1 Introduction

Cold spraying is a solid-state material deposition technology in which a high-pressure gas (air, nitrogen, or helium) is used to accelerate metal powders with particle sizes of 5–50 μm to a high velocity (300–1500 m/s). Materials are deposited in the solid state owing to severe plastic deformation of sprayed particles upon impact with the substrate [1,2,3]. The mechanism of solid particle adhesion has been explained in terms of adiabatic shear instability [4,5,6] as well as through hydrodynamic analysis of shock wave propagation within a deforming particle [7, 8].

Cold spraying was initially used as a low-temperature coating deposition method suitable for spraying soft metals such as copper, aluminum, and zinc [9]. However, the ability to produce thick near-net-shape deposits extended the scope of cold spray application. For example, cold spraying can also be used for local repairs of worn and corroded aluminum alloy, bronze, and copper parts [10]. Cold spraying is currently regarded as a low-temperature additive manufacturing technology with a relatively low spatial resolution compared to that of laser or electron beam additive manufacturing techniques but with a significantly higher production rate [11,12,13,14,15,16,17,18,19,20,21].

One of the major advantages of cold spraying is the possibility of building up titanium and titanium alloy deposits under ambient conditions [22,23,24,25,26,27,28,29,30,31,32]. The process temperature is relatively low; the nitrogen or helium jet temperature is as high as 1000 °C, and the particle impact temperature is as high as 600 °C. Consequently, cold spraying can limit the in-flight oxidation of individual particles and the entire deposit during production, which is especially important for titanium powders. However, owing to their high tensile strength and low density, titanium and titanium alloys are considered marginally suitable for cold spraying. An analysis of single-particle impacts showed that the optimum impact velocity required to produce low-porosity deposits of pure titanium and Ti6Al4V should exceed 800 and 1100 m/s, respectively [33]. However, acceleration of 10–50-μm particles to these high impact velocities within a nitrogen gas stream is considered to be a somewhat complicated task. For pure titanium, the required velocity can be obtained by applying nitrogen at a stagnation pressure of 4–5 MPa and a stagnation temperature of 1000 °C, which is impracticable for Ti6Al4V [24, 26, 27]. If helium is used instead of nitrogen, Ti6Al4V particles can be accelerated to speeds of 1100–1300 m/s. However, the Ti6Al4V deposits still exhibit some residual porosity even in this case [27]. Moreover, the high cost of helium significantly limits the application of cold spraying.

Note that typically the mechanical properties of as-built cold-sprayed parts differ considerably from those of the bulk materials. In particular, low ductility and toughness were reported for titanium, copper, aluminum, and stainless steel parts. This fragile behavior was explained by the fine grain size at the particle–particle interface and the limited material diffusion through the boundaries between adjacent particles [33]. Thus, heat treatment could be applied to the deposits to increase the material’s ductility. For low-porosity pure titanium deposits, heat treatment at 1000 °C increases the elongation from less than 1% to 7–8%. However, heat treatment does not affect the residual porosity. Similar observations were reported for Ti6Al4V parts [27]. Specifically, it was reported that as-sprayed deposits manufactured using nitrogen and helium showed ultimate tensile strength (UTS) values of ~ 200 and 500 MPa, respectively, and the elongation at break was < 1% in both cases. Heat treatment at 1000 °C for 4 h increased the UTS and elongation to ~ 420 MPa and 6% for the samples sprayed with nitrogen and to ~ 800 MPa and 6%, respectively, for the samples sprayed with helium. However, even the samples produced using helium showed UTS values that were significantly lower than those of the bulk materials (1100 MPa with 10–15% elongation), which could typically be explained by the residual porosity [27].

In this regard, hot isostatic pressing (HIP) of cold-sprayed titanium deposits seems to be a promising post-processing technique capable of improving the cohesion strength as well as the material density. During this process, the high pressure caused by hot pressurized gas around the part affects the stress around a single pore, inducing pore shrinkage, and accelerating the diffusion of material [34]. Two types of HIP are commonly used: pressing within deformable capsules and capsule-free pressing (Fig. 1) [35]. Capsule-free HIP can be used to densify parts with closed well-isolated pores. In this method, the part is placed in direct contact with the hot-pressurized gas inside the HIP vessel. The hot-pressurized environment inside the vessel induces gas absorption by the pore walls, followed by pore shrinkage. However, for high open porosity, the capsule-free method cannot provide efficient material densification owing to gas transfer between pores. In this case, HIP within deformable capsules (so-called encapsulated HIP treatment) can be applied. During encapsulated HIP, the part is closed in an impermeable capsule. The gas inside the capsule is removed by a vacuum pump before the capsule is placed inside the HIP vessel. The gas pressure causes capsule wall shrinkage during HIP [35].

