To study the deposition behavior of Al–basalt fiber on Al substrate, the deformation ability of Al powder was firstly examined through numerical simulation, and the impact behavior of a single Al particle was simulated. Secondly, the position of a single basalt particle after inserting a cohesive unit was examined to reveal the breaking behavior of a single basalt fiber. Finally, based on the numerical simulation results of single-particle deposition, 15 vol.%, and 30 vol.% basalts fiber-Al multi-particle deposition behavior was numerically investigated.
4.1. Model Verification
The powder-feeding gas was considered to be ideal state N2, the inlet pressure of the nozzle was 5 MPa, the inlet temperature was set to 380 °C, the powder-feeding gas pressure was set to 5.1 MPa, the powder-feeding gas temperature was set to 25 °C, and the nozzle pressure outside the outlet was set to 5.1 MPa. The environmental pressure and temperature were set to 0.101 MPa and 25 °C, respectively. Finally, the density-based solver was used for the iterative solution process until convergence.
The nozzle cross-sectional temperature and N
2 velocity distribution programs calculated by the computational fluid dynamics (CFD) software Fluent (ANSYS 2022 R1) are shown in
Figure 5a.
Figure 5b shows the changes in the velocity and temperature of basalt fiber along the nozzle
X-axis. Because of its cylindrical structure, basalt fiber is difficult to model with the DPM method, and its unique structure in the Laval nozzle is extremely complex. For this reason, we simplify its model to a spherical DPM model with an equivalent diameter and endow it with relevant physical properties of basalt fiber, such as heat conductivity, specific heat capacity, density, and Young’s modulus. It can be seen that the temperature of the basalt fiber along the
X-axis reaches a maximum value of 589 K, and the velocity reaches a maximum value of 768 m/s when it is close to the substrate. These data are further used as the particle velocity and temperature in the subsequent basalt fiber deposition simulations.
Figure 5c shows the velocity and temperature changes along the axis of the nozzle using N
2 as the powder-feeding gas. It is clear that the velocity and temperature of N
2 in the nozzle have a certain oscillation. This can be attributed to the fact that the intrinsic design pressure of the nozzle is 5 MPa, while the expansion ratio is greater than the optimal value, and an oblique shock wave is generated at the exit of the nozzle. However, this oblique shock wave has the strongest influence on the particles with a small diameter (1–5 μm) and has a minor effect on most of the particles used in this experiment with a diameter of 10–100 μm.
Figure 5d shows the correlation between the velocity and temperature of the particles (diameter: 10–100 μm) exiting the nozzle outlet and the particle size. In the numerical simulation of fluid–structure interaction using Fluent, the particles are simulated using the discrete phase model (DPM), which does not consider the particle pair reverse effects of airflow.
4.1.1. Al Single-Particle Deposition
To clarify the deformation ability of Al particles impacting the Al substrate, the impact behavior of single particles on the Al substrate was analyzed through the finite element method. In this simulation, the adhesion of particles to the substrate was not considered, and the speed and temperature of particles and the temperature of the substrate surface depended on the relevant CFD simulation results.
Figure 6 shows the morphology and simulation results of single Al particles with an average diameter of 55 μm deposited on an Al substrate (substrate temperature: 25 °C) under the parameters of 390 K and 580 m/s. The detailed boundary conditions of particles and substrate are described in
Figure 1a.
As shown in
Figure 6a, the Al particles and substrate suffer from large plastic deformation. Under the condition of high-speed impact, the particles undergo large deformation and combine with the substrate, and part of the material is squeezed away from the substrate due to the metal jet effect. The numerical simulation results are shown in
Figure 6b,c, The maximum local temperature and the maximum local stress of particles reach 728.794 K and 239.367 MPa, respectively. Suresh et al. [
40]. examined the microstructural evolution of single-grain Al particles during supersonic shock by quasi-coarse-grained dynamics (QCGD) simulations and identified the critical velocity for thermal softening at the periphery of the particle/substrate interface (i.e., adiabatic shear instability in the substrate/particles). The competing events of stress hardening and local softening at the interface under this velocity were used to explain the temperature and microstructural deformation behavior of Al particles during the deposition process, and it was proved that Al was effectively adaptable to material deformation before the occurrence of metal jets. According to the simulated plastic strain curve in
Figure 7b, it can be seen the that the deposition process of Al particles on the substrate involves high plastic strain. When the Al particles have kinetic energy, their plastic deformation rate is significantly higher than that of the substrate material. In the single-particle numerical simulation study by Lin et al. [
41]. the influence of interfacial bonding on the impact residual stress of the particles was dominant, which was higher than the impact response of the substrate material. This is consistent with the residual stress distribution shown by single-particle Al powder shock in this study. As shown in
Figure 7a, with the passage of the impact time, the stress rapidly rises from 0–10 ns. During this time, the Al particles and the substrate are just in contact, and most of the Al particles still retain their kinetic energy, as shown in
Figure 7d. With the progress of the deposition, the stress gradually becomes stable, most of the kinetic energy of the particles is converted into plastic strain energy, and a small part of the kinetic energy is converted into the internal contact elastic energy of Al powder to resist plastic deformation. During the entire deposition process, the particles have a high initial temperature (390 K), most of the heat is transferred to the substrate, and most of the heat generated during the particle impact deformation process is also converted into plastic strain energy.
