**Preface**

It is well known that research focused on the creation of specific materials that can withstand high temperatures, high wear, and high corrosion is essential for the industrial sector. Many issues in manufacturing are related to the wear and corrosion of parts, which serves as the primary cause of the failure of the entire product. According to their engineering use, the local characteristics of materials are altered during the design process. One effective method for extending a component's life is coating. Additional techniques include functionally graded materials, friction stir layer deposition, additive layer coatings, and friction stir coating. These methods can be used to solve problems related to the high wear and corrosion of an automobile's mating parts as well as enhance their mechanical and microstructural qualities. Techniques for plastic deformation, such as severe plastic deformation (SPD) and friction stir processing (FSP), are important factors to improve surface characteristics and a material's overall performance.

As a result, several advanced methods for depositing nano- and microstructured coatings by thermal layer deposition, microwave-assisted coatings, cold-spray and electroless coatings, along with well-known characterization approaches, have been presented here.

Other similar techniques not exactly related to the type of coating, but that still improve microstructural and mechanical surface properties in a similar way, are included in this volume. The readers of this volume can explore various experimental results that provide a wide range of processing and characterization methods. For example, friction stirring was used to create the surface composite Al359/Si3N4/Eggshell, and the influence of this surface composite on microstructural and tribological characteristics was examined. The management of dynamic and static parameters was examined in order to enhance the welding strength achieved using friction stir spot welding (FSSW), with an innovative control system based on fuzzy logic to optimise the process. Due to their inadequate tribological properties, titanium and titanium alloys are insufficient in friction and wear conditions. Because of their homogeneous thickness, better hardness, and superior corrosion resistance, electroless coatings have been discovered to be efficient as surface treatment techniques. The investigation of the tribological features of microwave-assisted g-C3N4/MoS2 nanocomposite coatings indicated a number of potential research directions. The results revealed that the addition of g-C3N4 in nanocomposites can reduce friction and improve wear life, producing superior results to those obtained with MoS2 alone. Biowaste, such as CBP (chicken bone powder), an example of a biowaste material produced in a poultry farm, contains particles that can be utilized to enhance the overall properties of a material. Surface composites may be an appropriate alternative material for structural and automobile components. Various applied loads, sliding speeds, and weight percents of additive graphitic carbon nitride in a molybdenum disulphide nanocomposite coating on a steel substrate were employed in tests to enhance the settings for Pin-on-Disc wear apparatus. A high-performance base oil additive compound called Liquid Crystals (LCs) was created to handle a variety of operating circumstances. Investigations were carried out into how the mesogenic phase temperatures range of LCs affected tribological features. These are some of the interesting findings that are discussed in this volume.

#### **Ashish Kumar Srivastava and Amit Rai Dixit** *Editors*

## *Editorial* **Research Progress in Metals and Alloys by Thermal Layering and Deposition**

**Ashish Kumar Srivastava 1,\* and Amit Rai Dixit 2,\***


Over the last 20 years, because of their superior hardness, chemical stability, and outstanding oxidation barrier, many coating systems have now been extensively researched using various deposition processes and employed for wear-resistant protection [1]. These surface coatings have been explored across a variety of technological disciplines, including aeronautics and transportation, tools and die, chemicals and petrochemicals, nuclear research, electronics, and healthcare. Coating technology applies single or several layers of an appropriate substance to a material surface without changing the bulk material's composition, allowing it to function in different environments and overcome challenges caused by abrasion, temperature, fatigue, corrosion, erosion, and friction [2]. Consistent research on innovative coatings has significantly contributed to worldwide economic advances over the past few decades. Depending on the exact use, different coating materials (such as ceramics, metals, composites, or polymers) and coating methods are used [3].

In the production of conventional devices, top-down procedures, such as etching and photolithography, are often used for patterning at the nano scale. However, bottom up procedures are increasingly being looked at as potential substitutes owing to physical restrictions on downscaling those processes. A vapor-phase process called atomic layers deposition is used to deposit thin films on different substrates through a series of selfcontained surface reactions [4,5]. Molecular layering (ML) and atomic layers epitaxy (ALE), two techniques that were initially presented in 1970s, form the foundations of ALD [6]. Graphite, graphene, and amorphous state carbon are examples of carbon-based inhibitors that are often deposited via solution methods, such as ion implantation, or chemical vapor deposition (CVD). However, it is challenging to use these techniques on substrates with high aspect ratios [7]. Similar to ALD, a molecular layers deposition (MLD) technique can produce conformal thin films of materials on three-dimensional objects [8]. Substantial thermal stresses occur between the coating and the metallic bond due to the latter's poor fracture toughness and comparatively low coefficient of thermal expansion. Consequently, the thermal cycle life of the single layer covering is often limited [9]. The multilayer concept is an efficient technique used to overcome such drawbacks and enhance shock life due to thermal loading [10]. Additionally, as an outer layer, the multilayer consists of an erosion-resistant layer, thermal barrier layer, corrosion- and oxidation-resistant layer, a layer controlling thermal stress, and a layer resisting diffusion [11].

A total of 13 papers were published in this Special Issue. From these 13 papers, a total of 8 papers focused on surface modification, surface treatment, microstructural examinations and mechanical properties. The published papers form comprehensive and sufficient learning materials that are sure to attract the attention of scholars in the manufacturing field. Authors also made significant efforts to produce qualitative research. For example, Kumar et al. [12] investigated the mechanical properties of nano-composites with a lower concentration of reinforcing nano particles, such as graphene, and ceramics in the matrix epoxy to assess the stability and damping properties of hybridized epoxy,

**Citation:** Srivastava, A.K.; Dixit, A.R. Research Progress in Metals and Alloys by Thermal Layering and Deposition. *Coatings* **2023**, *13*, 366. https://doi.org/10.3390/ coatings13020366

Received: 31 December 2022 Revised: 1 February 2023 Accepted: 3 February 2023 Published: 6 February 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

employing vibration methods to obtain precise findings. Mechanical testing, such as three-point bending, validates the efficiency of the impact hammer vibration method as a function of the Young's modulus of a nanocomposite. The nanocomposite of graphene contains 1% of epoxy, whereas the nanocomposite of ceramic contains 3% of epoxy. The frequency reduction in the heated environment was found to be much lower hybrid nanocomposites, but the decrease in pure epoxy was high; therefore, advancements in thermal and mechanical stability were found.

Srivastava et al. [13] investigated the surface tribological performance of Al359/Si3N4/ Eggshell by friction stir processing (FSP). In the past, researchers explored techniques that will help to develop lightweight materials with dimensional accuracy, a smooth surface finish, and less production cost. With all these abilities, it will become eco-friendly. An aluminum matrix composite is an advanced engineering material in this field of research. The impression of the microstructural properties is to develop a defect-free surface that will help with grain refinement and enhance the mechanical properties of the top surface of the base alloys. Si3N4 is used in defense applications due to its ballistic and mechanical properties. Si3N4 is evaluated as a standard reinforced material that has a low density, high melting point, high thermal stability, and good chemical stability. Eggshell is a valuable material used as a new, reinforced engineering material as it contains around 95% calcium carbonate (CaCO3), 3% phosphorus, and some other materials, such as zinc, potassium, sodium copper, magnesium, and iron. Many studies have been conducted to improve the mechanical and modified microstructures of matrix materials that have several disadvantages, such as porosity, solute redistribution, and solidification cracking. FSP is a popular technique that is used for versatile, solid-state processing, and green-energy-efficient techniques. It has negligible residual stresses and a refined microstructure, densification, and structural homogeneity. Generally, composites have a high coefficient of friction. Surface composites are made using a vertical milling machine. Al-6% Si3N4/Eggshell composites are undertaken to investigate frictional properties with the help of tribological tests. Light microscopy and FE-SEM equipped with EDS mapping are used in microstructural research. SiC, Al2O3, and B4C are reinforced in the metal matrix, improving their tensile strength, yield strength, and hardness but reducing ductility. Liquid- and solid-phase fabrication methods have been successfully conducted to form the desired composite.

Lashin et al. [14] investigated the regulation of dynamic and static factors in order to maximize the welding strength achieved via friction stir spot welding (FSSW). As a novel technique, a control system based on fuzzy logic was applied to improve the process. Static and dynamic conditions influenced the tensile strength of stir spot welding. The collected findings demonstrate that the fuzzy logic system was a simple and low-cost technology that could be applied in the prediction and optimization of FSSW strength [14]. Gunduz et al. [15] discovered that titanium and its alloys had insufficient friction and wear settings due to their poor tribological characteristics. Electroless coatings were discovered to be effective as surface enhancement treatments due to their homogenous thickening, higher hardness, and superior resistance to corrosion. The auto-catalytic reduction in coating processes significantly improved surface quality. To improve weak tribological properties, an electroless coating of Nickel-P-Gr was applied to a Ti6Al4V alloy, and samples were put to abrasion in a linear-type reciprocating ball on plate configuration to examine tribological characteristics. Its hardness increased by approximately 34% with a graphene-reinforced Nickel-P coating in the electroless coatings procedure, whereas it increased by about 73% with applied thermal treatments. Moreover, the substrate's wear rate was nearly 98% greater than the heat-treated nanocomposite coatings. The heat-treated nano composite coatings had the maximum wear resistance.

Saxena et al. [16] investigated the tribological properties of microwave-assisted g-C3N4/MoS2 nanocomposite coatings. The calcination approach was used to produce the g-C3N4 nanosheet, and the microwave-assisted method was used to developed nanocomposite with MoS2. A pin on disc setup was used to investigate tribological qualities. A morphological examination indicated that elements coexisted, and the layered structure of

MoS2 was evenly distributed over gC3N4. The presence of MoS2 nano particles reduced the pore volume and surface area of the composite, confirming that the pores of calcinated gc3n4 were filled by MoS2. The tribological properties of the nanocomposite were studied under various conditions, such as applied load, sliding speeds, and material compositions. The findings suggest that the inclusion of g-C3N4 in nano composites can minimize friction and enhance wear life. These results were superior to those obtained with MoS2 alone.

Kumar N et al. [17] used three waste materials as a reinforcement to develop surface composites of aluminum alloys using FSP. Currently composites are being used as substitutes for many alloys due to their high strength-to-weight ratio, hardness, and tensile strength. Surface composites can be suitable alternative material to structure and automotive components. Hybrid composites consist of two or more reinforcements, whereas a non-hybrid has only one reinforcement. Biowaste materials also have particles that can strengthen certain materials [18]. Chicken bone powder (CBP) is an example of biowaste material, which is produced in poultry farm. The use of waste in products and its disposal, are great challenges since they form unwanted and harmful. CBP is strengthened by its sufficient amounts of carbon and calcium. Walnut shell powder (WSP) is a green waste formed during the manufacturing of walnuts products, containing cellulose and lignin. It is able to enhance the bonding strength of existing material. Rice husk powder (RHP) is agricultural waste produced by farming. The storage of this waste remains a major issue; many farmers burn this waste, which causes significant pollution. These three biowastes are utilized in the matrix alloy of aluminum and processed with FSP to take advantage of grain refinement and surface treatment. As a result, tensile modulus, yield and ultimate strength, percentage elongation and elastic coefficient are improved by 20%–30% [19].

