Introduction

Modern artillery is developing toward higher chamber pressures, increased firing range, improved firing accuracy and extended barrel life. Gun bore surface can be progressively damaged by gunpowder, leading to an enlargement of inner barrel’s diameter. Ultimately, the effectiveness of the weapon will diminish due to a decrease in muzzle velocity and firing accuracy1,2,3.

Internal ballistics is a complex field that involves the study of strong and metastable chemical energetic material, which is capable of engendering hot gases containing a large amount of energy. The gunpowder ignition can bring about sharp pressure exceeding 300 MPa and extremely high temperature (ranging from 2600 to 3500 °C) in the combustion chamber, generating significant force on the projectile. The intense thermal erosion and chemical attack – referred to the thermal-chemical-mechanical gun bore erosion4, can directly impact the gun bore surface. This process weakens the internal compressive stress due to the autofrettage process5, induces martensitic transformation, and leads to the formation of detrimental external cracks6, pits, and chemically-affected zone7. Consequently, anti-erosion protective coatings serve as an effective means to isolate the erosion environment from the gun bore surface8,9, thereby extending the service life of the firearm.

Electrodeposited chromium coating is a crucial surface coating technology, which is widely applied in modern barrel weapons10,11,12. In order to access the failure mode of the potential coatings on large gun bore surfaces, the most effective method for evaluating the erosion life is through live fire shooting13,14. However, its approach involves high costs and risks, necessitating systematic organization of the firing range setup, ammunition configuration, and shooting commands. Additionally, external environmental conditions, such as weather and environment, must be taken into account. These complexities make coating assessments more challenging. Therefore, vented vessel test (VVT) has emerged as a prominent means of characterizing the erosion behavior of gun bore coatings15. As an alternation of the live-fire testing, the VVT process is explored to replicate the thermal-pressure coupling effect, acting on the gun barrel which suffers the thermo-chemical erosion from propellant combustion gases16. Underwood et al. 17 reported that the electrodeposited Cr coating caused thermal damage to the deep substrate, resulting in a thin layer of untempered martensite phase transformation near the substrate surface and grain-growth softening phenomenon after 100 VVT (260 ~ 310 MPa) firings. Mulligan et al. 18 noted that chromium exhibited the typical thermally induced “mud-flat” cracks after 100 VVT firings (results of Underwood et al.), while sputtered Cr coating, as a comparison, showed a much wider crack spacing with interspersed fine cracks. This indicates that magnetron-sputtering technology can obtain good coating adhesion in gun bore applications.

As a potential candidate for Cr substitute material, tantalum was often employed to try in large caliber gun, due to its excellent ductility, crack resistance, high temperature strength, high melting point, and outstanding corrosion resistance to gunpowder residues3,19,20,21. However, according to Mulligan’s research18, direct sputtered Ta coating on steel substrate can experience catastrophic adhesive failures after only one or two firing cycles. And this key influencing factor is the brittle β-Ta phase, which formed readily on the steel substrate at low deposition temperatures22,23,24,25. The brittle characteristics of β-Ta phase cannot bear the significant load (n×102 MPa) during VVT firing. Therefore, pure α-Ta phase on the steel substrate is crucial for good adhesion and effectively replacing Cr coating in hexavalent chromium electroplating, which often underperforms.

In this paper, Ta coating was prepared on the inner wall of annular steel tubes using high-power impulse magnetron sputtering (HiPIMS) technology. A pure α-Ta phase was obtained throughout the entire coating, which was meticulously examined at the surface and near-substrate regions. Subsequently, VVT firing was also explored to evaluate the adhesion and anti-erosion performance of the coatings. Finally, the failure mechanism after 10 VVT firings was systematically analyzed.

Results

Structure of as-deposited Ta coating

Figure 1a displays the SEM cross-sectional morphology of as-deposited Ta coating. The coating possesses dense microstructures with no visible pores. No crack or delamination emerges at the coating interfaces, implying the strong interfacial bonding between the coatings and PCrNi3MoVE steel substrate. The coating is uniform with about 25 μm in thickness on the sand-blasted rumpling surface.