Figure 7c shows the temperature changes in single-particle Al and the bonding part of the matrix with the process of deposition. Due to the large plastic deformation of single-particle aluminum, the ultra-high strain rate leads to a higher temperature change rate than that of the aluminum matrix in the bonding part.
According to the post-processed numerical simulation results for the deposition of single-particle Al powder, the plastic strain rate of Al powder is rather high during the CS deposition process, and the maximum value is
, which suggests that the Al powder has a good deformation ability. Schreibe et al. [
42] also found that Al particles had a rather high strain rate range, ranging from 10
−3 to 10
13, in the impact test of single Al particles, further indicating that Al powder exhibits an excellent deformation performance. In the numerical simulation of the deposition of basalt and Al powder mixed spray, it can be clearly observed that under the condition of high-speed impact, the edge of the particles is squeezed out of the metal jet when the Al powder is in contact with the substrate surface. These extruded Al powders in a high-temperature molten state can be well bonded to other particles and substrates in multi-particle mixed spray, which is beneficial to the adhesion and density of the coating. During the whole process, Al powder perfectly coats the basalt fiber in CS coating through metal interlocking, metal smelting, and other mechanisms and achieves better compactness at the same time, which can effectively improve the deposition efficiency of basalt fiber in the coating.
4.1.2. Cohesive Unit Suitability Verification
To verify the applicability of introducing a cohesive element to the numerical simulation of basalt fiber fracture, we designed the numerical simulation experiments of single-particle aluminum deposition and subsequent single-particle aluminum impact. Cohesive units have been applied to the deposition of cold-sprayed ceramic reinforced phases. Chakrabarty and Song [
43] initially used the smoothed particle hydrodynamics (SPH) method to simulate the brittle fracture behavior of ceramic particles after hitting the substrate. However, the ceramic morphology characterized by this method was quite different from the actual experimental results, and the SPH method was found to be mainly suitable for the Euler method. Then, to better compare the simulation and experimental results, Chakrabarty and Song [
26] generated grain boundaries in the ceramic particles by the Voronoi seed point method in the subsequent study and inserted condensed units. Additionally, CS ceramic deposits were destroyed.
In the model, we use basalt fiber to deposit parallel to the matrix plane. This is because when basalt fiber is deposited parallel to the matrix plane, its contact area with the matrix surface is the largest, which is the special case most prone to fracture. Of course, the positions and angles of basalt fiber deposition in front of the matrix are various, which we have taken into account in the subsequent numerical simulation of multi-particle mixed deposition.
It can be seen from the post-processing results of the numerical simulation shown in
Figure 8b that under the velocities of 500, 600, 700, and 800 m/s, the single basalt fiber does not break after hitting the substrate but rebounds on the substrate surface. Since the substrate is aluminum, which is a relatively soft material, the substrate deformation absorbs most of the energy during the collision of the stronger basalt fiber with the substrate, and the basalt fiber rebounds before reaching the material’s breaking strength of 3773 MPa. Accordingly, the subsequent impact of a single Al particle is examined. The particle diameter is 10–100 μm (based on laser particle size analysis). The average diameter is 55 μm, the particle speed is 600 m/s, and the temperature is 389 K.