Saxena et al. [20] attempted to improve the settings for the Pin-on-Disc wear apparatus. Various applied loads, sliding speeds, and the wt. of additive graphitic carbon nitride in a molybdenum disulfide nanocomposite coating on a steel substrate were used in the studies. ANOVA was performed to evaluate the effect of the interactions between different criteria. The maximum influences of applied loads on friction coefficient and wear were found to be approximately 59.6% and 41.4%, respectively, with sliding speed coming in second. The optimum conditions for the lowest coefficient of friction and wear depth in the nanocomposites were established to be at 15 N applied load, 750 mm/s sliding speed, and 9 wt.% of CN. Even at a 95% level of confidence, applied loads were determined to have the greatest impact on COF then sliding speed, whereas material composition had the greatest impact on wear. Taguchi technique and RSM results correlate well with experimental tests.

Wu et al. [21] created liquid crystal compounds (LCs) as a high-performance additive for a base oil that can fulfil a range of operating conditions. The influence of the mesogenic phase temperatures range of LC on tribological characteristics was investigated. Compared to the basic oil, the addition of LCs caused a significant decrease in the fictional coefficient (21.57%) and width of wear (31.82%). Furthermore, LCs exhibited improved tribological properties under temperature conditions of the mesogenic phase.

The above literature constitutes a summary of previously published research. However, some additional papers are still under extensive review to help them fit the scope of this Special Issue.

**Author Contributions:** Conceptualization, writing and editing, A.K.S.; Supervision, review and editing, A.R.D. All authors have read and agreed to the published version of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Review* **Cold Spray Coatings of Complex Concentrated Alloys: Critical Assessment of Milestones, Challenges, and Opportunities**

**Desmond Klenam 1,2, Tabiri Asumadu 3, Michael Bodunrin 2, Mobin Vandadi 4, Trevor Bond 5, Josias van der Merwe 2, Nima Rahbar 4,5 and Wole Soboyejo 5,\***


**Abstract:** Complex concentrated alloys (CCAs) are structural and functional materials of the future with excellent mechanical, physical, and chemical properties. Due to the equiatomic compositions of these alloys, cost can hinder scalability. Thus, the development of CCA-based coatings is critical for low-cost applications. The application of cold spray technology to CCAs is in its infancy with emphasis on transition elements of the periodic table. Current CCA-based cold spray coating systems showed better adhesion, cohesion, and mechanical properties than conventional one-principal element-based alloys. Comprehensive mechanical behavior, microstructural evolution, deformation, and cracking of cold spray CC-based coatings on the same and different substrates are reviewed. Techniques such as analytical models, finite element analysis, and molecular dynamic simulations are reviewed. The implications of the core effects (high configurational entropy and enthalpy of mixing, sluggish diffusion, severe lattice distortion, and cocktail behavior) and interfacial nanoscale oxides on the structural integrity of cold spray CCA-based coatings are discussed. The mechanisms of adiabatic heating, jetting, and mechanical interlocking, characteristics of cold spray, and areas for future research are highlighted.

**Keywords:** complex concentrated alloys; cold spray; adiabatic heating; solid-state coating; metallurgical bonding; mechanical interlocking; severe lattice distortion; Johnson–Cook model; finite element method; sluggish diffusion; microstructural evolution

#### **1. Introduction**

Complex concentrated alloys (CCAs) are the new frontier for the design of bulk materials and coatings for structural and functional applications [1–5]. This approach by design has disrupted ~5000 years of conventional and serendipitous material discovery processes. The CCA design approach is based on mixing elements in equal amounts without any distinct solvent and solute atoms [5]. This is counterintuitive to the well-established physical metallurgy of using one or two base elements, while systematically adding minute secondary alloying elements to induce the designed and desired properties. These properties could be mechanical, corrosional, thermal, magnetic, electrical, and physical. The approach deviates from exploring compositions at the corners of typical ternary and higher order phase diagrams. This new approach shows promise and is being touted as the future of materials, providing superior structural and functional properties to most conventional alloys [1].

**Citation:** Klenam, D.; Asumadu, T.; Bodunrin, M.; Vandadi, M.; Bond, T.; van der Merwe, J.; Rahbar, N.; Soboyejo, W. Cold Spray Coatings of Complex Concentrated Alloys: Critical Assessment of Milestones, Challenges, and Opportunities. *Coatings* **2023**, *13*, 538. https:// doi.org/10.3390/coatings13030538

Academic Editors: Ashish Kumar Srivastava and Amit Rai Dixit

Received: 30 January 2023 Revised: 21 February 2023 Accepted: 24 February 2023 Published: 1 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Cold spray is classified as a thermal spray coating technique and an additive manufacturing process [3,6–16]. The technique has been used to deposit coatings [7,9,17,18], to repair materials [6,8,19], and for design of bulk materials for various structural and functional applications [14–16]. The main advantages are: (i) easily applicable to a wide variety of ductile, dissimilar, and thermally sensitive materials, (ii) able to retain initial properties of the feedstock powder material, (iii) high flexibility and precise near-net shape manufacturing, (iv) deposits possess cold-worked microstructural features contributing to high hardness and relatively low porosity, and (v) superior deposition efficiency and rates compared to other types of conventional coating process. Lastly, the method has the unique characteristic of being operated under relatively low temperature (below the melting temperature) of the coating material. The layer-by-layer deposition satisfies the principles of additive manufacturing.

The inherent design principles of bulk materials from CCAs make them expensive compared to most conventional structural and functional alloys [1,2,20]. This hinders scalability and stiff competition from commercially available conventional alloys is inevitable [2,20]. To reduce overall costs, coatings using various techniques are being explored.

Cold spray technology is gaining traction and has been commercialized for various functional and structural applications as coatings and for bulk alloy design. It is instrumental for material repair. As one of the hotspot research areas, there has been increased and timely experimental and published literature on the subject. Since its discovery over 40 years ago in Russia, there have been thematic reviews, book chapters, and books on various aspects of cold spray technology [3,6,9,10,17,21–26]. These ranged from fundamental principles and applications to material perspectives [9,21–25]. Some of the reviews focused on various bonding mechanisms [6,17,27]. The latest review on the bonding mechanism focused on single-particle impact approaches [17]. Similarly, there have been lab-scale experimental approaches augmented with numerical simulations to understand the process better.

Bulk CCAs are expensive with stiff competition from dilute conventional alloys. The design of CCA-based coatings is one of the solutions to reduce the cost. Various types of CCA-based coatings have been developed and reviewed. The cold spray approach, since its proof of concept on CrCoFeMnNi alloy in 2019, is still in its infancy. Apart from a mini review of structural integrity and material aspects of CCA-based cold spray coating published by the authors [3], there is yet to be published a general overview, which critically assesses the current stage of knowledge in the area. This is ideal in highlighting areas for future research direction. By combining the advantages of CCA-based coating and the cold spray technique, robust and next generation multifunctional coatings can be developed and applied in various industries.

#### **2. Thermal Spray Technology—Brief Overview**

Thermal spray coating is a coating process that sprays the surface of any substrate with melted or heated particulate matter [28,29]. The thermal energy that heats or even melts the feedstock (precursor coating material) is generated via electrical or chemical means. The molten or heated (semi-molten) particulate matter possesses kinetic energy which helps in accelerating the particles onto the substrate. Adhesion and cohesion are induced by the high temperature and high velocity of the particle which in effect leads to severe plastic deformation. The deformed coating material produces a splat on the substrate, which is a pancake-like impacted particle. The bonding mechanisms are: (i) mechanically induced bonding of the coating particles' splatter on substrate with an interlocking effect, (ii) localized diffusion of the coating and the substrate material, and (iii) bonding is achieved from the interactions of the van der Waals forces of attraction. The main benefits are the ability to repair and strengthen damaged surfaces and high applicability to metals and ceramics. It also requires less heating for effective bonding at the coating and substrate interface. A schematic diagram showing the process is given in Figure 1.

**Figure 1.** Schematic diagram showing the thermal spray process.

Thermal spray coatings are in three categories based on application of combustion heat sources (flame, detonation gun, and high-velocity oxygen fuel spray), electrical energy sources (plasma or an arc) and thermal energy sources (kinetic, cold, and hypersonic spray) from gas expansion. The classification of thermal spray based on the interdependence of multiplicity of input variables is represented by a process map (Figure 2) [29]. These input variables include the particle temperature at optimized stand-off distance, the particle velocity, and feedstock particle size distribution with its associated physical and mechanical properties [29]. The process map provides a guide for the classes of materials and the suitability of the coating method and their effects on porosity and cohesion strength. The relatively high operation temperatures of conventional thermal spray coatings have some drawbacks. These include high tensile residual stresses, high oxidation rates, defects nucleated in the form of cracks, and phase transformation. These features affect the mechanical, chemical, and physical properties of the coating, hence some of these coatings cannot be used in temperature-sensitive operations. The development of cold spray coatings is one of the ways to minimize the challenges posed by other thermal spray processes.

**Figure 2.** Classification of thermal spray coating methods based on processing parameters such as particle velocity, feedstock particle size distribution, and particle temperature at optimized standoff distance.

#### **3. Cold Spray Technology: An Overview**

*3.1. Introduction and General Concepts*

The cold spray process of thermal spray coating was discovered and operationalized in 1980s. It is based on the principle that when a particle-laden supersonic gas jet impinges onto a solid substrate above a critical minimum particle velocity, there is a transition of the metallic particles from a state of abrasion to adhesion, which through plastic deformation are deposited on the substrate. These minute unmelted particles are typically in a size distribution range of approximately 1–50 μm. The system was fully commercialized after the Russian and United States patents were granted. The technological and innovation timeline of cold spray is given in Figure 3 with a project plateau of productivity by 2030.

**Figure 3.** Technology and innovation maturation timeline and projection for cold spray technology [9].

The operational principle is as follows with the schematic diagram shown in Figure 4. The whole architecture of the system is built on high pressure (Figure 4a) [15]. However, recent modification led to the low-pressure system as shown in Figure 4b. Based on Figure 4a, high-pressure gas, which is typically, air, nitrogen, or helium, enters the system and then splits into 5 and 95%. The 5% of the gas goes to the particle feeder adding particles to the flow, whereas the 95% goes to the gas heater. The gas is heated to ~500 to 1000 ◦C and the two streams recombine at the converging–diverging nozzle (the same as that used in rocket engines, which is called a de Laval nozzle) entrance, then accelerated through the nozzle. The velocity at the nozzle entrance is generally low which increases to as much as ~1000 m/s depending on the gas used. In the case of He, the velocity is very low as He is relatively lighter than nitrogen. The stream then impinges on the substrate while the gas turns away, whereas the particles crash on the surface of the substrate due to the high momentum they possess. The longer the flow is kept, the thicker the deposit. The working principle of cold spray is based on the conservation of energy. The impacting feedstock material possesses kinetic energy, which is converted to heat and deformation energies, causing localized severe plastic deformation upon impact. The heat energy is then transferred to the surface of the substrate, resulting in some adiabatic heating but not high enough to induce phase transformation.