Fig. 1
figure 1

Schematic of capsule-free (a) and encapsulated (b) HIP

Full size image

HIP treatment of cold-sprayed titanium alloys was first reported by Blose et al. [36, 37]. Unfortunately, they did not specify which type of HIP treatment was applied (encapsulated or capsule-free). Moreover, the mechanical properties of the deposits before and after HIP were not described.

Petrovskiy et al. recently showed that capsule-free HIP can be used to improve the mechanical properties of nitrogen-sprayed pure titanium cold-sprayed deposits [38]. It was demonstrated that the UTS and elongation at break increased from ~ 80 to ~ 450 MPa and from < 1 to 7%, respectively. However, some residual porosity was found in the samples after HIP. Finally, Chen et al. presented a comprehensive study of the effects of capsule-free HIP on the properties of Ti6Al4V cold-sprayed deposits produced using nitrogen and helium as working gas [39]. They proved that capsule-free HIP increased the UTS of the samples from ~ 100 to ~ 650 MPa and from ~ 380 to 950 MPa for the samples sprayed with nitrogen and helium, respectively. However, the elongation of the HIPed Ti6Al4V was low for both types of deposits (an engineering strain of 1.5–1.75%). Chen et al. concluded that capsule-free HIP cannot increase the tensile strength of the samples sprayed with nitrogen to those of bulk material owing to poor porosity elimination. For the samples sprayed with helium, HIP had a stronger effect owing to the lower porosity of the as-fabricated samples [39].

The effect of encapsulated HIP on the structure and tensile properties of nitrogen-sprayed cold-sprayed deposits is investigated in this study.

2 Materials and methods

Gas-atomized spherical Ti6Al4V titanium alloy powder was used in the experiments. Figure 2 shows scanning electron microscopy (SEM) images of the titanium powder. The particle size distribution (+ 5–95 μm) of the powder was higher than the standard typically used for cold spraying (+ 15–45 μm). The elemental distribution across the particle cross section is shown in Fig. 3. The analysis showed that the slight variation (near 0.6% wt.) of the element distribution in the particles.

Fig. 2
figure 2

SEM images of Ti6Al4V powder used in experiments

Full size image
Fig. 3
figure 3

Elemental distribution of a Ti6Al4V particle

Full size image

The samples were produced using commercial cold spray equipment (Impact 5/8, Impact Innovations, Germany) with the OUT-1 nozzle (https://www.impact-innovations.com). The spraying gun was fixed during deposition, and the substrate was displaced by a six-axis robot (KUKA, Germany). Nitrogen was applied as the working gas and powder carrier gas for cold spraying. Preliminary tests performed at working gas stagnation pressure p0 = 5 MPa showed that the powder deposition efficiency (DE) strongly depended on the gas stagnation temperature T0 (Fig. 4). The estimated maximum powder DE obtained at T0 = 1073 K was approximately 72%, which is lower than previously reported values [33]. The relatively low DE could be explained by the wide particle size distribution.

Fig. 4
figure 4

Influence of gas stagnation temperature on the deposition efficiency of Ti6Al4V powder

Full size image

The literature specifies that the optimum particle size for cold spraying is between 5 and 50 μm. Larger particles have a lower impact velocity, resulting in lower adherence and thus lower DE [1]. Here, selective particle deposition was observed; finer particles adhered to the surface, whereas the coarser ones rebounded. The observation was further confirmed by cross-sectional SEM analysis, which revealed that only particles less than ~ 50–60 μm in size appeared in the deposit. At the same time, the observation of cross-sections of the samples obtained at 1073 K revealed significant oxidation of particle boundaries, especially in the interlayer zones (Fig. 5). In order to minimize the sample oxidation but at the same time to keep the deposition efficiency at the value higher than 50%, the final samples were sprayed at compromised gas stagnation temperature T0 = 873 K. The final deposition parameters are listed in Table 1. Deposition was performed on the grinded aluminum plates with a thickness of 10 mm.

Fig. 5
figure 5

Cross-section of the coatings deposited at T0 = 1073 K

Full size image
Table 1 Spraying parameters
Full size table

The dimensions of the final deposits were approximately 60 × 60 × 40 mm3. When spraying was complete, the deposits were separated from the substrate by electrical discharge machining. Two batches of samples were produced. The samples in the first batch underwent encapsulated HIP, whereas those in the second batch were analyzed in the as-sprayed condition.

