1 Introduction

The widespread adoption of selective laser melting (SLM) additive manufacturing (AM) in various industries has opened new frontiers in the production of advanced materials and composites (Ref 1,2,3,4,5,6,7,8). Innovative applications include the development of functionally graded materials (FGM), where properties are systematically graded by adjusting the chemical composition to meet specific requirements (Ref 9,10,11,12,13,14). This enables the precise adjustment of mechanical properties, such as hardness and wear resistance, as well as other characteristics like resistance to high temperatures. SLM plays a crucial role in the production of FGMs, a capability not achievable with conventional manufacturing methods.

One example is the use of SLM to produce functionally graded materials consisting of 316L stainless steel (SS) and TiC ceramic particles. TiC, known for its hardness and stability at elevated temperatures, serves as a valuable complement to metal matrices. 316L SS, widely employed in various applications owing to its excellent corrosion resistance and formability, nevertheless faces challenges related to low strength and wear performance (Ref 15, 16).

The addition of ceramic particles, such as TiC, has been shown to be effective in producing wear-resistant composites (Ref 17,18,19,20,21,22,23). However, when reinforcing with ceramic particles, there is a trade-off between hardness and wear resistance on the one hand and ductility and toughness on the other hand. By grading the TiC volume fraction in the matrix using SLM, both the hardness and toughness of TiC and the ductility of the 316L SS matrix can be utilized.

In processes characterized by rapid cooling rates and significant thermal gradients, there is a risk of residual stresses in the components (Ref 24). This problem is even more pronounced in the SLM process, where the cooling rates remain very high despite the implementation of various scanning strategies to mitigate them (Ref 25,26,27,28,29,30). As a solution, the application of post-heat treatment annealing has been proposed to relive stress and improve mechanical properties (Ref 31,32,33,34,35,36).

While post-heat treatment annealing has been employed to relieve residual stresses in SLM-fabricated components, there is a notable absence of reports on its impact on multilayer composites produced by this method. The objective of this research is to investigate the influence of annealing heat treatment on the microstructure and mechanical properties of the multilayer 316L-TiC composite fabricated by the SLM process.

2 Experimental Procedure

2.1 Material

The raw materials utilized in this study included AISI 316L SS powder featuring particle sizes ranging from 15 to 50 μm and titanium carbide (TiC) powder with particles measuring less than 20 μm. The specifications of these powders adhered to the requirements of the SLM powder feeding system.

To achieve two distinct powder concentrations, namely, 316L-5TiC and 316L-10TiC (wt.%), a thorough mixing process was carried out using a turbo mixer for 4 h. This step involved meticulous blending of 316L and TiC powders. Figure 1(a), (b), and (c) displays scanning electron microscopy images of plane 316L and composite layers, respectively. As observed, the 316L and TiC particles are well mixed, with some submicron TiC particles adhering to the 316L particles, ensuring a homogeneous distribution of TiC in the 316L matrix after the SLM process. Before initiating the SLM process, the blended powders underwent a drying phase in an oven set at a temperature of 103 °C, lasting for 1 h. This meticulous preparation ensured the optimal condition of the powder feedstock for subsequent processing.

Fig. 1
figure 1

Scanning electron microscopy images of the powders: (a) 316L SS, (b) 316L SS mixed with 5 wt.% TiC, and (c) 316L SS mixed with 10 wt.% TiC

Full size image

2.2 Selective Laser Melting Process

A laboratory SLM system was used to produce multilayer samples consisting of three layers of 316L SS, 316L SS-5TiC, and 316L SS-10TiC (wt.%). The SLM process began with data preparation using CAD software and ended with the removal of the manufactured samples from the substrate. To protect against oxidation, the samples were produced in a controlled oxygen atmosphere using argon inert gas. Prior to production, the chamber was evacuated and filled with argon to reduce the oxygen content to a specific level.

In this study, the parallel line scan strategy was applied, which is widely accepted in SLM due to its easy programming and wide applicability. To avoid distortions and deformations in the resulting samples, a rotation angle of 67° in the laser direction was implemented for each successive layer during sample fabrication (Ref 37). This specific rotation angle ensures that the laser beam direction is not repeated over a significant number of layers (several thousand), thus minimizing residual stresses in the finished part. Figure 2 illustrates the scanning strategy used in this study.

Fig. 2
figure 2

The parallel line scanning strategy employed in the current study, features a 67° rotation at each new layer scanned

Full size image

2.3 Heat Treatment

Heat treatment was performed in a tube furnace by austenitizing at 1100 °C for 1 h, followed by cooling in air to room temperature under an argon gas atmosphere. At this temperature, the solubility of TiC in the matrix is almost negligible (Ref 38), and only phase transformations occur within the matrix.