According to the variation in the stress of a single basalt particle with time in
Figure 8b, the basalt fiber has the largest stress concentration effect when it just touches the substrate. However, most of the energy is absorbed by the substrate after the impact, and the remaining energy of the basalt fiber causes it to rebound on the substrate surface without further increasing the stress to reach the tensile strength of 3773 MPa, so it does not break while the substrate is deforming. It can be seen in
Figure 8a that the stress concentration caused by the basalt fiber when it just touches the substrate is equivalent to the stress concentration effect generated under the velocity of 600 m/s in
Figure 8b. However, with the further impact of the subsequent Al particles, the stress of the single-particle basalt fiber continues to rise and finally reaches 3891 MPa, which exceeds the tensile strength of the basalt fiber, and the material is then destroyed. Meanwhile, due to the continuous extrusion of Al particles, the basalt fiber does not rebound, and it is covered by Al particles and mechanically locked on the surface in contact with the substrate. To examine the interaction between CS particles, Yin et al. [
44] considered the subsequent consecutive impact of multiple particles to find that the effect of subsequent particles on the coating was rather significant. For successively impacted particles, the subsequent incident particles compacted the previously deposited particles, resulting in a coating with pores near the surface and a denser coating inside. This phenomenon verifies the applicability and accuracy of the cohesive parameter of single-particle basalt fiber and also shows that the fracture behavior of basalt fiber in the mixed spray is not only related to its properties, such as speed and temperature, but also to the continuous flow of Al particles. The impact on the basalt fiber morphology is also large.
4.2. Al–Basalt Fiber Co-Deposition
In our study, the finite element numerical analysis method of Fluent gas dynamics combined with ABAQUS explicit dynamics used has some limitations. The basalt fibers leaving the Laval nozzle have velocities in each degree of freedom direction, and therefore their morphologies are diverse. Modeling the analysis according to its velocity and morphology in each degree of freedom is a problem with a very large and difficult workload. We simplify the morphological diversity of basalt fibers by creating multiple poses of basalt fibers. Additionally, since the normal component of the velocity of the basalt fiber is the most dominant, we replace its velocity load with the normal component of the velocity.
For the simulation of multi-particle random deposition, the velocity and temperature of Al particles and basalt fiber before deposition are based on the CFD simulation data in the previous section. Next, a 15 vol.% and 30 vol.% basalt fiber-Al powder multi-particle random deposition simulation model has been established. According to laser particle size analysis, the number of Al particles and basalt fibers in the finite element model is shown in
Figure 9a,b.
Herein, the average diameter of basalt fiber was 15 μm, and the lengths were in the range of 60–148 μm (60, 80, and 100 μm), which were measured through the SEM images, and the kinetic analysis step was 650 ns.
To better observe the distribution of basalt fibers in the coating and avoid the inevitable loss of the fibers during the sample preparation process, the coated section is exposed by breaking, The SEM images and EDS analysis of the composite coating cross-section and surface are shown in
Figure 10.
According to the simulated deposition morphology (
Figure 11a,b), obvious metal smelting and mechanical interlocking effects exist between the Al particles (Al, basalt fiber deposition process dynamic video, can be downloaded from the support material link.). Through the SEM images of the fracture surface of the coating, we found the basalt fiber deposition morphology corresponding to that in the numerical simulation results, as shown in
Figure 11c–f. In the process of Al particle deposition, due to the adiabatic shear instability of the Al material, the plastic strain energy is dissipated, leading to the local temperature rise of the contact part between the Al particles and the matrix, inducing the thermal softening of the Al material in the contact part between the particles and the matrix, resulting in a jet effect similar to the flow of viscous materials.
Due to the jet effect of the Al material in contact with the substrate, the thermally softened Al particle materials contact each other. Some basalt fibers dispersed in aluminum powder materials bounce off the matrix after contacting the matrix, and some are covered by aluminum materials that are thermally softened and combined with each other. The particle materials without the metal jet effect will form a relatively closed space. A part of basalt fiber is locked in it. At the same time, due to the continuous impact of aluminum powder materials, the kinetic energy of particle impact causes basalt fiber fracture.
As shown in
Figure 12a, shows the stress changes in three basalt fibers of different lengths that first contact the matrix during the whole deposition process. Due to the impact of aluminum powder particles, the stress of basalt fibers oscillates. In the process of stress oscillation, fatigue damage caused by stress accumulates continuously. When the stress exceeds the breaking strength of the basalt fiber, the basalt fiber breaks, and the stress oscillation tends to decrease, which may be related to the change in the position of the basalt fiber under the impact of the aluminum particle kinetic energy. At the same time, the basalt fiber with a length of 100 μm reached the breaking strength at the earliest time, approximately 200 ns. The time for basalt fiber of 80 μm in length to reach the breaking strength is next, approximately 280 ns, and the time for basalt fiber of 60 μm in length to reach the breaking strength is the latest, approximately 320 ns.