Encapsulated HIP was performed using a Ti6Al4V capsule with a wall thickness of 1 mm. The capsule and samples were degasified before final welding. HIP was conducted in argon for 2 h at a static pressure of 110 MPa and temperature of 1173 K. When HIP was complete, the capsule was removed by electrical discharge machining.

Both batches of samples were subjected to a standard metallographic preparation procedure. The sample microstructure was analyzed using an optical microscope (Olympus BX51, Olympus, Japan) and a scanning electron microscope with different voltage 10–20 kV (Vega 3LMH, Tescan, Czechia). Elemental analysis was performed using an energy-dispersive X-ray microanalyzer (XCITE, Oxford instruments). The porosity was calculated using an image analysis method. In the image analysis, an average of 10 measurements over 5 images taken at magnifications × 20 and × 50 was calculated for as-sprayed and HIPed samples.

Tensile tests were performed in accordance with the ASTM E-8 standard. The samples were obtained by electrical discharge machining. Figure 6 shows the orientation of the tensile test samples relative to the initial specimen geometry. The strain direction during tensile testing was parallel to the nozzle pass direction. Ten tensile samples were produced: five for the as-sprayed coating and five for the coating after HIP treatment.

Fig. 6
figure 6

Orientation of the tensile test samples relative to the spraying direction (a) and photograph of the tensile test samples (b)

Full size image

3 Results

3.1 Structure

The samples were successfully produced. Surface examination did not reveal any cracks or other significant defects. Figure 7 shows SEM images of the as-sprayed deposits at different magnifications. The samples showed high porosity (~ 7.5%), as shown in the SEM images taken at low magnification (100×, 500×). The measured porosity was comparable to the values for cold-sprayed Ti6Al4V alloy obtained using nitrogen as a working gas with similar spraying parameters [27, 39]. The high porosity can generally be explained by insufficient particle acceleration and poor particle deformation upon impact. The particle boundaries are visible in the SEM images taken at higher magnification (2000×, 5000×). The total contact surface between adjacent particles was significantly decreased by pores concentrated at the particle–particle interfaces.

Fig. 7
figure 7

Cross-sectional SEM images of as-sprayed samples at different magnifications

Full size image

The samples had a dispersed (α + β) structure with fine grains, which is typical of Ti6Al4V alloy produced by rapid crystallization. The development of this structure can be attributed to rapid crystallization during powder fabrication and continuous dynamic recrystallization upon high-velocity impact during cold spraying, as previously reported [24, 27, 39]. Overall, the observations confirmed that the structure of the as-built samples was typical of Ti6Al4V deposits fabricated by cold spraying with nitrogen [24,25,26,27, 39].

The elemental distribution of the as-built deposits is presented in Fig. 8. The elemental distribution within the deposit is close to that of the particles before cold spraying. This result indicates that continuous dynamic recrystallization during particle deformation does not cause significant elemental diffusion.

Fig. 8
figure 8

Elemental distribution of as-built Ti6Al4V deposit

Full size image

Figure 9 shows the sample microstructure after encapsulated HIP. The sample porosity was decreased significantly by the encapsulated HIP treatment. Image analysis showed that the total porosity value did not exceed 0.2%. Considering the paper of Chen et al. [39], it is possible to conclude that encapsulated HIP is more efficient for densification of nitrogen-sprayed Ti6Al4V deposits than is capsule-free HIP.

Fig. 9
figure 9

Cross-sectional SEM images of HIP-treated samples at different magnifications

Full size image

This difference can be explained by the high percentage of open pores in the as-built deposits. Capsule-free HIP eliminates isolated pores, but the open pores that have channels in contact with the surface remain within the deposit. During encapsulated HIP, gas is removed from the open pores during degasification, and the high pressure applied by the walls of the deformable capsule causes the voids to shrink. The HIP-treated samples revealed the standard dispersed (α + β) microstructure typical of bulk Ti6Al4V after heat treatment under temperature conditions similar to those applied during HIP treatment. The particle boundaries are no longer visible in the SEM images.

Figures 10 and 11 show the chemical element distribution at different areas of the sample cross-section after encapsulated HIP. The high temperature and pressure during HIP contributed to a diffusion process that significantly changed the elemental distribution compared to that of the as-built samples. In particular, a vanadium-rich area localized at the grain boundaries, which corresponds to the area of β-phase formation during HIP, appeared.