Fig. 1: Characterization of as-deposited Ta coating on short smooth tube.
figure 1

a cross-sectional morphology by SEM; b XRD patterns; c cross-sectional morphology by TEM; d Enlarged morphology in the large red circles, and the corresponding NBDs in the small red circles.

Full size image

Figure 1b presents the phase composition of Ta coating on the inner tube. Pure α-Ta phase (PDF #04-0788) was identified with a preferential orientation of <110> peak. Unlike regular metals such as Fe, Cr, Ni, etc., some specific metals with high atomic number Z and density ρ, e.g. tantalum (Z = 73, ρ = 16.69 g cm−3), exhibit a high linear absorption coefficient for X-rays. This characteristic reflects that tantalum greatly reduces the penetration depth of X-rays26. Thus, this will lead to a reduced contribution from phases present in the core of other potential phases, which may be situated near the substrate-coating interface of a thick coating sample. For instance, β-Ta phase (PDF #25-1280) can be readily grown on steel substrate under sputtering conditions20,27. From previous studies28,29,30,31, preferential growth of β-Ta phase at the substrate surface, which is typically undetectable by XRD technology through thick coatings, is quite common. This phenomenon may be attributed to the epitaxial growth of oxygen-induced δ-TaO2 structure22. Hence, the positional sampling at the coating/substrate interface by FIB technology was carried out, in order to differentiate the phase composition by TEM observations.

Figures 1c and d show the bright-field TEM images at the Fe/Ta interface. The gradient refinement of grain size from substrate surface to deeper body, resulting from the sand blasting process on the steel substrate is evident. This refinement is intended to resist intrinsic crack propagation, which may be induced by the coating crack after a few firing rounds32. For the Ta coatings on tube surface, a pronounced columnar growth mode with a very high coating density can be observed and no voids can be detected between grain boundaries throughout the coating thickness. Figure 3d shows an enlarged version of framed zone in Fig. 1c. The width of the fine columns is about 50 nm and basically consistent from Fe/Ta interface to external surface, indicating that the Ta coatings were grown essentially with relatively high energetic species in HiPIMS technology. Nanobeam diffraction (NBD) patterns, obtained in white and dark regions complemented with bright field images, demonstrate that our deposition process gives coatings with a pure α-Ta phase. In other words, the phase composition of Ta coatings was maintained from the growth initiation till the end of coating deposition, according to the TEM and XRD results. No phase transformation happens during the whole HiPIMS deposition course. In contrast to SEM investigations which did not reveal a columnar structure in the coating, columnar features are visible in TEM micrographs. The length of columnar fragments ranges about 250 ~ 400 nm, as the nanocrystallites exhibit a random distribution of grain orientations. It is assumed that the incident ion energy of about 150 eV which is roughly relevant to the applied substrate bias, contributes to this phenomenon. Such high energetic ion bombardment can generate a significant number of defects, impeding local epitaxial growth on their individual columns and facilitating a re-nucleation mechanism like a tissue phase segregated at the crystallite boundaries33. Moreover, high energy delivered in the Ta coatings promotes their densification to an acceptable level of radiation damage and gas incorporation34.

Performance in VVT firing

Figure 2 presents photographs of the rifled tube after different VVT firings, with the flow direction of propellant gas indicated by the red arrows. Figures 2a to c display the samples after 1, 5, and 10 VVT firings, locally captured using an optical camera. There is no coating delamination on the undamaged substrate, demonstrating the strong adhesion of Ta coating to steel substrate, which is better than Mulligan’s results18. This observation further confirms that Ta coatings without any brittle β-Ta phase exhibit remarkable stability, and can withstand significant load of hundreds of megapascals during VVT firing. Figure 2d illustrates the condition of the edge after 10 VVT firings, oriented against the direction of the propellant gas. The metal luster, located well below the coating plane, is visible, indicating that the Ta coating surface has not been burnished by the explosive gunpowder gas. Consequently, the area exhibiting metallic luster is identified as the exposed steel substrate. This phenomenon can be attributed to the coating deposition process, where the inner surface of samples directly faced the columnar Ta target. The samples were positioned and secured by the sample holder, preventing Ta coating deposition at the edge. As a result, the substrate edge herein was straightly exposed to the gunpowder blast, making it susceptible to erosion. Figures 2e and f display the Ta coating sections of the lands and grooves after 10 VVT firings, as inspected under a stereomicroscope. Based on the surface status of Ta coatings, there are almost no eroding traces along the firing direction, and the nodulus-like morphology of as-deposited coating growth is still kept undestroyed and distinct. However, a certain number of bulge phenomenon can be seen, seemingly emerging a kind of uniform failure under the coating. This phenomenon will be analyzed in detail later.

