A Review on Corrosion and Wear of Additively Manufactured Alloys

Additive manufacturing (AM) refers to the manufacturing of three-dimensional (3D) objects or parts using a layer-by-layer addition of constituent materials often with the aid of computer software [1–3]. Some notable techniques in additively manufactured (AMed) processes include the laminated objective manufacturing (LOM), 3D printing, fused deposition modeling (FDM), stereolithography (SLA), laser metal deposition (LMD), and selective laser sintering (SLS) [4–12]. With AM, an enhanced range of surface finish and characteristics can be obtained on material surfaces that are not possible with conventional manufacturing processes [13,14]. As a result, AM is gaining rapid popularity as a preferred manufacturing process (Fig. 1). AMed materials find several applications such as making machine parts, construction materials, dentistry, and biomedical implants [15–18].

There are many challenges and outstanding research questions pertaining to the failure of AMed products [2,19]. With increased layer resolution, the production time increases significantly in AM [1]. Deficiencies in the interlayer bonding can lead to anisotropy in the mechanical properties of AMed materials. Moreover, many existing AMed processes only have a single material processing capability at a time [2]. Thus, without a thorough understanding of failure types, causes, related mechanisms, and their analysis, neither a strong theoretical framework nor a practical guideline for development is possible [3]. Under such a scenario, continued usage of AM could prove highly detrimental to the product life cycle, cost of manufacturing, product safety, and consumer satisfaction. In this context, there is an impending need to address the problems of failure of AMed alloys. In particular, failure analysis due to mechanical and electrochemical conditions needs a great deal of attention.

Additively manufactured alloys can fall into either bulk or surface. Bulk failure includes fracture, fatigue, creep, deformation, buckling, thermal shock, and stress cracking [4–10]. Failure from the surface can be corrosion or wear. Corrosion is one of the prime means of failure of AMed alloys [20,21]. Wegner et al. reported that a high surface roughness of AMed alloys, induced mainly due to the processing conditions, leads to accelerated corrosion rates locally [3]. Multiple local corrosion sites, in turn, result in crack initiation leading to a drastic reduction in the alloy’s fatigue strength. Defects arise during the manufacturing process or during the post-processing of alloys which make the materials vulnerable to corrosion. Recent research revealed insights from the corrosion behaviors in stainless steel, titanium alloys, aluminum alloys, and other alloys [11–15]. Galvanic corrosion can be used beneficially for energy harvesting [16]. It is also important to determine and optimize several heat treatment processes that impact the microstructural characteristics of alloys and consequently affect their corrosion response [17,18,20–25]. AMed alloys for use as biomaterials have special corrosion properties [22,25–27]. Such materials are typically used to support the function of biological tissue and find applications as implants in the human body. Major types of AMed alloys as biomaterials are those of stainless steel, cobalt-chromium, and titanium alloys [26,27]. The biocompatibility of such materials is a major issue that needs addressing [28–30].

The wear of AMed alloys is an important and critical aspect to investigate from a reliability standpoint [28]. Several factors like material microstructure, surface finish, wear type, and mechanisms involved are at interplay in AMed alloy wear [29]. The major wear types in AMed alloys are adhesive and abrasive wear. While abrasive wear is primarily related to contact type and hardness of surfaces in contact, adhesive wear depends on material microstructure in addition to harness. Poor surface finish can result in high surface roughness of AMed alloys which eventually lead to accelerated wear and a high lubricant demand. Some AMed alloys have an inherent microstructure that hinders the wear of the material. It is seen that metal compositing alters the microstructure and, thus, imparts strong anti-wear characteristics. In light of the significance of wear and its complexity, there is an impending need to understand various anti-wear properties specific to AMed alloy materials.

In this paper, corrosion and wear as a surface failure of AMed alloys have been identified and reviewed. Section 2 deals with the corrosion behavior and characterization techniques. The wear of AMed alloys has been included in Sec. 3. A section on the corrosion behavior of an increasingly important class of AMed alloys for biomaterials has also been included (Sec. 4). Recent research data from over 100 articles have been used to analyze the corrosion and wear behavior of AMed alloys. Thus, this work presents useful insights into the surface failure modes of AMed alloys, outstanding issues, and future research direction.

Many factors and parameters in AM have profound effects on corrosion behavior. The factors most often found to influence corrosion behavior are laser power and laser scanning speed for laser-based AM systems, as well as post-processing such as heat treatment and surface polishing [12,31–33]. These parameters influence many of the resultant alloy’s properties including microstructure, porosity, oxide layer thickness, and corrosion resistance, all of which impact corrosion behavior. The effect and extent of these properties on corrosion behavior largely depend on the material fabricated. Common alloys produced by AM are stainless steel alloys such as 304L and 316L stainless steel, titanium alloys such as Ti–6Al–4V, and aluminum alloys such as Al–Si10–Mg and AA2024, along with other more novel alloys [34]. These alloys have a wide range of applications due to their attractive mechanical properties and corrosion performances, which are often very different and sometimes superior to counterparts fabricated by traditional means. Table 1 displays the literature discussed in this review of corrosion behavior.

As shown in Table 1, the effects of heat treatment and surface finish on corrosion behavior were two of the most studied corrosion-related topics in recent times. Both heavily affect the corrosion behavior of the alloys and can either increase or decrease corrosion resistance depending on the process parameters. Heat treatments change microstructural features which corrosion behavior is largely based upon, while surface finishes can remove or create surface defects ideal for pitting initiation.

Stainless steels are generally very corrosion resistant due to a passivating oxide layer that forms Cr₂O₃ as a result of the chromium in stainless steel bonding with oxygen at the surface [66]. This chromium oxide layer is not electrically conductive, providing a physical barrier between the stainless steel and surrounding electrolyte, effectively preventing corrosion in many environments. However, uneven element distribution such as dendritic microstructure can occur in stainless steels which, if exposed to corrosive media, may experience pitting corrosion [35,67]. Additionally, the large thermal gradients and rapid solidification inherent during the AM process provide an ideal environment for elemental segregation and uneven distribution to occur. Austenitic stainless steels can also experience stress corrosion cracking (SCC) [10] and hydrogen embrittlement (HE) [39] in certain environments. In addition, both AM parameters and AM fabrication method used can produce varying corrosion resistances [38]. As such, some published results on corrosion of AMed alloys are contradictory.