**Figure 4.** Schematic diagram showing typical cold spray system showing (**a**) high-pressure and (**b**) low-pressure systems. Reprinted with permission from Ref. [3]. Copyright 2022, Elsevier.

The gas is heated to take advantage of the supersonic nozzle design. The nozzle converts the high thermal energy of the gas to kinetic energy at the exit of the nozzle. Thus, the process starts with a hot but slow mixture of gases and feedstock particles at the entrance of the nozzle and ends with a cold fast mixture at the exit. Generally, temperatures as high as 500 ◦C can be observed at the entrance of the nozzle but they reduce drastically to ~100 ◦C at the exit of the nozzle, resulting in no melting of the feedstock particles. Therefore, it is a solid-state process of surface coating where deposition of coating can occur from 0 to 800 ◦C, which is lower than the melting point of the feedstock.

#### *3.2. Processing Parameters of Cold Spray Coatings*

Typical cold spray processes have been used to manufacture various bulk mechanical components [10–16]. Among these components are flanges, cylindrical walls, tubes, heat exchanger array fins, and metal labels. The flexibility of the coating process and no detrimental effect on the substrate make it applicable for repair of worn parts of any mechanical component. As a form of additive manufacturing process, comparative analyses of the techniques with three methods of fusion-based additive manufacturing are given in Table 1. The comparison was also made using as-fabricated (AF) and heat treatment (HT) specimens.

**Table 1.** Comparison of cold spray technology with three methods under powder-based additive manufacturing process.



#### **Table 1.** *Cont.*

SLM—Selective Laser Melting, EBM—Electron Beam Melting, and LMD—Liquid Metal Deposition.

Gas propulsion parameters are required to accelerate the particles of the powder required to bombard the substrate to induce bonding. The parameters are the nature, type, temperature, and pressure of the gas system. For instance, the pressure of the gas is used to classify cold spray technology as either a high- or low-pressure cold spray system. The high-pressure cold spray system has pressures exceeding 1 MPa. The effects of processing parameters of cold spray coating are summarized in Table 2.

**Table 2.** Effects of the processing parameters on defects, adhesion, and deposition efficiency of cold spray coatings.


Note: ↑ shows an increase whereas ↓ shows a decrease in most observed experiments.

The effect of particle velocity on residual stresses in the cold spray method is inconclusive for most systems studied [30–38]. The three main residual stress states are peening, thermal, and quenching stresses [31,32,38]. The high-impact velocity of the impinging particles contributes to the peening effects [30]. The mechanism is likened to shot peening for increased fatigue strength in most structural components [38–44]. The quenching stresses are attributed to the compaction of impacted particles often impeded or restricted by the underlying coatings or substrate materials. Thermal stresses are due to differences in the coefficient of thermal expansion (CTE) of the substrate and coating materials. Based on the magnitude of the CTE mismatch of the coating and substrate, tensile or compressive residual stresses can be induced [38,45]. Typical compressive residual stresses are observed in the coatings and tensile stresses in the substrate when CTEsubstrate > CTEcoating [29]. Due to the low deposition temperature of the cold spray coating technique, the cumulative effects of thermal and quench stresses are negligible [29]. However, these effects are pronounced when the temperature of the substrate is above 400 ◦C and, vice versa, below 400 ◦C [29].

There is no clear and discernible relationship between residual stresses and impact velocity from both experiments and simulations [33–38,46]. In most instances, residual stresses increase with increasing particle velocity [33,34,36,37], but that is largely dependent on substrate-to-coating pairs and microstructural characteristics. However, this opens opportunities for further research to optimize the process parameters to obtain properties within acceptable limits. A typical example is Cu coatings on Cu substrate, where the difference in CTE is negligible [35]. The residual stresses increased with increasing impact velocity from 300 to 500 m/s. However, when particle velocities increased from 500 to 700 m/s, no discernible relationship was observed. For a Cu/Al substrate–coating pair, residual stresses increased for particle velocities between 300 and 500 m/s. The compressive residual stresses were increased for particle velocities between 500 and 700 m/s [35]. A similar trend of increasing residual stresses with increasing particle velocities has been reported for Ti6Al4V on Ti6Al4V [46]. The residual stress profiles for particle velocities from 700 to 800 m/s were comparable.

Particle size, distribution, and morphology: This is a useful parameter which affects the deposition behavior and quality of cold spray coatings [47]. It is characterized by the particle size, distribution, and morphology. The acceleration of large particles is poor due to the lower velocity of the particles compared to small particles. However, the larger the particle size, the easier it is to reach high temperatures, which contributes to reduced critical velocities [47]. A schematic diagram showing the effect of particle size on the particle impact velocity is given in Figure 5 [48]. There is always an optimum particle size range that impacts the substrate above the critical velocity and below erosion velocities. The optimized particle size range, which is specific to the coating material, leads to good adhesive, cohesive, and quality coatings.

**Figure 5.** Effect of particle size distribution on the particle impact velocity during a cold spray deposition [49].

#### *3.3. Bonding Mechanisms of Cold Spray of Metallic Materials*

The bonding mechanism of cold spray technology shows how the coating and substrate achieve adhesion and cohesion [17,27,47]. While it is known to be a solid–solid bond interaction, the overarching bonding mechanisms are still a bone of contention. However, the core mechanism is based on the ballistic impingement theory where particles are consolidated into coatings on substrate by ballistic impingement [50]. Two main conditions for the theory are the need for particles of certain sizes (below 5 mm) to be accelerated to certain critical velocities (50–3000 ms−1) to induce impact [50,51]. The high kinetic energy impingement of the particle on the substrate results in increased plastic strain and strain rates. Upon impact, the kinetic energy is converted to heat and sound which induce adiabatic heating leading to thermal softening and dynamic hardening effects. There are few reviews on the subject focusing on particle–substrate and particle–particle interactions and bonding mechanisms [52–57]. Metallurgical bonding has been observed using various microstructural characterization techniques [17,27,47,52,53,55–57]. These include scanning and transmission electron microscopy [55,58,59] and various empirical models [48,56,57] have been developed to explain the mechanism.

The schematic representation of the bonding mechanism of a typical cold spray process (tantalum particle on 4340 steel substrate) is given in Figure 6. The process commences with high kinetic energy of the impinging particles from the high velocities and high pressures, generating extreme plastic deformation at the surface of the contact region. This process disrupts any thin surface oxides and exposes clean surfaces of the substrate, aiding the metallurgical bonding process (Figure 6). As the impact progresses, the plastic flow of the materials at the interface extrudes much of the crushed pieces of oxide films to the periphery of the contact interface, enabling an intimate contact of newly exposed clean metal between the particle and substrate (Figure 6). Some of the oxides are removed from the interface altogether through "jetting" observed for particle impacts above the critical velocity. This is shown experimentally in a wide range of engineering materials. For a strong bond to be achieved in non-ductile materials, the systems require the addition of ductile matrix to induce metallurgical bonding and mechanical interlocking, which are critical for adhesion strength [53,54,57].

**Figure 6.** Schematic showing the bonding mechanism of a typical cold spray process.

The presence of oxide-free surfaces due to the initial particle–substrate is the precursor for metallurgical bonding. Surface preparation methods are used to create the oxide-free surfaces. Jetting contributes to removal of interfacial oxides. For many materials, the bulk plastic deformation of particle and substrate results in mechanical interlocking, a characteristic bonding mechanism. Typical bonding strength of cold spray coatings ranges from 10–60 MPa [57]. For the bonding mechanism to be effective, the following conditions should be met [47,52,55,57]:


The cold spray process is a solid-state process without any element of melting. However, there is localized heating when the particle impinges on the substrate. This is known as adiabatic shear instability (ASI). This is a critical contribution factor to the overall bonding mechanism and strength of the coating. The high strain rates at which the particles are accelerated onto the surface of the substrate result in the localized adiabatic heating.

#### **4. Computational and Numerical Simulation Models—An Overview**

Three main types of models are used to simulate the cold spray processes. These include the finite element-based models, constitutive equations, and models for describing the bonding at the particle–substrate and particle–particle interfaces [39,58]. A summary of the models and their applications to cold spray processes are highlighted.

#### *4.1. Modeling Approaches Using Numerical Simulations*

The numerical approaches used for simulating the cold spray processes focus on impact and deformation associated with particle–substrate and particle–particle interaction [17,27,60–73]. The main numerical methods are the Lagrangian [74,75], Eulerian [17,27,60,61,63,64], coupled Eulerian–Lagrangian [57,61,62,76,77], smooth-particle hydrodynamics [65–73], and molecular dynamics [17–19,38].

#### 4.1.1. Lagrangian Approach

The Lagrangian finite element-based approach applies mesh which deforms with the material under applied loads [59]. Tracking particle–substrate and particle–particle interaction is easy during impact even if complex boundary conditions are imposed [74,75]. It is also a good approach to trace any history-dependent variables, hence it is one of the pioneering approaches for simulating solid-state bonding processes. It is widely used due to reduction in computational time as the impact process is assumed as symmetric and quarter (axisymmetric) models are representative and cost effective [38].

The Lagrangian model is great for showing material jetting, which is characteristic for most impact-induced bonding. Although the jets are sharp and sharps are unphysical, they result in severe mesh distortions. This severe distortion of the meshes could compromise the simulation, resulting in premature termination and reducing computational accuracy significantly [60,75,78], sometimes at 19 ns [60]. Severe mesh distortion can be attributed to large strains imposed on the structure, resulting in failure, attributed to near-zero or negative element Jacobians. Localized strains could be as high as 450% in the jetting region, which could lead to overestimation of the bond modeling.

Ways to mitigate the mesh distortion have been explored [57,61,62]. One approach is the use of the arbitrary Lagrangian–Eulerian (ALE) which combines the Lagrangian and Eulerian analyses [59]. The ALE allows for the redefinition of the mesh in a continuous fashion arbitrarily while moving the mesh away from the material. The major challenges associated with the ALE technique are high computational cost [17,27,57,59], unrealistic deformation of particles at high impact velocities [61,62], inaccurate prediction of the particle– particle interactions, and overall reduction in equivalent plastic strains [17,27,60,63]. For the ALE method, the inherent interpolation errors are due to high strain gradients at the particle–substrate interface and adaptive remeshing [60]. In a nutshell, the pure Lagrangian approach is not the ideal technique for a high-strain cold spray modeling process.

#### 4.1.2. Eulerian Approach

The Eulerian approach to modeling cold spray processes improves on the inefficiencies of the pure Lagrangian formulation [27,61,64], thus solving the severe mess distortions and large deformations induced by high strains. The Eulerian approach allows the flow of material mass through stationary points where deformation is independent of the material [64]. This process assumes two overlapping meshes where one is fixed as the background and the other, mesh for the material which allows flows through it easily. The model is flexible, allowing for assigning different parameters to the different parts of the model. Due to the ease of flowability of the material in the mesh, it is easy to model the severe plastic deformation associated with the cold spray process. The error associated with the Eulerian approach is negligible compared to experimental data [64]. For instance, the predicted critical velocity of Cu was compared to the experimental value. It is the right method for modeling the material jetting process of the particle and the substrate. Generally, the demerits associated with this method are long simulation and computational times due to the fine mesh and severe plastic deformation, the difficulty with modifying boundary conditions, and the inability to modify contact properties such as the coefficient of

friction [64]. The inability to track the history-dependent variables in the Eulerian method and numerical dissipation are also possible shortcomings.