Fig. 2: Images after vented vessel test.
figure 2

a overall photo after 1 vented vessel firing; b overall photo after 5 vented vessel firings; c overall photo after 10 vented vessel firings; d photo of sample edge at the gunpowder end after 10 vented vessel firings; e photo of inner rifled tube after 10 vented vessel firings by stereomicroscope; f enlarged photo by stereomicroscope.

Full size image

Figure 3a shows the mass changes of the Ta coated samples after each firing round. During the initial six rounds, the mass is unchanged and even increased. This initial weight gain is maybe attributed to the residual gunpowder that is difficult to remove by regular cleaning. The return to original weight suggests that the adhered gunpowder was expelled by the following rounds, maintaining this stable state maintains till the sixth round. From the 7th to 10th rounds, the mass started to decrease with the total mass loss of 0.07 g, possibly because of the substrate depletion at the circular edge, as detected in Fig. 2d. In comparison, the average mass loss for the first 10 rounds is 5.9 g under a slightly low pressure of 300 MPa in our previous study35, which is approximately two orders of magnitude greater than that of Ta coating. This further proves the excellent anti-ablation capability of stable α-Ta phase. Figure 3b presents the XRD results after 10 VVT firings. The diffraction peaks can be in good accordance with tantalum oxides, carbides, and the reaction products are formed from tantalum interacting with nitrocellulose propellant. However, all α-Ta peaks were shifted towards higher angles, ascribing to the stress adsorption and accumulation in Ta coating by the 320 MPa compact. Figure 3c exhibits the metallographic photograph of cross-sectional morphology by the metallurgical microscope, etched by 4% nitric acid alcohol solution. The outermost white layer is the Ta coating with the thickness of still about 25 μm, with a few penetrating vertical cracks. Beneath Ta coating, white-bright zone can also be identified as cementite (Fe3C compound) with a thickness of about 74 μm. Another 74 μm thick region located further below the white-bright zone is chaotic decarbonized sorbite with a strip morphology. And then deeper is the substrate body, which remains unaffected by the transient heating from the exploded gunpowder. Therefore, the thickness of heat-affected zone beneath the 25 μm thick Ta coating reaches about 150 μm. Figure 3d lists the gradient hardness from coating surface to deeper substrate. In the Ta coating, the hardness was increased from Hv521.5 to Hv633.3 as the depth increased, indicating that tantalum coating is also slightly softened by the ultra-high temperature generated during the gunpowder explosion. At the Fe/Ta interface, the Vickers hardness should be at least Hv633.3, as this marks the boundary between the unaffected Ta and hard cementite zone. However, the hardness was fallen down to only Hv533.7, and the reason for this will be explored later.

Fig. 3: Characterization of Ta coatings after vented vessel test.
figure 3

a mass change of the Ta coated sample after each firing rounds in average; b XRD patterns after 10 vented vessel firings; c metallographic photograph of cross-sectional morphology; d cross-sectional hardness vs depth distribution from coating surface to deep substrate.