Fig. 10
figure 10

Results of EDS analysis of the samples after HIP

Full size image
Fig. 11
figure 11

Elemental distribution of Ti6Al4V after encapsulated HIP

Full size image

Electron backscatter diffraction (EBSD) analysis of the HIP-treated samples (Fig. 12) also revealed the (α + β) structure typical for Ti6Al4V alloy. The average grain size was 36.5 μm. In the Fig. 12b, the blue color indicates the fully crystallized grains, yellow corresponds to the polygonized grains, and red corresponds to deformed ones. Taking into account the EBSD results, one can conclude that the microstructure of the HIPed samples is almost recrystallized. No preferential deformation texture was found on the pole figures and inverse pole figures (Fig. 12d, e). Unfortunately, EBSD analysis of the as-built samples could not be performed owing to the very high porosity. However, the EBSD and SEM analyses of the HIP-treated sample were compared to the SEM analysis of the as-built sample, and the results indicated that HIP induced material recrystallization that resulted in significant grain growth.

Fig. 12
figure 12

EBSD image of the Ti6Al4V deposit after encapsulated HIP: a grain structure, b recrystallization structure, c IPF coloring structure, d pole figure, e inverse pole figure

Full size image

3.2 Tensile strength

The stress–strain curves and values of the tensile strength and strain at failure for the as-built and HIP-treated deposits are given in Fig. 13 and Table 2, respectively. The as-built samples exhibited brittle failure with zero plastic deformation. The average UTS and strain at failure of the as-sprayed samples were 67 MPa and 1%, respectively, which were much lower than the values reported for Ti6Al4V alloy (900–1100 MPa and 10–20%, respectively). The significant standard deviations (~ 35 MPa) were caused by the nonuniform distribution of defects within the tensile test samples. The samples were cut from different areas of the deposits with different number of pores. Consequently, the measured tensile strength varied dramatically.

Fig. 13
figure 13

Stress–strain curves of as-built and HIP-treated cold-sprayed Ti6Al4V deposits

Full size image
Table 2 Tensile strength of as-built and HIP-treated samples
Full size table

The encapsulated HIP treatment completely changed the tensile behavior of the cold-sprayed Ti6Al4V deposit. Both the UTS and elongation at failure increased significantly. The average tensile strength was 960 MPa, which is comparable to the reference values of the bulk materials. The standard deviation was almost the same as that of the as-built material (~ 40 MPa). The elongation at break of the HIP-treated samples also increased, from less than 1% to 13.5%. The significant enhancement of the tensile properties could be explained by the porosity decrease combined with elemental diffusion and material recrystallization induced by the high temperature and pressure during encapsulated HIP.

Figure 14 shows SEM images the fracture surfaces of the as-built and HIP-treated samples after tensile testing. The main failure mechanism of the as-built samples was failure at the particle–particle interface. The craters formed by particles pulled from the fracture surface are shown in Fig. 14a. The high porosity of the sample is also visible on the fracture surface. The brittle behavior was explained by the low cohesion strength of adjacent particles and low material ductility at the particle boundaries resulting from the fine grain size. A similar conclusion was reported in [27, 39].

Fig. 14
figure 14

SEM images of fracture surfaces of the as-built (a) and HIP-treated (b) tensile test samples

Full size image

The fracture surface of the tensile samples after the encapsulated HIP treatment is shown in Fig. 14b. In contrast to that of the as-built samples, the fracture surface of the HIP-treated samples exhibits a significantly more uniform surface morphology owing to obvious plastic deformation and elongation before final failure. No particle boundaries or craters are detected on the fracture surface; however, dimples are clearly visible on the fracture planes.

4 Conclusions

The high porosity of nitrogen-sprayed Ti6Al4V deposits is an important drawback limiting the application of cold spraying as an additive manufacturing technology. The high porosity could be explained by insufficient particle impact velocity and poor particle deformation. As a result, the deposits showed low ductility and UTS in comparison with the bulk material. Application of encapsulated HIP enhanced the mechanical properties of cold-sprayed parts. The results of this study are summarized as follows.

  1. 1.

    The high porosity (7.5%), low UTS (less than 70 MPa), and low ductility of the as-built samples were unacceptable for typical applications of parts manufactured from Ti6Al4V alloy (e.g., in the aerospace industry).

  2. 2.

    Application of encapsulated HIP for post-treatment of the cold-sprayed Ti6Al4V deposit manufactured using nitrogen as a working gas decreased the sample porosity to 0.2% and improved the tensile strength and ductility to values close to the reference ones (~ 950 MPa, with an elongation of 13.5%).

  3. 3.

    The mechanical property enhancement was explained by material recrystallization and pore shrinkage during HIP due to high temperature and high pressure, respectively.

  4. 4.

    Encapsulated HIP seems to be more efficient for porosity elimination in nitrogen-sprayed Ti6Al4V cold-sprayed deposits than capsule-free HIP.