Of primary study in stainless steels fabricated by AM is alloy microstructure in relation to the alloy’s properties. Stainless steels have a wide range of microstructures and a corresponding wide range of properties. AMed stainless steels often contain extremely irregular microstructures, such as a fine columnar sub-grain microstructure inside each grain [68]. Ganesh et al. associated the microstructure in 316L stainless steel produced by plasma transferred arc welding (PTAW) AM to the alloy’s pitting resistance [35]. Specifically, the authors found primary α-ferrite regions to be targeted for pitting corrosion, and that the sensitization caused by the AM process caused continuous corrosion along the grain boundaries. Kong et al. showed 316L stainless steel produced by selective laser melting (SLM) exhibited a thicker oxide film with lower Mo content than the equivalent traditionally quenched alloy [32]. These samples also displayed full γ-austenite-containing columnar sub-grains similar to what was previously mentioned [68], with grain size increasing with increasing laser power.

The stainless steel 316L fabricated by laser powder bed fusion (LPBF) can experience SCC in certain corrosive environments [10,37]. In cases of high-temperature water (288 °C), AMed 316L stainless steel with an anisotropic and non-equilibrium microstructure after stress relieving preferentially cracked in the build direction. SCC performance rivaled that of wrought 316L stainless steel with appropriate heat treatment. However, even fully annealed AMed 316L stainless steel experienced transgranular cracking. It was also found that oxides rich with Si at grain boundaries were preferentially dissolved, creating crack initiation sites and contributing to widespread crack branching.

The microstructure is heavily influenced by the heat treatment of 316L stainless steel fabricated by arc AM as Fig. 2 shows [36]. As-fabricated 316L stainless steel contains extremely fine vermicular δ and σ phases with the γ matrix as shown in Fig. 2(a). However, the σ phase disappears at heat treatments of 1100 °C and above, and the δ phase transforms into a granular morphology followed by complete dissolution with high enough heat treatment and time, creating a fully austenitic microstructure. From potentiodynamic polarization experiments, i_corr decreases and pitting potential increases with increasing heat treatment temperature and time, showing heat treatment indeed increases corrosion performance. This trend is primarily due to the depletion of the Cr-rich σ phase, since the boundary between the σ phase and the γ matrix is Cr-depleted, making the interface susceptible to pitting initiation.

Stainless steels 304 and 304L are also frequently studied due to their corrosion resistance. These stainless steels are similar to 316L in that they are both austenitic and have similar elemental compositions, although 304L does not have Mo like 316L has. They have similar relationships between laser power during AM and pitting potential in that increasing laser power increases pitting potential [32,69]. Increased α-ferrite content, however, decreases pitting potential in 304L stainless steel [69]. According to Baek et al., the direct metal tooling (DMT) process improves resistance of 304L stainless steel to HE in a gaseous hydrogen environment over that of traditionally rolled 304L stainless steel [39]. The authors speculated this was due to the microstructural textural differences between the fabrication techniques. Additionally, strain-induced austenite-to-martensite transformation in the hydrogen environment did not take place in the AMed alloy even though it did occur in the rolled alloy. The austenite’s stability under strain might also affect the alloy’s resistance to HE.

Stainless steels 304L and 316L fabricated by directed energy deposition (DED) share many microstructural characteristics with the same fabricated by SLM such as lack of fusion (LOF), gas porosity, and distributed oxide nano inclusions [41]. However, slower cooling of the DED method over SLM results in coarser grains and increased intracellular δ-ferrite, with increased printing power creating greater δ-ferrite content. This increased intracellular δ-ferrite content contributed to decreased corrosion resistance. Previous studies showed decreased oxide size because the SLM process increased breakdown potential (E_b) beyond that of wrought 304L stainless steel [12,40]. However, the authors found that this was not the case for the DED 304L stainless steel due to large amounts of δ-ferrite. The authors also found dissolution of the γ-austenite around finely distributed δ-ferrite, which created web-like covers over pits and nest-like pit walls.

AM of titanium alloys is highly attractive due to titanium’s excellent mechanical performance, corrosion properties, biocompatibility, and low machinability using traditional fabrication techniques. Titanium has an extremely high affinity for oxygen, thus when exposed to an environment containing oxygen, it oxidizes to form a TiO₂ oxide layer which is extremely resistant to corrosion in a similar manner to the oxide layer of stainless steel [70]. Due to its outstanding properties, AMed titanium is appealing to the aerospace and biomedical industries among others. However, the large thermal gradients and rapid solidification inherent to AM affect the alloy’s microstructure and thus corrosion performance.

The titanium alloy Ti–6Al–4V is the most common titanium alloy used in AM due to its widespread use in industrial applications using traditional fabrication methods, making up over 50% of all titanium alloys in the industry by weight [70]. However, Ti–6Al–4V produced by wire arc additive manufacturing (WAAM) has been shown to have anisotropic corrosion resistance in certain environments [42,44,47]. The cooling rate in the vertical direction is slower than that in the horizontal direction. This forms large, hard, lamellar grains of the α phase intertwined with a Widmanstätten pattern, increasing corrosion resistance in the vertical direction relative to the horizontal plane. Additionally, more acicular α’-martensite formed in the planes perpendicular to the build surface which corresponded to a lower corrosion rate. The authors also found heat treatment at 2 h decreased corrosion resistance, with greater temperatures corresponding to lower corrosion resistance. While the α’ phase was reduced with heat treatment, the α phase increased, which more heavily impacted corrosion resistance [44,45]. The difference in phases determined the diffusion mode and speed during corrosion due to their differing microstructures [71]. Compared to wrought Ti–6Al–4V, the WAAM counterpart showed only slightly decreased corrosion resistance. It was concluded that slower cooling decreases the α’ phase, which in turn decreases the corrosion rate. Interestingly, it was also found that the corrosion rate was higher in the vertical plane than in the horizontal plane when in an HCl solution due to increased α’ and decreased β’ phase in the vertical plane [42]. Meanwhile, Chandramohan et al. showed heat treatment decreased corrosion rate in H₂SO₄ and HCl environments but increased in an NaCl environment, indicating that solution has a strong effect on corrosion behavior as well [45].