2.4 Microstructural Observations

The fabricated specimens were subjected to extensive microstructural analysis employing conventional metallographic methods. Grinding was performed using silicon carbide abrasive papers with grit sizes ranging from 60 to 3000, followed by automated polishing using diamond paste with particle sizes of 1 and 0.25 µm. The specimens were then subjected to etching with Glyceregia reagent for the duration of 60-120 s to reveal the microstructural features. The examination and characterization of these features, along with a detailed analysis, was facilitated by a Leica DFC 450 microscope. SEM observations were performed using a Jeol JEM 6060 LV scanning electron microscope, with the operating voltage set between 5 and 15 kV. To maintain sufficient focus and image clarity, a working distance of 10 mm was maintained throughout the observations.

2.5 Mechanical Property Evaluations

The Vickers hardness of the layers was measured using a Qness Q30M hardness tester instrument at room temperature. Prior to testing, the surface of the sample was carefully prepared, and the indentations were made with a force of 9.8 N including a 10-s relaxation pause. Multiple indentations were made on each layer surface, and the average was reported.

Tensile test was performed on samples prepared from the layers of the produced composite in accordance with the JIS Z 2201 standard (Ref 39) at room temperature using an Instron 3369 universal testing machine with the crosshead speed set at 1 mm/min. Load and displacement data were recorded to characterize the stress–strain behavior of the specimens.

The wear test was performed at room temperature using a T10/20 TURQUDSE 2.0 ball on disk Tribometer. The test was performed at a speed of 0.1 m/s and a constant load of 15 N, with a specified wear distance of 500 m. In this test configuration, the specimens were disk shaped, while Al2O3 ceramic balls acted as the abrasive. The test procedure was based on the ASTM G 133-95 standard, and the wear behavior and the resulting wear tracks were analyzed and evaluated based on the wear tracks and the observed wear mechanisms. In addition, surface topography measurements were made on the worn samples using a Huvitz/HRM-300 Optic Profilometer.

3 Results and Discussion

3.1 Microstructure

Figure 3(a) and (b) shows optical micrographs of the as-polished 316L and 316L-TiC composite layers. Notably, the images show the absence of microporosities, indicating a dense and well-compacted microstructure. Furthermore, the TiC particles exhibit a highly homogeneous distribution throughout the composite layer, with no observable evidence of clustering or agglomeration. This uniform dispersion suggests effective incorporation of the ceramic phase, which could positively affect the mechanical properties of the material. Additionally, the interfaces at 316L/316L-5TiC and 316l-5TiC/316L-10TiC composite layers appear defect-free, exhibiting excellent continuity and adhesion between the distinct regions. This intimate interfacial contact is crucial for optimal load transfer and overall composite performance.

Fig. 3
figure 3

Images of fabricated composite layers in their as-polished condition: (a) 316L-5TiC/316L-10TiC and (b) 316L/316L-5TiC. Dashed lines indicate the interface location between the layers.

Full size image

Figure 4(a), (b), (c), (d), and (e) presents the micrographs of the 316L-105TiC, 316L-5TiC, and 316L layers, respectively, after etching with Glyceregia reagent before and after annealing post-heat treatment. Figure 4(b) shows the intricate details within the 316L layer. The melt lines are clearly visible, revealing the formation of melt pools created by the passage of the laser during the AM process. It is noteworthy that the average width of these melt pools, as seen in the image, exceeds the selected hatch spacing of 60 µm. This phenomenon is due to the inherent misalignment between the metallurgical surface and the laser direction with each new layer, a consequence of the rotation of the laser during subsequent layers.

Fig. 4
figure 4

Images of the fabricated layers after etching with Glyceregia reagent for (a) 316L-5TiC/316L-10TiC, (b) and (c) 316L layers before annealing, and (d) 316L-5TiC/316L-10TiC, and (e) 316L layers after annealing post-treatment. The plane surfaces of images are parallel to build direction.

Full size image

Figure 4(b) also reveals that grains extended beyond individual melt pools into subsequent layers, with some grains extending over several layers, indicating that certain austenite grains from the preceding layers were partially melted, followed by epitaxial grain growth on these partially melted grains. This observation is consistent with the epitaxial growth mode reported by other researchers (Ref 40, 41). Furthermore, the microscope image illustrates the directional growth of the grains from bottom to top, which is opposite to the direction of heat transfer. This indicates the interplay between thermal gradients and grain growth mechanisms during the AM process. It is worth noting that the grains are not visible in the TiC-containing layers (Fig. 4a) due to the influence of TiC particles that refine the grain structure.