Full size image

Figure 4a shows the surface morphology of Ta coatings after 10 VVT firings. Small cracks, with a density of 3 ~ 4 per millimeter, can be detected, though they are difficult to observe with a stereomicroscope. In other areas, the nodular morphology of as-deposited Ta coating remains largely intact, showing few signs of erosion. Figure 4b displays the cross-sectional morphology of Ta coatings after 10 VVT firings. At the surface of Ta coating, a chemically affected zone, which is reacted and corroded due to exposure to ultra-high temperature gunpowder, can be observed for less than 2.1 μm in thickness, which also can be supported by energy dispersion spectrum (EDS) mapping results in Fig. 4c. On both sides of vertical crack, the Ta coating remains also very compact with no reduction in coating thickness. In Fig. 4c, except oxygen enrichment at the Ta coating surface, carbon distribution in the steel substrate is not distinguishable in EDS analysis. Generally, the characteristic X-ray energy of low atomic number (Carbon, Z = 6) is quite low, and these low X-ray energies are easily absorbed by the substrate. Consequently, conducting quantitative analysis based on statistical theory becomes more difficult.

Fig. 4: Characterization of Ta coatings after 10 vented vessel firings by SEM observations.
figure 4

a surface morphology; b cross-sectional morphology; c EDS elemental mappings.

Full size image

Thus, EPMA was employed to precisely determine the C distribution, utilizing a soft X-ray analysis spectrometer. The corresponding EPMA at the Fe/Ta interface was explored, as shown in Fig. 5. In the Fe/Ta interfacial region, there is a significantly high concentration of carbon, influenced by external thermal cycling, leading to unbalanced behavior of uneven carbon distribution. The carbon enrichment in the substrate region near the interface is greater than that in deeper regions of the steel substrate, which is related to the presence of Fe3C cementite structure. It is noteworthy that carbon concentration is distinctly visible between Ta coating and steel substrate, resembling a gap traversing along the interface. This phenomenon may be attributed to two factors: i. decarbonization behavior from inner steel to the substrate surface; ii. separation of the coating from the substrate, followed by the filling of the resulting voids with the diamond polishing paste during metallographic sample preparation. In order to distinguish it, an FIB sampling was performed at a carbon site with comparatively low content, as the Ta coating is well bonded to the Fe substrate without cracks or delamination at this location. Here, the diamond polishing paste cannot enter into the interface, due to the high interfacial energy which remains intact during the metallographic sample preparation.

Fig. 5
figure 5

EPMA elemental maps and quantitative analysis of Fe, Ta, Cr and C at the Fe/Ta interface.

Full size image

Therefore, TEM observation at the upper site was shown in Fig. 6. The FIB sampling was carried out at the compact interface from the cross-sectional sample into deeper interface. No opportunity for polishing taste was provided inward, in order to avoid damaging the interface, as framed in Fig. 6a. Figure 6b exhibits the entire FIB sample. Some interfacial regions are tightly connected, while others exhibit significant separation. In the yellow framed zone, an enlarged high-angle annular dark-field (HAADF) in the scanning transmission election microscope (STEM) mode is analyzed in Fig. 6c. The carbon distribution mapping reveals distinct variations in carbon concentration, indicating that decarbonization behavior is prevalent not only at the interfacial separations but also within the tightly bonded zones. For the opening site, we also characterized the interfacial status, shown in Fig. 7. The steel substrate and Ta coating was completely separated by excessive carbon concentration, and stringy sub-micron-sized holes urged the Ta coating to deform and crack internally, indirectly confirming its good plastic deformation capability. Compared with Fig. 6c and Fig. 7a, the carbon diffusion behaviors are different. The carbon distribution in the compact Ta coating is relatively gradient, well obeying the Wagner diffusion model, while that in cracked Ta coating is concentrated at the crack borders, as shown in Figs. 7b–e. Furthermore, the carbon was well adhered with the Ta coating at the crack, which may inhibit future crack healing behavior during subsequent firing rounds. Moreover, this kind of opening hole at the Fe/Ta interface can also make the interfacial hardness decrease, because the loaded stress by Vicker’s Hardness will be relaxed by this kind of holes. That is the reason for hardness fall at the Fe/Ta interface.

Fig. 6: TEM observation of Ta coatings after vented vessel test.
figure 6

a FIB sampling site in the FIB-SEM view; b TEM observation of Fe/Ta interface site; c HAADF-STEM morphology at the yellow solid circle and its corresponding EDS mapping of Fe, Ta and C distribution.