#### 4.1.3. Combined (Coupled) Eulerian–Lagrangian Approach

This approach uses the Eulerian and Lagrangian approaches simultaneously to model aspects of the cold spray process [59,64,76,77]. This allows for the merits associated with each technique to be applied in a robust fashion, thus, solving the false positives or impractical results associated with particle–substrate deformation and mesh distortions associated with the Lagrangian method, while being able to track the particle–substrate interface, which is difficult under the Eulerian method. For the Eulerian part, the volume of fluid method is applied [77].

#### 4.1.4. Particle-Based Approach

The particle-based approach is ideal for modeling severe plastic deformation associated with the cold spray process [65–67]. Results from this technique are comparable to experiments that do not underestimate, or overestimate, as observed for most meshbased techniques [65,66,69]. The approach possesses merits that are not observed for the conventional mesh-based FEM Lagrangian and Eulerian methods [66,67]. A typical example is the smoothed particle hydrodynamics (SPH), a meshless discretization process that deviates from finite 3D meshed elements to discretized particles, the carrier of the information [65–68]. The particles are assigned state variables which describe material based on weight, velocity, and stress states. An authoritative review of the mathematical underpinnings of the SPH [66–70] and key advantages are as follows:


Considering that SPH is a mesh-free technique, an important factor is the interpolation theory. This is the process where continuum fluid dynamics laws of conservation are converted to integral equations with an interpolation function. While the application of the shape functions is difficult to implement, a weighting SPH kernel function according to Equation (1) is applied [19,71–73]. Thus, kernel approximation of the function f(r ) at specific positions is obtainable by using smooth kernels, W, which are integrated over a certain computational domain. Other governing equations and detailed assumptions are documented elsewhere [66,69,72].

$$\mathbf{f}(\mathbf{r}) = \int \mathbf{f}(\mathbf{r})' \mathbf{W}(\mathbf{r} - \mathbf{r}', \mathbf{h}) \, \mathrm{d}\mathbf{r}' \tag{1}$$

where: W = weighting function, h = support scale satisfying the normalization process when W(r − r , h)dr = 1, the delta function property achieved when smoothing length tends to zero is given as limh→<sup>0</sup> W(r − r , h) = δ(r − r ).

Some of the different interpolation kernels used in SPH modeling are given in Table 3 [66,79–81]. These kernel expressions cater for 2D domains only. Typical factors to consider in deciding the type to use are the desirable mathematical behavior and the lack of complexity to adequately describe the properties of the particle. Detailed derivation and the mathematical underpinnings of the kernel function have been summarized elsewhere [66]. The material parameters used for the SPH simulations include material properties (density, Young's modulus, and Poisson ratio), strength parameters (yield strength, hardening coefficient, strain rate constant, softening exponent, strain hardening exponent, heat capacity, thermal conductivity, melting and reference temperatures, reference strain rate) and equation of state (EOS) parameters (speed of sound, gradient of shock velocity, and Gruneisen coefficient).


**Table 3.** Typical interpolation kernels used for SPH modeling.

Some examples of SPH models for cold spray include SPH simulation of cold spray oblique impacting Cu particles which showed well-fitted results to the experimental data [72]. The issues of mesh distortions were completely avoided when the SPH method was applied [72]. There was reduced dependency on the particle weight in the case of the SPH technique compared to the mesh based Lagrangian methods [73]. The effect of interfacial oxide layers on the impact processes during the cold spray was investigated using the particle based SPH technique [65]. Excessive deformation, complex time evolution associated with free surfaces, and large thermal transport processes associated with the mesh based Lagrangian FE model were overcome using the SPH approach [65,66,69,70].

Ceramic particles were deposited using cold spray at room temperature and studied using the SPH [82]. The contact surfaces were free, which was vital in predicting critical velocities and their effects on deposition efficiency. Two main mechanisms were observed through simulation and experiment, which are fragmentation from the submicron- to the nanoscale and the bombardment of the submicron particles, providing enough bonding energy as pressure and thermal energy for the fragmented particles through the shock waves [82–84]. Although the deformation behavior was adequately simulated, the method was not appropriate for determining the critical and maximum velocities. The technique provided the basis for understanding the deposition behavior numerically.

An SPH and experimental study on cold spray between similar and dissimilar substrates was carried out [84,85]. The ratio of the deposition and rebound energy was used for the cold spray coating evaluation based on porosity rate, hardness, and bonding strength. The quality of the coating on similar metals was better than those of similar metals, as predicted based on high deposition energy and low rebound energy. The effect of interfacial oxides has also been studied using the SPH approach and the results correlated well with experiments [65]. Thus, the application of SPH to cold spraying has been successfully implemented in optimizing process parameters. However, the issues of computational cost and tensile instability are still a strategic research direction.

#### *4.2. Finite Element-Based Numerical Models for Cold Spray Processes*

Six main computational models have been used to explain and simulate cold spray particle impacts. These models are used for high strain rate plasticity of various feedstock metallic powders and the overall deposition properties. The models include the Johnson– Cook (JC) plasticity model [85,86], Zerilli–Armstrong (ZA) model, Voyiadjis–Abed (VA) model, Preston–Tonk–Wallace (PTW) model [87], Khan–Huang–Liang (KHL) model, and the Gao–Zhang (GZ) model. The shortcomings of the models led to modified models where initial assumptions are improved to fit predicted data to the experiment ones. The common and widely used model is the JC model which is built into the ABAQUS/Explicit finite element analysis platform.

#### 4.2.1. Johnson–Cook Model

This model was designed by Johnson and Cook and relates the flow stress to the strain rate, plastic strain, and temperature [38,86]. It is simple to use and successful in estimating flow stress more accurately for various engineering applications. The model has three parts focusing on strain hardening, strain rate hardening, and thermal softening as shown in Equation (2) [86]. The model is not suitable for predicting relatively high flow stresses at very high strain rate. In the case of Cu, the JC model fails when the strain rate exceeds 10−<sup>5</sup> s−<sup>1</sup> [86]. There is a linear relation between the work hardening and the strain rate on a logarithmic scale. This is indicative of the shortcoming of the JC model as it underestimates the material and mechanical behavior at very high strain rates.

$$\sigma = \left(\mathbf{A} + \mathbf{B}\varepsilon\_{\mathbb{P}}^{\mathbf{n}}\right) \left[1 + \mathbf{C}\ln\dot{\varepsilon}\_{\mathbb{P}}^{\*}\right] \left[1 - (\mathbf{T}^{\*})^{\mathbf{m}}\right] \tag{2}$$

The homologous temperature in the JC model, which is T∗, is given and defined in Equation (3), where the absolute, transition or reference, and melting temperatures are T, Tr, and Tm, respectively.

$$\mathbf{T}^\* = \begin{cases} 0, & \mathbf{T} \prec \mathbf{T\_r} \\ \frac{\mathbf{T} - \mathbf{T\_r}}{\mathbf{T\_m} - \mathbf{T\_r}}, & \mathbf{T\_r \prec T \prec T\_m} \\ 1, & \mathbf{T} \succ \mathbf{T\_m} \end{cases} \tag{3}$$

where: the parameters A, B, n, C, and m are material-dependent constants, ε<sup>p</sup> is an equivalent plastic strain, . ε ∗ <sup>p</sup> is an equivalent plastic strain (ratio of the plastic strain rate and reference strain rate).

The JC model is the most successful constitutive model for simulating cold spray behavior of various metallic alloys [38,48,85,86,88–92]. Typical applications of the JC model for Cu, Ni, and 316 steels with the respective JC parameters are given in Table 4. Similar studies have been carried out using refractory Ta powders on 4340 steel substrates [93] and the effect of nanoscale interfacial oxides on the deformation and cracking phenomena on cold spray Al 6061 powders [94]. These studies used experimental and computational approaches. An overview of some of the ferrous and non-ferrous cold spray coating phenomena, deformation mechanisms, material jetting, and the effects of various microstructural features has been presented [3,95–97].


**Table 4.** Typical JC model and general parameters used for numerical simulation of cold spray for Cu, Ni, and 316SS extracted from published literature [48,86,95–97].

To make up for the lapses in the original JC model, a modified JC model is proposed for relatively high strain rates [96–101]. This is given in Equations (4) and (5) [102,103]. The JC and modified JC models have fewer material constants compared to the typical Preston–Tonk–Wallace (PTW) model. These JC models are extensively used for various materials with their constants are easily accessible in the literature [102–105].

$$\sigma = \left(\mathbf{A} + \mathbf{B}\varepsilon\_{\mathrm{p}}^{\mathrm{n}}\right) \left[\mathbf{1} + \mathbf{C}\ln\frac{\dot{\varepsilon}\_{\mathrm{P}}}{\dot{\varepsilon}\_{0}} \left(\frac{\dot{\varepsilon}\_{\mathrm{P}}}{\dot{\varepsilon}\_{\mathrm{c}}}\right)^{\mathrm{D}}\right] \left[\mathbf{1} - \left(\frac{\mathbf{T} - \mathbf{T}\_{\mathrm{r}}}{\mathbf{T}\_{\mathrm{m}} - \mathbf{T}\_{\mathrm{r}}}\right)^{\mathrm{m}}\right] \tag{4}$$

$$\mathbf{D} = \begin{cases} \mathbf{x}\_{\prime} \dot{\boldsymbol{\varepsilon}}\_{\mathbf{P}} \ge \dot{\boldsymbol{\varepsilon}}\_{\mathbf{c}} \\ \mathbf{0}\_{\prime} \dot{\boldsymbol{\varepsilon}}\_{\mathbf{P}} \le \dot{\boldsymbol{\varepsilon}}\_{\mathbf{c}} \\ \dot{\boldsymbol{\varepsilon}}\_{\mathbf{c}} = \mathbf{y} \mathbf{s}^{-1} \end{cases} \tag{5}$$

where: D is a non-zero (x) parameter when . ε<sup>p</sup> (plastic strain rate) is within a certain critical strain rate (. <sup>ε</sup>c) value and . ε<sup>0</sup> is the reference strain rate.

A typical bilinear JC model, which is an improvement on the previous JC models with the capacity for two-stage rate sensitivity, has been used for many systems [94,106,107]. The bilinear model incorporates a second constant to improve the approximation of high strain rate sensitivity [107]. A typical bilinear JC model has been implemented in ABAQUS–Explicit for the study of interfacial oxide effects on cold spray of Al [94].