Figure 4(c) presents a magnified image of the 316L microstructure, revealing a characteristic cellular structure within the randomly oriented grains. This cellular morphology is a direct result of the exceptionally high cooling rates, typically ranging from 103 to 104 °C/s (Ref 42) during the SLM process. These rapid cooling rates, which are achieved due to the small melt pool size and the high laser travel speed, have a significant impact on the solidification process. At such intense cooling rates, the solidification process is so rapid that the formation of secondary dendrite arms is severely limited. This limitation contrasts with slower cooling scenarios where secondary arms readily develop, and eventually form a dendritic structure. Rapid heat removal promotes the formation of cellular interfaces along the solidification fronts. Stainless steel structures produced by SLM typically display three times the strength and improved mechanical properties of those produced by conventional methods. This improvement can be attributed to a refined cellular structure and higher dislocation density (Ref 43).

Figure 4(d) and (e) presents the annealed images of the fabricated layers after etching with Glyceregia. In Fig. 4(d), the austenite grain boundaries are not visible, similar to Fig. 4(a). In Fig. 4(e), the microstructure of the 316L layer reveals grown austenite grains, some of which showing annealing twinning. Very fine precipitates are observed in the microstructure after annealing treatment. Transmission electron microscopy studies conducted by Salman et al. (Ref 44) on the annealed microstructure of 316L produced by SLM revealed an amorphous structure for the precipitates, which are rich in Mn and Si. The cellular structure observed in the non-heat treated sample is not evident in the annealed sample. The studies indicated that cellular walls are composed of dislocation tangles accompanied by segregation of Mn, Mo, and Cr elements. Therefore, the disappearance of the cellular structure can be attributed to the reduction in dislocation density during annealing. The average grain size was measured to be 25 ± 7 µm in the annealed sample.

3.2 Mechanical Properties

Figure 5 displays the Vickers hardness (HV1) of fabricated layers, both in the as-built and annealed conditions. The hardness of each layer reflects an average of at least five measurements. The figure clearly shows a significant increase in the hardness of 316L layer when reinforced with TiC particles. In particular, the hardness values for 316L, 316L-5TiC, and 316L-10TiC layers are 245, 368, and 417 HV1, respectively. The fine cellular structure and high dislocation density within the austenite matrix contribute significantly to the hardness and strength of the material (Ref 45). These properties result from the rapid cooling rates inherent to the SLM process. The presence of TiC particles further enhances hardness through two mechanisms: particle strengthening and microstructural refinement. The hard TiC particles impede dislocation movement, preventing plastic deformation and increasing hardness. The submicron TiC particles refine the surrounding matrix, creating more grain boundaries that also imped dislocation movement (Ref 46). Figure 6 shows the scanning electron microscopy image of the 316L-10TiC matrix, where both submicron and larger TiC particles can be observed. AlMangour et al. (Ref 17) suggest that the formation of α/martensite during SLM may also contribute to increased hardness.

Fig. 5
figure 5

Hardness of 316L, 316L-5TiC, and 316L-10TiC layers in as-deposited and annealed conditions.

Full size image
Fig. 6
figure 6

SEM micrograph from the 316L-10TiC layer indicating the distribution of submicron-sized TiC particles along the larger particles

Full size image

Annealing resulted in a decrease in hardness of approximately 60-70 Vickers across all layers due to two primary microstructural factors: coarsening and dislocation recovery. Annealing promoted grain growth, coarsening the microstructure and reducing the number of grain boundaries, which act as barriers to dislocation movement and contribute to hardness. On the other hand, it facilitated dislocation recovery, reducing the overall density and consequently lowering the resistance to plastic deformation.

Despite the softening effect of annealing, layers containing TiC particles retain higher hardness compared to the unreinforced 316L layer. This is because the presence of TiC particles mitigates the softening mechanisms associated with annealing, while maintaining their strengthening influence through particle reinforcement.

Figure 7 presents the stress–strain curves of both single and multilayer 316L samples, with corresponding test results listed in Table 1. These results are presented for both the as-deposited and annealed post-treatment conditions. Samples were prepared parallel to the substrate for consistency. The 316L layer exhibits an impressive elongation of 45%, as shown in the figure. This exceptional ductility implies minimal microporosity or other defects that could negatively affect its formability. Notably, these properties exceed the minimum requirements of the ASTM A240 standard and even exceed data reported in the literature for 316L produced by SLM (Ref 47). In contrast, the multilayer sample with layers oriented parallel to the longitudinal direction exhibited brittle fracture behavior, as shown in Fig. 7. This brittleness is due to the high volume fraction of brittle TiC particles combined with the reinforced metal matrix within the composite layers, effectively eliminating any ductility. However, the annealing process successfully introduced some ductility, resulting in increased strength prior to the final fracture. Figure 7 further highlights the increased ductility observed in the 316L layer following the annealing post-treatment. This improvement can be attributed to the relaxation of residual stresses and the recovery of the dislocation microstructure, both of which contribute to improved material flexibility.