Figure 12c shows the strain evolution of three basalt fibers of different lengths during the deposition process. It can be seen that with the progress of the deposition process, the strain value will have a sudden change, and the time of the sudden change coincides with the time when the stress reaches the fracture strength as shown in
Figure 12a. These phenomena indicate that the length of basalt fiber will affect its existence form after deposition, and shorter basalt fiber is easier to maintain its pre-deposition structure after deposition.
Figure 12c,d shows the stress and strain changes in the substrate in contact with basalt fibers of three selected lengths during deposition and the temperature changes in the parts in contact with the substrate. Due to the impact of the deposited aluminum powder and basalt fiber, the kinetic energy of particles and fibers is absorbed by the matrix, resulting in the deformation of the matrix, the temperature is increased, and the stress value of the matrix is also subject to vibration during the deposition process.
Similarly, for mixed spraying with a basalt fiber content of 30 vol.%, we also selected three lengths of basalt fiber that first contacted the substrate. As shown in
Figure 13a,c, the stress and strain changes in the three lengths of basalt fibers during the deposition process are shown. Compared with mixed spraying of 15 vol.% basalt fiber, the spraying basalt fiber content of 30 vol.% reaches the breaking strength earlier, and the stress value is more violent. Due to the increase in basalt fiber content, it is inevitable that basalt fibers will contact each other during the deposition process. The contacted basalt fibers will reach the fracture strength faster under the continuous impact of aluminum powder particles.
After the post-processing of the numerical simulation results, the basalt fibers in the coating show four different morphologies: fracture, transverse crack, bent, and deformed to a small extent. Due to the thermal softening of local materials induced by the adiabatic shear instability effect, this part of aluminum materials produces jet phenomenon and combines during the bonding process with the matrix. Due to the thermal softening of local materials induced by adiabatic shear instability effect, during the bonding process between aluminum particles and the matrix, the metal aluminum with local thermal softening will produce a metal jet effect similar to the viscous material flow phenomenon, and the flowing aluminum metals will bond with each other.
At the same time, there are two situations: first, the thermal softened aluminum metal will cover the basalt fiber; second, the aluminum metal of the non-softened part will form a closed space to lock the basalt fiber. Under the continuous impact of subsequent aluminum particles, due to the low hardness of the thermally softened aluminum metal, the basalt fiber coated with aluminum will directly receive most of the impact kinetic energy of the particles, causing the basalt fiber with transverse cracks to continue to deform, and then a brittle fracture occurs. The aluminum metal forming the enclosed space has a certain strength, which can offset part of the kinetic energy impact. Therefore, basalt fibers locked by the confined space can maintain the existence of transverse cracks, bending, and other morphologies.
The basalt fibers extracted from the numerical simulation results are shown in
Figure 14. We found that basalt fibers characterized by bending and transverse cracks only appear in the coating area forming an enclosed space. Among the curved basalt fibers, most of them are 100 μm long, and only a few of them are 60 and 80 μm long. The basalt fiber covered by the heat-softened aluminum metal has the characteristics of brittle fracture, and most of the 80 and 100 μm basalt fibers split into three segments after brittle fracture, while the 60 μm basalt fibers generally split into two segments. This shows that the aspect ratio of basalt fiber will affect its morphology after deposition. Similarly, being in different positions of the coating (such as material covering and confined space locking) also affects the morphology of basalt fiber.
To explore the influence of the long-diameter ratio on basalt fiber deposition, a mixed-deposition experiment with a diameter of 12 μm and six long-diameter ratio gradients (2:1, 3:1, 4:1, 5:1, and 6:1) was designed. A total of 100 basaltic fiber mixed depositions of 10–100 μm diameter Al particles were employed, and the relevant powder particle temperature and velocity load were determined according to previous CFD numerical simulations.
As shown in
Figure 15, based on numerical simulations, we found that there was no brittle fracture of basalt fibers with a length-to-diameter ratio of 2:1. Cracks occur when the aspect ratio is 3:1, and when the length ratio is greater than 3:1, the basalt fibers are brittly fractured. The results show that the short-cut basalt fiber can reduce and avoid cracks and brittle fractures by adjusting the aspect ratio. Related scholars have found that common ceramic-reinforced phases such as Al
2O
3 [
16], WC-25Co [
45], WC-17Co [
22], and SiC [
17] show brittle fracture and cracking patterns after deposition in composite coatings. This is similar to the brittle fracture and transverse cracking that occurs when basalt fibers are deposited. However, the particle morphology determines that bending and deformation do not occur, which is a deposition morphology unique to basalt fiber deposition influenced by the aspect ratio.