Full size image
Fig. 7: Detailed TEM observations after vented vessel test.
figure 7

a Bright-field images; b HRTEM morphology in the green circle; c Magnified HRTEM morphology of the yellow circle; d Fast fourier transformation in the pink frame; e Fast fourier transformation in the red frame.

Full size image

Discussion

Tantalum alloy possesses many desirable material qualities, such as its high melting point (2996 °C) and exceptional strength at high temperatures, making it a powerful alternative to Cr plating for barrel coatings. β-Ta phase is brittle and unsuitable for protective coatings for gun bore. In contrast, the equilibrium α-Ta phase, with its thermodynamically stable body-centered cubic (bcc) structure and high flexible toughness, has been identified as a material of interest for scale-up testing by Benet Laboratories36. Recently, Cho et al. 21 obtained the α-Ta phase on steel by bombarding the substrate with more energetic ions, thereby avoiding the formation of the β-Ta phase that occurs under low-energy conditions. In our experiment, 150 eV Ta+ ion energy was utilized to deposit pure α-Ta phase, as confirmed by the TEM and XRD results in Fig. 1. Achieving a pure α-Ta phase throughout the entire coating is a fundamental requirement for ensuring its suitability for use in large gun barrels.

Live fire assessment is the most effective way but of great difficulty to evaluate the performance of this pure α-Ta coating. This challenge arises from the substantial economic costs associated with real experimental barrels and the large quantities of shells required. Therefore, cylindrical VVT firings should be adopted to simulate the exploding environment, as this approach is much more convenient to test the new coatings. In this paper, the single base propellant can produce the environmental temperature exceeding 3000 °C37 and pressure of approximately 320 MPa. Transient exploding temperature surpasses the melting point of Ta; however, the temperature absorbed by steel bore surface is less than 1350 °C35. Reaching the austenite transformation temperature and then cooling quickly, martensite transformation will take place, inducing brittle tendency for the gun bore without protective coatings. If Ta deposits on steel, the thermal convective characteristics will change. Thermal properties of steel and tantalum is listed in Table 1. From Table 1, it can be inferred that the heat generated by the gunpowder is radiated away 11% more than that of steel substrate, according to the emission values. Although the specific heat and thermal conductivity of Ta facilitate heat transfer, the thermal consumption during inward energy transformation definitely happens by tantalum coating with the high density (16.7 g/cm3). The total heat quantity decreased due to the comparatively high emission and heat transfer losses, resulting in a significant reduction of substrate temperature from the residual heat, shown in Fig. 8a. The temperature is far below the austenite transition point (Fig. 3c). Only the sorbite appears in the metallographic structure, which formed typically between Ac1 critical temperature of 773 °C and the Ac3 critical temperature of 884 °C38 for 32CrMo steel. The reduced temperature range compared to 1350 °C on steel bore in our previous research is at least 400 °C. Therefore, Ta coating greatly mitigates the heat-induced alterations in material properties. Within the temperature range of 773 ~ 884 °C, another degradation process, decarbonization, is active in medium carbon steel. In the high-temperature region of steel surface, the carbon atoms possess high activity, leading to outward diffusion of carbon. It results in the carbon enrichment near the surface of steel substrate, thus forming Fe3C cementite here and a decarbonized sorbite layer beneath.

Fig. 8: Failure mechanism of Fe/Ta system.
figure 8

a Thermal convective mode; b Failure mode.

Full size image
Table 1 Thermodynamic parameter of steel and Ta
Full size table

As discussed above, outward carbon diffusion can initially cause carbon to gather at substrate surface. The presence of tantalum coating prevents carbon loss. The transient heat convection to the Fe/Ta interface prompts a reaction between carbon and tantalum, resulting in the formation of B1-type TaC or A3-type TaC0.539 at temperatures exceeding 500 °C. Therefore, some carbon can diffuse into tantalum coating, as illustrated in Figs. 6c and 7a. The high-temperature duration is limited to only about 6.3 ms, as determined from the measured pressure curves in Fig. 10c. Here, we calculate the diffusion rates of C in Ta and Fe according to the Eq. (1)40 and (2)41, respectively.