Similar to stainless steel produced by AM, the AM process greatly affects the resultant microstructure and thus corrosion resistance of titanium alloys. Yang et al. compared in a NaCl solution the corrosion performance of SLM, SLM plus heat treatment, WAAM, and traditional rolling [46]. Two different heat treatments were studied: heat treatment 1 (HT1) was kept at 750 °C for 2 h, and heat treatment 2 (HT2) was kept at 1020 °C for 2 h. Corrosion resistance was highest for SLM + HT1, followed by traditional rolling, WAAM, SLM + HT2, and SLM, respectively. The authors also determined the type of constituent phase was the strongest influence on corrosion performance among microstructural factors, with grain size second and morphology third. Figure 3 shows the varying microstructures present in different fabrication/post-processing techniques. The authors reported the same acicular α’-martensite phase, located in this case inside columnar prior α grains in the SLM samples. With HT1 this α’-martensite decomposed into a fine mixture of lamellar α and β inside those columnar grains. With SLM + HT2, the columnar grains vanished and instead a coarse lamellar α + β structure appeared. In the WAAM samples, an α + β matrix with fine α lamellae was present, and the traditionally rolled samples yielded a typical bimodal structure. Previous studies showed β > α > α’ in terms of corrosion resistance in Ti–6Al–4V [42,43]. This explains the ranking of the corrosion resistance of the different processes, in addition to the effects of grain size and, to some extent, morphology.

Liang et al. studied the effects of high-temperature environments on the corrosion performance of Ti–6Al–4V fabricated by selective laser sintering (SLS) [48]. Both air and air–SO₂ environments were tested at 600 °C, and the AMed samples were compared to ones fabricated by traditional cold-rolling. X-Ray diffraction (XRD) analysis showed the corrosion products of the two fabrication methods were similar. The only change found was additional Al₂O₃ in the surface corrosion products of the SLS alloy in the air–SO₂ environment. Corrosion products were thicker on the AMed samples than the cold-rolled samples, and they were much thicker on samples of both fabrication methods in the air–SO₂ environment than in air. The sulfur species were more permeable, increasing corrosion. The cold-rolled titanium was comprised of its typical bimodal α and β structures, while the SLS titanium mainly consisted of the α’ martensite phase, corresponding to decreased corrosion resistance in the SLS titanium in concordance with Yang. et al.’s conclusions [46].

Additive manufacturing has the advantage of fabricating alloys with added elements in order to increase corrosion resistance. Meenashisundaram et al. have made some recent progress on the development of another titanium alloy fabricated by AM [49]. They recently developed Ti + Mg composites utilizing inkjet 3D printing to produce a titanium volume with 41.33% porosity and then filled the pores with Mg using a pressureless, capillary-mediated infiltration technique. After 1 h, the composite exhibited a much higher corrosion rate (937.48 µm/year) than the porous titanium on its own (137.79 µm/year) due primarily to the addition of Mg. However, even a corrosion rate of 1 mm/year is not high compared to that of other metal/Mg composites [72,73]. Mg is also biodegradable, and thus, this composite is ideal for orthopedic implants. Not only does its addition adjust pure titanium’s properties to approach that of bone but it also allows for Mg degradation over time with the resulting structure having good osteointegration [73–77]. Thus, the increased corrosion rate of the Ti + Mg composite is advantageous for such applications.

Aluminum alloys are also commonly used in AM due to their great corrosion performances and good strength-to-weight ratios. Aluminum’s corrosion resistance comes from its extremely high tendency to oxidize and form an Al₂O₃ oxide layer. Aluminum alloys containing Si and Mg are the largest group of all shaped cast parts and are used in many industries [78]. The addition of Si and Mg as major alloying elements creates a nearly eutectic aluminum alloy where solidification occurs concurrently for all phases, which is ideal for the AM industry [79,80], although anisotropic corrosion behavior still occurs due to irregular grain morphology [50]. The most common aluminum alloy used in AM is AlSi10Mg, which has good solidification shrinkage, casting, welding, hardening, and corrosion properties [52,55,81]. The mechanical properties of AlSi10Mg can be improved by AM relative to traditional casting, adding to its attractiveness [82–84]. Besides AlSi10Mg, AA2024 (Al–Cu–Mg) is also a popular aluminum alloy for AM [61,85]. It is known to have good fatigue resistance but lacking corrosion performance, although not much research on its corrosion behavior has been published.

Microstructures of aluminum alloys fabricated by AM often vary from traditional techniques, which has an obvious effect on corrosion performance, similar to other alloys fabricated by AM [52,54]. For AlSi10Mg, the grains are much finer than its die-cast counterpart [52]. This fine microstructure in combination with a consistent, fine Si particle distribution with no intermetallic was associated with the AM process. This along with the decreased Fe and Cu content was the reason for the increased corrosion performance. Rafieazad et al. found the microstructure to vary depending on heat treatment for samples fabricated by direct laser metal sintering (DMLS) [54]. As-fabricated samples and samples heat-treated at 200 °C for 3 h displayed fibrous intercellular Si eutectic phase networks in an α–Al matrix. However, increasing the heat treatment temperature to 300 °C disrupted the Si network, and at 350 °C, the Si particles combined and coarsened with heavily precipitated Si particles. Figure 4 displays a representative model of these changes along with scanning electron microscopic (SEM) images of the surface. These microstructural changes resulted in increased corrosion performance when increasing heat treatment temperature from 200 °C to 300 °C and then decreased corrosion performance when increasing from 300 °C to 350 °C. Zakay and Aghion showed similar results in samples fabricated by SLM [56], although Cabrini et al.’s research concluded LPBF parts exhibited an opposite pattern in terms of intergranular corrosion [53]. Namely, as-fabricated samples showed optimal corrosion performance, with performance decreasing for heat treatments up to 300 °C for 2 h and performance increasing when increasing heat treatment temperature from 300 °C to 500 °C for 2 h. These differences in corrosion performance likely resulted from the different corrosion tests performed: potentiodynamic polarization experiments [54,56] and salt spray tests [56] versus 24-hour submersion intergranular corrosion tests [53]. Additionally, Rafieazad et al. and Zakay and Aghion focused on general corrosion behavior while Cabrini et al. studied only intergranular corrosion. Another paper published by Cabrini et al. employed a statistical approach to evaluate the build direction and heat treatment temperature’s relationship with the corrosion behavior of AlSi10Mg fabricated by LPBF [57]. Interestingly, this paper contradicts the rest of the work mentioned herein that it found all heat treatments to be detrimental to corrosion performance with increasing temperature producing more detrimental effects and that build direction did not influence corrosion behavior. As such, the results of this study are somewhat questionable, and a larger number of samples and tests may have been necessary to produce results more in line with the rest of the literature.

