuk OO, Mecklenburg M, Pozuelo M, Roberts C, Ellison V, Prikhodko SV. Significant hardening effect of hightemperature aging of alloy Ti-6Al-4V composite reinforced with TiC. Mater Des 2023:112208.

[13] Prikhodko SV, Markovsky PE, Savvakin DG, Stasiuk OO, Ivasishin OM. Thermomechanical treatment of titanium based layered structures fabricated by blended elemental powder metallurgy. Mater Sci Forum 2018;941:1384e90.

[14] Kim KT, Yang HC. Densification behavior of titanium alloy powder under hot isostatic pressing. Powder Metall 2001;44:41e7.

[15] Dekhtyar AI, Bondarchuk VI, Nevdacha VV, Kotko AV. The effect of microstructure on porosity healing mechanism of powder near-b titanium alloys under hot isostatic pressing in a þ b-region: Ti-10V-2Fe-3Al. Mater Char 2020;165:110393.

[16] Ivasishin OM, Anokhin VM, Demidik AN, Savvakin DG. Cost-effective blended elemental powder metallurgy of titanium alloys for transportation application. Key Eng Mater 2000;188:55e62.

[17] Ivasishin OM, Moxson VS. Low cost titanium hydride powder metallurgy. In: Qian M, Froes SH, editors. Titanium powder metallurgy, science, technology and applications. Elsevier; 2015. p. 117e48.

[18] Sahay SS, Ravichandran KS, Atr R, Chenm B, Rubin J. Evolution of microstructure and phases in in situ processed Ti-TiB composites containing high volume fractions of TiB whiskers. J Mater Res 1999;14(11):4214e23.

[19] Poondla N, Srivatsan TS, Patnaik A, Petraroli M. A study of the microstructure and hardness of two titanium alloys: commercially pure and Tie6Ale4V. J Alloys Compd 2009;486(1e2):162e7.

[20] Zwicker U. Titan und Titanlegierungen Berlin. Heidelberg: Springer; 1974.

[21] Markovsky PE, Kovalchuk DV, Janiszewski J, Fikus B, Savvakin DG, Stasiuk OO, Oryshych DV, Skoryk MA, Nevmerzhytskyi VI, Bondarchuk VI. New approach for manufacturing Tie6Ale4Vþ40%TiC metal-matrix composites by 3D printing using conic electron beam and cored wire. Pt. 2: layered MMC/alloy materials, their main characteristics, and possible application as ballistic resistant materials. Usp. Fiz. Met. 2023;24(4):741e62.

[22] Davis AE, Kennedy JR, Strong D, Kovalchuk D, Porter S, Prangnell PB. Tailoring equiaxed b-grain structures in Ti-6Al-4V coaxial electron beam wire additive manufacturing. Materialia 2021;20:101202.

[23] Fanning J. Proceedings of titanium world conference 'titanium 99. Science and Technology'; 1999.

[24] Walicki M, Janiszewski J, Cieplak K. Estimation of ballistic performance of armor steels based on the split Hopkinson shear bar data. J Theor Appl Mech 2022;60(1):129e40.

[25] Diederen AM, Broos JPF, Trigt SNV, Peijen MCP. Ballistic protection against armour piercing projectiles using titanium base armour specialists' meeting on cost effective application of titanium alloys in military platforms. for NATO AVT Panel 2001.

[26] Bekci ML, Canpolat BH, Usta E, Güler MS, Cora ON. Ballistic performances of € Ramor 500 and Ramor 550 armor steels at mono and bilayered plate configurations. Eng. Sci. Technol. Int. J. 2021;24(4):990e5.

[27] SSAB, Armox Protection Plate, Protection Steel Buildings Data Sheet. https:// ssabwebsitecdn.azureedge.net/-/media/files/en/armox/armox-protectionsteel-in-buildings-en.pdf?m¼20170619110513.

[28] Backman ME, Goldsmith W. The mechanics of penetration of projectiles into targets. Int J Eng Sci 1978;16(1):1e99.

[29] Janiszewski J, Gra˛zka M, Tria DE, Surma Z, Fikus B. Laboratory investigations on perforation of 30PM steel plates. Problems of mechatronics. Armament, aviation. Safety Engineering 2016;7(2):19e40.

[30] Cheeseman B, Gooch W, Burkins M. Ballistic evaluation of aluminum 2139-T8. Preprint 24th int. Ballistics symposium (new orleans, LA, USA. 22e26 September 2008.

[31] Jones TL, DeLorme RD. Development of a ballistic specification for magnesium alloy AZ31B. 23rd int. Symposium on ballistics. Spain: Tarragona; April 2007. p. 16e20.

[32] Prikhodko SV, Savvakin DG, Markovsky PE, Stasuk OO, Penney J, Shirzadi AA, Davies PD, Davies HM. Diffusion bonding of TiC or TiB reinforced Tie6Ale4V matrix composites to conventional Tie6Ale4V alloy. Sci Technol Weld Join 2020;25(1):1e7.

[33] Prikhodko SV, Savvakin DG, Markovsky PE, Stasuk OO, Penney J, Enzingen N, Gaskil M, Deley F. Friction welding of conventional Ti-6Al-4V alloy with a Ti6Al-4V based metal matrix composite reinforced by TiC. Weld World 2021;65:415e28.

[34] Murray JL, Liao PK, Spear KE. The BTi (Boron-Titanium) system. Bul Alloy Phase Diagram 1986;7:550e5.