Introduction

Over several years, composite coatings have been very popular among scientists due to the search for innovative layers with enhanced properties and a wide range of coating possibilities available for production. Many works have revealed that the coexistence of two or more phases forming the cold-sprayed composite coating provides much better properties and structural features than single-phase coatings (Ref 1,2,3,4,5,6,7,8). To ensure the desired high mechanical and tribological properties, the deposits should provide excellent adhesion to the substrate and high cohesion.

One promising, innovative method for coating deposition is cold spraying which has a number of advantages such as low porosity, no change in phase composition after spraying and small residual stresses in absolute value (Ref 9). While the literature on cold-sprayed Cr3C2-NiCr coatings is limited (Ref 1, 3,4,5, 10, 11), Singh et al. (Ref 3) and Trelka et al. (Ref 5) showed that the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings obtained in this process are characterized by high density, high hardness and high abrasion resistance. Cold-sprayed Cr3C2-25(Ni20Cr) composite coatings have so far been investigated for microstructure and mechanical properties, including hardness and abrasion resistance. Trelka et al. (Ref 10) showed that these coatings formed during the cold spray process were characterized by low porosity (2 vol.%) and hardness up to 635 HV0.3. It was connected with the formation of the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings by plastic deformation of the metallic phase and embedment of the ceramic particles into the substrate and the previously deposited coating layers. Subsequent coating layers harden those deposited earlier as the outcome of the peening effect. A similar relationship was observed in the work of Trelka et al. (Ref 5), where coatings (Cr3C2-25(Ni20Cr))-5(Ni25C) sprayed on two different substrates: Al 7075 alloy and 1H18N9T steel were tested. The hardness increased with the distance from the deposit surface. The highest was in the zone near the coating-substrate interface, regardless of the substrate material. This was related to the severe deformation strengthening of the impact of Ni20Cr particles in these areas. A similar relationship was observed by Grujicic et al. (Ref 12), who studied the bonding mechanisms of cold-sprayed Cu coatings on the Al substrate.

For the broad application of these coating-substrate systems, it is necessary to investigate how they behave when an external load is applied, which may be reflected in three-point bending tests. This will enable identification of potential causes of damage to the coating-substrate system, e.g., poor cohesion of the particles forming the coating or poor adhesion of the deposit to the substrate. Gyansah et al. (Ref 13) studied the behavior of the cold-sprayed SiC/Al cermet coatings during three-point bending tests and observed zigzag cracks occurring perpendicularly to the applied force and crack branching mainly near the ceramic particles. In turn, Mayrhofer et al. (Ref 14) investigated the fracture toughness of Cr3C2-25(Ni20Cr) coatings produced by the HVOF spraying method. They observed intense cracking and crushing of ceramic particles during the three-point bending process. As presented by Gui et al. (Ref 15) and Koutsomichalis et al. (Ref 16), who investigated WC-Co-Cr composite coatings HVOF sprayed onto a steel substrate, after the three-point bending test, cracks were observed along the deposit because of the substrate deformation. During bending, the number of surface cracks per unit length decreased with increasing coating thickness. The deposit failed and fell off the substrate when the critical stress in crack propagation reached the damaged spots at the coating-substrate interface, which indicates that its cohesion was higher than adhesion. Hahn et al. (Ref 17) conducted four-point bending tests under cyclic mechanical loads on FeC0.8 and FeCrCB coatings produced by TWAS, HVOF and PTWA. They revealed many small fluctuations that did not cause destructive cracks in the deposits. In turn, a study performed by Bansal et al. (Ref 18) showed that the specific cracks of the Al2O3-13wt.%TiO2 ceramic coating produced by plasma spraying are influenced by the residual stresses occurring in the deposit after the spraying process. In this case, they were in the range of 120 MPa. Cold-sprayed coatings are usually characterized by a low state of residual stress (virtually zero), which results from the specificity of the process as has been demonstrated by Góral et al. (Ref 19).

Trelka et al. (Ref 5) showed that the (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings cold-sprayed on the steel substrate revealed higher adhesion (40.6 MPa) than those deposited on the Al 7075 alloy (34.7 MPa). In the literature, there are comprehensive results of adhesion tests of composite coatings made from the Cr3C2-25(Ni20Cr) powder but sprayed with a different method (Ref 20) or deposits of Sn + Al2O3, Zn + Al2O3, Al + Al2O3, Ni + Al2O3 or Cu + Al2O3 deposited using the LPCS method tested by Winnicki et al. (Ref 21). Thao et al. (Ref 20) showed that the plasma-sprayed Cr3C2-25(Ni20Cr) coatings deposited on a 16Mn steel substrate had an adhesion of 16.32 MPa, while the deposits tested by Winnicki et al. (Ref 21) had adhesion strength in the range from 6 to 55 MPa.