Besides heat treatment, the surface finish also has an impact on corrosion performance. Fathi et al. found sandblasting harmed corrosion performance for both DMLS-AlSi10Mg heat-treated alloy and die-cast A360 alloy. Meanwhile, the corrosion performance of grounded DMLS-AlSi10Mg surface varied with immersion time in corrosive media [58]. The ground surface initially experienced the highest corrosion current density for DMLS-AlSi10Mg. However, after 24 h of immersion time, the same surface experienced the lowest corrosion current density of any surface finish due to the development of a stable, dense, and thick passive film. The DMLS-AlSi10Mg also exhibited increased corrosion performance over die-cast A360 alloy with the same surface finishes. Fathi et al. also found increasing DMLS volumetric energy density reduced surface roughness, although the corrosion performance was related to the resultant microstructure rather than the surface finish [59]. Leon and Aghion observed increased corrosion resistance and low-cycle corrosion fatigue endurance for polished surfaces of AlSi10Mg samples produced by SLM relative to as-fabricated samples [51]. As-fabricated surfaces contained many inherent defects which acted as pit and crack initiation sites, although polishing removed these defects. Xing et al.’s research found ultrasonic peening treatments applied to the surface of AlSi10Mg samples fabricated by SLM increased the alloy’s resistance to SCC, in terms of both decreased attack severity and decreased corrosion rate [60]. This corrosion performance increase was related to increased surface hardness, conversion of surface stress from tensile to compressive stress, and reduced porosity.

Another common aluminum alloy used in AM is AA2024 that contains Cu and Mg as primary alloying elements. These alloying elements are essential to achieve the high mechanical properties of the alloy, but their addition creates a highly heterogeneous alloy that is very susceptible to localized corrosion [61,86–93]. However, localized corrosion of AA2024 fabricated by AM (AM2024) is distinct from wrought AA2024 largely due to the resultant microstructure of AM2024 [61]. These differences include the absence of micron-sized constituent particles in AM2024 which are present in AA2024 and the major second phase being θ-phase precipitation rather than S-phase in AM2024. As a result, AM2024 has increased pitting resistance and decreased corrosion rate relative to AA2024 due to the absence of the S-phase and elimination of micron-sized cathodic sites which develop in AA2024. However, even with the improvement of AM2024 overwrought AA2024, the alloy is still more susceptible to corrosion than other Al alloys such as AlSi10Mg. Cu is added to AA2024 for solution strengthening, but its addition increases the alloy’s susceptibility to pitting corrosion [94]. As such, the corrosion behavior of AM2024 is not a heavily researched topic.

While stainless steel, titanium, and aluminum alloys are the most common alloys in AM, other alloys such as nickel, iron, and copper metals and alloys have been studied in recent years. AMed alloys containing Ni such as Inconel 718 [62] and NiTi alloy [63] and their corrosion behavior have recently been researched. Ni et al. analyzed the effect of varying Nb content in LMD Inconel 718 on its corrosive behavior [62]. Increasing the Nb content from 1 wt% to 8 wt% changed grain morphology from columnar grains aligned with the build direction to refined and less columnar grains since Nb provided more nucleation sites. Carbide precipitates which acted as pit initiation sites were eliminated with increased Nb content due to the high affinity between Nb and C. The combination of these increased both corrosion resistance and pitting potential. Besides alloying elements, laser power and laser scan speed during fabrication also affect the corrosion behavior of Ni alloys, similar to other alloys. NiTi alloy’s corrosion performance produced by laser engineering net shaping (LENS) AM increased with increasing laser power and decreasing scanning speed [63]. Ni alloys fabricated by AM generally have a mixture of attractive mechanical properties and strong corrosion performance. As such, they will likely be researched more in the future.

Fe–Mn parts fabricated by AM have been studied for use in biomedical applications such as bone scaffolds due to the biodegradable nature of Fe with Mn increasing the degradation rate of Fe. This enables the design of temporary bone scaffold implants. Shuai et al. fabricated porous Fe–Mn bone scaffolds by SLM with sufficient mechanical properties for biomedical use [64]. The corrosion rate can be modified with the addition of Mn so that it does not corrode too fast (which causes premature failure) or too slow (which inhibits bone tissue growth). This is caused by the creation of micro-galvanic sites with the addition of Mn which creates a biphasic composition of γ and ɛ phases [95]. Zr-based bulk metallic glasses (BMG) fabricated by SLM is another topic of a recent study for biomedical applications [65,96]. The corrosion resistance of this SLM Zr-based BMG was higher than cast Zr-based BMG and Ti–6A–l4V in simulated body fluid in terms of both corrosion rate and passive current density. However, the authors failed to study the reasons for this corrosion performance. Its performance over Ti–6Al–4V may be due to a lack of grain boundaries which are known to generally be targeted for the selective corrosive attack [97]. AM also enables complex and highly porous shapes that lower the elastic modulus to replicate that of bone, which helps eliminate stress shielding and stabilizes bone-implant connection.

Other metals and alloys fabricated by AM such as highly pure copper [98], Ni–Al bronze [99], and Zi–Al matrix composites [100] have been researched in recent years, although their corrosion behavior has not been analyzed and as such is a topic for future study.