#### 4.2.2. Preston–Tonk–Wallace (PTW) Model

The PTW model was designed for estimating the material behavior of relatively high strain rates as shown in Equations (6) and (7) [87]. This is to correct the challenges associated with the JC model which breaks down at strain rates above 104/s [17,108]. This is a constitutive parametric model designed based on the dislocation movement and mechanisms during plastic deformation [108]. Typical PTW models have been used for wide strain rate ranges (10<sup>−</sup>3–1013/s) [17,108].

$$\sigma = 2 \left[ \pi\_{\sf s} + \alpha \ln \left[ 1 - \varphi \exp \left( -\beta - \frac{\theta \varepsilon\_{\sf P}}{\alpha \varphi} \right) \right] \right] \mu(\sf p, T) \tag{6}$$

$$
\alpha = \frac{\mathbf{S}\_0 - \mathbf{\tau}\_\mathbf{y}}{\mathbf{d}}, \boldsymbol{\beta} = \frac{\mathbf{\tau}\_\mathbf{s} - \mathbf{\tau}\_\mathbf{y}}{\alpha}, \boldsymbol{\varphi} = \exp(\boldsymbol{\beta}) - 1 \tag{7}
$$

where: τs= normalized work hardening saturation stress, s0 = saturation stress at 0 K, τy= normalized yield stress, θ = strain hardening rate, ε = equivalent plastic strain, d = dimensionless material constant. The "μ" is the shear modulus, which is a function of the temperature and easily estimated using the mechanical threshold stress (MTS) shear modulus model given in Equation (8) [109]. The τ<sup>s</sup> and τ<sup>y</sup> are given in Table 5.

$$
\mu(\mathbf{T}) = \mu\_0 - \frac{\mathbf{D}}{\exp\left(\frac{\mathbf{T}\_0}{\mathbf{T}}\right) - 1} \tag{8}
$$

where: μ<sup>0</sup> = shear modulus at 0 K, D and T0 = material constants, and T = temperature of the material [109]. The standard parameters for the PTW model are the strain rate dependence constant, strain hardening rate, strain hardening constant, yield stress constant at 0 K, yield stress constant at melting, medium strain rate constant, high strain rate constant, high strain rate exponent, saturation stress at 0 K, saturation strength at melting, temperature dependence constant, atomic mass, shear modulus, material constant (D), and temperature material constant [110].

**Table 5.** Parameters for normalized work hardening and normalized yield stress.


The PTW model is more complicated than the typical Johnson–Cook model, but there is more control of the material parameters. Recent investigations on dilute or conventional metallic materials such Cu and WC-Co and cold spray coating applications have been reported [104,108,111]. Plastic deformation during cold spray processes has been modeled using the PTW model for strain rates up to 107/s with comparable results to experiments [87,109]. The PTW has also been applied within the ALE framework due to the good fit at very high strain rates [102].

#### 4.2.3. The Zerilli–Armstrong (ZA) Model

The ZA model is for estimating plastic deformation behavior of materials, focusing on flow stress at relatively high temperatures [103,105,112]. The flow stress is given by the relation shown in Equation (9). The ZA model is a physical model and applies to a range of materials with varying crystal structures sensitive to temperature and strain rate. The exponential term in Equation (9) describes the thermal stress component which is derived experimentally. As the temperature tends to infinity, the thermal stress components become zero.

$$\sigma = \left(\mathbf{C}\_1 + \mathbf{C}\_2 \varepsilon^n\right) \exp\left\{-\left(\left(\mathbf{C}\_3 + \mathbf{C}\_4 \mathbf{T}^\*\right)\mathbf{T}^\* + \left(\mathbf{C}\_5 + \mathbf{C}\_6 \mathbf{T}^\*\right)\ln\dot{\varepsilon}^\*\right)\right\} \tag{9}$$

where: T<sup>∗</sup> = T − Tr, C1, C2, C3, C4, C5, C6, and n = material constants, ε = equivalent plastic strain, . ε <sup>∗</sup> = normalized equivalent plastic strain, T = absolute temperature, and Tr = reference temperature.

The modification of the ZA model is carried out to improve the temperature-dependent terms of the original model. The modified ZA model gives better results for temperatures above 300 K, and the work hardening is independent of strain rate and temperature, which is an assumption that drives the original model. The flow stress based on the modified ZA model is given in Equation (10).

$$\sigma = \text{B}\varepsilon\_{\text{P}}^{0.5} \left( 1 - \sqrt{\mathbf{x}} - \mathbf{x} + \sqrt{\mathbf{x}^3} \right) + \text{C}\_6 \tag{10}$$

#### 4.2.4. The Voyiadjis–Abed (VA) Model

The Voyiadjis–Abed model was designed as an improvement on the ZA model [113]. The prediction is more efficient for higher strain rates and higher temperatures than the ZA model. The model is summarized in Equation (11).

$$\sigma = \text{B}\varepsilon\_{\text{P}}^{\text{n}} \left( 1 - \left( \beta\_1 \text{T} - \beta\_2 \text{T} \ln \varepsilon\_{\text{P}} \right)^{1/\text{q}} \right)^{1/\text{P}} + \text{Y}\_{\text{a}} \tag{11}$$

where: <sup>ε</sup><sup>p</sup> = equivalent plastic strain, . ε<sup>p</sup> = plastic strain rate, T = temperature, and B, Ya, β1, β2, p, q, and n = material constants.

#### 4.2.5. The Other Types of Models

There are other types of models that have been used for estimating the mechanisms of cold spray methods [89]. These include the modified Zerilli–Armstrong [113,114], modified Khan–Huang–Liang (MKHL) [115,116], and the Gao–Zhang models [117]. The details of these models are summarized elsewhere [89].

The current state of the art, looking at numerical and experimental approaches for the design of cold spray coatings are discussed. Emphasis is on concentrated amorphous and crystalline coatings. The amorphous alloy coatings are mainly bulk metallic glasses (BMGs), whereas the crystalline materials are complex concentrated alloys (CCAs).

#### **5. Bulk Metallic Glass Cold Spray Coatings**

Bulk metallic glasses (BMGs) have gained traction since the 1990s [118–120]. These are mainly amorphous or partially crystallized materials with medium- to long-range disorder devoid of typical lattice defects such as dislocations and grain boundaries [120–126]. They undergo strain softening, localized shear instability, and easily rupture under applied loads. Most BMGs possess high elastic modulus and high strength [120]. There is a trade-off between strength and ductility, resulting in poor plasticity. Challenges associated with BMGs as structural materials include: (i) size limitation due to limited glass-forming abilities (GFAs) and very high cooling rates (>10<sup>5</sup> K/s) [123]. These alloys can be produced to about few centimeter sizes and are mainly wires, ribbons, and powders, restricting large scale and industrial applications [120–122]; and (ii) poor ambient temperature plasticity and deformation is constrained to concentrated shear zones [125,126]. Thus, they are not the right candidate for load-bearing structural applications [124,125].

Some of the main BMG coating systems include Al, Cu, Fe, Ni, and Zr. Characteristic features and suggested applications of these coatings are given in Table 6. Advantages of BMG cold spray coatings over traditional BMG coatings are:



**Table 6.** Thermal spray coating of various types of metallic glasses.

Overall, the distinctive combination of qualities provided by BMG cold spray coatings makes them a desirable option for a variety of applications, including those in the aerospace, automotive, energy, and biomedical industries.

Bulk metallic cold spray coatings are desirable due to the low temperature and solidstate deposition mechanism [54,140–143]. These contribute to effective reduction in porosity and retard recrystallization during cyclic thermal stresses. The low temperature leads to restraining of distortions which could arise from thermal stresses and oxidation phenomena. The process does induce compressive residual stresses in BMG cold spray coatings, while improving metallurgical bonding at the coating–substrate interface and between the various layers of the coatings [140–143]. These alloys have been investigated using numerical and experimental approaches, showing excellent functional and structural properties. The BMG coatings produced from cold spray with processing parameters are given in Table 7.


**Table 7.** Bulk metallic glass cold spray coatings on various substrates with processing parameters.

#### *5.1. Aluminum-Based BMG Cold Spray Coatings*

Aluminum bulk metallic glass alloys have low density, high specific strength (1000–1500 MPa) [154,155], great corrosion resistance, and high modulus. They are difficult to deform due to the disordered arrangement of atoms with the low critical resolved shear stresses of Al [120]. A few Al-based BMG coatings have been designed and studied [120,144,145] and their compositions are based on the ternary system (Al–TM–RE), where TM stands for transition metal (Ni, Co, or Fe) and RE stands for rare earth (La, Ce, Gd, or Y) [103]. The transition metals are used to accelerate the atomic packing process, whereas the rare earth elements induce glass formation [103,129,145,156–158]. Some of the main cold spray coating systems are Al-Co-Ce [129,156], Al-Ni-Ce [103], Al-Y-Ni-Co-Sc [145], and Al-Ni-Y-Co-La [144].

The wear and mechanical behavior of Al-based BMG cold spray coatings on Al substrate has been studied. The coating had high hardness due to the glass-forming properties and superior wear resistance was observed compared to the substrate. For corrosion resistance, the coatings were over five times better than the substrate when tested in NaCl solution. The coating has two main factors contributing to the high corrosion resistance. Due to the amorphous nature of the coatings and the chemical homogeneity, it prevents the onset of the formation of galvanic cells [154]. Furthermore, the corrosion resistance improves due to the presence of nanocrystals which accelerate the formation of protective passive films resulting from rapid diffusion of passive element to the interface between the substrate and coatings as well as the surface. This has also been confirmed experimentally [155,157,158]. These nanocrystals were confirmed with high-resolution transmission electron microscopy. Similar results were obtained for AlNiYCoLa on Al 7075 substrate [144]. The combined effect of the amorphous phase and the Al2O3 passive oxide layer was the contributing factor to the superior corrosion resistance.

Wear behavior is dependent on the volume fraction of the amorphous phase of the Al-based BMG coatings. The amorphous phase proportion, thickness, strength, porosity, and hardness are dependent on the processing parameters (stand-off distance, gas pressure, and temperature). Excellent wear resistance was observed for Al-based coating with ~81% amorphous phase. The coefficient of friction was also reduced by more than 30%. The wear mechanism of the coating was abrasive grooving with minute surface splat delamination.

#### *5.2. Copper-Based BMG Cold Spray Coatings*

Copper-based BMG alloys have structural and functional properties for various engineering applications [159–162]. These properties include corrosion resistance and high strength and hardness which are applicable in biomedical, aerospace, electronics, and sport equipment. The binary systems which are the base for Cu-based BMG are Cu-Ti, Cu-Ni, and Cu-Zr [159–162]. For the 54Cu–22Zr–18Ti–6Ni BMG coating on Al 6061 substrate, a

porosity below 5% was observed with coating thickness ranging from 300–400 μm [161]. The hardness of the coating was 412.8 HV, whereas the wear resistance was three times better than that of the pure Cu coatings. The adhesion strength of the BMG coatings was better than that of the pure Cu coatings due to the synergistic effects of the alloy elements controlling the deformation mechanisms of the substrate. Poor corrosion resistance of the BMG coatings was observed, which is attributed to the high porosity in the amorphous coating layer [161].

