Effect of Heat Treatment on Corrosion and Mechanical Properties of M789 Alloy Fabricated Using DED

Abstract

The directed energy deposition (DED) process offers potential advantages, such as a large building space, limited dilutions, narrow heat-affected zones (HAZ) and potentially improved surface properties. Moreover, heat treatments have been reported to significantly improve the properties of the as-built sample by modifying the microstructure. In this study, the influences of various combinations of heating and cryogenic treatments on the mechanical performance and corrosion resistance of DED M789 steel have been critically investigated. The microstructure and hardness were examined to discuss the characteristics of the M789 parts in the as-printed and heat-treated states. The corrosion rate was determined from the weight loss monitoring based on the seawater immersion condition. The microstructural results revealed the distortion of martensite lattice and the formation of nano-carbide precipitates after the cryogenic treatment. Moreover, the microhardness of the cryogenically treated M789 steel was found to be significantly higher which was attributed to the precipitate strengthening and elimination of retained austenite, resulting from the increased volume fraction of carbides due to cryogenic treatment. The corrosion characteristics were also modified by the heating/cryogenic treatments, and the substrate-to-deposit ratio of the corrosion sample also substantially affected the overall corrosion rate.

1. Introduction

Maraging steels are a class of low-carbon steels which are widely exploited in the aerospace, automotive and tooling sectors owing to their excellent combination of high strength, good toughness and superior weldability [1,2,3,4,5]. The strengthening of maraging steels is not dependent on carbon, due to its low content, but rather on the formation and types of the intermetallic during the aging process, which play significant role in determining the mechanical performance. Hence, various studies have deeply investigated the influence of alloying elements on the precipitation behavior, strengthening mechanisms and subsequent mechanical properties of maraging steels [6,7,8,9]. Moreover, the excellent weldability of maraging steels due to their low carbon content makes them highly suitable for additive manufacturing (AM), where rapid cooling rates can induce a martensitic structure during fabrication. Laser-directed energy deposition (DED) is one of the most widely utilized AM techniques which offers multiple advantages over other traditional manufacturing techniques such as near-net-shape parts, a low buy-to-fly ratio, large scale printing and, most importantly, the repair of worn or damaged parts [10,11,12,13,14,15,16]. DED is a powder-based deposition process, where a powder and laser are simultaneously passed on the substrate. The printing head deposits the metal powders on the substrate materials and the high-energy laser melts the powder with layer-by-layer symmetry. Due to higher cooling rates of molten metals during processing, DED manufactured parts comprise a distinct microstructure with columnar grains which shows unique properties [17]. Owing to those advantages, DED has been significantly exploited in the fabrication of diverse materials such as Ti, Ni, Al, and Fe [1,2,18,19,20,21].

Despite their tantalizing mechanical properties and excellent printability, the poor corrosion resistance of maraging steels significantly hampers their applicability in several sectors, especially under harsh conditions, as compared to conventional steels [22,23,24]. Hence, multiple studies have been carried out in recent years to design maraging steels using AM techniques with optimized properties that are comparable to 15-5 PH, 17-4 PH or duplex steels. Recently, the newly designed, cobalt-free, maraging steel grade known as M789 has been reported to demonstrate excellent corrosion resistance while maintaining adequate mechanical properties [25,26,27]. Turk et al. investigated the mechanical and corrosion characteristics of three maraging steels, W722, N700 and M789, produced via selective laser melting (SLM) [25]. Microstructural results revealed the typical layered structure in the as-built condition, which was altered to a recrystallized microstructure with a Fe-Ni (Co) martensitic microstructure after the solutioning and aging treatment. M789 exhibited exceptional mechanical and corrosion properties comparable to W722 and N700 maraging steels. Tian et al. examined the effect of solutioning (1000 °C for 1 h) and aging heat treatment (400–600 °C for 40–120 min holding time) on the microstructural evolution and terminal mechanical properties of M789 fabricated by laser powder bed fusion (LPBF) [7]. The highest tensile strength (i.e., 1798 MPa and 1019 MPa for heat treated and as-built, respectively) was achieved at 500 °C for 1 h aging treatment. Transmission electron microscopy (TEM) and atom probe tomography (APT) revealed the presence of plate-like and spherical Ni₃Ti precipitates which contributed to significant strengthening. It is pertinent to mention that most of the literature is focused on fabricating M789 steel using the LPBF; however, the research related to DED-manufactured M789 is still limited. Lek et al. critically investigated the influence of heat treatment on the mechanical and corrosion resistance of M789 fabricated by DED [28]. Electron back scattered diffraction (EBSD) revealed the martensitic structure in the as-built and heat-treated samples and scanning tunneling electron microscopy (STEM) and transmission Kikuchi diffraction (TKD) confirmed the presence of Ti- and Al-rich precipitates within the martensitic structure after the solution and aging treatment, which were responsible for high yield strength due to solid solution and grain boundary strengthening. Besides exceptional mechanical properties, DED-fabricated M789 also displayed higher pitting potential in the as-built condition as compared to that in the heat-treated condition. Since different cooling rates involved in LBF- and DED-fabricated M789 result in distinct microstructures and mechanical performance, it is extremely important to probe the microstructural evolution and its consequent effects on the mechanical performance of DED-fabricated M789.