While corrosion is very often harmful, it can also be applied in beneficial ways such as using galvanic corrosion to power a battery or using sacrificial anodes to prevent other parts from corroding. As such, it is worth mentioning some new applications of corrosion in AM. One such application is to corrode away support structures after a part is fabricated [101,102]. Overhangs with large angles from the vertical plane can typically not be achieved using general AM techniques, and even with advanced techniques, the angles may only reach 45–60 deg [103]. However, Hildreth et al. accomplished 90-deg overhangs using a support structure that was corroded away after fabrication in a sacrificial anode fashion [101]. An AISI-type 431 stainless steel bridge was fabricated with Metco 91 (mild carbon steel) as the support structure during fabrication. Nitric acid was used to efficiently dissolve supports with O₂ applied to accelerate the process. About 3.8 × 4.6 × 10 mm of carbon steel was removed with minimal removal of the stainless steel. Lefky et al. used sensitizing heat treatments and subsequent etching on LPBF Inconel 718 to remove all support structures [102]. Sensitization reduced corrosion resistance approximately 100–200 µm deep into all surfaces, making the removal process self-limiting which prevented damage to the part. The etching was performed using acid at an anodic potential which strongly targeted the sensitized regions over the unaffected material. While innovative applications of corrosion are not a topic of many studies currently, it may become popular with the continual development of new methods and precision in technology.

Electrochemical testing has been employed to characterize the corrosive performance of AMed alloys. It is an inexpensive and relatively quick method based on the charge-transfer process in an electrochemically controlled environment [104–108]. The various types of electrochemical corrosion tests are shown in Fig. 5. Electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization (PP) are the two major electrochemical tests performed to study and analyze the corrosion characteristics of AMed alloys.

In the current literature, the corrosion performance of AMed alloys does not have conclusive results. For instance, in a study by Sun et al., SLM 316L specimens were tested, and higher than normal passive current density was observed in a 0.9% NaCl electrolyte solution [109]. But in another study, for specimens prepared by the Laser Engineered Net Shaping (LENS) method, a lower passive current density was reported in the same electrolyte [110]. Hence, a slight variation in manufacturing methods and/or parameters can cause a significant change in corrosion performance. Sander et al. tested AMed multiple specimens of 316L using SLM by deliberately varying the laser scan speed and power. Pitting potential was observed to be 300 mV more for the 3D printed specimen compared to the wrought 316L specimen, indicating a higher corrosion resistance [12]. Kong et al. compared five samples of SLM 316L produced at different laser powers and a quenched sample, immersed in a simulated body fluid (SBF) solution [32]. Initially, the corrosion and passive current densities were similar for all six samples. However, during 288 h of immersion, considerable variations in pitting potential were recorded before the typical pitting potential was reached. The pitting potential increased with the laser beam power. The sample produced with the highest laser beam power (220W) displayed a higher pitting resistance compared to the quenching sample. It was hypothesized that in 3D printed specimens, the presence of impurities (MnS, Al-oxides, etc.) was influenced by SLM. They are either too fine or annihilated and unable to reform during the rapid solidification. Lodhi et al. prepared 316L stainless steel via SLM (AM 316L SS) and compared it to a wrought stainless steel sample to study its corrosive properties [111]. The two samples were tested in three different electrolytes: phosphate buffer saline (PBS), 0.9 M NaCl, and human serum. From the EIS test, the oxide film formed on the AM 316L SS sample was observed to be more compact, relatively less defective, and showed better stability as compared to the wrought sample (Figs. 6(a)–6(d)). They proposed that fine grain structure increases the corrosion resistance of stainless steel and enables rapid diffusion at the electrode–electrolyte interface. These results were corroborated by a cyclic potentiodynamic polarization test. The breakdown potential obtained for AM 316L SS was higher than that of wrought. The passive region did not show any fluctuations in current.

Titanium alloys behave similarly during electrochemical testing. SLM fabrication of titanium alloys expands on the corrosion-resistant properties shared between several titanium alloys that are procured through AM methods [42,44,45,112]. In a study done by Damborenea et al., the performance of SLM produced Ti–6Al–4V was compared with a wrought Ti–6Al–4V [113]. It was immersed in a phosphate-buffered solution (PBS) and tested. A sharp upsurge of current density was observed only for SLM Ti–6Al–4V. They concluded that the SLM Ti–6Al–4V surface was more vulnerable to the hydrated Ti⁺ ions formation.

With the intention of medicinal implant applications, Assis et al. investigated the corrosion performance of Ti–6Al–4V and Ti–6Al–7Nb alloys [114]. Potentiodynamic polarization curves were taken after immersion in Hank’s solution (72 h @ 37 °C). The scanning rate for the polarization test mentioned was 1 mV s⁻¹. Substantially, low current densities were found during potentiodynamic polarization testing, the lowest average being Ti–6Al–4V, with approximately 19 nA/cm². These alloys exhibited a high corrosion resistance, which was credited to the development of a bi-layered oxide film around the alloys. This film was found to shield the material from the corrosive activity on the inner layer but promote osseointegration with its porous structure on the outer layer. Further expanding on biomedical uses, a study by N.T.C. and Oliveira and Guastaldi aligns with the production of an oxide film as mentioned earlier that occurs during the electrochemical testing process [115]. Interestingly, the protective oxide film gradually developed during the test for each titanium alloy sample along with a high reading of impedance. Materials investigated were Ti–6Al–4V and Ti–13Nb–13Zr with either pure titanium or a variation of Ti–Mo which had a concentration of 6, 10, 15, or 20 wt%. The formation of TiO₂ film improved corrosion resistance not just in external environments but against internal human body conditions that commonly reduce the performance of the material in both electrostatic and mechanical functions. Potentiodynamic polarization testing indicates that a Ti–Mo of 6–20Mo wt% alloy presented greater corrosion resistance than a pure Titanium alloy in 35% HCl. It was observed that Ti–Mo generated optimal protection from the corrosive effects of a chloride solution, but the protective qualities attributed to Mo peaked at 15 wt% (Ti–15Mo). The samples shared a valve metal behavior with the proper anodic formation of porous oxides, all without pitting corrosion. Results were established by 360 total hours of immersion to Ringer physiological solution followed by a 1 mV s⁻¹ potentiodynamic polarization scan with a potential range of −8.0 to 5.0 V (Figs. 7(a) and 7(b)). Corrosive effects only became prevalent at the 5.0 V, and minimal values of passive current density for the materials lasted a significant duration of testing.