Additionally, by creating composite coatings containing a solid lubricant (for example, in the form of graphite, MoS2, etc.), it is possible to improve the operation properties of kinematic pairs in terms of lowering friction and increased durability, as was shown by Trelka et al. (Ref 5), Żórawski et al., (Ref 8), Góral et al. (Ref 19), Deng et al. (Ref 22), Cai et al. (Ref 23), Yi et al. (Ref 24), Song et al. (Ref 25). Coatings with the addition of solid lubricants are particularly desirable when the high reliability of tribological couplings is required, they are difficult or expensive to access, and in devices where the use of liquid lubricants is inadequate. Therefore, knowledge of how the deposits work during fatigue and wear conditions is essential.

In the light of the presented study, it seems justified to investigate the behavior of the systems composed of Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings cold sprayed on Al 7075 substrate under the influence of an applied external load. The mechanical and tribological properties of graphite have been the subject of many studies (Ref 5, 8, 19, 22, 23, 26, 27). Still, it is invaluable to determine how a solid lubricant such as graphite embedded in the cermet coating structure affects its mechanical properties and crack propagation. Therefore the subject of the paper is to investigate how the graphite content in the Cr3C2-25(Ni20Cr) cermet coatings affects their mechanical properties by understanding the cracking mechanism of the coating-substrate systems under external load and identification of their potential causes of damage, e.g., poor cohesion or adhesion.

Materials and Methods

The coatings were deposited on Al 7075 alloy substrates (flat bars with dimensions of 400 mm × 30 mm × 5 mm). The Cr3C2-25(Ni20Cr) deposits were cold sprayed with Cr3C2-25(Ni20Cr) Diamalloy 3004 powder (Oerlikon Metco Europe GmbH, the Polish Division, Poznań, Poland), which is a mixture of Cr3C2 and Ni20Cr (Fig. 1a). The (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings were deposited using 95 wt.% of Diamalloy 3004 and 5 wt.% of Ni25C Durabrade 2221 powder (Fig. 1b) (Oerlikon Metco Europe GmbH, the Polish Division, Poznań, Poland), where the graphite was coated with 4 to 8 μm Ni film. Graphite nickel plating is necessary in order to obtain the appropriate mass of the powder, which will contribute to the achievement of the critical velocity during spraying and thus allow the deposition of the solid lubricant in the coating. Cr3C2‐25(Ni20Cr) powder had grain size from 9.5 to 55.4 μm (d10 = 9.5 μm, d50 = 24.8 μm, d90 = 55.4 μm), while Ni25C powder—from 46.8 μm to 137.9 μm (d10 = 46.8 μm, d50 = 81.2 μm, d90 = 137.9 μm) (Ref 19). The cold spraying process was performed with an Impact Innovations 5/8 system (Impact Innovations GmbH, Rattenkirchen, Germany) mounted on a Fanuc M-20iA robotic arm (ZAP Robotyka, Ostrów Wielkopolski, Poland). The cold-sprayed parameters are presented in Table 1. The process parameters were selected based on the Taguchi method (orthogonal array, L9(34)) testing the following parameters: amount of Ni25C (5, 10, 15 wt.%), gas composition (N2, N2 + He, He), spraying distance (20, 30, 40 mm) and traverse speed of the gun (200, 300, 400 mm·s−1) on the deposit thickness, Cr3C2 and graphite content, hardness, Young’s modulus and friction coefficient.

Fig. 1
figure 1

The SEM cross-sections of the powders: Cr3C2-25(Ni20Cr) (a) and Ni25C (b)

Full size image
Table 1 Parameters of the cold spray process with the use of Impact Innovations 5/8 system
Full size table

The volume fractions of graphite, porosity and Cr3C2 in the coatings were measured using the binary segmentation method in ImageJ software based on the coatings cross-section microstructures (10 images). The phase analysis of the coatings was performed using a Bruker D8 Discover diffractometer with CoKα radiation (wavelength of 1.7903 Å) (Bruker AXS GmbH, Karlsruhe, Germany), Diffrac. EVA and HighScore Plus software and PDF-4 + database (Powder Diffraction File). The microstructure and chemical composition of the powders and coatings were characterized by scanning electron microscopy (FEI/Philips XL30, FEI Company, Oregon, USA) in the backscattered electron (BSE) mode to illustrate the phases occurring in the coatings. The hardness test (HV0.5) was carried out on a CSM Instruments SA device (Anton Paar GmFbH, Graz, Austria) under a static load of 4.903 N. An average from ten measurements was given as the result.