$$D=0.012\exp \left(\frac{-1.69\times {10}^{5}}{RT}\right)$$
(1)
$$D=0.022\exp \left(\frac{-8.42\times {10}^{5}}{RT}\right)$$
(2)

Through calculations at the selective temperature of 800 °C, the diffusion coefficients of C in Ta and Fe are 7.42 × 10−11 cm2/s and 1.77 × 10−6 cm2/s, respectively. It is evident that carbon diffusion rate in Fe is nearly 5 orders of magnitude faster than that in Ta. The significant disparity in interfacial carbon consumption by tantalum coating, compared to carbon evolution from the steel substrate, can inevitably lead to carbon accumulation beneath Ta coating. This is consistent with the phenomena of Figs. 5, 6 and 7a. Since the characterization of gunpowder explosion is transient heating and pressurization, the pressure in all directions may be similar at the lands according to the aerodynamic theory42. However, the temperature distribution can vary significantly during the non-equilibrium course, which lasts for only a few milliseconds. Consequently, the state of carbon evolution beneath the coating varies considerably. In most areas, only thin layer of carbon precipitates on steel surface, as shown in Figs. 3c and 4b. In contrast, substantial carbon accumulation in a few localized areas can cause the tantalum coating to bulge upward, resulting in the deformation depicted in Figs. 2e and f. The overall schematic diagram illustrating the coating failure mode is presented in Fig. 8b. Large-scale coating deformation can be induced by carbon blister and piling-up, resulting in unavoidable cracks like Fig. 7a. Rapid carbon diffusion contributes to the interfacial irregularities, which propagate along the cracks. Some carbon may adhere to the crack walls of Ta coating (Figs. 7b to e), while the remainder can directly be ejected outside, creating the opening holes as seen in Fig. 2f. Therefore, the primary failure mode of Ta coating of 25 μm thick under VVT firings is decarbonization under gunpowder explosions, and the external collapse of interfacial carbon will lead to the coating bulges and opening cracks.

Methods

Coating preparation

The present study utilized a PCrNi3MoVE gun steel substrate, with its chemical composition detailed in Table 2. Ring-shaped samples were machined into specially designed rifled rings for VVT firing and smooth regular rings for the characterization of as-deposited Ta coating, and the machining dimensions of rifled rings and the actual ring samples for VVT were shown in Fig. 9. Figures 9a–d show the detailed size of the rifled sample, the real rifled sample before deposition, the Ta coated rifled sample and the Ta coated smooth tube, respectively. The smooth, short tube samples (Fig. 9d) were used to investigate the morphologies and phase structures, as shown in Fig. 9. Initially, all the rifled samples were polished gradually until 800# honing stone by reciprocating mold polishing machine, and the regular samples were polished gradually till the 800# abrasive paper on the lathe. Then, all the rings were sand-blasted by white alumina of 220 mesh at a pressure of 0.4 MPa from a distance of 100 mm. Finally, all the samples were cleaned ultrasonically in an ethanol solution for about 10 minutes, before being mounted around a cylindrical Ta target (purity: 99.995%, external diameter: 40 mm, length: 600 mm) with a uniform target-substrate distance of 20 mm. Rotational speed of the permanent magnet inside the cylindrical target is rotating with a speed of 8 rad/min.

Fig. 9: Sample dimensions and preparations.
figure 9

a shape of the specimen of rifled tube, and the dimension unit here is millimeter; b photo of the actual rifled tube; c photo of Ta coated rifled tube; d photo of smooth tube.