Corrosion studies of AMed aluminum alloys have generally been restricted to the Al–Si type of alloys. Revilla et al. studied the corrosion behavior of AM ALSi10Mg alloys manufactured via SLM with varying parameters [116]. Four samples were made from recycled powder, one was made from new powder. The reference material used was a cast Al alloy (AA4420). The samples were immersed in a solution of 0.1 M NaCl and tested. The values of corrosion potential obtained from the potentiodynamic polarization test were similar for all samples in the range of −0.6 V to 0.7 V versus Ag/AgCl. The authors hypothesized that this could be because of the similar chemical composition of all six samples. They also stated that the results obtained represent surface behavior instead of a localized effect. Leon et al. also obtained similar results [117]. An SLM-manufactured AlSi10Mg alloy was compared with a cast Al alloy A360.2. In both cases, the corrosion resistance of the SLM AlSi10Mg alloys was only slightly greater than that of the cast Al alloy. This behavior was attributed to the different microstructures of both types of alloys. Corrosion behavior of DMLS-produced Al alloys (as-printed and as-ground) against casting Al alloy (A360.1) was investigated by Fathi et al. [52]. The electrolyte employed was an aerated 3.5 wt. NaCl solution. In the potentiodynamic polarization test, the DMLS specimen exhibited higher corrosion resistance compared to cast Al alloy. This performance was attributed to the volume fraction variation of intermetallic compounds, the micro-constituents (i.e., Si), and the homogeneous corrosion layer on its surface. The EIS test indicated that the as-printed DMLS Al–Si10–Mg alloy exhibited the highest corrosion resistance. The test results are shown in Figs. 8(a) and 8(b).

Since heat treatment affects the corrosion performance of AMed alloys by changing microstructural characteristics, it is important to determine appropriate heat treatments for various alloys. However, not only do different alloys respond differently to heat treatments due to the underlying compositions of each respective alloy, but the AM process itself influences microstructure greatly. As such, this section compares the published results on the heat treatments and resultant corrosion performances of 316L stainless steel, Ti–6Al–4V, and AlSi10Mg alloys.

Corrosion performance varies not only between alloys and heat treatments but also between publications. Figure 9 displays the effects of heat treatment temperatures on corrosion current density and corrosion potential for various alloys from recently published works. Note that for Fig. 9(a), the y-axis is corrosion current density on a log-based scale with units of A/cm², while the x-axis is the temperature with which the respective samples underwent heat treatment. For Fig. 9(b), the y-axis is corrosion potential with respect to a standard hydrogen electrode (SHE). All data are based on potentiodynamic polarization experiments in 3.5 wt% NaCl solution with the exception of one of the experiments with AlSi10Mg, where Harrison solution was used (3.5 wt% (NH₄)₂SO₄ + 0.5wt% NaCl) [56]. Counter and reference electrodes were either platinum and SCE, respectively [36,44,46,56] or graphite and Ag/AgCl, respectively [45,54]. Additionally, gas metal arc AM [36], SLM [44,46,56], and DMLS [45,54] fabrication techniques were used. Note that for Ti–6Al–4V [45], the heat treatment displayed at 900 °C was followed by heat treatment at 650 °C.

While publications provide somewhat differing results, some trends can be gathered. Heat treatment temperature influences the corrosion rate, although different alloys are affected to varying degrees. Of all alloys and publications shown in Fig. 9(a), 316L stainless steel is least affected by heat treatment, with all heat treatments slightly improving corrosion performance. The 316L stainless steel also has some of the lowest corrosion rates, with a corrosion current density of less than 10⁻⁷ A/cm² for all points. Meanwhile, aluminum alloys tended to be more corrosive with a corrosion current density of about 10⁻⁷–4 × 10⁻⁶ A/cm². The difference in corrosion performance of the different publications on AlSi10Mg is likely due to the different publications using different electrolytes. Certain heat treatments are beneficial to corrosion performance for these aluminum alloys while others are damaging due to the effect heat treatment has on the Si distribution in an α–Al matrix, as discussed previously in Sec. 2.1.3. Ti–6Al–4V has a low corrosion current density of 10⁻⁸–10⁻⁷ A/cm² for two publications [44,45], but a high and dramatically changing corrosion current density of 2 × 10⁻⁷–8 × 10⁻⁶ A/cm² for another publication [46]. Additionally, heat treatments always degrade corrosion performance for the same two publications [44,45], while greatly improving the corrosion performance at heat treatment of 750 °C in the other publication [46]. The reasons for these discrepancies with Ti–6Al–4V are unknown, although they might be due to the implementation of different AM machines, fabrication parameters, or the orientation of the corroded surface with respect to the build direction.

Corrosion potential indicates the thermodynamics of the corrosion reaction. As such, higher corrosion potentials often show a tendency against corrosion, while lower corrosion potentials often show a tendency toward corrosion. Some alloys such as stainless steels have a low corrosion potential while they are active. However, after brief exposure to many corrosive environments, stainless steel forms a passivation layer, which results in a much higher corrosion potential. Nevertheless, if the corrosive environment can penetrate the passive layer, corrosion can be quite severe despite the high corrosion potential. Thus, corrosion potential is not always an accurate description of the direction of the corrosion reaction. Note that there was no corrosion potential data recorded for the AlSi10Mg alloy [56]. Passivated 316L stainless steel has some of the highest corrosion potentials for all heat treatments, signifying its superior corrosion resistance in these tests. Meanwhile, the potentials for Ti–6Al–4V are second, and AlSi10Mg is the lowest for all recorded heat treatments. The large differences between publications for both corrosion current density and corrosion potential show that there are many influences on corrosion performance outside of heat treatment. More research needs to be performed to precisely define the optimal heat treatments for different alloys in relation to corrosion performance, and careful experiments need to be run to compare alloys.