The three-point bending tests were performed under constant velocity and a cyclic load for three samples of each system to show the influence of graphite on the coating-substrate adhesion and susceptibility to cracking. To distinguish the tests, the one performed with a constant load was called three-point bending test and the other with a cyclic one-three-point bending test under a cyclic load. The samples for both three-point bending tests had the same dimensions. The specimens for the three-point bending tests were cut from the coating-substrate system and ground on abrasive papers and polished on diamond suspensions with a finishing gradation of 1 µm. The dimensions of the samples were 1.2 mm × 3 mm × 24 mm, where 0.6 mm was the deposit and 0.6 mm was the substrate. The samples were prepared so that the thickness of the coating and substrate was the same. In this way, their influence on the obtained results was eliminated. Samples were measured using a micrometer screw and an electric calliper. The scheme of the sample and the three-point bending test is shown in Fig. 2a. The three-point bending test was performed using an INSTRON 6025 (Instron, Massachusetts, USA) modernized by Zwick/Roell with a computer-controlled mandrel traverse speed equipped with a system for a three-point bending test. The test was performed at room temperature. The former (defined as a counter-sample in bending tests according to the ISO 7438:2016(E) standard) moving into the substrate-coating system with a constant speed of 0.001 mm·s−1 increased the applied force until the coating cracked. The investigations were carried out for three samples of each coating-substrate system. The three-point bending test under a cyclic load was performed at 25 °C and 200 °C under the load of 1 N with an amplitude of changes of 0.5 N and frequency of 0.5 Hz using a 402 F1 Hyperion Thermomechanical Analyser (Netzsch, Selb, Germany). The number of cycles was 5000. The scheme of the force distribution versus time is presented in Fig. 2b. The experiment consisted in subjecting the tested coating-substrate systems to stresses and strains. Three-point bending tests conducted under such fluctuating conditions can cause cracks at the interface or near the coating-substrate junction. The tests were carried out in two stages to check which compressive or tensile stresses dominate the material. Tests were investigated on a coating-substrate system, where the force was applied to the substrate side, then on the specimen with the force applied to the coating side.

Fig. 2
figure 2

The scheme of sample for the three-point bending tests (dimensions in mm) (a) and the distribution of force during the test under constant velocity (a) and a cyclic load (b)

Full size image

The adhesion of the coatings to the substrates was measured using a PosiTest AT-A device (DeFelsko, NY, USA) with an electronically controlled hydraulic pump and the pull-off method, according to the ISO 4624: 2016 standard (Ref 28). An adhesive FM 1000 pad (Solvay, Brussels, Belgium) dedicated to this test was placed on the examined coating surface and then pressed by the aluminum dolly with a 0.3 MPa (Fig. 3). The samples prepared in this way were heated at 180 °C for 1 h and cooled in a furnace. Then, the deposit around the adhesive dolly was separated by milling around to a depth of 2 mm. The PosiTest AT-A smoothly increased the pull-off force until the maximum value, occurring just before the sample detachment from the substrate, was recorded. The pull-off tests were performed for three coating-substrate systems of each type, and an average value from these measurements was the result.

Fig. 3
figure 3

The aluminum dolly pressed to the coating surface (a), adhesive FM 1000 pad (b) and prepared sample with the separated area outside the aluminum dolly (c)

Full size image

Tests of the wear index and friction coefficient of three deposits of each type were carried out with using a ball-on-disc T-21 tribotester (The Institute for Sustainable Technologies, Radom, Poland). The ball was a 6 mm-diameter sintered Si3N4 sphere, its linear sliding speed was 0.1 m·s−1, the radii were 5, 7 and 8.5 mm, and the number of cycles was 20,000. The results presented in the paper are the average value from measurements performed for three radii. The surface of the coating was ground and polished on diamond polishing suspensions with a finishing gradation of 1 μm. The investigations were performed under 5 N load and at two temperatures of 25 °C and 250 °C. The chamber was heated with the samples and the temperature was maintained for one hour using a thermocouple. The friction and wear characteristics were tested in accordance with the ISO 20808:2016(E) standard (Ref 29). The friction coefficient average CoF was calculated based on registered friction force values collected during the test, according to formula (1).

$$\mathrm{CoF}=\frac{{F}_{\mathrm{Tav}}}{F}\left[-\right]$$
(1)

where CoF—friction coefficient average, FTav—force friction averaged measured during the test, F—sample loading force.

The wear index was determined according to formula (2):

$$W_{{\text{v}}} = \frac{V}{{F_{{\text{n}}} \cdot s}}\left[ {\frac{{{\text{mm}}^{3} }}{{{\text{N}} \cdot {\text{m}}}}} \right]$$
(2)

where V—volume of worn material [mm3], Fn—normal force exerting pressure on the sample [N], s—wear path [m].

Results and Discussion

Characterization of the Coatings

The X-ray diffraction patterns for the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings are presented in Fig. 4. The (Cr3C2-25(Ni20Cr))-5(Ni25C) coating apart from the Cr3C2 and Ni-Cr (Cr0.25Ni0.75, PDF-4+ 04-003-7001) phases revealed in the deposits without a solid lubricant, also contains graphite and nickel phases (Ref 19). The presence of the nickel and graphite phases in the (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings is related to the phase composition of the used powder, e.g., Cr3C2, Ni20Cr (Cr0.25Ni0.75), Ni, C (graphite). The cross-sections of the powders are shown in Fig. 1. This is related to the cold spray process, one of those advantages is no change in phase composition in the resulting coatings.