Full size image
Table 2 Chemical composition of the PCrNi3MoVA gun steel
Full size table

The chamber was evacuated to a vacuum lower than 5 × 10−2 Pa, then heated to 240 °C. A base pressure of <3 × 10−3 Pa was achieved before sputtering. Then, the substrates were etched in Ar plasma for 30 min prior to deposition for adhesion enhancement, removal of native oxide and contaminates on the steel surface. After that, the Ar partial pressure was maintained at 4.5 × 10−1 Pa. A high-power impulse magnetron sputtering (HiPIMS, HiPIMS-400A, Xinbo Technology (Dongguan) Co., Ltd. China) power supply was used in this study. The voltage peak, pulse frequency and pulse width were set as 950 V, 1000 Hz and 100 μs, respectively. Correspondingly, the average power and peak current were adjusted as 4.2 kW and 160 A, respectively. The pulsed bias on the substrate was set at 150 V, and the deposition time was lasted for 4.5 h. After all the deposition procedure, the samples were cooled down to room temperature in vacuum. The Ta coated samples of rifled and smooth tube were exhibited in Figs. 1c and 1d, respectively.

VVT firing

The VVT firing was conducted using a vented vessel device, as illustrated in the schematic diagram (Fig. 10a) and actual device photo (Fig. 10b). As depicted in Fig. 10a, the vented vessel device is roughly composed of three chambers, namely, the combustion chamber, the test chamber and the venting chamber. Detailed principles of VVT firing course have been described elsewhere35. A single base propellant is used during our VVT firing. Nitrocellulose, an energetic material, is commonly used as the primary component in the formulation of single base propellants. For each firing round, 180 g propellant is loaded into the 600-mL combustion chamber. Figure 2c exhibits a comparison of the pressure curves of three different firing courses. The consistent trend of the three curves illustrates that the vented vessel device can provide a similar explosive environment with a transient peak pressure of approximately 320 MPa. Although the ignition responses varied slightly for several milliseconds, the gunpowder action on the tube samples remained nearly identical for each round. After each firing, the eroded surface was cleaned with a non-woven fabric soaked in alcohol, and the samples were carefully weighed on an electronic balance with an accuracy of 0.01 g. The mass change was calculated to evaluate the coating performance. In this study, each specimen underwent 10 matching rounds of firing.

Fig. 10: Detailed diagram and descriptions of vented vessel test.
figure 10

a principal diagram of vented vessel device (1. Igniter plug; 2. Vented vessel body; 3/7. Pressure gauge; 4. Combustion chamber; 5. Test specimen; 6. Cylindrical mandrel; 8. Seal component; 9. Pressure controlled sheet); b photo of the actual vented vessel device; c Measured pressure curves for vented vessel tests.

Full size image

Characterization

The as-deposited rings and rifled samples after VVT firing were cut by electric spark cutting equipment into small pieces. Some pieces were ultrasonically cleaned by alcohol for 5 min, in order to observe the surface morphology after drying in air. From Fig. 2f, the ordinary and bulge parts were cut over precisely, mounted into resin and then grinded till 1.5 grit polishing paste carefully. After that, in order to clear the polishing paste, the mounted samples were ultrasonically cleared out in the double-distilled water for 10 min and then alcohol for 5 min, and finally dry in cold air. This is ready for cross-sectional inspections. An optical microscope (Observer. Z1m, Zeiss), a stereomicroscope (Stemi 508, Zeiss), and a field emission scanning electron microscopy (SEM, accelerating voltage: 15 kV, JSM-7800F, Electronics, Japan) coupled with an energy dispersive spectrometer (EDS), were employed to observe the surface and cross-sectional morphologies. An electron probe micro-analyzer (EPMA, JEM-IHP200F, JEOL, Japan) was used to determine the elemental distributions at the interface for the cross-sectional samples after 10 vented vessel test. A Vickers hardness tester (Digivicker 1000 A, load: 25 g, loading time: 10 s, Mega instruments) was used to assess the cross-sectional gradient hardness from the surface area to deep substrate. X-ray diffraction (XRD, Smart Lab 9kw, Rigaku, Cu Ka radiation, 45 kV, 200 mA) was conducted to analyze the sample surface before and after VVT firing. Micro-area samples at the Fe/Ta interface before and after VVT firing were extracted using a focused ion beam (FIB, Thermo Scientific Scios 2), in order to observe the details of coating structure. A transmission electron microscope (TEM, Talos F200X G2) was used to obtain detailed information about these micro-area samples.