This section discusses some aspects of the wear of AMed materials, especially the AMed metallic materials. It is divided into four sections: the first section discusses some fundamentals of wear and wears mechanisms, the second section discusses the surface finish of AMed parts and its influence on the wear-rate of AMed parts, the third section discusses the microstructure of the AMed parts and its influence on the wear mechanisms, and the fourth section discusses one effective way to enhance the anti-wear property that is special to the additive manufacturing process.

Wear is one of the most important aspects of surface damage. It will reduce the life of a component, remove material, and lead to catastrophic failure in severe cases. The study of wear was started at the dawn of engineering science during the renaissance. The advancement of the surface science and scientific instrumentations in the last century has led to the great advancement of wear analysis and prevention. If the AMed part requires to be in moving contact with some other machine parts, its wear properties should be studied.

Wear can take many forms which are categorized as “wear types.” Some common wear types are abrasive wear, adhesive wear, erosive wear, and corrosive wear. The most studied wear types on the AMed part are abrasive wear and adhesive wear. Abrasive wear is highly related to the hardness and the type of contact. This type of wear is caused by direct or indirect mechanical interactions. The contacting material can “plow” and “cut” into the surfaces during contact. When the contact between two surfaces is loaded repeatedly, this cyclic load can cause crack growth underneath the surface. These sub-surface damages can grow and create fragmentations. Meanwhile, adhesive wear is related to not only the hardness but also the material crystal structure and microstructure. This type of wear is caused by the material transfer between contacting bodies. It is caused by the adhesive force exceeding the strength of the material. This type of wear often increases the roughness of the contacting surfaces.

The poor surface finish that is common in the AMed part can make wear easier. Generally, wear is increased when the surface roughness is increased [118]. This effect is especially prominent when the roughness is high (Fig. 10). The AMed metallic parts have a roughness in the range of 10s of microns [119]. Even though this roughness can be reduced by carefully controlling the printing parameters, it is still too large for contacting parts [120]. This high roughness not only can cause a higher wear-rate at dry sliding [118,121,122] but also require higher viscosity of the oil when lubricated to achieve proper lubrication [123]. In the case of steel with high surface roughness, the initial wear is linearly correlated to the surface roughness [124].

To improve the surface finish of the AMed part, some surface finishing techniques can be applied. The mechanical polishing methods such as barrel finishing and shot peening can effectively reduce the surface roughness [125]. The ultrasonic burnishing technique uses a metallic pin made from hard materials. That pin is driven by ultrasonic vibration and repeatedly pressed down on the printed metal surfaces [125–127]. It can decrease the surface finish Ra from 18 µm to 3.5 µm for AMed AlSi10Mg alloy [126]. However, this method can only polish the outside of the material. When the AMed part is a non-magnetic material, magnetically driven magneto-rheological fluid can be used to polish the part. This can reduce the surface roughness on par to the ultrasonic method, but can reach into the inside surfaces [128]. Printing patterns on the surface can also improve the tribological and wear performance of a part but it is limited by the resolution of the AM equipment [129].

The microstructure of an AMed part can influence its wear behavior. Stainless steel’s resistance to oxidation makes it susceptible to severe adhesive wear by adhesion and metal transfer [130,131]. Its homogenous austenite structure as shown in Fig. 11(a) also encourages the growth of adhesive wear damage [130]. AMed stainless steel, compared to its wrought or cast counterparts, has a fine and inhomogeneous microstructure [132–134] (Fig. 11(b)). This microstructure makes AMed parts have double the hardness compared to wrought steel [132], reducing both abrasive wear and adhesive wear. In the case when AMed stainless steel has a similar hardness to wrought stainless steel, its microstructure also makes it more wear-resistant [135]. The segregation of chromium carbide, while bad for corrosion resistance, greatly improves its wear resistance. The AMed aluminum alloy AlMgSc also shows better wear resistance compared to the as-cast alloy [136]. The cellular structure in this case also reduces wear.

The unique microstructure makes the AMed material have anisotropic wear performance [137,138]. In both the case of aluminum alloys and stainless steel, when the sliding direction coincides with the direction of printing (the direction of cellular structures), the rate of wear increases. This direction also has the highest friction, indicating the material may be more prone to adhesive wear in this direction. In Fig. 12, this effect is illustrated.

The AMed metal part can have unfavorable microstructures that make it prone to wear. The pores formed with unfavorable printing parameters for Inconel 718 can double the wear-rate compared to the parts printed with optimized printing parameters [139] (Fig. 13). The low-density porous structure also leads to the increment of the wear in the AMed TiC/Inconel 718 [139]. The enlarged grains formed by the arc additive manufacturing process for stainless steel reduce the hardness and wear resistance [140]. For stainless steel or Co–Cr alloys, the repetitive melting process can lead to intergranular segregation of chromium carbide (Cr₂₃C₆) [141]. These segregated carbide particles decrease the hardness of materials and can cause third body wear [141].

Post-printing heat treatment can improve the anti-wear performance of AMed parts. The solutionizing process can remove the carbide in the metal matrix of Co alloy [141,142]. This process can decrease the wear-rate by one-third for printed Co–Cr–Mo alloy [141] (Fig. 14). The laser-remelted Co–Cr–Mo alloy also has an improved wear resistance by forming a hard coating on the surface [143]. Another way to improve the microstructure is to incorporate some alloying elements. The Co–Cr–W alloy with 3 wt% of Cu has a 30% increase in wear resistance because it increases the volume percentage of the harder ε-phase [144].

The wear resistance of the AMed parts can also be improved by compositing hard particles to form a metal matrix composite (MMC) [145]. AM processes such as SLM have been proved to be able to form Metal matrix composites in situ or ex situ [146]. The inclusions of the particles can effectively improve the wear resistance in two major mechanisms: one is introducing the heterogenous which improves the adhesive wear behavior, another and more important way is it increases the hardness of the AMed part significantly. The TiC nanoparticle-reinforced AlSi10Mg has a high hardness of 181.2HV and high wear resistance [147]. This nanoparticle can also be used to reinforce the Al–15Si alloy and increase its wear resistance [148]. Graphene nanoplatelet-reinforced AlSi10Mg has half of the wear-rate of AlSi10Mg alloy [149]. The wear is suppressed by the graphene platelets wrapping around the wear debris.