Fig. 4
figure 4

X-ray diffraction patterns for the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings

Full size image

Figure 5 shows the cross-section microstructures of the coatings. The Cr3C2-25(Ni20Cr) deposit microstructure was comprised of Cr3C2 ceramic particles (thin arrows, gray) distributed uniformly in the Ni20Cr matrix (light gray) (Fig. 5a). The plastically deformed metallic Ni20Cr particles were in the form of strips tightly adjacent to each other. The large Cr3C2 particles fragmented during spraying. The dispersed ceramic particles were distributed approximately evenly throughout the coating. The mechanism for forming the composite Cr3C2-25(Ni20Cr) coatings in the cold spray process has been described by Wolfe et al. (Ref 1), Poza et al. (Ref 2) Fernandez et al. (Ref 4), Trelka et al. (Ref 5), Góral et al. (Ref 11), Grujicic et al. (Ref 12) and Sevillano et al. (Ref 30). The mechanism of formation of cold-sprayed deposit was purely mechanical (plastic deformation of metallic particles and blocking hard ceramic particles). The Cr3C2-25(Ni20Cr) coatings comprised the Ni20Cr matrix and Cr3C2 ceramics in the structure, while the (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings additionally had graphite and Ni phase. Hence, various bonding mechanisms in the deposits were observed. Similar to the report by Góral et al. (Ref 11) and Trelka et al. (Ref 5), who showed a detailed description of the mechanism of the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings formation, the metallic matrix was in the form of strongly deformed and elongated well-bonded grains of Ni20Cr phase. The phenomenon responsible for this bonding was adiabatic shear instability and the associated localization of the plastic flow at the interfacial zone. The plasticization occurred only in the outer part of metallic particles as a result of heat released by the strong impact of Ni20Cr particles on the substrate or previously deposited particles. Additionally, adiabatic shear combined with high pressures favored the formation of tight contact surfaces, which allowed for achieving good adhesion of the coatings due to the “surface scrubbing” jets providing clean contact surfaces (Ref 11, 12). It is also responsible for coating cohesion and its low pore content. The Cr3C2-25(Ni20Cr) coatings had a porosity value of 1.7 ± 0.4 vol.%, which was mainly an artifact due to the falling off of ceramic particles during sample preparation. In the (Cr3C2-25(Ni20Cr))-5(Ni25C) deposits containing solid lubricant, no pores were observed. Moreover, the (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings comprised 1.6 ± 0.4 vol.% of graphite which is a similar value to the porosity of coatings without graphite (Table 2). Furthermore, the graphite (thick arrows, black) rounded by the Ni shell (bright areas near the graphite) was co-deposited between strongly elongated Ni20Cr splats and near the crushed ceramic particles, Fig. 5b. The addition of graphite caused a reduction in deposition rate (from 23.1 and 21.3 g·min−1), which was probably related to the removal of the deposited graphite from the surface by much harder Cr3C2 and Ni20Cr particles. The deposition efficiency for Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings was 24.3% and 22.4%, respectively. It follows that 75.7% of powder without and 77.6% of this with the graphite were reflected from the substrate. Ceramic particles cannot plastically deform; therefore, they were embedded in the metallic phase and bonded to the matrix by blocking. They also formed micro-asperities that favored the bonding of incoming alloy particles and improved the contact surface of the coating with the substrate, Fig. 5. The ceramic particle’s impact on the metal particle, and its rebounding from the already applied layer, especially co-deposited ceramic, hardened the substrate and the metal matrix. The content of Cr3C2 in the deposits was nearly halved compared to the feedstock powder, which suggested the coexistence of embedment and rebounding mechanisms during the coatings’ formation. Both Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings showed the same amounts of ceramic particles 31.8 ± 1.8 vol.% and 31.5 ± 1.9 vol.%, respectively. This agreed with the results regarding the Ni-Al2O3 cermet coatings shown by Sevillano et al. (Ref 30). Another mechanism, which played a crucial role in deposit creation, was the mechanical interlocking involving the mechanical trapping of particles by the substrate or previously deposited layer, resulting in a mutual geometric interlocking (Ref 2). This was observed especially near the substrate zone in the form of Ni20Cr particle-scale rivet-like interlocks.

Fig. 5
figure 5

SEM-BSE cross-section microstructures of the Cr3C2-25(Ni20Cr) (a) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coating (b); Cr3C2—thin arrows, graphite-thick arrows

Full size image
Table 2 Porosity, graphite and Cr3C2 content in cold-sprayed coatings
Full size table

In Fig. 6 the microstructure of the coating (Cr3C2-25(Ni20Cr))-5(Ni25C) with the map of the chemical composition of occurring elements: C (graphite), Ni and Cr, is presented. As shown in Fig. 6, the gray precipitates are Cr3C2 carbides, the black region corresponds to graphite, and all around is a Ni0.75Cr0.25 matrix. Moreover, close to the solid lubricant Ni regions are visible, which results from the nature of the used cold-sprayed particles.