Binary 50Cu–50Zr [146] and ternary Cu–Ti–Ni BMG cold spray coating systems have been studied for biomedical applications [163]. In the case of the equiatomic binary Cu–Zr coatings, the gas temperature was critical for the deposition efficiency. The main zones were no bonding, weak bonding, great bonding, and viscous flow. Due to the poor adhesion behavior associated with no bonding and weak bonding, there was no coating on the substrate. The deposition was carried out at 600 ◦C and 800 ◦C with the highest hardness and cohesive strength at 800 ◦C.

The BMG coatings of 50Cu–50—x(Ti)–xNi were produced from low-energy ball milling (mechanical alloying) and cold spray [163]. The milling process reduced the particle sizes and offered some mixing, whereas the cold spray was used to deposit the coatings on austenite stainless steel substrate. The 50Cu–20Ti–30Ni and 51Cu–17Ti–13Ni coatings had the best wear resistance and relatively low coefficient of friction (0.32–0.45), with ~50% reduction in friction coefficient on 304 steel substrate (Figure 7) [163]. The coating sufficiently inhibited the formation of biofilm, thus being a promising coating for biomedical applications. This was mainly due to enhancement of antimicrobial effects of Ni by the relatively high amounts of Ti and Cu.

**Figure 7.** Comparison of the coefficient of friction of CuTiNi BMG cold spray coating with 304 steel substrate where (a) is the untreated 304 substrate, (b) the 304 substrate with Cu-based metallic glass coatings and (c) the experimental set-up for the wear testing [163].

Numerical and experimental approaches have been used to study the nanocrystallization of CuNiTiZr BMG coatings [147]. The numerical and experimental results showed that the kinetic energy of the impacting particles has a significant impact on the activation energy for nucleation and the fraction of crystallinity in BMG coatings [147]. The highvelocity impact resulted in a reduction in the free energy barrier and an increase in the driving force for the phase transition from amorphous to crystalline due to the kinetic energy of the particles. When the CuNiTiZr BMG was subjected to the cold spray technique, the microstructural characterization showed that the nanocrystallization of the substance was related to the strain energy produced by rapid plastic deformation.

#### *5.3. Iron-Based BMG Cold Spray Coatings*

These alloys have unique structural properties such as wear resistance and great strength behavior [120,134,164] and functional properties such as superior glass-forming ability, corrosion resistance [165–167], and soft magnetic properties. Initial works include production of cold spray coatings of Fe-Cr-W-Mo-Mn-C-Si-B-Zr on Al substrate showing ultralow porosity along the nominal microcracks [148,165]. Microhardness of the BMG cold spray coating is ~639 HV0.3, which is about ten times higher than that of the Al substrate [148]. The degree of amorphization was poor as most of the particles were not fully amorphous. The drawback of the coating on the substrate is the small thickness of ~0.2 mm, which is not durable for long-term and aggressive corrosion or oxidation environments. The bonding mechanism is attributed to adiabatic shear instability which leads to metallurgical bonding at the BMG particles and Al substrate [143,148]. Localized deformation was observed at the particle–particle boundary and softening resulted from adiabatic deformation within the feedstock particles during the formation of the splat [88,168–170]. For typical BMG coatings, the deposition efficiency (DE) is one of the main parameters that ensure structural integrity and the quality of the coating [141,150,166,171]. The DE is also influenced by processing parameters such as gas temperature, stand-off distance, substrate temperature, and process gas pressure (Table 2) [150,166,171].

The tribological behavior of BMG coatings produced from cold spray on various metallic substrates has been studied [120,143,149]. For FeCSiBPCrAlMo coatings, a lower coefficient of friction was observed compared to bearing steel [149]. Similarly, the wear behavior of cold spray coatings of FeCrMoCBY is better than that of BMG coatings produced using other thermal spray methods [143]. The superior tribological behavior of cold spray coatings of BMG compared to high-velocity air fuel (HVAF) was due to excellent adhesion at the particle–particle and particle–substrate interface [18,172,173] and lack of oxidation due to the low heat input. The CS coating also had ~2% porosity and high hardness due to grain refinement resulting from the peening effect associated with cold spray [173–175].

#### *5.4. Nickel-Based BMG Cold Spray Coatings*

Typical Ni-based BMG coatings have high strength, high thermal stability, and great corrosion resistance [120,176,177]. The Ni-based BMG systems are from the ternary compositions of Ni–Ti–Zr, Ni–Zr–Al, and Ni–Nb–Ti with slight modification of the microstructure with minute addition of metalloids (Si or B) [120,176,177]. Some of the major compositions which have been studied extensively include 57Ni–18Ti–20Zr–3Si–2Sn [176–178], 59Ni–20Zr–16Ti–2Si–3Sn [151,178], and 53Ni–20Nb–10Ti–8Zr–6Co–3Cu [177], where the compositions are according to weight percent.

The main properties of interest for Ni-based BMG cold spray coatings on different substrates are wear and corrosion resistance [120,145,151,152,179,180]. Cold spray coatings showed better wear and corrosion resistance than other thermal spray techniques such as vacuum plasma and high-velocity oxy-fuel spraying. This is mainly due to negligible phase transformation of the BMG produced by cold spray [151]. The propelling gas plays a critical role in the deposition efficiency of the BMG on mild steel [103]. By switching from nitrogen to helium, there was significant improvement in deposition efficiency [151,152]. The high drag force of helium and the high heating rate of feedstock led to increased splat to crater ratio, enhancing mechanical properties and bond strength [120]. This has been observed in NiTiZrSiSn BMG coating [145,151,152,179,180].

#### *5.5. Zirconium-Based BMG Cold Spray Coatings*

Zirconium-based bulk metallic glass (BMG) cold coatings are protective coatings applied to a variety of substrates [118,120,121,123,181]. These amorphous metallic coatings have a unique combination of properties, including high strength, high corrosion resistance, excellent wear resistance, and excellent biocompatibility for biomedical applications [103,120,153,182,183]. These amorphous alloys have a wide supercooled region with excellent glass-forming abilities. The common compositions are based on the quaternary

and quinary systems of Zr–Cu–Al–Ni and Zr–Cu–Al–Ni–Ti [118,120,181]. There have been some additions of Be and Fe for excellent mechanical properties, corrosion and wear resistance [120,181].

The mechanical properties such as hardness and strength are functions of operating temperature and the volume fraction of the amorphous phase. For a typical ZrCuAl-NiTi/Cu metallic glass coating, the hardness increased from 500 HV at 400 ◦C to 620 HV with a 50◦C increase in temperature.

#### **6. Complex Concentrated Alloys—Brief Overview**

Complex concentrated alloys are a broad class of structural and functional materials which are defined by their design strategy [5,184]. These encompass the medium-entropy alloys, high-entropy alloys (HEAs), and multiple principal element-based alloys (Figure 8). These alloys, by composition, occupy the central region of a typical phase or composition diagram as shown in Figure 9. The high-entropy alloys have multiple principal alloying elements with at least five components. They are in near-equiatomic ratios ranging from 5–35 at% [184]. The design philosophy of HEAs is premised on the assumption that forming solid solutions by constituent elements leads to high configurational entropy or entropy of mixing.

**Figure 8.** Complex concentrated alloys and their derivatives.

**Figure 9.** Typical illustration of dilute (compositions at the corners) and complex concentrated (compositions at and near the center) alloys.

These alloys extend the frontiers for alloy design beyond the conventional dilute alloying concentrations with potentially different properties for structural and functional applications [1–3,20,185]. The excellent structural and functional properties are attributed to four core effects.

#### *6.1. Core Effects of Complex Concentrated Alloys*

There are four main core effects associated with CCAs/HEAs, which are high configurational entropy of stabilization, sluggish diffusion, severe lattice distortions, and cocktail effects. The core effects and the relationship between the science and the physical metallurgy of CCAs are schematically shown in Figure 10. The entropy effect focuses on the thermodynamics; sluggish diffusion drives the kinetic phenomenon; severe lattice distortions focus on deformation theory, solid-state physics, and the strengthening mechanisms. The cocktail effects relate to how the structural and functional properties are impacted. The three core effects of high configurational entropy of stabilization, sluggish diffusions, and severe lattice distortions have been a bone of contention and require further investigation. They are not overarching and there is limited evidence that the "core effects" play significant roles in determining the unusual properties of CCAs. A brief description of the core effects is summarized.

**Figure 10.** The interrelationship between core effects, the science, and physical metallurgy of CCAs.

High configurational entropy and enthalpy of mixing: The high entropy and enthalpy of mixing effects deviate from the Gibbs phase rule stabilizing phase proportions. This contributes to the observed properties, which are not common to most conventional oneor two-principal element-based materials. Increasing the constituent elements lowers the Gibbs free energy of the system, which outweighs the driving force for detrimental intermetallic phases and restricts their formation. Thus, most CCAs/HEAs should have simple and stabilized solid-solution phases. Configurational entropy and the enthalpy of mixing are related by Boltzmann's relation (Equation (12)).

$$
\Delta \mathbf{S}\_{\rm conf} = \mathbf{k} \ln \mathbf{w} \tag{12}
$$

where: k = Boltzmann's constant, w = possible ways of mixing available energy.

Initial families of CCAs/HEAs supported the assumption of high configurational entropy of mixing. Recent families of HEAs deviate from the possibility of forming stabilized solid solutions across the full range of temperatures. For example, the Cantor alloy (CrCoFeMnNi) is observed to be stable as a face centered cubic solid solution at high temperatures. However, precipitates of tetragonal sigma phases can be observed at intermediate temperatures. A few five-component CCAs are stable to intermetallic formation, hence the high configurational entropy stabilized is not an overarching effect and not quite strong enough for the observed properties in typical CCAs/HEAs. Common additions leading to the decomposition of the solid-solution phases are Ti, V, Cr, Cu, Al, and Mo due to their size difference and entropy. The concept of Hume-Rothery rules still applies to the dominance of enthalpy of formation. Thus, more investigation is required to establish the underlying factors and physical metallurgy principles.

Sluggish diffusion: Based on the atomic environment, there is slow diffusion due to variation in the potential energy of the lattices. This results in low kinetic transformation leading to sluggish diffusion effects. Slow diffusion of the mixing atoms results in variation of lattice potential energy due to differences in the atomic environment. Due to many constituent elements, some of the atoms are easily trapped, contributing to the sluggish diffusion phenomenon. This increases recrystallization temperature, reduces particle coarsening rate, and then leads to nanocrystalline structure.

Severe lattice distortion: This phenomenon is attributed to differences in size, bond energy, and crystal structures, resulting in mismatch [186–189]. It promotes the overall reaction rate [186,188]. Thus, localized and severely distorted lattices are observed but do not have the same effects of broadening X-ray diffraction peaks as in the case of dislocation defects [186,188]. The severe lattice distortions result in more incoherent scattering with reduced intensity of the XRD peaks [190,191]. The best way to measure the severe lattice distortions and quantify them is by a pair distribution function from the total scattering data, an extension of XRD, and neutron diffraction measurements [190–192]. In Figure 11, the weighted average distance to the nearest neighbors (interatomic distance) is shown and the width of the peaks estimates how much distortion occurs within the structure. Neutron diffraction has been used to estimate the lattice distortions in the Cantor alloy which is given in Figure 12 [186,188,189,193].