In this study, the microstructural, mechanical and corrosion characteristics of maraging steel M789, which was fabricated using DED, were investigated. Various heat treatment and cooling media were exploited to probe the influence of heating and cooling regimes on the microstructure and related properties of DED-fabricated M789 steel. Hence, finding the optimized conditions for the best performance of DED-fabricated M789 steel can significantly increase its utilization in diverse conditions.

2. Materials and Methods

2.1. Powder Materials

The argon-atomized pre-alloyed M789 powder material supplied by BÖHLER Edelstahl GmbH and Co KG (Kapfenberg, Austria) was used as the feedstock powder. The chemical composition of the M789 powder was examined using inductively coupled plasma–optical emission spectroscopy (ICP-OES) and is listed in

Table 1. Chemical composition of pre-alloyed maraging steel M789 powder, which complied with the ASTM A693-16.

Elements	Fe	C	Si	Mn	Cr	Mo	Ni	Cu	Ti	Al
Wt. %	Bal.	0.01	0.44	0.03	12.25	1.04	10.45	<0.02	1.00	0.62

.

The particle fraction ranged approximately from 45 μm to 90 μm. Each powder feedstock was characterized using standard test methods (ASTM B822/ISO 13322-2:2006 [29]), including flowability and apparent density under the ASTM B-213-17 [30] and ASTM B-212-13 [31], respectively. The results are listed in

Table 2. Summary of standard powder measurements and size distribution for the initial M789 feedstocks used in L-DED process.

Power Characterization		Powder Size Distribution (μm)
Apparent density (g/cm3)	Hall flow rate (s/50 g)	D10	D50	D90
3.4	22.0	42.0	54.1	71.3

.

2.2. Hybrid Laser Direct Energy Deposition System

The M789 parts were printed using a hybrid laser direct deposition system, which incorporated a custom-built five-axis stage combined with a milling machine (Hwacheon Machinery Co., Ltd., Seoul, Republic of Korea). The machine included a conventional 3-axis deposition head, as well as a 2-axis tilt-rotate, as shown in Figure 1.

The specimens were fabricated using a zig-zag laser scanning strategy that incorporated a 180° rotate scanning approach, whereby the laser scanning lines were tiled with 180° between each laser. The builds were all produced using a laser power of 385 W, laser scan speed of 1000 mm/min, and a powder feed rate of 6.6 g/min. To eliminate surface irregularities and defects, each layer was deposited in the normal direction (ND) with a hatch spacing of 0.5 mm and a layer thickness of 0.25 mm. No shield gas was used to control the environment of the build chamber.

2.3. Heat Treatments

The experimental heat treatment scheme was formulated according to ref. [7] and is illustrated in detail in Figure 2. Solution and aging treatments were implemented as part of the heat treatment scheme, as shown in Figure 2a,d. It is noteworthy that the cryogenic treatment (i.e., submerging the material in liquid nitrogen) was deliberately included for comparison, as is depicted in Figure 2b,c. The samples are referred to as SAT (Solution treatment at 1050 °C for 1 h and aging treatment at 500 °C for 3 h), SCAT (Solution treatment and cryogenic treatment at −196 °C for 24 h before the aging treatment), DAT (Directed aging treatment) and CAT (Cryogenic treatment before the aging treatment), based on the naming conventions established by this heat treatment scheme. These designations correspond to the solution and aging treatment, solution and aging treatment with cryogenic treatment, aging treatment with cryogenic treatment, and direct aging treatment, respectively.

Initially, the samples were subjected to solid solution treatment at 1050 °C for 1 h to ensure adequate dissolution of the solute atoms into the matrix. The heated samples were then rapidly quenched by water after solution treatment to obtain supersaturated lath martensite (Figure 2a,b). The cryogenic treatment was further utilized to attain an approximately fully martensitic transformation before the retained austenite became mechanically stable (Figure 2b). Subsequently, the samples were subjected to aging treatment at a heating rate of 10 °C/min up to 500 °C for 3 h to introduce substantial nanosized M_xC strengthening particles. The samples that were omitted to solution treatment were further designated as cryogenically treated aged samples (Figure 2c) and directed aged samples (Figure 2d), respectively.