From the numerical simulation results after the deposition of basalt fiber with two contents in
Figure 16, we found that increasing the basalt fiber content in the mixed raw materials can indeed increase its probability of deposition. However, the two deposition forms of material covering and confined space locking also indicate that the uniform dispersion of basalt fibers in mixed raw materials also plays an important role in improving the deposition probability of basalt fibers.
According to the simulated 1/2X cross-sectional strain and stress cloud (
Figure 16), the basalt fibers deposited in the coating are completely covered by molten Al powder and are mechanically locked, which is consistent with
Figure 14. This SEM images of related samples in
Figure 16b,d are in agreement with this observation. According to the simulated cross-sectional equivalent plastic strain of coating with 15 vol.% basalt fiber (
Figure 16a), the deposited basalt fibers have fewer voids around the coating, and the coating is relatively dense. However, under the equivalent plastic strain of the coating section with 30 vol.% of basalt fiber (
Figure 16c), more basalt fiber deposition is observed, but the pores around the basalt fibers are larger, and the deformation degree of Al powder is higher than that in the bigger coating with 15 vol.% basalt fiber. This indicates that the increase in the content of basalt fiber in the mixed powder affects the density and energy of mixed powder deposition.
4.3. Coating Performance
The hardness of the coatings with basalt contents of 15 vol.% and 30 vol.% was measured by a Rockwell hardness gauge (Model: FangYuan XHRD-150). To ensure the accuracy of the results, 10 test points were randomly selected on the coating surface with a size of 40 × 30 mm
2. Data were collected for each test point, and the average Rockwell hardness of the two coatings was obtained.
where
is the residual indentation depth increment under the initial testing force F0 after removing the main testing force; the
value of the ball’s Rockwell scale is 130; C is constant, generally C = 0.002 mm. The relevant experimental data are shown in
Figure 17.
According to the data of the Rockwell hardness test, the Al–basalt powder composite coating has a high hardness, and with the increase in the basalt content, the Rockwell hardness increases from HRC38.9 to HRC49 Its hardness is similar to that of WC-17Co composite coating prepared by Emmanuel et al. [
22]. This suggests that the basalt fiber reinforcement phase improves the hardness of the composite coating.
In the friction and wear test data, the stable value of the dynamic friction coefficient is close to the WC-25Co wear-resistant coating prepared by Dosta et al. [
45].
Figure 18a,b show the friction-wear test results, which reveal an interesting phenomenon. Specifically, as the load increases, the kinetic coefficient of friction of the composite coating shows a decreasing trend, which is the opposite of the behavior in the general metal friction-wear test. Similarly, under the same load condition, the kinetic coefficient of friction of composite coating with 30 vol.% basalt fiber is higher than that of coating with 15 vol.% basalt fiber.
We found that the dynamic friction factor of the composite coating in the wear test had oscillations, especially in the composite coating with 30% high basalt fiber content. In addition, with the increase in the load applied by the spherical grinding head in the wear experiment, the vibration tends to decrease and gradually stabilize. Therefore, we observed and characterized the worn surface of the composite coating by SEM, and the results are shown in
Figure 19.
As shown in
Figure 19a, the color of the coating compacted by the spherical grinding head in the wear experiment is white. With the reciprocating motion of the spherical grinding head, the coating starts to wear gradually. In this process, the enclosed space formed by no thermally softened Al particles is gradually exposed in the form of pits. At the same time, the basalt fibers in the coating also begin to be exposed and fall off, leaving pits at the falling position, as shown in
Figure 19b. This will change the surface roughness of the coating and make the dynamic friction coefficient vibrate. Different from the metal material wear test, this vibration will exist until the coating is worn through and contacts the metal substrate. At the same time, the detached basalt is in constant contact with the spherical grinding head in the reciprocating motion. Repeated rolling reduces the size of basalt fiber, similar to the wear particles in the wear experiment. Some studies have shown that when the size of abrasive particles is less than 50 μm [
46], the sliding friction will change into rolling friction, which will reduce the dynamic friction coefficient. This can also explain the problem that the dynamic friction coefficient decreases before the coating wears through. With the increase in basalt content, the number of basalt shed and exposed in the wear test will also increase, resulting in more pits on the coating surface, making the vibration of the dynamic friction coefficient more obvious.