Fig. 6
figure 6

SEM-BSE cross-section microstructure of the (Cr3C2-25(Ni20Cr))-5(Ni25C) coating and maps of Cr, Ni and C distribution

Full size image

Three-Point Bending Test

The three-point bending test of the substrate-coating systems caused a crack to appear in the coating in the place where more and more stress was generated by the former moving into the sample. With increasing force a crack propagated across with the deposit perpendicular to the surface until it reached the coating-substrate interface. Then it started running along the interface in a direction perpendicular to the original one. The crack spread until the system became unstable because of the failure of the entire deposit, as evidenced by the sudden collapse of the curve in Fig. 7. The cermet coating was completely damaged. The aluminum alloy substrate remained undamaged. The mechanism of the crack formation during three-point bending tests was the same in both cermet deposits. Nickel provided additional ductility, and therefore the flexural strength of the (Cr3C2-25(Ni20Cr)-5(Ni25C) coating was higher, Fig. 7. The cracks mainly propagated near the chromium carbides. Extensive cracking of Cr3C2 ceramic particles in the HVOF Cr3C2-25(Ni20Cr) coatings during three-point bending tests was also observed by Mayhofer et al. (Ref 14). Bansal et al. (Ref 18) showed a similar cracking mechanism in ceramic composite Al2O3-13wt.%TiO2 coatings plasma-sprayed on mild steel resulting from three-point bending, i.e., a single vertical crack propagating from the coating surface toward the deposit-substrate interface, then extending horizontally along with the deposit. The fracture at the interface evenly separated the coating from the substrate, with no residue of embedded coating particles in the substrate (Ref 18). The examined coating-substrate systems revealed a significant difference in the force value causing destabilization of the system and detachment of the coating from the substrate. The set with the (Cr3C2-25(Ni20Cr))-5(Ni25C) deposit was damaged at a stress level that was 53% greater than that for the Cr3C2-25(Ni20Cr), i.e., 2.3 ± 0.04 MPa and 1.5 ± 0.02 MPa, respectively, Fig. 7. Moreover, the stability loss of the Cr3C2-25(Ni20Cr) system occurred when the fracture reached the interface, where a small horizontal fracture was observed, Fig. 8a. In the case of the coating containing graphite, the horizontal crack was larger and ran along the interface, Fig. 8d. The results showed that the graphite built in the deposit structure affected the increase in the bending force of the cermet coating-substrate system and enhanced the resistance to crack propagation. The graphite-containing coating was less brittle than the coating without the solid lubricant and therefore possessed higher fracture toughness. In contrast to the composite coatings tested by Koutsomichalis et al. (Ref 16) and Gui et al. (Ref 15), the deposits did not fall off the substrate during the three-point bending test, which proves the higher adhesion of the coatings tested by us. Koutsomichalis et al. (Ref 16) noted that when the critical stress for crack propagation was reached in defect sites at the interface, the entire coating fell and peeled off the substrate material. If the crack propagation is limited to a localized region, the fracture of the coating follows. They showed that if the energy causing propagation of the crack dissipates, the fracture cannot extend beyond the substrate material—it terminates at the deposit (Ref 16). The force generated during our tests caused stresses in the sample, which caused cracks in the coatings perpendicular to their surface. These cracks occurred only in the coatings because the Al 7075 substrate had much greater ductility in comparison with cold-sprayed coatings. The resulting stresses caused further development of cracks at the interface, i.e., along the substrate-coating boundary, which was the weakest element of this system. In turn, Gyansah et al. (Ref 13) showed the presence of a zigzag crack path in cold-sprayed SiC/Al deposits after three-point bending tests. Their coating cracking occurred mainly in the vicinity of the ceramic particles. The SiC particles acted as a stress concentration point and strengthening mechanism for crack initiations and propagations, respectively.

Fig. 7
figure 7

Strength versus displacement of the pin during the three-point bending test of the Cr3C2-25(Ni20Cr)—Al 7075 alloy and (Cr3C2-25(Ni20Cr))-5(Ni25C)—Al 7075 alloy systems

Full size image
Fig. 8
figure 8

SEM-BSE cross-section microstructures of the systems with the Cr3C2-25(Ni20Cr) (a-c) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coating (d-f) after the three-point bending test; the applied force to the substrate side

Full size image

The Three-Point Bending Test Under a Cyclic Load

The three-point bending tests of the coating-substrate systems were performed to investigate the mechanisms of interfacial fracture, as well as to determine the effect of low cyclic loads at 25 °C and 200 °C on the adhesion of the coating to the substrate materials. The distribution of forces and stresses in the samples during the tests had a direct impact on the above-mentioned damage. The obtained results of the storage modulus (E′) and damping coefficient (tgδ) provided a better understanding of the actual changes at the interface. The changes in these parameters were treated mainly as information about the crack appearance. Sample test conditions are shown in Fig. 9. The applied force was marked in black, while the behavior of the coating-substrate system after the applied force was marked in red.