2.4. Evaluation of Corrosion Rates

The corrosion rate was evaluated by the immersion test in electrolyte solutions at 25 °C for 1 to 4 days. Mirror-polished samples with a 1 × 1 cm² exposed area were completely immersed in a 0.1 M NaCl solution. Corrosion behavior was characterized following immersion and every 24 h thereafter for a total of 4 scans per sample. The corrosion rate relative to the weight loss test can be quantified using the following equation:

$$v=\frac{W\times K}{\rho \times S\times t}$$

where v is the corrosion rate (mm/y), W represents the weight loss of the samples (g), K constant (8.76 × 10⁴) ρ is density (g/cm³), S is the exposed surface area (cm²) and t is the immersion time (h).

2.5. Metallography and Microscopic Characterization

The M789 parts were mechanically ground and polished according to standard metallographic procedure and subsequently etched with Fry’s reagent (1 g CuCl₂ + 50 mL HCl + 25 mL HNO₃ + 150 mL H₂O) [32]. The microstructures of the samples in various thermal states were analyzed using a Zeiss Axioscope A1 optical microscope (Carl Zeiss, Oberkochen, Germany) and a JEOL 7800F field emission scanning electron microscope (JEOL Ltd., Tokyo, Japan) operating at 20 kV. The aggregation of alloying elements at the microscale was qualitatively determined using Aztec Energy Dispersive X-ray Spectroscopy (EDX).

The composition and volume fraction of austenite after immersion were identified using a Rigaku SmartLab X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) with a Cu-Kα anticathode operated at a voltage of 45 kV and current of 200 mA. The radiation was scanned over a 2θ angle range of 40° to 90° to identify the phase present in the as-fabricated and heat-treated samples, with a scanning rate of 2° min⁻¹ and step size of 0.020°. The raw data were then processed to examine the dislocation density as a function of peak broadening using the Williamson–Hall method [33].

A microhardness profile was measured using a HMV-G Vickers hardness testing machine (Shimadzu Corporation, Tokyo, Japan) with a load of HV1 (9.807 N) and a dwelling time of 10 s.

3. Results and Discussion

The optical and SEM micrographs depicted in Figure 3 provide an overview of the microstructures of the as-fabricated SAT, SCAT, DAT and CAT samples. The as-fabricated sample displays fish-scale like melt pool boundaries along the construction direction that alternate with layers (Figure 3a). Melt pool boundaries are produced by the scanning strategy with 90° rotations between each adjacent layer. The fish scale-like melt pool boundaries become invisible after the SAT and SCAT samples (Figure 3b,c) but can still be observed after aging treatment (Figure 3d,e).

Figure 4 provides a detailed view of the SEM microstructures of SAT, SCAT, DAT and CAT samples. Specifically, Figure 4a shows a lath martensite structure, which is a type of microstructure that is commonly observed in quenched and tempered steels. This solidified dendrite structure is characterized by a thin, elongated shape, with a high degree of alignment between neighboring martensite laths. As seen in Figure 4b,c, the distribution of nano-carbide precipitation was observed in SCAT samples, which suggests that the martensitic lath bundles were refined. Figure 4c,d show that there is a clear distribution of spherical carbide particles within the matrix. This is in contrast to the SAT and SCAT samples, where such particles were not observed. This distribution pattern suggests that the DAT and CAT samples underwent a different type of transformation after cryogenic treatment. One of the key observations that can be made from these SEM microstructures is the broad-to-flack transformation of martensite. Specifically, the cryogenic treatment process appears to lead to the distortion of the martensite lattice and the precipitation of carbon atoms.

Figure 5 presents the XRD results of the as-fabricated, SAT, SCAT, DAT and CAT M789 samples. It reveals that the dominant martensitic phases of (110)_α, (200)_α, (211)_α and (202)_α diffraction peaks were obtained in the as-fabricated M789 samples. However, the XRD pattern of the CAT and DAT samples exhibits the presence of an additional (111)_γ peak, suggesting the emergence of an austenite phase as a result of the aging treatment. The reversion of martensite to austenite generates additional quantities of austenite [28]. The formation of reverted austenite in the aging process is facilitated by the nickel contents exceeding 10% [34]. Conversely, the absence of (111)γ peaks in the SAT and SCAT samples indicates that only a small amount of retained austenite may exist within the matrix as a consequence of the solution annealing process. Enlarged (211)_α peaks are shown in order to explore the effect of cryogenic stage on the martensite structure, as shown in Figure 5b. Overall, 2θ of (211)_α of the as-fabricated sample increases after heat treatment and the martensite lattice constant decreases. The transformation of retained austenite after cryogenic treatment differs from that of the traditional process samples, owing to the smaller grain size resulting from the DED process [35]. The as-fabricated M789 exhibits a high density of dislocations, which can impede the transformation of retained austenite (see Figure 6 and

Table 3. XRD diffraction pattern parameters and calculated dislocation density of DED M789 steels.