**Figure 11.** Schematics showing pair-wise distribution function for estimating the severe lattice distortions in CCAs.

**Figure 12.** Normalized pair distribution function 3 for typical Ni-based alloys compared with the CrMnFeCoNi alloy based on the first six coordination shells [189].

Based on the pair distribution function, any severe distortions are easily quantified. Each peak in Figure 12 corresponds to a coordination shell around the atom at radius r and the area depends on the interaction of the atomic species and the respective coordination number [189]. In the case of severely distorted lattices, there is an increment in the width of individual peaks due to the noticeable displacement of the atoms from the original lattices or ideal positions. Once these displacements are significant within the coordination numbers with low "r" shells, the cumulative effect at a larger "r" causes more broadening of the peaks, resulting in poorly discernible features as opposed to that which is observed for the alloys in Figure 12. In Figure 12, no significant differences are observed for most of the peaks for the alloys studied, thus well-defined lattices are observed with no severe lattice distortions.

Cocktail effect: This is the overall effect from composition–processing–microstructure of the alloy. This effect is different from traditional or conventional alloys. Although this effect is not clear, it violates the rule-of-mixtures considering solute and solvent atoms are not easily discernible. There are unusual physical, mechanical, and chemical properties of the resulting CCAs, which are better than the average constituent elements. This is mainly attributed to the interaction between constituent elements and the indirect impact of these elements.

#### *6.2. Complex Concentrated Alloy Coatings*

Bulk CCAs are expensive due to the high amounts of constituent elements. This is a factor that could hinder massive application and scalability of bulk CCAs for structural and functional applications [1,2,20]. This has necessitated the need for the design of CCAbased coatings to solve the overall cost problem [22]. The superior functional properties of CCAs/HEAs such as wear, irradiation, thermal stability, corrosion, and oxidation resistance are linked to the associated mechanisms and how they differ from different classes of nonconventional coatings. The various approaches to depositing CCAs on various substrates have been reviewed [22]. The fabrication processes are mainly plasma/laser deposition, vapor deposition, and thermal spraying as shown in Figure 13 [22]. These coatings induce surface properties required to improve mechanical, physical, and chemical properties of

engineering components. Reviews focusing on laser and vapor depositions of CCAs/HEAbased alloys are available elsewhere [10,28]. Most of the CCA-based coatings that have been explored using the various techniques on different metallic substrates are given in Table 8.

**Figure 13.** Typical fabrication methods of CCA/HEA coatings [194].

6.2.1. Complex Concentrated Alloy Cold Spray Coatings—An Outlook

Cold spray coatings have been used to deposit CCAs/HEAs on various substrate materials. These cold sprayed CCA/HEA-based coatings have shown excellent particleto-substrate bonding, particle-to-particle bonding, adhesion strength, cohesive strength, relatively low porosity, and thickness of a few millimeters. These properties are essential in enhancing the physical, mechanical, and corrosion properties of these coatings in ways that induce structural integrity.

Trend analyses of publications with the keyword "cold spray + high entropy alloy" on Scopus and Web of Science were merged. A total of 37 peer-reviewed papers were found between 2019 and 2022, and their respective citations are shown in Figure 14. The first paper on cold spray of CrCoFeNiMn on an Al substrate was published in 2019. There is an increase in publications with an astronomical increase in citations since the pioneering works in March and June of 2019. In January 2023, two papers were published on the subject with 10 citations recorded based on the merged data from Scopus and Web of Science. The main papers are given in Table 9. The general theme is on the microstructural characterization of CCA-based cold spray coatings with a few focusing on the deposition of alloyed CCA-based powders on different metallic substrates. This has been successful as cold spray is a great way of depositing metals and intermetallics on various metallic substrates with ease.

**Figure 14.** Yearly distribution of published journal articles indexed in Scopus and Web of Science with the corresponding citations from 2019 to 2022.

#### 6.2.2. Mechanical Properties of Cold Sprayed CCAs on Different Metallic Substrates

Hardening mechanisms resulting from work hardening and grain refinements are associated with cold spray coatings in most conventional and complex concentrated alloys [9,27,55,56,88,89,195,196]. Deformation of the coating particles is due to high energy densities resulting in localized adiabatic shear [88,169,170]. The deformations are functions of particle distribution, morphology, size, and mechanical behavior of particles and processing parameters. The general behavior is also assisted by a peening effect of the coating particles, contributing to the enhanced mechanical properties. When the strain energy release rate is higher than the adhesive energy, propagating defects such as cracks interact with the interface layer, resulting in debonding of the coating. This ultimately leads to delamination, a process of the coating separating from the substrate, which is also due to poor wettability.

Strength and hardness properties: Strength and hardness of coatings are functions of composition of the substrate and coating material [27,105,112,197,198], the coating technique with optimized process parameters [91,199–201], quality, and microstructural features of the coating [28,202–206]. The bombardment of particle feedstock on the substrate induces severe plastic deformation [9,27,55,56,88,89,195,196]. This increases the strength and hardness of the cold sprayed coatings and the predominant strengthening mechanisms are mainly work hardening and grain refinement [88,168–170] as reported for CrMnFe-CoNi [200] and CrFeNiMn [207] feedstock pre-alloyed through atomization. The porosity of the CrMnFeCoNi was observed to be ~0.47%, whereas that of CrFeNiMn was ~3.3%. The relatively low porosity in most CCA-based cold spray particles compared to other thermal spray processes shows the effect of the process on strengthening mechanisms.




**Table 8.** *Cont.*


*Coatings* **2023** , *13*, 538

**Table 8.** *Cont.*

2020

Nanostructured

AlNiCoFeCrTi

high-entropy

 coating performed by cold spray

 [270]


**Table 9.** *Cont.*



The core effects of CCAs, such as high configuration entropy of stabilization and enthalpy of mixing, sluggish diffusion, severe lattice distortion, and the cocktail effect due to interaction between atoms, could be contributing effects to the overall behavior of CCA-based coatings. The comparison of hardness of CCA-based coatings based on various thermal spray approaches is given in Figure 15 [28]. There are discernible trends for the approaches with the HVOF and APS having the highest hardness trends. Factors contributing the high hardness can be attributed to the approach used in producing the feedstock. For most of the HVOF and APS approaches, the feedstock was produced using mechanical alloying and high-pressure mechanical blending coupled with tubular mixing. This resulted in reducing the overall grain sizes to ~10–40 μm. The strengthening mechanisms for these processes were a combination of solid solution strengthening, interstitial strengthening, precipitation hardening, cohesive strengthening, and dispersion strengthening. This observation is further strengthened by the mixed phases associated with methods used in producing the feedstock.

**Figure 15.** Variation of hardness of various CCA-based coatings through various thermal spray approaches [28].

Most of the coating phases for warm spraying, HVOF, and APS processes were mixtures of FCC, BCC, B2, and carbides and oxides of constituent elements. For instance, there is hardness variation between cold sprayed and atmospheric plasma spray coatings. The additional hardness of APS was attributed to additional effects from mixed phases of FCC and BCC, and how these phases interact with one-, two-, and three-dimensional defects across various microstructural length scales. The presence of oxide phases induced by the high-temperature APS is also a contributing factor to the higher hardness

than with the cold spray with predominantly a single phase with no formation of oxides or other precipitates. The comparison of the deposition of Cantor alloy (CrCoFeMnNi) using APS and cold spray showed an interesting trend. The cold sprayed coating had relatively high hardness of ~333 HV compared to the APS technique, which had a hardness of ~273 HV. This is due to dynamic recrystallization at highly deformed regions resulting from the plastic deformation, increased dislocation density coupled with numerous grain boundaries [19,51,84,86,163,187], and their interactions with various microstructural features and defects (Figure 16). The effect of the overarching mechanisms for adhesion and cohesion, which is metallurgical interlocking, was a contributing factor (Figure 17).

**Figure 16.** Schematic representation of different types of microstructural defects.

**Figure 17.** Microstructural characterization of Cantor alloy deposited on Al 6082 substrate showing (**a**) surface morphology and (**b**) cross-sectional view with the mechanical interlocking phenomena. Reprinted/adapted with permission from Ref. [200]. Copyright 2019, Elsevier.

The 6061Al and CrCoFeNi composites designed by using cold spray have showed enhanced strength and hardness after friction stir processing [279]. The replacement of conventional ceramic materials with CCAs is due to poor interface wettability between Al matrix composites reinforced with ceramics [289]. The cold spray process induced geometrically necessary dislocations (GNDs) of 17 × 1015 <sup>m</sup>−<sup>2</sup> with an average grain size of 24 <sup>μ</sup>m. The GNDs were further reduced to 9.8 × 1015 <sup>m</sup>−<sup>2</sup> with average grain size of ~4 μm, resulting in great hardness and strength properties [279]. Although the strength and hardness behavior of CCA-based cold spray coatings are promising, these properties can be enhanced further by optimizing the process parameters of feedstock, cold spray processing parameters, and post-treatment processes.

Cold sprayed CrMnFeNi has been applied by varying the process parameters (feeding rate and gas pressure) with high hardness of ~304 HV0.3 [27]. This value was achieved at gas pressure of 60 bar and feed fate of 1 rpm. The hardness was attributed to work hardening from the severe plastic deformation from the high-impact process of the cold spray technique [19,54,174]. In the case of CrCoFeNi cold spray coatings on AZ91 Mg alloys [284], the hardness was ~315 HV, exceeding conventional Al- and Zn-based cold spray coatings on Mg alloys. The enhanced hardness is due to grain refinement and the high hardness behavior of the CrCoFeNi compared to the Zr and Al alloys [290–297]. Apart from the peening effect resulting in grain refinement of the Mg surfaces, the alpha phase of AZ91 alloy also undergoes dynamic recrystallization contributing to the grain refinement [284].

Fatigue behavior and residual stresses: Fatigue performance of cold spray coatings is enhanced by properties of particles, processing parameters, and metallurgical factors [18]. Excellent fatigue and adhesion strengths were observed for FeCoCrNiMn alloy sprayed on metallic substrate. This was due to dense FeCoCrNiMn coating with optimized process parameters resulting in pancake-like appearance and high flattening ratio and severe grain refinement from the high-energy accelerated particles [18]. The peening effect resulting from the bombardment of the particles on the metallic surface induced high compressive residual stresses [173–175], which led to the high fatigue resistance and the observed adhesion strength [18,172,173]. From the XRD measurement, the maximum compressive residual stress was 200 MPa, about 100 μm from the coating–substrate interface.

Cracks were nucleated at the surface defects [18,172,174]. The first stages of deformation were along the trans-particle paths which were then followed with interparticle and mixed trans-particle fracture mechanisms. The differences in fracture mechanisms are dependent on microstructural features at the coating–substrate and particle–particle interfaces [18,172,174], due to the likelihood of geometric and continuous dynamic recrystallization at these interfaces [170].