Fig. 9
figure 9

Example of the conditions of cyclic three-point bending tests

Full size image

The first three-point cyclic bending test was carried out at 25 °C on the Cr3C2-25(Ni20Cr)—Al 7075 system, where the force was applied from the substrate side, the same as during the three-point bending tests with constant velocity. The obtained results showed no changes in both the storage modulus (E') and damping coefficient (tgδ) (Fig. 10a). Additionally, no cracks in the microstructure were observed (Fig. 10b). The force applied from the substrate side meant the stresses generated in the material were suppressed by the Al 7075 alloy and system damage effects were not observed. Therefore subsequent tests were carried out on systems where the force was applied from the coating side. The tests were carried out on the Cr3C2-25(Ni20Cr)-Al 7075 alloy and (Cr3C2-25(Ni20Cr))-5(Ni25C)-Al 7075 alloy systems at 25 °C (Fig. 11 and 12).

Fig. 10
figure 10

The obtained storage modulus (E′) and damping coefficient (tgδ) (a), and microstructure (b) of the Cr3C2-25(Ni20Cr)—Al 7075 system at 25 °C; the applied force to the substrate side

Full size image
Fig. 11
figure 11

The obtained storage modulus (E′) and damping coefficient (tgδ) for the Cr3C2-25(Ni20Cr)—Al 7075 alloy (a) and (Cr3C2-25(Ni20Cr))-5(Ni25C)—Al 7075 alloy (b) systems at 25 °C; the applied force to the coating side

Full size image
Fig. 12
figure 12

The SEM microstructure of the cross-section of Cr3C2-25(Ni20Cr) on Al 7075 alloy (a) and (Cr3C2-25(Ni20Cr))-5(Ni25C) on Al 7075 alloy (b) after three-point cyclic bending tests at 25 °C; the applied force to the coating side

Full size image

The recorded curves of the storage modulus and damping coefficient for the examined systems did not reveal significant changes corresponding to the appearance of the cracking observed at the interfaces. In Fig. 11, the damping coefficient for both samples is on the same level. However, the storage modulus for sample Cr3C2-25(Ni20Cr)-Al 7075 alloy is higher than that (Cr3C2-25(Ni20Cr))-5(Ni25C)-Al 7075, and in its microstructure, the crack at the interface is observed (Fig. 12a). Due to the fact that the analyzed (Cr3C2-25(Ni20Cr))-5(Ni25C)-Al 7075 substrate system did not reveal significant changes in its microstructure, the experiment was performed at the elevated temperature—of 200 °C. The choice of this temperature was dictated by to desire to predict the system behavior in the working conditions, not causing the significant destruction of their structure. The determined storage modulus and tgδ of the systems are shown in Fig. 13. The value of the E’ for both samples was slightly lower than for systems examined at 25 °C. The observed in Fig. 13a damping coefficient for samples without graphite was twice as high; however, its values for both systems were at a very low level, and therefore they were considered comparable. In the case of the Cr3C2-25-(Ni20Cr) system, three large fractures (up to 400 µm) at the interface in the middle part of the sample were observed. One of them is shown in Fig. 14a. In the system with a deposit containing graphite, one crack of about 300 µm appeared at the interface, as shown in Fig. 14b. The similar character of the storage modulus and damping coefficient at two temperatures is connected to the fact that up to 200 °C in the examined systems (both in the cermet coatings and in Al 7075 alloy) no phase transformation occurred (Ref 31, 32). In addition, a significantly limited propagation of cracks in the coatings with the solid lubricant coated by nickel was observed.

Fig. 13
figure 13

The obtained storage modulus (E′) and damping coefficient (tgδ) for the Cr3C2-25(Ni20Cr)—Al 7075 (a) and (Cr3C2-25(Ni20Cr))-5(Ni25C)—Al 7075 (b) systems at 200 °C; the applied force to the coating side

Full size image
Fig. 14
figure 14

The SEM microstructure of the cross-section of Cr3C2-25(Ni20Cr) on Al alloy (a) and (Cr3C2-25(Ni20Cr))-5(Ni25C) on Al alloy (b) after three-point cyclic bending tests at 200 °C; the applied force to the coating side