Heat Treatment States	Parameter	Fe-α’ (110)	Fe-α’ (200)	Fe-α’ (211)	Fe-α’ (202)	Micro Strain (ε × 10−3)	Dislocation Density (δ × 10−3, nm−2)
As-fabricated	2θ (°)	44.49	64.54	81.90	98.43	3.89	4.02
	FWHM (°)	0.41	0.79	0.82	1.40
SAT	2θ (°)	44.50	64.68	82.06	98.62	5.96	9.43
	FWHM (°)	0.46	1.02	1.04	2.01
SCAT	2θ (°)	44.49	64.57	82.00	98.60	6.75	1.21
	FWHM (°)	0.46	1.01	1.11	1.36
DAT	2θ (°)	44.54	64.71	82.05	98.58	3.19	2.70
	FWHM (°)	0.43	0.67	0.79	1.03
CAT	2θ (°)	44.50	64.61	82.04	98.48	5.80	8.93
	FWHM (°)	0.45	0.97	1.02	1.13

). However, cryogenic treatment before the aging treatment results in a more stable austenite phase due to reduced space constraints [36]. Consequently, cryogenic treatment can effectively reduce the content of retained austenite by transforming it to martensite, resulting in a homogeneously dispersed fine carbide structure and a more favorable residual stress distribution [37].

Figure 7a presents the cross-sectional distribution of micro-hardness for the AISI 1045 substrate and as-fabricated M789 samples processed via DED. The micro-hardness values of the as-fabricated M789 samples were uniform within each layer, with an average value of approximately 286 HV and 172 HV for the M789 and AISI 1045 substrates, respectively. The interface between M789 and the substrate was characterized by a sharp drop in hardness after heat treatment.

The hardness profile after heat treatment is shown in Figure 7b. The substrate hardness values were observed to decrease, whereas the M789 steel exhibited a significant increase in hardness variation from 420 HV to 500 HV. These observations can be attributed to precipitate strengthening and the elimination of retained austenite, resulting from the increased volume fraction of carbides due to cryogenic treatment and more severe quenching. The martensite phase resulting from the treatment also had a lower carbon content, which led to a tougher matrix phase. The transformation of retained austenite, a soft constituent, into martensite compensated for the loss of hardness due to the carbon-depleted martensite.

Based on these findings, it can be concluded that controlling the intrinsic heat treatment during DED is crucial for achieving an apparent heterogeneity in microstructure and mechanical properties.

To investigate the influence of various processing regimes on the corrosion resistance of M789, an immersion test was conducted in the 0.1 M NaCl solution with various exposure times, as shown in Figure 8. The corrosion samples were cut with different substrate-to-deposit ratios (70:30 (C1) and 50:50 (C2)) to probe the effect of deposit on the electrochemical characteristics. The results revealed that for the C1 sample (Figure 8a), SCAT had the highest corrosion rate (0.175 mm/y) as compared to the other processed samples after 24 h immersion. The DAT sample showed a significant improvement in the corrosion resistance as compared to the as-fabricated (AB) sample. By increasing the immersion time to 48 h and 96 h, the corrosion rates were decreased and the difference in the corrosion resistance of various samples was also minimized. The significant decline in the corrosion rates after a long immersion time (96 h) can be attributed to the formation of passive films on the metastable pits which suppressed the interaction of the surface with the saline environment. In addition, when the deposit ratio was increased (C2 sample), all the samples displayed a remarkable improvement in the corrosion resistance with almost similar corrosion rates, indicating the superior corrosion characteristics of deposit materials as compared to the substrate (Figure 8b).

4. Conclusions

In this study, M789 maraging steel was fabricated using DED techniques. The built samples were subjected to various combinations of treatments including solution treatment, aging and/or cryogenic quenching. The detailed microstructural characterization and corrosion resistance were investigated for as-built and processed samples to evaluate the effect of treatment on the properties. The results showed that the cryogenic treatment resulted in martensite lattice distortion and the emergence of nanocarbide precipitates. Consequently, the microhardness of cryogenically treated samples was recorded to be significantly higher than other samples due to precipitation strengthening and a decrease in the fraction of retained austenite. The DAT sample showed the highest corrosion resistance as compared to other processed samples. Moreover, for increased immersion times (48 h and 96 h), the corrosion rates were decreased, which can be attributed to the formation of passive films on the metastable pits which suppressed the interaction of the surface with the saline environment. Lastly, the corrosion sample with a higher substrate-to-deposit ratio showed a remarkable corrosion resistance, confirming the better corrosion characteristics of the deposit material as compared to the substrate.