How these composites improve the wear resistance of different alloys will be described in the following sections categorized as titanium alloys, aluminum alloys, and steel alloys.

Titanium alloys and titanium-based composite materials draw considerable interest in the additive manufacturing community. Its high corrosion resistance [70], biocompatibility [150], low density, and excellent mechanical behavior under high-temperature range [151] make it a highly desirable choice for applications such as aviation, engine, and medical applications. However, the wear resistance of common engineering titanium alloys is poor [152]. It is recommended to use anodization and coating when titanium alloys are to be used in tribological contacts [152]. The SLM process only slightly increases the wear resistance of the titanium alloys such as Ti–6A–4V [153,154]. This can attribute to the high roughness offsetting the increased hardness gained from the SLM process. Table 2 shows results from literature on the wear-rate reduction of various AMed titanium alloys paired with a variety of fillers.

Aluminum alloys, despite some good properties such as low weight and high strength-to-weight ratio, have poor wear resistance. However, the use of aluminum alloys to fabricate MMC to improve its tribological behavior was started in the 1970s [159–161]. The aluminum MMC has great potential for use in industries such as automotive parts.

The additive manufacturing process producing aluminum MMC material has improved tribological properties. This improvement, together with the improved hardness from the additive manufacturing process, finally allows the aluminum alloys to be used in applications such as bearings and moving parts. Literature results of aluminum alloys fabricated by SLM paired with a range of fillers and the corresponding wear-rate reduction are shown in Table 3.

Steel alloys, especially stainless steels, are one of the most important engineering materials that are widely fabricated by the additive manufacturing process. Unlike titanium alloys, the hardness of steel alloys is much higher compared to their cast or forged counterpart. That hardness increase can decrease the wear-rate of the steel alloys if a favorable additive manufacturing parameter, such as energy density and scanning speed, is employed.

Additively manufactured austenitic stainless steel has poor wear resistance, despite its high corrosion resistance. Its low hardness makes it form abrasive wear more easily, and its homogenous microstructure also makes it prone to adhesive wear [130]. Adding composite materials also proves to be an effective way to improve its wear resistance [165]. Table 4 displays published results on AMed 316L stainless steel and P20 steel paired with some fillers and the resultant wear-rate reductions.

In summary, through a discussion of corrosion and wear failure modes of AMed alloys, we have found the following. Control of alloy microstructure is still in its beginning phases of development, but the science behind the microstructures present in AMed alloys is continually advancing. Along with these advancements brings more innovative and advanced alloys and fabrication methods. Among the alloys fabricated by AM, stainless steel, titanium, and aluminum stainless steel alloys have made some of the greatest improvements in understanding in recent years, especially in how AM affects and creates microstructures and how these microstructures affect corrosion performance. Some key findings in recent years are listed in the following.

In terms of corrosion:

• Corrosion performance of additively manufactured alloys is frequently better than that of the same alloys fabricated by traditional techniques. This is largely due to the unique microstructures present as a result of rapid solidification and large thermal gradients during the AM process. These thermal gradients are greater in the horizontal plane than build direction, forming columnar grains which in turn creates anisotropic corrosion resistance.

• AM corrosion performance depends largely on fabrication process parameters and post-processing. Laser powers which are too high or too low are detrimental to corrosion resistance. Laser scanning speed also holds a strong influence on corrosion, but its effect depends on the alloy used. In terms of post-processing procedures, sandblasting has been found to decrease corrosion resistance while polishing increases it. Data on heat treatment are somewhat contradictory, but it appears that it increases the corrosion resistance of 316L stainless steel while having positive or negative effects on Ti–6Al–4V and Al–Si10–Mg, depending on the temperature used.

In terms of wear:

• For AM, the most important wear prevention technique is the fabrication of metal matrix composite materials.

• Some popular filler particles and their effects on anti-wear properties of AMed materials were discussed.

• AMed parts have superior anti-wear properties compared to cast counterparts. Depending on feasibility, AMed parts can be a better option for reducing wear. Printing parameters and heat treatment can optimize the wear resistance.

• The existing issues in the wear of AMed alloys require more research and innovation. For instance, titanium oxide protective film is an interesting development, but it requires more exploration.

• Need for precision on AMed alloys: The performance of alloys fabricated by AM inherently varies due to the many parameters which exist. While this variation is impossible to remove with current technology, statistical analysis such as factorial designs on the fabrication parameters can provide useful information on how to decrease these variations. Some parameters such as laser power and scanning speed have already been studied to an extent, but the results these studied have not shown they hold true across different additive manufacturing processes.

• Consideration of anisotropic behavior: Anisotropic corrosion performance should be taken into consideration during the design of AMed alloys. Since corrosion behavior varies depending on orientation, proper care must happen to orient surfaces that have the most corrosion resistance such that they are exposed to the most corrosive environments. Alternatively, the anisotropic microstructures of AMed alloys need to be removed through the development of new processes.

• Needs to understand heat treatment effects on AMed alloys: The published results of the effects of heat treatment on the corrosion performance of AMed materials vary greatly. Discrepancies in literature may be due to small sample sizes (often three samples or less per test are used). Also, a variety of different methods of AM are used, which may influence the effect of heat treatment on corrosion. This problem can be solved by developing a database containing many datapoints for each variable, thus reducing the effect that statistical variation has on corrosion performance. Also, one way to potentially improve the effect of heat treatment on corrosion performance is by creating a method that removes the anisotropic microstructures and thus corrosion behavior of these alloys.

• Improving AMed material performance using nanoparticles: Incorporating nanoparticles in the metal matrix is an effective way to improve AMed metallic parts. The inclusion of particles enhanced the hardness of the printed part and resulted in a more wear-resistant material.

• Failure analysis of AMed steel: There are many processes for AMed stainless steel. But variations in parameters and processes, however slight, produce different results. The behavior of AMed stainless steel needs to be analyzed and understood.

Peter Renner was supported by the National Science Foundation (NSF) Graduate Research Fellowship.

There are no conflict of interest.