Full size image

During the three-point bending test under a cyclic load, the deformation developed on the microscale level and grew for each cycle until it reached a critical length and caused deposit-substrate bond damage. The cracks after the conducted experiments were only in the coating-substrate interface area (Fig. 12 and 14). This was correlated with occurred stresses, which were higher in the stiffer coating compared with the ductile Al 7075 alloy. The results indicated that the Cr3C2-25(Ni20Cr)—Al 7075 system has degraded more. The Cr3C2-25(Ni20Cr) coating with the addition of graphite caused a reduction of cracks at the deposit-substrate interface. Moreover, the experiments showed that no cracks occurred with low stress applied in the substrate, which is promising to take into account industrial application. Hahn et al. (Ref 17) also examined similar systems using four-point cyclic bending tests when the force was applied to both the coating and the substrate side. Tests under cyclic mechanical stresses were performed on different coatings deposited on the Al 6060 alloy. FeC0.8 deposits were produced by the TWAS and HVOF spraying methods, and the FeCrCB coating was produced using the PTWA method. Their tests showed a large number of small fluctuations, suggesting the presence of a large number of small cracks, regardless of the position of the test specimens. Those authors (Ref 17) observed that the tested coatings appeared to be capable of a slight deformation and tolerated a certain amount of crack initiation and propagation without destructive cracking and failure of the whole system. In the analyzed systems, the cracks were initiated and propagated at the interface. However, the coating-substrate system, despite the cracks at the interface, was able to withstand further bending, which indicated the system's stability.

Hardness of the Cermet Coatings

The Cr3C2-25(Ni20Cr) coating showed a higher hardness (579 ± 25 HV0.5) than the deposit containing the graphite embedded in the structure (549 ± 39 HV0.5), Fig. 15. The incorporation of the solid lubricant lowers the hardness by ~ 30 HV0.5 compared to the Cr3C2-25(Ni20Cr) composite coatings. It also increases the plasticity of the coatings, as evidenced by the results obtained during the three-point cyclic bending test. The hardness of the coatings was influenced both by the content of the solid lubricant coated by nickel, which was present in a greater amount by mass than graphite, and the distribution of the ceramic phase particles in the metallic matrix, as well as strain hardening of metallic Ni20Cr particles during cold spraying. Composite coatings were characterized by homogeneity due to the uniform distribution of the ceramic phase in the metal matrix, while, on the other hand, the use of a powder mixture of different materials caused local differences in the results of hardness values.

Fig. 15
figure 15

The hardness of the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings

Full size image

The microhardness of cold-sprayed Cr3C2-25(Ni20Cr) coatings has been the subject of many studies (Ref 1, 3, 10, 11), where investigations of the influence of the powder morphology differences, the different methods of their production and the various parameters of the cold spray process were examined. The coatings studied by Wolfe et al. (Ref 1) were characterized by lower hardness—from 277 to 575 HV0.3 compared to the deposits tested in this work. It may be related to the different morphology of the TAFA 1375 V powder used in the cited work, compared with the Diamalloy 3004 powder. Singh et al. (Ref 3) tested Cr3C2-NiCr coatings made of the same type of powder as presented here but with the use of other parameters of the spraying process, e.g., different gas mixture, temperature and spray distance. They obtained a higher hardness of the coatings—875.67 HV0.2, which may suggest an influence of the process conditions. The microhardness HV0.3 of Cr3C2-25(Ni20Cr) coatings produced with the use of Diamalloy 3004 powder carried out by Góral et al. (Ref 11), and Trelka et al. (Ref 10) was about 10% higher than shown in this work. However, the parameters of cold spraying were also different. In turn, Żórawski et al. (Ref 8) measured the HV0.3 microhardness of cold-sprayed Cr3C2-Ni20Cr/graphite coatings depending on the various process parameters, i.e., gas composition, graphite content, spraying distance and traverse speed of the gun. They showed that the hardness of the coatings ranged from 334.4 to 492.2 HV0.3, depending on the combination of parameters used. The obtained values were lower than for the deposits tested in this work.

Ball-on-Disc Test of Coatings

To determine the wear index and friction coefficient, the ball-on-disc test was performed. Table 3 shows the tribological properties of the deposits measured at two different temperatures: 25 and 250 °C, and under 5 N load. The tests showed a similar wear index of both coatings. Its slightly higher values (8% at 25 °C and 7% at 250 °C) for the deposits with the graphite were in the range of errors and smaller than the typical accuracy of wear test results. The same effect was observed by Hanyou et al. (Ref 33), during ball-on-disc test for the laser thermal sprayed Cr3Cr2-NiCr coatings with MoS2 as a solid lubricant. Hanyou et al. (Ref 33) observed a decrease in the wear index of coatings at room temperature by 40% after tests with a load of 4 N. That wear index was five times lower than that obtained in this work. Such difference could be caused by using other production methods and the higher solid lubricant content (10%). Wear index of coatings at 250 °C was about six times higher compared to that obtained at 25 °C. The ball-on-disc test showed that the coatings with the solid lubricant examined at increased temperature revealed a reduced friction coefficient value.

Table 3 The wear index and friction coefficient of the Cr3C2-25(Ni20Cr) and (Cr3C2‐25(Ni20Cr))‐5(Ni25C) coatings were determined under 5 N load
Full size table

Figure 16 shows the wear tracks of the coatings measured after the ball-on-disk tests. The Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings showed a similar wear index. The images of the wear tracks and the values of the wear index correspond to the wear profiles presented in Fig. 16e. The deposits tested at 250 °C had deeper and wider wear profiles than those tested at 25 °C.