Fig. 7
figure 7

Stress–strain curves for 316L and multilayer samples in as-deposited and annealed conditions

Full size image
Table 1 Ultimate tensile strength (UTS), yield strength (YS), and elongation of 316L and multilayer SLM samples before and after annealing post-heat treatment
Full size table

Figure 8 presents the weight loss of the fabricated layers in as-deposited and annealed post-heat treatment conditions, along with the corresponding friction coefficient profiles. The weight loss decreases from the 316L layer to the composite layers containing 5 and 10% TiC, respectively. In addition, the weight loss is higher for the layers after the annealing post-heat treatment. A similar trend is observed for the friction coefficient.

Fig. 8
figure 8

(a) Weight loss of 316L, 316L-5TiC, and 316L-10TiC layers in as-deposited and annealed conditions after wear test and (b) the corresponding friction coefficients

Full size image

Friction coefficient profiles can be divided into two regions: running-in and steady state, under dry wear conditions (Ref 48, 49). The initial surface roughness determines the running-in time until the steady state is reached. Based on the friction coefficient profiles in Fig. 8, the running-in distance is very short, and there is not much difference among samples, considering that they were polished before the test and had a smooth surface. After this stage, the wear reaches the steady-state stage. The friction coefficient values shown in the figure are the average of the profile for the entire wear distance. Higher friction coefficient values correspond to more severe contact between surfaces, which consequently increases stress and weight loss. Several factors such as the amount of load, number of cycles or wear distance, and area of contact influence the wear behavior of the material. In the samples of this study, these parameters were identical. However, the hardness of the matrix and the amount of TiC ceramic particles were different among samples. The annealing post-heat treatment also influenced the residual stresses, which could influence the wear performance. The higher the initial hardness, the better the wear performance. This explains the lower weight loss of TiC-containing layers and, in contrast, the higher weight loss for annealed samples. The annealing treatment generally increased the weight loss in all layers, although the effect was less pronounced in TiC-containing layers. This suggests that while annealing may slightly soften the matrix, the presence of TiC particles still provides superior wear resistance compared to the unreinforced 316L layer even after heat treatment.

Figure 9 (a), (b), (c), (d), (e), and (f) reveals the wear tracks of 316L, 316L-5TiC, and 316L-10TiC layers in both as-deposited and annealed states. The presence of debris on the worn surfaces, confirmed by EDS analysis to be oxygen-rich oxidized particles, suggests material removal during wear. Notably, these particles are more abundant in the 316L layer, indicating greater material loss due to its lower hardness without TiC reinforcement. Furthermore, deeper scars are evident in the 316L wear track compared to the TiC-reinforced counterparts. This observation directly points to the effectiveness of TiC particles in mitigating wear damage and promoting smoother surfaces. Based on these images, abrasive wear appears to be the dominant mechanism, with a higher intensity in the 316L layer. The results shown in Fig. 9 are consistent with the visual observations and show a significant reduction in weight loss with increasing TiC content. The worn surfaces of the annealed layers were analyzed using optical profilometer images, as shown in Fig. 9(g), (h), and (i) for the 316L, 316L-5TiC, and 316L-10TiC layers, respectively. These images provide further insight into the surface topography and are consistent with the observed wear patterns and loss of each layer.

Fig. 9
figure 9

Scanning electron microscope images of wear tracks for (a) 316L, (b) 316L-5TiC, and (c) 316L-10TiC layers in as-deposited condition and (d)-(f) in annealed conditions, respectively. (g)-(i) Surface topography of the wear tracks corresponding to 316L, 316L-5TiC, and 316L-10TiC layers in annealed condition, respectively

Full size image

4 Conclusions

Investigation of the influence of annealing heat treatment on the microstructure and mechanical properties of 316L-TiC multilayer composite produced by selective laser melting leads to the following conclusions:

  • The interface between 316L stainless steel and TiC-containing layers was found to be defect-free.

  • The TiC-reinforced layers exhibit significantly increased hardness compared to the unreinforced 316L layer, even after annealing. This underscores the robustness of TiC particle reinforcement in maintaining material strength and durability.

  • The remarkable ductility of the 316L layer observed in tensile tests exceeded ASTM standards and published data in the literature. The multilayer sample exhibited brittle behavior. However, annealing improved ductility and strength due to stress relaxation and microstructure recovery.

  • The TiC reinforcement effectively improved the wear performance of the layers. The observed decrease in weight loss with increasing TiC content, despite the influence of annealing treatments, highlights the superior wear resistance provided by TiC reinforcement.