The fracture behavior was mainly mixed mode with brittle characteristics during crack propagation and localized fatigue striation characteristics and localized ductile microvoid coalescence resulting from multiple crack initiation sites. The undeformed splat particles and microvoids were stress concentration points accelerating crack propagation rates [172,174].

Corrosion properties: The corrosion and electrochemical behavior of CCA cold spray coatings on various metallic substrates are comparable to their laser cladded or electric arc cladded CCA counterparts. In the case of CrCoFeNi cold spray coatings on AZ91 Mg alloys, an Ecorr of ~−290 mVSCE was recorded [284]. The presence of mixed and multilayer oxides and passive films of FeO, Cr2O3, and NiO retarded the rate of dissolution. This reduced the overall corrosion rate, thus improving the corrosion resistance. After 28 days of immersion, about 1% weight loss was recorded and a localized corrosion mechanism was observed with micropits of ~1 μm on average [284]. The highly uniform and dense microstructure coupled with stable passive oxide films provide the shielding effect needed and contributed to significant improvement in corrosion resistance.

For most coatings derived from Cantor alloys and their derivatives with relatively high Cr, Mo, and Ni contents, the corrosion resistance is comparable and, in some instances, better than conventional ferritic and austenitic stainless steels. There is still substantial work required and investigations to be carried out in this area.

#### 6.2.3. Microstructural Characterization of Cold Sprayed CCAs

The microstructural defects associated with cold spray deposits, as for any other thermal spray coatings, include porosity, voids, delamination, spalling, interface contamination, cracks (transverse and interlamellar), pull-outs, oxide clusters, and metallic inclusions. Details of these defects and brief definitions are given in Table 10. Generally, oxides are associated with metallic-based coating systems. By understanding the nature, types, and geometries of the defects in cold spray coatings, a critical mechanical property, fracture toughness, can be assessed and quantified. Fracture toughness is used to estimate the stresses required to propagate or accelerate pre-existing defects or flaws. Generally, the main mechanism of the propagation of cracks or defects highlighted in Table 10 is by intersplat decohesion or intersplat cracking [28,29,297]. This can be further exacerbated by cracks or pores interlinking, pore compaction, and the sliding of splat. There can also be issues of bifurcation of cracks and secondary crack formation contributing to increased fracture surface area [28,29].

**Table 10.** Typical defects and characteristic features associated with thermal spray processes. Reprinted/adapted with permission from Ref. [29]. Copyright 2023, Taylor and Francis.


#### **7. Areas for Future Research Direction and Implications**

The areas for future research directions should focus on materials, methods, and mechanisms for CCA-based cold spray coating. They are schematically shown in Figure 18. For the materials, focus should be on the 3D transition metal based CCAs and the refractory CCAs. For the methods, the two main cold spray techniques: high-pressure and lowpressure cold spray techniques, should be explored for depositing CCA-based particles on similar and dissimilar metallic materials. The mechanisms of cold spray coatings for structural integrity and failure modes still require some amount of work. Strategic research directions in each thematic area are discussed in the following.

**Figure 18.** Future perspectives for cold sprayed CCA-based coatings for structural and functional applications.

#### *7.1. Materials*

The application of cold spray on dilute metallic alloys as feedstock and substrate has been well studied, although there is still room for improvement. Generally, the process parameters have also been optimized for some systems, especially Cu, Al and Al-based alloys, Ti and Ti-based alloys, Ta and Ta-based alloys, and ferrous alloys. Concerted efforts in simulation-driven cold spray processes have also been carried out, which are thoroughly reviewed in Section 4.

The field of CCAs is less than 30 years old and most of the alloy families show better structural and functional properties than dilute or conventional alloy systems [185,298–300]. As coating material, the classes of CCAs fall under the broad category of 3D transition and refractory metals. When properly developed, these classes of CCAs can be used across the temperature spectrum—from ambient to intermediate to high and ultrahigh temperatures. This will be critical for applications in the thermal barrier coatings (TBCs) for power generation and propulsion [28,29,205,301]. Thus, the CCA-based coatings are best suited for a combination of ambient and high-temperature properties such as strength, oxidation, ductility, thermal stability, and wear resistance. A proof of concept has been carried out using NiCoCrAlSi CCAs through the high-velocity oxy-fuel and air plasma spray (APS) processes as a metallic bond coating for TBCs [205].

Extensive research in the literature showed the elements that are frequently being used for various types of thermal spray applications. They are given in Figure 19 with Cu, Co, Ni, Al, and Fe being the most frequently used elements. Although most of the elements are 3D transition metals, two of the refractory base metals were also used, which could be attributed to the high-temperature oxidation behavior. Figure 19 shows less than a tenth of the stable metallic elements on the periodic table that could easily be explored for next generation coating materials. This is one of the identified gaps which requires urgent attention and could contribute significantly to understanding the emerging fields of low-cost CCA-based coatings, where the concept of scrap and nature-mixing pre-alloy design concepts can be used.

**Figure 19.** Frequency of use of elements on the periodic table for CCA-based thermal spray coatings as of 2020 [28].

The design and development of new BMG alloys and cold spray coatings with tailored properties are needed for various structural and functional properties. These properties, such as improved mechanical strength, corrosion resistance, and biocompatibility, are critical to produce robust and next generation bulk metallic glass coatings in a costeffective manner.

#### *7.2. Methods*

The two types for cold spray coating are given in Figure 4. These methods need further investigation for depositing pre-alloyed cold spray coating feedstock on various similar and dissimilar substrates of CCAs. Currently, the high-pressure technique is mostly used for most of the coating processes with varying process parameters. Although the results for some classes of CCAs are impressive, the application of cold spray to CCAs is still in its infancy since the first process on CrCoFeMnNi was carried out in 2019 [200]. The search suggests less than 35 research-based articles have been published from 2019 to January 2023. The majority used the high-pressure cold spray approach to deposit the coatings.

Pre-processing of pre-alloyed CCA powders is one area requiring immediate attention. The current powder metallurgy routes such as gas and water atomization and mechanical alloying coupled with spheroidization lead to challenges with oxidation. Approaches such as the ultrasonic atomization process are being explored but could be quite expensive and not easily scalable. Other factors hindering scalability have been highlighted [2,20,185].

Full-scale and detailed mechanical and microstructural characterizations of the particle–substrate and particle–particle interactions need to be ascertained. This will contribute to understanding the mechanism associated with the bond strength. For mechanical characterization, little has been carried out. Therefore, techniques such as tensile adhesion, stud-pull, pin and ring, shear load, peel, laser shock adhesion, scratch, and interfacial indentation tests at the nano- and microscale will be necessary. By conducting thorough mechanical and microstructural assessments, deeper understanding of the physical metallurgy and the mechanisms of these cold spray coatings and how to use them in industrial applications can be implemented effectively.

Porosity is one of the main defects associated with thermal spray coating processes. There is the need to understand the effect of porosity on structural integrity of CCA-based cold spray coatings. Thus, various porosity measurement techniques such as mercury intrusion porosimetry (MIP), gas (He) pycnometry, water absorption or Archimedes (immersion) method, X-ray computed microtomography (CMT), ultrasmall-angle X-ray scattering (US-AXS), and electrochemical tests need to be established for the coatings. The working principles of these techniques are reviewed elsewhere [29]. Similarly, measurement of residual stresses using conventional non-destructive methods such as X-ray, synchrotron, laser excitation, and neutron diffraction techniques should also be explored. The implications of these results will contribute to explaining the role of defects in overall structural integrity and overarching mechanisms.

Computational and numerical modeling approaches to studying various aspects of CCA-based cold spray coatings require further attention. Most of these approaches have been adequately discussed for conventional alloys and the highlightes are summarized in Section 4. These simulations and physic-based systems can be adapted for CCAs by revisiting the underlying assumptions. By using these computational approaches synergistically with experiments, the concept of rational development of coatings is more guided leveraging from the merits of both approaches. This builds into the material innovation infrastructure where computations, data science and experimental tools are harnessed to design alloys and coatings within acceptable timelines reducing the material lifecycle (concepts to scalable material or coating) significantly.

#### *7.3. Mechanisms*

The main underlying mechanisms of cold spray from microstructural perspectives have been metallurgical interlocking and adiabatic heating. This has been shown for most of the dilute or conventional principal element-based alloys. Similar results have been observed for some CCAs, especially the Cantor alloys and their derivatives. This was expected as the compositions are close to highly alloyed stainless steels, Ni-based superalloys, and Ti-based alloys. However, there is the need to investigate the mechanism of bonding and failure modes as these mechanisms, especially solid-state deformation of particles and bonding to the substrate and each other, are yet to be fully understood. The effects and contributions of various microstructural defects are listed in Figure 14 and Table 9.

The interaction between the cocktail of constituent elements and defects such as stacking faults and dislocations and the effects on bonding mechanisms needs further attention. This will necessitate and drive the next frontiers of cold spray coatings to tailor the coating and substrate microstructures.

There is the need to focus research on process optimization for these classes of coatings. This will require understanding of the process parameters such as spray distance, gas pressure, and powder feed rate. The optimized process parameters can lead to improvements in coating properties, such as density, thickness, and adhesion strength. The coating properties can be studied using advanced mechanical and microstructural characterization techniques to establish the relevant nano- and micromechanisms for better understanding of the performance of BMG and CCA cold spray coatings.

Large-scale applications of these coatings are also areas for further research. Issues that could hinder scalability should be investigated and studied for applications in biomedical, aerospace, and automotive industries. Scaling up is relevant to exploring factors that could have been missed during miniature-sized coating fabrications. Translating laboratory-scale curiosity and science to large-scale industrial applications is critical for BMG and CCA cold spray coatings. This will ensure that issues around standardization of coating systems and mechanisms are investigated. This will allow for consistency and uniformity while providing new insights for further research.

#### **8. Summary and Concluding Remarks**

A comprehensive overview of cold spray technology as applied to the emerging fields of complex concentrated alloys is presented. Based on the literature and discussions, the summary and concluding remarks are as follows:


resulting from the bombardment of the feedstock particles onto the substrate. These mechanisms are different from the precipitation hardening, solid solution strengthening, dislocation strengthening, and oxide formation associated with other forms of thermal spray coating techniques. The deformation mechanisms, which are dependent on the energy density, result in localized adiabatic shear as a function of properties of the particles and the process parameters.


**Author Contributions:** Conceptualization, D.K., M.B., J.v.d.M., N.R. and W.S., methodology, D.K., T.A., T.B., M.V. and M.B.; formal analysis, D.K., T.A., M.V. and M.B.; investigation, D.K., T.A., M.V. and M.B.; resources, M.B., J.v.d.M., N.R. and W.S.; writing—original draft preparation, D.K., T.A., T.B., M.B. and M.V.; writing—review and editing, D.K., M.B., J.v.d.M., N.R. and W.S.; visualization, D.K., T.A., T.B. and M.V.; supervision, J.v.d.M., N.R. and W.S.; project administration, N.R. and W.S.; funding acquisition, N.R. and W.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the Army Research Laboratory (ARL) Grant No. W911NF–19–2– 0108. The contributions from the Worcester Polytechnic Institute Global fellowship is also acknowledged.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