Fig. 16
figure 16

The surface topography of wear tracks of the coatings tested under 5 N load; Cr3C2-25(Ni20Cr) coating, 25 °C (a), Cr3C2-25(Ni20Cr) coating, 250 °C (b), (Cr3C2-25(Ni20Cr))-5(Ni25C) coating, 25 °C (c), (Cr3C2-25(Ni20Cr))-5(Ni25C) coating, 250 °C (d), and depth vs. width of the deposit wear tracks (e)

Full size image

Adhesion of the Coatings

Coating adhesion measurements using the pull-off method were performed to determine the adhesion of the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings to the Al 7075 substrate. The pull-off tests of the coatings showed a similar adhesion for both coatings: 36 ± 5 MPa for Cr3C2-25(Ni20Cr) and 35 ± 1 MPa for (Cr3C2-25(Ni20Cr))-5(Ni25C). However, the coating without the graphite remained intact, and detachment occurred in the glue zone, while the (Cr3C2-25(Ni20Cr))-5(Ni25C) coating partially detached from the substrate, as is shown in Fig. 17. The samples after the performed tests showed a partially adhesive failure of the coating from the substrate in each of the three tests carried out. Thao et al. (Ref 20) revealed the adhesion of plasma-sprayed Cr3C2-20%NiCr composite coatings deposited on 16Mn steel to be 16.32 MPa (averaged over three results). This result was less than half the results obtained for the Cr3C2-25(Ni20Cr) and (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings cold sprayed on the Al 7075 alloy. Winnicki et al. (Ref 21) also determined the adhesion of cermet coatings deposited using the LPCS method. They deposited Sn + Al2O3, Zn + Al2O3, Al + Al2O3, Ni + Al2O3, Cu (electrolytic) + Al2O3 and Cu (spherical) + Al2O3 coatings on two different substrates: AA 1350 aluminum alloy and M1E copper alloy. The results showed that the adhesion of the coatings deposited on the AA1350 alloy ranged from 6 to 55 MPa, depending on the coating material used. After the tests were performed, various types of fractures were observed: adhesive, adhesive-cohesive, cohesive and failure in the adhesive slice. The research presented in this work showed that the Cr3C2-25(Ni20Cr) coatings were characterized by the fracturing type in the adhesive slice zone, while the (Cr3C2-25(Ni20Cr))-5(Ni25C) deposits were partially adhesive (the deposit was not completely detached from the substrate). These results suggested that the graphite presence reduced the coating adhesion, as indicated by glue failure in the test. The coating detachment from the substrate is also visible after three-point bending test, Fig. 8(d–f). However, the (Cr3C2-25(Ni20Cr))-5(Ni25C) coatings revealed higher flexural strength needed for their cracking which indicates that they showed the higher bonding force of the particles formed the cermet deposit. This recommends their use in industry as deposits working under heavy wear conditions.

Fig. 17
figure 17

The types of fractures after pull-off tests: failure in the adhesive slice of the Cr3C2-25(Ni20Cr) coating (a), partially adhesive failure of the (Cr3C2-25(Ni20Cr))-5(Ni25C) coating (b)

Full size image

Conclusions

The work describes the difference in the microstructure, mechanical and tribological properties of the systems composed of the Cr3C2-25(Ni20Cr) or (Cr3C2-25(Ni20Cr))-5(Ni25C) composite coating deposited on Al 7075 alloy. The susceptibility to cracking during three-point bending tests, adhesion, hardness and wear resistance of these cermet deposit-substrate systems was investigated. The results showed that even a small amount of the graphite surrounded by nickel embedded in the cermet coating structure significantly affects the resistance of the deposit to cracking under an applied external load. The following conclusions have been drawn from the test results.

  1. 1.

    During three-point bending, all the examined coating-substrate systems revealed cracks appearing at the coating surface and propagating through the deposit in the direction of the applied load to the coating-substrate interface, where they ran parallel to the substrate. In the (Cr3C2-25(Ni20Cr))-(Ni25C), the force required to destroy the durability of the coating-substrate system was about 53% higher than that without the solid lubricant.

  2. 2.

    The three-point bending under cyclic load showed that graphite embedded in the cermet coating prevented the formation of crack nuclei at room temperature and reduced the size of cracks in the deposit at 200 °C.

  3. 3.

    Both the cold-sprayed cermet coatings were characterized by a homogeneous and compact microstructure.

  4. 4.

    The graphite incorporation in the deposit structure resulted in a hardness value lower by 5.5% than that of the coating without it.

  5. 5.

    Coatings containing the solid lubricant showed a similar wear index as the Cr3C2-25(Ni20Cr) deposits.

  6. 6.

    The Cr3C2-25(Ni20Cr) coating—Al 7075 substrate system revealed higher adhesion (above 36 MPa) which was indicated by the breakage entirely within the glue.